41451 Sustaining and Sharing Economic Growth V.2 in Tanzania Contents of CD ROM Part I: Background Papers for Specific Book Chapters Chapter 1 A Decade of Reforms, Macroeconomic Stability, and Economic Growth Utz, Robert. 2005. "Tanzania: Recent Growth Performance and Prospects." World Bank, Washington, DC. Chapter 2 The Challenge of Reducing Poverty in Tanzania Kessy, Flora. 2005. "Rural Income Dynamics in Kagera Region, Tanzania." Economic and Social Research Foundation, Dar es Salaam. Levin, Jorgen, and Robert Mhamba, and Jorgen Levin. 2005. "Economic Growth, Sectoral Linkages, and Poverty Reduction in Tanzania" World Bank, Washington, DC. Mkenda, Adolf F. 2005. "Population Growth, Economic Growth and Welfare Distribution: An Overview of Theory, Empirical Evidence and Implications to Tanzania." University of Dar es Salaam. Simonsen, Marianne., and Louise. Fox. 2005. "A Profile of Poverty in Tanzania." World Bank, Washington, DC. Chapter 3 Spatial Dimensions of Growth and Poverty Reduction Mpango, Philip. 2005. "Spatial Dimensions of Growth and Poverty Reduction in Tanzania Mainland." World Bank, Washington, DC. van Dijk, Meine Pieter. 2006. "Urban Rural Dynamics in Tanzania, through Informal Redistribution Mechanisms." Institute for Water Education, Delft, the Netherlands. Chapter 4 Outlook on Growth and Poverty Reduction Demombynes, Gabriel, and Johannes G. M. Hoogeveen. 2004. "Growth, Inequality and Simulated Poverty Paths for Tanzania, 1992­2002." Policy Research Working Paper 3432, World Bank, Washington, DC. Chapter 5 Agricultural Productivity and Shared Growth Makki, Shiva S., IJsbrand de Jong, and Henry Mahoo. 2005. "Smallholder Ground Water Irrigation in Tanzania." World Bank. Washington DC. Chapter 6 Fostering Growth, Export Competitiveness, and Employment in the Manufacturing Sector Chandra, Vandana, Pooja Kacker, and Ying Li. 2005. "Tanzania: Growth, Exports and Employment in the Manufacturing Sector." World Bank, Washington, DC. Chapter 9 Fostering Innovation, Productivity, and Technological Change Aubert, Jean-Eric, and Godwill Wanga. 2005. "Innovation in Tanzania: Insights, Issues and Policies." World Bank, Washington, DC. infoDev. 2005. "Improving Competitiveness in Tanzania: The Role of Information and Communication Technologies." OTF Group, Inc., Watertown, MA. Skof, Annabella. 2006. "Constraints to Technology Access in Tanzanian Horticulture: A Case Study of Barriers to the Introduction of Improved Seed and Pest Control Technologies." World Bank, Washington, DC. Utz, Anuja. 2006. "Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy." Knowledge for Development Program, World Bank Institute, Washington, DC. Chapter 10 Enhancing the Business Environment Kopicki, Ronald. 2004. "Supply Chain Development in Tanzania: An Assessment of Three Products/Commodities." World Bank. Washington DC. (Executive Summary, Ch.1: Overview, Ch. 2: Maize, Ch. 3: Sugar, Ch. 4: Fish, Ch. 5: Lessons Learned) Chapter 11 Harnessing Natural Resources for Sustainable Growth COWI. 2005. "Natural Resource Based Growth: Summary Paper." World Bank, Washington, DC. (Executive Summary, Part 1: Literature Review, Part 2: Uncaptured Growth Potential-- Forestry, Wildlife, and Marine Fisheries, Part 3: Success Stories of Growth? Mining, Freshwater Fisheries and Tourism) Chapter 12 Enhancing the Capacity of the Poor to Participate in Growth Alderman, Harold, Hans Hoogeveen, and Mariacristina Rossi. 2005. "Reducing Child Malnutrition in Tanzania: Combined Effects of Income Growth and Program Interventions." World Bank, Washington, DC. Christiaensen, Luc, Vivian Hoffmann, and Alexander Sarris. 2006. "Coffee Price Risk in Perspective: Household Vulnerability among Rural Coffee Growing Smallholders in Tanzania." World Bank, Washington, DC. Hoogeveen, Johannes G. 2004. "The Distributional Impact of the PEDP in Rural Kilimanjaro." World Bank, Washington, DC. Hoogeveen, Johannes G. 2005. "Risk, Growth and Transfers: Prioritizing Policies in a Low- Income Environment with Risk--The Case of Tanzania." World Bank, Washington, DC. Kilama, Blandina, and Wietze Lindenboom. 2004. "Trends in Malnutrition in Tanzania." Research on Poverty Alleviation, Dar es Salaam. Mkenda, Adolf F. 2004. "The Benefits of Malnutrition Interventions: Empirical Evidence and Lessons to Tanzania." University of Dar es Salaam. Tanzania Food and Nutrition Centre. 2004. "Causes of Malnutrition and Tanzania's Nutrition Programs: Past and Present." World Bank, Washington, DC. Part II: Statistical Tables Appendix A Population and Demographics Appendix B The Economy Appendix C Exports and Imports Appendix D External Debt Appendix E Central Government Revenue and Expenditure Appendix F Monetary Situation Appendix G Agricultural Production Appendix H Employment, Labor, and Production in the Manufacturing Sector Appendix I Consumer Prices and Cost of Living Part III: Summary of Main Findings and Recommendations Preliminary Draft Tanzania Recent Growth Performance and Prospects May 22, 2005 Robert Utz AFTP2 1 1. Economic Growth since 1990.................................................................................... 5 Tanzania's Growth Performance in International Comparison .................................. 7 Sources of Growth ...................................................................................................... 8 Structural Change ..................................................................................................... 11 2. External Account Developments ............................................................................. 14 3. Investment Trends since 1990 ................................................................................. 23 Aggregate Capital Formation.................................................................................... 23 Capital Formation in the Public Sector..................................................................... 25 Private Sector Investment ......................................................................................... 28 4. Investment and Savings ........................................................................................... 31 The Macroeconomic Background: the Real Economy............................................. 31 Aspects of Money and Credit ................................................................................... 33 External Financing.................................................................................................... 37 5. Education and Human Resources ............................................................................ 42 6. Investment, Growth, and Factor Productivtiy.......................................................... 46 5. What is the Binding Constraint................................................................................ 51 Returns...................................................................................................................... 51 Cost of Finance ......................................................................................................... 53 Hypotheses - Key Constraints to Achieving Higher Rates of Growth ..................... 53 6. Economic Growth Perspectives............................................................................... 55 Growth Scenarios...................................................................................................... 55 Policy-based Growth Projections.............................................................................. 57 Input-based Projections............................................................................................. 60 Sectoral Projections .................................................................................................. 64 2 LIST OF FIGURES Figure 1. Annual growth of real GDP at f.c., 1960-2003 .................................................. 5 Figure 2. Value Added and Exports of Manufactures as % of GDP .............................. 10 Figure 3. Nominal and Real Exchange Rate Indices, 1990 - 2004.................................. 14 Figure 4. Trade, Current Account, and Overall Balance of Payments Surplus ............... 15 Figure 5. Trade, Service, and Income Balance................................................................. 16 Figure 6. Credits and debits ­ goods, services, and income, 1990 - 2003....................... 17 Figure 7. Terms of trade and export and import price indices, 1990 - 2003 ................... 18 Figure 8. Exports of Goods and Services (in m.US$)...................................................... 19 Figure 9. Merchandise Exports-Traditional and Non-traditional (in m. US$) ................ 19 Figure 10. Capital Formation by Private and Public Sectors, (as share of GDP, current prices)........................................................................................................................ 24 Figure 11. Capital Formation (as share of GDP, at 1992 prices)..................................... 24 Figure 12. Directly productive and other investment (share of total), 1993-2003 .......... 25 Figure 13. Capital Formation by the public sector, (percent of GDP, current prices)..... 26 Figure 14. Central Government Capital Formation and Development Assistance, 1985- 2003........................................................................................................................... 26 Figure 15. Central Government Capital Formation as a share of Total Central Government Expenditure, 1991-2003....................................................................... 27 Figure 16. Sectoral shares in development expenditure, FY04 ....................................... 28 Figure 17. Private sector and foreign direct investment (% of GDP)............................... 28 Figure 18. Imports of Capital Goods (% of GDP), 1997-2003........................................ 29 Figure 19. Gross National and Domestic Saving (% of GDP), 1990- 2003.................... 33 Figure 20. Real interest rates (T-bills, lending, and saving), 1993-2003......................... 37 Figure 21. Foreign Direct Investment, Flows and Stocks, 1985-2003 ............................ 39 Figure 22. FDI as a percentage of investment in selected countries................................ 40 Figure 23. International Investor country risk rating for Tanzania, 1979 -2004 ............. 42 Figure 24. Adult Literacy Rates, 1960-2000 ................................................................... 43 Figure 25. Predicted earnings based on preferred specification - Tanzania .................... 44 Figure 26. Economic growth and investment in Tanzania, moving 5 year average 1965- 2003........................................................................................................................... 47 Figure 27. Private Sector Capital Formation, (% of GDP, average 1999-2003) ............. 48 Figure 28. Cost of inefficiencies in business environment as % of sales, various countries ................................................................................................................................... 52 Figure 29. Average annual per capita GDP growth for five year periods, 1960-2003.... 55 Figure 30. Average per capita real GDP growth in 185 countries during the last decade. ................................................................................................................................... 56 Figure 31. Projections of GDP per capita and poverty 2003-2025.................................. 57 Figure 32. Average Years of Schooling, 1970-2002 ....................................................... 60 Figure 33. Contribution of capital accumulation to growth, with 4 % growth of output per worker and 20% investment of total output........................................................ 62 3 LIST OF TABLES Table 1. Real GDP growth rates ........................................................................................ 7 Table 2. Sources of Growth (expenditure)......................................................................... 9 Table 3. Sources of Growth (production) ......................................................................... 11 Table 4. Structural Change of the Tanzanian Economy, 1988-2003............................... 12 Table 5. Monetary and Non-monetary GDP, Growth and Share in GDP, 1988-2003 ..... 12 Table 6. Composition of exports (in percentage)............................................................. 19 Table 7. Constraints to private sector activities............................................................... 30 Table 8. Investment and the Resource Balance in % of GDP ......................................... 31 Table 9. Financial variables in % of GDP ....................................................................... 34 Table 10: Commercial Bank Lending to Some Sectors, % of Total Domestic Loans...... 35 Table 11. External Sources of Finance, 1997-2003......................................................... 38 Table 12. Companies listed at Dar es Salaam Stock Exchange, March 2004 ................. 38 Table 13. Foreign Direct Investment by Sector, Stocks and Flows, 1998 and 1999....... 39 Table 14. Investors' perceptions on factors influencing investment decisions ............... 41 Table 15. General Rating of Investment Policies by Sector............................................ 41 Table 16. Decomposition of Tanzania's Growth 1986-2003, Depreciation of initial capital stock by 0, 25 and 50 percent........................................................................ 49 Table 17. Decomposition of Tanzania's Growth 1995-2003, Depreciation of initial capital stock by 0, 25 and 50 percent........................................................................ 49 Table 18. Sources of Growth, Regions, 1990-2000......................................................... 50 Table 19. Projections of per capita GDP and share of population below poverty line, 2010-2025 ................................................................................................................. 57 Table 20. Policy based growth projections...................................................................... 59 Table 21. Impact of additional years of schooling on economic growth......................... 61 Table 22. Contribution of investment to growth (average over 10 years)....................... 62 Table 23. Growth and Total Factor Productivity in Selected East Asian Countries, 1960- 1994........................................................................................................................... 63 Table 24. Overall, input based projections ...................................................................... 63 Table 25. Structural Transformation, selected countries 1980-1998............................... 64 Table 26. Scenarios for economic growth and structural transformation........................ 64 4 1. ECONOMIC GROWTH SINCE 1990 1. During the past decade, Tanzania has seen an acceleration in economic growth from 0.4 percent in 1993 to well above five percent in recent years. Sustaining and possibly further increasing growth is thus one of the key challenges facing policy makers in Tanzania. Broadening the impact of growth is an additional concern, which has received prominence in Tanzania's efforts to reduce poverty. The recent pattern of growth differs significantly from Tanzania's historical growth experience in several aspects. Figure 1. Annual growth of real GDP at f.c., 1960-2003 14.0% 12.0% 10.0% 8.0% 6.0% 4.0% 2.0% 0.0% -2.0% -4.0% -6.0% 4 1960 1963 1966 1969 1972 1975 1978 1981 198 1987 1990 1993 1996 1999 2002 GDPf.c. 5 per. Mov. Avg. (GDPf.c.) 2. First, the nature of volatility during the past decade seems to be markedly different from that in previous decades. Until the early 1990s, Tanzania's economic growth performance was characterized by fairly large year to year fluctuations. Since 1993, growth has been steadily accelerating. This is the more remarkable, as Tanzania experienced severe external shocks during that period, including floods and droughts. This increased resilience to external shocks is likely to be the result of several related factors. Firstly, as the economy has been liberalized, farmers and private sector agents operate with more flexibility to react to external shocks and take advantage of market opportunities. For example, given Tanzania's size and diversity of agricultural production, weather shocks typically only affect certain parts of Tanzania and the liberalization of domestic and international trade provides greater scope to react to price signals. 3. Second, looking at long-term trends, it seems that 1984 was a turning point in the secular trend in Tanzania's economic performance. Until 1984, Tanzania appeared to be on a downward trajectory of its long term growth. 1984 was indeed also a turning point in Tanzania's history. President Nyerere admitted the failure of his socialist policies and Tanzania embarked on a market-oriented reform program. Looking at 5 Tanzania's economic performance in the long run provides a strong endorsement for the market oriented reform course pursued since then by successive governments. The reversal of the negative long term trend and the gradual acceleration in growth are consistent with the gradual broadening and deepening of reforms that has taken place and the cautious, but steady private sector response to these reforms. The temporary slump in economic growth during the early 1990s, which was caused by a weakening of macro- economic policies, points towards the fragility of growth in poor countries like Tanzania and the continued need to make macro-economic stability a cornerstone of Tanzania's growth strategy. 4. In the mid-nineties, Tanzania resumed its reform course with a clear and sustained commitment to macro-economic stability through sound fiscal and monetary policies as the foundation for economic growth. Macro-economic stabilization was accompanied by wide ranging structural reforms, including privatization of state owned enterprises, liberalization of the agriculture sector, efforts to improve the business environment, and strengthening of public expenditure management. These reforms have resulted in sustained high growth, which in the last five years was above five percent annually. Box 1. Overview of Structural Reforms in Tanzania Financial Sector. During the past decade, the financial sector has seen significant change. From being the sole preserve of state-owned financial institutions, it has gone through a process of privatization and been opened up to new entrants. These reforms have already shown some results in the form of narrowed interest spreads and a fairly rapid increase in credit to the private sector. Aside from the unfinished privatization agenda, the main challenges for the sector include further reduction in interest spreads and enhanced access to credit by the private sector. The recent establishment of a credit rating agency is a further step in enhancing the efficiency of financial intermediation in Tanzania. Parastatal Sector. Tanzania has been aggressively implementing its privatization agenda. By the end of June 2004, 169 divestitures had been completed and 57 entities have been put into receivership with LART. The number of parastatal enterprises with central government involvement has now been reduced to 47 of which 8 are under privatization and most of which represent minority shares in joint ventures. Major outstanding reform issues refer to the privatization of some financial sector entities such as NMB and NIC as well as that of public utilities such as TANESCO. Trade Policies and Institutions. Reforms of trade policies have taken mainly place in the context of regional agreements, including SADC and EAC. Joint Bank/Fund work on tariff and non-tariff barriers has been recently carried out. In addition, the Bank is also supporting a DTIS and the CEM will draw on that work. Factor Markets (Labor and Land). Revisions of land and labor legislation has been completed, with most of the emphasis on the reform of institutions. Relatively less work has been done on the potential impact of land and labor market rigidities on economic growth and the potential contribution of reforms to enhanced economic growth. Infrastructure ( Power Sector and Transport). With regard to the transport sector the establishment of the executive agency TANROADS with responsibility for the trunk road network has been a major step forward. However, a clear separation of responsibility between the Ministry of Works, TANROADS, and districts has not yet been implemented and this hampers effective road maintenance and development activities. The current formulation of a new Road Act provides a real opportunity to establish a more appropriate policy and institutional framework and provide the basis for accelerated infrastructure development. Detailed work on the restructuring of the power sector has been carried out but the implementation of the restructuring has been delayed, partly as a consequence in international energy market. The performance of 6 TANESCO has been significantly improved under a performance based management contract. Nonetheless, reform of the policy and institutional framework for the power sector is essential to ensure the effectiveness of future investments in the sector. Public Institutions Interfacing with the Private Sector. Excessive red tape and government interference in private sector activities are one of the main constraints to private sector development in Tanzania. The government has started reviewing regulations, focusing on removing obstacles and re-organizing the most important tasks of government. In practical terms this means (1) harmonization of local government taxation to remove excessive tax burden on private enterprise; (2) streamlining of work permit procedures; (3) review and amendment of licensing legislation to reduce the cost of business establishment and continuation; (4) review and revision of export-import procedures to reduce time costs and corruption- related costs; and (5) design and implementation of a program for enhancing access to commercial courts by SMEs. Tanzania has reformed the legal framework for regulatory institutions which are currently being established. The effectiveness of these regulatory institutions, especially given the current oversight arrangements, needs to be closely monitored. 5. Finally, throughout Tanzania's post-independence history, annual growth rates have rarely exceeded 6 percent, which underlines the significance of the recent acceleration in economic growth, but also cautions against overly optimistic expectations of significantly higher growth rates at a sustained basis. Tanzania's Growth Performance in International Comparison 6. During the last five years, Tanzania's economic growth has been significantly above the average for sub-Saharan Africa. Indeed, with an average growth rate of 5.2 percent during that period, Tanzania was close to the average performance of South Asia (5.4%) and South East Asia (5.6%). However, while Tanzania was able to catch up to the South and East Asian nations with respect to real economic growth, when measured on a per capita basis a relative wide gap remains, reflecting different rates of population growth. On a per-capita basis, Tanzania grew by a respectable 2.7 percent. However, south and east Asia grew by 3.6 and 4.6 percent respectively. Table 1. Real GDP growth rates Countries and regions 1988-93 1993-98 1998-03 1988-93 1993-98 1998-03 Countries Annual Real GDP Growth Annual Real Per-Capita GDP Growth Tanzania 3.6 3.0 5.2 0.4 0.1 2.7 Uganda 6.0 7.6 5.8 2.1 4.7 3.1 Kenya 3.1 2.5 1.0 0.1 0.0 -1.2 Ghana 4.6 4.3 4.4 2.0 1.5 2.4 Cote d'Ivoire 0.3 13.0 -0.3 -3.1 9.7 -1.8 Regions Sub-Saharan Africa 1.3 2.9 3.0 -1.4 0.3 0.6 East Asia and Pacific 8.6 7.7 5.6 6.9 6.4 4.6 South Asia 5.3 5.7 5.4 3.2 3.7 3.6 Latin America and Caribbean 2.1 3.6 1.2 0.2 1.9 -0.3 7 Source: World Bank (various incl. WDI) Sources of Growth 7. On the expenditure side, Table 2 suggests that exports have been a significant engine of growth in recent years, growing at an average of about 19 percent during the past five years and contributing 4.2 percentage points of Tanzania's growth performance. This presents a significant change in comparison to the preceding decade, when exports contributed only about one percentage point. However, more detailed analysis shows that the growth in exports was primarily driven by increased exports of gold. Since mining activities are import intensive, the contribution of mining to growth has been rather modest in recent years. 8. Government investment has also seen a turn-around. After shrinking between 88-98, mainly as the consequence of the fiscal consolidation program, the period 98-03 saw government investment increase by an annual average of 23.9 percent and contributing about 1.0 percent of Tanzania's economic growth. Most of the increase in government investment is financed from increases in development assistance. Government consumption has kept pace with economic growth and benefited also from increased general budget support by donors. 9. Private sector investment on the other hand, has only been growing slowly throughout the past decade, following fairly rapid growth during the period 1989- 93. The fact that GDP grew faster than investment suggests a continuous improvement in productivity in recent years. However, it is likely that in order to sustain economic growth, private investment activity will have to pick up. 10. Private consumption has grown at a slower pace than GDP during the past five years. Consumption growth has slowed down from 4.6 percent annually during the period 93-98 to four percent during the period 1998-2003. The fact that economic growth exceeds consumption growth implies an increase in savings which provides increased funding for private sector investment. 11. Imports have also been on the rise, although more slowly than exports. Given that the value of imports is still one and a half times that of exports, the stimulus to growth arising from the export side has been offset by increased imports. However, it is important to note that imports have facilitated the acceleration in exports and government investment rather than representing purely a leakage of the national economy. 12. On aggregate, an interesting picture of Tanzania's recent growth performance emerges. Aid financed public sector investment seems to have been the primary driver of economic growth in Tanzania during the past five years. The growth impulses have translated into growth in consumption and also higher demand for imports. However, the strong dependence on external developments of Tanzania's growth path also highlights the vulnerability of the economy to external shocks. The fact that growth in investment has been lagging behind overall economic growth gives cause for concern 8 and may develop into a real constraint to economic growth, once the potential for efficiency gains driven by economic reforms is exhausted. Table 2. Sources of Growth (expenditure) Avg. Ann. Growth Rate Avg. Contr. To Growth ECONOMIC ACTIVITY 1988-93 1993-98 1998-03 1988-93 1993-98 1998-03 GDP (real 1992 LC) 2.9% 3.4% 5.8% Private Consumption 3.2% 4.6% 4.0% 2.6% 3.9% 3.6% Government Consumption -0.9% 5.9% 6.0% -0.2% 1.1% 1.3% GDI 5.4% -0.8% 8.6% 1.2% -0.2% 1.9% Prvt. fixed investment 19.4% 1.8% 2.8% 1.5% 0.3% 0.4% Govt. fixed investment -8.8% -8.0% 23.9% -1.1% -0.6% 1.0% Exports GNFS 11.1% 4.2% 18.9% 1.5% 0.7% 4.2% Imports GNFS 4.7% -1.2% 11.8% -2.2% 0.5% -5.8% Statistical discrepancy -0.2% -3.0% 0.0% 13. Sectoral growth rates as shown in have accelerated across the board during the past five years. Industry has been the most dynamic sector with gold mining activity continuing its rapid expansion. However, despite continued high growth rates of value added in the mining sector, its overall contribution to economic growth remains small, given the small share of mining in overall GDP. The fact that value added in the mining sector is small reflects the capital and import intensive nature of mining activities. Furthermore, government revenue from mining activities remains limited, as royalties are only 3 percent of revenue and mining companies remain virtually tax exempt under an incentive regime that allows the accelerated full expensing of investments and only limited administrative capacity to effectively monitor gold exports and transfer pricing within multinational enterprises. 14. The construction sector grew by an average of 9.6 percent during the past five years and thus making an important contribution to Tanzania's overall growth performance. Part of the rapid growth in construction is attributable to public investment in infrastructure, but there is also increased investment in residential and business structures. 15. The manufacturing sector has started to recover, growing at an average of 6 percent per year over the past five years. However, the share of manufacturing value added and exports in GDP is still significantly below their level in the early 1990s. The balance of payments shows a relative sharp decline in merchandise exports over the past decade, which could indicate a loss of competitiveness of the traded sectors versus the non-traded and "booming sectors". However, in the Tanzanian case, it is also possible that this decline in exports of manufactures which coincided with a general decline in manufacturing activity, could be the result of the restructuring of the economy which saw the exit of state-owned enterprises in the context of Tanzania's privatization program. 9 Figure 2. Value Added and Exports of Manufactures as % of GDP 10.0% 9.0% 8.0% 7.0% PDGfo 6.0% 5.0% % 4.0% 3.0% 2.0% 1.0% 0.0% 6 1990 1991 1992 1993 1994 1995 199 1997 1998 1999 2000 2001 2002 2003 2004 Manufactures Exports (% of GDP) Manufacturing, value added (% of GDP) 16. Growth of agriculture averaged 4.4 percent during the past five years, which is about one percentage point higher than during the previous decade. Within agriculture, fishing was the most dynamic sector. However, crops remain the mainstay of the Tanzanian economy and output grew by 4.4 percent during the past five year, contributing 1.6 percentage points to Tanzania's overall growth. 17. The service sector accounts for about 35 percent of Tanzania's economy. It grew by about 5.7 percent during the past five years, which represents a significant improvement when compared to growth of the sector during the previous decade. Growth was particularly strong in the areas of trade, tourism, transport, and communication. 10 Table 3. Sources of Growth (production) Avg. Ann. Growth Rate Avg. Contr. To Growth ECONOMIC ACTIVITY 1988-93 1993-98 1998-03 1988-93 1993-98 1998-03 Agriculture 3.4% 3.2% 4.4% 1.7% 1.6% 2.1% Crops 3.6% 3.4% 4.4% 1.3% 1.3% 1.6% Livestock 2.8% 2.3% 3.8% 0.2% 0.2% 0.2% Forestry and Hunting 2.8% 2.4% 3.6% 0.1% 0.1% 0.1% Fishing 3.2% 3.8% 5.9% 0.1% 0.1% 0.2% Industry 1.6% 4.6% 7.8% 0.3% 0.7% 1.4% Mining and Quarrying 11.4% 16.0% 13.7% 0.1% 0.2% 0.3% Manufacturing 1.5% 3.8% 6.0% 0.1% 0.3% 0.5% Electricity and Water 5.4% 5.3% 4.1% 0.1% 0.1% 0.1% Electricity 6.5% 5.8% 4.2% 0.1% 0.1% 0.1% Water -0.6% 1.7% 3.5% 0.0% 0.0% 0.0% Construction -1.1% 2.0% 9.6% -0.1% 0.1% 0.5% Services 2.3% 2.8% 5.7% 0.8% 1.0% 2.0% Trade, Hotels and Restaurants 2.2% 3.6% 6.5% 0.3% 0.6% 1.1% Transport and Communication 3.8% 3.8% 5.9% 0.2% 0.2% 0.3% Financial and Business Services 4.2% 3.3% 4.3% 0.4% 0.3% 0.4% Finance and Insurance 5.4% 3.6% 3.5% 0.2% 0.1% 0.1% Real Estate 3.5% 3.1% 4.7% 0.2% 0.2% 0.3% Business Services 3.3% 4.5% 5.2% 0.0% 0.0% 0.0% Public Admin. and Other Services 2.7% 0.9% 3.7% 0.2% 0.1% 0.3% Public Administration 1.7% -1.1% 2.2% 0.1% -0.1% 0.1% Education 5.3% 4.5% 6.1% 0.1% 0.1% 0.1% Health 4.2% 3.7% 5.3% 0.0% 0.0% 0.0% Other Services 4.8% 4.9% 6.1% 0.1% 0.1% 0.1% Less Fin. Services (ind.measured) 8.5% 4.3% 2.7% -0.4% -0.2% -0.1% Total GDP (factor cost) 2.7% 3.3% 5.4% 2.7% 3.3% 5.4% Structural Change 18. The past five years have witnessed an acceleration in structural change of the Tanzanian economy. The relatively fast growth of industry has led to an increase by two percentage points in its contribution to GDP and it accounts now for 18.5 percent of GDP. Most of the increased contribution of industry to GDP is attributable to the expansion of the mining and construction sectors. Gains in manufacturing were more modest. On the other hand, the share of agriculture has fallen by more than two percentage points from 49.1 percent in 1998 to 46.8 percent in 2003. This pattern of 11 structural change is consistent with the change most economies experience as they develop and is thus a positive sign. Table 4. Structural Change of the Tanzanian Economy, 1988-2003 Sector Average Annual Growth Rate Share in GDP 1988-93 1993-98 1998-03 1988 1993 1998 2003 Agriculture 3.4% 3.2% 4.4% 47.6% 49.3% 49.1% 46.8% Industry 1.6% 4.6% 7.8% 16.4% 15.5% 16.5% 18.5% Services 2.3% 2.8% 5.7% 35.9% 35.2% 34.3% 34.8% Total GDP (factor cost) 2.7% 3.3% 5.4% 100.0% 100.0% 100.0% 100.0% 19. Monetization of the economy has also increased during the past five years. As shown in Table 5, Tanzania's growth acceleration was concentrated on the monetary sector, whose share in GDP increased from 72.6 percent in 1993 to 75 percent in 2003. Growth in non-monetary GDP is only slightly above the rate of population growth. Non-monetary GDP captures mainly non-monetary agriculture, but also includes construction and owner­occupied dwellings. Table 5. Monetary and Non-monetary GDP, Growth and Share in GDP, 1988-2003 Sector Average Annual Growth Rate Share in GDP 1988-93 1993-98 1998-03 1993 1998 2003 Monetary GDP 2.6% 3.4% 6.0% 72.6% 73.0% 75.0% Non-Monetary GDP 3.2% 3.0% 3.8% 27.4% 27.0% 25.0% Total GDP (factor cost) 2.7% 3.3% 5.4% 100.0% 100.0% 100.0% 12 Box 2. The Size of the Shadow Economy in Tanzania Alongside the "official" economy, there exists a "shadow" economy or an "informal sector, " which escapes detection in the official estimates of GDP. This includes illegal activities such as trade with stolen goods, manufacturing and sale of illegal drugs, or smuggling as well as legal activities that are not reported, primarily for tax evasion and avoidance reasons. Schneider (2004) estimates the size of the shadow economy for 145 countries, using the following definition: "The shadow economy includes all market- based legal production of goods and services that are deliberately concealed from public authorities for the following reasons: (1) to avoid payment of income, value added or other taxes; (2) to avoid payment of social security contributions, (3) to avoid having to meet certain legal labor market standards, such as minimum wages, maximum working hours, safety standards, etc., and (4) to avoid complying with certain administrative procedures, such as completing statistical questionnaires or other administrative forms. The study suggest that in 2002/03, the size of the shadow economy in Tanzania was about 60 percent of the official GDP and thus among the largest shadow economies in the Africa region and world wide. Even though estimates of the size of the shadow economy are to be taken with a grain of salt, the high share estimated for Tanzania certainly underlines the urgency of reviewing the incentive environment which drives economic activity outside the formal economy. In this context, government's "Property and Business Formalization Programme" is an important element of Tanzania's growth strategy. Country Size of Shadow Economy (% of off. GDP) Region/Country Group Size of Shadow Economy (% of off. GDP) Tanzania 60% Africa 42% Kenya 36% Asia 31% Uganda 45% South America 43% 13 Mozambique 42% Central Europe and FSU 38% South Africa 30% OECD 16% Source: Schneider (2004) 2. EXTERNAL ACCOUNT DEVELOPMENTS 20. We start the discussion of external account development with a review of key price and exchange rate developments. The nominal effective exchange rate (NEER) depreciated sharply between 1990 and 1995, reflecting macro-economic instability and high rates of inflation in Tanzania. The onset of reforms in 1995 initiates a period of relative stability and a slowdown of the depreciation of the Tanzanian Shilling against other currencies. This period lasts until 2002 when the NEER starts to depreciates more rapidly. However, this recent period of depreciation of the Tanzanian Shilling is primarily due to the appreciation of the Euro and the South African Rand in international currency markets rather than domestic developments. Figure 3. Nominal and Real Exchange Rate Indices, 1990 - 2004 Nominal exchange rate index Euro, UK Pound, US$, Real and Nominal Effective Exchange Rate, 1990- and South African Rand, 1990 - 2004 2004 300.00 250% 230% 250.00 210% 190% 200.00 170% 150% 150.00 130% 110% 100.00 90% 70% 50.00 50% 4 4 1990 1992 1994 1996 1998 2000 2002 200 1990 1992 1994 1996 1998 2000 2002 200 USDTZ UKPTZ EUROTZ RANDTZ NEER REER Source: IMF data base 21. Developments in the nominal effective exchange rate are reflected in the real effective exchange rate (REER), where three distinct periods between 1990 and 2004 are distinguishable. From 1990 until 1995, when the NEER depreciated by almost 60 percent, the REER remained remarkably stable since the nominal depreciation was offset 14 by the high rates of domestic inflation. Subsequently, when the NEER stabilized, the REER appreciated by about 50 percent between 1995 and 1998 and remained at the strongly appreciated level until around 2001. In 2002, the REER started to depreciate in parallel to the depreciation of the NEER. According to recent work by the IMF (2004), the appreciation of the REER led to an overvaluation of the exchange rate. The recent depreciation led to a return of the exchange rate to its equilibrium value, i.e., is consistent with long run equilibrium in domestic goods and labor markets and a sustainable current account. 22. The appreciation of the REER between 1995 and 1998 was due to a variety of developments. Primarily, it seems to be linked to the initiation of macro-economic stabilization by the authorities through tighter fiscal and monetary management, which dampened inflationary expectations. The reasons for the recent depreciation of the exchange rate are less apparent. Indeed, a number of factors such as recent increases in foreign aid inflows in the form of budgetary support and the increase in exports of gold are likely to exert upward pressure on the exchange rate.1 Figure 4. Trade, Current Account, and Overall Balance of Payments Surplus 10.0% 5.0% 0.0% -5.0% -10.0% PDGfo -15.0% % -20.0% -25.0% -30.0% -35.0% -40.0% 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Current account (excl. transfers) Overall balance Current account (incl. transfers) 23. Figure 4 shows the current account (excl. and inc. of current transfers), and the overall balance of payments surplus for the period 1990 ­ 2003. All three key external balances have been improving significantly since 1993. The current account deficit (excl. grants) has declined from 35 percent of GDP in 1993 to only 7 percent in 2002. Current transfers were relatively stable and the current account balance after grants improved 1IMF (2004) finds that foreign aid inflows, government consumption, and relative productivity lead to an appreciation of the exchange rate, while increased openness leads to a depreciation of the exchange rate. 15 from a deficit of 26 percent in 1993 to a small surplus in 2002. Finally, the difference between the current account and the overall balance of payments are capital flows, which in the case of Tanzania is primarily foreign direct investment , capital grants, and net borrowing (on concessional terms). Again these have been fairly stable during the past decade and the improvement in the overall balance of payments from a deficit of 15 percent in 1993 to a surplus of 4 percent of GDP in 2003 is due to underlying developments in exports and imports of goods and services. 24. It is interesting to observe, that the improvement in the external accounts coincides with an extended period of strong appreciation of the real effective exchange rate. It is likely, that the positive effect of overall macro-economic stabilization and structural reforms outweighed the exchange rate appreciation. Figure 5. Trade, Service, and Income Balance -40.0% -35.0% -30.0% -25.0% PDGfo -20.0% % -15.0% -10.0% -5.0% 0.0% 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Trade balance Service balance Income balance 25. Figure 5 shows that the improvement in the balance on trade, service, and income is due to improvements in each of three components, with the improvements in the trade and service accounts being the most significant. 16 Figure 6. Credits and debits ­ goods, services, and income, 1990 - 2003 35.0% 20.0% 18.0% 30.0% 16.0% 25.0% 14.0% PD 20.0% 12.0% Gfo 10.0% 15.0% PDGfo % % 8.0% 10.0% 6.0% 4.0% 5.0% 2.0% 0.0% 0.0% 1990 1992 1994 1996 1998 2000 2 200 1990 1992 1994 1996 1998 2000 2002 Goods exports: f.o.b. Goods imports: f.o.b. Services: credit Services: debit 6.0% 5.0% 4.0% PD Gfo 3.0% % 2.0% 1.0% 0.0% 1990 1992 1994 1996 1998 2000 2002 Income: credit Income: debit 26. The improvement in the trade balance is almost entirely due to a reduction of imports as a percentage of GDP, while exports have actually declined initially and only recovered to the 1993 level in 2002. Developments in the service account mirror those in the trade account. The drop in payments for services reflects the fall in payments for insurance and freight related to the decline in imports. Revenue from the exports of services, i.e., primarily tourism, on the other hand have been stagnant as a percentage of GDP. Finally, the improvement in the income account is primarily due to the reduction of interest payments as a consequence of debt relief, rescheduling, and a greater share of concessional debt. 27. Available data on prices of exports and imports suggest relatively large fluctuations in export and import prices during most of the 1990s, which translates into large fluctuations in the terms of trade. However, on average, during the past five years the terms of trade have been more favorable for Tanzania than in the preceding years. This improvement in the terms of trade coincides with the recent depreciation of the real effective exchange rate. 17 Figure 7. Terms of trade and export and import price indices, 1990 - 2003 160 160 140 140 120 120 100 100 80 80 60 60 40 40 1990 1992 1994 1996 1998 2000 2002 2004 9901 9921 9941 9961 9981 0002 0022 0042 Export Unit Value Index (1992=100) Terms of Trade index (1992=100) Import Unit value index (1992=100) Source: IMF data base 28. Export performance during the 1990-2003 period could be distinguished into four phases: sluggish growth during 1990-93, followed by fast growth during 1993-96, decline in 1996-99, then recovery from 2000 onwards (Figure 8).2 During the fast growth period of 1993-96, the strong export performance can be attributed to both merchandise and service exports, with the latter actually registering a much stronger performance, albeit from a low base. The rapid growth in services exports, which tripled their share from around 3 percent in 1990 to over 9 percent in 1996 was due entirely to the rapid growth in tourist receipts, which rose ten-fold from $47m. in 1990 to over $470m. in 1996, reflecting an annual growth of nearly 47 percent. Both merchandise and service exports declined between 1996 to 1999, while the recovery from 2000 onwards was due mostly to the strong performance in merchandise exports as the growth in services exports remained sluggish. 29. During the 1990s, traditional and non-traditional exports had very similar performances, mirroring overall merchandise export performance of sluggish growth followed by acceleration, then decline (Figure 2). However, from 2000 onwards, the performance of these two categories of exports diverged dramatically, with the sharp recovery in merchandise exports since then being entirely attributed to the acceleration in non-traditional exports, which more than compensated for the continued decline in traditional exports. The divergence in performance between these two categories in the last few years has resulted in a significant shift in the composition of Tanzanian exports, from around half each in 1990, to nearly 80 percent being made up of non-traditional exports in 2003 (Table 6). 2This an the following paragraphs on export performance are from the DTIS concept note by Helena Tang. (2004) 18 Figure 8. Exports of Goods and Services (in m.US$) 1800.00 1600.00 1400.00 1200.00 1000.00 800.00 600.00 400.00 200.00 0.00 0 2 4 6 8 0 2 199 199 199 199 199 200 200 Exports of Goods Exports of Services Tourism Exports of Gds & Services Source: IMF BOP Statistics. 30. The impressive rise in non-traditional exports has been due practically to one export commodity alone--gold. Indeed, 70 percent of the recovery in non-traditional exports since 1998 was due to gold exports, which had risen from nil in the first half of the 1990s to being the single largest export item in 2003, making up nearly 40 percent of all exports and single-handedly propelling non-traditional exports to the dominant export category. Aside from gold, fish and fish products have also emerged as a significant export item that has contributed to the recovery of non-traditional exports, rising from 7.5 percent of total exports in 1997 to 13.5 percent in 2003. Manufactured exports have also recovered in 2003, although they are still below the levels (both in dollar terms and in terms of shares of total exports) reached in the mid-1990s. Figure 9. Merchandise Exports-Traditional and Non-traditional (in m. US$) 1200.0 1000.0 800.0 600.0 400.0 200.0 0.0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Traditional Exports Non-traditional Exports Total Exports Source: Economic Survey 2003, Government of Tanzania. Table 6. Composition of exports (in percentage) 19 1990 1995 1996 1997 1998 1999 2000 2001 2002 2003 Traditional Exports 50.2 56.2 57.1 57.8 60.5 55.4 44.1 29.8 22.8 21.4 Coffee 21.0 20.9 17.8 15.9 18.5 14.1 12.6 7.4 3.9 4.8 Cotton 18.7 17.6 16.4 17.3 8.1 5.2 5.7 4.3 3.2 4.5 Tea 5.4 3.4 3.0 4.2 5.2 4.5 4.9 3.7 3.3 2.4 Tobacco 2.7 4.0 6.4 7.1 9.4 8.0 5.8 4.6 6.2 4.0 Cashew Nuts 1.4 9.4 12.8 12.1 18.2 18.6 12.7 7.3 5.2 4.0 Sisal 1.0 0.9 0.7 1.2 1.2 1.3 0.8 0.9 0.7 0.6 Cloves 3.7 1.5 1.6 0.4 1.0 Non-traditional Exports 49.8 43.8 42.9 42.2 39.5 44.6 55.9 70.2 77.2 78.6 Minerals & metals 6.8 6.6 7.3 6.8 4.5 13.5 26.9 38.9 42.5 43.2 of which: gold 0.2 0.5 6.4 17.0 32.7 37.8 38.6 Other mineral 6.8 6.6 7.3 6.6 4.0 7.1 9.9 6.2 4.7 4.6 Manufactured 18.2 16.0 16.1 14.8 6.1 5.5 6.5 7.2 7.3 9.6 Fish & products 7.5 12.6 10.4 11.5 12.5 12.9 13.5 Horticulture 0.7 1.5 1.6 1.5 1.4 1.2 1.1 Other 24.8 21.3 19.5 12.4 14.8 13.5 9.5 10.2 13.2 11.2 Source: Economic Survey 2003, Government of Tanzania. 31. Exports of agricultural products more than doubled between 1990 and 1997. However, since than they have dropped dramatically and are now not much higher than they were at the beginning of the 1990s. Exports of individual agricultural commodities broadly followed this pattern. The decline in exports of coffee was gradual since 1995, while exports of cotton virtually collapsed in 1998 falling by more than half during that year. Exports of cashew nuts continued to expand until 1999, but have declined almost by half since then. What is particularly perplexing is that the drop in agricultural exports coincided with the acceleration in overall economic growth. To put agricultural exports into perspective, we note that their total value is only about one fifth of aid inflows to Tanzania. 20 Figure 10. Main agricultural exports (US$ million), 1990-2004 500.0 400.0 lionli 300.0 m $ 200.0 US100.0 0.0 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 Coffee value (million $) Cotton value (million $) Cashew nuts value (millions of $) Tea value (million $) Tobacco value ($) Sisal value ($) Cloves value ($) Source: IMF data base 21 Figure 11. Volume and Unit Values of Coffee, Cotton, Tea, and Cashew Nut exports, 1990-2004 Coffee Cotton 3500.0 80.0 2000.0 100.0 3000.0 70.0 1800.0 90.0 ) )n 1600.0 80.0 2500.0 60.0 )not onst /to$SU( 1400.0 70.0 50.0 ds ofs 2000.0 an 1200.0 60.0 eulavti 40.0 us on)t/$SU( 1000.0 50.0 and 1500.0 ho(t us 30.0 e 800.0 40.0 hot( 1000.0 600.0 30.0 e Un 20.0 muloV uelavtinU 400.0 20.0 500.0 10.0 muloV 200.0 10.0 0.0 0.0 0.0 0.0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 90 92 94 96 98 00 02 04 19 19 19 19 19 20 20 20 Unit value (US$/ton) Volume (thousands ton) Unit value (US$ per ton) Volume (thousands ton) Tea Cashew nuts 2000.0 30.0 1200.0 180.0 )not/ 160.0 25.0 1000.0 on)t 140.0 1500.0 800.0 120.0 S$U( 20.0 )notsdn )no/t$SU( ndsa 100.0 600.0 e 1000.0 15.0 80.0 luavitnU 400.0 60.0 ousht( 10.0 sauoht( e e 500.0 200.0 40.0 5.0 mul eulavtinU 20.0 Vo 0.0 0.0 moloV 0.0 0.0 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 199 199 199 199 199 199 199 199 199 199 200 200 200 200 200 Unit value (US$ per ton) Volume (thousands ton) Unit value (US$ per ton) Volume (thousands ton) Source: IMF data base 32. Exports of manufactured goods also experienced large fluctuations during the past decade. Particularly noteworthy is the drop by almost two thirds in 1998. Since then, manufacturing exports have been growing relatively fast, although the total value of manufactured exports is still small at around US$ 80 million in 2003. Figure 12. Exports of Manufactured Goods (US$ million), 1990-2004 140.0 120.0 100.0 n io 80.0 ill M $SU 60.0 40.0 20.0 0.0 90 91 92 93 94 95 96 97 98 99 00 01 02 03 19 19 19 19 19 19 19 19 19 19 20 20 20 20 2004 Source: IMF data base 22 33. Several conclusions and/or implications can be drawn from the stylized facts of export performance outlined above. Much of this export growth was due to gold exports alone. This raises the issue of the sustainability of such export growth. It also raises the issue of the income distributional impact of such growth, as proceeds from such export revenues probably do not reach large parts of the population, and even less so the poor. Both of these implications, in turn, point to the need for greater export diversification. There are signs that such diversification has begun--witness the increase in fish and fish products and horticulture exports--but it needs to be sustained, as is the recent recovery in manufactured exports. 34. Export diversification could also come from services exports, and in particular tourism, which has the potential to play a much more important role in boosting overall economic growth in Tanzania than it had done during the second half of the 1990s. Notwithstanding the recovery in tourism since 2000, by 2003 tourism receipts of US$450m. were still below the peak of $500m. attained in 1995. Expansion of the tourist sector--which aside from being labor-intensive also has a high income multiplier and strong forward and backward linkages--would also have the beneficial effects of generating employment and incomes beyond the sector itself3. 35. The decline in virtually all agricultural crop exports, in light of the fact that agriculture generates 70 percent of employment and that 90 percent of the poor reside in the rural areas, further underlines concerns over the distributional impact of the recent export recovery. Recovery of such agricultural crop exports would contribute not only to overall growth but also certainly poverty reduction. 36. Finally the stagnation in tourism earnings since 1995 is also disappointing, given Tanzania's potential. 3. INVESTMENT TRENDS SINCE 1990 Aggregate Capital Formation 37. For most of the 1990s, capital formation has been on a declining trend, from about 30 percent of GDP in 1991 to 18 percent in 1997. Since then, investment has recovered, reaching 25 percent of GDP in 2002. As Figure 13 clearly shows, the recovery in overall investment since 1997 is entirely driven by a recovery in investment by the public sector, while investment by the private sector continued its downward trend. This gives reason for concern, since much of the literature on economic growth suggests that private investment is closely tied to economic growth, while the linkage between public investment and growth is much weaker. 3One study (Kweka, 2001) found that the output multiplier for tourism in Tanzania is 1.8, which means that an investment of 1m. shillings leads to an increase in 1.8m. shillings to the economy. The same study also found that the output multiplier effect for tourism is higher than that for agriculture, manufacturing, and other services. Further, tourism was found to lead the other three sectors in terms of backward and forward linkages, and was second to agriculture in terms of inter-sector effects among 23 sectors. 23 Figure 13. Capital Formation by Private and Public Sectors, (as share of GDP, current prices) 35.0% 30.0% 25.0% Total Public Sector: 20.0% Private Sector ++ 15.0% Total Capital 10.0% Formation 5.0% 0.0% 85 87 89 91 93 95 97 99 01 03 19 19 19 19 19 19 19 19 20 20 Figure 14. Capital Formation (as share of GDP, at 1992 prices) 35.0% 30.0% Buildings 25.0% Other Works 20.0% 15.0% Equipment 10.0% Total Capital 5.0% Formation 0.0% 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 38. Figure 14 shows the evolution of capital formation by type of investment. Capital formation includes residential and rural own-account buildings and a variety of public expenditures (roads and bridges, water supply) which may be crucial in permitting private enterprise to flourish but are not themselves directly productive4. One proxy for productive investment is to take the sum of investment in equipment and non-residential buildings. The recovery since 1997 affected primarily investment in buildings and other works. Investment in equipment has been stagnant about 12 percent of GDP and dropped to 10 percent more recently. Figure 15 shows that the share of directly productive investment in total investment has been declining during the past decade by 4I.e. they do not produce outputs that are recorded in the GDP figures. 24 ten percentage points from 63 percent in 1993 to 53 percent in 2003. This raises questions as to the appropriate balance between public and private sector investment. A question that is also of direct importance to the donor community is whether public funds should be primarily be used to fund public expenditures and public investment or whether there is greater scope to use donor funds to generate private sector investment, e.g., by compensating for revenue losses if investment tax incentive schemes were introduced. Figure 15. Directly productive and other investment (share of total), 1993-2003 120% 100% 80% Other 60% Directly Productive 40% 20% 0% 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Capital Formation in the Public Sector 39. Public investment has seen significant changes during the past decade. It peaked around 1990, reaching almost 12 percent of GDP primarily as a result of large investments by the parastatal sector. Subsequently, as government embarked on a privatization process, investment by parastatals and total public investment declined dramatically. By 1999, when the privatization process had significantly advanced, parastatal investment has fallen below one percent of GDP and remained at that level. 25 Figure 16. Capital Formation by the public sector, (percent of GDP, current prices) 14.0% 12.0% 10.0% Central Gov't 8.0% Parastatals 6.0% Institutions+ Total Public Sector: 4.0% 2.0% 0.0% 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 40. Central Government investment started to increase in the early 1990s, but fiscal restructuring and the withholding of aid by donors led to a slump in public investment in the mid-nineties. However, starting in 1998, central government investment has increased rapidly and reached almost seven percent of GDP in 2003. Figure 17 shows very clearly the close relationship between central government capital formation and inflows of development assistance. As has been shown in various PERs (e.g. World Bank, 2003), about 90 percent of the development budget in Tanzania is financed by foreign aid. Figure 17. Central Government Capital Formation and Development Assistance, 1985-2003 8.0% 7.0% 6.0% 5.0% Central Gov't 4.0% Grants 3.0% 2.0% 1.0% 0.0% 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 41. It is also interesting to note that in the course of fiscal adjustment, development expenditure was the category that was most affected. As Figure 18 shows, development 26 expenditure as a share of total expenditure declined significantly during the mid nineties, but has recovered since. Figure 18. Central Government Capital Formation as a share of Total Central Government Expenditure, 1991-2003 40.0% 35.0% 30.0% 25.0% 20.0% 15.0% 10.0% 5.0% 0.0% 91 92 93 94 5 96 97 98 99 00 1 02 03 19 19 19 19 199 19 19 19 19 20 200 20 20 42. With regard to the sectoral distribution of development expenditures Figure 19 shows that development spending is disbursed widely across sector, with no clearly apparent prioritization of investments in infrastructure for transport and energy, which are key constraints to economic growth. The literature on the relationship between public investment and economic growth suggests that only certain types of public investment, in particular infrastructure, are growth enhancing. A recent investment climate assessment (The World Bank, 2004) highlights an unreliable power supply as one of the main private sector constraints. Work on rural development highlights the importance of rural infrastructure, in particular rural roads, as key determinants of growth in rural areas by linking farmers to markets. 27 Figure 19. Sectoral shares in development expenditure, FY04 GENERAL PUBLIC SERVICES 5% OTHER ECONOMIC PUBLIC ORDER AND 12% SAFETY 4% EDUCATION TRANSPORTATION AND 13% COMMUNICATION 17% HEALTH 7% HOUSING AND COMMUNITY AMENITY MINING, MINERAL 16% ,MANUFACTURING AND CONSTRUCTION AGRICULTURE, 19% FORESTRY, FISHING AND HUNTING 7% Private Sector Investment 43. National accounts figures show a steady decline in private sector investment from about 20 percent of GDP in 1995 to less than 13 percent of GDP in recent years, which is close to the international average. It is also of interest to note that the decline in overall private sector investment coincided with the increase in foreign direct investment since the mid nineties, which peaked in 2000 exceeding 5 percent of GDP, but has since returned to the previous level of around 3 percent of GDP. Figure 20. Private sector and foreign direct investment (% of GDP) 25.0% 20.0% 15.0% Private Sector 10.0% FDI 5.0% 0.0% 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 28 44. A survey of direct foreign investment carried out in 2001 provides more detailed information on FDI. The value of the foreign investment stock was estimated to have risen to US$ 2.6 bn in 1999. Mining is the biggest recipient of FDI and accounts for about 40 percent of the stock of FDI. Manufacturing accounted for 22 percent and tourism for 13 percent of the stock of FDI at the end of 1999. Agriculture, despite its importance and potential, has not yet been able to attract any significant amounts of FDI. 45. Another source of information on investment trends are data on imports of capital goods shown in Figure 21. After exceeding 14 percent of GDP in 1992, imports of capital goods declined dramatically to about four percent of GDP in 1997. Subsequently, there was a slight recovery to about 7 percent of GDP in 2003. These figures are broadly consistent with the investment trends above. However, the share of imported capital goods in total investment and in investment in equipment has declined quite dramatically. Figure 21. Imports of Capital Goods (% of GDP), 1997-2003 Imports of Capital Goods 0.18 0.16 0.14 f.c.ta 0.12 0.1 GDPfo 0.08 0.06 %0.04 0.02 0 1986 1987 1988 1989 1990* 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 29 Table 7. Constraints to private sector activities Respondents' Evaluation of General Constraints to operation % of firms evaluating constraint as "major" or "very severe" Tanzania Exporter Non- exporter Bold = greater concern for exporters Tax rates 72.1 72.2 72.1 Electricity 57.6 61.1 56.8 Cost of Financing 56.2 42.6 59.5 Tax Administration 54.7 61.1 53.2 Corruption 50.0 55.6 48.7 Access to Financing 47.1 29.6 51.4 Macroeconomic Instability 42.0 42.6 41.9 Competition from Imports 34.8 38.9 33.8 Customs and Trade Regulations 30.8 46.3 27.0 Regulatory Policy Uncertainty 30.8 22.2 32.9 Business Licensing and Operating Permits 26.8 24.1 27.5 Crime, theft and disorder 25.0 29.6 23.9 Skills and Education of Available Workers 24.6 24.1 24.8 Access to Land 24.3 18.5 25.7 Anti-competitive or informal practices 23.9 27.8 23.0 Transportation 22.5 35.2 19.4 Legal system/Conflict resolution 19.6 25.9 18.0 Labor Regulations 12.0 14.8 11.3 Telecommunications 11.6 13.0 11.3 Source: RPED Tanzania Investment Climate Assessment, 2003 46. A recently completed investment climate survey (RPED, 2004) pinpoints various constraints to private sector investment (Table 7). Taxation, cost and access to finance, availability of electricity, and corruption are seen by more than 50 percent of the surveyed firms to be a major or very severe constraint. These are important entry points for improving Tanzania's business climate to foster increased economic activity. 47. However, it is important to keep in mind that the investment climate survey provides a snapshot of current constraints to private sector profitability, but does not offer a comprehensive assessment of factors that could constrain economic growth. For example, existing firms may have chosen their location based on the availability of access to transport and the ICA would show relative satisfaction with the existing infrastructure. However, lack of transport infrastructure might nonetheless be a major constraint for the economic development of regions that do not have access to infrastructure consistent with their economic potential. A similar situation could prevail with respect to skills and education of available workers, where the available skills might be sufficient for current levels of economic activity, but provide little scope for expansion of economic activity, in particular in "knowledge intensive" sectors such as manufacturing or business services. 30 4. INVESTMENT AND SAVINGS The Macroeconomic Background: the Real Economy 48. Table 8 is derived from the most recent National Accounts figures (Economic Survey, 2003). It gives a number of key macroeconomic magnitudes in per cent of GDP at current market prices (the relevant measure when we are interested in the proportionate allocation of available resources). The growth of GDP at factor cost at constant prices is also included. Table 8. Investment and the Resource Balance in % of GDP Investment Exports Imports Import Net Foreign National Domestic GDP (GFCF) Surplus Flow Saving Saving Saving Growth I X M B=M- From B-F S=I+F- DS=I-B % X ROW B F (1) (2) (3) (4) (5) (6) (7) (8) (9) 1990 26.1% 12.6% 37.5% 24.8% 16.5% 8.3% 17.8% 1.3% 6.2% 1991 26.3% 10.3% 33.6% 23.4% 12.9% 10.5% 15.9% 3.0% 2.8% 1992 27.2% 12.4% 39.4% 26.9% 14.8% 12.2% 15.1% 0.3% 1.8% 1993 25.1% 18.0% 47.7% 29.7% 9.5% 20.2% 4.9% -4.6% 0.4% 1994 24.6% 20.6% 43.6% 23.0% 9.5% 13.6% 11.1% 1.6% 1.4% 1995 19.8% 24.1% 41.5% 17.4% 8.0% 9.5% 10.3% 2.4% 3.6% 1996 16.6% 19.9% 31.9% 12.0% 7.0% 5.0% 11.7% 4.6% 4.2% 1997 14.9% 16.2% 25.7% 9.5% 7.6% 1.9% 13.0% 5.4% 3.3% 1998 16.2% 13.4% 28.1% 14.7% 6.9% 7.8% 8.4% 1.5% 4.0% 1999 15.5% 13.8% 26.5% 12.7% 5.9% 6.8% 8.7% 2.8% 4.7% 2000 17.6% 14.6% 23.1% 8.4% 6.8% 1.6% 16.0% 9.2% 4.8% 2001 17.0% 15.5% 23.7% 8.2% 7.2% 1.0% 16.0% 8.8% 5.8% 2002 19.1% 16.1% 22.3% 6.2% 7.8% -1.6% 20.7% 12.9% 6.2% 2003 18.6% 17.8% 26.7% 9.0% 12.2% -3.3% 21.9% 9.7% 5.6% Source: author's calculations based on Economic Survey (2003) 49. The table shows that the Tanzanian economy underwent dramatic changes in recent years · As discussed in detail before, the investment ratio (column 1) saw a sharp drop during the first half of the 1990s declining from a high of 27.2 percent in 1992 to a low of 14.9 percent. In recent year's, investment has started to recover and stood at 18.6 percent of GDP in 2003. While in earlier years the investment ratio seems to have had little impact on economic growth, in recent years the recovery of investment does coincide with an acceleration in economic growth. This suggests that previously, investment productivity was low, partly because much of it was accounted for by parastatals. In more recent years, where investment is either by the private sector or by the public sector for the provision of infrastructure services, the link between investment and growth appears to have been strengthened. 31 · Imports declined by half from 48 percent of GDP in 1993 to 27 percent in 2003. Exports also experienced huge fluctuations during the past decade. After peaking in 1995 at about 25 percent of GDP, they declined to 13.4 percent by 1998 but have recovered since then and stand now at 17.8 percent of GDP. · Primarily as a reflection of the sharp contraction of imports, the import surplus or trade deficit (column4) has been declining sharply since the mid nineties. While it stood at close to 30 percent of GDP in 1993, in 2003 it has shrunk to 9 percent of GDP. It is interesting to note that until around 1995, the imports surplus was in magnitude similar to investment, i.e., investment was wholly supplied from the excess of imports over exports, while domestic production was exhausted by consumption. Subsequently as the import surplus started to decline, a growing share of investment has been contributed by domestic output. In the past five year, on average less than 50 percent of investment has been covered by the imports surplus, while the rest has been provided by domestic output. · Part of the import surplus is financed from the excess of current transfers from the rest of the world (such as grants) over net factor income paid to the rest of the world (such as interest on external public debt). This net flow is given in column 5 and represents primarily grant assistance to Tanzania from bi-and multilateral donors. Until 1999, net flows from the rest of the world have been declining. However, in recent years these flows have seen a marked recovery and stood at 12.2 percent of GDP in 2003. · The residue of the import surplus is financed by borrowing ­ "foreign saving" - and is given in column 6. Until 1999, borrowing has played an important role in financing the import surplus. However, during the past four years, borrowing has been small or even negative in the past two years, reflecting both the decline in the import surplus as well as the increase in net flows from the rest of the world. · National saving (column 7) is obtained as the residual after subtracting foreign saving from investment. 5 with national savings being on a downward trend until throughout the 1990s falling to 8.7 percent in 1999. However, 2000 saw almost a doubling in national saving to 16 percent of GDP and has subsequently risen to reach 21.9 percent in 2003. · Domestic saving (column 8) shows a similar evolution to that of national saving and the increase witnessed during the past four years is quite remarkable. 5There is always a "residual error" in the National Accounts, a discrepancy between the value of GDP estimated from the sides of production and expenditure. Since 1990, on average the statistical discrepancy was only 0.1 percent of GDP, but it fluctuated between ­3.6 percent of GDP to +2.5 percent of GDP. In more recent years, the statistical discrepancy has been more stable and positive with an average of 1.4 percent of GDP. The series reported in column 7 assumes that the measured expenditures in the first three columns are correct so that the error lies in the measurement of output, or most likely, consumption. If the discrepancy reflects errors in measuring one or more of the expenditure magnitudes in the first three columns, the discrepancy would have to be added to the saving figure. However, no matter where the national accounts discrepancy is allocated, both series have the same broad "shape." 32 Figure 22. Gross National and Domestic Saving (% of GDP), 1990- 2003 15 10 PD 5 Gfo % 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 -5 -10 Gross domestic savings (% of GDP) Gross national savings, including NCTR (% of GDP) Aspects of Money and Credit 50. Sound monetary policy and a sound financial sector are important elements in providing an enabling environment for economic growth and investment. This has also been one of the corner stones of Tanzania's very successful stabilization strategy, which led to the decline in the rate of inflation from a high level of 27.4 percent 1995 to around five percent in recent years. With regard to investment, the interaction between government borrowing and private sector borrowing is of particular interest, since large government borrowing typically raises concerns about crowding out of private sector borrowing. 33 Table 9. Financial variables in % of GDP Monetary Survey 1997 1998 1999 2000 2001 2002 2003 Net foreign assets 8% 8% 7% 9% 12% 13% 17% Net domestic assets 12% 11% 10% 8% 6% 6% 3% Domestic Credit 9% 9% 9% 9% 8% 7% 8% Net claims on government 5% 5% 5% 5% 3% 2% 2% Credit to private sector 3% 4% 4% 4% 5% 5% 6% Other Items (net) 3% 2% 0% -1% -2% 1% -2% of which: liquidity paper (issued by BOT) 0% -1% -3% -4% -2% -1% -1% Broad money (M3) 20% 18% 16% 17% 18% 19% 21% Currency in circulation 6% 6% 4% 5% 5% 5% 4% Deposits 10% 10% 9% 8% 9% 10% 10% Foreign currency deposits 4% 3% 3% 4% 5% 5% 6% Commercial Banks Net foreign assets 5% 5% 5% 6% 6% 6% 6% Foreign assets 5% 5% 5% 6% 7% 6% 6% Foreign liabilities 0% 0% 0% 0% 0% 0% 0% Net domestic assets 9% 8% 8% 8% 8% 11% 11% Reserves 1% 2% 2% 2% 2% 2% 2% Domestic credit 9% 10% 10% 10% 8% 10% 10% Claims on government (net) 5% 5% 5% 5% 3% 4% 2% Claims on private sector 3% 4% 5% 5% 5% 6% 8% Other items net -1% -3% -3% -4% -2% -1% -1% Demand deposits 4% 4% 4% 4% 4% 5% 5% Time and Savings deposits 6% 5% 5% 5% 6% 6% 6% Foreign currency deposits 4% 3% 4% 4% 5% 6% 6% 51. 34 Table 9 presents monetary statistics for the last nine years, again using GDP as a scaling device6. Government efforts to maintain monetary stability are reflected in a relatively stable aggregate money supply (M3) during the past decade, although the increase in this ratio from 16 percent in 1998 to 21 percent suggests some financial deepening of the economy. Nonetheless, the money to GDP ratio is still low in Tanzania compared to other African countries. M2, which is another measure of the money supply, stood at 21.6 percent of GDP in Tanzania. For Kenya it was 40 percent, Mozambique 29.6 percent, and Uganda 19. percent. 52. Behind this aggregate stability, some dramatic changes in the structure of assets and liabilities are visible. Net foreign assets have more than doubled between 1997 and 2003. This is primarily attributable to the rapid increase in international reserves held by the Bank of Tanzania. On the other hand, net domestic assets of the financial sector have declined dramatically from twelve percent in 1997 to three percent of GDP in 2003. A sharp reduction in lending to government by the Bank of Tanzania as well as a reduction in outstanding liquidity paper by the Bank of Tanzania are the driving forces behind the overall decline in net domestic assets. 53. Financial intermediation by commercial banks has seen an increase from 14 percent of GDP in 1997 to about 17 percent in 2003. However, the most dramatic developments occurred with respect to the structure of credit. Overall credit by commercial banks remained fairly stable at around 9 percent of GDP between 1997 and 2003. However, behind this overall stability is a dramatic change in the composition of credit. In 1997, the bulk of credit (5% of GDP) still went to the public sector, while credit to the private sector only amounted to 3 percent of GDP. In subsequent years, credit to government declined steadily as a result of prudent fiscal policy and repayment of domestic lending by government in most years. This decline in credit to government was compensated by an increase in credit to the private sector, which reached 8 percent of GDP in 2003. Table 10: Commercial Bank Lending to Some Sectors, % of Total Domestic Loans Average Share in total lending growth rate 1997 2000 2003 1997-2003 Public Sector 3% 2% 2% 20% Agricultural production 8% 6% 12% 38% Mining and manufacturing 24% 31% 26% 30% Building and construction 2% 3% 5% 44% Transportation 8% 13% 9% 31% Tourism 1% 1% 2% 40% Marketing of agricultural produce 1% 0% 0% -100% Export of agricultural produce 2% 0% 0% -100% Trade in capital goods 0% 0% 0% -100% All other trade 24% 26% 23% 27% Specified financial institutions 0% 2% 4% 93% Other 27% 13% 17% 18% 6End of year value as per cent of GDP for the preceding year. 35 Total 100% 100% 100% 28% 54. Mining and manufacturing and trade account for almost 50 percent of bank lending. Agricultural production and the transportation sector account together for another 20 percent of lending. Credit between 1997 and 2003 has grown at an average annual rate of 24 percent. Credit to agricultural production, building and construction, tourism, and to specified financial institutions saw the fastest increases in credit. It is also interesting to note that credit for marketing and export of agricultural produce completely disappeared during that period. It is also worthwhile to note that the increase in credit to the private sector is not reflected in an increase in private sector investment. This could be due to the fact that most of the credit serves to finance working capital, which is supported by the large share of credit going to trade and agricultural production. The sectoral pattern of increases in credit is also consistent with the sectoral growth patterns, where the fastest growing sectors also experienced the largest increase in credit. 55. Additional information on the impact of credit financing is provided by the recently completed investment climate assessment, which provides information on the financing of enterprises. The survey showed that only 20 percent of firms reported having loans from financing institutions. Only 16 percent of investment is financed through bank lending in Tanzania, while about 68 percent of investment is financed through retained earnings. 56. Figure 12 shows real interest rate developments during the past decade. Real interest rates on savings have been negative until 1998 and only slightly positive in recent years. The developments in real savings rate seem to be closely related to domestic saving, which also has been rising as real interest rates have been increasing. Another observation from Figure 23 is the large interest rate spread between lending and savings rates. Overall, this picture reflects the low level of development of the financial sector in Tanzania. In particular, access to credit is limited to a small number of enterprises with solid collateral in key urban areas, while small and medium size enterprises and firms located outside the main urban areas are virtually excluded from access to credit. It is also worth noting that high lending rates coincided with excess liquidity in the system in recent years. 36 Figure 23. Real interest rates (T-bills, lending, and saving), 1993-2003 20% 15% 10% 5% 0% -5% -10% 199 199 199 1996 199 3 4 5 7 199 199 200 2001 200 200 8 9 0 2 3 Real interest rate - T-bills Real interest rate - lending Real interest rate - saving External Financing 57. Current and capital transfers and concessional borrowing by government are the primary sources of external finance for Tanzania. Foreign borrowing by the private sector is rather limited, with access limited to a few large companies in Tanzania. 37 Table 11. External Sources of Finance, 1997-2003 1997 1998 1999 2000 2001 2002 2003 Grants and other transfers 9.2% 8.1% 7.0% 8.0% 8.0% 7.4% 10.5% Net Government External Borrowing 0.1% -3.8% 1.2% 1.4% -2.6% 0.4% 3.7% Net bank borrowing/lending -0.9% -0.2% 0.2% -1.5% -0.8% 0.0% -0.2% FDI 2.1% 3.3% 6.0% 5.1% 3.5% 2.6% 2.6% Other 1.0% 0.4% -0.1% -1.8% 0.3% 0.0% -1.0% Source: IMF 58. Even though the Dar es Salaam stock exchange has been opened up to foreign participation in 2003, portfolio investment is still very small. This reflects partly the small number of companies (six) traded and the relative unattractiveness of the Dar es Salaam stock exchange in terms of liquidity. Dividend yields are insufficient to compensate for exchange rate risks. Table 12. Companies listed at Dar es Salaam Stock Exchange, March 2004 NAVPS Company EPS Tshs DPS Tshs P/E D.Yield Tshs TBL 161.52 107 8.05 10 377.91 TOL - - 16.5 - 158.34 TATEPA 14.77 40 38.6 7 290.12 TCC 235.6 219 7.22 13 465.69 SIMBA 100.16 100 7.49 14 325.74 DAHACO 63.21 50.56 9.02 9 155.16 59. It is also important to note that for most of the past decades there have been capital outflows from the banking sector, although in the aforementioned capital flow survey, only very few banks indicate that they borrow or lend abroad. The capital outflows by the banking sector are related to developments in domestic financial markets, which until recently have been characterized by significant levels of excess liquidity. 60. Foreign direct investment has gained in importance in recent years and finances about 20 percent of Tanzania's investment. Figure 10 shows FDI inflows as a percentage of gross fixed capital formation and the stock of foreign capital as a percentage of GDP. The graph shows that immediately upon the resumption of reforms in 1995 FDI jumped to about 15 percent of total Gross Fixed Capital Formation and has remained at that level in subsequent years. FDI peaked between 1999 and 2001, primarily on account of large inflows in the mining sector and from the privatization of the National Bank of Commerce (2000) and Tanzania Telecommunications Ltd. (2001). 38 Figure 24. Foreign Direct Investment, Flows and Stocks, 1985-2003 45 40 35 30 25 20 15 10 5 0 -5 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 Inflows as a percentage of GFCF Inward stock as a percentage of GDP 61. The latest detailed information on FDI is from an survey of private capital flows carried out in 2001. The mining sector accounts for the bulk of FDI. However, other sectors such as manufacturing, trade, tourism, and agriculture have also been able to attract sizeable amounts of FDI, at least in comparison to other African countries. In the medium to long-term, it is likely that FDI in these sectors is an important driver of structural change and knowledge transfer to the Tanzanian economy. Table 13. Foreign Direct Investment by Sector, Stocks and Flows, 1998 and 1999 Stock Flow 1998 1999 Mining and Quarrying 503.6 848.9 345.3 Manufacturing 406.4 475.4 69 Wholesale, Retail Trade, Catering, and Accommodation 251.7 281.4 29.7 Services Agriculture, Hunting, and Forestry 119.5 151.4 31.9 Finance Insurance, Real Estate, and Business Services 133.9 147.7 13.8 Construction 106.9 120.8 13.9 Transport, Storage, and Communications 49.5 60 10.5 Others 29.5 31.6 2.1 Community, Social and Personal Services 1.4 1.8 0.4 Electricity, Gas, and Water 35.4 35.4 0 Total 1637.8 2154.4 516.6 39 Figure 25. FDI as a percentage of investment in selected countries Ireland Azerbaijan Bolivia Bulgaria Georgia Armenia Côte d'Ivoire Chile M ozambique Hungary Uganda Cambodia United Republic of Tanzania Colombia Venezuela Poland M exico Viet Nam South Africa Romania Portugal M alaysia M auritius El Salvador Thailand Ghana Philippines Senegal Turkey Egypt Kenya Greece Indonesia -20.0 0.0 20.0 40.0 60.0 80.0 100.0 62. In international comparison (Figure 25), Tanzania is among the relative attractive destinations for FDI. The survey of investors also includes questions on their perceptions on factors affecting the investment climate. Government stability, availability of banking services, investment incentives, and the macro-economic situation are among factors that make Tanzania an attractive destination for foreign investors. On the other hand, bureaucracy, electricity supply and tariffs, corruption, high bank lending rates and non-availability of credit, depreciation of the Tanzanian Shilling, HIV/AIDS, and land law administration exert a negative influence on investors decisions. 40 Table 14. Investors' perceptions on factors influencing investment decisions Positive Neutral Negativ Overall positive assessment Government Stability 82% 16% 3% Banking Services 64% 22% 14% Investment Incentives 52% 28% 20% Monetary Policy 51% 26% 24% Domestic Political Scenario 48% 41% 11% Fiscal Policy 46% 20% 34% Labor Stability 41% 38% 22% Overall negative assessment Bureaucracy 14% 11% 76% Electricity Tariff 11% 19% 70% Corruption 9% 21% 70% Electricity Supply 32% 13% 55% Tax collection efficiency 27% 22% 52% Interest Rate 18% 32% 50% Exchange Rate 18% 32% 50% HIV/AIDS 2% 55% 42% Land law administration 20% 43% 36% Source: BoT et al. (2001) 63. On a sectoral basis, investor perceptions of investment policies are for most sectors quite positive (Table 15). Only investors in the agriculture and construction sector have a less favorable view of the investment climate in their sectors. This is quite important in view of Government's objective to accelerate agricultural production and points to the need for increased Government effort in this area. 64. Another interesting aspect is that the large majorities of investors plan to expand their investments in the medium term, primarily because of the socio-economic stability. Table 15. General Rating of Investment Policies by Sector Sector Favorable Community, Social, and Personal Services 80% Wholesale, Retail, Accomodation, and Tourism 79% Transport and Storage and Communication 79% Mining and Quarry 75% Financing, Insurance, and Real Estate and Business Services 75% Manufacturing 72% Agriculture, Hunting, and Forestry 55% Construction 53% Source: BoT et al. (2001) 65. Assessments of country risk by publications such as Euromoney, International Investor, or the International Country Risk Guide provide another perspective on a countriy's attractiveness for investors. Figure 26 shows that Tanzania's risk rating by the International Investor publication has continuously improved since the introduction of 41 market based reforms. In fact, the movement in the risk rating seems to mirror quite closely Tanzania's growth performance. Figure 26. International Investor country risk rating for Tanzania, 1979 -2004 30 60.0 25 70.0 ero 20 15 80.0 gniknaR Sc10 90.0 5 0 100.0 evtialeR 1979 1982 1984 1987 1989 1992 1994 1997 1999 2002 Score Relative Rank Source: International Investor, various issues 5. EDUCATION AND HUMAN RESOURCES7 66. Education is the fundamental enabler of the knowledge economy. Well educated and skilled people are key for creating, sharing disseminating and using knowledge effectively. Ideally, the knowledge economy requires an education system which is flexible ­ starting from basic education that provides the foundation for learning, to secondary and tertiary education that can develop core skills, including technical ones, that encourage creative and critical thinking critical for problem-solving and innovation, to a system of lifelong learning. A lifelong learning system is a system which encompasses learning throughout the life cycle (from early childhood to retirement) and includes formal training (schools, training institutions, universities); non-formal learning (on-the-job and household training), and informal learning (skills learned from family members or people in the community). The basic elements of such a system are comprehensiveness, new basic skills (acting autonomously, using tools interactively, and functioning in socially heterogeneous groups), multiple pathways and multiple providers. 67. Developing countries such as Tanzania face many challenges in developing such systems. These include expanding coverage to achieve universal access to basic education as well as increased access to secondary and provision; strengthening tertiary education, which is generally weak; improving the linkages between formal and informal education systems and the labor market; and raising the quality of learning. 68. To its credit, Tanzania has made some progress in education with respect to neighboring countries. For example, Figure 27 shows that in 2001, Tanzania's adult literacy rate at 77 percent was higher than that of Uganda (69 percent), but lower than 7This section draws on a background study to the CEM by A. Adhar "Fostering Innovation, Productivity, and Technological Change ­ Tanzania in the Knowledge Economy," October 22, 2004. 42 that of Kenya, whose adult literacy rate at 84 percent is higher than that in Botswana (79 percent) and is close to the levels of South Africa (86 percent). Figure 27. Adult Literacy Rates, 1960-2000 Adult Literacy Rates 85 Mauritius 75 South Africa Kenya 65 Malaysia Botswana 55 Uganda Tanzania 45 Ghana Tanzania Ghana 35 Kenya Uganda Botswana Mauritius Malaysia South Africa 25 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 Source: World Bank SIMA Database 69. These achievements can be attributed to several factors. In recent years, the Tanzanian government has recognized the need to raise educational levels in the population as a necessary condition for enhancing economic growth. Appropriate programs for primary and secondary education are being put in place which are intended to enhance access and increase the quality of education. Key measures so far have included the abolition of primary school fees in 2001, significant increases in budgetary funding for primary education, and the implementation of the Primary Education Development Program. Tanzania's gross enrollment ratios (GERs) for primary education increased from 100.4 percent in 2002 to 105.3 percent in 2003. At the same time, the country has also been making efforts to improve the quality and equity in the education system. The primary school drop out rate decreased from 5.5 percent in 2000 to 3.8 percent in 2003. The pass rate in standard 7 examinations increased from 24.9 percent in 2001 to 27.1 percent in 2002, and the transition rate from primary to secondary education also rose from 21 percent in 2001 to 21.7 in 2002. 70. Moving on to secondary enrollments, Tanzania's GERs stood at a low 6 percent as compared to 31 percent in Kenya.8 In order to increase secondary education, the government has launched the Secondary Education Development Program (SEDP) which has among its aims, increasing the proportion of the relevant age group completing secondary education, expanding enrollments with equity, improving learning outcomes of students, especially among girls, and enabling the public administration to manage secondary education more effectively. It also includes institutional reforms and capacity building at central, region, district, and school levels for more efficient operation of the secondary education system. 8Source: World Bank SIMA database. 43 71. In terms of tertiary education, Tanzania's performance is again weak: tertiary GERs stood at 1 percent as compared to 3 percent and 4 percent, respectively, for Uganda and Kenya in 2001. Botswana and South Africa has tertiary GERs of and 5 and 15 percent, respectively in 2001. In the 2000/01 academic year there were 6,117 students at the University of Dar es Salaam, and 13,442 altogether in the country's three universities (University of Dar es Salaam, Sokoine University of Agriculture, and the Tanzania Open University). In April 2001 an Education Fund was established to sponsor children from very poor families to complete higher education.9 But in the last ten years, a number of private universities have also emerged and today the country has nine private universities, mostly of denominational nature and small in size, which award diplomas in areas such as financial and business management, wildlife management, community development, social welfare and cooperatives, and transport and media operations.10 72. A recent paper has also looked at the returns to education in the manufacturing sector in Tanzania. The findings suggest that the marginal returns to education for primary and secondary education are rather limited. However, the data suggests a sharp increase in the returns for people who have a tertiary education. One of the implications of these findings is that the impact on economic growth of more primary and secondary education is likely to be small, but greater investment in tertiary education has higher payoffs in terms of economic growth.11 Figure 28. Predicted earnings based on preferred specification - Tanzania Source: Mans Soderbom et al. (2004) 73. Government has been the major financier of technical and vocational education and training (VET), with assistance from donors. But the VET system in Tanzania faces several problems including inefficient resource utilization, inequitable distribution of 9Source: EIU: Tanzania Country Profile 2003. 10Background note on Tanzania for the Policy Workshop on Knowledge for Development, ESRF, 2002. 11Mans Soderbom, Francis Teal, Anthony Wambugu and Godius Kahyarar, "The Dynamics of Returns to Education in Kenyan and Tanzanian Manufacturing," Centre for the Study of African Economies (CSAE), University of Oxford, CSAE WPS/2003-17, January 2004. 44 educational opportunities, poor labor market linkages, and a lack of coordination between donors and the government. The unsustainable costs of training appear to be caused not only by low capacity utilization, but also by low student-to-faculty ratios, while inequitable distribution of education opportunities do exist, biased towards primary schools of wealthier backgrounds. Recognizing that the VET system had failed to produce graduates who were suited for the labor market, policy changes were introduced in 1996 that emphasize the government's continued responsibility in the provision and financing of more and better basic education, coupled with a reduction in untargeted subsidies through increased cost-sharing, liberalization of private education and training at all levels, and decentralization of authority. The Vocational Training and Education Authority (VETA) that was set up in 1994 is working to ensure that training provided is responsive to the labor market.12 74. Brain drain: Africa is a capital scarce region and loss of this limited resource is widely considered detrimental to the prospects of sustained growth and development. There is a significant parallel to this problem on the side of human capital. Weakness in human capital and particularly skill deficiency is a drag on investment and growth in Africa. Progress in overcoming shortages of skilled and trained manpower seems to be disappointingly slow, despite substantial resources devoted by both governments and donors to this effort during the last three decades. This deficiency is sustained at a time when Africa is losing a very significant proportion of its skilled and professional manpower to other markets and increasingly depending on expatriates for many vital functions. Although comparatively the Africa region is the smallest source of immigration to the developed world, a high proportion of its migrants is made up of highly skilled professionals. For example, it has been estimated that for a number of African countries, more than 30 percent of its highly skilled professionals are lost to the OECD countries. Nearly 88 percent of adults who emigrate from Africa to the United States have a high school education or higher. There are more African scientists and engineers working in the United States than there are in Africa. The emigration of technically skilled people has left 20,000 scientists and engineers in Africa servicing 600 million people. 75. Tanzania is no stranger to the brain drain phenomenon. The most vulnerable occupations at national level include medicine, accountancy, law, engineering, and science based occupations. As a proxy for the national picture, data from two premier institutions of higher learning provide some interesting evidence of this phenomenon. For the University of Dar es Salaam, with a teaching staff of about 861 people, about 149 members of the staff left the university between 1990 and March 2002, which is about 17.3 percent of staff. The majority of those who left were from the Faculty of Arts and Social Sciences (38), followed by Medicine (17), Engineering (13), Law (11), Science (10) and Commerce (9). The ranks at which most of them exited were at Senior Lecturer and Lecturers level. For Sokoine University of Agriculture, with a staff of 239 people, about 50 members of staff left in the same period (1990-2002), which accounts for 21 12Indermit Gill and Amit Dar, "Vocational education and training in Tanzania : finance and relevance issues in transition," Country Study Summary 19781, October 1998, the World Bank. 45 percent of the staff, and again the majority that left were either Lecturers or Senior Lecturers, the prime ranks for academic production.13 76. In the field of medicine, Tanzania is again facing a massive skills loss, especially of doctors and scientists. Low salaries for doctors is the principle reason driving the brain drain--and those that remain seek higher wages in private hospitals in large urban centers, leading to a lack of doctors in some of the country's district hospitals. In a bid to increase the number of health professionals in the country, the Tanzanian government has recently promised to cover all training costs for medical students in both public and private universities.14 77. Threat of HIV/AIDS: HIV/AIDS also poses significant risks for human resource development in Tanzania. A HIV/AIDS indicator survey carried out in 2004 estimated the prevalence rate of HIV/AIDS inTanzania at seven percent of the population between 15 and 49. Studies that focus on Africa calculate the annual loss of GDP to range between a modest decline of 0.3 percent to a high of 1.5 percent in GDP growth annually. But a new report argues that the costs are likely to be much higher. Previous estimates overlooked the impact of HIV/AIDS on children if one or both parents die. In countries facing an HIV/AIDS epidemic on the same scale as South Africa, for example, if nothing is done quickly to fight the epidemic, they could face economic collapse within several generations, with family incomes being cut in half.15 Tanzania thus need to continue taking steps to curb the scourge of HIV/AIDS. 6. INVESTMENT, GROWTH, AND FACTOR PRODUCTIVTIY 78. While investment plays an important role in a country's growth process, the link between investment and growth is by no means automatic. Figure 29 plots five year averages for the rate of real GDP growth and the ratio of gross fixed capital formation to GDP16 and it is evident that there is no simple positive relation between the investment ratio and economic growth.17 The investment ratio has been increasing fairly steadily until the mid nineties, while economic growth has been on a downward trend for most of this period. 13Background note on Tanzania for the Policy Workshop on Knowledge for Development, ESRF, 2002. 14Deodatus Balile, "Tanzania Pledges Support for Science Training," June 19, 2003, SciDev.net (http://www.scidev.net/dossiers/index.cfm?fuseaction=dossierReadItem&type=1&itemid=871&language= 1&dossier=10) 15Clive Bell, Hans Gersbach and Shanta Devarajan, "The Long-run Economic Costs of AIDS: Theory and an Application to South Africa, World Bank, 2003. 16At factor prices in constant 1992 prices 17Simple regression analysis actually shows a significant negative relationship between growth and investment. 46 Figure 29. Economic growth and investment in Tanzania, moving 5 year average 1965-2003 0.08 0.3 0.07 0.25 ) %(ht 0.06 0.2 0.05 ow gr 0.04 0.15 PDG PDG/FC 0.03 0.1 GF ealr0.02 0.05 0.01 0 0 1960 1963 1966 1969 1972 1975 1978 1981 1984 1987 1990 1993 1996 1999 2002 grow th GFCF/GDP(total) GFCF/GDP(private) 79. While there are reasons to believe that the relationship between public investment and growth may be weak, Figure 29 also suggests that the relationship between private sector investment and growth is equally ambiguous. However, it is worth noting that this finding is not specific to Tanzania, but much of the recent economic literature (e.g., Easterly and Levine, 2001) on economic growth also confirms this lack of a direct relationship between economic growth and investment. Figure 30 plots the average growth rates for the period 1999-2003 for a sample of 122 countries against their five year average investment ratios and it is apparent that there is no direct relationship between the level of investment and economic growth.18 The implication of this apparent lack of a strong relationship between investment and growth is however not that investment does not matter. It rather points towards the importance of the quality of investment, especially public sector investment which is not ruled by market forces, the inherent risk of investment decisions, and the environment in which the investment takes place. In addition, it also highlights the role of total factor productivity as an important factor determining economic growth. 18Statistical regression analysis of the data for the sample of 122 countries shows a statistically significant, but small positive relationship between investment ratios and economic growth. 47 Figure 30. Private Sector Capital Formation, (% of GDP, average 1999-2003) 40.0 14.0 12.0 35.0 10.0 30.0 8.0 25.0 6.0 4.0 20.0 2.0 15.0 0.0 10.0 -2.0 -4.0 5.0 -6.0 0.0 -8.0 Investment Growth 80. Growth accounting is a widely used method to separate the contribution of physical and human capital as well as total factor productivity to economic growth. The analysis requires estimates of the physical and the human capital stock.19 Human capital is typically estimated on the basis of average years of schooling. 81. The capital stock is generally estimated by using the perpetual inventory method. After estimating the capital stock for an initial period20, estimates for subsequent periods are obtained by assuming a certain rate of annual depreciation21 and by adding new investment. However, there are many instances where the value of an investment does not automatically add an equivalent amount to the capital stock.22 Among the reasons for this are poor investment decisions, or over priced investment cost. For Tanzania, there is indeed a strong likelihood that government investment did not add to the productive capital stock. Examples for this abound, such as the Morogoro shoe factory which never produced at more than 4 percent of its production capacity. Once government decided to privatize, the sale value of privatized enterprises, if they could be sold at all, was typically small and in many cases negative. In the case of investments in infrastructure, it is also quite likely that the rate of depreciation exceeded the commonly assumed rate of 5 percent on account of insufficient investments in maintenance and rehabilitation. 82. 1985, when Tanzania abandoned its socialist policies and started to introduce market reforms might thus be an appropriate year for the re-estimation of the capital 19Estimates of human capital are based on Collins and Bosworth (2003) who provide data up to 2000 by averaging time series for years of schooling assembled by Barro and Lee and Cohen and Soto and applying a rate of return of seven percent to years of schooling. Estimates of the capital stock are also drawn from Collins and Bosworth (2003). 20Nehru and Dhareshwar (1993) provide estimates of the capital stock for a large number of countries, including Tanzania. 21We assume an annual rate of depreciation of 5%. 22For a detailed discussion see Pritchett (1996) 48 stock.23 Given the paucity of data, we examine two scenarios, where in 1985 the actual capital stock is assumed to be only 50 and 75 percent of the estimated capital stock. With respect to the growth accounting exercise, a lower initial capital stock implies reduced absolute amounts of depreciation and thus higher levels of net investment24 and additions to the capital stock. In turn, the share of growth attributed to capital will be higher and the share attributed to factor productivity will be lower, the lower the initial capital stock. The contribution of education remains unaffected by changes in the capital stock.25 Table 16. Decomposition of Tanzania's Growth 1986-2003, Depreciation of initial capital stock by 0, 25 and 50 percent. Contribution of Adjustment in Output per Physical Education Factor Capital Stock Worker Capital Productivity 0% 0.96% 0.07% 0.48% 0.41% 25% 0.96% 0.36% 0.48% 0.12% 50% 0.96% 0.82% 0.48% -0.34% Source: Author's calculations 83. As Table 16 shows, the relative contribution of physical capital and factor productivity changes quite dramatically with respect to the assumption of the initial capital stock. Using available estimates of the capital stock attributes most of the economic growth experienced between 1986 and 2003 to increases in factor productivity while the contribution of capital formation was rather small. However, if we assume that existing statistics overestimate the capital stock in 1985, the picture changes dramatically. If the capital stock in 1985 was only half of its estimated value, then most of the growth experienced since 1985 is attributable to additions to the capital stock and increases in factor productivity play only a minor role. Table 17. Decomposition of Tanzania's Growth 1995-2003, Depreciation of initial capital stock by 0, 25 and 50 percent. Contribution of Adjustment in Output per Physical Education Factor Capital Stock Worker Capital Productivity 0% 2.27% -0.05% 0.70% 1.61% 25% 2.27% 0.21% 0.70% 1.36% 50% 2.27% 0.57% 0.70% 0.99% Source: Author's calculations 84. Our adjustments to the initial capital stock affect also the decomposition of growth for the more recent period. Under the original assumptions on the capital stock, 23Since even after 1985 until the mid 1990s, public sector investment and investment by parastatals was relatively large, it might indeed be appropriate to discount up to 50 percent of public sector investment during this period. 24Net investment is gross investment minus depreciation. 25This is purely a consequence of the underlying growth accounting methodology. It is likely that a lower capital-qualified labor ratio would imply higher returns to capital and lower returns to qualified labor. 49 the growth accounting results suggest that investment levels were insufficient to maintain Tanzania's capital stock and to generate economic growth. Assumptions of a smaller initial capital stock yield a picture where physical capital investment actually led to an increase in the capital stock per worker and thus made a positive contribution to growth. While the assumptions on the initial capital stock also affect total factor productivity, the overall picture remains one where improvements in factor productivity made important contributions to economic growth in recent years. Table 18. Sources of Growth, Regions, 1990-2000 Contribution of Output per Physical Education Factor Productivity Worker Capital Tanzania* 1.8% -0.1% -0.5% 0.5% 0.6%-1.2% World 3.5 1.2 0.3 1.9 Industrial 1.5 0.8 0.2 0.5 Countries China 8.8 3.2 0.3 5.1 East Asia 3.4 2.3 0.5 0.5 Less China Latin 0.9 0.2 0.3 0.4 America South Asia 2.8 1.2 0.4 1.2 Africa -0.2 -0.1 0.4 -0.5 Middle East 0.8 0.3 0.5 0.0 * Data for Tanzania refer to the period 1993-2003 Source: data for Tanzania ­ author's calculations, other data - Bosworth, B. and Susan Collins. The Empirics of Growth: An Update. Processed. Brookings Institution. 9/22/2003. 85. To interpret the findings of the growth accounting exercise for Tanzania, comparisons to other countries are useful. Table 18 compares the contribution of capital, education, and factor productivity to growth in Tanzania with the results obtained for other regions. Several interesting facts are apparent from this table. Firstly, average growth rates varied significantly during the 1990s. The Asian economies performed strongest led by China, where output per worker grew by 8.8 percent. On the other end of the spectrum the African countries in the sample saw their output per worker decline by 0.2 percent. The wide variation in growth rates is reflected in similarly large variations in the contributions of capital and factor productivity. For the Asian economies, capital formation clearly played an important role and it appears that in general faster growth of output per worker is associated with higher growth in capital per worker. On the other hand, the contribution of factor productivity seems to be less clear cut. For example, in the case of China, almost 60 percent of its growth are explained by enhanced factor productivity while in the case of the other East Asian economies, only about 20 percent of growth are explained by increases in factor productivity. The contribution of education to growth shows less variability across regions. The contribution ranges in a rather narrow band between 0.2 percent and 0.5 percent. 50 86. The increase in average years of education during the past decade is reflected in a relative large contribution of human capital to growth. However, compared to other countries, Tanzania's performance during the 1990's appears to be plagued by several weaknesses. Firstly, the contribution of capital formation seems to be low, even when we allow for the fact that the initial capital stock may have been lower than suggested by the commonly used estimates. In addition, the data also suggest scope for greater increases in factor productivity although estimates for more recent years actually suggest that factor productivity has indeed increased in Tanzania to levels achieved by other regions. However, it is likely that this recent upturn in factor productivity is the result of demand side factors and efficiency gains from reforms that have moved Tanzania closer to its production frontier, but there is little evidence that the increased factor productivity indeed represents technological change that would on a sustainable basis increase Tanzania's productive capacity. 5. WHAT IS THE BINDING CONSTRAINT 87. Drawing on the information and analysis in the preceding sections, we now review potential constraints to economic growth using the framework developed by Hausmann, Rodrik, and Velasco (2004). This framework proposes that systematic analysis of private returns to investment and of the cost of finance should allow to identify binding constraints to growth, on which policy makers should focus their attention. Returns 88. The 2005 WDR contained an assessment of how various elements of the business environment impact on the cost of firms. For Tanzania, the cost of an inefficient business environment are estimated to be very high in international comparison amounting to 25 percent of sales, including the cost of contract enforcement difficulties, regulation, bribes, crime, and unreliable infrastructure. Figure 31 suggests that the costs of bribes and of unreliable infrastructure are particularly high in Tanzania. In the following paragraphs we summarize the assessment of factors affecting returns and their appropriability. 51 Figure 31. Cost of inefficiencies in business environment as % of sales, various countries Source: World Bank, WDR 2005 89. Micro Risks. Crime is relatively low in Tanzania, but the cost to businesses seem nonetheless to be high. Tanzania certainly suffers from the general negative perception of Africa. Corruption is high, both according the TI corruption perception index and the Kaufmann et. al governance ratings and the 2005 WDR shows that the share of revenue spent on bribes is relatively high in Tanzania. 90. Macro Risks. Tanzania has developed a solid track record of political and economic stability. A financial sector assessment indicates that systemic risk in the sector is low. 91. Lack of R&D. Innovation occurs primarily through the acquisition and adoption of foreign technology, although agricultural research institutes have been successful in developing a variety of seeds that are well adapted to the Tanzanian environment. While there are a few success stories of successful enterprises developing new (to Tanzania) areas of production such as the export of high quality fish, horticulture, and cashew nuts, mining, retail supermarkets, overall structural change and technology adoption has been slow. 92. Externalities, spillovers, coordination failure. Economic policy making has been successful in achieving macro-economic stabilization, with the Ministry of Finance and the Bank of Tanzania setting the economic policy agenda. However, there is an apparent need for greater focus on coordinating policy formulation and implementation for structural transformation and economic growth. There are few or now formal mechanisms in place for managing the effect of externalities and spillovers on innovation. However, informal mechanisms and relationships with the authorities 52 developed in the course of introducing innovations may in fact limit the spread of innovations and thus limit externalities and spillovers. While this could be beneficial for the individual innovator, it limits the spread of innovations and its impact on economic growth. 93. High tax rates, inefficient tax structure. Tax rates are moderate and the tax structure is relatively efficient with a VAT as the main source of revenue. 94. Insufficient infrastructure. Electric power is only available in urban areas and even there the power supply is unreliable, forcing most companies to have back-up generators. Road infrastructure is underdeveloped. Although in the most recent investment climate assessment transport is not among the main constraints, it is likely that agriculture sector growth. Access to cell phones has improved significantly in recent years, however, international telephone charges remain high. 95. Human Resources. The share of the population with post-primary education in Tanzania is among the lowest in Africa. Returns to education are low for primary and secondary education, but increase significantly for higher education. This could suggest that availability of qualified manpower is among the key growth constraints. Cost of Finance 96. Local Finance: Real returns on savings are low, but there are high interest spreads which indicate inefficiencies in the system. In particular access to credit is limited to relatively large and well established companies in the major urban areas. Interviews with banks and business people suggest that there is a high unmet demand for credit, partly because collateral requirements by the Bank of Tanzania are too high. 97. International Finance: Foreign direct investment has been significant in recent years, in particular in the mining sector. Tanzania's risk rating has improved, but is still low in international comparison. Hypotheses - Key Constraints to Achieving Higher Rates of Growth 98. Tanzania's growth acceleration is driven by the implementation of a broad based reform program since the mid 1980s, supported by foreign aid inflows. The emergence of Tanzania as a major gold producer in the late 1990s plays a relatively minor role in explaining Tanzania's growth performance. 99. Large regional differences in factors such as geography (climate, soil fertility, distance to markets), production structure, and access to infrastructure and banking services translate into large regional differences in growth performance. This in turn suggests that constraints to growth may differ according to sector, product, or region. 100. Returns are low because of poor infrastructure, especially power and transport. Poor infrastructure is generally believed to be among the key constraints to 53 economic growth in Tanzania. Limited access and unreliable power supply reduce the competitiveness of Tanzanian companies. Poor infrastructure, especially roads, is seen as a key constraint to agriculture. 101. Limited access to capital prevents the full exploitation of investment opportunities. Real interest rates on savings are now positive, but still very low while lending rates are high and interest spreads are large. However, access to capital as a binding constraint on economic growth arises because of a underdeveloped financial sectors where large parts of the economy are excluded from access to credit. 102. Sustaining growth requires improved rates of technological change and innovation. Our interpretation of Tanzania's recent growth performance is that it is primarily driven by efficiency gains which brought Tanzania closer to its production frontier. As the opportunities for efficiency gains become more limited, technological change (in particular in agriculture) and the identification of new areas of comparative advantage will become more important. 103. Corruption and bureaucracy affect the appropriability of returns. Investor surveys show that these factors lower the attractiveness of Tanzania for investors and that they lower private returns by as much as 25 percent. 104. Risk and lack of credit constrain rural growth.26 Tanzania's rural areas may be characterized by the highly imperfect credit markets and high (climatic and disease) risk. On top of that, many investments are lumpy. In such an environment, and driven by precautionary motives, asset poor households have a large incentive to smooth consumption and to invest in safe but low return income processes. Since most households are poor, one is likely to find --in an aggregate analysis, a low rate of return on human and physical assets for rural areas. From this one would conclude that credit is not a binding constraint. 105. Yet, if one could do the analysis for households that do not have to worry about consumption smoothing or lumpiness of investments (say the asset rich), one is likely to find that the returns to human and physical are high (see e.g. Dercon 1996) . So unless we identify risk in combination with imperfect credit markets as a binding constraint beforehand, we may not pick it up at all and miss an important part of the story. 106. Hence the analysis for rural areas should make a distinction between those likely to be bound by concerns of consumption smoothing (who make a costly efficiency- security trade-off in their income decisions) and those who are not bound by such a constraint. If one finds large differences between both groups this can be interpreted as evidence to support that risk exposure and credit constraints are binding constraints. ASSESSMENT OF HYPOTHESIS 26Suggested by Hans Hoogeveen, HDSNP 54 6. ECONOMIC GROWTH PERSPECTIVES 107. In this section we examine various growth scenarios. We assess the likelihood of Tanzania achieving and sustaining high growth rates and the requirements of sustained high growth in terms of investment in both human and physical capital. We also take a look at the implications of sustained high growth for structural transformation. The discussion of sectoral growth rates will provide the basis for an assessment as to whether macro-economic growth projections are consistent with the sectoral prospects for continued high economic growth. Growth Scenarios 108. We start the discussion with some simple illustrations of what the impact of various growth rates on per capita income and poverty levels in Tanzania would be. The scenarios we consider are annual growth rates of real per capita GDP of 2%, 4%, 6%, and 8 percent. Before proceeding to the projections, it is informative to compare these growth scenarios to Tanzania's historical growth performance. Since independence, the average per capita growth rate was a mere 0.7 percent and since the introduction of reforms in 1985, per capita growth has averaged 0.9 percent. As Figure 32 shows, the highest per capita growth rates that Tanzania sustained over a five year period was slightly above 3 percent. However, in most periods it was much less. Figure 32. Average annual per capita GDP growth for five year periods, 1960-2003 4.0% 3.0% 2.0% 1.0% 0.0% -1.0% -2.0% -3.0% 1960- 1964- 1969- 1974- 1979- 1984- 1989- 1994- 1999- 1963 1968 1973 1978 1983 1988 1993 1998 2003 109. International experience provides a more encouraging perspective on Tanzania's growth prospects. During the last ten years (1994-2003), 9 countries have been able to achieve an average growth rate of per capita GDP of 6 percent or more, 12 countries at rates between 4 and 6 percent, and 64 countries have been growing between 2 and 4 percent. Thus based on international experience, growth rates between 2 ­ 8 percent seem to be within the realm of the possible. 55 Figure 33. Average per capita real GDP growth in 185 countries during the last decade. 80 70 60 50 40 30 1999-2003 20 1994-2003 10 0 > 6% 5% 3% 1% -1% -3% -5% < -6% 110. A strand of the economic literature (e.g., Sachs and Warner, 1995?) argues that geographical factors play an important role in determining a country's long term growth prospects. While some of the geographic and locational factors that are thought to limit economic growth are present in Tanzania, overall it seems that Tanzania's location and geographical features are not likely to constrain Tanzania to permanently low growth rates (add Sachs Warner simulations, potential story) 111. Table 19 shows projections of per capita GDP as well as associated poverty rates for the growth scenarios discussed above. If per capita GDP were to grow at 2 percent annually, it would increase from currently US$ 290 to $ 448 by 2025. If it grew on the other hand by 8% annually, GDP would reach US$ 1577. The results of these calculations are quite sobering, since they imply that even with growth rates that are significantly higher that past performance, Tanzania will still be a relatively poor country in twenty years. On the other hand, a more positive interpretation would be that many countries with income levels between US$1000 and US$1500 are quite similar to Tanzania. This would suggest that reaching such income levels within a relatively short period of time and the implied high growth rates may indeed be achievable. 112. We now turn to the question of what impact various growth scenarios would have on poverty. At growth rates of 2 percent and 8 percent, poverty levels would decline to 13% and well below 10 percent, respectively. The projections of poverty levels need to be considered with a grain of salt, especially for estimates in the outer year's and in cases where poverty has declined significantly. This is due to the fact that inequality is likely to increase with faster growth and the income elasticity of poverty is likely to fall as poverty declines. In making these projections it is important to note that countries with per capita incomes of US$ 1000 ­ US$1500 have typically much higher poverty rates than those projected by our poverty simulations. 113. Even current poverty levels in Tanzania are below what would be expected given Tanzania's per capita income. According to WDR 2005 which uses which uses the share of the population with a daily income of less than a dollar (purchasing power adjusted), poverty in Tanzania in 2001 was 19.9% with a per capita GDP at purchasing power parity 56 of $ 600. Compare this to Madagascar with a PPP GDP $ 800 and poverty of 49.1%, Malawi with a PPP GDP of $ 600 and poverty of41.7%, or Kenya with PPP GDP of $ 1020, poverty 23%. So even without projecting into the future, poverty in Tanzania seems out of line with its GDP (at least in comparison to nearby countries). The issue is further complicated that PPP GDP for Tanzania seems to be considerably underestimated when compared to Uganda and Kenya. Nonetheless, part of the difference is due to low inequality in Tanzania. Tanzania's Gini coeeficient, a measure of inequality is 0.38, while for Madagascar it is 0.48, Malawi 0.50, and Kenya 0.51. If we would increase the Gini for TZ from 0.34 to 0.45 (using HBS estimates rather than WDR estimates), poverty would increase from, 19.9% to 33.3%. A substantial increase, but poverty would still be relatively low. At least from a regional perspective. Table 19. Projections of per capita GDP and share of population below poverty line, 2010-2025 Year Average annual real GDP growth rate 2% 4% 6% 8% 25% 19% 13% 10% % below poverty line 2010$333 $382 $436 $497 GDP per capita 21% 11% <10% <10% 2015$368 $464 $584 $730 16% <10% <10% <10% 2020$406 $565 $781 $1073 13% <10% <10% <10% 2025$448 $687 $1045 $1577 Source: author's calculations Figure 34. Projections of GDP per capita and poverty 2003-2025 1600 40% 1400 2 35% 1200 4 30% 6 1000 2 25% 4 8 800 6 20% 600 8 15% 400 10% 200 5% 0 0% 04 06 08 10 12 14 16 18 0 22 24 01 03 05 07 09 11 13 15 17 19 21 23 20 20 20 20 20 20 20 20 202 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Policy-based Projections 114. Policies and institutions are an important factor in the determination of growth. This section uses the estimated relationship between economic growth and various indicators of the quality of policies and institutions in Tanzania to provide an assessment of Tanzania's growth potential. 115. The basis for these projections provides work undertaken at the World Bank (Easterly, 1999) that links aggregate assessments of policies and institutions to economic 57 growth. This research indicates that policy based predictions generally are closer to eventual outcomes than traditional projections based on assumptions concerning investment and investment productivity. In addition to the policy indicator, a variety of conditioning variables, including initial per-capita GDP, telephones per capita (a measure of infrastructure), primary enrollment (a measure of human capital), and lagged per- capita GDP growth, are used to predict future growth. Box 3. Impact of changes in policies and institutions on economic growth The following list compiled by William Easterly provides examples of specific improvements in policies and institutions that are estimated to lead to a one percentage point increase in economic growth: · Increase of 1.2 years in average schooling of labor force · An increase in secondary enrollment of 40 percentage points · A reduction of 28 percentage points in the share of central bank credit in total credit · An increase of 50 percentage points in financial depth (M2/GDP) · An increase of 1.7 % of GDP in public investment in transport and communication · A fall in inflation of 26 percentage points · A reduction in the government budget deficit of 4.3 percentage points of GDP · A fall in the black market premium on the exchange rate of 36 percentage points · An increase in (exports+imports)/GDP of 40 percentage points · A fall in government consumption/GDP of 8 percentage points. · An increase in foreign direct investment/GDP of 1.25 percentage points Compiled by William Easterly based on the following sources: Barro 1991, Barro and Lee 1993, King and Levine 1993, Easterly and Rebelo 1993, Fischer 1993, Easterly and Levine 1997, Easterly, Loayza, and Montiel 1997, Borensztein, De Gregorio, and Lee 1994. 116. The underlying model is estimated from cross-country regressions of per-capita GDP growth in 1990-97 for between 75 and 90 countries, depending on data availability. Since assessments of country policy and institutions may differ depending on which criteria are used, we estimate the model using four different assessments of policies and institutions: - the World Bank's Country Policy and Institutional Assessment27 - the Institutional Investor assessment28 27The World Bank's Country Policy and Institutional Assessment (CPIA) is based on Bank economists' and sector specialists' rating of 20 items in the areas of management of inflation and current account, structural policies, policies for social inclusion and equity, and public sector management and institutions. 58 - the ICRG assessment29, and - the Euromoney assessment.30 117. The original quantitative assessments of each of these four institutions are re- scaled so as to measure the quality of institutions and policies on a range from 1-6, where 1 indicates poor quality and 6 high quality. Table 20 presents projections of the last CEM which were based on the country ratings in 1999 and updated projections based on the ratings in 2004. In 1999, we projected that per-capita growth would be around 1.7 percent if policies and institutions remained unchanged but would increase to about 2.5 percent if there were improvements. As can be seen from Table 20, the country risk ratings by all four agencies show an improvement of 0.5 points between 1999 and 2004. Actual annual per capita GDP growth for this period was 3.2 percent and thus exceeded the 1999 projections by 0.7 Table 20. Policy based growth projections Projections Projections Rating 1999 Constant Improve +0.5 Rating 2004 Constant Improve+0.5 CPIA 3.5 1.8% 2.9% 3.9 5.0 6.1 EUROMONEY 2.2 1.7% 2.3% 2.9 4.2 4.9 ICRG 3.4 1.4% 1.9% 3.9 4.4 4.9 INSTITUTIONAL 2.0 1.9% 3.1% 2.2 3.5 4.8 AVERAGE 2.8 1.7% 2.5% 3.2 4.3 5.2 Source: Author's calculations 118. Based on the country ratings for 2004, the projections of per-capita growth range between 3.5% and 5.0 percent, with an average of all four projections of 4.3 percent. If there are further improvements in institutions and policies, the projections suggest that even per capita growth rates of over 5 percent per annum are feasible. 28Institutional Investor credit ratings are based on a survey of leading international bankers who are asked to rate each country on a scale from zero to 100 (where 100 presents maximum creditworthiness). Institutional Investor averages these ratings, attaching greater weights to respondents with greater world wide exposure and more sophisticated country analysis system. 29ICRG compiles monthly data on a variety of political, financial, and economic risk factors to calculate risk indices in each of these categories as well as a composite risk index. Five financial, thirteen political, and six economic factors are used. Political risk assessment scores are based on subjective staff analysis of available information. Economic risk assessment scores are based upon objective analysis of quantitative data and financial risk assessment scores are based upon analysis of a mix of quantitative and qualitative information. The political risk measure is given twice the weight of financial and economic risk. 30Euromoney country risk scores are based on the weighted average of quantitative indicators in nine categories: political risk (25%), economic performance (25%), debt indicators (10%), debt in default or rescheduled (10%), credit ratings (10%), access to bank finance (5%), access to short-term finance (5%), access to capital markets (5%), and discount on forfeiting (5%). For items on which no data are available, the rating is 0, which might introduce a down-ward bias for countries like Tanzania, for which data availability is often poor. 59 Input-based Projections 119. An alternative way to assess Tanzania's growth prospects is to use the growth accounting framework and look at the likely development of the immediate determinants of growth, i.e., human resources, physical capital, and total factor productivity. Contribution of increased investment in human resources 120. The average years of education of the workforce is generally used as the basic indicator of the quality of human resources in an economy. According to Cohen and Soto (2001), in 2000 (Figure 35), Tanzania's average years of schooling (3.4) was higher than that of Uganda (3.22), but lower than in Kenya (5.08), and far from that in South Africa (7.22). Looking forward, it is useful to look at the experience of other economies to assess the possible progress in this indicator. Figure 35 suggests that countries such as Malaysia have been able to increase average years of schooling by about one and a half years per decade. Other countries that have also invested heavily in education such as Mauritius or Kenya have seen increases by about 1 year per decade. Uganda, Tanzania or Ghana have in the past seen much slower increases in average years of schooling. However, recent aggressive efforts to expand access to education, if sustained, are likely to result in a rapid increase in years of schooling over the coming decades Figure 35. Average Years of Schooling, 1970-2002 Average Years of Schooling (Cohen-Soto) 9 Tanzania Ghana Kenya Uganda Mauritius Malaysia Malaysia 8 South Africa Mauritius 7 6 Kenya 5 South Africa Ghana 4 Tanzania 3 Uganda 2 1 1960 1970 1980 1990 2000 Source: Cohen and Soto, OECD 2001 121. In the following we thus look at the likely impact on economic growth of increases by 0.5, 1 and 1.5 years of schooling per decade. Table 21 shows that such increases in years of schooling would result in contributions to economic growth of 0.8, 1.4 and 2.1 percentage points. In interpreting these results, it is important to recognize 60 that years of schooling respond only gradually to increases in enrollment as more educated younger cohorts replace less educated, older cohorts. Thus, even if a policy of universal primary education is in effect, the share of the population with no education will decline only gradually. Our low case scenario for growth in years of schooling assumes that universal primary education is not achieved and that the increases in post- primary education are modest. The medium an high growth scenarios on the other hand, assume that universal primary enrolment is achieved and that there is also a substantial increase in post-primary education. Table 21. Impact of additional years of schooling on economic growth. Highest level of education low medium fast reached 1992 2001 2011 2011 2011 No 24.9 25.2 20 15 15 Adult Only 3.3 2.1 2 2 2 Primary 1-4 15.2 11.9 11 11 8 Primary 5-8 50.7 54.6 58 58 55 Form 1-4 3.9 4.6 6 9 11 Form 5-6 0.3 1.4 2 3 5 Dimploma/University 0.2 0.4 1 2 4 Average years of education 3.8 4.2 4.7 5.2 5.7 Contribution to growth 0.7 0.8 1.4 2.1 Note: Years of schooling are calculated based on HBS figures and differ from Data by Cohen and Soto (2001) due to different methodology used. Contribution of increases in investment to economic growth 122. In the growth accounting framework, the relationship between growth and investment is more complex than in a simple incremental capital output ratio (ICOR) model. Changes in the stock of capital depend on the rate of capital depreciation, the share of GDP invested, and the level of GDP. For example, Figure 36 illustrates that for a given growth rate of output (4%) and investment ratio (20%), the contribution of investment to GDP growth increases over time. This is due to the fact that with increasing GDP, the real amount of investment increases, even with a constant investment ratio. 61 Figure 36. Contribution of capital accumulation to growth, with 4 % growth of output per worker and 20% investment of total output 3.00% 2.50% 2.00% 1.50% 1.00% 0.50% 0.00% 5 9 3 7 1 5 200 2007 200 2011 201 2015 201 2019 202 2023 202 123. Table 22 shows the projected contribution of investment to growth for various combinations of overall growth rates and investment ratios. The higher the investment ratio and the growth rate, the higher is the contribution of investment to growth. For example, if output per worker is stagnant and the investment ratio 18 percent, the contribution of investment to growth is 0.3 percent. On the other hand, in a high growth, high investment scenario, the average contribution could be 2.7 percent. Of course, these projections need to be considered carefully since investment spending does not automatically translate into an equivalent amount of productive capital. The composition of total investment into private and public investment and the share of public investment used for directly productive purposes (i.e., infrastructure) are important factors that affect the relationship between investment and its contribution to economic growth. Table 22. Contribution of investment to growth (average over 10 years) Investment as share of GDP 16% 18% 20% 22% 0% -0.1% 0.3% 0.8% 1.2% Growt 2% 0.3% 0.8% 1.2% 1.6% h rate 4% 0.7% 1.2% 1.7% 2.1% 6% 1.2% 1.7% 2.2% 2.7% Contribution of Total Factor Productivity to Economic Growth. 124. Enhancing factor productivity refers to the ability to obtain more output from a given amount of resources, be it human or physical capital. In the long run, there are essentially two types of constraints to productivity: policy imposed constraints and distortions, and knowledge constraints. 62 125. In the short run, productivity can also be constrained by demand, leading to the under-utilization of resources. However, in the absence of other constraints, such demand constraints would be either temporary or the economy would adjust appropriately. Table 23. Growth and Total Factor Productivity in Selected East Asian Countries, 1960-1994 Indonesia Korea Malaysia Thailand Taiwan Percentage points per year Growth of output per 3.4% 5.7% 5.7% 5.0% 5.8% worker Contribution of total 0.8% 1.5% 1.5% 1.8% 2.0% factor productivity Source: Bosworth and Collins (1996) 126. Cross country studies on factor productivity provide an indication of possible productivity growth rates for Tanzania in the long run. A study by Prescott and Parente (1999) indicates that such measures could lead to an threefold increase in output through the more efficient use of existing capital and the adoption of new technologies. Looking at the experience of South-East Asian countries, some studies indicate that about half of their fast growth is attributable to productivity gains. More conservative estimates (Bosworth and Collins, 1996) still attribute between 0.8 and 2.0 percentage point of their growth of output per capita to increases in total factor productivity. In assessing Tanzania's growth potential, it appears thus clearly possible that about 1 percentage point of growth per worker could come from productivity gains. In the immediate future, the contribution of productivity gains to output growth can be expected to be even higher, given the potential for more productive use of the existing capital stock. Assessment of growth potential based on projected factor accumulation and productivity increases 127. This section aggregates the individual projections of the potential contribution of physical capital accumulation, education, and total factor productivity to total growth. We present three scenarios, which describe which use the lower and upper estimates of the contribution of each factor as well as an intermediate scenario. Table 24. Overall, input based projections Low Medium high Physical Capital 0.3 1.2 2.7 Education 0.8 1.4 2.1 Total Factor 0.5 1.0 1.5 Productivity Total 1.6 3.6 6.3 Note: Low: investment/output ratio of 16 percent at 2 % output growth, increase in average years of schooling by 0.5 years 63 Medium: investment/output ratio of 18 percent at 4 % output growth, increase in average years of schooling by 1.0 year High: investment/output ratio of 22 percent at 6 % output growth, increase in average years of schooling by 1.5 years 128. The resulting aggregate output projections suggest that the policy based projections of 4.3 to 5.2 percent seem to be achievable, although the input requirements in terms of investment, education, and factor productivity are quite demanding. In interpreting the results shown in Table 24 it is important to note that even the "low" scenario requires that the performance of the last decade is maintained. Any reversals clearly could result in much lower growth. Sectoral Projections 129. Development and economic growth is characterized by structural transformation of the economy. The share of agriculture typically declines, while that of industry and services increases. Table 25 shows the changes in the composition of GDP for a number of countries, which illustrates the potential magnitude of transformation possible over a twenty year period. Fast growing countries such as Ghana, Thailand, India, and Indonesia experienced relative rapid structural transformation, while in slow growing economies such as Kenya or Tanzania, the pace of structural transformation was also more limited. Table 25. Structural Transformation, selected countries 1980-1998 Share in GDP of Agriculture Industry Manufacturing Services 1980 1998 1980 1998 1980 1998 1980 1998 Ghana 58% 37% 12% 25% 8% 8% 30% 38% India 38% 25% 24% 30% 16% 19% 39% 45% Kenya 33% 29% 21% 16% 13% 10% 47% 55% Thailand 23% 11% 29% 40% 22% 29% 48% 49% Indonesia 24% 16% 42% 43% 13% 26% 34% 41% Tanzania 45% 46% 18% 14% 12% 7% 37% 40% Source: WDR, 1999 130. For Tanzania, we examine three scenarios. The baseline scenario projects current sectoral growth rates into the future. We compare this scenario to one with higher aggregate growth of and faster structural transformation and one with lower growth and limited structural transformation. Table 26. Scenarios for economic growth and structural transformation Slow Growth Medium Growth Fast Growth Share Avg. real Share in Avg. real Share in Avg. real Share in in GDP growth rate GDP growth rate GDP growth rate GDP 2003 2004-2025 2025 2004-2025 2025 2004-2025 2025 Agriculture 46.8% 3.3% 40% 3.7% 33% 5.0% 25% 64 Industry 18.5% 4.8% 22% 7.8% 30% 10.9% 33% Services 34.8% 4.4% 38% 5.7% 37% 8.9% 42% Total 100.0% 4.0% 100% 5.4% 100% 8.0% 100% 131. Under the baseline scenario, the share of agriculture in GDP drops to 33% by 2025 while the shares of industry and services increase to 30 and 37 percent respectively. Comparing these projections with international experience suggests that these sectoral projections are consistent with the pattern of structural transformation observed in other economies, although it seems likely that industry and services grow at similar rates, which would imply a slightly higher share of services and a slightly lower share of industry in GDP by 2025. 132. One of the key characteristics of structural transformation is that the industry and service sectors have to grow faster than agriculture. The underlying process consists of agricultural surplus which is invested in the industry and service sector and that higher productivity in the agriculture allows the movement of labor from agriculture to other sectors. 65 References Collins, Susan and Barry Bosworth. Economic Growth in East Asia: Accumulation versus Assimilation. Brookings Papers on Economic Activity, 2:1996. IMF. Tanzania ­ Selected Issues and Statistical Appendix. July 22, 2004. Hausmann, Ricardo, Dani Rodrik, Andreas Velasco. Growth Diagnostics. September 2004. Mans Soderbom, Francis Teal, Anthony Wambugu and Godius Kahyarar. The Dynamics of Returns to Education in Kenyan and Tanzanian Manufacturing. Centre for the Study of African Economies (CSAE), University of Oxford, CSAE WPS/2003-17, January 2004. Rodrik,Dani. Saving Transitions. Draft. July 1998. RPED. Investment Climate Assessment ­ Constraints on Enterprise Performance and Growth in Tanzania. Draft. April 14, 2004. Schneider, Friedrich. The Size of the Shadow Economies of 145 Countries all over the World: First Results over the Period 1999 to 2003. IZA Discussion Paper Series. IZA DP No. 1431. December 2004. The World Bank. United Republic of Tanzania ­ Public Expenditure Review FY03. June 2003. 66 1. Introduction The Tanzanian economy has changed considerably during the decade covering 1991-2001. This paper provides an assessment of the monetary and non-monetary poverty and explores the correlates of poverty at the micro-level using the Household Budget Surveys (HBSs) from 1991/1992 and 2000/2001. These datasets along with a comprehensive discussion of poverty on a descriptive level are presented in the HBS report (2002). Information includes household expenditure and consumption, education levels, asset holdings, and access to services. Furthermore, the 2000/2001 HBS has detailed information on health status and food security. The sample from the first HBS consists of 4328 households (rural, urban, and Dar es Salaam), whereas the most recent HBS holds information on a much larger sample, namely 22,178 households. Both are representative on the stratum level. We perform reduced form regression analysis of the correlates of household consumption and investigate whether the observable characteristics in hand contribute to explaining observed trends in poverty and inequality. The paper pays special attention to rural-urban differences. 2. Trends in monetary poverty and inequality 2.1. Monetary poverty: Levels and changes As pointed out by Demombynes and Hoogeveen (2004) the period from 1991-2001 has been characterised by low growth during the first half of the decade and high growth during the last and this development has, in total, translated into a decline in monetary poverty. The development in poverty can be seen from Table 1 that shows the fall from 38.6% to 35.6% in headcount poverty using the basic needs poverty line and per adult equivalent consumption. Individual and regional weighting along with adjustments made for price variation over time are implemented as in the HBS report (2002). [Box 2.1 on Consumption aggregate and the poverty line around here] The change in poverty conceals large regional differences in levels of poverty as can be seen from Table 2 and 3. We consider 7 regional zones: Coastal (Tanga, Morogoro, Pwani), Northern Highlands (Arusha, Kilimanjaro), Lake (Tabora, Rukwa, Kigoma, Shinyanga, Kagera, Mwanza, 1 Mara), Central (Dodoma, Singida), Southern Highlands (Iringa, Mbeya, Rukwa), South (Lindi, Mtwara, Ruvuma), and Dar es Salaam. As mentioned above the data are only representative on the stratum level. Hence, the estimates on zone level are likely to be imprecise and should be interpreted with great care. This is also clear from the bounds on the confidence intervals. Significant declines in poverty are observed in the Southern Highlands and Dar es Salaam whereas the Northern Highlands has experienced a significant increase in poverty. Interestingly, the Southern Highlands was one of the poorest zones and the Northern Highlands was the most well-off zone in 1991/1992. Both in 1991/1992 and 2000/2001 do we observe lower levels of poverty in urban areas, especially in Dar es Salaam and the Southern Highlands, when compared to rural areas. This is illustrated in Table 4, Figure 1, Figure 2, Table 5, and Table 6. Furthermore, even though all areas have lower average levels of poverty in the latest survey, the decrease in poverty is by far the largest in Dar es Salaam and urban parts of the Southern Highlands as well. The decomposition of the poverty level into zones and urban-rural areas shows that the increase in poverty in the Northern Highlands is driven by an increase in poverty for rural areas whereas the decrease in poverty in the Southern Highlands is observed for both urban and rural areas. The geographic composition of the population has changed too: A larger share of the population lives in Dar es Salaam at the time of the most recent survey. Obviously, the poverty line lies in the second quintile of the consumption value distribution, yet there is great variation between the distributions of households across Dar es Salaam, other urban, and rural areas. This is illustrated in Table 7. The share of households in each quintile in rural areas is roughly 20%, which is not surprising since approximately 80% of the households reside here and thus dominate the distribution. Contrary to this, a much larger proportion of the households in Dar es Salaam and other urban areas is in the top quintile. The change in poverty can be decomposed into a growth component, an inequality component, and a residual.1 Table 8 shows such decomposition for mainland Tanzania, Dar es Salaam, other urban, and rural areas. Growth has clearly decreased poverty but at the same time, higher inequality has 1See Datt and Ravallion (1992) 2 worked in the opposite direction. This is true for all areas but the magnitudes of both effects are by far the largest in Dar es Salaam. Alternatively, the change in poverty can be decomposed into regional composition effects: There are inter-regional effects, population shift effects, and interaction effects.2 This is depicted in Table 9 for Dar es Salaam, other urban, and rural areas. Most of the change over the decade has happened within these areas, yet 12.0% comes from population shift effects. This fits with the fact that a significant larger share of the 2000/2001 population lives in Dar es Salaam where the reduction in poverty has been the largest. 2.2. Levels and trends in inequality Not only has poverty levels changed during the decade, inequality has changed as well. We calculate a number of percentiles of the consumption value distribution (measured by the value of per adult equivalent consumption) to shed light on the distribution of consumption value distribution among Tanzanian households. For example, the 50th percentile of the consumption value distribution is the level of consumption where 50% of the households have lower consumption and 50% have higher consumption. Fractions between different percentiles of the consumption value distribution depicted in Table 10 are fairly constant over time, showing that there has not been much change in the distribution of means in the economy. The only sizeable changes are seen with regards to the top of the distribution: Individuals on the 90th percentile have become better-off compared to both the 50th and the 10th percentile. Table 11 and 12 show Gini coefficients and Theil indices for different population groups. Both the Gini coefficient and the Theil index indicate an increase in inequality but neither measure has changed considerably. In general, the level of inequality is low compared to other African countries.3 Considering urban-rural levels of inequality in 1991/1992 we see that Dar es Salaam is by far the most equal of the three, yet this changes complete over the decade. We observe a small reduction in inequality in rural areas and a minor increase in other urban areas but Dar es Salaam becomes much more unequal. As with headcount poverty there is substantial regional variation in inequality both in terms of level and change over time. Again these estimates should be interpreted with care. In 1991/1992 the poor Southern Highlands has the highest level of inequality within the 2See Huppi and Ravallion (1991) 3 regions whereas the well-off Northern Highlands has the lowest levels. In 2000/2001 this has changed such that the Central area has the highest level of inequality. The Southern Highlands experience a significant decrease and the Northern Highlands a significant increase in inequality. There is variation in the level of inequality for households with different levels of education and number of household members but the variation is not particularly systematic. The Theil index can be decomposed into within- and between-group inequality.4 Table 13 shows the share of the between-group inequality of total inequality for different groups of the population. Clearly, the within-group estimate contributes the bulk of the inequality no matter what grouping we consider. However, the shares of inequality between groups with different levels of educations and for households of different sizes are much larger than the shares of between-group inequality for other groupings. Furthermore, except for number of household members, the share of the between-group inequality has increased over time. 3. Non-monetary poverty and access to services Before we turn to our reduced form regression analysis of the correlates of household consumption it is worthwhile to consider other measures of poverty. 3.1. Non-monetary poverty: Levels and changes Table 14 shows (farming and fishing related) asset holdings by poverty status. There are virtually no differences between poor and non-poor households and means for the group of poor are higher than for non-poor. This is not surprising given that the type of assets considered is farming related. In Table 15 we investigate this further by considering asset holding across the consumption value distribution. This shows that the higher the consumption of the household, the smaller the probability of owning farming related assets. The tendency is much clearer though, for access to communication and information equipment such as telephone, radio, and television as depicted in Table 16: The poor have to a much smaller degree access to such devices and even though ownership has increased overall, non-poor households have become even better off in this dimension over time. Table 17 shows access across the consumption value distribution, and this supports the findings from Table 16. 3Mozambique, 1997: 0.40, Uganda, 1999: 0.43, Kenya, 1997: 0.45. 4 Table 18a, 18b, and 18c show housing quality for Mainland Tanzania, Dar es Salaam, other urban, and rural areas measured by type of foundation, type of floor, and type of roof. Housing quality has increased overall but as shown in Table 19, 20, and 21, richer households live in better housing and well-off household have become better off over time. Another important non-monetary measure is food security. Unfortunately, this information is only available in 2000/2001. Table 22 shows that perceived problems with satisfying food needs is highly correlated with actual poverty status. Furthermore, as shown in Table 23 there is large variation in food patterns across rural and urban areas. In particular, high protein types of food such as meat, fish, eggs, and beans are most often consumed in Dar es Salaam whereas the consumption of dairy is higher in rural areas. Our final measures of non-monetary poverty concern incidence and types of sickness along with types of treatment. A significant share of the Tanzanian population is sick during a 4-week period and the majority, 63.8%, is sick with malaria. This is shown in Table 24. Furthermore, there is hardly any difference between the share of sick among poor and non-poor in a given area but the incidence of illness is particularly pronounced in rural areas. Table 25 shows the distribution of the incidence of poverty across the consumption value distribution. Again, there is no indication that the incidence of sickness varies across the consumption value distribution. Furthermore, Table 26a, 26b, and 26c document that sick poor and non-poor are equally likely to use all types of health care but there are significant differences between rural and urban areas. Table 27 shows the use of health care across the consumption value distribution. The richest 20% are more likely to use private hospitals and less likely to use traditional healers than the poorest 20% and the difference is significant. This is not captured when considering only differences between poor and non-poor. The confidence intervals bounding the means are generally large. Table 28 shows that impediments to use of health care such as distance and lack of financial means are no more common among poor than among non-poor. However, as shown in Table 29, the 4See Shorrocks (1984). The decompositions are implemented using S. P. Jenkins' Stata program, ineqdeco 5 richest 20% are more likely to state that they consider medical care too expensive or too far away compared to the poorest 20% of the population and the difference is significant. 3.2. Access to services In terms of access to different services there seems to be some variation across poor and non-poor households. Table 30a and 30b present average distance to health care services, primary schools, markets, transport, and water in dry seasons by poverty status. The discrepancies are, however, not as clear-cut as one might expect. In general, the group of poor lives further away from all types of services but in most cases the differences are very small. Distances to primary schools have increased overall whereas distances to markets have decreased. Table 31 depicts distances by quintiles. Here it becomes clear that the increase in distance to primary school is due to an increase for the poorest 40% of the population. Similarly, the decrease in distance to markets only holds for the richest 60%. Table 32 shows use of financial services among poor and non-poor. This is likely to be highly correlated with access to this type of service. First of all poor households have much lower usage of financial services, be it formal or informal. The difference even seems to have increased over time. Secondly, usage has decreased overall. Table 33 shows use of financial services by quintiles and the conclusions from this table corresponds to those from Table 32. Being poor unambiguously translates into unfavourable housing services: Table 34 shows the share of households with access to piped water and electricity and the tendency is clear. Furthermore, the expansion of services seems to have benefited the non-poor. This is particularly clear from Table 35 that depicts access by quintiles. Finally, we consider school enrolment. Table 36 shows enrolment of 7-13 year olds by area and gender. Enrolment is higher for girls in 1991/1992 but increases over time for all areas but more so for boys resulting in similar enrolment rates for both genders in 2000/2001. Furthermore, enrolment is much higher in Dar es Salaam than elsewhere. Table 37 shows enrolment by quintiles. We see that enrolment generally increases with value of consumption and improves for all groups over time. One exception is households in the lowest quintile in 1991/1992. These households have 6 exceptionally high enrolment rates compared to the other quintiles but this may be a reporting problem. In conclusion, there has not been much increase in welfare measures outside Dar es Salaam, neither in terms of monetary poverty, non-monetary poverty, nor access to services. In fact, non-monetary poverty measures and access to services show a worse story than the monetary measures. This is extremely important as this makes the poor even more vulnerable to negative shocks and potentially affects poverty status of future generations. One important exception from this is school enrolment rates. Simultaneously, inequality has increased, yet this is completely driven by an increase in Dar es Salaam. Actually, there is a slight indication that inequality has decreased in rural areas. In the following we will take a closer look at the structure of income and consumption and investigate to what degree we can explain the development in monetary poverty: Do individuals in urban areas have particularly favorable characteristics with regards to reducing the likelihood of poverty or is the environment advantageous in the sense that returns to individual characteristics, for example schooling, are higher? Furthermore, can changes in observable characteristics or returns rationalize the variations in inequality? 4. The structure of consumption and income An important addition to the analysis of poverty is the examination of the structure of consumption and income within households. The questionnaire from 1991/1992 lacks detailed questions on consumption and income compared to the 2000/2001 survey. Therefore, we only consider 2000/2001 information. Table 38 presents the composition of consumption where medical care, education, water, and transport have been added back into the consumption measure. On average, the share of food out of total consumption amounts to 72.4%. Medical care, education, and water sum to 3.2%, 4.1%, and 5.0%, respectively. Table 39 demonstrates that these shares are not constant over the consumption value distribution. In particular, the food share seems to decrease with higher total consumption: The share of food for the lowest quintile is 74.3% and 62.8% for the highest quintile. Table 40 depicts the shares of different sources of income out of total income. Here, we have added back the value of non-monetary consumption to total income from the questionnaire. The majority 7 of income in mainland Tanzania stems from agriculture followed by non-agricultural self employment and paid employment. However, this varies greatly with area: Income from agriculture amounts to only 2.0% in Dar es Salaam where paid employment contributes the largest share. Conversely, in rural areas agriculture dominates and paid employment sum to 6.6%. Interestingly, households in Dar es Salaam and other urban seem to rely more on remittances as a source of income than households in rural areas. Hence the average flow of income runs from rural areas to urban areas. Sources of income vary with the distribution of the value of consumption as well as shown in Table 41. Agriculture contributes the majority of income for poor households whereas paid employment is more important for richer households. Concerning remittances, there does not seem to be differences across quintiles. To sum up, we find that the share of food out of total consumption is decreasing with the value of consumption. Still, the value of food consumption out of total consumption is quite high for all quintiles. This means that not only poor households are vulnerable to negative income shocks in the sense that their food consumption could be severely affected. Furthermore, poor households rely on income from agriculture whereas richer households are more likely to have income from paid employment or self employment. We will touch on this in Section 5 as well. 5. Descriptive statistics This section discusses household characteristics in relation to poverty and considers the question of whether households in urban and rural areas differ with regards to the levels of observable characteristics. Table 42 presents levels of poverty according to the number of household members. Not surprisingly, higher numbers of household members correlates with higher levels of poverty. The variation is large ranging from 5.8% for households with only one member to 57.2% for households with 10 or more members. There does not seem to be much change over time with regards to the level of poverty for a given household size but there is a tendency for households to be smaller in 2000/2001. 8 One reasonable explanation for the positive correlation between the number of household members and poverty is that larger households have more dependants, i.e. older people or children. Table 43 shows the relationship between the number of children aged 5 or younger and poverty. Here we find that the presence of more young children in the household is correlated with higher levels of poverty. Furthermore, a larger share of the population live in households with more children in 2000/2001 compared to 1991/1992. Table 44 shows poverty rates by the civil status of the head of the household. There are no substantial disparities in poverty between married and widowed heads of households. Divorcées, however, have much lower levels of poverty than the other two groups. There is some change in the composition of household heads by civil status over time. The significantly higher share of households with a widowed head is interesting and possibly related to the HIV/AIDS pandemic. Table 45 depicts a clear positive correlation between level of completed schooling and poverty. Importantly, a larger share of the population lives in a household with a more educated head in 2000/2001 compared to 1991/1992. Table 46 shows levels of completed schooling for head of household by quintiles. This table confirms the findings from Table 45. Employment patterns differ substantially between poor and non-poor as well.5 This is shown in Table 47. Poverty levels are significantly higher among farmers and fishermen as well as for family workers and household heads without economic activity compared to paid employees and (non agricultural) self employed. Interestingly, households are moving out of farming or fishing in 2000/2001 compared to the earlier survey and poverty levels among this group seem to have gone down slightly as well. At the same time population shares of paid employees and self-employed have increased considerably. Since farmers and fishermen together constitute the largest employment group this development is likely to be among the main drivers of the total reduction of poverty. Table 48 shows employment patterns by quintiles. Households in the lowest quintiles dominate farming and fishing whereas the richest 20% of the population are more likely to be in paid or self employment. 5The employment information is not directly comparable in the two surveys as discussed in HBS report (2002). Hence, caution should be taken when interpreting the changes over time. 9 Poverty and employment patterns for the head of the households give useful information on both income sources and vulnerability. A slightly different measure is the main source of income shown in Table 49. This table documents that there are large discrepancies between the employment of the head of the household and sources of income. In particular, the share of the population with main income from crops or livestock is much smaller than the share with farming or fishing as main employment. In other words: Secondary employment activities as well as income generated by activities of other household members are important as well. Another indirect measure of potential income and vulnerability is business involvement. Table 50 shows poverty by main type of business of the head of the household. There is not much change over time in the share below the poverty line within a particular type of business except for wholesale and retail where poverty has decreased significantly. Furthermore, the poor are less likely to have business at all. Finally, the share of the population without any business has increased over time. This is consistent with the increase in paid employment. Table 51 shows the type of business of the head of household in Dar es Salaam, other urban, and rural areas. The share without any business is the highest in rural areas and if the household is involved in any business it is most likely agriculture. Finally, Table 52 depicts main type of business by quintiles. The main message is that business ownership is increasing in the value of consumption. To sum up, the poor and non-poor households have remarkably different characteristics with regards to family composition as well as type of and access to income sources. 5.1. Rural-urban differences The question is now whether some of trends described above vary by rural and urban areas. Table 53, 54, and 55 show statistics on selected characteristics and it is worth noticing that rural and urban households have become more alike in terms of both household composition and with regards to economic variables. Rural households have more members on average, but household sizes have decreased in all areas and the decrease has been stronger in rural areas. Rural households do also have more children under the age of 6 but an increase in average numbers of young children among households in Dar es Salaam and other urban areas has made all household more comparable in this dimension as well. Regarding the level of education, it is clearly higher in Dar es Salaam in both surveys but the differences between rural and urban areas seem to decrease over time. 10 Based on this, it is likely that differences in observables, especially differences in level of education, across rural and urban households may explain part of the dissimilitude in poverty levels. Furthermore, since the distributions of observables seem to have become more alike across rural and urban areas the contribution from observables towards the change in inequality ­ if any ­ must go through changes in returns. 6. Correlates of poverty In this section we perform reduced form regression analysis of the correlates of household consumption. We regress log per adult equivalent real consumption on a set of explanatory variables, cf. Datt and Joliffe (1999). The coefficients should be interpreted as the percentage change in per adult equivalent real consumption resulting from a marginal change in the explanatory variable. In the following we discuss the rationale for our included variables and the intuition behind the estimated parameters. None of the effects have a straight forward interpretation and policy advice based on the results should be careful. The analysis does, nevertheless, present a significant addition to the descriptive analysis discussed above. The main advantage of such analysis is that it allows us to isolate the effect of a specific variable holding all other (observable) factors constant. For example, households living in Dar es Salaam are less likely to be poor than households living in rural areas. Yet the level of education is also higher in the former compared to the latter. Regression analysis allows us to separate these two effects. In the regression analysis, we include a set of household composition variables; schooling of the head of the household; a set of observables includes asset information such as number of fields, number of livestock, and ownership of a plough; and a set of employment information variables for the head of the household. Most of the variables are potentially endogenous to consumption and the estimated coefficients should not be interpreted as causal effects. See appendix A1 for further discussion of the included variables. All regressions control for regional effects through inclusion of dummies for the 20 regions that together constitute the 7 zones, see Section 2 above. We estimate separate regressions for Dar es Salaam, other urban, and rural areas in both 1991/1992 and 2000/2001.6 Firstly, we estimate the model without employment information included. Table 56 6We reject the simple model with no variation in coefficients across areas and years. 11 and 57a, 57b, and 57c present these results in two different ways. Table 56 shows the regressions for Dar es Salaam, other urban and rural areas where we have included a dummy for 2000/2001. Furthermore, this dummy is interacted with all explanatory variables and the full set is included. These interaction terms should be interpreted as the differences in 2000/2001 from the 1991/1992 coefficients. Table 57a, 57b, and 57c present regressions for each area in each year. This model is equivalent to the one in Table 56 but the coefficients from the 2000/2001 regressions are now in levels. These coefficients can alternatively be found by adding the level effect in Table 56 to the 2000/2001 interacting term for all variables. In short: Table 56 stresses the changes in coefficients between the two surveys whereas Table 57a, 57b, and 57c stress the levels of the coefficients. We find that household composition effects consumption even though consumption is measured in per adult equivalents. The net effect of more household members is negative independent of age but generally less negative in rural areas. Furthermore, the effect of more adult male members (aged 15- 59) is significantly larger (the coefficient is more negative) than the effect of more adult female members in Dar es Salaam. There is generally no change over time in the effects of family composition. Age of the head of the household has no effect, neither in 1991/1992 nor in 2000/2001. A different model specification with age dummies did not give any economically and statistically significant effects either. To address some of the vulnerability issues discussed below we included in earlier estimated models a dummy for female headed households, civil status variables, and a dummy for raising a foster child along with interaction terms between these. The female headed household dummy did not come out significant no matter the specification and the same hold for civil status, whereas prevalence of a foster child came out significantly positive indicating that households with foster children are families who have the economic means to do so. We chose not to include this due to the obvious endogenity problems related to having a foster child. Education in general leads to significantly higher consumption in all areas compared to no schooling, which is the excluded category.7 Furthermore, more education leads to even higher consumption. This is extremely positive given the recent policy of providing free primary schooling. However, returns may change if the level of education changes for the population in general. The effects of schooling are higher in 1991/1992 in other urban and rural areas compared 7We include type of schooling instead of just years of schooling to allow for non-linear effects of education. 12 to Dar es Salaam. This relation changes in 2000/2001, though, where the effects are more comparable across areas and even higher in Dar for primary schooling and post secondary schooling.8 Agricultural related assets such as number of livestock, fields, and ownership of plough do not have strong effects on per adult equivalent real consumption in 1991/1992, though all are positive in other urban and rural areas and turn significantly positive in 2000/2001 in other urban and rural areas. The variable may be endogenous, though. In particular, the effect of owning a plough is negative in Dar es Salaam in 2000/2001. However, it is highly unlikely that ownership of a plough causes lower consumption (except for those households that lower consumption temporarily to be able to purchase a plough) in general. It is much more likely that the variable picks up part of the effect of being involved in agriculture. The same reasoning could be applied to the effect in other urban and rural areas. The depressing result does therefore not mean that a policy of subsidising or providing ploughs, for example, would not have a positive effect on consumption but we would need other estimation techniques along with better information such as panel data to evaluate such policies. Apart from this set of variables we have tried with a wide range of functional forms of `close to service variables' such as distance to roads, distance to markets, and distance to public transport but did not find any significant effects. Like effects of assets discussed above this does not necessarily mean that infrastructure does not have positive effects on consumption. Our second set of regressions includes employment information. These regressions are depicted in Table 57, 58a, 58b, and 58c. The included categories are paid employment, self employment, family employment, and no employment. The excluded category is agriculture. In 1991 there are no significant differences in returns to different types of employment in Dar es Salaam, whereas family employment leads to significantly higher consumption in other urban and rural areas and paid employment increases consumption in rural areas. This pattern changes completely in 2000/2001. Now, both paid and self employment lead to significantly higher consumption and the returns are almost twice as high in Dar es Salaam compared to other urban and rural areas. Furthermore, no employment reduces consumption in rural areas and family employment increases consumption in 8We reject the simple model with no variation in returns to education across areas and years. 13 other urban areas. Counter intuitively, no employment increases consumption in Dar es Salaam. This could, for example, be caused by differences in the definition of family employment and no employment. The inclusion of the employment information causes very little change in the other parameters. We conclude that Dar es Salaam has pulled away from the other areas with regards to returns to high level schooling and employment. In particular, returns to post secondary education and paid employment exceed returns elsewhere. Along with higher levels of schooling in general and a much larger concentration of paid employees and self employed in the capital area (see the chapter on the labor force analysis) this helps explain why Dar es Salaam fared much better in 2000/2001 compared to other areas. 7. Vulnerability Vulnerability means the probability of being poorer in the future compared to today. Since neither the HBS surveys nor the Integrated Labor Force Survey (ILFS) constitute a panel we cannot perform a proper analysis of the determinants of vulnerability. However, several other sources provide information on identification of vulnerable groups in Tanzania and impoverishing forces, i.e. factors that reduce wellbeing. In the following we will discuss these findings and relate this to the analysis presented above. The Poverty and Human Development Report, World Bank (2003a), the ESRF report (2004), and the Tanzania Participatory Poverty Assessment, World Bank (2003b) all present work on vulnerability in Tanzania. Potential impoverishing forces are economic, environmental, governance, socio-cultural, health, and life-cycle factors. Economic forces stem from changes in national or international policies whereas environmental factors include droughts, flooding, earthquakes, or deforestation. Governance forces are linked to the responsibilities of the government such as prevention of corruption or tax procedures, socio-cultural forces are connected to traditional ways of living the prohibit changes accommodating shocks and health factors include malnutrition and epidemics. Finally, life-cycle forces are introduced through the special needs of for example children, pregnant women, and older people. 14 ESRF (2004) is based on a qualitative survey in the Kagera Region and deals with rural income dynamics. The report stresses coffee booms, introduction of new crops and introduction of new economic opportunities such as tobacco and mining, lumbering and fishing as particularly important for growth and livestock is considered to be an important precautionary means of saving. Furthermore, hard work, cooperation, and human capital investment are thought to be important for upward mobility whereas chronic illness, lack of education, and lack of social services are examples of factors that are related to downward mobility. Some groups in the population have characteristics that make them especially vulnerable. This is known as poverty traps. These groups are identified as children, persons with disabilities, youths (unemployed, youths with unreliable income in general, and female youths), elderly persons, people with chronic diseases, women and in particular widows, and drug addicts as well as alcoholics. The Poverty and Human Development Report, World Bank (2003) proposes a list of measures of extreme vulnerability, namely the proportion of households with less than one meal per day, number of days of sickness, chronic illness, proportion of orphaned children, and the proportion of orphans in the labor force who do not attend school. With the HBS data sets we cannot without a panel data set address all of these potential vulnerability measures, nor test whether they in fact increase the probability of being poor in the future. In particular it is not possible to make comparisons across the surveys based on the information on food insecurity and illness presented above since it is only available in 2000/2001. Given that limited calorie or low-protein intake increase vulnerability it seems, however, that households in rural areas are more vulnerable than households in urban areas, see Table 22. Furthermore, the incidence of sickness is much higher in rural areas, see Table 23. The information on the share of households with a widowed head is comparable over time and the trend is depressing: From Table 43 we see that the share has increased significantly over time.9 8. Conclusion 9Note, though, that the earlier version of the regression analysis did not find that households with widowed heads are more poor than other households. 15 We have analysed the HBS surveys from 1991/1992 and 2000/2001 on poverty. This analysis documents that the increase in welfare measures between 1991/1992 and 2000/2001 outside Dar es Salaam has been depressing. In fact, we have not seen much development neither in terms of monetary poverty, non-monetary poverty, nor with regards to access to services. Furthermore, non- monetary poverty measures and access to services show a worse story than the monetary measures. This is extremely important as this makes the poor even more vulnerable to negative shocks and potentially affects poverty status of future generations. One important exception from this is school enrolment rates. Simultaneously, inequality has increased, yet this is completely driven by an increase in Dar es Salaam. Actually, there is a slight indication that inequality has decreased in rural areas. Several features vary between poor and non-poor households. Firstly, the share of food out of total consumption is decreasing with the value of consumption making poor households more vulnerable to negative income shocks in the sense that their food consumption could be severely affected. Furthermore, poor households rely on income from agriculture whereas richer households are more likely to have income from paid employment or self employment. Also, poor and non-poor households have remarkably different characteristics with regards to family composition as well as type of education and employment pattern. Finally, the distributions of observable characteristics vary between rural and urban areas: Urban households have higher educated heads and are more likely to be paid employees or (non agricultural) self employed. The regression analysis shows that Dar es Salaam has pulled away from the other areas with regards to returns to high level schooling and employment. In particular, returns to post secondary education and paid employment exceed returns elsewhere. Along with higher levels of schooling in general and a much larger concentration of paid employees and self employed in the capital area this helps explain why Dar es Salaam fared much better in 2000/2001 compared to other areas. Box 2.1. The consumption aggregate and the poverty line The consumption aggregate used to calculate per adult equivalent consumption as well as the poverty lines is described in the HBS report (2002). The consumption aggregate sums the value of all items consumed. Durables and non-consumption expenditure are excluded. Furthermore, due to increased costs of health, education, water, postage, and telephone charges that "reflect an increase 16 in payment for similar services" (HBS report (2002), p. 132) these are also excluded. The consumption measure is standardised to 28 days and adjustments are made for the age and gender profile of each household to achieve per adult equivalent consumption. Finally, since prices vary with time and across regions the Fisher Ideal Index is applied to the consumption aggregates (and the poverty lines) to adjust for this, see the HBS report (2002) for further details. We follow the HBS report (2002) and use the basic needs poverty line. This poverty line includes both food expenditure and costs of other essential items of expenditure. Figure B1 shows that the consumption value distribution in 1991/1992 first-order statistically dominates the distribution in 2000/2001. This means all FGT (Foster, Green, and Thorbecke (1984)) poverty measures, including the headcount, poverty gap, and squared poverty gap are higher in 1991/1992 for all poverty lines, see Coudouel, Hentschel, and Wodon (2004). A1. Conditioning set in regression analysis We include a set of household composition variables even though consumption is measured in adult equivalent real consumption. Effects of household composition are difficult to interpret because a household member represents both a potential cost and a source of income. Furthermore, household composition is highly likely to be endogenous to consumption: On the one hand children present potential future income and may therefore be seen as an investment for poor households that do not have access to saving, cf. the discussion above. Following this, poor household are more likely to have children than non-poor households. On the other hand are children themselves a costly consumption good that requires income. This makes it more likely that non-poor households have children or that they take care of relatives' offspring (or relatives in general). Leaving the information out is likely to bias other parameter estimates. See Lanjouw and Ravallion (1995) A second set of variables holds information on completed level of schooling of the head of the household. This information is likely to be very important for poverty status, vis-à-vis the descriptive analysis presented above. Furthermore, education is a hard policy instrument. Schooling is potentially endogenous to consumption because of intergenerational links both between schooling and inheritance in general, cf. Behrman (2002). The effect of schooling is therefore likely to be overestimated. However, completed level of schooling is predetermined to consumption in this period and we are thus free of simultaneity bias. 17 A third set of observables includes asset information such as number of fields, number of livestock, and ownership of a plough. Estimating these effects are particularly interesting for policy makers: If ownership increases consumption policy makers may consider subsidising purchases. However, our analysis only provides an initial indication of the benefits of such a policy and a more thorough analysis of costs and benefits should be carried out. One reason for this is that the inclusion of asset information in the regression analysis is problematic: Ownership requires that the household had means in the past to acquire the asset. This means that the household was relatively well-off at some point in time. It is highly likely that households that were non-poor in the past are well-off in the future. Therefore, we may overestimate the effects of asset holdings. Furthermore, assets work as precautionary savings, especially when households do not have access to financial institutions or are unable to obtain a loan. Therefore, during periods of economic hardship households may sell off assets whereas during booms households are likely to purchase assets. This works in the opposite direction as the former argument: If the household sells off assets they can cover more expenses and have higher consumption but the researcher observes them to have fewer assets. Therefore, the effect may be underestimated. In total it is difficult to say which effect is dominating. Finally, we include a set of employment information variables for the head of the household (paid employment, self employment, employment within the extended family, and no employment) to see whether type of employment is associated with consumption. Clearly, type of employment may be affected by poverty status: It may be difficult to find a job at all if the household is extremely poor. Similarly, well-off households may have much better networks and access to the labor market. Effects of paid employment, for example, may therefore be overestimated. 18 List of references Behrman, Jere R. and Mark R. Rosenzweig (2002) Does Increasing Women's Schooling Raise the Schooling of the Next Generation?, American Economic Review 92(1), pp. 323-334. Coudouel, Aline, Jesko Hentschel and Quentin Wodon (2002), Poverty Measurement and Analysis, in the PRSP Sourcebook, World Bank, Washington D.C. Datt, G. and D. Joliffe (1999), Determinants of Poverty in Egypt: 1997, Discussion Paper no. 75, International Food Policy Research Institute, Washington D.C. Datt, G., and M. Ravallion (1992), Growth and Redistribution Components of Changes in Poverty Measures: A Decomposition with Applications to Brazil and India in the 1980's, Journal of Development Economics 38, pp. 275­95. Demombynes, G. and J. G. Hoogeveen (2004), Growth, Inequality and Simulated Paths for Tanzania, 1992-2001, mimeo, World Bank, Washington D.C. ESRF (2004), Rural Income Dynamics in Kagera Region, Tanzania Foster, J.E., J. Greer, E. Thorbecke (1984), A Class of Decomposable Poverty Indices, Econometrica 52, pp.761-766. Huppi, M. and M. Ravallion, (1991), Measuring Changes in Poverty: A Methodological Case Study of Indonesia during an Adjustment Period, World Bank Economic Review 5(1), pp. 57-82. 19 Lanjouw, P. and M. Ravallion (1995), Poverty and Household Size, Economic Journal 105, pp. 1415-34. Mkai (2002), Household Budget Survey 2000/2001, National Bureau of Statistics Tanzania Shorrocks, A. F., 1984, Inequality Decomposition by Population Subgroup, Econometrica 52, pp. 1369-85. World Bank (2003a), Poverty and Human Development Report World Bank (2003b), Tanzania Participatory Poverty Assessment 20 Tables for Tanzania Country Economic Memorandum Note: For all tables it holds that · Bold signifies statistical difference across years at the 5% level · Monetary poverty is based on adult equivalent scale consumption and the basic needs poverty line (excluding expenditure to medicine and education) List of Tables Table 1: Share below the poverty line across years p. 4 Table 2: Share below the poverty line by zones and year p. 4 Table 3: Difference in share below the poverty line across years by zones p. 4 Table 4: Share below the poverty line by Dar es Salaam, other urban, and rural p. 4 Table 5: Share below the poverty line by zones, rural-urban, and year p. 6 Table 6: Difference in share below the poverty line across years by zones and rural-urban p. 7 Table 7: Consumption value distribution by Dar es Salaam, other urban, and rural areas p. 8 Table 8: Decomposition of change in poverty p. 8 Table 9: Decomposition of change in poverty due to changes in regional composition p. 9 Table 10: Relative quintiles p. 9 Table 11: Gini coefficients p. 10 Table 12: Theil indices p. 11 Table 13: Decomposition of Theil, share of between-group inequality p. 11 Table 14: Asset holdings by poverty status p. 12 Table 15: Asset holdings by quintiles p. 12 Table 16: Access to means of communication by poverty status p. 13 Table 17: Access to means of communication by quintiles p. 13 Table 18a: Housing quality, Mainland Tanzania and Dar es Salaam p. 14 Table 18b: Housing quality, Other urban and Rural p. 14 Table 19: Type of foundation by quintiles p. 15 Table 20: Type of floor by quintiles p. 16 Table 21: Type of roof by quintiles p. 16 Table 22: Perceptions of problems with satisfying food needs and monetary poverty, 2000/2001 only p. 17 Table 23: Food patterns in Dar, other urban, and rural areas, 2000/2001 only p. 17 Table 24: Incidence and types of sickness, 2000/2001 only p. 17 Table 25: Incidence of sickness in general, malaria, and diarrhea by quintiles, 2000/2001 only p. 18 Table 26a: Type of health care provider in case of sickness, 2000/2001 only p. 19 Table 26b: Type of health care provider in case of sickness, 2000/2001 only p. 19 Table 26c: Type of health care provider in case of sickness, 2000/2001 only p. 20 Table 27: Type of health care provider in case of sickness by quintiles, 2000/2001 only p. 21 Table 28: Reason for not using health care, 2000/2001 only p. 22 21 Table 29: Reasons for not using health care by quintiles, 2000/2001 only p. 22 Table 30a: Distance to services by poverty status, 1991/1992 p. 23 Table 30b: Distance to services by poverty status, 1991/1992 p. 23 Table 31: Access to services by quintiles p. 24 Table 32a: Use of financial services by poverty status, 1991/1992 p. 25 Table 32b: Use of financial services by poverty status, 1991/1992 p. 25 Table 33: Use of financial services by quintiles p. 26 Table 34a: Access to piped water and electricity, 1991/1992 p. 26 Table 34b: Access to piped water and electricity, 2000/2001 p. 26 Table 35: Access to piped water and electricity by quintiles p. 27 Table 36: School enrolment, 7-13 year olds p. 27 Table 37: School enrolment by quintiles, 7-13 year olds p. 28 Table 38: Composition of consumption (incl. medical care, education, water, and transport), 2000/2001 p. 29 Table 39: Composition of consumption by quintiles (incl. medical care, education, water, and transport), 2000/2001 p. 29 Table 40: Composition of income sources by area (value of non-monetary food consumption added back), 2000/2001 p. 30 Table 41: Composition of income sources by quintiles (value of non-monetary food consumption added back), 2000/2001 p. 31 Table 42: Poverty by number of household members p. 32 Table 43: Poverty by number of children aged 5 or younger p. 32 Table 44: Poverty by civil status of head of household p. 33 Table 45: Poverty by level of schooling of head of household p. 33 Table 46: Levels of completed schooling of head of household by quintiles p. 34 Table 47: Poverty by employment of head of household p. 35 Table 48: Employment of head of household by quintiles p. 35 Table 49: Main source of income p. 36 Table 50: Poverty by type of business p. 36 Table 51: Main type of business by Dar es Salaam, other urban, and rural areas p. 37 Table 52: Main type of business by Dar es Salaam, other urban, and rural areas p. 38 Table 53: Number of household member by Dar es Salaam, other urban, and rural areas p. 39 Table 54: Number of children aged 5 or younger by Dar es Salaam, other urban, and rural areas p. 39 Table 55: Average length of education of head of household by Dar es Salaam, other urban, and rural areas p. 39 Table 56: Regression analysis, log per adult equivalent real consumption, 2000/2001 interaction terms (regional dummies included) p. 40 Table 57a: Regression analysis, log per adult equivalent consumption, Dar es Salaam, coefficients in levels (regional dummies included) p. 42 Table 57b: Regression analysis, log per adult equivalent consumption, other urban, coefficients in levels (regional dummies included) p. 43 Table 57c: Regression analysis, log per adult equivalent consumption, rural, coefficients in levels (regional dummies included) p. 44 Table 58: Regression analysis, log per adult equivalent real consumption, 22 2000/2001 interaction terms (regional dummies included) p. 45 Table 59a: Regression analysis, log per adult equivalent consumption, Dar es Salaam, coefficients in levels (regional dummies included) p. 47 Table 59b: Regression analysis, log per adult equivalent consumption, other urban, coefficients in levels (regional dummies included) p. 48 Table 59c: Regression analysis, log per adult equivalent consumption, rural, coefficients in levels (regional dummies included) p. 49 List of figures Figure 1: Share below the poverty line by Dar, other urban, rural, and mainland Tanzania p. 5 Figure 2: Difference in share below the poverty line by Dar, other urban, rural, and mainland Tanzania p. 6 Figure B1: First order stochastic dominance p. 50 23 Table 1. Share below the poverty line (adult equivalent share) across years Poverty 95% Confidence interval Year Low High 1991/1992 38.6% 34.2% 43.0% 2000/2001 35.3% 32.2% 38.4% *Bold signifies that the difference across years is significant at the 5% level Table 2. Share below the poverty line by zones and year 1991/1992 2000/2001 95% Confidence interval Population 95% Confidence interval Population Zone Poverty Low High share Poverty* Low High share* Coastal 40.0% 25.9% 54.0% 12.8% 34.7% 28.1% 41.4% 12.8% Nrn. highlands 20.2% 11.8% 28.7% 10.1% 36.1% 26.0% 46.1% 11.0% Lake 37.0% 29.2% 44.7% 35.1% 39.0% 33.3% 44.7% 37.4% Central 48.8% 34.0% 63.7% 9.4% 42.4% 33.9% 50.9% 8.3% Srn. highlands 46.6% 36.1% 57.1% 15.3% 25.8% 19.8% 31.8% 14.0% South 43.9% 33.1% 54.6% 11.9% 43.2% 32.5% 54.0% 9.1% Dar es Salaam 28.1% 22.6% 33.7% 5.3% 17.6% 12.4% 22.9% 7.4% *Bold signifies that the difference across years is significant at the 5% level Table 3. Difference in share below the poverty line across years by zones Difference Zone in poverty Coastal -5.2% Nrn. highlands 15.8% Lake 2.0% Central -6.4% Srn. highlands -20.8% South -0.6% Dar es Salaam -10.5% *Bold signifies that the difference across years is significant at the 5% level Table 4. Share below the poverty line by Dar es Salaam, other urban, and rural 1991/1992 2000/2001 95% Confidence interval Population 95% Confidence interval Population Poverty Low High share Poverty* Low High share* Dar es Salaam 28.1% 22.6% 33.7% 5.3% 17.6% 12.4% 22.9% 7.4% Other urban 28.7% 18.7% 38.6% 12.6% 25.9% 21.9% 30.0% 13.6% Rural 40.8% 35.7% 46.0% 82.1% 38.6% 34.9% 42.3% 79.0% *Bold signifies that the difference across years is significant at the 5% level 24 Figure 1. Share below the poverty line by Dar, other urban, rural, and mainland Tanzania Poverty rate 50% 40% 1991/1992 30% 2000/2001 20% 10% 0% Dar es Other Rural Total Salaam urban 25 Figure 2. Difference in share below the poverty line by Dar, other urban, rural, and mainland Tanzania Dar es Other Salaam urban Rural Total 0.0000 -0.0200 -0.0400 -0.0600 -0.0800 -0.1000 -0.1200 Table 5. Share below the poverty line by zones, rural-urban, and year 1991/1992 2000/2001 95% Confidence interval Population 95 % Confidence interval Population Zone Poverty Low High share Poverty* Low High share* Urban: Coastal 17.9% 11.3% 24.4% 2.0% 24.9% 17.2% 32.5% 2.1% Nrn. highlands 18.4% 10.3% 26.5% 0.8% 20.8% 12.0% 29.6% 1.6% Lake 15.2% 3.9% 26.5% 3.8% 32.7% 26.3% 39.2% 4.4% Central 34.5% 14.1% 54.8% 0.4% 24.1% 16.7% 31.4% 1.0% Srn. highlands 42.7% 26.9% 58.5% 4.3% 16.9% 8.9% 25.0% 2.8% South 42.6% 29.3% 55.9% 1.2% 31.2% 25.6% 36.7% 1.6% Dar es Salaam 28.1% 22.6% 33.7% 5.3% 17.6% 12.4% 22.9% 7.4% Rural: Coastal 44.1% 28.2% 59.9% 10.8% 36.7% 29.0% 44.5% 10.7% Nrn. highlands 20.4% 11.2% 29.6% 9.3% 38.7% 27.5% 50.0% 9.4% Lake 39.6% 31.4% 47.7% 31.3% 39.8% 33.4% 46.2% 32.9% Central 49.5% 34.0% 65.1% 9.0% 44.8% 35.5% 54.2% 7.3% Srn. highlands 48.2% 34.9% 61.4% 11.0% 28.0% 21.1% 34.9% 11.2% South 44.0% 32.1% 55.9% 10.7% 45.7% 33.0% 58.4% 7.5% *Bold signifies that the difference across years is significant at the 5% level 26 Table 6. Difference in share below the poverty line across years by zones and rural-urban Difference Zone in poverty Urban: Coastal 7.0% Nrn. highlands 2.4% Lake 17.5% Central -10.4% Srn. highlands -25.8% South -11.5% Dar es Salaam -10.5% Rural: Coastal -7.3% Nrn. highlands 18.3% Lake 0.3% Central -4.7% Srn. highlands -20.2% South 1.7% *Bold signifies that the difference across years is significant at the 5% level 27 Table 7. Consumption value distribution by Dar es Salaam, other urban, and rural areas 1991/1992 2000/2001 Population 95% Confidence interval Population 95% Confidence interval Share Low High Share Low High Dar es Salaam 0-20% 12.7% 8.9% 16.5% 8.2% 4.8% 11.6% 20-40% 16.5% 13.2% 19.8% 12.2% 7.8% 16.6% 40-60% 22.8% 18.9% 26.6% 17.3% 13.1% 21.5% 60-80% 24.4% 21.0% 27.9% 19.2% 15.9% 22.6% 80-100% 23.5% 18.9% 28.2% 43.0% 35.5% 50.4% Other urban 0-20% 13.1% 4.8% 21.3% 14.4% 10.9% 17.8% 20-40% 16.6% 12.4% 20.8% 15.6% 13.4% 17.7% 40-60% 19.1% 11.6% 26.6% 14.9% 12.5% 17.3% 60-80% 19.9% 15.2% 24.6% 22.7% 20.9% 24.6% 80-100% 31.3% 22.3% 40.3% 32.4% 27.4% 37.4% Rural 0-20% 21.5% 17.1% 26.0% 22.1% 18.7% 25.4% 20-40% 20.8% 17.7% 23.8% 21.5% 19.5% 23.5% 40-60% 20.0% 17.6% 22.4% 21.1% 18.9% 23.4% 60-80% 19.7% 16.9% 22.6% 19.6% 17.4% 21.8% 80-100% 18.0% 14.6% 21.5% 15.7% 13.5% 18.0% *Bold signifies that the difference across years is significant at the 5% level Table 8. Decomposition of change in poverty Country level Dar es Salaam Other urban Rural Poverty 1991 38.6% 28.1% 28.7% 40.8% Poverty 2001 35.4% 17.6% 26.0% 38.7% Variation 2001-1991 -3.2% -10.5% -2.7% -2.1% Growth impact -8.4% -18.4% -6.6% -5.3% Inequality impact 5.5% 12.4% 4.0% 2.7% Residual -0.2% -4.5% -0.2% 0.6% 28 Table 9. Decomposition of change in poverty due to changes in regional composition Poverty 1991 0.3860 Poverty 2001 0.3542 Population share Absolute Percentage in period 1 change change Dar es Salaam 5.4% -0.6% 17.7% Other urban 12.6% -0.3% 10.8% Rural 82.1% -1.7% 53.8% Total Intra-sectoral effect -2.6% 82.3% Population-shift effect -0.4% 12.0% Interaction effect -0.2% 5.7% Change in poverty -3.2% 100.0% Table 10. Relative quintiles 1991/1992 2000/2001 p90/p10 4.4770 4.6660 p90/p50 2.1740 2.2470 p10/p50 0.4860 0.4820 p75/p50 2.1420 2.1430 p75/p25 1.4940 1.4950 p25/p50 0.6980 0.6970 29 Table 11. Gini coefficients 1991/1992 2000/2001 95% Confidence interval** 95% Confidence interval** Gini Low High Gini* Low High Mainland 0.3304 0.3017 0.3479 0.3373 0.3233 0.3483 Dar es Salaam 0.2974 0.2796 0.3166 0.3441 0.3162 0.3751 Other urban 0.3403 0.3084 0.3841 0.3511 0.3381 0.3702 Rural 0.3270 0.3069 0.3481 0.3190 0.3004 0.3501 Coastal 0.3447 0.3099 0.3888 0.3040 0.2868 0.3322 Nrn. highlands 0.2649 0.2383 0.2971 0.3160 0.2905 0.3482 Lake 0.3344 0.2865 0.3788 0.3421 0.3177 0.3693 Central 0.3013 0.1982 0.3093 0.3433 0.2771 0.4538 Srn. highlands 0.3727 0.3296 0.4270 0.3163 0.2948 0.3470 South 0.2936 0.2655 0.3502 0.3337 0.3022 0.3653 No primary educ. 0.3029 0.2779 0.3486 0.3053 0.2769 0.3222 Some primary educ. 0.3357 0.3153 0.3713 0.3180 0.3049 0.3484 Compl. primary educ. 0.3305 0.2728 0.3665 0.3163 0.2737 0.3605 Some secondary educ. 0.2837 0.2353 0.3279 0.3068 0.2701 0.3418 Compl secondary educ. 0.2910 0.2499 0.3283 0.3402 0.3097 0.3775 Post secondary educ. 0.3148 0.2696 0.4064 0.3561 0.2910 0.4084 Adult eduation only 0.2938 0.2546 0.3597 0.3009 0.2475 0.3546 Household members: 1 0.3452 0.3045 0.3932 0.3190 0.3009 0.3513 2 0.2903 0.2551 0.3279 0.3135 0.2889 0.3490 3 0.2824 0.2458 0.3206 0.2892 0.2617 0.3072 4 0.3134 0.2794 0.3586 0.3025 0.2777 0.3251 5 0.2905 0.2620 0.3234 0.3051 0.2827 0.3287 6 0.3002 0.2587 0.3467 0.3183 0.2927 0.3376 7 0.2936 0.2625 0.3369 0.3052 0.2841 0.3641 8 0.2666 0.2375 0.2980 0.3117 0.2781 0.3911 9 0.2835 0.2553 0.3381 0.3033 0.2746 0.3340 10 or more 0.3087 0.2731 0.3519 0.3195 0.2728 0.4234 *Bold signifies that the difference across years is significant at the 5% level **Bias corrected confidence interval, 50 repetitions 30 Table 12. Theil indices 1991/1992 2000/2001 95% Confidence interval** 95% Confidence interval** Theil Low High Theil* Low High Mainland 0.185 0.171 0.220 0.199 0.180 0.222 Dar es Salaam 0.152 0.133 0.183 0.208 0.172 0.277 Other urban 0.201 0.164 0.266 0.214 0.191 0.239 Rural 0.184 0.161 0.201 0.177 0.140 0.207 Coastal 0.204 0.162 0.252 0.165 0.136 0.210 Nrn. highlands 0.120 0.097 0.154 0.170 0.124 0.235 Lake 0.191 0.161 0.231 0.207 0.187 0.269 Central 0.161 0.119 0.231 0.185 0.117 0.285 Srn. highlands 0.244 0.190 0.290 0.169 0.130 0.201 South 0.157 0.105 0.223 0.197 0.178 0.281 No primary educ. 0.154 0.135 0.190 0.158 0.135 0.189 Some primary educ. 0.195 0.164 0.232 0.169 0.157 0.214 Compl. primary educ. 0.183 0.151 0.228 0.208 0.141 0.493 Some secondary educ. 0.135 0.091 0.174 0.155 0.122 0.197 Compl secondary educ. 0.146 0.111 0.225 0.197 0.160 0.245 Post secondary educ. 0.196 0.122 0.298 0.229 0.175 0.312 Adult eduation only 0.151 0.113 0.190 0.148 0.120 0.183 Household members: 1 0.197 0.158 0.264 0.177 0.149 0.222 2 0.142 0.105 0.199 0.178 0.145 0.236 3 0.131 0.114 0.156 0.148 0.122 0.184 4 0.164 0.125 0.222 0.154 0.141 0.176 5 0.139 0.118 0.167 0.167 0.143 0.195 6 0.156 0.126 0.215 0.169 0.146 0.203 7 0.145 0.114 0.176 0.163 0.128 0.213 8 0.120 0.092 0.146 0.170 0.130 0.234 9 0.140 0.085 0.222 0.151 0.125 0.186 10 or more 0.169 0.129 0.235 0.202 0.134 0.413 *Bold signifies that the difference across years is significant at the 5% level **Bias corrected confidence interval, 50 repetitions Table 13. Decomposition of Theil, share of between-group inequality Groups 1991/1992 2000/2001 Dar - other urban - rural 1.95% 6.52% Geographical zones 2.79% 5.19% Education levels 4.70% 12.02% Household sizes 18.24% 15.77% 31 Table 14. Asset holdings by poverty status 1991/1992 2000/2001 95% Confidence interval Population 95% Confidence interval Population Share Low High share Share* Low High share* Fields 88.6% 84.6% 92.5% 82.1% 85.7% 83.7% 87.7% 78.1% Fish nets 0.0% 0.0% 0.0% 3.0% 2.4% 0.4% 4.4% 2.4% Livestock 45.1% 39.2% 51.0% 46.3% 50.4% 46.5% 54.4% 41.4% *Bold signifies that the difference across years is significant at the 5% level Table 15. Asset holdings by quintiles 1991/1992 2000/2001 Share 95% Confidence interval Share* 95% Confidence interval Low High Low High Fields 0-20% 88.4% 84.9% 92.0% 85.8% 83.4% 88.1% 20-40% 83.2% 78.0% 88.5% 84.5% 81.7% 87.2% 40-60% 84.9% 82.0% 87.9% 79.5% 75.5% 83.4% 60-80% 80.9% 75.9% 86.0% 71.0% 66.0% 76.0% 80-100% 73.4% 66.1% 80.8% 58.9% 51.4% 66.5% Fishnets 0-20% 0.0% 0.0% 0.0% 2.8% 1.5% 4.0% 20-40% 0.1% -0.1% 0.2% 2.0% 1.1% 2.9% 40-60% 0.0% 0.0% 0.0% 2.2% 0.9% 3.4% 60-80% 0.0% 0.0% 0.0% 2.0% 1.0% 3.0% 80-100% 0.0% 0.0% 0.0% 2.7% 0.0% 5.4% Livestock 0-20% 44.9% 38.4% 51.4% 49.9% 45.5% 54.2% 20-40% 46.9% 39.4% 54.4% 46.7% 41.7% 51.7% 40-60% 50.9% 43.5% 58.3% 41.1% 36.0% 46.1% 60-80% 43.3% 35.0% 51.5% 34.0% 28.8% 39.3% 80-100% 39.9% 31.2% 48.5% 33.3% 26.4% 40.3% *Bold signifies that the difference across years is significant at the 5% level 32 Table 16. Access to means of communication by poverty status 1991/1992 2000/2001 Poor Non-poor Poor Non-poor 95% Confidence interval 95% Confidence interval 95% Confidence interval 95% Confidence interval Share Low High Share Low High Share Low High Share Low High Radio 32.0% 26.7% 37.3% 45.1% 40.2% 49.9% 44.3% 40.5% 48.1% 63.3% 60.5% 66.1% Telephone 0.2% 0.0% 0.5% 0.9% 0.5% 1.2% 0.2% 0.1% 0.3% 2.4% 1.7% 3.0% TV 0.0% 0.0% 0.0% 0.3% 0.1% 0.4% 0.4% 0.2% 0.6% 4.7% 3.8% 5.7% *Bold signifies that the difference across years is significant at the 5% level Table 17. Access to means of communication by quintiles 1991/1992 2000/2001 Share 95% Confidence interval Share* 95% Confidence interval Low High Low High Radio 0-20% 31.5% 26.3% 36.7% 45.6% 42.1% 49.2% 20-40% 38.6% 32.0% 45.3% 56.6% 52.2% 60.9% 40-60% 44.6% 37.6% 51.6% 62.7% 57.9% 67.5% 60-80% 54.8% 47.0% 62.6% 71.6% 67.8% 75.4% 80-100% 56.1% 48.6% 63.7% 80.9% 76.8% 85.0% Telephone 0-20% 0.2% 0.0% 0.5% 0.2% 0.1% 0.3% 20-40% 0.4% 0.1% 0.7% 1.0% 0.4% 1.5% 40-60% 0.7% 0.2% 1.1% 1.1% 0.7% 1.5% 60-80% 1.1% 0.1% 2.1% 3.5% 2.0% 5.0% 80-100% 2.7% 1.1% 4.2% 10.0% 6.5% 13.5% TV 0-20% 0.0% 0.0% 0.0% 0.4% 0.3% 0.6% 20-40% 0.1% 0.0% 0.2% 2.2% 1.5% 3.0% 40-60% 0.3% 0.1% 0.5% 2.9% 2.1% 3.7% 60-80% 0.5% -0.3% 1.3% 7.4% 5.2% 9.6% 80-100% 0.4% 0.0% 0.8% 16.3% 11.4% 21.3% *Bold signifies that the difference across years is significant at the 5% level 33 Table 18a. Housing quality, Mainland Tanzania and Dar es Salaam Mainland Tanzania Dar es Salaam 1991/1992 2000/2001 1991/1992 2000/2001 95% Confidence interval 95% Confidence interval 95% Confidence interval 95% Confidence interval Share Low High Share Low High Share Low High Share Low High Foundation: Concrete 18.3% 13.2% 23.4% 25.2% 22.1% 28.4% 64.3% 56.7% 72.0% 82.2% 76.4% 88.0% Stones or other 15.4% 12.3% 18.6% 18.2% 15.7% 20.8% 19.6% 14.9% 24.2% 9.3% 5.1% 13.4% No foundation 65.7% 59.9% 71.6% 56.2% 52.1% 60.3% 16.1% 10.0% 22.2% 8.2% 4.5% 11.8% Floor: Earth 83.7% 81.5% 85.9% 77.3% 74.8% 79.7% 14.6% 10.3% 18.8% 7.3% 3.5% 11.1% Concrete 14.7% 12.6% 16.7% 21.5% 19.1% 23.8% 84.0% 79.8% 88.2% 90.7% 86.2% 95.1% Roof: Durable 28.8% 24.7% 32.8% 38.2% 34.8% 41.5% 98.2% 96.7% 99.7% 97.5% 94.8% 100.1% Non-durable 70.3% 66.1% 74.5% 61.2% 57.8% 64.6% 1.6% 0.2% 3.1% 2.2% -0.4% 4.8% Table 18b. Housing quality, Other urban and Rural Other urban Rural 1991/1992 2000/2001 1991/1992 2000/2001 95% Confidence interval 95% Confidence interval 95% Confidence interval 95% Confidence interval Share Low High Share Low High Share Low High Share Low High Foundation: Concrete 34.1% 22.5% 45.8% 41.2% 34.1% 48.3% 13.8% 8.0% 19.6% 18.3% 14.9% 21.7% Stones or other 33.2% 23.7% 42.8% 30.1% 25.7% 34.5% 12.8% 9.4% 16.3% 17.1% 14.1% 20.2% No foundation 32.6% 22.5% 42.6% 28.4% 22.4% 34.3% 72.8% 66.1% 79.4% 64.3% 59.6% 69.0% Floor: Earth 48.6% 36.6% 60.6% 42.8% 39.1% 46.4% 92.0% 90.1% 93.9% 88.1% 85.9% 90.2% Concrete 49.9% 37.8% 61.9% 56.3% 52.7% 59.9% 6.3% 4.7% 7.9% 10.7% 8.6% 12.7% Roof: Durable 69.7% 55.7% 83.8% 79.1% 74.1% 84.2% 19.7% 15.3% 24.0% 27.2% 23.7% 30.6% Non-durable 29.1% 15.3% 42.9% 20.2% 15.1% 25.2% 79.4% 74.9% 83.9% 72.1% 68.6% 75.6% 34 Table 19. Type of foundation by quintiles 1991/1992 2000/2001 Share 95% Confidence interval Share* 95% Confidence interval Low High Low High Stones or other 0-20% 14.5% 10.1% 18.9% 14.8% 11.9% 17.7% 20-40% 11.8% 8.0% 15.6% 23.1% 18.5% 27.6% 40-60% 15.4% 11.3% 19.5% 18.2% 14.8% 21.5% 60-80% 19.9% 14.7% 25.2% 19.7% 15.6% 23.8% 80-100% 18.9% 12.0% 25.8% 17.3% 13.2% 21.4% Concrete 0-20% 13.7% 8.0% 19.4% 17.2% 13.2% 21.2% 20-40% 17.3% 11.8% 22.8% 18.9% 15.2% 22.6% 40-60% 18.8% 12.7% 24.9% 28.8% 23.9% 33.6% 60-80% 19.1% 12.5% 25.8% 36.3% 31.1% 41.5% 80-100% 30.9% 18.4% 43.4% 46.5% 38.7% 54.3% No foundation 0-20% 71.0% 64.0% 78.0% 67.6% 62.5% 72.8% 20-40% 70.6% 63.6% 77.6% 57.6% 52.2% 63.1% 40-60% 65.1% 58.1% 72.2% 52.9% 47.4% 58.5% 60-80% 60.5% 52.3% 68.7% 43.7% 38.2% 49.3% 80-100% 49.9% 37.0% 62.9% 35.9% 27.2% 44.7% *Bold signifies that the difference across years is significant at the 5% level 35 Table 20. Type of floor by quintiles 1991/1992 2000/2001 Share 95% Confidence interval Share* 95% Confidence interval Low High Low High Earth Floor 0-20% 90.1% 87.5% 92.6% 89.0% 87.0% 90.9% 20-40% 86.6% 82.8% 90.3% 83.0% 79.8% 86.3% 40-60% 81.0% 77.4% 84.6% 74.7% 70.4% 79.0% 60-80% 79.8% 75.8% 83.7% 62.6% 57.9% 67.3% 80-100% 69.6% 61.5% 77.7% 47.7% 41.6% 53.9% Concrete 0-20% 7.7% 5.8% 9.6% 9.9% 8.0% 11.8% 20-40% 11.8% 8.2% 15.3% 16.0% 12.8% 19.2% 40-60% 17.3% 13.9% 20.7% 23.3% 19.6% 27.0% 60-80% 19.4% 15.6% 23.2% 36.3% 31.6% 40.9% 80-100% 29.3% 21.5% 37.1% 50.9% 45.1% 56.8% *Bold signifies that the difference across years is significant at the 5% level Table 21. Type of roof by quintiles 1991/1992 2000/2001 Share 95% Confidence interval Share* 95% Confidence interval Low High Low High Non-durable 0-20% 80.0% 75.9% 84.1% 75.3% 71.5% 79.1% 20-40% 75.9% 70.4% 81.5% 67.7% 63.2% 72.1% 40-60% 65.9% 59.2% 72.5% 56.3% 51.2% 61.4% 60-80% 59.3% 52.9% 65.7% 42.4% 36.6% 48.1% 80-100% 53.3% 43.1% 63.6% 31.4% 25.6% 37.1% Durable 0-20% 18.8% 15.0% 22.6% 23.9% 20.2% 27.7% 20-40% 23.6% 18.0% 29.1% 31.7% 27.4% 36.1% 40-60% 32.7% 26.3% 39.1% 43.4% 38.4% 48.5% 60-80% 39.6% 33.3% 45.8% 56.6% 50.8% 62.3% 80-100% 46.3% 36.1% 56.6% 67.8% 62.1% 73.6% *Bold signifies that the difference across years is significant at the 5% level 36 Table 22. Perceptions of problems with satisfying food needs and monetary poverty, 2000/2001 only 95% Confidence interval Problems with satisfying food needs Poverty Low High Never 26.3% 12.5% 40.2% Seldom 36.2% 23.0% 49.4% Sometime 39.1% 23.8% 54.4% Often 47.2% 33.5% 60.9% Always 46.7% 34.5% 58.9% Table 23. Food patterns in Dar, other urban, and rural areas, 2000/2001 only 95% Confidence interval 95% Confidence interval 95% Confidence interval 95% Confidence interval Meat Low High Fish Low High Eggs Low High Dairy Low High Dar es Salaam 2.1 1.8 2.4 2.1 1.9 2.3 0.7 0.6 0.9 1.6 1.3 2.0 Other urban 1.8 1.7 1.9 2.3 2.1 2.4 0.4 0.3 0.5 1.7 1.5 2.0 Rural 1.2 1.1 1.3 1.9 1.7 2.1 0.3 0.2 0.3 1.9 1.7 2.2 Table 24. Incidence and types of sickness, 2000/2001 only Sick during last 4 weeks Sick with malaria during last 4 weeks Sick with diarrhea during last 4 weeks Share of 95% Confidence interval Share of 95% Confidence interval Share of 95% Confidence interval individuals Low High sick Low High sick Low High Tanzania 26.5% 25.0% 28.0% 63.8% 61.7% 65.8% 11.6% 10.2% 13.0% Dar es Salaam 19.4% 15.9% 22.8% 65.4% 59.5% 71.4% 6.5% 4.5% 8.6% Other urban 23.4% 22.1% 24.7% 67.2% 64.9% 69.5% 9.1% 7.9% 10.4% Rural 27.7% 25.9% 29.5% 63.2% 60.8% 65.6% 12.3% 10.6% 14.0% Non-poor 27.2% 25.5% 28.8% 65.6% 63.4% 67.8% 11.7% 11.7% 11.7% Poor 25.3% 23.1% 27.6% 60.2% 56.6% 63.8% 11.4% 9.1% 13.6% 37 Table 25. Incidence of sickness in general, malaria, and diarrhea by quintiles, 2000/2001 only 2000/2001 Share 95% Confidence interval Low High Sick (1) 0-20% 25.7% 23.5% 27.9% 20-40% 27.0% 24.8% 29.1% 40-60% 26.4% 23.6% 29.2% 60-80% 28.7% 25.8% 31.5% 80-100% 25.6% 21.9% 29.4% Sick with malaria (2) 0-20% 61.0% 57.6% 64.4% 20-40% 63.5% 60.5% 66.5% 40-60% 64.9% 61.2% 68.6% 60-80% 68.4% 63.1% 73.6% 80-100% 68.6% 61.9% 75.3% Sick with diarrhea (2) 0-20% 11.2% 9.2% 13.2% 20-40% 11.5% 9.3% 13.8% 40-60% 12.5% 9.7% 15.3% 60-80% 12.1% 9.4% 14.7% 80-100% 10.3% 6.8% 13.7% (1): Share of individuals (2): Share of sick 38 Table 26a. Type of health care provider in case of sickness, 2000/2001 only Private hospital Public hospital Community or dispensary or dispensary health center Share of 95% Confidence interval Share of 95% Confidence interval Share of 95% Confidence interval individuals Low High individuals Low High individuals Low High Tanzania 21.9% 19.6% 24.2% 41.9% 38.1% 45.7% 11.0% 8.1% 13.9% Dar es Salaam 46.5% 39.7% 53.2% 41.2% 30.9% 51.5% 6.2% 0.4% 11.9% Other urban 31.1% 27.7% 34.5% 38.0% 34.1% 41.9% 9.3% 6.8% 11.8% Rural 18.4% 15.9% 21.0% 42.5% 37.9% 47.1% 11.7% 8.2% 15.2% Non-poor 22.2% 19.5% 24.8% 42.6% 38.2% 47.0% 11.4% 7.9% 14.8% Poor 21.3% 17.9% 24.6% 40.3% 35.0% 45.6% 10.3% 6.9% 13.8% Table 26b. Type of health care provider in case of sickness, 2000/2001 only Private doctor Traditional Regional or dentist healer hospital Share of 95% Confidence interval Share of 95% Confidence interval Share of 95% Confidence interval individuals Low High individuals Low High individuals Low High Tanzania 6.7% 5.1% 8.4% 14.1% 11.8% 16.3% 3.3% 2.5% 4.0% Dar es Salaam 2.0% 0.9% 3.1% 2.4% 0.8% 4.1% 3.2% 1.2% 5.1% Other urban 5.1% 3.9% 6.2% 5.4% 3.9% 6.9% 12.0% 8.6% 15.5% Rural 7.4% 5.4% 9.4% 16.4% 13.6% 19.2% 1.8% 1.2% 2.5% Non-poor 6.7% 4.8% 8.6% 12.4% 10.1% 14.7% 3.6% 2.7% 4.6% Poor 6.9% 4.5% 9.4% 17.6% 14.0% 21.2% 2.4% 1.7% 3.2% 39 Table 26c. Type of health care provider in case of sickness, 2000/2001 only Missionary hospital Share of 95% Confidence interval individuals Low High Tanzania 9.3% 7.3% 11.3% Dar es Salaam 1.1% 0.2% 1.9% Other urban 6.6% 5.1% 8.2% Rural 10.4% 7.9% 12.8% Non-poor 9.8% 7.5% 12.2% Poor 8.1% 5.8% 10.4% 40 Table 27. Type of health care provider in case of sickness by quintiles, 2000/2001 only 2000/2001 Share 95% Confidence interval Low High Private Hospital 0-20% 20.7% 17.5% 23.9% 20-40% 18.4% 14.7% 22.0% 40-60% 22.2% 17.8% 26.6% 60-80% 24.8% 20.2% 29.4% 80-100% 33.6% 26.3% 40.9% Public Hospital 0-20% 39.8% 34.7% 44.9% 20-40% 46.1% 39.8% 52.4% 40-60% 42.7% 35.9% 49.5% 60-80% 38.5% 31.3% 45.7% 80-100% 41.2% 29.5% 52.9% Community Health Center 0-20% 10.8% 7.3% 14.3% 20-40% 10.6% 6.5% 14.6% 40-60% 13.5% 6.5% 20.5% 60-80% 11.6% 6.5% 16.7% 80-100% 7.4% 3.7% 11.1% Private Doctor 0-20% 6.9% 4.7% 9.2% 20-40% 4.3% 2.7% 5.9% 40-60% 5.8% 3.1% 8.5% 60-80% 11.9% 5.4% 18.3% 80-100% 6.2% 3.2% 9.3% Traditional Healer 0-20% 16.7% 13.3% 20.1% 20-40% 15.8% 11.8% 19.9% 40-60% 12.8% 9.0% 16.6% 60-80% 9.0% 5.3% 12.7% 80-100% 7.9% 3.8% 12.0% Regional Hospital 0-20% 2.4% 1.7% 3.2% 20-40% 3.1% 1.6% 4.6% 40-60% 2.8% 1.3% 4.3% 60-80% 5.0% 2.8% 7.1% 80-100% 5.7% 3.5% 7.8% Missionary Hospital 0-20% 8.1% 5.9% 10.3% 20-40% 10.7% 7.0% 14.3% 40-60% 9.0% 5.5% 12.4% 60-80% 9.2% 6.0% 12.4% 80-100% 11.6% 3.9% 19.4% 41 Table 28. Reason for not using health care, 2000/2001 only No Too Too far need expensive away Share of 95% Confidence interval Share of 95% Confidence interval Share of 95% Confidence interval individuals Low High individuals Low High individuals Low High Tanzania 91.2% 90.2% 92.3% 7.1% 6.1% 8.1% 1.7% 1.3% 2.2% Dar es Salaam 93.9% 90.9% 96.9% 6.7% 3.8% 9.7% 0.3% -0.1% 0.7% Other urban 93.1% 92.0% 94.2% 7.8% 6.4% 9.2% 0.4% 0.2% 0.7% Rural 90.7% 89.3% 92.0% 7.0% 5.7% 8.2% 2.1% 1.6% 2.7% Non-poor 91.7% 90.5% 92.9% 6.2% 5.1% 7.3% 1.7% 1.2% 2.2% Poor 90.4% 88.7% 92.1% 8.6% 6.8% 10.3% 1.8% 1.2% 2.4% Table 29. Reasons for not using health care by quintiles, 2000/2001 only 2000/2001 Share 95% Confidence interval Low High No Need 0-20% 90.4% 88.8% 92.0% 20-40% 91.1% 89.1% 93.1% 40-60% 92.3% 90.6% 93.9% 60-80% 92.7% 90.9% 94.4% 80-100% 92.1% 89.4% 94.9% Too Expensive 0-20% 8.4% 6.8% 10.1% 20-40% 6.3% 4.4% 8.2% 40-60% 6.2% 4.7% 7.7% 60-80% 6.4% 4.8% 8.0% 80-100% 4.5% 3.0% 6.1% Too Far Away 0-20% 1.8% 1.2% 2.4% 20-40% 2.0% 1.2% 2.9% 40-60% 1.8% 1.1% 2.6% 60-80% 1.5% 0.8% 2.2% 80-100% 0.5% 0.2% 0.8% 42 Table 30a. Distance to services by poverty status, 1991/1992 1991/1992 Poor Non-poor 95% Confidence interval 95% Confidence interval Km Low High Km Low High Distance to nearest hospital 24.7 20.2 29.3 17.7 14.7 20.6 Distance to nearest primary school 1.5 1.1 1.8 1.4 1.2 1.7 Distance to nearest market 4.9 3.5 6.2 4.2 3.2 5.1 Distance to public transport 5.6 3.9 7.2 4.7 3.4 5.9 Table 30b. Distance to services by poverty status, 2000/2001 2000/2001 Poor Non-poor 95% Confidence interval 95% Confidence interval Km Low High Km Low High Distance to nearest hospital 25.6 21.5 29.7 19.0 15.9 22.0 Distance to nearest primary school 2.0 1.7 2.3 1.9 1.5 2.2 Distance to nearest market 3.3 2.7 4.0 2.9 2.4 3.4 Distance to public transport 5.7 4.8 6.7 4.0 3.4 4.6 *Bold signifies that the difference across years is significant at the 5% level 43 Table 31. Access to services by quintiles 1991/1992 2000/2001 Km 95% Confidence interval Km* 95% Confidence interval Low High Low High Hospital 0-20% 24.7 20.2 29.3 25.2 21.2 29.2 20-40% 20.1 16.2 24.0 22.8 17.6 28.1 40-60% 16.4 13.3 19.5 18.0 14.4 21.5 60-80% 15.1 12.0 18.2 15.4 12.4 18.3 80-100% 16.0 11.9 20.1 12.1 8.9 15.2 Market 0-20% 4.8 3.5 6.0 3.5 2.8 4.1 20-40% 4.5 3.4 5.6 3.5 2.8 4.1 40-60% 3.9 3.0 4.9 2.6 2.0 3.3 60-80% 4.5 2.8 6.1 2.0 1.5 2.6 80-100% 3.3 2.2 4.4 1.7 1.0 2.4 Primary School 0-20% 1.4 1.1 1.7 2.0 1.7 2.3 20-40% 1.5 1.2 1.8 2.4 1.6 3.3 40-60% 1.4 1.1 1.7 1.5 1.3 1.8 60-80% 1.4 1.0 1.7 1.5 1.2 1.8 80-100% 1.6 1.1 2.0 1.4 0.9 2.0 Public Transport 0-20% 5.5 3.9 7.1 5.8 4.8 6.7 20-40% 5.5 3.6 7.4 4.8 3.9 5.6 40-60% 3.6 2.7 4.5 3.6 3.0 4.2 60-80% 4.8 3.2 6.5 2.9 2.3 3.6 80-100% 4.4 2.6 6.3 3.0 1.4 4.7 *Bold signifies that the difference across years is significant at the 5% level 44 Table 32a. Use of financial services by poverty status, 1991/1992 1991/1992 Poor Non-poor 95% Confidence interval 95% Confidence interval Share Low High Share Low High Savings 13.3% 9.4% 17.1% 23.7% 20.1% 27.3% Loans 0.6% 0.1% 1.1% 1.6% 1.0% 2.3% Informal savings 5.1% 2.4% 7.8% 5.9% 3.6% 8.3% Table 32b. Use of financial services by poverty status, 2000/2001 2000/2001 Poor Non-poor 95% Confidence interval 95% Confidence interval Share Low High Share Low High Savings 2.7% 1.7% 3.7% 10.0% 8.6% 11.4% Loans 0.4% 0.0% 0.8% 0.9% 0.6% 1.3% Informal savings 2.4% 1.2% 3.6% 5.3% 3.4% 7.1% 45 Table 33. Use of financial services by quintiles 1991/1992 2000/2001 Share 95% Confidence interval Share* 95% Confidence interval Low High Low High Formal Savings 0-20% 13.3% 9.4% 17.1% 2.9% 2.0% 3.8% 20-40% 18.4% 14.0% 22.8% 5.9% 4.3% 7.6% 40-60% 25.1% 19.6% 30.6% 8.1% 6.2% 9.9% 60-80% 27.4% 21.7% 33.2% 14.4% 11.3% 17.5% 80-100% 33.2% 26.9% 39.5% 25.1% 18.7% 31.4% Bank Loan 0-20% 0.6% 0.1% 1.1% 0.5% 0.1% 0.8% 20-40% 1.7% 0.5% 3.0% 0.5% 0.2% 0.9% 40-60% 1.7% 0.7% 2.8% 0.8% 0.2% 1.4% 60-80% 1.5% 0.6% 2.5% 1.9% 0.3% 3.5% 80-100% 1.6% 0.5% 2.6% 1.1% 0.5% 1.7% Informal Savings 0-20% 4.9% 2.3% 7.6% 2.6% 1.4% 3.7% 20-40% 4.7% 2.3% 7.2% 3.7% 2.2% 5.2% 40-60% 6.6% 3.7% 9.5% 6.4% 3.0% 9.9% 60-80% 6.3% 1.9% 10.7% 6.3% 3.7% 8.9% 80-100% 8.4% 1.2% 15.6% 7.2% 3.8% 10.6% *Bold signifies that the difference across years is significant at the 5% level Table 34a. Access to piped water and electricity, 1991/1992 1991/1992 Poor Non-poor 95% Confidence interval 95% Confidence interval Share Low High Share Low High Piped water 33.2% 24.1% 42.3% 32.2% 25.8% 38.7% Electricity 4.9% 3.1% 6.6% 9.6% 7.7% 11.4% Table 34b. Access to piped water and electricity, 2000/2001 2000/2001 Poor Non-poor 95% Confidence interval 95% Confidence interval Share Low High Share Low High Piped water 30.7% 25.1% 36.2% 43.9% 39.3% 48.5% Electricity 4.4% 3.2% 5.7% 13.3% 11.5% 15.0% 46 Table 35. Access to piped water and electricity by quintiles 1991/1992 2000/2001 Share 95% Confidence interval Share* 95% Confidence interval Low High Low High Piped Water 0-20% 33.0% 24.0% 42.0% 31.2% 25.9% 36.5% 20-40% 28.3% 21.0% 35.6% 36.1% 30.7% 41.4% 40-60% 31.2% 23.8% 38.6% 44.2% 38.3% 50.2% 60-80% 39.6% 30.5% 48.6% 52.6% 46.6% 58.7% 80-100% 37.2% 25.4% 48.9% 61.4% 53.0% 69.9% Electricity 0-20% 4.7% 3.0% 6.4% 4.5% 3.3% 5.6% 20-40% 7.2% 4.6% 9.8% 7.9% 6.3% 9.5% 40-60% 9.1% 6.3% 11.9% 11.4% 8.9% 13.8% 60-80% 11.7% 7.9% 15.5% 20.0% 16.2% 23.8% 80-100% 16.2% 9.8% 22.5% 31.4% 24.9% 38.0% *Bold signifies that the difference across years is significant at the 5% level Table 36. School enrolment, 7-13 year olds 1991/1992 2000/2001 Share 95% Confidence interval Share 95% Confidence interval Low High Low High Enrolment rate Tanzania 51.5% 44.9% 58.1% 61.5% 58.7% 64.2% Dar es Salaam 87.1% 80.2% 94.0% 76.0% 71.2% 80.7% Other urban 62.6% 51.0% 74.2% 76.2% 72.5% 79.9% Rural 48.2% 40.8% 55.7% 58.0% 54.8% 61.3% Enrolment, boys Tanzania 49.7% 42.4% 57.0% 60.7% 57.5% 63.9% Dar es Salaam 87.6% 80.7% 94.6% 75.0% 66.8% 83.2% Other urban 64.3% 49.5% 79.1% 76.2% 71.4% 81.0% Rural 45.4% 37.1% 53.8% 57.4% 53.7% 61.1% Enrolment, girls Tanzania 53.5% 46.2% 60.9% 62.4% 59.1% 65.8% Dar es Salaam 86.5% 78.9% 94.1% 76.8% 69.9% 83.7% Other urban 60.3% 47.2% 73.4% 76.2% 72.7% 79.6% Rural 51.2% 42.8% 59.5% 58.9% 54.9% 63.0% *Bold signifies that the difference across years is significant at the 5% level 47 Table 37. School enrolment by quintiles, 7-13 year olds 1991/1992 2000/2001 Share 95% Confidence interval Share* 95% Confidence interval Low High Low High Enrolment 0-20% 58.6% 50.0% 67.2% 55.3% 51.8% 58.8% 20-40% 45.3% 36.3% 54.4% 60.4% 55.3% 65.4% 40-60% 41.4% 32.3% 50.5% 69.5% 64.3% 74.8% 60-80% 51.2% 39.7% 62.8% 73.6% 68.2% 79.0% 80-100% 63.6% 44.9% 82.3% 78.6% 69.9% 87.3% Enrolment, boys 0-20% 57.0% 46.8% 67.2% 54.4% 50.0% 58.8% 20-40% 42.6% 31.9% 53.3% 59.6% 54.2% 65.1% 40-60% 40.7% 30.9% 50.4% 68.6% 61.5% 75.6% 60-80% 54.5% 40.4% 68.7% 70.8% 63.8% 77.9% 80-100% 52.4% 29.2% 75.7% 86.1% 76.2% 95.9% Enrolment, girls 0-20% 60.2% 50.9% 69.6% 56.3% 51.6% 61.0% 20-40% 48.3% 37.5% 59.2% 61.1% 54.6% 67.6% 40-60% 42.5% 30.7% 54.2% 70.6% 63.8% 77.4% 60-80% 47.4% 33.3% 61.5% 76.3% 69.9% 82.8% 80-100% 71.6% 52.0% 91.1% 75.9% 65.3% 86.5% *Bold signifies that the difference across years is significant at the 5% level 48 Table 38. Composition of consumption (incl. medical care, education, water, and transport), 2000/2001 2000/2001 Share 95% Confidence interval Low High Food 72.4% 71.6% 73.2% Medical 3.2% 3.0% 3.5% Education 4.1% 3.8% 4.5% Water 5.0% 3.3% 6.7% Table 39. Composition of consumption by quintiles (incl. medical care, education, water, and transport), 2000/2001 2000/2001 Share 95% Confidence interval Low High Food: 0-20% 74.3% 73.3% 75.4% 20-40% 73.8% 72.8% 74.8% 40-60% 72.0% 70.6% 73.3% 60-80% 69.4% 68.0% 70.7% 80-100% 62.8% 60.0% 65.6% Medical: 0-20% 3.6% 3.2% 3.9% 20-40% 2.9% 2.6% 3.2% 40-60% 3.1% 2.6% 3.5% 60-80% 3.2% 2.9% 3.5% 80-100% 3.3% 2.7% 3.9% Education: 0-20% 4.4% 3.8% 4.9% 20-40% 3.7% 3.2% 4.1% 40-60% 3.8% 3.1% 4.5% 60-80% 4.0% 3.5% 4.6% 80-100% 5.3% 3.6% 6.9% Water: 0-20% 5.5% 2.0% 8.9% 20-40% 5.8% 3.4% 8.3% 40-60% 4.8% 3.8% 5.7% 60-80% 4.1% 3.0% 5.1% 80-100% 3.2% 2.1% 4.3% 49 Table 40. Composition of income sources by area (value of non-monetary food consumption added back), 2000/2001 2000/2001 Share of 95% Confidence interval total income Low High Mainland Tanzania: Employment 10.5% 8.8% 11.5% Self employment 17.3% 15.6% 18.9% Agriculture 37.5% 35.7% 39.3% Transfers (incl. remittances) 5.2% 4.6% 5.7% Remittances 5.0% 4.5% 5.5% Other sources 5.4% 4.9% 6.0% Dar es Salaam: Employment 35.4% 28.8% 41.4% Self employment 30.4% 24.5% 36.3% Agriculture 2.0% 1.0% 2.9% Transfers (incl. remittances) 7.9% 5.8% 10.0% Remittances 7.8% 5.7% 10.0% Other sources 10.9% 7.9% 13.9% Other urban: Employment 22.0% 19.6% 23.5% Self employment 29.8% 27.8% 31.9% Agriculture 17.4% 15.0% 19.8% Transfers (incl. remittances) 7.7% 6.6% 8.8% Remittances 7.2% 6.3% 8.0% Other sources 9.4% 8.3% 10.6% Rural: Employment 6.6% 4.9% 7.6% Self employment 14.2% 12.3% 16.1% Agriculture 43.6% 41.7% 45.5% Transfers (incl. remittances) 4.5% 3.9% 5.1% Remittances 4.4% 3.9% 5.0% Other sources 4.3% 3.8% 4.9% 50 Table 41. Composition of income sources by quintiles (value of non-monetary food consumption added back), 2000/2001 2000/2001 Share of 95% Confidence interval total income Low High Employment 0-20% 6.9% 5.1% 8.0% 20-40% 8.0% 6.3% 9.1% 40-60% 12.7% 9.0% 15.8% 60-80% 15.7% 12.8% 17.5% 80-100% 26.0% 19.5% 31.4% Self employment 0-20% 14.3% 12.7% 15.8% 20-40% 16.5% 14.6% 18.4% 40-60% 19.7% 16.1% 23.2% 60-80% 21.6% 18.7% 24.4% 80-100% 24.3% 17.5% 31.0% Agriculture 0-20% 41.2% 39.2% 43.2% 20-40% 40.1% 37.6% 42.6% 40-60% 35.3% 32.5% 38.2% 60-80% 31.1% 27.4% 34.8% 80-100% 23.1% 18.6% 27.6% Transfers (incl. remittances) 0-20% 4.6% 4.0% 5.2% 20-40% 5.3% 4.4% 6.2% 40-60% 5.1% 4.3% 5.9% 60-80% 6.5% 4.9% 8.1% 80-100% 6.0% 4.3% 7.7% Remittances 0-20% 4.5% 3.9% 5.1% 20-40% 5.2% 4.3% 6.1% 40-60% 4.8% 4.1% 5.6% 60-80% 6.3% 4.8% 7.8% 80-100% 5.9% 4.2% 7.7% Other sources 0-20% 4.9% 4.3% 5.6% 20-40% 4.6% 3.8% 5.4% 40-60% 6.1% 4.8% 7.5% 60-80% 6.9% 5.6% 8.2% 80-100% 6.9% 5.1% 8.7% 51 Table 42. Poverty by number of household members 1991/1992 2000/2001 95% Confidence interval Population 95% Confidence interval Population Poverty Low High share Poverty* Low High share* 1 5.8% 1.7% 9.9% 1.4% 5.3% 3.1% 7.5% 1.9% 2 10.7% 3.9% 17.6% 3.4% 11.0% 6.8% 15.2% 4.2% 3 12.9% 7.6% 18.1% 6.2% 15.5% 11.5% 19.5% 9.6% 4 20.4% 13.7% 27.0% 9.1% 20.8% 16.6% 25.0% 12.8% 5 27.0% 20.5% 33.4% 10.8% 27.4% 22.9% 32.0% 13.8% 6 38.3% 30.9% 45.7% 12.4% 35.0% 29.2% 40.8% 13.8% 7 44.0% 35.6% 52.5% 11.9% 45.7% 39.2% 52.2% 11.9% 8 45.2% 36.4% 53.9% 10.0% 44.4% 37.2% 51.6% 8.5% 9 35.7% 25.0% 46.3% 8.3% 47.9% 40.2% 55.6% 6.0% 10 or more 57.2% 48.8% 65.5% 26.5% 56.6% 49.5% 63.6% 17.5% *Bold signifies that the difference across years is significant at the 5% level Table 43. Poverty by number of children aged 5 or younger 1991/1992 2000/2001 95% Confidence interval Population 95% Confidence interval Population Poverty Low High share Poverty* Low High share* 0 32.0% 27.6% 36.3% 36.8% 30.8% 27.2% 34.5% 32.9% 1 39.8% 34.2% 45.3% 35.6% 32.3% 28.2% 36.4% 32.7% 2 45.3% 36.8% 53.8% 18.0% 40.5% 35.8% 45.2% 23.2% 3 42.3% 26.1% 58.4% 5.4% 44.5% 34.9% 54.1% 6.8% 4 or more 53.4% 36.5% 70.3% 4.2% 50.1% 32.3% 67.9% 4.4% *Bold signifies that the difference across years is significant at the 5% level 52 Table 44. Poverty by civil status of head of household 1991/1992 2000/2001 95% Confidence interval Population 95% Confidence interval Population Poverty Low High share Poverty* Low High share* Never married 13.3% 7.2% 19.4% 3.8% 20.7% 16.3% 25.1% 5.5% Married 40.2% 35.6% 44.8% 87.6% 36.3% 33.0% 39.6% 81.6% Divorced 29.0% 17.4% 40.6% 3.7% 28.8% 22.1% 35.5% 5.1% Widowed 37.3% 25.2% 49.4% 2.5% 40.0% 31.9% 48.2% 7.8% *Bold signifies that the difference across years is significant at the 5% level Table 45. Poverty by level of schooling of head of household 1991/1992 2000/2001 95% Confidence interval Population 95% Confidence interval Population Poverty Low High share Poverty* Low High share* No primary educ. 45.6% 39.1% 52.2% 27.2% 50.7% 45.5% 55.8% 25.7% Some primary educ. 36.6% 31.5% 41.7% 55.0% 31.5% 28.1% 34.8% 59.2% Compl. primary educ. 34.2% 22.5% 46.0% 4.4% 29.6% 17.6% 41.7% 2.7% Some secondary educ. 21.2% 9.9% 32.5% 1.6% 14.3% 6.9% 21.6% 1.7% Compl secondary educ. 9.6% 4.5% 14.8% 2.3% 12.6% 8.2% 17.0% 3.5% Post secondary educ. 10.7% 2.7% 18.6% 2.1% 10.6% 0.1% 21.2% 3.3% Adult eduation only 51.0% 38.2% 63.7% 7.4% 46.5% 34.4% 58.6% 4.0% *Bold signifies that the difference across years is significant at the 5% level 53 Table 46. Levels of completed schooling of head of household by quintiles 1991/1992 2000/2001 Share 95% Confidence interval Share* 95% Confidence interval Low High Low High No Primary 0-20% 32.4% 27.2% 37.6% 35.3% 31.4% 39.3% 20-40% 29.3% 23.8% 34.7% 23.9% 20.8% 27.1% 40-60% 22.9% 17.4% 28.3% 20.9% 16.8% 25.1% 60-80% 19.9% 14.6% 25.2% 12.8% 10.3% 15.3% 80-100% 13.9% 8.5% 19.4% 8.6% 4.6% 12.5% Some Primary 0-20% 52.1% 46.7% 57.5% 54.1% 50.0% 58.2% 20-40% 54.2% 48.0% 60.4% 63.1% 59.3% 66.9% 40-60% 56.6% 50.5% 62.6% 64.2% 59.7% 68.7% 60-80% 60.1% 53.2% 67.0% 64.7% 60.5% 68.9% 80-100% 61.5% 52.8% 70.1% 55.0% 48.7% 61.3% Completed Primary 0-20% 3.8% 2.0% 5.6% 2.5% 1.3% 3.6% 20-40% 3.9% 2.1% 5.6% 2.9% 1.8% 4.0% 40-60% 4.6% 2.4% 6.7% 2.0% 1.3% 2.7% 60-80% 5.4% 2.8% 8.1% 3.8% 1.3% 6.2% 80-100% 7.1% 3.2% 11.1% 3.0% 1.1% 4.9% Some Secondary 0-20% 0.8% 0.3% 1.4% 0.8% 0.4% 1.2% 20-40% 0.9% 0.4% 1.5% 1.1% 0.4% 1.9% 40-60% 2.3% 1.1% 3.5% 2.3% 1.4% 3.2% 60-80% 3.5% 1.7% 5.2% 3.3% 1.9% 4.6% 80-100% 3.1% 0.9% 5.4% 4.9% 2.3% 7.4% Completed Secondary 0-20% 0.6% 0.3% 0.9% 1.3% 0.9% 1.8% 20-40% 2.3% 0.7% 3.9% 3.3% 2.2% 4.4% 40-60% 3.9% 2.4% 5.4% 3.1% 2.3% 4.0% 60-80% 3.9% 2.2% 5.5% 6.8% 5.0% 8.5% 80-100% 5.3% 3.2% 7.4% 12.5% 7.9% 17.0% Post Secondary 0-20% 0.6% 0.1% 1.0% 1.0% 0.0% 2.0% 20-40% 2.8% 0.8% 4.8% 2.0% 1.3% 2.8% 40-60% 2.8% 1.3% 4.3% 3.6% 2.5% 4.7% 60-80% 3.1% 1.6% 4.7% 6.1% 4.3% 7.9% 80-100% 5.5% 2.8% 8.2% 15.0% 10.9% 19.1% Adult education only 0-20% 9.7% 6.4% 13.0% 5.0% 2.9% 7.0% 20-40% 6.6% 3.6% 9.7% 3.7% 1.7% 5.6% 40-60% 7.0% 3.9% 10.1% 3.8% 1.7% 5.9% 60-80% 4.2% 0.9% 7.4% 2.6% 0.3% 5.0% 80-100% 3.5% 0.7% 6.3% 1.1% 0.1% 2.1% *Bold signifies that the difference across years is significant at the 5% level 54 Table 47. Poverty by employment of head of household 1991/1992 2000/2001 95% Confidence interval Population 95% Confidence interval Population Poverty Low High share Poverty* Low High share* Farming/fishing 42.3% 37.2% 47.5% 77.9% 39.6% 35.9% 43.4% 70.5% Paid employee 18.6% 13.0% 24.1% 6.9% 16.4% 12.2% 20.6% 11.1% Self employed 19.8% 11.2% 28.5% 5.9% 20.9% 17.6% 24.2% 10.9% Family worker 31.3% 22.8% 39.8% 6.0% 43.4% 32.9% 54.0% 2.4% No activity 38.6% 24.0% 53.3% 3.3% 44.5% 36.4% 52.5% 5.0% *Bold signifies that the difference across years is significant at the 5% level Table 48. Employment of head of household by quintiles 1991/1992 2000/2001 Share 95% Confidence interval Share* 95% Confidence interval Low High Low High Farming/Fishing 0-20% 85.6% 82.2% 89.0% 78.5% 75.8% 81.2% 20-40% 77.9% 73.2% 82.6% 74.7% 71.0% 78.3% 40-60% 72.9% 68.5% 77.2% 66.9% 63.0% 70.9% 60-80% 68.1% 63.0% 73.3% 57.4% 51.4% 63.3% 80-100% 63.3% 55.0% 71.7% 42.0% 34.4% 49.6% Paid Employment 0-20% 3.2% 2.3% 4.2% 5.7% 4.3% 7.2% 20-40% 6.7% 4.2% 9.3% 8.6% 6.7% 10.4% 40-60% 9.9% 7.2% 12.7% 14.3% 11.3% 17.2% 60-80% 11.1% 8.0% 14.2% 18.5% 15.3% 21.6% 80-100% 13.5% 8.3% 18.7% 31.2% 25.1% 37.3% Self Employment 0-20% 3.0% 1.5% 4.5% 6.9% 5.4% 8.4% 20-40% 7.2% 4.8% 9.6% 9.7% 7.7% 11.7% 40-60% 7.1% 4.3% 9.9% 11.9% 9.4% 14.5% 60-80% 7.9% 5.5% 10.4% 19.0% 13.9% 24.1% 80-100% 11.1% 6.8% 15.4% 21.6% 16.7% 26.5% Family Employment 0-20% 4.9% 3.0% 6.9% 2.8% 1.9% 3.7% 20-40% 5.0% 3.1% 7.0% 2.0% 1.3% 2.7% 40-60% 6.0% 4.2% 7.9% 2.7% 1.7% 3.6% 60-80% 9.0% 6.3% 11.6% 1.7% 1.1% 2.3% 80-100% 10.3% 6.9% 13.6% 1.7% 1.0% 2.5% No Economic Activity 0-20% 3.2% 1.7% 4.7% 6.1% 4.8% 7.5% 20-40% 3.1% 1.2% 5.1% 5.1% 3.4% 6.8% 40-60% 4.0% 1.9% 6.1% 4.2% 2.9% 5.4% 60-80% 3.8% 1.4% 6.3% 3.5% 2.3% 4.7% 80-100% 1.9% 0.5% 3.2% 3.5% 0.5% 6.5% *Bold signifies that the difference across years is significant at the 5% level 55 Table 49. Main source of income 1991/1992 2000/2001 95% Confidence interval Population 95% Confidence interval Population Poverty Low High share Poverty* Low High share* Crop 48.5% 42.1% 54.9% 40.2% 39.8% 35.8% 43.8% 58.6% Livestock 35.3% 22.5% 48.2% 5.6% 52.2% 39.3% 65.1% 5.7% Business 37.0% 29.7% 44.2% 24.4% 23.6% 18.6% 28.5% 12.6% Wages 31.3% 23.5% 39.2% 9.5% 14.8% 10.9% 18.7% 9.1% Other 25.2% 20.0% 30.4% 20.4% 33.5% 27.7% 39.2% 13.9% *Bold signifies that the difference across years is significant at the 5% level Table 50. Poverty by type of business 1991/1992 2000/2001 95% Confidence interval Population 95% Confidence interval Population Poverty Low High share Poverty* Low High share* Mainland Tanzania: No business 42.4% 36.6% 48.1% 53.9% 39.3% 33.6% 45.0% 64.2% Agriculture 33.0% 26.9% 39.2% 27.4% 31.2% 25.1% 37.4% 21.5% Wholesale 34.0% 27.0% 41.0% 13.9% 23.6% 16.5% 30.6% 11.4% Other 32.7% 20.5% 44.9% 3.8% 34.0% 21.8% 46.2% 3.0% *Bold signifies that the difference across years is significant at the 5% level 56 Table 51. Main type of business by Dar es Salaam, other urban, and rural areas 1991/1992 2000/2001 Share 95% Confidence interval Share 95% Confidence interval Low High Low High Dar es Salaam: No business 65.1% 60.6% 69.6% 50.4% 44.0% 56.8% Agriculture 11.1% 8.2% 14.1% 11.8% 8.2% 15.5% Wholesale 18.8% 15.3% 22.3% 16.7% 13.5% 19.8% Other 5.0% 3.3% 6.6% 21.0% 14.5% 27.6% Other urban: No business 47.3% 37.1% 57.6% 44.6% 40.8% 48.3% Agriculture 28.7% 17.7% 39.6% 21.0% 18.5% 23.5% Wholesale 19.9% 13.7% 26.1% 19.0% 17.2% 20.7% Other 4.1% 1.9% 6.4% 15.5% 13.9% 17.2% Rural: No business 59.8% 54.2% 65.4% 59.0% 54.6% 63.5% Agriculture 25.5% 19.9% 31.1% 21.3% 17.6% 25.1% Wholesale 11.4% 9.0% 13.9% 8.8% 7.1% 10.5% Other 3.3% 1.4% 5.1% 10.8% 8.7% 13.0% *Bold signifies that the difference across years is significant at the 5% level 57 Table 52. Main type of business by quintiles 1991/1992 2000/2001 Share 95% Confidence interval Share* 95% Confidence interval Low High Low High No business 0-20% 67.4% 59.8% 74.9% 66.2% 60.0% 72.4% 20-40% 60.7% 53.6% 67.7% 56.8% 51.6% 61.9% 40-60% 55.5% 48.7% 62.3% 55.6% 49.8% 61.4% 60-80% 53.7% 47.5% 59.9% 54.1% 48.9% 59.4% 80-100% 55.3% 47.4% 63.2% 49.3% 44.5% 54.1% Agriculture 0-20% 20.8% 13.7% 27.9% 15.4% 10.0% 20.7% 20-40% 21.7% 15.5% 27.9% 22.2% 17.9% 26.5% 40-60% 30.3% 23.9% 36.8% 22.2% 17.3% 27.1% 60-80% 27.3% 20.7% 33.8% 20.9% 16.6% 25.2% 80-100% 25.5% 18.6% 32.5% 22.3% 17.6% 26.9% Wholesale and Retail 0-20% 9.5% 6.2% 12.9% 6.4% 4.5% 8.4% 20-40% 13.7% 8.7% 18.8% 8.5% 6.4% 10.6% 40-60% 11.2% 7.1% 15.3% 8.9% 6.6% 11.2% 60-80% 16.1% 12.2% 20.0% 13.3% 10.0% 16.5% 80-100% 13.8% 10.9% 16.8% 16.8% 14.3% 19.2% Other 0-20% 2.3% 0.2% 4.4% 12.0% 8.2% 15.8% 20-40% 3.9% 1.6% 6.2% 12.6% 9.5% 15.7% 40-60% 3.0% -0.7% 6.7% 13.3% 10.0% 16.6% 60-80% 2.9% 1.6% 4.2% 11.7% 9.4% 13.9% 80-100% 5.3% 2.8% 7.8% 11.7% 9.3% 14.1% *Bold signifies that the difference across years is significant at the 5% level 58 Table 53. Number of household member by Dar es Salaam, other urban, and rural areas 95% Confidence interval 95% Confidence interval Members Low High Members Low High Dar es Salaam 6.6 6.0 7.3 6.1 5.7 6.5 Other urban 7.0 6.2 7.8 6.3 6.0 6.6 Rural 8.1 7.6 8.6 7.0 6.6 7.3 *Bold signifies that the difference across years is significant at the 5% level Table 54. Number of children aged 5 or younger by Dar es Salaam, other urban, and rural areas 95% Confidence interval 95% Confidence interval Children Low High Children Low High Dar es Salaam 0.5 0.4 0.6 0.8 0.6 0.9 Other urban 0.9 0.7 1.2 1.0 0.9 1.0 Rural 1.2 1.0 1.3 1.3 1.2 1.4 *Bold signifies that the difference across years is significant at the 5% level Table 55. Average length of education of head of household by Dar es Salaam, other urban, and rural areas 95% Confidence interval 95% Confidence interval Years Low High Years Low High Dar es Salaam 7.4 6.8 8.1 7.8 7.1 8.4 Other urban 5.8 5.3 6.3 6.5 6.3 6.8 Rural 3.5 3.2 3.8 4.3 4.0 4.5 *Bold signifies that the difference across years is significant at the 5% level 59 Table 56. Regression analysis, log per adult equivalent real consumption, 2000/2001 interaction terms (regional dummies included) Dar es Salaam Other urban Rural 95% Confidence interval 95% Confidence interval 95% Confidence interval Variable Estimate Low High Estimate Low High Estimate Low High y2001 -0.20 -0.74 0.35 0.08 -0.52 0.68 0.49 0.04 0.95 HH members sq. 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 0.00 Age HHH 0.00 -0.01 0.02 0.02 -0.01 0.04 0.00 -0.01 0.02 Age HHH sq. 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Age HHH missing -0.52 -0.85 -0.20 -0.04 -0.62 0.55 -0.36 -0.73 0.02 No. HH members aged 0-9 -0.17 -0.20 -0.14 -0.17 -0.20 -0.13 -0.12 -0.14 -0.09 No. HH members aged 10-14 -0.25 -0.28 -0.21 -0.25 -0.31 -0.20 -0.17 -0.19 -0.14 No. HH members aged 60 or above -0.23 -0.32 -0.13 -0.12 -0.23 -0.01 -0.06 -0.12 0.00 No. female HH members aged 15-59 -0.20 -0.23 -0.17 -0.21 -0.26 -0.17 -0.16 -0.19 -0.13 No. male HH members aged 15-59 -0.27 -0.31 -0.22 -0.22 -0.27 -0.18 -0.13 -0.16 -0.10 Some primary education 0.02 -0.07 0.11 0.23 0.12 0.34 0.14 0.07 0.21 Completed primary education 0.02 -0.10 0.13 0.36 0.22 0.49 0.32 0.19 0.45 Some secondary education 0.17 0.03 0.30 0.22 -0.03 0.47 0.24 0.02 0.45 Completed secondary education 0.14 0.03 0.25 0.42 0.29 0.55 0.37 0.24 0.50 Post secondary education 0.41 0.29 0.54 0.45 0.31 0.59 0.46 0.31 0.60 Adult education only 0.00 -0.17 0.17 0.14 -0.05 0.33 0.04 -0.07 0.15 No. livestock 0.00 0.00 0.01 0.01 0.00 0.01 0.00 0.00 0.01 No. fields 0.03 -0.03 0.09 0.00 -0.02 0.02 0.00 -0.01 0.01 Plough 0.19 0.01 0.38 0.03 -0.18 0.24 0.09 -0.03 0.21 Plough missing -0.58 -0.80 -0.35 -0.12 -0.58 0.34 -0.02 -0.37 0.33 y2001 interaction terms: 0.00 0.00 0.00 0.00 0.00 0.00 HH members sq. 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Age HHH 0.00 -0.03 0.02 -0.02 -0.04 0.01 -0.01 -0.03 0.01 Age HHH sq. 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Age HHH missing 0.51 -0.25 1.28 -0.30 -0.96 0.36 0.35 -0.16 0.86 No. HH members aged 0-9 -0.04 -0.09 0.01 -0.01 -0.05 0.03 -0.02 -0.04 0.01 No. HH members aged 10-14 0.02 -0.04 0.08 0.03 -0.03 0.09 -0.01 -0.04 0.02 No. HH members aged 60 or above 0.01 -0.15 0.16 -0.05 -0.17 0.07 -0.08 -0.14 -0.01 No. female HH members aged 15-59 0.04 -0.01 0.08 0.07 0.02 0.12 0.03 0.00 0.07 No. male HH members aged 15-59 0.03 -0.03 0.09 0.04 -0.01 0.09 0.00 -0.04 0.03 Some primary education 0.37 0.14 0.61 0.02 -0.09 0.13 0.07 0.00 0.15 Completed primary education 0.66 0.33 1.00 0.01 -0.14 0.16 0.12 -0.06 0.29 Some secondary education 0.35 0.07 0.64 0.31 0.06 0.57 0.37 0.13 0.61 Completed secondary education 0.47 0.21 0.73 0.18 0.04 0.32 0.35 0.19 0.51 Post secondary education 0.56 0.27 0.85 0.32 0.16 0.47 0.44 0.26 0.61 Adult education only 0.52 0.09 0.96 -0.10 -0.31 0.10 0.01 -0.12 0.14 60 Table 56. Continued. Regression analysis, log per adult equivalent real consumption (regional dummies included) Dar es Salaam Other urban Rural 95% Confidence interval 95% Confidence interval 95% Confidence interval Variable Estimate Low High Estimate Low High Estimate Low High No. livestock 0.01 0.00 0.01 0.00 -0.01 0.00 0.00 0.00 0.00 No. fields -0.01 -0.07 0.06 0.00 -0.02 0.03 0.00 0.00 0.01 Plough -0.73 -1.03 -0.43 0.14 -0.09 0.37 0.04 -0.10 0.17 Plough missing -0.03 -0.56 0.50 -0.26 -0.75 0.23 -0.52 -0.99 -0.05 Constant 9.17 8.82 9.53 8.80 8.21 9.39 8.42 7.99 8.85 # observations 2349 14872 9778 R squared 0.46 0.34 0.29 61 Table 57a. Regression analysis, log per adult equivalent consumption, Dar es Salaam, coefficients in levels (regional dummies included) 1991/1992 2000/2001 95% Confidence interval 95% Confidence interval Variable Estimate Low High Estimate Low High HH members sq. 0.01 0.01 0.01 0.01 0.01 0.01 Age HHH 0.00 -0.01 0.02 0.00 -0.02 0.02 Age HHH sq. 0.00 0.00 0.00 0.00 0.00 0.00 Age HHH missing -0.52 -0.85 -0.20 -0.01 -0.70 0.68 No. HH members aged 0-9 -0.17 -0.20 -0.14 -0.21 -0.25 -0.17 No. HH members aged 10-14 -0.25 -0.28 -0.21 -0.23 -0.27 -0.18 No. HH members aged 60 or above -0.23 -0.32 -0.13 -0.22 -0.35 -0.10 No. female HH members aged 15-59 -0.20 -0.23 -0.17 -0.16 -0.20 -0.13 No. male HH members aged 15-59 -0.27 -0.31 -0.22 -0.24 -0.28 -0.20 Some primary education 0.02 -0.07 0.11 0.39 0.17 0.61 Completed primary education 0.02 -0.10 0.13 0.68 0.37 0.99 Some secondary education 0.17 0.03 0.30 0.52 0.27 0.77 Completed secondary education 0.14 0.03 0.25 0.61 0.37 0.84 Post secondary education 0.41 0.29 0.54 0.97 0.71 1.23 Adult education only 0.00 -0.17 0.17 0.53 0.12 0.93 No. livestock 0.00 0.00 0.01 0.01 0.01 0.02 No. fields 0.03 -0.03 0.09 0.02 0.00 0.05 Plough 0.19 0.01 0.38 -0.54 -0.77 -0.31 Plough missing -0.58 -0.80 -0.35 -0.61 -1.09 -0.13 Constant 9.17 8.82 9.53 8.97 8.56 9.39 # observations 1124 1225 R squared 0.48 0.37 62 Table 57b. Regression analysis, log per adult equivalent consumption, other urban, coefficients in levels (regional dummies included) 1991/1992 2000/2001 95% Confidence interval 95% Confidence interval Variable Estimate Low High Estimate Low High HH members sq. 0.01 0.01 0.01 0.01 0.01 0.01 Age HHH 0.02 -0.01 0.04 0.00 0.00 0.01 Age HHH sq. 0.00 0.00 0.00 0.00 0.00 0.00 Age HHH missing -0.04 -0.62 0.55 -0.34 -0.65 -0.03 No. HH members aged 0-9 -0.17 -0.20 -0.13 -0.18 -0.19 -0.16 No. HH members aged 10-14 -0.25 -0.31 -0.20 -0.22 -0.24 -0.20 No. HH members aged 60 or above -0.12 -0.23 -0.01 -0.17 -0.20 -0.14 No. female HH members aged 15-59 -0.21 -0.26 -0.17 -0.14 -0.16 -0.13 No. male HH members aged 15-59 -0.22 -0.27 -0.18 -0.18 -0.20 -0.16 Some primary education 0.23 0.12 0.34 0.25 0.21 0.28 Completed primary education 0.36 0.22 0.50 0.37 0.31 0.43 Some secondary education 0.22 -0.03 0.47 0.53 0.47 0.60 Completed secondary education 0.42 0.29 0.55 0.60 0.55 0.65 Post secondary education 0.45 0.30 0.59 0.76 0.70 0.82 Adult education only 0.14 -0.05 0.33 0.04 -0.04 0.11 No. livestock 0.01 0.00 0.01 0.00 0.00 0.01 No. fields 0.00 -0.02 0.02 0.00 0.00 0.01 Plough 0.03 -0.18 0.24 0.18 0.08 0.27 Plough missing -0.12 -0.58 0.34 -0.38 -0.55 -0.20 Constant 8.80 8.21 9.39 8.88 8.77 9.00 # observations 1489 13383 R squared 0.41 0.33 63 Table 57c. Regression analysis, rural, log per adult equivalent consumption, coefficients in levels (regional dummies included) 1991/1992 2000/2001 95% Confidence interval 95% Confidence interval Variable Estimate Low High Estimate Low High HH members sq. 0.00 0.00 0.00 0.00 0.00 0.00 Age HHH 0.00 -0.01 0.02 -0.01 -0.01 0.00 Age HHH sq. 0.00 0.00 0.00 0.00 0.00 0.00 Age HHH missing -0.36 -0.73 0.02 -0.01 -0.36 0.34 No. HH members aged 0-9 -0.12 -0.14 -0.09 -0.13 -0.15 -0.12 No. HH members aged 10-14 -0.17 -0.19 -0.14 -0.18 -0.19 -0.16 No. HH members aged 60 or above -0.06 -0.12 0.00 -0.14 -0.17 -0.10 No. female HH members aged 15-59 -0.16 -0.19 -0.13 -0.12 -0.14 -0.11 No. male HH members aged 15-59 -0.13 -0.16 -0.10 -0.14 -0.16 -0.12 Some primary education 0.14 0.07 0.21 0.21 0.18 0.25 Completed primary education 0.32 0.19 0.45 0.44 0.33 0.55 Some secondary education 0.24 0.02 0.45 0.61 0.51 0.70 Completed secondary education 0.37 0.24 0.50 0.72 0.63 0.81 Post secondary education 0.46 0.31 0.60 0.89 0.80 0.99 Adult education only 0.04 -0.07 0.15 0.05 -0.02 0.12 No. livestock 0.00 0.00 0.01 0.00 0.00 0.00 No. fields 0.00 -0.01 0.01 0.00 0.00 0.01 Plough 0.09 -0.03 0.21 0.12 0.07 0.18 Plough missing -0.02 -0.37 0.33 -0.54 -0.85 -0.23 Constant 8.42 7.99 8.85 8.91 8.77 9.06 # observations 2210 7568 R squared 0.31 0.28 64 Table 58. Regression analysis, log per adult equivalent real consumption, with employment information (regional dummies included) Dar es Salaam Other urban Rural 95% Confidence interval 95% Confidence interval 95% Confidence interval Variable Estimate Low High Estimate Low High Estimate Low High y2001 -0.68 -1.31 -0.06 -0.34 -0.68 0.00 0.38 -0.02 0.78 HH members sq. 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 0.00 Age HHH 0.00 -0.02 0.02 0.00 -0.01 0.02 0.00 -0.01 0.01 Age HHH sq. 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Age HHH missing -0.56 -0.90 -0.23 -0.39 -0.67 -0.11 -0.45 -0.76 -0.14 No. HH members aged 0-9 -0.17 -0.20 -0.14 -0.17 -0.20 -0.13 -0.12 -0.14 -0.09 No. HH members aged 10-14 -0.24 -0.28 -0.21 -0.24 -0.29 -0.18 -0.16 -0.19 -0.14 No. HH members aged 60 or above -0.23 -0.33 -0.13 -0.15 -0.26 -0.04 -0.06 -0.12 -0.01 No. female HH members aged 15-59 -0.20 -0.23 -0.17 -0.22 -0.26 -0.18 -0.16 -0.18 -0.13 No. male HH members aged 15-59 -0.27 -0.31 -0.22 -0.23 -0.28 -0.19 -0.13 -0.16 -0.10 Some primary education 0.01 -0.08 0.11 0.19 0.10 0.28 0.12 0.06 0.19 Completed primary education 0.00 -0.12 0.12 0.33 0.20 0.46 0.27 0.13 0.40 Some secondary education 0.15 0.01 0.29 0.27 0.10 0.43 0.20 0.00 0.40 Completed secondary education 0.12 0.00 0.24 0.38 0.26 0.49 0.27 0.11 0.44 Post secondary education 0.38 0.24 0.52 0.42 0.29 0.55 0.35 0.20 0.51 Adult education only 0.00 -0.19 0.20 0.16 -0.02 0.34 0.04 -0.07 0.15 Paid employment 0.02 -0.12 0.16 0.08 -0.02 0.18 0.14 0.04 0.24 Self employment 0.03 -0.13 0.19 0.06 -0.04 0.15 0.09 -0.01 0.19 Family employment -0.02 -0.16 0.13 0.15 0.06 0.25 0.18 0.02 0.35 No employment -0.09 -0.31 0.14 0.16 -0.05 0.38 -0.19 -0.44 0.05 No. livestock 0.00 0.00 0.01 0.01 0.00 0.01 0.00 0.00 0.01 No. fields 0.03 -0.02 0.09 0.00 -0.02 0.02 0.00 -0.01 0.01 Plough 0.20 -0.01 0.41 0.04 -0.16 0.25 0.09 -0.02 0.21 Plough missing -0.59 -0.81 -0.36 -0.11 -0.59 0.38 -0.06 -0.41 0.30 65 Table 58. Continued. Regression analysis, log per adult equivalent real consumption, with employment information (regional dummies included) Dar es Salaam Other urban Rural 95% Confidence interval 95% Confidence interval 95% Confidence interval Variable Estimate Low High Estimate Low High Estimate Low High y2001 interaction terms: HH members sq. 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Age HHH 0.01 -0.02 0.03 0.00 -0.01 0.01 -0.01 -0.02 0.01 Age HHH sq. 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Age HHH missing 0.69 -0.03 1.40 0.15 -0.27 0.58 0.43 -0.03 0.90 No. HH members aged 0-9 -0.04 -0.09 0.01 -0.01 -0.05 0.03 -0.01 -0.04 0.02 No. HH members aged 10-14 0.01 -0.05 0.07 0.02 -0.04 0.07 -0.01 -0.04 0.02 No. HH members aged 60 or above 0.03 -0.12 0.18 -0.01 -0.12 0.10 -0.08 -0.14 -0.01 No. female HH members aged 15-59 0.03 -0.02 0.07 0.08 0.03 0.12 0.03 0.00 0.07 No. male HH members aged 15-59 0.02 -0.04 0.08 0.05 0.00 0.10 0.00 -0.04 0.03 Some primary education 0.30 0.10 0.49 0.03 -0.07 0.12 0.07 -0.01 0.15 Completed primary education 0.59 0.30 0.87 -0.01 -0.15 0.14 0.13 -0.04 0.30 Some secondary education 0.27 0.03 0.52 0.19 0.01 0.37 0.32 0.09 0.54 Completed secondary education 0.40 0.18 0.62 0.14 0.01 0.26 0.34 0.15 0.53 Post secondary education 0.47 0.21 0.72 0.24 0.10 0.39 0.39 0.21 0.58 Adult education only 0.46 0.04 0.88 -0.13 -0.33 0.06 0.00 -0.13 0.13 Paid employment 0.42 0.18 0.66 0.14 0.04 0.25 0.06 -0.06 0.17 Self employment 0.37 0.12 0.63 0.19 0.09 0.29 0.11 -0.01 0.22 Family employment 0.27 -0.04 0.59 -0.02 -0.14 0.09 -0.21 -0.39 -0.02 No employment 0.39 0.07 0.71 -0.15 -0.37 0.07 0.07 -0.18 0.33 No. livestock 0.01 0.00 0.01 0.00 -0.01 0.00 0.00 0.00 0.00 No. fields 0.01 -0.05 0.07 0.00 -0.02 0.02 0.00 0.00 0.01 Plough -0.55 -0.91 -0.19 0.15 -0.07 0.38 0.04 -0.09 0.17 Plough missing -0.01 -0.54 0.51 -0.28 -0.79 0.23 -0.45 -0.93 0.02 Constant 9.24 8.84 9.64 9.06 8.74 9.38 8.52 8.15 8.90 # observations 2349 14872 9778 R squared 0.47 0.37 0.30 66 Table 59a. Regression analysis, log per adult equivalent consumption, Dar es Salaam, coefficients in levels, with employment information (regional dummies included) 1991/1992 2000/2001 95% Confidence interval 95% Confidence interval Variable Estimate Low High Estimate Low High HH members sq. 0.01 0.01 0.01 0.01 0.01 0.01 Age HHH 0.00 -0.02 0.02 0.01 -0.01 0.02 Age HHH sq. 0.00 0.00 0.00 0.00 0.00 0.00 Age HHH missing -0.56 -0.91 -0.22 0.12 -0.51 0.76 No. HH members aged 0-9 -0.17 -0.20 -0.14 -0.21 -0.25 -0.17 No. HH members aged 10-14 -0.24 -0.28 -0.21 -0.24 -0.28 -0.19 No. HH members aged 60 or above -0.23 -0.33 -0.13 -0.20 -0.31 -0.09 No. female HH members aged 15-59 -0.20 -0.23 -0.17 -0.17 -0.21 -0.14 No. male HH members aged 15-59 -0.27 -0.31 -0.22 -0.25 -0.29 -0.21 Some primary education 0.01 -0.08 0.11 0.31 0.14 0.48 Completed primary education 0.00 -0.12 0.12 0.59 0.33 0.85 Some secondary education 0.15 0.01 0.29 0.42 0.22 0.62 Completed secondary education 0.12 0.00 0.24 0.52 0.33 0.71 Post secondary education 0.38 0.24 0.53 0.85 0.63 1.07 Adult education only 0.00 -0.19 0.20 0.46 0.09 0.84 Paid employment 0.02 -0.12 0.17 0.44 0.25 0.64 Self employment 0.03 -0.13 0.19 0.40 0.20 0.61 Family employment -0.02 -0.17 0.14 0.26 -0.02 0.54 No employment -0.09 -0.32 0.14 0.30 0.07 0.53 No. livestock 0.00 0.00 0.01 0.01 0.01 0.02 No. fields 0.03 -0.02 0.09 0.04 0.01 0.07 Plough 0.20 -0.01 0.41 -0.35 -0.64 -0.06 Plough missing -0.59 -0.82 -0.36 -0.60 -1.08 -0.12 Constant 9.24 8.84 9.64 8.56 8.07 9.05 # observations 1124 1225 R squared 0.48 0.40 67 Table 59b. Regression analysis, log per adult equivalent consumption, other urban, coefficients in levels, with employment information (regional dummies included) 1991/1992 2000/2001 95% Confidence interval 95% Confidence interval Variable Estimate Low High Estimate Low High HH members sq. 0.01 0.01 0.01 0.01 0.01 0.01 Age HHH 0.00 -0.01 0.02 0.00 0.00 0.01 Age HHH sq. 0.00 0.00 0.00 0.00 0.00 0.00 Age HHH missing -0.39 -0.67 -0.10 -0.23 -0.55 0.08 No. HH members aged 0-9 -0.17 -0.20 -0.13 -0.17 -0.19 -0.16 No. HH members aged 10-14 -0.24 -0.29 -0.18 -0.22 -0.24 -0.20 No. HH members aged 60 or above -0.15 -0.26 -0.04 -0.16 -0.19 -0.13 No. female HH members aged 15-59 -0.22 -0.26 -0.18 -0.14 -0.16 -0.13 No. male HH members aged 15-59 -0.23 -0.28 -0.18 -0.18 -0.20 -0.16 Some primary education 0.19 0.09 0.28 0.21 0.18 0.25 Completed primary education 0.33 0.20 0.46 0.32 0.26 0.39 Some secondary education 0.27 0.10 0.44 0.45 0.39 0.52 Completed secondary education 0.38 0.26 0.49 0.51 0.46 0.57 Post secondary education 0.42 0.29 0.55 0.66 0.60 0.72 Adult education only 0.16 -0.03 0.34 0.02 -0.06 0.10 Paid employment 0.08 -0.02 0.18 0.22 0.19 0.26 Self employment 0.06 -0.04 0.16 0.25 0.21 0.28 Family employment 0.15 0.06 0.25 0.13 0.06 0.20 No employment 0.16 -0.05 0.38 0.01 -0.04 0.06 No. livestock 0.01 0.00 0.01 0.01 0.00 0.01 No. fields 0.00 -0.02 0.02 0.01 0.00 0.01 Plough 0.04 -0.17 0.25 0.19 0.10 0.29 Plough missing -0.11 -0.60 0.39 -0.39 -0.56 -0.22 Constant 9.06 8.74 9.38 8.72 8.60 8.84 # observations 1489 13383 R squared 0.40 0.36 68 Table 59c. Regression analysis, log per adult equivalent consumption, rural, coefficients in levels, with employment information (regional dummies included) 1991/1992 2000/2001 95% Confidence interval 95% Confidence interval Variable Estimate Low High Estimate Low High HH members sq. 0.00 0.00 0.00 0.00 0.00 0.00 Age HHH 0.00 -0.01 0.01 -0.01 -0.01 0.00 Age HHH sq. 0.00 0.00 0.00 0.00 0.00 0.00 Age HHH missing -0.45 -0.77 -0.14 -0.02 -0.37 0.33 No. HH members aged 0-9 -0.12 -0.14 -0.09 -0.13 -0.15 -0.12 No. HH members aged 10-14 -0.16 -0.19 -0.14 -0.18 -0.19 -0.16 No. HH members aged 60 or above -0.06 -0.12 -0.01 -0.14 -0.17 -0.10 No. female HH members aged 15-59 -0.16 -0.18 -0.13 -0.12 -0.14 -0.10 No. male HH members aged 15-59 -0.13 -0.16 -0.10 -0.14 -0.16 -0.12 Some primary education 0.12 0.06 0.19 0.20 0.16 0.23 Completed primary education 0.27 0.13 0.40 0.39 0.28 0.50 Some secondary education 0.20 0.00 0.41 0.52 0.42 0.61 Completed secondary education 0.27 0.11 0.44 0.61 0.52 0.70 Post secondary education 0.35 0.19 0.51 0.75 0.65 0.85 Adult education only 0.04 -0.07 0.15 0.03 -0.03 0.10 Paid employment 0.14 0.04 0.24 0.20 0.14 0.25 Self employment 0.09 -0.02 0.19 0.19 0.14 0.25 Family employment 0.18 0.02 0.35 -0.02 -0.10 0.06 No employment -0.19 -0.44 0.05 -0.12 -0.19 -0.06 No. livestock 0.00 0.00 0.01 0.00 0.00 0.00 No. fields 0.00 -0.01 0.01 0.00 0.00 0.01 Plough 0.09 -0.02 0.21 0.13 0.07 0.19 Plough missing -0.06 -0.42 0.30 -0.51 -0.82 -0.20 Constant 8.52 8.15 8.90 8.90 8.76 9.05 # observations 2210 7568 R squared 0.32 0.29 69 Figure B1. First order stochastic dominance rpc1991 rpc2001 1 .75 eulav =35 years), women adults (>35 years), and mixed youth group (<35 years). Each group had 5-8 respondents. The guide solicited information on the following themes; Change in Activity Possibility Set/Income Earning Opportunities To focus the attention of participants on issues related to income mobility, three groups of respondents were asked to describe the activity choices that are available in 2004 and in 1994; crops grown more/less than in 1994 and what has caused the differences; changes in markets, access to inputs, traders, prices over the past 10 years; introduction of new crops; and migration issues. The respondents were further requested to put the changes on time line in order to show the trend for community and explain whether the time trend in 2004 is above or below 1994 and at important changes in trend, whether it is above or below 1994. In the time line, changes in public services between 1994 and 2004 were added. Questions addressed in relation to public services include services that are available now (extension, village health worker, school, dispensary, road, hospital, water) compared to 2004. The respondents were further requested to picture three cases of individuals, who have gone up, gone down and chronically poor/trapped in poverty, and provide characteristics of individuals in each group (factors responsible for such a situation). Probing on what constraints chronically poor people from taking advantage of the existing opportunities, and the role of assets, membership to formal and informal groups, personal characteristics (education, alcoholic etc), family size, political connections, and income-generating activities (farmer/business/formal employment) on growth was done. Objectives of this session were: (i) To assess changes in income generation over time, the reasons for these change, and those most affected (positively or negatively) by these changes. (ii) To assess the role that outside influences play in driving economic mobility (upward or downward). ____________________________________________________________________ 10 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank (iii) To obtain an understanding of local views on the characteristics of wealthy and poor households and of the reasons as to why these characteristics are associated to wealth or poverty. (iv) To initiate a discussion of why some households are able to succeed in gaining wealth while others fall behind economically or are unable to improve their situation. (v) To assess local awareness of existing economic opportunities both inside and outside the village and of barriers that influence and constraint economic mobility. Identification and Sorting of shocks Following an initial identification of factors that prevent change to happen, a discussion about shocks (as we understand them) was initiated. The major community and idiosyncratic shocks were identified. Probing was done on climatic, economic, illness or mortality, violence and/or crime, political, and service related shocks. After identifying the shocks, respondents were asked to categorize and rank the categories of shocks. Ranking was done based on the importance of the shock, that is, whether the shock affected majority of the community members. Coping Strategies This session intended to elicit how people cope with shocks. Hypothetical examples were used to solicit information on coping with a sudden shock requiring a specific amount of money. Four hypothetical cases were used: raising a specific amount overnight for an emergency illness; raising the same amount within two months; raising the same amount within two months for a problem that has hit a household together with several others in the village; and raising the same amount overnight for a positive event such as sending a child to school6. After identifying the coping strategies, respondents were requested to categorize various responses and rank the categories. The essence of this exercise was to learn what strategies are potentially successful and used, why others are not used in particular circumstances, and whether used strategies differ by status of an individual (rich/poor, old young etc). Potential for Collective Action Through focus group interviews the role of groups, associations, networks and interpersonal relationships in enhancing economic progress was investigated. Participants were asked to 6 Amount was varied at each village depending on perceived economic situation of the village. ____________________________________________________________________ 11 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank describe membership in local groups, and associations in order to elicit joint actions that are undertaken, and to tell how these are beneficial for households well-being and income generation. Issues examined include the following: (i) Important of social capital for economic progress of individuals and of the community as a whole; the groups that benefit most, excluded from joint activities or its benefits, or refuse to participate. (ii) The change in social capital over time; whether new groups, organizations or collectives have evolved within the village, or whether some have disappeared, and the reasons for these changes. (ii) The Quantitative Component The KHDS collected detailed information from about 800 households in nearly 50 communities during 4 rounds held between 1991-1994. This survey collected a wealth of information at household level on (1) demographic characteristics, (2) health status, and health seeking behavior, (3) nutritional status, (4) mortality and related expenditures, (5) human capital--enrollments and education expenditures, (6) fertility and contraceptive use, (7) time use in the labor force, and other productive activities, (8) income levels and sources, (9) assets and durable goods, (10) consumption expenditure, (11) savings, debts and transfers and (12) characteristics of non-resident parents and children. In addition, questionnaires were administered to community key informants (such as leaders and influential people), at nearby health facility, school, and market. These households were revisited in 2004 to provide a unique opportunity for longitudinal data stretching a long period of time. Consequently, the qualitative part of this study can only inform future KHDS surveys. During fieldwork, the current and past community questionnaires from the KHDS, along with the full data set, were used to inform the field team about community characteristics, now and in the past. This information is critical at assessing the quality of retrospective responses, and helps to put case histories in perspective. Further, the information in the community questionnaire is used to assess the generalizability of the focus group findings. ____________________________________________________________________ 12 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank 4.0 STUDY FINDINGS 4.1 Community Perception of Economic Mobility Over Time 4.1.1 Factors for Economic Mobility7 The surveyed communities were requested to picture their economic mobility and the factors associated with the observed economic situation. For the three villages surveyed (Villages A, C and D), the groups, unanimously, agreed that their communities have experienced an upward economic growth in the period of study. The same was deducted from case interviews at Village B. Major factors for growth include good coffee prices, introduction of new cash crops such as tobacco, and new economic opportunities such as mining, trading, and formation of groups aimed at helping each other. Livestock was considered as an important precautionary saving. Other factors featured from individual life histories include formal employment and remittances/transfers (Appendix 2). Despite falling prices, coffee is still regarded as an important source of income by majority of villagers because coffee is still the main cash crop whereas banana is a major food crop. In Village D, tobacco is slowly replacing coffee. Agriculture The 1994-1997 period was mentioned as the very good period economically. Coffee performed very well in terms of production and prices. In addition, the influx of refugees resulted to increased market mainly for bananas and beans. Majority of villagers located close to Ugandan boarder sold their coffee in Uganda and the price ranged from TShs 30,000 to a maximum of TShs 40,000 per bag of coffee estimated to have about 70-80 kilograms. The price in the Ugandan market is still good compared to that in Tanzania. At the time of survey, a bag of coffee was sold at TShs 14,000 as observed at Village C. Cooperative Union [Kagera Cooperative Union (KCU), Karagwe District Cooperative Union (KDCU) and Biharamulo Cooperative Union (BCU)] branches located in the surveyed villages were not operating very well at the time of survey. This is due to the fact that villagers do not wish to sell their coffee through these Unions. The reasons given include low prices, and selling on credit-- sometimes farmers are not paid instantly. Coffee and cotton were the major cash crops in Village D before the collapse of cooperative unions. Cotton like other major cash crops such as coffee, and sisal collapsed after trade liberalization that resulted to short and/or long term closure of cooperative unions. For cotton in particular, liberalization and privatization of textile industries resulted to closure/partial functioning of some of the industries (some are being revived). This was also aggravated by cheap textile imports. In 1998 tobacco was introduced in Village D. Since then tobacco has 7 The negative factors are discussed in section 4.3 ____________________________________________________________________ 13 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank performed very well in terms of production and price. At the time of survey, the first grade of tobacco was earning TShs 1010 per kilogram. Even the last grade of tobacco (grade 7) was sold at TShs 190, a price that is much better compared to coffee price. Tobacco grades fall between grade 1 to grade 4 for majority of farmers and an average farmer could harvest more than 100 kilograms per season. Despite cash crop production, food crops are also traded in the community and in the nearby districts and regions. These include bananas, beans, and cassava. For example, in 2003, cassava flour fetched a very good price in Uganda market. An interesting finding regarding new development in agriculture is participation of farmers in contract farming. Some farmers at Village D have revived their cotton farms and they have signed a contract to produce and supply requisite quantity of cotton per year to a certain private trader8. For example, a group of 4 farmers have signed a contract to produce 5 tons of cotton per year. However, other farmers depend on seasonal private buyers. Off farm Income Generation Activities Fishing was found to be a major off farm income generating activity and a major protein source for majority of Village A households. Villagers residing in Village B also participate in fish retailing. Both men and women participate in the fishing business, even though with clear cut gender role distinctions. Men participate in actual fishing in the lake and wholesale, while women are involved in fish retailing, and processing. The older cohort of men and women (>35 years) mentioned that fishing is basically men's activity due to traditional reasons whereby a woman is considered impure and thus she can pollute the water if she goes in it for fishing. In actual fact, women are not involved in actual fishing even when they possess fishing equipment such as fishnets and boats. They have to put the equipment on rent. However, a younger cohort of women (<35 years) had different perception regarding participation of women in actual fishing. They mentioned that, given capital and opportunity, they could participate in actual fishing just like men. New other income generation avenues have emerged in the study area. The most notable one is mining. In 2000, gold was discovered in Village D and mining activities started right away. A significant segment of the population is involved in mining process by participating in different activities. The youth participates in the actual mining but adults sell a piece of their land (a radius of about 20 square meters is sold at TShs 5,000) for mining activities once the land has been identified as a potential area for gold deposits. Mining activities are associated with increased trade on the mining sites. Although women do not participate in actual mining, they benefit from mining activities by providing grass to thatch miners' houses, fetching water for sale, and food vending. 8 Respondents could not tell the name of the trader. ____________________________________________________________________ 14 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank Establishment of famous development groups named Vigombe was mentioned to be instrumental in the upward growth observed in Village C. Through these groups members help each other in terms of providing capital to start a business, secure a new plot of land, or buy livestock. In addition, the group members contribute household items such as mattress, blanket, kitchen utensils etc as per the request of a member. The constitution allows members to request for whatever they need and monthly visits are arranged to provide/contribute what the member has requested (cash, in kinds such as household items etc). The visit rotates until every member is reached, that is, one complete cycle. Then the cycle starts again or the Kigombe (singular) is dissolved. It should be noted that, every member should have got the same commodities at the completion of the cycle. Whatever a member contributes, is reciprocated when his/her turn comes. In case the Kigombe agreed to contribute money, every member would have got the same amount of money by the end of the cycle. Local brewing is another source of income for a big segment of population studied. Since the banana market is flimsy, bananas are converted into different types of local brew. Majority of community members in all villages studied are involved in the local brew chain either as banana providers, actual possessors, or consumers of the final product. Both men and women participate in different processes of local brewing. Whereas men are responsible in softening the bananas and distilling, women are involved in other processes such as milling the sorghum, boiling and mixing the ingredients, and selling the brew. In Village C, local brew named konyagi is also traded across the boarder to Uganda. It is worth noting that local brewing is an activity predominantly practiced among women, especially widows. This is because the business can be started using a small capital and the raw materials for the same are also cheaply available in the village. The type of bananas used for local brewing is different from the ones used as staple food. Thus, local brewing is not associated with food insecurity resulted from decreased staple food available at household level as a result of using staples for local brew in Villages A, B and C where banana is the main staple. However, local brewing has been associated with food insecurity for some households in Village D because sorghum is both main staple and main ingredient in two famous types of local brew. Other important economic activities include charcoal making, lumbering, pottery which is performed mostly by women, beekeeping, quarrying, hunting, trading in small merchandise in kiosks, and food vending. Trade liberalization has resulted to mushrooming of micro businesses, with consumables coming closer. Nevertheless, quarrying is considered an inferior activity because it demands a lot of physical strength and pays less. ____________________________________________________________________ 15 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank Livestock Keeping Livestock keeping, especially of goat and local breeds of cattle, is another important economic activity. Animals are very important source of manure because artificial fertilizers are not available in the villages surveyed. Animals are also important in case of emergency situation and during periods of economic hardships (see section 4.4). Some villagers at Village C have obtained the new breed of dairy goat from Tanganyika Christian Refugee Services (TCRS), a Non-Governmental Organization (NGO) operating in refugees affected areas. In Village A, contrary to the older cohort, the younger one considers animal husbandry as a promising venture and some of the respondents are already involved in this undertaking. Kagera Dairy and Development Trust (KADADET), a local NGO is active in providing dairy cattle to those who apply and qualify to take care of them. Although livestock keeping especially of goat and local breeds of cattle is common in Village D, villagers have shifted to keeping dairy cattle. Early adopters in the village were two and others are waiting eagerly for their pay from this season's tobacco to buy dairy cattle. The majority keep goats and it is estimated that even a very poor family has at least 2 goats. The Role of Refugees in the Household Economy Refugees have been associated with positive and negative effects. The influx of refugees has resulted to the construction of the road to Village C in 1994 and rehabilitation of Nyakazi- Kigoma highway that passes through Village D in 1999 under the Special Program for Refugees Affected Areas (SPRAA) project. The SPRAA which was a project funded by European Union (EU) in collaboration with Tanzanian government was aimed at improving Tanzania's ability to cope with the refugee movements generated by the continuous crisis in the Great Lakes region. Roads to Villages A and B were rehabilitated under Tanzania Social Action Fund (TASAF) projects. These roads opened up the villages to external trade and were important for transportation of foods from the village to the refugee camps. The demand for bananas, beans, cassava flour, and local brew increased significantly during the refugee epoch. Some villagers benefited from the refugees' labor in the farm and in grazing animals, for in kind payments such as bananas and beans or for low pay. For instance, in Village D, refugees were charging TShs 6,000 for digging one acre of land but the same would cost about TShs 12,000 for a laborer in the village. However, refugees have also been associated with negative impacts. There are remnants of refugees in Kagera rain forests and cases of vehicles ambush are still been reported. Refugees are also associated with theft, especially of bananas and bringing guns into the villages. Further, they are responsible for the current observed environmental degradation as they did massive forest clearing for settlement and firewood. ____________________________________________________________________ 16 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank Social Services The major service related shock in the villages surveyed is the rehabilitation of the main roads to the villages. As mentioned earlier, main roads to Village C and highway through Village D were rehabilitated during the refugees' era but rehabilitation of the roads to Villages A and B was initiated under TASAF. Another community service introduced in the study period is the water project--in 1998-99 under Health, Sanitation and Water (HESAWA) project. At Village D, 4 community taps were constructed. Further, 4 community wells were constructed to serve the communities located far away from the sources of gravity water. Nevertheless, spring and rivers are important sources for some households residing far from the community water sources. Community taps were also constructed at Village C but villagers do not use the water from the taps because that water is hard. This water is mainly used in local brewing. Water harvesting during the rain season is very popular in Village C. During the dry season, they use water from natural streams and river. HESAWA project offered to rehabilitate the gravitational water system in Village A with the support from Bukoba Rural Local Council Authority and the community but the project did not take off. In the same community Partage offered to reconstruct wells but only two were done (see section 4.5 for details regarding Village A water project). There is a gradual change in the quality of education from 1994-2004. More children are enrolled in primary school and increased number pass the Primary School Leaving Examination (PSLE). In 2002 the Primary Education Development Program (PEDP) came in. Villagers witnessed rehabilitation of old classrooms and construction of new ones, and increased enrollment at standard one. Communities in Villages C and D are were in the process of contributing money for construction of community secondary schools at the time of survey. Due to the demand for saving and credit institutions, Village C has started a Savings and Credit Cooperative Society (SACCOS) that was to start operating in June 2004. This project was supported by TCRS. At the time of survey, members had already contributed the initial amount required (collateral) as per constitution. Villagers were very excited about the new established SACCOS because cases of burglary have been increasing in the village and it is no longer safe to keep money in the house. Other activities supported by TCRS include rehabilitation of classrooms and construction of school pit latrines, provision of health education, for example, ways to fight the spread of HIV/AIDS, and provision of small development oriented credits. United Nations Children Fund (UNICEF) has also been active in educating the community on community participation in helping orphans and children from poor families, and education on HIV/AIDS. ____________________________________________________________________ 17 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank 4.1.2 New Economic Opportunities Vanilla farming was introduced in the study area in 1997 as an alternative for coffee, even though it has not been fully adopted. Adoption of vanilla is taking a slow pace because of the risk the villagers feel is embedded in the activity. Having experienced a catastrophic fall in banana production because of the newly introduced banana fertilizer and pesticide, farmers are reluctant in going into new agricultural ventures. Villagers are not sure of the consequences of intercropping vanilla with bananas. Consequently, the early adopters of the vanilla crop have planted it at the periphery of their farms. Another problem mentioned for vanilla adoption is related to securing the seedlings for both vanilla and the recommended anchoring trees. At Village A the price for vanilla seedling ranges from TShs 300-500, whereas that for the recommended anchoring tree ranged from TShs 200-300 each. As a way of raising income and creating awareness on vanilla farming, Village A primary school has established a vanilla farm. This farm has acted as a pilot/experimental plot for households residing in Village A. Performance of that farm will be a catalyst or impediment for adoption of vanilla by a bigger segment of the community. Other crops such as tobacco and certain species of trees named mironge have been introduced in Village D. For tobacco, the dealers provide seedlings, fertilizer and pesticides on cash, and provide market for the output. Mironge tree has several uses; the leaves are used as vegetable, the bark for production of wax; seeds for medicine and for extraction of oil; and roots as medicine. Farmers have also adopted new variety of sorghum named Tegemeo. This variety that matures only after 3 months was introduced in 1999 by the Village/Ward Extension Officer. New varieties of bananas are still on experimental plots. Lumbering and carpentry activities are also growing. Households have started planting trees, not only for firewood but also for economic reasons. For instance, villagers in Village C are planting grevillea species for shade, timber and firewood. They obtain seedlings from private seedling plots. They integrate grevillea into their banana/coffee farms but they are not sure of the consequences in banana/coffee production. They, however, know that planting Eucalyptus species has a negative consequence on agricultural production because they tend to dehydrate the soil. Thus, they do not integrate Eucalyptus in their farms. Increased vocational training for youth and demand for modern furniture and houses has resulted into increased carpentry and masonry activities. Despite the fact that lumbering and vanilla farming are potential lucrative ventures, lack of land and initial capital deter poor households from grabbing these opportunities. Information asymmetry on availability of markets also impedes efforts towards such investments. For instance, while one group of respondents mentioned that one-kilogram of vanilla is sold at about TShs 15,000, others mentioned the price twice as much. Information on good crop husbandry is also not extended to majority of farmers despite the fact that there is an ____________________________________________________________________ 18 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank Agricultural Extension Officer located in the Ward. Theft cases have also increased with increased vanilla farming. Vanilla demands serious security from planting to harvesting because people steal even the seedlings. Another new opportunity for youth in the contemporary Kagera is business. Youths are involved in cross boarder trade and establishment of kiosk in the villages has gained momentum. Activities like tailoring and mechanics have also grown over time. People own bicycles and motorcycles and they could easily be repaired in the village in case of minor mechanical problems. Nonetheless, cross boarder trade has been associated with increased primary school drop out for boys. In Village C, it was found that some boys do drop out in order to participate in the business especially in assisting traders to cross the Kagera River with their merchandise. Fishing is now a more promising avenue as a result of external market in Mwanza, Uganda and the Democratic Republic of Congo (DRC). After trade liberalization several fish processing plants were established, especially in Mwanza, and this has resulted in the increased demand for fish, like the Nile Perch. However, lack of capital deters the youth from practicing modern fishing. For instance, investing in successful fishing activity, one needs about 80 fishnets and each costs about TShs 5,500, a canoe that costs about TShs 150,000, and funds to carter for miscellaneous costs such as diesel, kerosene, pressure lamps, etc. This is money that a normal young person in the village cannot obtain. What some youths have done is joining together, contributing some money and having a joint fishing group. However, the majority works for the rich people in Uganda. Mining activities started in Village D in 2000. Although mining is an economic viable activity, the consequence of such activity on the environment and future sustainability of the livelihood of those who have sold their farms to miners is a subject of further research. A new endeavor envisaged by the youth and women in Village D is participation in drama and traditional dance for pay. At the time of survey, women have established a group that participates in entertaining people in different ceremonies, and in campaigns against HIV/AIDS for pay. 4.1.3 Summary of the Communities' Economic Performance The communities studied have experienced ups and downs in their economic performance. However, they are at a higher state economically, compared to 19949. This is mainly due to the coffee boom and good coffee prices in 1994-1997, increased trading opportunities, introduction of new economic ventures such as mining, introduction of new cash crops, and 9 It is worth noting that it was hard for respondents to remember the exact year when a specific event occurred. This may result to some inconsistencies in years reported in this report. ____________________________________________________________________ 19 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank formation of development groups aimed at helping each other. The importance of remittances and formal employment are also echoed in the life histories. Boxes 1 to 3 present factors for economic mobility for three communities studied. A community consensus for Village B is not available but more than half of respondents who provided their life histories have experienced upward mobility. Appendices 3 and 4 show the community and individual economic mobility graphs respectively. The y-axis represents communities'/households' well-being and the x-axis shows years when specific event occurred. If the graph shows 2004 to be below 1994 this means that relatively, the community/respondent is worse off in 2004. The money obtained from the coffee, tobacco, and mining boom was used to build houses, acquire new plots of land and livestock, sending children to school, buying bicycles, radios, furniture, motorcycle, and about two villagers in Village C bought cars. Paying bride price was also mentioned at Villages C and D. Respondents admitted that the 10 years period saw an economic turning point for the majority of villagers. It is around this time when iron roofed houses sprouted in the villages and villagers who did not have household items such as mattresses acquired them. Box 1: Village A Community Economic Mobility Graph (1994-2004) 1994-1996 · The community experienced economic growth due to high coffee price and coffee boom, and availability of cheap labor (migrants and natives). 1996-1999 · This period was characterized by high incidence of HIV/AIDS illness and deaths, loss of labor due to repatriation of the refugees and HIV/AIDS deaths · Increased number of orphans due to HIV/AIDS deaths · Falling coffee price that resulted to poor farming management (some villagers neglected their coffee farms). 1999-2004 · The community responded to HV/AIDS shock by forming several economic and social groups · Several Non Governmental Organizations (NGOs), charity organizations and government institutions came in, for instance Partage, World Vision, Bukoba District Rural Development Program (BDRDP), and TASAF · There was introduction of new activities such as vanilla farming, proliferation of small businesses, lumbering, and intensification of fishing activities · The road to the village was rehabilitated. · Social service projects such as the water projects under HESAWA and Partage, and construction of classrooms under Primary Education Development Program were implemented (PEDP). ____________________________________________________________________ 20 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank Box 2: Village C Community Economic Mobility Graph (1994-2004) 1994-1997 · The community experienced upward economic growth due to high coffee price in Uganda market, the coffee boom, and good weather conditions. · Good market for bananas, beans, peas, maize, in Uganda and refugees affected areas. 1998-2000 · El Nino rains destroyed a lot of crops. The rains were followed by serious episodes of malaria, which resulted to unprecedented deaths in the community. The most vulnerable groups were children and elderly. · In 1999 water project was executed in the village · In 1999 TCRS started executing some programs in the community. 2001-2002 · The economy was in recession due to unsatisfactory coffee prices, and poor market for other commodities · Primary Education Development Program started to be implemented at the community primary school. 2003-2004 · Drought, followed by windy storm and hail rain · Satisfactory coffee prices · Establishment of Savings and Credit Cooperative Society (SACCOS) · The community started contributing funds for construction of community secondary school · United Nations Children Fund (UNICEF) started executing educational programs in the village ____________________________________________________________________ 21 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank Box 3: Village D Community Economic Mobility Graph (1994-2004) 1993-1994 · Drought which was followed by serious famine 1995-1998 · The economy was growing slowly due to moderate coffee and cotton prices at BCU. El Nino rains resulted to good agricultural performance. · In 1998 tobacco was introduced in the study area. 1999-2000 · The economy slowed down due to mild drought and beans pests. · Introduction of the dairy cattle · In 1998-99 the gravitational water project was executed in the village 2000-2004 · Upward economic mobility was experienced due to establishment of mining activities in the village, and booming of tobacco harvests. · Due to satisfactory cotton prices, some farmers started revitalizing their cotton farms and some have entered into contract farming. · The community started contributing funds for construction of community secondary school · Village D primary school classrooms were constructed · Primary Education Development Program started to be implemented in the community · Major shock in the period was bean pests 4.2 Individual Economic Profiles The respondents participating in FGDs were requested to picture three cases of individuals, who have gone up, gone down and chronically poor/trapped in poverty, and provide characteristics of individuals in each group (factors responsible for such a situation). The results from different groups of respondents are summarized in Table 4. Upward Mobility Hard work and cooperation feature as most important factors for growth as mentioned by the three groups of respondents. Increased human capital investment (education), good health and good planning and execution of the planned activities are other important aspects of growth mentioned by the youth and adults groups10. Other factors for upward mobility include having right ideas, entrepreneurship, being a businessman/woman, having capital to 10 Some respondents expressed their concerns about the educated unemployed and questioned the importance of educating their children if they cannot secure a formal job after graduation. The ____________________________________________________________________ 22 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank start a business, and perseverance on what one does, that is, not to despair even if one is making a temporary loss. Tolerance is very important especially for activities such as mining. Since agriculture is the backbone of the economy of the villages surveyed, having a piece of land is another important factor for growth. Hard work is negated by ill behaviors such as drunkenness, laziness, adultery, and idleness. Downward Mobility Downward mobility was associated with chronic illness such as the opportunistic infections resulting from HIV/AIDS and which renders the person in question debilitated, and thus unable to work. Besides, opportunistic infections such as Tuberculosis (TB) are also responsible in pulling individuals down because a lot of financial and non-financial resources such as time are used in taking care of the illness. A sick person is unable to work and at the same time, savings are eroded in buying medical supplies. Ill behaviors such as drunkenness, adultery, laziness, and extravagancy are also culprits for downward fall. Although local brewing is considered as an important source of income for some households in the village, it is also an impoverishing force for the consumers. Other factors include lack of education and this was related to poor planning resulted from being illiterate, being robbed, witchcraft, landlessness, being a thief, and increased demographic load. The youth group underlined the fact that after getting married and starts having children, ones economic graph stagers a bit before it takes an upward turn again. This is also clearly demonstrated in the life histories (Appendix 2). Being a thief is associated with downward mobility because if a thief is caught, he/she is fined or jailed. The fine could cost a fortune and if jailed, it takes time for thieves to be accepted back in the community and establish a viable economic venture after being paroled. Repeated crime cases may result to chronic poverty. Another important factor for downward mobility was associated with involvement in a Police case as a result of selling coffee in Ugandan market or committing other crimes. Since this is illegal business, Policemen sometimes arrest villagers while smuggling coffee and the whole consignment becomes confiscated. This is clearly demonstrated by life history C11 (Appendix 2) whereby the respondent lost 5 bags of coffee as bribe. Unfortunately, this was more than half of what he had harvested that season. educated unemployed are seen as useless in the sense that they can't participate in subsistence agriculture, as it is too low for their status, thus they remain unemployed. ____________________________________________________________________ 23 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank Table 4: Characteristics of Upward and Downward Mobile, and Chronically Poor Individuals Factors Category Mixed Adult Women Adult Youth 1.Upward 1. Hard work 1. Education 1. Hard work Mobility 2. Right ideas 2. Cooperation 2. Entrepreneurship 3. Cooperation 3. Businessman/ 3. Education 4. Good plans and execution Woman 4. Cooperation of planned activities 4. Having capital 5. Perseverance 5. Education 5. Hard work 6. Good health 6. Perseverance and/or 6. Not drunkard 7. Good plans and tolerance execution of planned 7. Having land activities 8. Good Health 8. Not drunkard 2.Downward 1. Laziness/Not working hard 1. Involvement in a 1. Drunkenness Mobility 2. Illness Police case 2. Adultery 3. Lack of education 2. Illness, such as 3. Being robbed 4. Drunkenness and/or taking mental problems and 4. Illness marijuana HIV/AIDS 5. Increased 5. Being bewitched 3. Landlessness demographic load 6. Being robbed 4. Extravagancy 7. Being a thief 5. Being robbed 8. Landlessness 6. Laziness 9. Adultery 7. Drunkenness 8. Adultery 3.Chronically 1. Laziness 1. Lack of education 1.Laziness Poor/Trapped in 2. Landlessness 2. Laziness/not working 2. Individuals living Poverty 3. Idleness hard with HIV/AIDS 4. Lack of good plans 3. Drunkenness 3. Drunkenness 5. Drunkenness and/or taking 4. Long term/chronic and/or taking marijuana illness, for instance, marijuana 6. Uncooperative mental problem 4. Idleness 7. Being a thief 5. Poor planning 5. Uncooperative 8. Desperation 6. Idleness 6. Adulterer 9. Being bewitched 7. Being handicapped 7. Being a thief 10. Illness, such as mental 8. Lack of social problems and HIV/AIDS services and social capital especially for elderly Chronic Poverty ____________________________________________________________________ 24 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank Almost the same factors for downward mobility were also associated with chronic poverty. New factors include lack of social services and social capital especially for elderly, idleness desperation, uncooperative, and being handicapped. Witchcraft was also strongly echoed by the adult group. They argued that one can be a hard worker but the fruits of his/her efforts are hampered by witchcraft. Also, in mining activities there are a lot of myths about how to get the biggest stone. This also featured in the Tanzania Participatory Poverty Assessment (TzPPA) as an impoverishing force under cultural beliefs and practices. Consequently, visiting witchdoctors is a common phenomenon among miners. Elderly were termed to be more vulnerable to chronic poverty because there are no social services and social capital for elderly. Health services are available at a fee and sometimes elderly cannot afford to pay. The village did not have any social arrangements for different vulnerable groups such as elderly, widow and orphans. These groups are left in the hands of the extended family. At Village C chronically poor people were termed to be very uncooperative and they are not member of development groups--Vigombe. However, if they apply they are registered on the condition that they will receive the contributions last. Respondents argued that the type of activities done by different individuals do not determine the path of growth but rather the pace of growth. They agreed that every activity, if well planned and done, it will led to upward mobility. Nevertheless, business related activities were associated with a greater pace of upward mobility compared to agriculture because they are not seasonal11. Unforeseen weather conditions and market asymmetries make agriculture a very risky endeavor. 4.3 Community and Idiosyncratic Shocks Major community shocks were identified through the focus group discussions whereas idiosyncratic shocks were identified from the life histories. Although shocks were found to be common across surveyed villages, the importance of each differs per village. Table 5 summarizes and ranks the major community shocks in Villages C and D. Market Related Shocks In Villages A and C, the three groups of respondents mentioned market access problem to be the most important community shock. As mentioned earlier, coffee market declined after 1997 and the price has been fluctuating since then. Banana market is also a problem. During village transect walk, research team observed some bananas left to rot in the farms. This is especially the case with Matoke, the staple type of banana that is not used in local brewing. Further, respondents mentioned that sometimes maize is used to feed chicken because there is 11 However, sales could be low during low agricultural season because agricultural dependent households will have less to expend. ____________________________________________________________________ 25 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank no market for the same. Market for cotton is also a problem but the situation is improving though private traders visiting the village. Tobacco market is readily available. Table 5: Categorization and Rankling of Shocks Village C Village D Type of shocks Mixed Women Mixed Mixed Women Mixed Adult Adult Youth Adult Adult Youth 1. Market related shocks, for instance 1 2 1 4 - - market for coffee, cotton, and banana 2. Illness related shocks such as 2 1 2 2 2 3 malaria, TB and HIV/AIDS 3. Weather related shocks such as 4 5 3 1 1 2 drought, wind storm, and El Nino 4. Crime and violence related shocks, 5 3 5 5 5 4 for example, theft 5. Mortality related shocks 7 4 4 - - 5 6. Agricultural related shocks such as 3 - 7 3 4 1 crop pests 7. Governance related shocks 6 - 6 6 6 - 8. Behavioral related shocks - - - - 3 6 Weather Related Shocks The most striking shock at Village D was drought. The village is hit intermittently and this results into chronic food insecurity for some households in the study area. El Nino rains and the drought that followed it were major community shocks at Villages B and C. The severity of the rains lies mainly in the malaria epidemic and the deaths resulted thereafter. Hail rain and windy storm that wiped out bananas in some farms were also reported at Village C. Illness-Related Shocks Illness was mentioned as a second important shock for the surveyed villages. Malaria was a problem for all villages surveyed whereas HIV/AIDS was a major health problem in Villages A and B. Other diseases mentioned include gout (especially for women), stomach pain as a result of worms, and TB. Although cases of HIV/AIDS have been reported in Villages C and D, respondents mentioned that it is not a current threat but rather a potential future threat because of the known infections. Agricultural Related Shocks Crop pests especially bean pests have been a persistent problem at Village D from year 2000. Coffee Berry Disease (CBD) was also mentioned at Villages A and B. Availability of agricultural inputs especially fertilizer for food crops such as maize is also a problem for all ____________________________________________________________________ 26 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank the villages surveyed. After the collapse of the cooperative unions, all agricultural inputs and subsidies disappeared with them. Mortality Related Shocks Except for the high mortality rate experienced after the El Nino rains due to increased malaria cases, mortality was ranked the least among the community shocks in Villages C and D. In Village C, the community experienced unprecedented number of deaths among children and elderly after the rains. In Village D, malaria is a persistent/chronic health problem but in 2000 the village experienced high child mortality as a result of malaria outbreak. Villagers were not able to explain the cause of such an outbreak but they claimed that it happened during the dry season. HIV/AIDS was associated with high mortality rate among the productive and reproductive age group in Villages A and B. The village has been devastated by the pandemic as it was among the first villages affected in early 1980s when the first cases of HIV were identified in Tanzania. HIV/AIDS deaths have resulted to decreased labor that results to decreased farm management. Increased number of orphans in the wake of HIV is also apparent. Crime Related Shocks Theft was a problem in the villages studied as mentioned in the group discussions and in the life histories. Major possessions stolen are coffee, livestock especially goat, and the type of banana that is used for local brewing. The study team witnessed two coffee thieves being punished at Village C headquarters at the time of survey. As mentioned earlier, some respondents have been arrested by Police while smuggling coffee to Uganda. This has resulted to confiscation of the consignment of the perpetrator. The downward fall resulted from this act is apparent. In Village D, armed banditry associated with refugees invades homes and shops at night to steal major possessions including livestock. Early this year they invaded the mining sites and disappeared with a lot of possessions. Governance Related Shocks Three examples in relation to governance are cited. One is the problem encountered in establishment of a savings and lending society at Village D. Their Councilor asked the villagers to contribute TShs 1,000 aimed for establishment of the same. The organization has not been established at the time of survey and people had started demanding their money back, but they were not been successful. Villagers perceived this as cheating--one form of corruption. Corrupt acts were also associated with smuggling coffee to Uganda. Respondents mentioned that sometimes the perpetrators are forced to give "something" as a bail. Another problem mentioned is in relation to community participation in community projects such as ____________________________________________________________________ 27 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank construction of classrooms, water project and road rehabilitation. Hamlet leaders are supposed to mobilize villagers for such activities. However, participation from the villagers is sometimes not forthcoming and some of the projects backfired. This is exemplified by Village A well project (see section 4.5 for Village A water project). Behavioral Related Shocks Women group alleged their husbands to be coming back home very late at night and drunk. A case was reported at Village D where some drunken men have dared to visit the village dispensary to harass birth attendants who were busy helping women to deliver. Drunkenness is a problem widespread in Tanzania. In the Poverty Reduction Strategy (PRS) consultative meetings, about one quarter of women expressed concern about laziness and drunkenness among men (URT, 2000). Idiosyncratic Shocks It is worth noting that while mortality and illnesses such as malaria and opportunistic infections resulting from HIV/AIDS are community wide problems, the same shocks hit individuals differently. Consequently, they can as well be categorized as idiosyncratic shocks as echoed in the life histories. Other idiosyncratic shocks as identified in the life histories include injury, disappearance of an important household financial supporter, accusation over witchcraft, family conflicts, acquisition of second wife etc. Mitigation strategies to witchcraft accusations often result in actions that are equally damaging in the welfare of the accused households. Running away from home so as to avoid physical abuse and the threat of death was common in Village A. The community was found to demolish the houses of the accused, and sometimes cutting down all the banana trees in the farms of the accused. One household experienced that problem, fled to the nearby forest, and the members spent about 8 months hiding (see life history A9, Appendix 2). The members could come home during daytime, cook whatever is available, eat and flee to spend the night in the forest. The reason behind such fleeing behavior was that abuses are done at night because even though the community consents to such violence, nobody would like to be seen doing that, as it can result into personal hatred. Political connectedness was associated with positive and negative impacts. The former village Chairman at Village A narrated one negative impact he experienced. He was involved in a series of court cases because a thief was bitten to death in the village during his regime. He uses a significant amount of his money and time attending the court cases. Nevertheless, positive impacts were also cited (see life history A7, Appendix 2). The respondent was able to secure a job because of his political influence. ____________________________________________________________________ 28 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank 4.4 Coping Strategies Hypothetical examples were used to solicit information on coping with a sudden shock requiring a specific amount of money. As mentioned earlier, four hypothetical cases involving raising specific amount of money overnight for an emergency illness, within two months, within two months for a problem that has hit a household together with several others in the village, and overnight for positive event such as sending a child to school were used. Table 6 summarizes and ranks different coping strategies. Selling livestock, especially goats, is an important coping strategy adopted for a sudden shock and even for anticipated shocks. Further, cattle are most of time used as a collateral to solicit a loan from rich people rather than been sold. However, for anticipated shocks or when one has a longer period to act, cattle is sold a priori in the nearby cattle market whereby one gets a better price compared to selling to a rich man in the village. Table 6: Categories and Ranking of Coping Strategies Village C Village D Categories Mixed Mixed Mixed Mixed Women Women Adult Youth Adult Youth 1. Selling food and cash crops, e.g. 2 2 2 - - - coffee/butura, beans etc. 2. Selling a piece of land 3 1 - 3. Selling/mortgaging livestock 1 3 1 - - 4 4.Borrowing 6 4 5 4 1 3 5. Savings/Cash 5 - 4 2 - 1 6. Asking for a help 8 6 6 5 4 - 7. Selling labor/doing small business 7 5 3 3 3 5 such as local brewing 8. Selling/mortgaging assets, e.g. 4 - - 1 2 2 bicycle and furniture 10. Do nothing 9 7 7 6 5 6 Another important coping strategy adopted, especially in the short span, is selling crops especially coffee. Respondents in Village C agreed that coffee traders are available in the village anytime--only that a person with emergency situation does not have room for bargaining. Coffee is also sold as butura. This is a name given to un-harvested coffee, in either form, that is, flowers, unripe cherries, or ripe cherries but still not harvested. Traders participating in buying butura can even buy coffee flowers. Although selling butura is common for coffee, other crops such as cotton, maize and beans could also be sold as butura. However, no readily available markets for food crop buturas. Further, when an individual is in hardship, one can mortgage un-harvested coffee for a specific sum to be paid back in a ____________________________________________________________________ 29 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank specified period. If the person in question is unable to pay back within the agreed period, the creditor harvests the coffee when it gets ripe. Most of the time coffee traders buy butura at a very low price compared to what they later on accrue. However, there are situations although seldom, whereby the trader gets a loss after harvesting coffee bought as butura. Although selling butura is widely practiced, it is considered as a humiliating act, and thus practices by poor households. For Village D, as Table 6 portrays, using available cash in the house is the first strategy adopted by majority of the villagers. The recent developments mentioned earlier have made the villagers affluent and as such having savings in their homestead is common. Selling/mortgaging assets was another important strategy12. Selling goat, maize, beans and sorghum during emergency period is common. Although intermittent, some households sell green cotton, maize and cassava. Borrowing from friends and relatives featured prominently in Village D. This is associated with the fact that there are several sources of money in the village, and thus people have money to dispense as loans. Village A community has responded to mortality shock by forming economic development and funeral groups. Every respondent in the group discussion and individuals who narrated their life histories were members of at least one funeral society. Women join more than one group compared to men. The joining is related to women's caring role in the community: women are required to offer labor during the death moments. In order to earn income to cater for domestic demands and requirements of the funeral societies, women have formed economic development groups such as communal farming and savings and credit groups. Further, in Village A, the respondents agreed unanimously that few households always have readily available cash in the house and thus borrowing is not the best option for a unexpected/swift shock. As such, at times of unexpected shock, households respond by mortgaging a piece of land and other assets such as bicycles. It is worth noting however that, majority of the households do not have assets and they depend on their piece of land "kibanja" to counter any unexpected shock. Whereas mortgaging a piece of land to counter shock is common, selling of land is rare. This is because land ownership in this area is customary based: the land belongs to the clan and clan members must consent before any land is sold. Selling or mortgaging of the land is considered as a last option as it is a destructive coping mechanism that could lead to destitution of the household in question. Coping strategies for the three hypothetical cases (accessing funds to carter for negative shock) were almost similar except that given more time, respondents could opt for other avenues such as selling labor, making local brew or borrowing from multiple friends to raise the money. Selling labor has two aspects: one may borrow and pay back by supplying labor; or selling labor for cash. For a community wide shock, seeking help or borrowing money 12 Note that the mixed adult group lumped livestock and agricultural produce under assets. ____________________________________________________________________ 30 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank from friends and relatives residing in the nearby villages, and selling cattle in the nearby cattle market were mentioned as feasible alternatives. Although "do nothing" is ranked the last by all the groups, it was mentioned by the majority as the option when everybody else in the village is facing the same problem. Respondents expressed the inevitability of exogenous assistance from the government or NGOs operating in the study area to combat such a situation. For funds to cater for positive events, respondents mentioned seeking help from NGOs such as TCRS or village government as the best option. Further, for organized clans, clansmen could solve such problem. For education purposes, Village D was found to have village educational fund. Each household contributes TShs 500 per year for such a purpose. This money is used to support students from impoverished households, and who have been selected to join secondary schools but their parents could not afford to pay the fees amounting to TShs 40,00013 per year. The instituted TShs 40,000 per year has been a setback towards joining secondary school by pupils from poor household as box 4 portrays. The money is also used to support other education related activities such as construction of community secondary school. Another alternative support for a positive event is to request the village government to hold a harambee (this is a name given to collective contributions made upon a request) aimed at solving such a problem. This has been done in instances where a child is selected to join secondary school but the parents could not raise the fees. Another source of funds for educational purposes is The National Education Fund. This Fund has been instrumental in supporting students from most vulnerable households in different parts of Tanzania but the surveyed villages have not heard of such a Fund. Box 4: Access to Education: Deprivation of Children from Poor Households Nyakato, a girl child from a poor household passed the Primary School Leaving Examination (PSLE) in 2002 at Village D primary school. However, her father did not show any enthusiasm in that because he did not have money to pay for the school fees (TShs 40,000 per year) plus other contributions. The village government assisted in paying the fees for the first year on the agreement that her father will try hard to support the subsequent three years of study (Form II - Form IV). She joined secondary school but only finished Form I because her father was not able to pay the fees for the second year and she had to drop out. In year 2003, her brother passed the PSLE but he never joined Form I because his father could not pay. The village government was not in a position to offer a help because it assumed that the boy would just do Form I and drop out just like her sister. In 2004, the father secured some money and sent the daughter to a vocational school in the nearby region. The son was still at home at the time of survey. The pathetic part is that the village government and even the school have not heard of The National Education Fund to support children from most vulnerable households. The Fund is intended to support three children from every Ward. This is the situation in other areas in Tanzania. The Education Sector Public Expenditure Review 2004 indicates that the number of students enrolled in Form I in the past five years is consistently lower than the number selected (URT, 2004c). This means that passing PSLE does not guarantee children from poor households to access secondary education, although pass rate at PSLE is a PRS indicator. 13The need to channel more funds to the National Education Fund and disseminate the information This is just contribution for school fees. Contributions for other items such as uniforms, exercise about the Fund widely is vital. books etc makes the amount quite significant. ____________________________________________________________________ 31 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank In responding to HIV/AIDS shock and poverty situation, Village A community received some external assistance mainly from Partage, a French originated non-governmental organization based in Bukoba. Other organizations that have come into assistance include the World Vision, Bukoba District Rural Development Program (BDRDP), and TASAF. Partage is active in providing some housing materials and food supplies to affected households, that is, households that have lost the adult bread earner. Furthermore, the organization supports school and medical costs for HIV/AIDS orphans and individuals living with HIV/AIDS. In making the lives of orphans sustainable, Partage has been helping the affected households in securing an off-farm income-generating source such as dairy cattle keeping and vocational training for orphans. BDRDP is a development program funded by the Dutch to finance and undertake activities geared towards reduction of poverty in rural areas. In Village A, BDRDP has assisted in the building of two bridges. TASAF is currently assisting in the construction/repair of the road and dispensary in the area. Local people are gaining income from the jobs offered by the TASAF initiatives. Increasingly, villagers are investing in the schooling of their children as a coping strategy for old age. From the life histories, it was reveled that remittances from children residing in the village or working elsewhere contribute significantly to the welfare of the households (Appendix 2, life histories A2, A4, B2 etc). The educated children have also taken up the role of investing in the education of their younger siblings. The respondents expressed differences in coping strategies adopted by different groups of people in the community. They agreed that rich persons have more options than poor persons, including having cash in the house or bank, having money to buy coffee or butura, and having assets to mortgage. They can also borrow money from friends. However, the poor have limited options. They cannot borrow, as most likely they do not have a collateral14, and they have no assets to mortgage. The only option is to sell whatever they have, including the land. Further, in most cases it is the poor who sell butura. Selling land and butura are erosive coping strategies as they push the person in question down the ladder. Another option for the poor people was selling labor. Respondents mentioned the famous phrase in Tanzania "the poor man's capital is his own labor." Asking for a help, and do nothing are other alternatives for the poor. Respondents had different opinions regarding the coping strategies adopted by the youth and elderly. The adult group thought of an elderly as an already impoverished person who cannot cope, and who would most likely ask for help. Moreover, if he/she tries to borrow, she/he will be stigmatized because people will be questioning why he/she did not save when the days 14 There is a high probability of default among the poor if collateral is not asked. ____________________________________________________________________ 32 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank were good! On the contrary, the youth group thought that coping for old people should not be a problem because older people are wise, they have savings and assets, and they receive remittance from their children. It was however cautioned that, "elderly" or "youth" is not a uniform category. There are old people who are rich and those who are poor, and youth who are strong and hard working, but others are idle and drunkards. The study team tried to solicit information on whether households prepare themselves for an expected death, for instance, of a household member who is living with HIV/AIDS, and whether funeral arrangements for a person who died of HIV related opportunistic infections are different from the one died of other causes. Respondents in the focus groups mentioned that no apriori preparations are made for a foreseen death and funeral arrangements are the same irrespective of the cause of death. However, individuals living with HIV/AIDS prepare themselves for death. The preparation is in relation to future survival of their children. The story of one individual living with HIV/AIDS is narrated in Box 5. Box 5: Living Positively with HIV/AIDS Koku, a 54 years primary school teacher was married as a second wife to a relatively prominent civil servant. The prospective husband lied to her that he was not married but latter she found out that she was the second wife. Her husband married a third wife in 1995. Unfortunately, the third wife was infected by HIV/AIDS and died in 1998. Koku has to divorce her husband because the husband was not cooperative when she asked him to go for HIV testing. She went for HIV testing in 1998 and she was found HIV positive. She had redone it in 2001 and 2003 and the same HIV status was established. She has been living positively with HIV/AIDS. From when she was first diagnosed to be HIV positive, she has taken several precautions including: not having sexual intercourse; attending HIV/AIDS seminars, meetings and workshops; using antiretroviral therapy; and checking her diet. Apart from being HIV positive, she is diabetic. She combines both local and hospital drugs to deal with her health problems. What disturbs her most is the fact that she told her divorced husband to go for blood screening and he refused. The man is still promiscuous and he continues to produce children. She has done several things to prepare herself in case she dies: she very regularly reviews her will every time she acquires anything valuable; she has kept a good sum of money in the bank; she has bought another banana plantation for her children, and she has built a house for them. Her biggest worry is how her children will live if she dies abruptly. The youngest child is 13 years of age and he is attending primary school (standard five). The eldest son is twenty and he is his third year of secondary school. The other two who are 16 and 17 years are in their first and second years of their secondary school education, respectively. 4.5 Collective Actions and Community Forms of Organization Information regarding the collective actions in the village and presence of different local forms of organization such as funeral groups was sought from respondents in focus group ____________________________________________________________________ 33 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank discussions. At Ward and Village levels, villagers were found to be participating collectively in activities such as road, school, and dispensary construction, well drilling and construction, and security related activities. In addition, villagers were found to be belonging to different organizations on individual or household basis. Three types of local organizations were identified. These include funeral societies, development oriented organizations, (Vigombe), clan, religious, and NGOs related groups such as groups formed to meet specific requirement for obtaining a loan from a certain NGO. Funeral societies and vigombe are the most popular organizations in the study area. Community Collective Actions A big part of the social sector development activities was found to be externally funded. The local community contributes in kinds, for instance by providing labor for construction, and collecting sand and stones. Involvement of the community in these activities is aimed at making the community own those projects and thus utilize the outcomes sustainably. The gravitational water systems in Villages A and D were rehabilitated under HESAWA project. In Village A, villagers were requested to contribute a total of TShs 250,000 and cement for the project whereby the Bukoba District Council was to contribute about TShs 4 million. The money was to be channeled through HESAWA for topping up. The project had not taken a serious pace by the time of survey because the District Council has not contributed the required sum. Community wells and gravitational water system were also constructed by HESAWA in Villages C and D. Partage offered to reconstruct wells in Village A under the conditions that villagers in different clusters using the same well should collect the recommended type of stones for such activity and contribute construction labor. This was a three years project but only two communities were able to meet the requirements and thus had their wells constructed. It was however not clear why some clusters were reluctant in participating in the project. It was, nevertheless argued that community leaders in those clusters were not aggressive enough in persuading the villagers to participate. The more plausible reason is the laxness of villagers that stems from the socialist mentality whereby the state used to provide all social services free of charge. Thus, community participation in its own development process is still novel in those communities. Another reason was related to societal norms and beliefs. In one of the community well, water level went down after construction and this was attributed to spirits (mizimu). Consequently, the well was demolished. The communities that constructed wells have devised a mechanism to exclude the communities that were reluctant to participate. They also punish community members who refuse to participate in well cleaning. In order to exclude the former, they lock the tape and for those reluctant to work, they have to pay a fine of TShs 500. The money collected is used for minor repairs. ____________________________________________________________________ 34 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank Villagers were also participating in implementing PEDP. In all the villages surveyed, they have contributed stones, sand, and construction labor. Villages C and D have started contributing money for construction of community secondary schools. TASAF has been keen in involving the community in planning and in executing programs aimed for their own development. It involved the community in prioritizing and selecting the most needed community service. Although Village A did not have a dispensary at the time of survey and villagers have been aspiring for the same for quite sometime, villagers ranked road construction highly because road construction offers employment opportunities; good road eases transportation; and it increases accessibility to internal and external markets. It also results to availability of goods and services and establishment of business centers along the road. The construction process was participatory and the villagers who participated were paid a total of TShs 1,000 per day for about 12 days, but it was mandatory to contribute TShs 500 after the 12th day for development purposes. Even though priority was given to widow/widowers and youths, it turned out that majority of the participants in this project were women. Some of the reasons cited include: · The number of women in the village is higher than that of men, and the village has a lot of widows. · Most men died of HIV/AIDS as they used to have multiple partners even after knowing that they were infected--some spreading the virus deliberately. This is exemplified by life history A8 (Appendix 2) · For youth, this avenue was considered low paying as one could earn more than that amount by participating in fishing activities · Women feel responsible for their children, as some men are alcoholic, and they tend to push their children to their mothers if asked for some money. The surveyed communities were found to be very keen in maintaining discipline and piece in the village. Respondents cited cases whereby thieves were beaten to death or punished to work in the community project, for instance, in road construction; demolishing of their property such as a house and crops; and banishment of murders and sorcerers. In some instances, the Local Government Authority (LGA) revoked the punishment for accused but the community took a strong stand against that. One respondent in Village A lamented: ".....if the Local Government Authority wants these individuals to return to this village, then they should find a place for all other villagers to migrate to and leave the killers here...." Having strong rules against delinquents has resulted to reduced incidences of theft from internal, although thieves from other areas and refugees still disturb the villagers. However, ____________________________________________________________________ 35 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank the accused sorcerers have been punished without necessary establishing that they are culprits. Funeral Societies Except for Village D, every respondent interviewed individually, or participated in focus group discussions was a member of at least one funeral group. There are numerous funeral societies with diverse constitutional requirements. Although these groups have been in Village A for quite sometimes, respondents argued that many sprouted in the HIV/AIDS era. Many of them were formed in reaction to increased number of deaths related to HIV/AIDS pandemic. Whereby in the past the clan could deliver requisite services for funerals of the clan members, the clan groups succumbed to pressure of HIV/AIDS deaths, and thus the community realized that HIV/AIDS was a community problem rather than clan, household, or individual problem. Further, insufficient labor power, and decreased farm productivity has resulted to decreased capacity of a single household to carter for funeral meals. While there were about two funeral groups in 1994 in Village A, the number has increased to more than three and eight male and female oriented organizations respectively. For males, Plot is the biggest organization having more than 100 members in every hamlet. It is an organization that has as representative the head of the household. Thus, membership is open for a household with representatives as men only although widows are allowed in by virtue of being the head of their households. If the head of the household is the member of the plot, whenever there is a funeral the wife has to participate accordingly in offering different services at the place of death. Every member is supposed to contribute a bunch of banana, TShs 100, and a bundle of firewood for every funeral occurred in the house of the member. At the time of survey, Village C had a total of 26 registered funeral societies (15 percent increase from 2002), with diverse constitutional requirements. Majority of respondents belonged to one funeral group. In case the group loses an insured member (in most cases a father, mother, husband, wife, father in law, mother in law, or a child) members in different groups contribute financial and non-financial resources such as time and in kinds such food, and firewood according to their constitution. Box 6 describes requirements for a group named Rwenkanja as a case study. More information on specific roles of funeral societies is provided in the life histories (Appendix 2). Despite the fact that each organization has its own constitution, there is an umbrella organization that brings together a number of organizations. For instance, the plots from different hamlets have a joint committee that can rule out matters that seem to go beyond an individual plot, matters such as misunderstandings. The same applies to women's groups. The umbrella organization is used as a "Court Marshall" to solve disputes originating from different organizations. The umbrella organization decides on the fate of the individual who has transgressed, and thus facing the threat of been excluded from the organization. ____________________________________________________________________ 36 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank Nonetheless, if the member does not abide to the constitution, for instance, by failing to contribute for four consecutive funerals (for a plot), the member is automatically excluded. Some funeral societies have fallen apart due to internal conflicts. Example is Abagambakamoi (Abagambakamoi means, literally, "those who speak one thing, that is, solidarity)," in Village A, which was an organization for well to do people. They used to contribute higher fees and each member had to donate one crate of beer and 20 liters of local brew, named rubisi, per funeral. This organization collapsed because members could no longer afford funeral expenses due to economic crisis and increased deaths as a result of HIV/AIDS. The lesson learnt from this is the formation of smaller organizations, demanding small contributions that majority could afford. Better off households could then join more than one group. Contrary to other villages surveyed, there were no funeral groups in Village D. The reason given is that death is still considered as communal problem involving everybody in the village and not only a specific group. In every hamlet there is Health and Disaster Management Committee and it is responsible for organizing decent burial for deceased. All community members contribute to the burial ceremony at their will. Respondents agreed that mortality rate in Village D has gone down. They had high mortality rate in the 1980s/1990s due to measles and epidemics such as meningitis. Box 6: Multiple Roles of a Funeral Group Rwenkanja is a funeral-based organization based in Village C and having 56 members at the time of survey. The group has been divided into two groups Rwenkanja A and Rwenkanja B due to increased recruitment and in order to facilitate management of group's activities. Although the group is funeral based, it also conducts development and wedding related activities. Members participate in communal farming for TShs 1,000 per piece of work. They also cultivate their own crops especially peas for sale and for using during funeral ceremonies. During funeral or wedding ceremony of an insured member of the group, men bring water, and a bundle of firewood whereas women bring one bunch of banana and one kilogram of beans or peas. The group is also in good relation with other groups and its sends a representative to functions of the group they are in good relation with. The organization has bought some household items such as cooking pots, buckets, plates etc and uniforms for use during funeral or wedding ceremonies. The group also offers soft loan to its members facing emergency situation. Entrance fee for the beginners was TShs 2,000 but it was increased to TShs 3,000 for latecomers in order to compensate for the already acquired items. A member who has failed to attend a funeral and cannot provide a convincing reason for such an act is fined TShs 1,000. Members could resign from the group voluntarily, but could also be expelled from the organization if they fail to abide to constitution. An expelled member is compensated based on the value of group's items/assets at that particular time. ____________________________________________________________________ 37 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank Economic Development Organizations Although some funeral societies performed both social and development related activities, there are specific development related groups named Vigombe as noted in Village C. Vigombe spouted from 1997 onwards. The villagers learnt this from individuals who have traveled to other areas of Tanzania and Uganda. The early adopters put the idea into practice and later on majority of villagers join the suite by forming several other Vigombe's with different requirements. Nonetheless, Vigombe are not registered because most of the time they got dissolved just after one round. Majority of the interviewed women belonged to more than one group, the maximum being four. Farming and savings groups were also popular among women in Villages A and C. Development groups were also available in Village D, although not many. These include; · Youth farming groups for money. · Businessmen/women groups for provision of capital to its members. The interest rate of 5 percent is charged for any loan disbursed. · Catholic Church Choir group: The group conducts communal farming for money. It also provides loans to its members at 5 percent interest rate but for non-members the interest rate is 20 percent. · Women development groups have also been established with the motive of accessing loan from government and non-government institutions but none has been successful. Clan, NGOs and Religious Related Organizations Clans have been instrumental in supporting widows and orphans. The clans men make sure that each orphan is provided a home after the death of both parents and that widows have a land to till. In Village A, the Bukoba District Council had a program to assist orphans in the past, but at the time of survey there was no District Council organized program for such endeavor. However, NGOs like Partage have been very active to that effect. Saidia Wazee Tanzania (SAWATA) [meaning help the elderly, Tanzania], an NGO helping older people was also found to have a branch in Village C, but it was not very active. Only 4 members of the group have received assistance from the national office. The religious organization found in all villages surveyed is Wanawake Wakatoliki Tanzania (WAWATA), which is a Roman Catholic based women organization. The organization started as a tool for sensitizing on Catholic values of solidarity. Women in this group are involved in different social and economic activities but these differ by village. In Village A ____________________________________________________________________ 38 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank the group used to participate in income generating activities such as collective farming for earning income. However, due to mismanagement of funds by some leaders, these income- generating activities were halted. At the time of survey, the organization was found to be concentrating on wedding and funeral activities. In case there is a funeral, women in the group contribute kichane cha ndizi (one banana fruit comb), grass, and TShs 100. 4.6 Some Community Characteristics For each KHDS village, community questionnaire was administered in KHDS I and KHDS II. The study team gathered a team of key informants from each village who collectively responded on specific community characteristics. Key informants included: Village Executive Officers (VEOs), Village Chairpersons, influential people such as religions leaders and teachers, farmers, health workers, and leaders of some development and burial societies. Information from the KHDS II community questionnaires was studied and in this section, it is aligned with the qualitative data. Table 7 summarizes some of community characteristics from the community questionnaires (2004). Economic Mobility For the four villages studied, the major economic activity was farming. The second main activity was trading for Villages B and D but it was fishing for Village A and livestock keeping for Village C. Trading was the third important activity for Villages A and C. Except for Village A, key informants indicated that people in their communities are better off in 2004 compared to 1994. The reasons given for such improvement include good weather that resulted to bumper harvests, open market which could be associated with good crop prices at a particular time, better services (water supply and roads), and more economic opportunities. For Village A, the community considers itself in worse off position in 2004 compared to 1994. The reasons given include AIDS epidemic, declining coffee prices, and inflation. Except for Village A, the other three communities agreed that it is easier to get a job now in the village compared to the past 10 years. This is supported by the qualitative data that shows that new economic opportunities have emerged in the study village and villagers have been able to diversify their economic activities. Despite the upward growth, the communities experienced two major shocks. These are low crop prices and drought. In coping with these shocks community members diversified their activities by introducing new businesses and new crops, and selling assets. Some received remittances from family members and friends. Only Village B reported to have Agricultural Extension Officer in the village. For the other villages, the office of Agricultural Extension Officer was located within 6 km, 13 km, and 10 km for Villages A, C, and D respectively from the village center. In most instances, Extension Officers are located at Ward level and they are shared among several villages. This ____________________________________________________________________ 39 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank explains why there were some complaints about the performance of Extension Officers in the focus group discussions. Table 7: Some Community Characteristics from the Community Questionnaires (2004) Variables Village A Village B Village C Village D 1. Major - Farming - Farming - Farming - Farming economic - Fishing - Trade - Trade - Trade activity 2. Economic Downward Upward Upward Upward mobility - Open market 3. Reasons for - AIDS epidemic - Better services - Good weather - Coffee the observed - Declining crop (water supply and - Better services smuggling economic prices roads) (water supply and - Good weather path - Inflation - More economic roads) opportunities Passable Not passable Passable 4. Condition of Passable throughout the throughout the throughout the the road throughout the year year year year 5. Availability of public Available Available None Available transport 6. Sources of Spring, river, Spring, river, lake water during Well without pump Well with pump lake and and rainwater the dry season rainwater 7. Sources of Spring, river, water during Spring, river, lake Well without pump Well with pump lake and the rainy and rainwater rainwater season 8. Availability None None None None of a Bank 9. Distance to the nearest 50 4 100 80 Bank (Km) 10. Job More difficult Easier Easier Easier availability 11. Major - Drought (1995) - Storm (2003) - Drought (1994) - Drought (1998) disasters - Epidemic (1996) - Flood (1998) - Epidemic (1996) Rehabilitation of Construction of 12. Community Construction of Construction of dispensary and classrooms and collective classrooms and classrooms construction of participation in a actions well drilling classrooms water project 13. Major - Malaria - Malaria - Malaria - Malaria community - HIV/AIDS - Tuberculosis - HIV/AIDS - Tuberculosis health - Upper respiratory - Intestinal - Tuberculosis - HIV/AIDS problems infection parasites ____________________________________________________________________ 40 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank Road Infrastructure Except for Village C, roads to the surveyed villages are passable throughout the year. This was also echoed in the qualitative study. Public transport is also available in those three villages. The sloppy terrain of Village C puts it at disadvantage especially during rainy season. Social Services Only a few households had piped water system but different other sources are used in the four villages surveyed. The main sources of water during both dry and rainy season are well without pump (Village A), spring, river, lake and rainwater (Villages C and D), and well with pump (Village B). As pointed out in the focus group discussions, no formal banks were available in the study area. Banks are located between 50-100 km from the village except for Village B that has a nearby bank located at 4 km from the village. Each village surveyed had at least a dispensary, health center or a village health worker located in the village. However, hospitals are located far from the village and transportation is needed to the nearby hospital except for Village B that is located 4 km from a nearby hospital. The range is to the nearest hospital is 25­100 km (25 km, 80 km, and 100 km for Villages A, D and C respectively). The mentioned health services have been in the village before 1994 and no new health service was introduced in these areas in the period under study. Just like in the group discussions, malaria featured as the most important health problem affecting all the villages surveyed. The second problem was HIV/AIDS in Villages A and B and TB in Villages C and D. Intestinal problems featured as the third health problem in Village D and this supports the life history findings--cases of hook and luke worms were reported in the life histories. Collective Actions Community members have participated in collective actions such as construction of classrooms and water projects. This supports the focus group discussions findings. Different forms of organization were reported in the community questionnaire. These include funeral and health insurance groups and income generation groups. Others include faith and women's based groups. The only group found to cut across the village to individual level is faith-based group. Other groups were mostly formed at individual level. The life histories (Appendix 2) present a diverse range of groups of which respondents are members. As with qualitative data, no funeral groups were reported at Village D. ____________________________________________________________________ 41 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank In conclusion, we see that numerical data conformed to focus group data in most cases, that is, the numbers and words are telling the same story. One intrinsic value of qualitative data lies in explaining the numbers. For instance, it is indicated in Village B community questionnaire that new economic opportunities have contributed to economic growth. The focus group data went further by pointing out what these opportunities are, and more so in life histories, individuals' experiences in tapping these opportunities are propounded. Further, the community questionnaires do not indicate new economic opportunities as been instrumental to upward mobility for Villages A, C and D. However, it has been clearly indicated in the focus group discussions and life histories that, new economic opportunities such as gold mining and new crops have emerged. This points to the strength of qualitative survey that allows for open-ended checklists and probing for issues that are not captured by the quantitative research. ____________________________________________________________________ 42 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank 5.0 CONCLUSIONS AND POLICY ISSUES The main objective of this research was to collect qualitative data essential for better understanding of determinants of economic mobility and formulate hypotheses that could be tested using quantitative data and/or identify areas for further research using quantitative data. The study was conducted in four KHDS villages selected from four districts in Kagera. The villages were purposively selected to reflect villages that have been hit by a large negative shock such as HIV/AIDS incidence, positive development such as road construction, and differential access to markets resulted from either isolation or being close to town. The qualitative work, both the life histories and focus group interviews present some interesting patterns. We cannot generalize from these findings because we visited only a small number of villages. Nevertheless, these findings can be used as starting point for more detailed hypotheses that can be investigated quantitatively. Below are some of the key findings and potential policy conclusions from this study. 5.1 Economic Opportunities On Farm Income Generation All the villages surveyed accentuated the importance of agriculture for both income and food. Households relied on their plantation with coffee and tobacco as cash crops and banana, cassava, sorghum, and beans as main food crops. Despite fall in price for major cash crops such as coffee and cotton, farmers were found to be recipients of new high value cash crop such as vanilla and tobacco. Revitalization of cotton was also underway after assurance of market from private traders. However, various factors contribute to the lack of interest in farming activities in the contemporary study area. These include; · Thin market for agricultural produce resulted to inaccessibility of some areas. · Low coffee prices and the monopoly of the cooperative unions (KCU, KDCU, BCU) notwithstanding the fact that prices in Uganda are still reasonable. · Extension Agents did not appear to play a major role in promoting new crops, for instance, vanilla. This has resulted to high uncertainty on adoption of the crop. In order to make agriculture a more attractive endeavor, improvement of rural and district roads and across boarder roads network to enhance access to farm and agricultural markets is of the essence. Commitment of the government to release monthly funds to specifically improve transport link to the North Western and Southern regions of the country should be fulfilled so that planned activities which include management of trunk, regional, district, and ____________________________________________________________________ 43 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank feeder roads are carried on. Further, intensification of agriculture by introducing high value crops is crucial. This could be effected by strengthening the extension services and strategic research, and by institutionalizing participatory methodologies at grassroots level.15 Off-farm Income Generation Activities All villages visited were rural, yet the respondents underscored the importance of off farm income generation activities and a number of activities appear to be picking up. Not only are these activities picking up, but also are considered important means for economic growth. This is the general situation in Tanzania especially after the collapse of markets for major cash crops. It was further found that the degree to which off farm activities are carried out is strongly related to the village's "macro economic environment," in the sense that, once money is available, it is often spent on products produced by village entrepreneurs. The village economic environment is determined by: agriculture; market access and roads; and presence of public work and other works such as mining so that money goes around the village because people are engaged in labor activities. Hence, intensification of off-farm income earning opportunities in line with improved agriculture and market access is necessary for economic growth. Formal Employment and Transfers Formal employment was found to be a factor for economic growth on its own right and as one source of start up capital for farm and off farm income earning activities. In addition to that, remittances were found to be an important source of income for a significant number of households as featured in the life histories. Better off get more in remittances because their children are better educated and have better jobs. Remittances may be so important that poor people that cannot educate their children well are in a poverty trap. The importance of remittances especially for the elderly but also for others underscores the importance of considering the rural areas in conjunction with developments in urban areas where most formal and informal sector jobs are found. Education is therefore a pillar for formal employment and remittances. Accordingly, importance of funding education for children from impoverished families cannot be overemphasized. The need to channel more funds to the National Education Fund and disseminate widely the information regarding the Fund to the rural areas so that children from poor households benefit from it is imperative. Livestock Keeping Livestock keeping especially of goat was found to be very important as precautionary saving. Nonetheless, villagers in Villages A and D showed their eagerness in keeping dairy cattle for 15Example of such participatory methodologies is "shamba darasa (a farm classroom)" where farmers learn by doing. ____________________________________________________________________ 44 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank both manure and milk. Soils in Kagera are old and in the long run, they will not be able to sustain the farming systems. Consequently, facilitation of availability of dairy cattle is a policy option. Examples from districts covered by Heifer project show that the famous project named "kopa ng'ombe, lipa ng'ombe" meaning "borrow a cow, pay back a cow" is very successful. Introduction of dairy cattle go in line with strengthening veterinary investigation centers to carry out effective disease surveillance and early warning system, analyze, and process animal disease data as chicken and goat diseases were reported widely. Community Participation Involving the communities in their own development should be an agenda of all development programs. This involves planning with the beneficiaries and makes them become part of the programs. This enhances management and sustainability of programs. This is also important in changing mindset about what the government and other institutions could offer and what is the role of community in question. The approach by TASAF is the best practice. Who Does What? An interesting pattern can be distinguished with regard to who carries out which type of activity: · Poor are mostly engaged in labor intensive activities (fishing, casual labor, brewing, smuggling, lumbering using traditional tools), and keeping small ruminants such as goats and traditional chicken. · Middle income participates in crafts such as masonry, carpentry, bicycle repair, and trading/transporting but they have limited start up capital. · Rich are engaged in cross boarder trade (sometimes they buy coffee in bulky from the villages for sale in Uganda), own rental houses, pharmacies, crafts, bar, keep dairy cattle and poultry, and they often have formal jobs. So where all are engaged in non-farm income generating activities, the poor only have their labor to offer. This underscores the importance of health for the poor as echoed in the factors for upward mobility. Start-up capital seems an important reason why different households do not engage in different off farm activities. Several idiosyncratic sources such as formal jobs, inheritance and assistance from NGOs are cited. Remarkable is the absence of a savings scheme in this list. The fieldwork suggests that accumulating savings may be very difficult because banks are absent, and it is dangerous to keep too much cash in the house because of fire and theft. Livestock may be the best means of accumulation, yet not in all places do households keep ____________________________________________________________________ 45 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank livestock, for instance, Village A. In other places, like Villages C and D much livestock often gets stolen. Interesting developments in this regard are (i) the start-up of a savings and credit co-operative in Village C and (ii) the few cases where funeral and development societies also provide loans for projects. The obvious policy issue here is facilitation of development of SACCOS in the study area. Very successful stories on how loans from SACCOS and financial intermediaries such as Promotion of Rural Initiative and Development Enterprises (PRIDE) have moved beneficiaries up the ladder have been cited even in remote areas such as Mtwara Rural (ESRF, 2003). 5.2 Downward Pushing Economic Forces Illness Illness was one of the major factors for downward mobility. Although AIDS has been in the study area for a long period of time and it has devastated several households, malaria was mentioned as number one health problem in all the villages surveyed. Unprecedented numbers of deaths resulted from malaria were experienced after the El Nino rains. Untimely illness and deaths can set people off on a downward economic path for several reasons: · Medical expenses incurred during illness reduce household savings and assets. · Illness decreases ones own labor power directly, or indirectly because others have to devote their time to care taking. · Especially for elderly spouses, the death of a husband or his disappearance appears to have irreversible and very negative consequences. · Care taking for orphans puts strain on a household's resources. · Death leads to a loss of remittances and other assistance if the deceased was important in someone's informal assistance network. · Death frequently leads to other problems and conflicts around one's will or inheritance. In various instances these turned out to be major (and expensive) occurrences that end up in court. As a consequence, future research should investigate the quantitative importance of these mechanisms but more importantly, thorough campaigns on the use of mosquito nets (treated and non-treated) are needed in Kagera. It was well established in the focus group discussions that villagers do not use mosquito nets simply because of negligence, rather than low purchasing power. In Village A, a health worker in the group propounded the opportunity ____________________________________________________________________ 46 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank cost of not using a mosquito net, and the respondents were astounded. Campaigns on clean water are also crucial as several cases of stomach problems related to hook and luke worms were reported especially in Village D. Drunkenness Alcohol abuse featured as a structural factor for downward mobility for majority. This is also echoed in the Shinyanga Human Development Report, PRSP, and TzPPA (URT, 1998; URT, 2000; URT, 2004a). Alcohol abuse erodes households' income, and may result to unintended physical and sexual abuses that may further results to HIV infection. The widespread argument behind alcohol abuse is that people in rural areas have nothing to do especially during the agricultural off-season, thus they engage in drinking habits in order to break the boredom. Thus, alcohol abuse is structural in the sense that it is caused by the structure of the economy, not individual behavior. However, research has also shown that people do drink (especially men) even during agricultural season and that is why some villages have set by laws on what time the local brew pubs should be opened, and the owners of the pubs who violate these by law plus the perpetrators are fined. Hence, rural poverty alleviation strategy has to consider intensification of off-farm activities but this study also underscores the importance of having such by laws in place in the study area. On the contrary, local brewing was found to be a very important source of off farm income. This is because of low markets for bananas. The only option for farmers is to convert them to local brew. The importance of improving market access not only for cash crops but also for food crops cant not be overemphasized. Theft Theft or cheating was mentioned surprisingly, a large number of times. Theft appeared to be worse in some areas. For instance it was mentioned much more frequently in the life histories in Village C. In all instances it had pretty devastating consequences, especially when items were stolen from poor people, and the consequences appeared to be long lasting. In Village D, theft is also a big problem especially after establishment of mine activities. Further, theft has also contributed to low adoption of vanilla. In PRSP, personal security is identified under human capabilities, survival and well being. It also featured in the TzPPA where research participants concluded that acts of crime and violence have caused social and material and bodily ill-being. It has also impaired participation in income generating activities as people live in constant fear brought about by rising crime, theft etc. The government has taken various initiatives to improve efficiency and fairness in delivery of legal and judicial services, for instance, recruitment of resident magistrates, primary court magistrates and State Attorneys, and institution of Human Rights and Good Governance Commission in 2002. However, little has been done to increase personal safety of an ordinary person and no ____________________________________________________________________ 47 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank indicators have been set for monitoring system. Thus, improved security would be a good pro-poor strategy. Governance Factors Various causes related to governance were mentioned. At individual levels, examples are very different, but in all cases one should keep in mind that they cause quite some hardship. Examples such as border adjustment, not allowed to establish a fish farm, arrested for smuggling and the bribe associated with it etc are some of examples cited (Life histories A5, B8, C11--Appendix 2). At community level, reluctance in participating in community projects and cheating were mentioned. Another more structural aspect is related to the absence of Extension Workers, leading to unclarity around new, potentially profitable crops like vanilla, unnecessary animal diseases etc. Similarly, though less explicitly stated in the life histories is the limited functioning of village health workers. This can be inferred from the reluctance to use mosquito nets in Villages A, C and D where so many malaria deaths occurred. Finally, in all villages located near the boarder households preferred to sell their coffee to the Uganda (illegally) rather than to the local cooperative unions, as they are expected to (and which have a monopoly). The reasons are that prices in Uganda are much higher and payment is timely. If the liberalization of coffee markets in Kagera had been implemented fully and competition allowed, there would be no reason for farmers to prefer to sell on the fully liberalized, Uganda market. By forcing farmers to engage in illegal activities to capture a reasonable margin for their crop, farmers are disenfranchised of much needed income and run the risk of imprisonment or the loss of a large share of their harvest through bribe or confiscation if caught smuggling. Note that there is little economic rationale to force farmers to sell to the local cooperative union. The example of the episode with very high coffee (and banana) prices (in 1994- 1996/97) showed that this led to much investments in house, additional plots, and durables. It also enhanced the general level of economic activity (masonry and carpentry activities) and thus led to the presence of more off farm business activities. In other words, it makes a lot of sense to pass on coffee prices to the farmer. They know how to best invest the money earned. Other Negative Factors Remarkable factors include suspicion over witchcraft and subsequent punishment, and family conflicts. Suspicion over witchcraft has resulted to hash treatment of the accused and the consequences are devastating. Accusation over witchcraft is associated with decline in material, social and bodily well-being. Family conflicts have resulted to detrimental effects such as prolonged court cases and destruction of assets such as house. ____________________________________________________________________ 48 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank 5.3 Areas for Future Research · Deaths have been associated with downward economic mobility through different mechanisms. Investigating quantitatively the mechanisms through which deaths push some households down the spiral and how others hold up is one research agenda. · The high rate of mortality has lead to many funeral groups. They existed in the past, but the large number and high degree of organization is new. In few instances, the funeral societies also take on other, developmental tasks. An interesting research question is whether more funeral groups will evolve into organizations that promote development. · There are many households in Village C that reported legs and hands problems/gout leading sometimes to amputation. The number of cases is atypical, and given the seriousness of the problem, follow up is needed. · There are remarkable large numbers of divorces and adultery cases. Intra and inter household factors associated with this and the consequences to the parties involved is another research area. · Quantitative analysis of the changes in the role of transfers on household welfare using the two KHDS surveys is important. This will build on the work by Lundeberg and Over (2000) which used only the 1994 KHDS data. · Environmental impact assessment is also needed in the mining sites at Village D. Although mining is the current profitable endeavor, the long-term consequences may be disquieting. ____________________________________________________________________ 49 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank REFERENCES ESRF. (2003). "Supply, Demand and Utilization of Financial Services in Tanzania." Research Report Submitted to The World Bank. Dar e s Salaam: ESRF. Lundberg, M and M. Over. (2000). "Transfers and Household Welfare in Kagera." Draft Paper: Michigan State University and the World Bank. Madey, D. L. (1982). "Some Benefits of Integrating Qualitative and Quantitative Methods in Program Evaluation, with Illustration." Educational Evaluation and Policy Analysis, 4:223- 236. Rao, V. (1998). "Wife-Abuse, its Causes and its Impact on Intra-Household Resource Allocation in Rural Karnataka: A `Participatory Econometric Analysis'." In Gender, Population and Development. Edited by M. Krishnaraj, R. M. Sudarshan and A. Shariff. Delhi: Oxford University Press. Temu, A.E and J. M. Due.2000. "Participatory Appraisal Approaches versus Sample Survey Data Collection: A Case of Smallholder Farmers Well-being Ranking in Njombe District, Tanzania." Journal of African Economies 19(1): 44-62. URT. (1998). "Participatory Poverty Assessment--Shinyanga Region." Dar es Salaam: UNDP. URT. (2000). "Poverty Reduction Strategy Paper." Dar es Salaam: Government Printers. URT. (2002a). "The Economic Survey." Dar es Salaam: President's Office--Planning and Privatization. URT. (2002b). "Household Budget Survey 2000/01." Dar es Salaam: National Bureau of Statistics. URT. (2003a). "Kagera Region Socio-economic Profile." Dar es Salaam and Bukoba: National Bureau of Statistics and Kagera Regional Commissioner's Office. URT. (2003b). "2002 Population and Housing Census." Dar es Salaam: National Bureau of Statistics. URT. (2003c). "Education Sector Public Expenditure Review 2003/04." Dar es Salaam: Ministry of Education and Culture and Ministry of Science, Technology and Higher Education. URT. (2004a). "Vulnerability and Resilience to Poverty in Tanzania: Causes, Consequences and Policy Implications." TzPPA Main Report. Dar es Salaam: Vice President's Office. ____________________________________________________________________ 50 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank URT. (2004b). National AIDS Control Program: HIV/AIDS/STI Surveillance Report ­ Report No 17. Dar es Salaam: Ministry of Heath. ____________________________________________________________________ 51 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank APPENDICES Appendix 1: Kagera Region: Administrative Divisions Appendix 2: Life Histories from KHDS Respondents Appendix 3: Community Economic Mobility Graphs Appendix 4: Individual Economic Mobility Graphs Appendix 5: Village Field Reports ____________________________________________________________________ 52 Rural Income Dynamics in Kagera Region, Tanzania: A Report Prepared for the World Bank Revised Background Document SPATIAL DIMENSIONS OF GROWTH AND POVERTY REDUCTION IN TANZANIA MAINLAND 1. Introduction Observed diversity across and within countries has attracted the attention of development scholars and policy makers for centuries. Understanding the causes of geographically uneven distribution of economic growth and skewed income distribution is vital for at least two reasons. First, it is considered to be the key for unlocking secrets of how to kindle growth of regions still lagging behind and sustain growth in the better-off regions. As Krugman (1991) put it, if we want to understand differences in national growth rates, a good place to start is by examining differences in regional growth. Secondly, extreme economic gaps among different regions are potential flash points for social and political instability1. Consequently, the challenge of advancing the kind of growth that creates benefits throughout society has been and continues to feature prominently on the development agenda in many countries. This chapter attempts to shed some light on the regional dimensions of economic growth in Tanzania, based on economic indicators for 1992-2003 across the country's twenty one regions2. The main objectives are to document and explain cross-regional variations in incomes3, identify the most common barriers to economic growth at the regional level, and suggest regional policy options for improving economic growth performance. 2. Topographic Features Tanzania Mainland is located in East Africa, bordering the Indian Ocean between Kenya and Mozambique (1,424 km coastline). The country also shares borders with the Democratic Republic of Congo (DRC), Zambia, Malawi, Uganda, Rwanda and Burundi. The total land area is about 943,000 km2 while 62,000 km2 is covered by water, including five major lakes namely Victoria, Tanganyika, Nyasa, Rukwa and Eyasi. The country is traversed by seven major river basins which include Rufiji, Wami, Pangani, Ruvu, Ruvuma, Kagera and Malagarasi. Other land marks include mountains notably Mount Kilimanjaro which is the highest point (5,895 m) and Mount Meru (4,566 m), as well as the Great Rift Valley - a branch of which runs through central Tanzania, and another branch through north-west and down to the Usangu plains in the south where the two converge. The climate varies from tropical along the coastal plains, to semi-arid in the central plateau, and temperate highlands located in the north and south. Natural resources include gold, diamonds, tanzanite and other gem stones, nickel, tin, phosphates, iron ore, coal, salt, soda ash, natural gas, hydropower, wildlife, and fishery. Natural hazards include drought, and tsetse flies. 3. Data Sources and Limitations Regional GDP data used here are from the National Income Accounts of Tanzania compiled by the National Bureau of Statistics (NBS). However, it is important to note that the currently available regional GDPs (at current prices) are derived by distributing national GDP into regional GDPs, using proxies such as crop production (compiled by the statistics unit in the Ministry of Agriculture and Food Security), and manufacturing and mining value-added from the regions. Employment is used as the proxy for other sectors. The problem is that the accuracy and coverage of these proxies is rather limited. Consequently, regional growth rates are not reliable for lack of appropriate regional deflators. Although work has now begun to 1 Concentration of economic activities around certain centers is also often associated with other costs such as traffic congestion and pollution. 2 The GDP data used here covers only 20 regions because Manyara region was created only recently (2003) by splitting Arusha region into two. Therefore, the GDP for Arusha covers the two regions. 3Economic activity is unevenly spread even within individual regions. 1 build-up a new regional GDP series, using a combination of the product and income approaches, starting with a few regions (Kilimanjaro, Kagera and Dodoma) and expected to be disaggregated to the district level, the exercise is still in progress. Supplementary data is therefore utilized in this chapter in view of the limitations of available regional GDP. Supporting data are obtained from the relevant ministries, departments and agencies (MDAs), notably the President's Office Planning and Privatization (PO-PP), President's Office Regional Administration and Local Government (PO-RALG), Ministry of Finance (MoF), Ministry of Agriculture and Food Security (MAFS), Ministry of Communications and Transport (MCT), Ministry of Works (MoW), Ministry of Education and Culture, Tanzania Revenue Authority (TRA), Tanzania Investment Centre (TIC) and Bank of Tanzania. 4. Overall Regional Growth Patterns Development literature suggests an expansive list of candidates (economic and non- economic) of why some regions could be excluded from sharing in growth. They encompass, (i) specialization in activities characterized by decreasing returns to scale (costs increase more than proportionately as the volume of production increases), (ii) inefficient use of resources, (iii) quality of governance, institutions and organization of production, (iv) differences in accumulation of knowledge drawn from experience and innovation, (v) productive capacity (vi) state of the financial services, (vii) population growth and migrations, (viii) openness to international trade, (ix) public policy, (x) natural resource endowments, (xi) burden of diseases, (xii) differences in agro-climatic and geographic conditions, (xiii) remoteness from markets and related state of transport and communication infrastructure, as well as (xiv) cultural, political and historical factors. Like many other countries, Tanzania is characterized by highly uneven regional distribution of economic activity and incomes, with Dar es Salaam dominating over all other regions. Regional GDP estimates show that about 52 percent of the annual national GDP for the period 1992-2003 was produced in only six regions, namely Dar es Salaam, Mwanza, Shinyanga, Arusha, Mbeya and Iringa (red-shaded in Map 1) out of the 20 regions. Dar es Salaam region alone has consistently contributed highest to the national GDP (equal to the contribution to national GDP of the bottom six regions combined!). In 2003 the GDP per capita of Dar es Salaam was 2.3 times that of the poorest region (Coast) and almost double the national average. By contrast, the peripheral regions of Coast, Lindi, Kigoma, Mtwara, Mara and Dodoma in the central part (yellow-shaded in Map 1) produce only little. Together, they accounted for a mere average of 18 percent of the annual national GDP over the same period4. The dominancy of Dar es Salaam is attributable to three major factors. First, Dar es Salaam is the defacto seat of Government and therefore has the highest concentration of political power, resources and related support/ancillary activities. Second, the city is the major port and commercial/financial capital and therefore more connected to the global economy. The port is the main conduit of export and import trade not only for Tanzania, but also serves neighboring countries (Uganda, Rwanda, Burundi, Eastern DRC, Zambia and Malawi) that are landlocked. Third, Dar es Salaam has the highest concentration of manufacturing and service industries and only little of traditional, low productivity agriculture. In general, there are many reasons behind the observed spatial inequalities in Tanzania, including economic and non economic factors. Non economic factors relate mainly to historical reasons especially colonial legacy as well as culture. For example, some regions such as Kigoma, Rukwa, and Mtwara were designated by the colonial administration as labor reserves. It was not until after independence that some investments in human capital development (schools, hospitals) were directed to such regions. It is also the case that due to 4Regional GDP data for Tanzania could be overstating regional income disparities partly because production in the regions which contribute little to national GDP is largely for subsistence and is not fully captured in market- based GDP numbers. The purchasing power of the shilling also tends to be higher in the poorer regions. 2 differences in opportunities, some regions such as Kilimanjaro have over the years developed an entrepreneurial culture than others. Regarding economic factors, the overall pattern described above points to the rather obvious fact that regional income levels tend to mirror the nature and concentration of economic activities. Specifically, major contributors to national GDP turn out to be regions which also have the highest concentration of non-agricultural economic activities and services (manufacturing; mining and quarrying; production and distribution of electricity, gas and water; trade; tourism; financial and business services). The main characteristic of these activities is that they are service or processing-oriented and therefore adding value. The annual survey of industrial production and employment statistics show that 2/3 out of the top- six contributors to the national GDP (Dar es Salaam, Mwanza, Arusha, Mbeya) also turn out to be among the top-five in terms of regional shares in overall industrial activity as indicated by the number of industrial establishments, employment and value added (Table a). This is in contrast to poorer regions where over 2/3 of the employed population is found to be earning a living from traditional agriculture, livestock rearing and fishing. Map 1: Main and Least Contributors to GDP by Region Table a: Distribution of Industrial Establishments, Workers & Value Added in 2000 (%) Workers Workers Value Value (2000) (2004) Added Added Region Establishments (2000) (2004) Dodoma 0.2 0.2 0.2 0.6 0.3 Arusha 9.3 6.8 6.6 4.3 7.3 Kilimanjaro 4.4 8.1 7.9 6.4 3.4 Tanga 10.7 10.2 9.5 8.3 9.9 Morogoro 2.9 14.3 11.6 4.3 2.9 Coast 0.2 0.0 0.0 0.0 0.0 3 Dar es Salaam 43.4 25.9 33.5 41.6 59.0 Lindi 1.0 0.1 0.1 0.6 0.1 Mtwara 0.6 0.1 0.1 0.1 0.0 Ruvuma 1.9 2.6 0.0 0.8 0.4 Iringa 2.9 15.4 13.2 4.2 2.8 Mbeya 3.2 3.6 3.4 3.3 6.1 Singida 0.4 0.1 0.1 0.2 0.6 Tabora 1.0 1.2 1.9 2.8 2.1 Rukwa 0.2 0.1 0.1 0.1 0.0 Kigoma 0.4 0.2 0.2 0.3 0.1 Shinyanga 2.9 2.9 2.9 6.0 0.9 Kagera 4.8 1.8 1.8 2.6 1.8 Mwanza 7.2 5.0 6.8 9.2 2.2 Mara 2.7 0.8 0.1 4.4 0.0 Tanzania Mainland 100 100 100 100 VA = VA = 441,482 701,057 Workers = Workers = TShs. TShs. Establishments = 52584,589 89,826 Million Million Source: Economic Survey (2004) and Annual Survey of Industrial Production Vol III (2003) The decomposition of income by source (Table b) also suggests that regions/areas that are performing better in terms of contribution to national GDP and per capita income, are mainly urban, which depend much less on agriculture as a source of income. Table b: Composition of Income Sources by Area (% Share of Total Income) Income Source Dar es Salaam Other Rural Tanzania Urban Mainland Employment 35.4 22.0 6.6 10.5 Self Employment 30.4 29.8 14.2 17.3 Agriculture 2.0 17.4 43.6 37.5 Transfers 7.9 7.7 4.5 5.2 Other 10.9 9.4 4.3 5.4 Source: Household Budget Survey 2000/01 The importance of non-agricultural activities in regional economies is also clearly reflected in many other socio-economic indicators, including the quantity of electricity sold by region, regional distribution of investment projects registered by the Tanzania Investment Centre, and tax regimes of the regions and of local government authorities. Chart 1 indicates that on average, the bulk (46%) of electricity sold in Mainland Tanzania is consumed in Dar es Salaam. Other major consumers of electricity (Arusha, Tanga, Kilimanjaro, Morogoro, Iringa, Mwanza, Mbeya and Shinyanga) are, as already alluded to, the same regions with the highest concentration of industrial activity. By contrast, the poorer regions of Coast, Lindi, Rukwa and Kigoma together consume only 1.5 percent of the total quantity of electricity sold in the Mainland. The regional distribution of projects registered by the Tanzania Investment Centre (TIC) from 1990 to June 2002 indicates that 1,187 (58%) of the total 2063 projects were for Dar es Salaam region. Other regions which attracted more investment projects were Arusha (11%), Mwanza (7%), and Tanga (4%). Most of the investment projects were in manufacturing (43%), agriculture & livestock development (7 %), construction (7%), services (6%) and tourism (5%). Revenue collection records kept by the Tanzania Revenue Authority (TRA) also reveal that in 2003/04 about 85 percent of total tax revenue region-wise came from Dar es Salaam alone. 4 Other key contributors to total tax revenue included Arusha (3.2%), Tanga (2.8%), Mwanza (2.4%), and Kilimanjaro (2.1%). Analogously, revenue statistics recently compiled by the President's Office Regional Administration and Local Government (PO-RALG) for local government authorities (LGAs) on mainland Tanzania show that in 2002 LGAs own revenue sources amounted to Tshs.48.4 billion, out of which Tshs.26.1 billion (equivalent to 54%) were collected by LGAs in Dares salaam, Mwanza, Mbeya, Arusha and Kilimanjaro regions. The above revenue collection pattern partly reflects the fact that all the five regions, border neighboring countries and are therefore major entry points for customs and excisable goods, as well as hubs of trade and commerce. In addition, these regions are also home to most of the manufacturing industries and are correspondingly, the main sources of VAT and other business income taxes. In 2003/04 the four regions accounted for about 82 percent of total income tax collections and 92.5 percent of customs and excise revenue. In a nutshell, the main message that can be drawn from the exposition above is that the more a region has of value- adding type activities, the greater the size and dynamism of its economy. Chart 1: Average Quantity of Electricity Sold 1993-2003 (Percent) Dar-es-Salaam Arusha Tanga Kilimanjaro Morogoro Iringa Mwanza Mbeya Shinyanga Dodoma Tabora Kagera Mara Mtwara Singida Ruvuma Kigoma Rukwa Lindi Coast 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0 50.0 Source: Economic Survey, 2004 Climate and uneven natural resource endowments have also had a strong bearing on economic growth of different regions. This fact is perhaps most borne out by the recent growth of mining, concentrated around Shinyanga, Mwanza, Arusha, and Mbeya regions; and tourism in Arusha, Manyara, Kilimanjaro and Morogoro regions. Similarly, differences in agricultural production (crops grown, volume, farm productivity, and relative unit prices) have played an important role in shaping regional growth pattern in Tanzania. Basic agricultural statistics published by the Ministry of Agriculture and Food Security (MAFS) show, as expected, that there are marked regional differences in the crops cultivated, largely dictated by climatic conditions. For example, cotton production and related activities are found to be concentrated in regions around Lake Victoria. Between 1994/95 and 2000/02 Shinyanga, Mwanza and Mara regions alone produced about 88.6% of Tanzania's total production of cotton annually. Analogously, the major coffee producing regions are Kagera, Kilimanjaro, Mbeya, Ruvuma and Arusha which together produce over 90 percent of Tanzania's total production of coffee. Most of cashew nuts are produced in Mtwara region which singly accounts for 2/3 of the total 5 national cashew nut production. Regarding production of food grains, Mbeya, Iringa, Rukwa, Ruvuma and Arusha regions are the main producers of maize, paddy, and wheat as well as legumes. As for livestock, the District Integrated Agricultural Survey carried out in 1998/99 indicates that Shinyanga, Mwanza, Tabora, Singida, Arusha, Mara and Dodoma regions account for over ¾ of the total number of cattle, goats and sheep in the country. It can therefore be expected that, variations in the combination, quantity produced and unit prices for the various crops have a bearing on respective incomes, which translate into regional differences in regional economic growth. In addition, there are also significant variations in individual crop-yield per hectare across regions, as well as over time within a particular region. In general, the data suggests that regional income levels are extremely low in regions where not only smallholder agriculture is the dominant economic activity, but also that the area under each crop, average farm-size and productivity are very low. For example, most of the regions that contribute least to national GDP, are found to also have very low productivity (and below the national average) for all the major food crops compared to the richer regions (see chart 2 for maize, paddy). There is also some evidence that regions which used to be major producers of traditional export crops, have suffered major falls in real incomes partly as a result of a decline in average farm size/ production (Chart 3) and prices. This development is particularly notable in the case of the coffee-growing Kilimanjaro region, where there has been growing population pressure on agricultural land5. According to the 2002 population census, Kilimanjaro region had the highest population density (104) after Dar es Salaam region (table b). A recent study by Oxfam (2005) indicates that declining international coffee prices have had adverse consequences on smallholder coffee farmers in Kilimanjaro region. They include, farmers exiting from growing coffee or turning to food crops and horticulture, as well as migration to towns and mining centers. Likewise, coffee traders gave up/closed their business. However, for some regions (such as Mwanza, Shinyanga, and Kilimanjaro), the decline in real incomes, due to productivity decline for the major cash crops, was partly mitigated by substitution from coffee to other crops and new opportunities, business/trading in other regions and disposal of assets. In addition, there was growth of other sources of income particularly from new investments in mining, fish processing and related services around Lake Victoria. In this regard, it seems plausible that the coffee producing regions in the southern highlands zone did better than the Northern highlands because the agricultural economy of the former is more diversified, with a wide range of agricultural sources of income (tea, pyrethrum, coffee, maize, potatoes, beans, paddy) compared to the Northern highlands zone which is dominated by coffee and relatively much less of the other crops. Therefore, a key message here is that the composition of production, and responses of the population to new challenges (including changes in relative prices) and opportunities, have had a strong bearing on the relative growth performance of different regions. The observed regional income diversity in Tanzania has also to do, at least in part, with demographics. Data from the 1988 and 2002 population census (Table c) indicate that a number of the regions (for example Kigoma and Dar es Salaam) that lie at the bottom/top of the scale in GDP per capita terms, also have the lowest/highest working age group respectively. Available data also indicate significant inter-regional migration, typically from rural to urban areas and mining centers. Dar es Salaam tops the list of regions that recorded positive net migration, followed by Arusha, Tabora, Mbeya, Morogoro, Rukwa and Shinyanga. It is interesting to note that, with the exception of Kilimanjaro and Iringa regions, the top-six regions where people moving out outnumbered those moving in, also rank among the lowest in per capita GDP terms. These are Coast, Kigoma, Dodoma and Mtwara regions. 5 According to the Kilimanjaro Development Forum (KDF), average land holding in Kilimanjaro region has declined by over 50% during the past 50 years and as a result the land has become less and less productive for the key crops (coffee, bananas), given little or no improvements in land use. 6 Therefore, regions with a relatively smaller and declining workforce due to migration and other demographics are likely to lag behind. Chart 2a: Average Maize YieldPer Hectare FY95-FY01 2500 2000 1500 Kilograms 1000 500 0 ya a a o ra rage a anza gera Arusha nyanga Mt ra st wa Lindi Mbe Iringa Rukw KigomRuvumalimanjar ogoro Ma Tanga Ka Coa Ki Mor Ave Tabor Mw SingidaDodoma Shi Regions Source: Statistical Unit of the Ministry of Agriculture & Food Security Chart 2b: Average Paddy YieldPer Hectare FY95-FY01 3500 3000 2500 2000 1500 Kilograms1000 500 0 o a a a ra ra Kilimanjar ya anza rage wa Lindi gera Ka Ma Coast/D Iringa'Sal aam Mbe Arusha Rukw Tanga Kigom Tabor Mt nyanga Mor ogoroMw RuvumaAve SingidaDodoma Shi Regions Source: Statistical Unit of the Ministry of Agriculture & Food Security Chart 3a: Production of Seed Cotton ('000 Tons) Chart 3b: Coffee Production ('000 Tons) 45 140 40 120 35 100 30 80 Mwanza 25 Kilimanjaro Kagera 60 Shinyanga 20 Mbeya 40 15 20 10 0 19 19 19 19 19 19 20 20 20 5 94/ 95/ 96/ 97/ 98/ 99/ 00/ 01/ 02/ 0 19 19 19 19 19 19 20 20 20 95 96 97 98 99 00 01 02 03 94 95 96 97 98 99 00 01 02 /95 /96 /97 /98 /99 /00 /01 /02 /03 7 Source: MAFS (2005), Basic Data for Agriculture Sector 1995/96-2002/03 Table c: Regional Population Dynamics Population Population Population Population Growth Growth Rate Net Migration Density Population Average H-hold Distribution Rate 78-88 88-02 1988 Census 1988 Density 2002 Size 2002 Region 1988* 2002* Dodoma 5.5 5.1 2.4 2.3 -101,085 30 41 4.5 Arusha 6.0 3.8 3.8 4.0 141,724 20 35 4.5 K'njaro 4.9 4.1 2.1 1.6 -124,383 83 104 4.6 Tanga 5.8 4.9 2.1 1.8 -52,168 48 61 4.6 Morogoro 5.6 5.2 2.6 2.6 30,437 17 25 4.6 Pwani 2.8 2.6 2.1 2.4 -103,912 20 27 4.4 Dar 6.0 7.4 4.8 4.3 500,621 977 1793 4.2 Lindi 2.9 2.4 2 1.4 -49,831 10 12 4.1 Mtwara 3.9 3.4 1.4 1.7 -98,689 53 68 3.8 Ruvuma 3.5 3.3 3.4 2.5 -15,219 12 18 4.8 Iringa 5.3 4.5 2.7 1.5 -120,198 21 26 4.3 Mbeya 6.5 6.2 3.1 2.4 46,999 25 34 4.2 Singida 3.5 3.2 2.5 2.3 -63,880 16 22 5.0 Tabora 4.6 5.1 2.4 3.6 66,370 14 23 5.9 Rukwa 3.1 3.4 4.3 3.6 38,305 10 17 5.1 Kigoma 3.8 5.0 2.8 4.8 -102,923 23 45 6.9 Shinyanga 7.8 8.4 2.9 3.3 6,763 35 55 6.3 Kagera 6.0 6.1 2.7 3.1 -5,980 47 72 5.2 Mwanza 8.3 8.8 2.6 3.2 -33,504 96 150 5.9 Mara 4.3 4.1 2.9 2.5 -39,878 50 70 5.5 Manyara N/A 3.8 13 23 5.2 Tanzania Mainland 100.0 100.0 3 3.1 26 38 4.9 Regional growth pattern in Tanzania also reflects changes in government policies particularly related to fiscal and trade regimes. A recent study using evidence from the 2000/01 HBS (Fan et.al., 2005) indicates that there is an opportunity in Tanzania to improve on the growth and poverty impacts of public expenditure through better regional targeting. Specifically, investment in rural roads is found to have a larger impact on per capita incomes in the western, central and southern regions of Tanzania and much less elsewhere. A ranking of average actual regional expenditure (Votes 70-95) for 1995/96 ­ 2003/04, indicates that the top-five regions (Kilimanjaro, Mwanza, Mbeya, Arusha and Tanga) together received about 32.6 percent of annual total recurrent expenditure for regions as compared to an average of 16.8 percent received by the bottom-five regions (Lindi, Rukwa, Coast, Singida, Mtwara and Kigoma). This is largely explained by the fact that the former have relatively more schools and health facilities compared to other regions, which translates into higher expenditure on personal emoluments (PE) for teachers and health workers, as well as for other charges (OC). It is interesting that regions which receive the largest share of recurrent expenditure also rank above average in terms of contribution to national GDP. With regard to development expenditure by region, the regularity is less clear. Regions that on average received a higher share of development expenditure were Kagera, Mara, Mwanza, Iringa and Morogoro, while the regions that got the smallest share of regional development expenditure, included Dar es Salaam, Shinyanga, Kilimanjaro, Tabora and Lindi. However, what is clear is that in general, the regions that currently contribute least to national GDP (Coast, Lindi, Kigoma, Mtwara, Dodoma) have also received relatively little public resources for both recurrent and capital expenditure6. As such, it would be more efficient to allocate investment expenditure by region based on estimated rates of returns. With regard to trade policy, it could be argued that during 6 This analysis gives only a partial picture since a significant amount of recurrent and development expenditure in the regions goes through ministerial votes. Unfortunately the budget books do not permit disaggregation of ministerial votes by region. 8 the era of the control regime (1967-86), Dar es Salaam (and a few urban centres) as the headquarters of most of the parastatals, did benefit more from parastatal sector operations compared to peripheral regions with smaller branches. Restrictions also most likely forced private firms to locate in Dar es Salaam where they could easily manoeuvre with the controls. By contrast, trade liberalization/removal of trade monopolies did create a window of opportunity for growth of other regions particularly in the southern highlands. This is partly supported by re-emergence of active private sector in crop/food grain marketing and distribution (e.g. cashew nuts in the coastal regions; as well as new private investment in the tea sector, and stable production of grains, which are dominant in the Southern highlands and much less prone to decline in world market prices compared to coffee ­the main export crop of the northern zone. The pattern of economic growth in Tanzania also partly reflects differences in the level of human capital development. The average level of education and general skills available in a region, are critically important because they are fundamental to private sector development. In particular, the level of human capital development dictates the capacity of individual regions to learn and adopt better land use systems; new farming practices, and introduce new high value crops or venture into other business opportunities as they emerge. The general level of education is also paramount to the extent that it prescribes the quality of leadership in a region down to the village level. Indeed, while it is rather common to find village and district council chairpersons who are retired senior civil servants (Permanent Secretaries, teachers etc.) in the relatively well to do regions, this is a quite rare occurrence in the poorer regions. It is also interesting to note that the formation of effective and operational civic development forums has begun taking root in regions with a better human capital base (Kilimanjaro, Mbeya, Iringa, and Ruvuma). These development forums are aimed at building consensus among stakeholders of a particular region on binding constraints and articulation of a shared strategy for faster progress. Similarly, the current drive in some regions to form community banks to deal with the problem of lack of financial capital is seen to be much stronger in regions that have, among others, a better education base (Dar es Salaam, Mbeya, and Kilimanjaro). Basic education statistics from the Ministry of Education and Culture and indicators drawn from the Poverty and Human Development Report 2002 (Table d) largely conform to the regional variations in incomes. Overall, Dar es Salaam, Mwanza, Kilimanjaro, Mbeya and Ruvuma regions still rank among the top-five in terms of human capital development while, Lindi, Mtwara, Kigoma, Dodoma and Coast regions generally lie at the bottom of the spectrum. Table d: Selected Indices of Human Capital Development by Region Mean Life Expectancy at birth Monthly (years) Consumption 1988 No. of per capita Adult literacy rate Total NER Secondary (`000 Tshs) REGION (% age 15 and above) 2004 Schools 2004 2000 Dar es Salaam 91 93.1 87 21.9 50 Arusha 78 91.9 107 10.3 57 Rukwa 68 87.9 33 6.7 45 Ruvuma 84 99.3 48 9.6 49 Iringa 81 99.1 82 11.2 45 Shinyanga 55 86.3 51 8.0 50 Mwanza 65 99.5 75 8.1 48 Singida 71 85.0 35 6.9 55 Mbeya 79 99.3 104 12.6 47 Tabora 65 68.2 41 10.4 53 Morogoro 72 81.9 58 10.0 46 Lindi 58 84.1 19 9.5 47 Coast 61 94.5 40 10.5 48 Mtwara 68 94.2 41 12.4 46 Mara 76 100.0 59 8.0 47 Kilimanjaro 85 100.0 160 11.2 59 Tanga 67 97.9 83 9.3 49 9 Dodoma 66 76.3 50 8.5 46 Kigoma 71 77.2 38 7.3 48 Kagera 64 86.8 70 9.0 45 Sources: Poverty & Human Development Report (2002), Ministry of Education and Culture (2004) 5. Emerging Regional Income Trends Although Tanzania's overall real GDP growth has been rising steadily from 3.6% in 1995 to 6.2% in 2002 and slightly down to 5.6% in 2003, the regional per capita incomes have tended to vary significantly over time. This is partly revealed by gaps between the top-five and bottom-five regions in terms of average GDP per capita between 1992 and 2003 (table e), coefficients of variation of GDP per capita for the twenty regions of Mainland Tanzania, and nominal GDP growth rates. Significant regional income growth differences are also shown by variations in factors like agricultural production and productivity, manufacturing value-added, industrial employment, number of commercial vehicles operators, and new investments registered, all of which have an important bearing on income growth. Table e: Gaps Between the Top-Five and Bottom-Five Regions in Average GDP Per Capita 1992-95 1996-99 2000-03 Top-five 107,274 222,230 332,315 Bottom-five 51,433 109,559 177,468 National Average 74,043 159,180 247,030 Gap 55,841 112,671 154,848 Source: Staff computations In the quest for shared growth, some stakeholders in Tanzania have argued for special considerations to be given to the seemingly disadvantaged regions so as to narrow down regional differences. However, traditional growth theory suggests that as long as there are no barriers hindering mobility of goods and factors of production, per capita incomes across regions tend to converge, as poorer regions grow faster until they catch-up with the richer ones. One simple indicator of the widening or narrowing regional income gaps is the coefficients of variation (CV) derived as the standard deviation as a percentage of the mean. The CV constructed for the twenty regions of Tanzania for the period 1980-2003 (Fig 1) suggest that, income differential relative to the national average has declined over the last two decades. Indeed, while on average the per capita income of the top-five regions doubled between 1992 and 2003, that of the bottom-five regions increased slightly faster (2.5 times) over the same period. Average regional per capita GDP ranking for 1992-95 and 1996-99 show that Dar es Salaam, Arusha, Rukwa, Ruvuma and Iringa regions maintained their position as the richest-five in per capita GDP terms, while Kagera, Kigoma, Dodoma, Tanga and Kilimanjaro show-up at the bottom. However, average regional GDP per capita ranking for 2000-03 reveals an interesting development whereby Mtwara and Mwanza regions seem to have gained to become part of the top-five, while Rukwa and Ruvuma regions dropped out. There are a number of possible explanations for the apparent convergence of regional per capita incomes in Tanzania. They include rural-urban migration particularly into Dar es Salaam, remittances, and emergence of other regional growth poles. Migration facilitates convergence to the extent that it reduces the spread in real wages between urban and rural areas. Data on regional shares of new investments registered by TIC does indicate that part of the story has to do with some regions such as Mwanza, Arusha and Shinyanga having attracted more new investment particularly in the areas of mining, fish processing, transportation, tourism and related services, and manufacturing. This is well corroborated by changes in regional shares in income tax revenue between the late 1990s and early 2000s (table f). 10 Table f: Regional Shares in Total Annual Income Tax Revenue (Percent) REGION 1996/97 ­ 1999/00 2000/01 ­ 2003/04 Dar es salaam 78.5 67.7 Arusha 5.3 7.3 Mwanza 3.0 6.0 Shinyanga 0.9 3.9 Morogoro 2.1 3.3 Tanga 2.3 3.1 Kilimanjaro 2.5 2.9 Other regions 5.4 5.8 Source: Tanzania Revenue Authority In the case of Mtwara region, the economic vibrancy seems to have been triggered mainly by new investments in the cashew nut industry, supported by improved private sector participation in the supply of needed inputs, as well as purchasing and marketing of raw cashew nuts for export to the Indian market. Cashew nut exports more than doubled from about 47,000 tons in the mid 1990s to 121,000 tons in 2000. By contrast, the recent decline in average GDP per capita in Ruvuma and Rukwa most probably had to do with the decline in agricultural production, especially of maize, due to extended drought in 2002/03. Fig. 1: Inter-regional Output Disparities in Tanzania 1980-2003 (Coefficients of Variation of GDP per Capita) 1.00 0.90 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 1975 1980 1985 1990 1995 2000 2005 Source: Tanzania CEM 2000 and Staff computations 11 Ranking of Average GDP Per Capita By Region Ranking of Average GDP Per Capita By Region 1996-99 2000-03 Kagera Kigoma Dodoma Dodoma Tabora Kilima nja ro Coast Ta nga Kilimanjaro Ma ra Singida Coa st Tanga Lindi Morogoro Morogoro Mbe ya Lindi Ta bora Mara S ingida Region Mbeya Mwa nza Shinyanga S hinya nga Ruvuma Ta nza nia Ave ra ge Tanzania Average Mtwa ra Rukwa Ruvuma Mwanza Iringa Arusha Rukwa Iringa Arusha Mtwara Da r e s S a la a m Dar es Salaam 0 100000 200000 300000 400000 500000 0 100000 200000 300000 400000 500000 600000 Tshs Tshs Source: National Income Accounts The composition of income sources derived from the HBS 200/01 also indicate that transfers could explain, at least in part, the regional variations in income growth seen in Tanzania. Overall direct transfers (including remittances) for Tanzania Mainland are estimated at 5.2% of total income. However, transfers are higher for urban areas (about 7.8%) compared to rural areas (4.5%). Income growth can therefore be expected to be higher in regions where anecdotal evidence points to large transfers (e.g. Kilimanjaro) than those regions where transfers are negligible . Table g: Regional GDP Per Capita Level Ranking REGION 1992-95 19 96-99 2000-03 Dar es Salaam 1 1 1 Arusha 2 2 4 Rukwa 3 3 6 Ruvuma 4 5 7 Iringa 5 4 3 Shinyanga 6 7 8 Mwanza 7 8 5 Singida 8 9 14 Mbeya 9 11 9 Tabora 10 10 17 Morogoro 11 12 12 Lindi 12 13 11 Coast 13 14 16 Mtwara 14 6 2 Mara 15 15 10 Kilimanjaro 16 17 15 Tanga 17 16 13 Dodoma 18 18 18 Kigoma 19 19 20 Kagera 20 20 19 Source: Staff computations 12 The apparent convergence of regional per capita incomes not withstanding, the case for region-specific interventions to stir growth and development remains. As already indicated, different regions of Tanzania have shared very unequally in the overall positive economic growth rates recorded in the latter half of the 1990s, translating directly into regional differences in poverty status. This is partly captured by trends in regional GDP per capita which indicate that during the 1992-99, only five regions (Dar es Salaam, Arusha, Rukwa, Ruvuma, and Iringa) out of twenty regions of Tanzania Mainland recorded GDP per capita above the national average. Unequal share of regions in overall growth and welfare is also reflected in the poverty headcount ratios from the HBS 1991/92 and 2000/01 which indicate that regional differences in poverty status are such that poverty is more severe in rural Tanzania than in urban areas (Table h). Given that the national average GDP per capita of Tshs.247,030 for 2000-2003 (equivalent to about US$225) remains very low, dependence on market forces per se, some of which (like migration flows) are rather weak, to address substantial regional disparities is certainly inadequate. Deliberate interventions are therefore needed particularly by promoting and supporting regional private sector development initiatives, including ensuring a properly functioning regulatory framework. In the case of the poorer regions, initial focus will have to be put on investment in basic infrastructure (transport, communications, power and water) and related maintenance. Table h: Regional Poverty Status Food Basic Needs 1991/02 2000/01 1991/02 2000/01 Dar es Salaam 13.6 7.5 28.1 17.6 Other Urban 15.0 13.2 28.7 25.8 Rural 23.1 20.4 40.8 38.7 Total 21.6 18.7 38.6 35.7 Source: Household Budget Survey 1991/92 and 2000/01 6. Major Constraints to Regional Growth: The foregoing exposition of regional development patterns in Tanzania suggests that regional variations in incomes are related to at least five factors. These include, (i) the size of the most dynamic sector (i.e. non-agricultural production), (ii) Agriculture production and productivity, (iii) level of human capital development, (iv) public policy, and (v) demographics. In the light of the above, an important question therefore is what are the most formidable obstacles to growth in some of the regions? Two broad categories of constraints can be identified in this respect. One category relates to cross-cutting constraints, which include poor economic infrastructure, lack of innovation and use of rudimentary technology in production, poor financial services, and institutional and policy related constraints. The second broad category constitutes region-specific constraints to growth. Insights obtained from visiting some regions and concerns expressed in region-specific investors forums7, indicate that poor economic infrastructure (particularly transportation, and unreliability of power and water supply) is perhaps the most binding constraint to economic growth in most regions. A poorly maintained road network, and sporadic power and supply are perceived to serious obstacles to attracting private investment, business and entrepreneurial development. It is also seen to impede access to information, knowledge and markets (in other regions of Tanzania and in neighboring countries). An examination of the road transport infrastructure system of Tanzania (Map 2) compared to the distribution of new investment registered by the Tanzania Investment Centre (TIC) and regional shares of national GDP readily support to this view. Regions that contribute most to national GDP not only have higher road density (measured by the ratio of road network to the region's land 7Regional investors forums have been held for Kilimanjaro, Mtwara Development corridor, Mwanza, Kigoma (August 2002) and Tanga. 13 area) and commercial vehicles (table i) but also managed to attract more investment. By contrast, remoteness (measured by distance of the regional headquarters from the main market of Dar es Salaam), negligible road network, poor power supply and telecommunications tend to draw investment away from the poorer regions and thereby exacerbate the observed regional income disparities. The regional distribution of other infrastructure especially power, water supply and telecommunications is seen to be concentrated largely in Dar es Salaam, and other regions such as Arusha and Mwanza but are virtually absent in the southern (Lindi, Mtwara) and western regions (Rukwa, Kigoma). Table i: Regional Contribution to GDP, Distribution of New Investments and Road Network Projects Road Distance Commercial Vehicle Contribution Registered by Road from DSM Road Network Operators Region to GDP (%) TIC 1990-2002 Density (km) (km) 1999 Arusha 7.7 218 9.5 647 7846 1177 Coast 2.0 48 9.7 40 3151 1027 Dar es Salaam 16.9 1187 36.2 0 504 15024 Dodoma 3.5 16 10.0 479 4111 658 Iringa 5.6 36 11.1 501 6317 742 Kagera 3.8 13 13.8 1425 3926 1084 Kigoma 2.5 9 5.2 1442 1927 201 Kilimanjaro 4.1 61 21.1 562 2813 3030 Lindi 2.1 16 5.3 459 3521 Mara 3.3 33 10.8 1369 2110 614 Mbeya 5.8 39 6.9 851 4182 736 Morogoro 4.8 65 5.0 196 3519 275 Mtwara 3.8 14 23.0 565 3844 68 Mwanza 9.7 144 19.7 1164 3861 1538 Rukwa 3.2 3 5.8 1186 3986 208 Ruvuma 3.2 8 6.0 992 3787 844 Shinyanga 7.8 45 8.8 1001 4451 408 Singida 2.6 6 5.4 709 2666 313 Tabora 3.4 12 6.7 1039 5089 129 Tanga 4.4 90 14.9 354 3997 899 Sources: Ministry of Communications & Transport (2002), and TIC The second major constraint is lack of innovation and continued use of rudimentary technology in production (in farming, livestock keeping, fishing and artisanal mining). 14 Production tools used by the majority are prehistoric, with only limited use of animal power in agriculture and traditional irrigation. Introduction of new value crops (like vanilla, spices, etc.) has not taken place on a large scale except in a few regions around Arusha, Kilimanjaro and Iringa regions. Average household farm sizes are also very small (0.5 ­ 2 hectares) in most of the poorer regions. The third key constraint is poor financial services especially lack of access to credit for farmers and SMEs in most regions ­ Typically, the poorer regions are found to have very few banks. Even where credit facilities are available from the few banks, lack of collateral and bankable projects stand out as major bottlenecks to accessing financial capital. As a result, most people who invest in productive activities (agriculture, real estate, SMEs etc.) use only their meager cash savings. Furthermore, only a small proportion of regional savings mobilized by the banking system are actually utilized within that region for investment purposes. Most SMEs and individuals are not well organized and have great difficulties coming-up with good business plans. Fourth are institutional and policy-related constraints. In some regions, an unwieldy number of local government taxes and fees continue to be a big bother to business activity. The rationalization of local government taxes and abolition of nuisance taxes by central government has apparently not been fully implemented at local level. Apparently, many people in the regions are not aware of the tax reforms and local authorities continue to collect a large variety of taxes and levies. For example, in 2000 Newala District Council imposed as many as eleven different kinds of levies on cashew farmers, ranging from development levy to crop levy, primary society fees, district development fund, village levy, water fund, district ceremonies fund, self reliance fund, input purchase fund, health fund and regional ceremonies fund. Other institutional related constraints have simply to do with inefficient regulation. In a number of regions prices offered to the farmers turn out to be extremely low and often dictated by middlemen-traders partly because the latter tend to take advantage of lack of alternative markets and poor storage facilities to buy crops on the basis of arbitrary units (e.g. containers/bowls, tins etc.) to their advantage instead of using standard weigh-scales to measure the agricultural produce offered for sale by the poor peasants. In the case of coffee, tea, cotton and cashew nuts, although in principle crop boards are supposed to regulate the respective industry for quality control, and supervision of crop purchasing to protect farmers against unruly trading practices, evidence from various regions (Mtwara, Mbeya) indicate that this regulatory function is only weakly performed by crop boards. Finally, there are region-specific constraints to growth. For example, Kigoma region faces unique population growth dynamics, including severe refugee pressures (the official number of refugees is estimated at 350,000 - 400,000) resulting from civil strife in Burundi and DRC. Apart from destruction of the environmental, the volume of transport and trade has actually declined since the onset of the conflict in DRC and Burundi, mostly on account of increased insecurity. Kigoma region also registered the highest inter-census population growth rate of 4.8 for 1988-2002, thereby dashing any prospects of early demographic transition to lower dependency ratios. Migration of the youth to urban centres is apparently denying the region significant workforce. Another example of a region-specific constraint is the high prevalence of tsetse flies, a phenomenon which has severely held back development of the livestock sub- sector in Kigoma and Rukwa regions. It is estimated that as much as 70 percent of Kigoma region's land area is tsetse fly infested. As a result, currently only 0.3 million hectares of land is used for grazing out of about 1.5 million hectares of land which is suitable for grazing livestock. Other region-specific constraints include land shortage particularly in Kilimanjaro, remoteness for regions located in the periphery (Kigoma, Rukwa, Kagera, Mara and Lindi), and shortage of reliable water sources particularly for the regions located in the central plateau (Dodoma, Singida). 15 7. Some Tentative Implications This chapter has highlighted significant income disparities among the twenty one administrative regions of Tanzania. It has also pointed to the most formidable obstacles to growth, some of which are region-specific and others that cut across. It is also quite apparent that the regions also have different opportunities; some partly dictated by location and climate, natural resource endowments and even by economic history. The difficult questions then become: What does regional differentiation imply going forward? What can individual regions do to ignite growth? What are the roles of different players (especially government and private sector)? Need to evolve region-specific interventions based on a geographically differentiated growth strategy: The marked variations in economic activities and incomes across the administrative regions of Tanzania suggest that there is perhaps, need to evolve region- specific interventions that are based on geographically differentiated growth strategies. However, regions that are located in proximity and judged to be held back by a common set of constraints to growth (e.g. low level of human capital development, small market etc.), these could team up to develop a joint or zonal growth strategy. Such regional/zonal action plans should in principle be built around economic potentials and opportunities of each region and focus on tackling the most critical constraints hindering full utilization of those potentials. Expressed differently, the strategies should target at promoting regional specialization based on an individual region's resource base and entrepreneurial skills. For example, regions/zones whose people have a long history of ingenuity in specific productive activities and trades, such as production of particular crops, commerce, livestock keeping, fishery, handicrafts et cetera, the private sector in those regions should be nurtured and encouraged to concentrate on how to do things differently in these areas, focusing on value-addition, and exploitation of available and emerging opportunities. Since traditional agriculture turns out to be the least dynamic, more support and facilitation to private sector need to be directed at agro- processing, packaging, and storage to tap intra/inter-regional markets and those in neighboring countries. Similarly, individual regions need to develop strategies and incentives to attract capital, starting with local investment from those who have roots in that region and beyond. Improving agricultural production and productivity across the regions of Tanzania is imperative: In addition to region-specific interventions, a deliberate effort to improve agricultural production and productivity across the regions of Tanzania is imperative for growth. This will entail doing more to promote and instill the discipline of adhering to basic agro-practices on crop husbandry, use of quality seeds, and use of fertilizers and so on. In a number of regions, increasing average farm sizes alone can make a big difference on total production and incomes. The whole issue of technology and organization of production, dissemination of information on markets, and popularization of new innovations need to be given a big push. Key roles for private sector and government: As already alluded to, the emergence of a strong private sector in each region is central to any strategy to deliver shared-growth. Therefore nurturing SMEs in the regions need to be given top priority. A geographically differentiated growth strategy however, does also require government to continue playing midfield. Four areas are particularly important for government in this respect. These include, expenditure on provision of physical infrastructure, enforcing regulations, promotion of public-private partnership, and assisting regions in dealing with constraints that are way beyond an individual region's ability such as the burden of refugees and diseases (HIV/AIDS, epidemics, tsetse flies etc.) or where interventions have a bearing on other regions, such as encouraging people to move from areas of severe population pressure on land to other sparsely populated regions. 16 URBAN RURAL DYNAMICS IN TANZANIA, THROUGH INFORMAL REDISTRIBUTION MECHANISMS Draft Final report for the World Bank Meine Pieter van Dijk, UNESCO-IHE Institute for Water Education, POBox 3015 2601 DA Delft the Netherlands, mail m.vandijk@unesco- ihe.org tel 0031152151779 fax 0031152122921 Version September 30, 2006 Table of contents Abbreviations.....................................................................................................................iii Tables.................................................................................................................................. v Boxes.................................................................................................................................. vi Figures ............................................................................................................................... vi Executive Summary..........................................................................................................vii 1 Introduction...................................................................................................................... 1 2 The context of economic restructuring and urban and regional growth.......................... 1 2.1 Introduction............................................................................................................... 1 2.2 Economic success of Tanzania ................................................................................. 1 2.3 How does growth affect the poor?............................................................................ 3 2.4 Explanations for the spread of growth in Tanzania .................................................. 5 2.5 The Tanzanian tiger .................................................................................................. 6 2.6 No success, just increased prices of commodities and donor darling....................... 7 2.7 Conclusions............................................................................................................... 9 3 Decentralization and its impact on urban poverty in Tanzania........................................ 9 3. 1 Introduction: what to expect from decentralization? ............................................... 9 3.2. The theoretical framework..................................................................................... 10 3.3. The history of decentralization in Tanzania........................................................... 12 3.4. Description of the roles of different levels of government.................................... 13 3.5. Local Government Reform Programs.................................................................... 15 3.6 The evidence: what actually happens in Tanzania.................................................. 18 3.7. Testing the different theories about the effects of decentralization....................... 21 3.8 Conclusions............................................................................................................. 23 4. Infrastructural bottlenecks for urban economic development?..................................... 24 4.1 Introduction............................................................................................................. 24 4.2 An overview of the infrastructure........................................................................... 24 4.3 Conclusions............................................................................................................. 31 5 Urban rural dynamics in Tanzania................................................................................. 31 5.1 Introduction............................................................................................................. 31 5.3 Why call it informal flows? .................................................................................... 34 5.4 Why study rural urban linkages? ............................................................................ 35 5.5 Why do we want to study rural-urban linkages differently?................................... 36 5.6 Relevant questions in the Tanzanian case............................................................... 36 5.7 More evidence of positive developments in the rural areas.................................... 39 5.8 Conclusions and recommendations......................................................................... 42 6. The role of the informal sector to spread development beyond Dar es Salaam ........... 43 i 6.1 Introduction............................................................................................................. 43 6.2. Global cities and peripheral nodes......................................................................... 44 6.3 The context of economic restructuring and urban and regional growth................. 45 6.4 Role of urban informal sector in spreading development and alleviating poverty. 45 6.5 Rural informal sector activities............................................................................... 47 6.6 Different activities in the urban informal sector..................................................... 48 6.7 Mechanisms in the urban informal sector: flows of people.................................... 49 6.8 Informal flows of money ........................................................................................ 50 6.9 Flows of goods and services ................................................................................... 50 6.10 Promoting on the role of the informal sector........................................................ 52 6.11 Current policies of the government for the informal sector and SMEs................ 52 6.12 Formalization process........................................................................................... 53 6.13 Credit or loans....................................................................................................... 54 6.14 Export promotion.................................................................................................. 55 6.15 Other SME policies............................................................................................... 55 6.16 Conclusions on promotion policies....................................................................... 56 6.17 Conclusions on global city or local node.............................................................. 57 7 Dar es Salaam, a dynamic capital with three local governments................................... 57 7.1 Introduction............................................................................................................. 57 7.2 Activities at different levels of government............................................................ 58 7.3 Sources of urban growth: issues in the private sector............................................. 63 7.4 The economy of Dar es Salaam .............................................................................. 68 7.5 Sources of revenue of Dar es Salaam ..................................................................... 71 7.6 Conclusions on Dar es Salaam................................................................................ 72 8. Regional growth in Tanzania: local-global interactions in a low-income country....... 74 8.1 Introduction............................................................................................................. 74 8.2 Research questions.................................................................................................. 74 8.3 Theoretical framework for regional development studies ...................................... 75 8.4 A dynamic and a stagnating region......................................................................... 77 8.5 Kilimanjaro region.................................................................................................. 79 8.6 The Arusha region................................................................................................... 80 8.7 The Kilimanjaro and Arusha regions compared..................................................... 80 8.8 Conclusions on booming and declining regions and cities in Tanzania................. 85 8.9 Conclusions and policy implications ...................................................................... 87 9. A potentially promising economy is negotiating the terms of globalization................ 87 9.1 Introduction............................................................................................................. 87 9.2 Tanzanian economic policies.................................................................................. 88 9.3 Trade and the role of China .................................................................................... 90 9.4 Frameworks in which Tanzania negotiates the conditions for globalization.......... 90 9.5 Current ongoing trade negotiations......................................................................... 91 9.6 The dilemma EAC or SADC .................................................................................. 93 9.7 Consequences.......................................................................................................... 94 9.8 Conclusions............................................................................................................. 94 10 Conclusions.................................................................................................................. 95 11 Some recommendations............................................................................................... 96 12 References.................................................................................................................... 97 ii Annex 1 Political developments and changing economic policies in Tanzania............. 102 Notes ............................................................................................................................... 108 Abbreviations ACP Africa, Caribbean and Pacific countries ACSAP Anti-Corruption Strategy and Action Plan AfDB African Development Bank AGOAAfrican Growth and Opportunity Act BDS Businesses development support system BEST Business Environment Strengthening for Tanzania Program BRELA Business Registration and Licensing Agency BSS Business Support Services CBO Community Based Organization CEM Country Economic Memorandum CSRC Civil Service Reform Commission COMESA Common Market for Eastern and Southern Africa CRDB Cooperative and Rural Development Bank CSRP Civil Service Reform Program CTI Confederation of Tanzanian Industries DANIDA Danish International Development Agency DAS District Administrative Secretary DC District Commissioner DFID Department of International Development DTIF Dar es Salaam International Trade Fair DTIS Diagnostic Trade Integration Study EAC East African Community EDF European Development Fund EPA Economic Partnership Arrangements EPZ Export Processing Zone ERP Economic Recovery Program ESAP Economic Social Adjustment Program ESRF Economic and Social Research Forum EU European Union FDI Foreign Direct Investment FSRP Financial Sector Reform Program GDP Gross Domestic Product GVC Global Value Chain HIPC Highly Indebted Poor Countries HIV Aids virus ICT Information and Communication Technology IHE Institute for Infrastructural, Hydraulic and Environmental engineering ILO International Labour Organisation IFMS Integrated Financial Management System IMF International Monetary Fund IRP Integrated Road Program JDI Japan Development Institute iii KICAMP Kinondoni Integrated Coastal Area Management Program LGA Local Government Authorities LGCDGS Local Government Capital Development Grant System LGRP Local Government Reform Program LGSP Local Government Support Project MDA Ministries, Departments, Agencies MFA Multi Fibre Agreement MIT Ministry of Industry and Trade MPs Members of Parliament NSGRP National Strategy for Growth and Reduction of Poverty NBS National Bureau of Statistics NESP National Economic Survival Program NGO Non Governmental Organization NMB National Micro-finance Bank NPRS National Poverty reduction Strategy PPP Private Public Partnership PSD Private Sector Development PRSC Parastatal Sector Reform Commission PSRG Parastatal Sector Reform Program PRSC Public Sector Reform Commission PSRG Public Sector Reform Program PRSP Poverty Reduction Strategy Paper PSRP Public Sector Reform Program RABO Cooperative Dutch Raifeissen farmers Bank RARP Regional Administration Restructuring Program RAS Regional Administration Secretary RC Regional Commissioner RDC Rural District Councils REP Research on Enterprises Program RS Regional Secretariat REPOA Research on Poverty Alleviation SACU Southern African Customs Union SADC Southern African Development Community SAP Structural Adjustment Program SDI Special Development Initiative SIDA Swedish International Development Agency SIDO Small Industry Development Organisation SMEs Small and Medium Enterprises SNV Netherlands Development Organization SPS Sanitary and Phytosanitary Standards SSA Sub-Saharan Africa TANESCO Tanzanian National Electricity Supply Company TASAF Tanzania Social Action Fund TAZAMA Tanzanian Mining Activities TAZARA Tanzania Zambia Railway Authority TCCIA Tanzania Chamber of Commerce, Industry and Agriculture iv TDTC Technology Development and Transfer Centre THA Tanzanian Harbour Authority TIC Tanzania Investment Centre TNBC Tanzania National Business Council TPSF Tanzania Private Sector Foundation TRA Tanzania Revenue Authority TRC Tanzanian Railway Corporation TSh Tanzanian shilling UN United Nations UNESCO United Nations Education, Science & Cooperation Organization UNFPA United Nations Fund for Population Activities UK United Kingdom URT United Republic of Tanzania US United States VAT Value Added Tax WB The World Bank WDC Ward Development Committees WDP Ward Development Plan WEO Ward Extension Officers WTO World Trade Organization Tables 1. Poverty reduction between 1991 and 2001 2. Contribution of different sectors to export in 2003 and 2004 3 Local procurement of goods and services by mining companies in Tanzania 4. Per capita income data for the five regions 5. Grants and own sources of revenue for the Arusha and Dar es Salaam regions 2001-04 6. Remittances as a contribution to total monthly income per household 7. The terms informal sector and informal flows compared 8. How informal flows affect different locations in Tanzania 9. Most important rural informal sector activities in Tanzania 10. Most important informal sector activities in Dar es Salaam 11. Urbanization trend in Tanzania over 1967-2002 12. Positive and negative factors for the urban informal sector according to Tripp 13. Dar Es Salaam population by municipality and sex (2002) 14. Dar es Salaam by municipalities, divisions, wards, streets, villages and hamlets 15. Growth rates of the urban population in selected countries, 1970-1980 16. Regional GDP at current prices: Arusha, Kilimanajaro and Dar es Salaam 17. Sources of revenue of the city 18. Comparison of own and grants collected and received by municipalities 19. Dar es Salaam versus Nairobi 20. Regional GDP at current T. Sh.: Arusha, Kilimanajaro and Dar es Salaam 21. Kilimanjaro and Arusha compared on key variables 22. Specific factors explaining performance of Kilimanjaro and Arusha region v Boxes 1. Energy 2. Road network 3. Rail network 4. Air transport 5. Communication 6. Port and maritime system 7. Terms of reference for such a study of the potential of the port of Dar es Salaam for the country and the sub-region 8. More property tax collected 9. Rural development initiatives mentioned in the newspapers 10. Global cities 11. BRELA reasons to register your business 12. Factors explaining the success of the High Potential Asian Economies 13. Suggested variables for the comparison between the two regions 14. Different levels of economic integration Figures 1. New conceptualization urban rural dynamics 2. Urban rural dynamics: flows 3. City Population distribution by age (2002) 4. Population by age groups vi Executive Summary Tanzania is sometimes called "a rare example of African success", but the Financial Times (3-8-2005) adds that, although "the country has done most things right (it) remains mired in poverty". For example, the business climate has greatly improved during the ten-year president Benjamin Mkapa was in power. The country has also witnessed a fairly good macro economic performance (more than 4 per cent growth per year since 1998 and 6.2 percent economic growth per year during the 2002-2005 period). Tanzania belongs to the top of African pro-business reform countries. According to the Financial Times (6-9-2006) it simplified business regulation, improved property rights and made it easier to start an enterprise. This conclusion is based on a World Bank IFC report, which found that the cost to register a new enterprise reduced 40 percent. Through a change of the legislation small investors also receive better legal protection. Importing goods has become easier by using a computerized system. The average number of days to import goods went down from 51 to 39 days. For export the average went from 30 to 24 days. Finally the taxes to be paid for this kind of transactions went down 3 percent. The Tanzanian population also grew with 2.9 percent per year between 1988 and 2002 and population growth declined to 2.1% per year only in 2005 (World Bank, 2006a). This means that a reasonable economic growth of 6.2 percent during the last three years (5.0% in 2005, World Bank, 2006a) translates into a per capita growth of about three percent. On top of that economic growth is not distributed equally. Dar es Salaam, Tanzania's capital, functions as a center of growth, while some other regions are still stagnating. Only in Dar es Salaam there has been real poverty reduction (see table 3). Section 2 reviews different explanations of Tanzania's success and the possible lack of growth in different regions of the country, while in the third section we go deeper into the urban-rural dynamics to explain the distribution of growth over the country. In the next section we suggest how to develop these flows more, while in section 5 the role of the urban informal sector in the Tanzanian economy is highlighted. Finally we draw some conclusions in section 6 2. Different explanations of Tanzania's success 2.1 Introduction Eight explanations of the economic success of Tanzania, but the possible lack of sread of this growth will be reviewed: · There is no success, prices of commodities and development aid have increased (2.2); · Tanzania and in particular Dar es Salaam benefited from economic restructuring (2.3) · There is growth, but the income differentials between Dar and rural areas becomes bigger (2.4); · The effects of decentralization on different regions are important (2.5); vii · The role of the urban informal sector in spreading economic growth is important (2.6); · There is not enough infrastructure outside the capital to spread growth (2.7); · The important role of private and partially foreign investments in sectors like mining and tourism contributes to Tanzania's success (2.8); · Globalization is threatening, but also provides new opportunities (2.9). 2.2 There is no success, prices of commodities have increased and donor darling Some argue that there is no success, prices of commodities have increased and that means the value of exports has increased, but not the quantities produced, while the government and donors spend a lot of money in Dar es Salaam, which stays there. A lot is probably spent in the center and there is a clear link with how development is managed. Jacobs (1970) concludes that the relation between the city and its hinterland can be very fruitful, if properly managed with the urban manager as the key actor. It is true that the government and donors spend a lot in Dar es Salaam. However, does the fact that the Government and donors spend a lot in Dar es Salaam mean that the other regions get no chance to develop? The center is driving development. Much depends on the mechanisms to transmit the positive effects of these expenditures and investments. It is also true that the prices of commodities have increased substantially between 2002 and 2006. The following table gives an indication of some of the trends. Volumes have also increased. 2002 was a bad year with low commodity prices. Since then tea, cotton and tobacco went up, but now coffee is going down again. Among traditional exports cotton increased by 52%, tobacco by 48% and cashew nuts by 28.6% in 2004, due to favorable weather conditions and timely availability and proper usage of agricultural inputs. In some cases the purchasing price for the farmers should increase to get more money into the rural areas. Table 1 Contribution of different sectors to export in 2003 and 2004 Year Traditional exports Non traditional Total value exports 2003 US$ 220.5 million US$ 908.7 million US$ 1.129.2 million 2004 US$ 292.3 million US$1,042.6 million The value went up An increase of An increase of by 18.2% between 32.6% 14.7% 2003 and 2004 Share 21.9(19.5% in 2003) 78.1 (80.5% in 100 2003) Source: Economic survey 2004. In the category Non traditional exports in particular the price of gold increased substantially in the world market. This commodity alone contributed 47.2%. In 1998 non traditional exports were only 232.2 million. The year ending June 2006 total exports on an annual basis amounted US$ 1.736 million, representing an 8.9 % improvement over the previous year (Guardian, 26-8-2006). Tourism and gold account for 51% of total exports. viii The improvement was due to improved prices and export volumes, in particular the volume of cotton and cloves exported increased substantially. Of course the positive effects of the increased commodity prices were partly offset because Tanzania also had to pay more for its imported oil. However, the country is not facing a major balance of payments crisis, although the value of Tanzanian shilling was somewhat slipping during the summer of 2006. The conclusion is that increased prices and more aid are not enough to explain the good performance of the economy since 2002. 2.3 Dar es Salaam the focal point of the Tanzanian economy The economic crisis of the eighties led to reforms in the nineties, although the implementation of the reform policies was not always straightforward. Inflation increased and no poverty strategy existed. In 1995 president Mkapa came to power and he was reelected for five years in 2000. He introduced serious reforms and under his government the Poverty Reduction Strategy Paper (PRSP) was prepared in 2000. He also pushed decentralization of government and restored trust in the Tanzanian economy. The new president, after a year in office still has a good press, despite the drought in 2005 and the resulting energy crisis. August 2006 a meeting of Tanzania's regional leaders was held for the second time in 2006. The president delivered an important speech and the event shows the authorities take decentralization more serious and we will try to assess the extent to which it contributed to regional development.i In "Development vision 2025" Tanzania presents a strategic view. It wants to be semi- industrious by 2015 and the country has an industrial development policy, which is a good basis for such a future. The role of the manufacturing sector should go from 8 to 15 percent of GDP by 2020. This requires upgrading technology, human resources development and reforms at the local government level. So far most of the reforms concentrated on making the government more effective and efficient at the national level. There is also an ambitious economic development program dubbed "Tanzania Mini-tiger plan 2020" (The Citizen, 11-7-2005). It aims at fast-tracking the achievements promised in the Development vision 2025. The minister of State in President's office (Planning and privatization) told Members of Parliament that the new program seeks to replicate economic successes of South East Asian countries. It complements the ongoing second phase of the National Poverty Reduction Strategy (NPRS). The program aims at increasing the national per capita income from US$ 340 (World Bank, 2006a) to 1000 and at creating 2 to 3 million new jobs by 2020. The Minister noted that the project involves all sectors of the economy and requires quick decisions, the reason it operates under the President's office. Tanzania has achieved political stability and economic growth; the challenge is now to spread the effects of the resulting economic growth. It will be argued that largely informal flows play an important role in this process. These flows of people, ideas, money, goods and services bring development to other parts of the country. A more positive approach to these flows means attention will be given to the issue of rising the productivity of economic activities and to the need for these entrepreneurs, cities and regions to compete in the global economy. Skills, entrepreneurship and existing financial ix mechanisms can be further developed to improve the competitiveness of the Tanzanian players the global economy. Dar es Salaam counts about 3 million inhabitants and has the highest population density in the country (1793 people per km2). The capital benefits from a number of favorable conditions. It lies at a cross road of important routes in East Africa and has a good natural port, which can serve a much larger area than the capital region. The city has a more developed infrastructure than other Tanzanian cities and benefits from the presence of a number of government activities, although Parliament has its seat in Dodoma. Here we analyze the growth in population, in tax revenues and summarize other indicators of the dominant position of Dar es Salaam in table 3, before drawing some conclusions. In population terms Dar es Salaam is one of the fastest growing cities in Sub Saharan Africa. The city population is estimated to be approximately 3 million in the year 2005, with a growth rate of 4.3% percent. The relatively high population growth rate is due to increased birth rates, immigration rates, and more significantly by transient population. Of the three Municipalities, Kinondoni had the highest population with a total of 1,083,913 people (43 percent) followed by Temeke with 768,451 people (31 percent) and Ilala with 634,924 people (26 percent; see table 2). Table 2 Dar Es Salaam population by municipality and sex ( 2002) Municipality Male Female Total % of the Growth rate total 1988-2002 Kinondoni 547,081 536,832 1,083,913 43 4.1 Temeke 387,364 381,087 768,451 31 4.6 Ilala 320,408 314,516 634,924 26 4.6 Total 1,254,853 1,232,435 2,487,288 100 4.3 Source: URT (2002): Population and housing census 2002. Rural to urban migration played a big role in the urbanization process in Tanzania in the first two decades of the country's independence. The major factor for rural youths to move to urban areas has been the search of a better life especially aspiration for salaried jobs, access to services and social facilities. A lot of people seem to go directly to Dar es Salaam (Mlingi and Asey, 2005: 71). There is less this two-step approach of going to other cities first in Tanzania. For Dar es Salaam City, the number of people born outside Dar es Salaam region was 56.3 percent in 2004. This means that more than half of the city's population was migrants. This shows that migration has had an impact on the rapid population growth of the major urban center in the country. The Tanzania Revenue Authority (TRA) is an example of the economic success of the reform policies. It collected 25 million US$ per month in 1995. Ten years later this is 160 million! With about 200 million the government could cover an important part of its budget instead of being dependent on foreign donors, who currently finance more than 40 percent of Tanzania's budget, although even if Tanzania were to increase revenue to US$200 million or more, it still would not be able to adequately finance infrastructure etc. and still need to benefit from ODA. The increased tax effort is even more x remarkable, since import duties in general have gone down as a consequence of the Uruguay round of the World Trade Organization. The revenue to GDP ratio is still low, so we could also describe what we found as the recovery from a really bad situation rather than a remarkable success story. Local governments have also done better (Heymans and Kumar, 2002). They collect the development levy (which was abolished a few year's ago), land tax and a cess levy on the value added tax (0.3%), while for national government the income, consumption (VAT of 20%) and import duties are very important. The Ministry of Finance has also made the tax system slightly easier, by abolishing a number of taxes, or by taking them away from local government and by providing an income transfer to local governments instead.ii The revealing figure about taxes is that 82 percent is collected in Dar es Salaam and only 18 percent in the rest of the country! Given the TRA is in the capital, given the money is generated there and the ministries are there, it can be expected that a substantial part of the revenue is also spent in the capital. Very little is known about the micro implications of economic growth for the poor living in different cities and regions in the interior of Tanzania (Dekker and Eeckhout, 2005). The existing inequalities come out very clear in the Household budget survey 2000/01 (NBS, 2002b). The question is what explains the difference and what can be done to spread welfare, rather than trying to stop it from accumulating in the capital. Tanzania calculates the poverty line on the basis of one dollar per day and works with a basic needs line (2200 calories per person per day plus some expenditure on for example clothing as well). The percentage of poor according to the poverty line decreased from 39 to 31 percent in a decade (NBS, 2002). Less than one out of every five Tanzanians lived below the basic needs poverty line (19 percent, down from 39 percent in 1991). However, poverty diminished in particular in Dar es Salaam from 28 to 18 percent between 1999 and 2001. In the other urban areas the decline was from 29 to 26 percent. At the same time the inequality increased, in particular in Dar es Salaam it became quite pronounced (the Gini coefficient increased from 0.30 to 0.36, which is above the national average). These figures combined with the rapid population growth of Dar es Salaam through migration suggest that Dar has indeed been an important engine of poverty reduction in Tanzania. Demombeys and Hoogeveen (2004) have shown that poverty in fact increased during the first part of the nineties, only to decrease in the second part. What still needs to be explained is why according to these indicators there was a considerable decline in urban poverty (in particular in Dar es Salaam) and only a slight reduction in rural poverty. The picture is even more complicated if certain regions are going down (for example the Kilimanjaro region), while others are gradually improving (for example the Arusha region). Even within these regions the picture may differ as illustrated by the decline in some of the districts in the Arusha region and the increase in wealth in others. Notoriously poor regions are Lindi, Singinda and Shinyanga. xi The picture becomes more nuanced if other variables are taken in to consideration. For example the access to clean drinking water has increased between 1991 and 2001, in particular in the urban areas, but not in Dar es Salaam. Sanitation facilities show a different pattern, with a slight decrease in the rural areas and a general increase in the urban areas. However, the number of people per sleeping room has declined, also due to a decline in household size. Here the exception is Dar es Salam, where the number of people per room has increased, despite the decline in household size (NBS, 2000: 16). If improved foundations, roofing and walls of houses are taken into consideration the rural areas do better than the capital. More often in the rural areas people have built a foundation for their house, improved the walls and used galvanized metal sheets for the roofing (NBS, 2002). This is also the case in the other cities, while in Dar es Salaam the percentage of people with improved roofs was already relatively high and hardly changed. Improved foundations, walls and roofing are typical micro level indicators of the spread of economic growth. In terms of access to services the number of people living less than one kilometer from a primary school in 2001 has decreased compared to 1991. Also access to clinics and pharmacies (also measured in terms of distance) has not improved in the rural areas and other towns between 1991 and 2001, but it has improved in Dar es Salaam. Public transport provides the same picture, an increase in access (measured as less than one kilometer from a bus stop) in the capital and a decrease in the rural areas. Finally, access to markets has improved between 1991 and 2001. This in particular the case in the other urban centers and slightly less in Dar es Salaam. However, in the rural areas access to markets has decreased, which may be the result of a concentration process. Looking at education would be another way of comparing the rural and urban areas. Here the number of people between 20 and 30 years of age that have finished primary education went down in the rural areas. In the urban areas the percentage of educated people in this age bracket increased. This may also be an indication of rural-urban migration for educational purposes and as such would be an illustration of what we call the informal flow of people in the country. Also for secondary education the increase of successful candidates is in the urban areas. However, the number is very low, being only 6 percent of the age group 20 to 30 years of age. Agriculture has been growing quite constantly at about 5 percent (PRSWC, 2003). The conclusion is that poverty measured in monetary and basic need terms (in the narrow sense of the word) may have decreased more in Dar es Salaam. That doesn't mean there are no signs of improved living conditions in the rural areas and other cities as well. If the broader definition of poverty is used of the Human development report (UNDP, 2004), it can be shown that in this decade the access to education and health services has not really improved. In the restructuring process, spreading services has suffered from budget cuts. Other indicators of Dar es Salaam's role are summarized in table 3. Tanzania's capital count almost half of all manufacturing establishments and about one third of all establishments in the country! Table 3 gives the impression of cumulative causation: all xii these factors together give Dar es Salaam a lead over other parts of the country. Given the dominant role of the capital on most indicators, the two questions that deserve attention are: what explains the economic success of Dar es Salaam and what are the main factors explaining the growth and income differential between Dar es Salaam and the rest of Tanzania? What explains the success of Dar es Salaam? Dar es Salaam enjoys some vary favorable conditions (Van Dijk, 2006a): · Strong growth of population · High tax revenues, the city generates 82% of all taxes in Tanzania · Its share of GDP is 17.5% while it counts only 13% of the total population · Manufacturing units 1405 out of 3431 · Establishments 9397 out of 28910 · Licenses 18106 out of 81661 · Buses 6600 out of 11279 · 62% of households involved in informal activities · Location plus largest part government's services and formal sector, port · Enormous dynamics, decentralization resulted in more competitive municipalities · Increase in total factor productivity 6.2% · Growth exports and 61% households in informal activities Table 3 Indicators of Dar es Salaam dominant role Variable Dar es Salaam Other urban Rural Comment Share of Compared to 13 GDP 17.5% % of population Number of In total manufacturing 1405 3431 Total One third of all establishments 9397 28910 Number of licenses 2002/3 18.106 81.661 in rest Taxes 82% Only 18 in rest Poverty From 28 to 18 From 29 to 26 From 41 to 39 From 22 to 19 reduction 91-01 percent (-10%) percent (-3 %) percent (-2%) percent (-3%) Average hh size 4.8 4.9 5.9 Aver hh smaller Working age 62.9% 57.0% 50.3% Largest share in (15-64) Dar es Salaam Younger people 25.6% 22.2% 18.6% More in 15-44 year range Remittances 3.5% 3.2% 2.1% 2.4% of budget Number of buses 6600 in 2004 In total 11.279 Source: NBS (2002, 2005). In Dar es Salaam you find the big businesses or some very small indigenous business. Mining and tourism are flourishing but informal sector incomes are very low. In xiii purchasing power parity it is slightly better. However, the differences are too big! We will go deeper into some of the factors explaining the success of Dar es Salaam. Dar es Salaam not only has the advantage of its location, it also benefits from the presence of the largest part of the government's services and of all formal sector economic activities in Tanzania. The city has a decentralized administrative structure, where the three local governments that make up the city have elected politicians as their mayors. The local governments work together at the city level and are part of the larger Dar region. Dar es Salaam shows an enormous dynamics. In particular the three municipalities have become very active in developing social policies and in promoting development. Although they don't really formulate their own economic policies, they started recognizing the importance of having economic activities and creating the conditions for such activities, including the informal sector. In this respect the three municipalities can serve as examples for local governments in the interior of the country. NBS (2005) analyzes as the major source of growth in Tanzania an increase of total factor productivity between 1986 and 2004 of 6.2. Main sources of growth expenditure- wise are exports (although, the growth of exports is mainly attributable to mining, whose contribution to growth is still rather small). Through the port the capital city plays a more important role than being a national capital. The port is in an excellent position to receive more goods or channel goods from neighboring countries like Burundi, Rwanda, Uganda, Mozambique and Malawi. The major issues for the members of the Confederation of Tanzanian Industries (CTI) are the economic policies of the government and the level of taxes. CTI wants the government to abolish cross-subsidies for tariffs of utilities. They want to discuss taxes and by-laws (on parking, on waste collection, pollution, etc.) with the government. Finally, they consider that changes in the labor laws are necessary, but have been achieved. Some labor laws dated from 1964. Competition for their members comes from Kenya and South Africa. However, CTI consider the presence of South Africa in certain sectors (for example in mining and telecom) as very beneficial. For example, prices of mobile operators have come down substantially. The overview of the problems of Dar es Salaam and the disjoint way in which the different municipalities deal with their problems show the importance of undertaking a strategic planning exercise for cities and regions. Dar es Salaam is involved in preparing a plan, which will clearly bring out the urgency of certain bottlenecks (City Director, 2004). Dar es Salaam is a peripheral node in a global system. It should take benefit from its location and try to serve a larger area because of the benefits it would bring to its population. Given the functionality of Dar es Salaam as the country's main port, administrative center and center of economic activities the city can play a very important role in the further development of the country. xiv 2.4 Main factors explaining differences between Dar and the rest of Tanzania What are the main factors explaining the growth and income differential between Dar es Salaam and the rest of Tanzania? A drop in poverty can be seen in table 3, in particular in Dar es Salaam, but also in the other urban areas and even in the rural areas. Sustainability and expansion of food and cash crop production in the Dar es Salaam region appears difficult to be achieved. Some of the existing problems and obstacles facing this sector are: poor farming methods and equipment, non existence of data about farmers and crop yield, great increase of people especially youths, high rate of urban expansion leading to decrease of agricultural land, land degradation due to soil erosion, inadequate knowledge on improving farming techniques by farmers, outbreak of crop pests and diseases affecting plants and crops, and shortage of inputs including better quality seeds and fertilizers. Other factors will also be discussed in the following sub-sections. Rutasitara (2002: 85-86) mentions as characteristics of the non-agricultural activities that these are typical sole proprietorship activities, mostly managed by family members. The activities tend to be intertwined with the agricultural calendar and usually use local inputs. Income may be unstable and hired labor is limited. However, there are low capital requirements to enter into the rural non-agricultural activities and the technology tends to be simple. These activities are easily differentiated by gender, offering employment (and hence income) to the disadvantaged. Rural households with non-agricultural activities suffer less of poverty (Rutasitara, 2002). He gives characteristics of these activities, which can be considered the origin of many urban informal sector activities and hence we will draw on this study. The importance of non-farm activities is very much linked to the possibility of markets for these products. These products can be marketed more easily than agricultural products, which all come on the market at the same time and require the farmers to deal with only one party (the private trader, or some marketing board), who has a much better understanding of the price. Plus agricultural products face transport and storage cost; it requires more information and more capital. However, it is more difficult to get the message to the rural population that non-farm activities are more rewarding. It also requires knowledge and technologies and some small investments. Micro finance schemes can be very useful (see Gallardo et al., 2005). A lot of these products and services end up in cities and towns. The rural informal sector, or these non-agricultural activities are important examples of rural-urban linkages (Bagachwa, 1999). He notes that a dynamic rural informal sector not only employs 21 percent of the nation's labor force, but also provides estimates of the value added of the sector. He considers the sector (including the urban informal sector) contributes one third of the total official economy's value added. Dar es Salaam is very different from the rest of the country. Rapid development in this city is largely the result of better macroeconomic policies, agglomeration effects and a restored confidence in the Tanzanian economy. Such policies affect development directly and indirectly. Better governance helps a lot indirectly. Changes in pricing and taxation may help the agricultural sector directly. Unfortunately decentralization also led to too xv much taxation in the beginning and it is Central government that has pushed the local governments to put order in their house. They abolished certain taxes and forced the local governments to rationalize tax collection by focusing on a limited number of taxes and complementing their budgets with income transfers. The following table summarizes some of the positive and negative factors of Dar es Salaam versus Nairobi. Most of these factors affect both Dar es Salaam/urban areas and rural areas and we might also use them to explain why Tanzania has been growing faster than Kenya during the past decade. All Local Government Authorities (LGAs), which includes the City, Municipality, Town and District Councils, have a similar organizational structure irrespective of differences in population, resource-base and infrastructure. The LGA can also create local incentives. It would be interesting to have an infrastructure master plan for Dar es Salaam, since there doesn't seem to be a vision. The strategic plan, which is currently prepared at the City Council level (City Director, 2004), will help to develop such a perspective and can guide further development. Table 4 Dar es Salaam versus Nairobi Dar es Salaam positive trends Nairobi negative factors Politically stable Political crisis at local government level Decentralization stimulates better urban No real decentralization and poor urban governance and management governance and management Not as strictly regulated as before No policy with respect to informal sector Opportunities for informal sector Keeping of cattle in the city is not allowed Potential of the harbor Serious traffic congestion and urban Possibilities for urban agriculture pollution A lot of donor money is available Land scarcity Space for expansion Urban violence Source: this study and Karanja (2005) for Nairobi. The City Director (2004) notes in the City profile that there is inadequate knowledge about the dynamics of the informal sector, that the city has inadequate strategically located areas for allocating space to informal sector operators, but at the same time he admits that a robust informal sector will increase employment and improve living conditions of those who are economically marginalized. Reading the profile it seems the capital doesn't yet know how to deal with this sector, which is a pity, given its importance. ILO (1991) makes an effort to estimate employment in the informal sector. The report suggest 2 million people are involved, which would be 2.5 times the number of people with formal wage employment. According to the ILO employment in the informal sector would be growing at 3.5 percent per year. The integrated labor force survey 2000/1 (NBS, 2002a) measured informal sector activities by the number of households involved. xvi It turns out that in Dar es Salaam 62 percent of the households are involved informal sector activities. The percentage is 61 for other urban centers (used to be 42). In the rural areas 27 percent of the households have informal sector activities (used to be 21; URT 2002). In terms of number of people the labor force survey found it would be 1,439,847 people doing it as a main activity and 1,363,010 having informal sector activities as a secondary activity. This suggests that currently a much larger number of people are involved in this sector than according to ILO (1991). Tripp (1997) stresses the importance of women in the urban informal sector. NBS (2002a) provides the figures. In Dar es Salaam and in the rural areas there are slightly more males than females in both the main and secondary activities. In the other urban areas, more females were employed than males in both the main and secondary activities. The rural areas are the mirror image of Dar es Salaam: fewer young and less educated people, less tax revenues and lower per capita expenditure, fewer establishments, etc. The informal sector has become a mechanism to generate and redistribute growth. Hence we should understand better how it functions. The informal sector is the result of and resulting in flows of people, money and goods. These flows, partially generated by the urban informal sector, like migration, remittances and smuggling, will be discussed. 2.5 The positive effects of decentralization The next explanation concerns the effects of decentralization. It is concluded that decentralization stimulates better urban governance and management, but decentralization benefits in particular those local authorities that already have a strong economic base. One effect of decentralization is that more regional initiatives are taking place as has been shown by a comparison between Arusha and Kilimanjaro districts. Selecting two regions in the north for more detailed study, we found that the Arusha region is booming because of tourism, but also because of agriculture and trade links with Kenya and Uganda. Wheat, coffee and flowers (ten Dutch horticulturalists and at least one Frenchman), plus mining are mentioned, although the latter concerns mainly some small-scale activities. Tourism is an activity, which takes place outside the major cities and it increase has had a positive impact on rural demand. There are advantages and disadvantages to decentralization. Local governments need to develop the economic basis, but it is good to make the responsible people at this level accountable through elections. Promoting economic activities is not yet a policy at the local level. But supplying more facilities/services provision is. The good thing is that politicians know they will be held accountable for what they have promised. One of the remarkable achievements of Tanzania is the improved accountancy system for decentralized projects. There is a general problem that many of the good people are at the central level. The Ministry of Finance has trained 400 accountants for rural development projects. Many benefits were expected from more decentralization, in particular more democratization, improved management and a reduction of the regional disparities. However, the implementation of decentralization strategy after 1997 was not totally satisfactory (Steffensen et al., 2004 and Mitullah, 2005). Many shortcomings were noted, xvii partly because by then the central government continued to maintain strong presence at regional level. However the process has strengthened the democratization process by introducing mechanisms of accountability and by forcing local governments to compete for funds. An effect of decentralization is that more regional initiatives are taking place as has been shown by a comparison between Arusha and Kilimanjaro districts. Also the three municipalities in Dar es Salaam have been most active in developing economic policies, for example providing space and infrastructure. Other local governments have been able to do something about solid waste collection. We argue that decentralization has contributed to local development, because of the flows of goods, services, capital, ideas and people it has generated. There is still some resistance against the implementation of decentralization policies at the national level. Also the accountability in the process can be improved. The pace of harmonizing the roles between the different levels of government in Tanzania was slow. Sectoral ministries are reluctant to give away authority. The Local Government reform program also needs to be marketed more. After 2008 the LGSP project hopefully has created the momentum to do things better at the local government level. The responsibility is now with the local government and empowerment of the local governments should be the result of this project. By then this approach should also be mainstreamed. It should be day-to-day government business. The LGSP helped to build capacity at the local government level. Decentralization is the framework for further development activities and the citizens and politicians should use it as such. A point of critique is the rather arbitrary local tax increases. Decentralization provides the opportunities; the challenge is now to design mechanisms to get capital and entrepreneurial talent to rural areas. It was hoped that decentralization would have diminished the rural urban differences. That could have led to less migration and hence solve the problems of big cities like Dar es Salaam. There are however no indications, despite a slight improvement of the income differences, that the number of migrants or their destination has changed over time. Local governments have received more autonomy in the framework of decentralization. Local governments can and will do more, although they are not able to take up all the responsibilities fully yet, if only because of the lack of sufficient local finance. The local government reform component of the reform program, which really started in 1996 and the local government grant system and increased tax collection have provided the local level with more funds. 2.6 Spread of economic growth is more important than shown in available statistics Another explanation focused on the role of the urban informal sector in spreading economic growth, which may be more important than what would be expected from the available statistics. What is the role of the urban informal sector in spreading economic growth? There are estimates of the informal sector employing up to 90 percent of the xviii labor force (if non estate agriculture would also be included). A more realistic figure of 58% is mentioned in several World Bank publications. It means that labor is allocated, investments are made and raw materials are used without really being measured for almost half of the economy. It involves flows of people and money, not shown directly in the statistics, but known to many people. We have concentrated on how this sector affects Dar es Salaam and other Tanzanian cities and regions. The urban informal sector contributed to the growth of Dar es Salaam. However, it also contributes to spreading economic growth beyond the capital, through informal flows of people, money and goods and services throughout the country and even abroad. China recently adjusted its GDP 17 percent upward after including these activities (Financial Times, 14-12-2005). This sector developed from an illegal sector in the Nyerere time to a recognized source of entrepreneurship under the current leadership. The opportunities then were linked to the policy of nationalizing economic units and the resulting shortages and distortions in the economy around 1995. Currently the entrepreneurs in this sector explore the opportunities provided by a more liberal economic environment and more booming formal sector activities. Table 5 How informal flows in affect different locations in Tanzania Factor\region Dar es Salaam Arusha Kilmanjaro Flows of people Migration Strong inward As well in as out of Mainly out of the migration of young the region region motivated & skilled Lifetime migrants 1,319,360 249,971 113,743 2004 Flows of money Loc. tax revenues: Increasing Increasing Declining Remittances: In and out Coming in and Coming into the Increasing going out when out region to support migration those left behind Investments: High Investment increase Decreasing Government expenditures: High Medium Limited Flow of goods and services Local products and In and out, In and out, from Flows are not as small scale services including from abroad, partially important as before abroad illegal from Kenya Source for migration: Mbonile (2004). If more than 50 percent of the Tanzanian economy is informal, the current dynamics should also be reflected in the informal sector. The six percent economic growth will also apply to this sector, because a lot of money is spent in the sector and most of the money generated in the sector will be spent informally and bring about multiplier effects. Hence, the question is what determines the further development of the informal sector? Where do xix the people come from? How do they finance their activities? What is the role of the government in their development? An impression of some relevant factors is given in table 5. For each flow we try to indicate how it works out in Dar es Salaam, the Arusha and the Kilimanjaro region. Table 6 gives an impression of the activities in the informal sector of Dar es Salaam. Very different activities in this sector have very different problems and hence there is no one size fits all solution. ILO (2005) suggests getting the informal sector out of the margin into the mainstream. This requires removing bottlenecks and introducing some specific facilities. The urban informal sector may help families to grow out of poverty, but it is also very much a subsistence sector. The question is how to get these people to invest and to expand their businesses. An impetus from outside may mean more for the local economy, than efforts to develop policies for the informal sector in the capital. Such a push could be the training received, a trip to another city or country or financial support from a family member living elsewhere. The informal sector may have its own dynamics, but currently lack of credit facilities or fear for taxation restrains its development (Kweka et al., 2004). Table 6 Most important informal sector activities in Dar es Salaam Small business license holders in Dar es Manufacturing activities Salaam: trade Operating a stall for assorted items Tailors Selling firewood Making buns, rice cakes and pastries Selling buns, rice cakes and pastries Working as a carpenter Vending fruit Working as a cobbler Selling fried chips, chicken, eggs and or Repairing watches barbecued meat Slaughtering animals Vending fish Welding Source: Trip (1997). 2.7 Not enough infrastructures outside the capital Some argue that there is not enough infrastructure outside the capital. Infrastructure delivers inputs to farms, helps to market agricultural produce, distributes raw materials and facilitates both domestic and international trade. There is a direct link between economic development and the growth rate of infrastructure. The rehabilitation of physical infrastructure, particularly the trunk and regional roads under the Integrated Road Program (IRP) during the period improved significantly marketing networks. Recently the government started working on the privatization of the railroads, which hopefully leads to an improved functioning and could also be a big boost for the harbor of Dar es Salaam. xx A lot of money is spent on improving the infrastructure and transport (World Bank, 2006b). Transport contributes to poverty alleviation by assuring cost-effective transport of goods and people. It also links the country to the global economy and assists neighboring landlocked country to connect to the world. World Bank (2006b) provides an overview of the transport infrastructure and its management. The lack of electricity of a good railway system and the relative high minimum wages mean that not many foreign investors are interested in setting up a factory in Tanzania. The cost of bringing a container to Rwanda is still higher then shipping one from the port of Dar es Salaam to Japan. The positive effects of infrastructure built by different levels of government and of the expenditures of local governments works in favor of Dar es Salaam, to the extent that the capital has received more infrastructure and collected more taxes than any other city or region in Tanzania, but the infrastructure is gradually improving. 2.8 Private and foreign investments in booming sectors like mining and tourism Finally, the role of private and partially foreign investments in booming sectors like mining and tourism has been studied. These sectors contribute to the growth of regions or cities through the multiplier effects of these expenditures. However, the companies and projects use a lot of facilities in Dar es Salaam. Also large amounts of development aid are spent there before generating development in those regions, which are lucky to get it. The exact links between foreign direct investment (FDI), aid and regional development in Tanzania are difficult to establish. It is also difficult to assess where exactly the FDI and development aid is spent. The role of private and partially foreign investments in booming sectors like mining and tourism is certainly important. Also the steep increase of the horticulture sector is driven by foreign direct investment. Foreign investments amounted to US$ 470 million investments in the country in 2004, this is more than the direct investment in Uganda and Kenya in that year, namely US$ 237 and 46 million respectively. 2.9 Globalization is threatening, but also provides new opportunities Globalization is certainly threatening, but we will argue that it also provides new opportunities for countries like Tanzania. In a more liberalized world economy a lot of specialized agricultural products could also be exported and the country has an interest to anticipate such developments, which can be expected in the framework of the Doha round and the current negotiations with the EU. In 2003 the Ministry of Industry and Trade formulated a Trade policy for a competitive economy and export-led growth (MIT, 2003). MIT (2003) attaches to trade a pivotal role and doesn't discuss autarky, import substitution and protection, but rather globalization and further integration. It is quite a comprehensive document and it really goes for competition, regional integration and multilateral trading, but as will be shown the current situation has become extremely complicated because of a breakdown of the World Trade Organization (WTO) negotiations, pressure to finish the inter-regional schemes, in particular the economic partnership agreements (EPA) with the EU and the suggestion made to choose between the Southern African Development Community xxi (SADC) and East African Community (EAC). In the meanwhile the EAC has started an ambitious program in the direction of African unity, skipping a number of stages, which are normally in between full economic integration and a creating a free trade area. When at the beginning of 2006 a EU official said Tanzania had to choose between EAC or SADC he was reprimanded. During a recent meeting of SADC some members phrased the same question and Tanzania will have to answer this. In its trade strategy it assumes it can follow both tracks, but both groupings have changed gear and want to go much further than originally announced. The EAC wants complete unification of the country and SADC intends to reach the next stage of integration in a relatively short period. Also in August 2006 the 12th meeting of the EAC Council of ministers has taken place. According to the Deputy Secretary General of EAC the community faces a big challenge of meeting the people's expectations. Already the launch of the EAC Custom Union had heightened expectations of increased productivity and trade in the region. People expected wealth creation and an improvement of people's standards of living. Negotiations concern the admission of Rwanda and Burundi to the Community and a protocol on the free movement of persons, labor and services (Citizen, 25-8-06). SADC will carry out a plan of transforming the block into a common market. The timetable is a free market by 2008, a custom union by 2010 and a common market by 2015. The SADC versus EAC discussion comes in parliament in October 2006. The arguments used are that Tanzania's roots are more in the Southern African countries and in supporting their struggle against apartheid. In EAC Tanzania is very much afraid of Kenya and the experience with the first EAC has proved them right. The private sector sees more opportunities to do business in EAC and COMESA and fear South Africa in SADC. To play a role in the global economy it will be important that Tanzania becomes internationally competitive. In these important negotiations the country determines the rules of its participation in the global economy. Tanzania has already benefited from globalization, but it could benefit even more if it negotiates the best conditions for its participation. 3. The study of rural urban linkages 3.1 A new conceptualization The challenge concerning the relations between a city and its hinterland is to maximize the positive interactions between the rural areas and the towns and vice versa. The informal sector plays an important role in the rural-urban relations. There are in Tanzania informal flows of people, money, goods and services (the 5 flows: Goods-Ideas-Money- People-Services or GIMPS), which spread development over the country. We will use the term informal flows of people (the migrant comes and returns with skills, ideas, equipment and goods), money, ideas, goods and services. This conceptualization goes beyond the traditional registration of goods exported from or imported in the rural areas and beyond the informal sector concept. The idea is shown in figure 1 xxii Urban Goods Rural system system with its own Ideas absorbing dynamics Money these flows People Services Economists usually study rural-urban linkages in terms of `exports' of goods from the rural areas and imports from the urban areas. We want to study rural-urban linkages in a different way: 1. To avoid thinking in terms of two separate systems, an urban and a rural, served by two types of experts and functioning more or less isolated from each other. 2. To find out whether the urban system can stimulate the rural and the rural can stimulate the urban and 3. To discover to possibilities of spreading development given in many countries the development expenditures tend to be spent in cities and bring about a process of development over there. In the cities most institutions spending the money are based. 3.2 Do these flows lead to rural development? Using this framework, the following relevant questions can be asked do these flows lead to rural development in the case of Tanzania? Many ideas flow back to the rural areas, driven by return migration, trade, remittances and new technologies and in particular the possibilities of modern telecommunication. In many villages one may find a copy shop or an internet cafe Most investment is driven by households, since banks provide no loans to these people and remittances are very important. Reviewing the evidence, we found that effects can be assessed: · Not through comparing the household incomes in the household survey s of 1990 and 2001; and · Not using the data on remittances, which are more important in the urban areas. However, these remittances are not really measured in the 2001 household survey because they are net and we do not know for which purpose they have been used; xxiii · But it may work, since decentralization leads to better regional and urban governance (example above and the evidence for Arusha and Dar es Salaam); · Tanzania has many pro-poor rural policies (Owen and Teal, 2005); · More infrastructure and transport is built outside Dar es Salaam; · One notes improvement in assets in the rural areas (see NBS, 2005); · The agricultural survey also shows improvements like diversification and the use of new technologies; · The importance of return migration should not be underestimated. Because of its high population growth Tanzania has a relatively young population. According to the 1988 census 46 percent of the population was under 15 years of age. The majority of the population of Dar es Salaam however is between fifteen and sixty- four years, which is the working age. There are few elderly people (only 2% are above 65 years). The total population increased from 7.5 million in 1948 to 34.5 million in 2002 (URT, 2003). Migration is common in Tanzania. People are moving from high-density areas to less populated areas and people are opting for living in the cities. An important transmission mechanism for skills and income is the labor market and its functioning is influenced by good urban management (Van Dijk, 2006b). One could look at these migrants as physical capital, in the form of human bodies, being brought to urban areas with more potential, where this capital will generate a higher rate of return than in the rural areas. People are reacting to incentives! There are also international flows of people. The number of people leaving Tanzania is not known, but Mbonile (2004) estimates international migration towards Tanzania to be 236,775 in 2002, based on the 2002 population census. 12 percent of these migrants settle in Dar es Salaam, 2.1 percent in Arusha and 2.3 percent in Kilimanjaro. In total the census reveals 4,671,641 lifetime migrants, or even twice as much if in and out-migration movements are counted separately. He also notes big differences per region. Rural-urban migration has contributed to growth of the urban informal sector. The major pull factor has been the search for a better life especially aspiration for employment opportunities, access to better services and social facilities. This in turn has put pressure on the utilization of undeveloped and developed facilities in Tanzania's cities. Rapid growth of squatter settlements, peri-urban shantytowns and unsanitary environmental conditions are some of the common are a growing health and social problem in these cities. Remittances, loans and savings are components of the flow of money to the urban informal sector. Remittances are a tricky concept, but the amounts involved are substantial. Conceptually remittances are transfers without an exchange, while otherwise goods are exchanged for barter or money. NBS (2002: 155) registers net remittances (received minus provided) but doesn't give a clue how the money is used. Remittances and social networks are the key to survival strategies in the rural and urban areas. It turns out that net remittances make up 3.5 percent of the monthly income per household in Dar es Salaam. In the rural areas the percentage is 2.1 and in other urban centers 3.2. The average for Mainland Tanzania is 2.4 percent (table 3). xxiv The urban informal sector mobilizes a lot of funds and generates income opportunities for the poor. The current policy trend is formalization, where poor people are given titles for land and other assets (their enterprise), which may make it easier for them to obtain loans. Tripp (1997: 164) distinguishes different informal marketing arrangements determining the flows of goods: between members of one village or of a nearby village. Secondly, shuttle markets, where the products are taken by bus or truck to a nearby town and then interregional trade where they are sold elsewhere and other goods are brought back. Finally there are open urban markets and export markets, where goods can arrive formally or informally. The Quarterly report of industrial commodities (NBS, 2002b) provides data on industrial production in 1985, 2000 and 2003. However, the location of the production and the destination of the products is not indicated. If this is the case for formal sector products it can hardly be expected that such data are available for informal sector. However, the two sectors compete, although the CTI doesn't consider the competition from the informal sector as very damaging for their members. In fact they want the informal entrepreneurs to become members of CTI. These flows of goods are sustained through small-scale transportation and often would be called smuggling because the formal requirements are not met. These informal movements of goods and services `exported' are not covered, the information is well known, for example food crops are going from Tanzania to Kenya and cattle is going on hoof. Other products flowing out of Tanzania are: coffee, tea and cashew nuts, but also hides and precious stones (Tripp, 1997: 87). Industrial products enter from Kenya usually essential goods (for example textiles through Mombassa). The government claims it contains smuggling. That is true for certain products, but in other areas there are problems. One way to reduce it was reducing the duties in the framework of different trade agreements. The International Monetary Fund published a study showing that 50 billion of Tanzanian shillings per year is lost as government revenues in Tanzania, because of smuggling along the Indian Ocean. Positive developments can be observed in the rural areas. For example the increase in horticulture (stimulated by a number of foreign investors). Secondly the Regional administrative secretariats are now assumed to deliver economic development supports services that focus on production related activities in agriculture, livestock, commerce, industry and natural resource sectors. Of course there is abject poverty in Dar es Salaam and in many of the country's regions (for example the article on Tanga region in Citizen 25-8-06). However, there are also a number of real positive developments. In the first place new pres I will distinguish some general favorable conditions and some factors that favor in particular the rural areas. 3.3 General favorable conditions Generable favorable conditions in Tanzania are: 1. A coherent vision on developments and trade policies. xxv The plan 2025 was very ambitious but it serves the purpose of mobilizing the people. Japan is still pushing the mini-tiger plan, which is based on the example of Malaysia. In that country local ownership, the role of local businessmen and indigenous investments were important. The National Strategy for Growth and Reduction of Poverty (NSGRP) is a good oversight of what the country does and wants to achieve. Tanzania really has a strategy for the future, in fact it has several plans. For example besides the plan 2025 and the mini tiger plan it tries to implement De Soto's formalization approach for the informal sector. The Tanzanians may have listened to too many people and they may try to do too many things at the same time. However, the result has been higher economic growth and an attractive, because more investors-friendly investment climate. Tanzania still faces constraints like insufficient infrastructure and not enough attention to technical education, but the intentions are there. The government has developed a private sector development strategyiii and it will also have a Small and Medium Enterprises (SME) policy and has already a trade policy, but these policies are not always implemented. 2. More attention for decentralization (see above) 3. More foreign direct investment (above) 4. More property tax collected (see above) 5. More joint ventures with US and Chinese companies are promoted. Tanzania woes Chinese investors and although these are mainly interested in selling their products and buying raw material in Africa, the last mission made some promises. According to the Guardian (24-8-6) the vice president has pledged government support to investors trying to realize their goals. The Chinese answered that they seek investments in business industry, minerals, agriculture and irrigation and has helped Tanzania with the development of an Export Processing Zone (EPZ). It is not so clear why China would invest in Tanzania other than for raw materials. You cannot expect them to play the role that Japan has played for many East Asian tigers. Technical education has also been neglected, meaning that companies have to train their own people. This is a factor increasing the costs of foreign investors. To achieve more joint ventures with the US a trade mission to the US is planned and Tanzania has achieved collaboration with the Brooklyn Chamber of Commerce (BBC) to promote investment and trade ties between the US and Tanzania (Capital, October 1, 2006). The BBC director announced projects ranging from manufacturing, horticulture, infrastructure to services and the investors hope to benefit of the opportunities offered to them under the African Growth and Opportunities Act (AGOA). 6. The government stimulates universities to pay more attention to improving `business skills' of students and staff xxvi Special programs have been developed at Dar Es Salaam University and also several institutes have specialized in providing appropriate technology for these self-starters. 7. Several attempts are undertaken to improve market access as well as productivity. A number of activities envisage improving competitiveness. Tanzania's Business Environment Strengthening project (BEST) for example is providing support to private business organizations in Tanzania. They can get grants to advocate for changes in the business environment. Money can also be obtained to allow such organizations to improve their capacity to advocate more effectively for changes in Government policy, regulations and so forth. 3.4 Special factors promoting regional and rural development Some of the factors that favor in particular the rural areas are: 1. An improved infrastructure (above) 2. Many rural development initiatives (NBS, 2005) 3. Activities to improve the competitiveness of agriculture (see above) 4. Efforts to extend the micro credit network (already mentioned) 5. Indicators showing increased activity in the rural areas These indicators range from the sales of telephone cards to consumption of construction materials. The impression is that in the rural areas a lot of phone cards are sold and everybody is calling. That is a good sign that people are connected. The number of buses would also be a good indicator. All the consumption goods that one finds in the urban areas can also be bought in rural areas. Decentralization has provided an infrastructure. There are local multipliers for all these expenses. 6. A trend to diversification For example the rise of Jatropha biofuels (Van Eijck, 2006) or the surge in horticultural investments. 3.5 Factors which could constrain the expected dynamic development Some factors which could constrain the expected dynamic development 1. Increased corruption; 2. Continuing to hinder entrepreneurs. There are lots of examples ranging form the gold mines to the small entrepreneurs who have to move their businesses out of the center of Dar es Salaam or Arusha; 3. Socialist-oriented land policies (Citizen 25-8-06); 4. Conflicts between different population groups. We realize that we may have left out certain negative and positive aspects of rural areas in this conceptualization: xxvii a. The environmental impact, for example as measured by the footprint approach; b. The criminal relations: smuggling, illegal woodcutting, prostitution, etc.; and c. The recreational aspects of staying in the rural areas, to the extent that this is not measured as a service. Many may prefer to go there for a holiday or retirement. 4. How can Tanzania benefit more from these flows? How can Tanzania benefit more from these flows? Alternative model: use informal sector to further spread development. Decentralization provides the opportunities; the challenge is now to design mechanisms to get capital and entrepreneurial talent to rural areas. The BEST project with the Tanzania Private Sector Foundation will be taken over by the World Bank, which will be using this structure to develop SMEs. There is 100 million dollar available, which is a strong increase in budget from an organization that currently handles 25 to something like 60 to 70 million dollar per year! September 2006 there was a mid term review of the BEST project. It showed that to adjust laws and achieve all the objectives of the project really takes time. The Better Regulation Unit in the Ministry of Planning had to fight for its position and the chief officer has been replaced. So there are not as many concrete results as hoped. So far the building is ready for something bigger with a more concrete output. The advantage of working with the Tanzania Private Sector Foundation is that the organization manages to speak with one voice in the negotiations with the governments. The World Bank will use this structure for an enterprises development program, but only a part of it is for SME. US$ 40 million will go into developing a proper business environment, dealing with things like registration: land, your company, etc. US$ 10 million is for access to finance and US$ 50 million goes to enterprise development and is mainly matching existing initiatives. There are about three or four serious NGOs in the micro finance sector in Tanzania. The National Micro Finance Bank may be one of the banks to be supported. The World Bank supported project wants to give matching grants to companies that try to be competitive. It will not select markets or take the entrepreneurs by the hand. Rather give them a grant if they want to export, or if they need information about standards for flowers. It would be co-financing to assure that they are really interested as well. For the Tanzanian bureaucracy it is important that the president is in favor of small enterprise development and decentralization. Many officials still see enterprises only as units to be taxed. There is a need for more non-Asian black entrepreneurs, but it is all a question of networks, which are more difficult to access for an African entrepreneur. Cultural diversity is generally considered an asset for a country or a city. In Tanzania it is also important to maintain the multicultural character of the society. Even if this means that black entrepreneurs need preferential treatment, if only because entrepreneurs from Europe or Asia have better access to business support services (BSS). xxviii For the SME the choice between SADC and EAC is not an important issue. Tanzania can do The Tiger model, it is just a question of increasing the efficiency levels. That requires that they continue to improve the business climate. It is possible to think in terms of an alternative approach to the urban rural dynamics. This approach would facilitate these flows and try to maximize their positive effect, just like the government is trying to maximize the effects of a gold mining company for the rural areas where it is working. Stimulating the flows would mean: · No longer local economic development, but linking local productive activities to global formal/informal value chains! · Not: more protection because it will not work, but anticipating what globalization will bring; · Rather find out where Tanzania can be competitive, for example in sectors like horticulture, mining and tourism; · Develop literacy and professional skills; · Stimulate entrepreneurship; · Promote innovation in agriculture and in off-farm employment (non agricultural activities); · Reinforce rural urban relations; · Increase the purchasing power of the people .in rural areas Some policy recommendations can be formulated: 1. Tanzania should move from an improved business climate to entrepreneurship development in the urban and rural areas, for small entrepreneurs and farmers The recommendation is to support the current reform process and in particular instilling the confidence and skills in small entrepreneurs and farmers and supply them with ideas how to make a living. This requires involvement of the private sector, of NGOs and organizations of producers themselves. 2. It should maximize the local effects of investments Finally the question is how can the rural areas benefit even more from these urban-rural flows? The challenge is to maximize the positive interactions between the urban and the rural areas. Sectors like flowers and horticulture, mining and tourism need to be stimulated to maximize the local impact. However, it is not desirable to go against the effects of globalization. Hence the European countries should try to re-activitate the Doha round and turn it in to a real development round, as was promised at the beginning. The following table gives an impression of local procurement of goods and services by mining companies in Tanzania. It gives an illustration that the current policy of increasing the impact of these industries for the rural areas can be successful. The local expenditures have increased since 1999, but could increase even more in the near future. The figure has quadrupled in 6 years, but the Tanzanian government has recently renegotiated the contract with the owner of the gold mines (Citizen, 23-8-2006: 1-2) to assure a greater impact of the mines in the regions where they work. xxix The results are that in the first place the company will pay the government annually US$ 7 million before starting paying corporate tax. Secondly, the company pays each of the relevant district councils an annual US$ 200,000. The company assured the President that they would now buy as many locally made products as will be available. It had been calculated that the firm sourced over seas for an amount of US$ 655.4 million during these six years. The firm will pay 30 percent corporate tax and improve the relations with local small miners. This is an example of maximizing the local effects of investment! Table 7 Local procurement of goods and services by mining companies in Tanzania Year Local procurement of goods and services 1999 US$ 36.1 million 2000 104.7 2001 109.5 2002 125.4 2003 146.2 2004 133.5 3. Cities can be engines of development for the rural development process Many business support services, new ideas, technologies and expertise are located in cities. The challenge is to strengthen the spontaneous process of flows between urban and rural and maximize its potential positive effects. Coming back to Jacobs (1970) these cities provide ideas, products and markets to the rural areas and in this way can contribute to their development. It would not be wise to isolate the two systems too much. 4. More positive effects of the flows requires change in policies, attitudes and investment: · Policies: for example improve purchasing power through off-farm employment in labor intensive rural public employment schemes to build required infrastructure · Attitude: stimulate decentralization, promote entrepreneurship and facilitate flows · Investments: more in education, health, infrastructure, micro finance and innovation 5. Promoting the urban informal sector 5.1 Introduction The urban informal sector also plays an important role and on top of that has a tremendous potential to spread development. We have to look for those activities that can grow out of informality when they reach a certain level. If asked many small entrepreneurs claim a lack of finance is their major problem. The question is whether loans are the real bottleneck. Other problems for micro and small entrepreneurs are a lack of space, infrastructure, market outlets, skills and of innovative capacity. Current policies of the government for the informal sector and small enterprises in Tanzania include: xxx formalization policies 5.2, credit or loans 5.3; export promotion 5.4 and other SME policies 5.5 The new president is very much in favor of local small businesses, so SME development it is a political priority now. From 2004 the small entrepreneurs can get a license fee without paying. Also the taxation system is more smooth and predictable. Tanzania finds it difficult to compete in an international context, where markets are flooded with cheap Chinese products. This point leads to the suggestion to really look where the country's competitive advantage is. Tourism, horticulture, growing fruits and vegetables and developing resource based industries (like processing of Tanzanite) are more likely sectors than exports of manufactured products, or advanced services. 5.2 Formalization process In order to empower informal sector, the government has prepared the Property and Business Formalization Program (PBFP). The Program seeks to facilitate the transformation of informal (extralegal) properties and businesses into formal legal entities. Its aim is to enable Tanzanians to use the formalized resources as collateral in accessing credit from financial institutions and thereby enhance economic growth and ultimately reduce poverty. This requires according to De Soto (1989) title deeds. However, the question is what would be the value of such a deed in a village? Does Tanzania have the legal system to allow financial institutions to sell such property in case of non-repayment of loans? The formalization work has already started in Dar es Salaam. The Business Registration and Licensing Agency (BRELA) has been created. It was established in 1997 and launched at the end of 1999. The Agency is responsible for business facilitation and regulation. Its mission is to regulate and facilitate business operation in conjunction with other partner institutions in Tanzania, to ensure that they are in accordance with sound business and commercial principles. Their slogan is that BRELA provides business legality in Tanzania. Besides registration they are also granting patents, overseeing copyrights and involved in business and industrial licensing. Mainstreaming informal enterprises into formal sector faces some challenges, which include: inadequate knowledge of the dynamics of the informal sector, and inadequate strategically located areas for allocating space to the informal sector operators. The PBFP program is currently limited to the first of two phases, one of Diagnosis. The second phase of Reform design will start when the evaluation of the first phase is completed. In Tanzania there is a lot of talking about formalization of the informal sector. In practice this does not mean much more than abolishing the license fee if your revenue is below 20 million Tanzanian Shilling (TSh). Secondly, small entrepreneurs don't have to renew the license, like in the past. If the revenue is over 20 million, they pay 20,000 TSh, but only once. For tax purposes, if an enterprise doesn't have records, it pays a lump sum. If it keeps records, it pays a percentage. xxxi 5.3 Credit or loans Loans in Tanzania go to large local firms (55%), multinational firms (without mining, which is self supporting, 20%), SMEs (10%) and consumers (15%). Agricultural gets only 10 percent of this the rest going to trading and manufacturing. A revealing figure is the growth of the loan portfolio. With an economic growth of 6 percent, credit is growing at 25 percent per year. There is no mechanism in place to provide agricultural credit. This is a reason for some people to argue in favor of an agricultural bank, but the bankers we talked to consider this is too interventionist. Micro finance is not very much developed in Tanzania (Gallardo et al., 2005). Micro credit can play an important role for that micro and small enterprise development in the future. This is not the way the formal sector banks will go, but there is scope for alternative micro finance institutions in Tanzania to expand their business. Despite its name, the National Micro-finance Bank (NMB) is for example not very much into lending to small and micro entrepreneurs at the moment. Investing by Tanzanian banks in risky business is not encouraged because of the high rate of interest paid on treasury bills (9 to 10 percent and risk free, while inflation is only 4 percent). Micro loans is currently not interesting for the formal sector banks, because of the cost involved in lending to one entrepreneur would be a few hundred dollars. With a saving component it may become interesting in the future. Savings in the informal sector are not registered because they don't pass through the formal banking system. Again micro finance and saving schemes could make these amounts visible. 5.4 Export promotion Several efforts are made to promote exports by small and medium enterprises in Tanzania. The Ministry of Industry and Trade with support from Denmark (DANIDA) focuses on a limited number of medium scale companies, which are willing and able to export and consider new markets (outside the region or Africa). The project with Danish support helps these enterprises to overcome the barriers to export and to focus on quality requirements, helping them to satisfy international standards. For practical reasons the project focuses on a limited number of cities. The project does provide grants to help Tanzanian companies to export. However, they don't deal with the informal sector, but there is some kind of continuum between informal micro and formal SME. They consider the main bottlenecks in the sector are regulatory and administrative constraints, taxes and compliance cost. Although the administrative system is decentralized, entrepreneurs often have to go to Dar es Salaam for tax purposes or to obtain an import license. 5.5 Other SME policies Innovation is not a big thing in the small enterprise sector in Tanzania. A lot of entrepreneurs are mainly copying what is happening elsewhere. Also incubator centers don't exist yet. Still there are many Tanzanian producers in sectors like cloth, vegetable oil, soap, etc., which could benefit from innovation or incubation centers (Bongenaar and Szirmai, 2003). Van Dijk and Sandee (eds, 2003) suggest using the term innovation in the broad sense of introducing new ideas, new products and ways of working or financing or xxxii selling the production of micro and small enterprises. The Technology Development and Transfer Center (TDTC) at the university of Dar es Salaam now produces sieving machines for small scale mining (Sunday Observer 7-7-2005) and could be used for developing/adapting technologies for other activities as well. Although most of these informal sector enterprises are small, they trigger local economic development. There are also export associations in Tanzania, even if these sectoral institutions are not always called export associations. For coffee, horticulture, sisal and mining such associations do exist and could be used to increase exports from SMEs as well. Also Trade Fair sales can help SMEs. The ILO promotes developing the urban informal sector in East Africa by promoting subcontracting of solid waste collection and treatment to small enterprises (Mitullah, 2005). Preparing training manuals is part of their approach to the issue. Local organizations develop the material. Other ideas mentioned in ILO (2004) to promote the informal sector are: a new pact between Central government and small businesses; improving their access to public procurement; development of entrepreneurship and more efficient institutional arrangements 5.6 Conclusions on the promotion of the urban informal sector The lack of finance for small and medium enterprises (SMEs) and micro enterprises in the informal sector could slow the growth of these activities and hence the economic growth of other cities and regions. The process of formalizing the informal sector, which has started in Tanzania under the influence of De Soto (1989), has some dangers. One is that it will kill the sector, whose secret is that they are not paying all the taxes and satisfying all the rules and regulation developed for the formal sector. Secondly, it is noted that the large number of activities currently undertaken for this sector may stimulate people to actually go to the cities and try to benefit from these programs. Hence it could increase rural urban migration and should be organized in rural areas and small towns. Finally there will always be an informal sector, even if you make registration easier and paying taxes more transparent and justifiable. The dynamics is often new people starting economic activities without bothering to register or to pay taxes. Most informal sector activities don't believe in formalization and hence they want to stay out of the tax and credit web. However, the dynamic part of the informal sector will eventually make the transition. The approach of De Soto, giving everybody titles, is not a sufficient condition for successful small enterprise development. A lot more (education, training, markets, innovation) is necessary to turn the small and micro entrepreneurs into dynamic medium size enterprises. Another point of departure for the promotion of the informal sector would be to take the real problems of these small entrepreneurs as the point of departure. In Trip (1997) one finds a number of positive and negative factors, summarized in table 8. Micro and small businesses in Tanzania face a lot of competition and opportunities in the global economy. They have undergone a process of adjustment, with positive effects for regional development and poverty eradication. However, further development of these activities requires a friendly national and international environment to be able to survive xxxiii and to develop their businesses. The authorities have at least put in place a lot more positive regulatory framework and trade negotiations put more emphasis on the role of these activities. If the contribution of these activities to the economy is really recognized and if the informal sector is more systematically promoted it really help to spread development over the country. Table 8 Positive and negative factors for the urban informal sector Positive Negative Rotating credit societies 117 Licensing restrictions 103 Networking strategies 122-24 Oppressive laws 103 Women taking control over their own lives Harassment 124 117 Battle over grounds 158 The moral economy model 127 One can add: taxes, lack of space, lack of Ending enforcement policies 193 credit, etc. Source: Tripp (1997, with page numbers). 6. Conclusions: prospects for sustained growth Dar es Salaam is not a global city in the sense of Sassen (1991). Such a city would have an important services sector and commands value chains. However, even if cities don't manage to become global cities they may still be important as peripheral nodes in the world capitalist system (Van Dijk, 2006b). This seems to be the case of Dar es Salaam, the capital of Tanzania. The city serves a large hinterland and can reinforce easily its regional function in East Africa. Good urban management plays an important role in these processes. There are other policy reforms necessary, like pushing decentralization even more, spreading infrastructure, services and promoting the role of the private sector outside the capital. An alternative model of development suggested for example by the ILO (1972) and De Soto (1989) would be to develop the informal sector further and use is as a mechanism to spread development These units are not always registered, do not pay all the different taxes or the legal minimum wage, but they certainly make up a local sector, creating a lot of employment and with a lot of entrepreneurial capacity. Tanzania should be tapping into it. Most small-scale activities are still producing for the domestic market and some for informal exports. The availability of finance is by the formal sector bankers in Tanzania not considered the bottleneck. However, money and institutions are available; the issue is to design mechanism (micro finance and guarantee arrangements) to get the capital into the rural areas and small towns and to reach the farmers and the small entrepreneurs. Different levels of government can also mobilize more money through partnerships with the commercial and non-commercial private parties. There are several actions necessary to sustain the growth of Dar es Salaam. They range from improving the infrastructure to attracting more economic activities. The challenge is xxxiv to demonstrate how the key reforms ­ macro stability, liberalization of prices, exchange rate, and interest rates, privatization, etc. had a different impact on Dar es Salaam, other urban and rural areas. The conclusion is that this may not yet be fully the case, but it is important that the ideas float and they inspire a certain confidence: we can do it! Trust in a country, its government and its economy is an important factor determining its chances to develop in Tanzania. xxxv URBAN RURAL DYNAMICS IN TANZANIA, THROUGH INFORMAL REDISTRIBUTION MECHANISMS 1 Introduction Tanzania has witnessed a fairly good macro economic performance of over 4 percent per year since 1998 (6.2 percent economic growth during 2002-04), but the impact on poverty seems limited. Only in Dar es Salaam there was clearly observable poverty reduction (World Bank, 2005). The issue addressed in this study is: why did poverty alleviation happen in Dar es Salaam and hardly in the rest of the country? Secondly, what are the transmission mechanisms distributing the positive effects of economic growth over different cities and regions? Finally, how does this economic growth affect the poor in other cities and rural areas? In chapter 2 the context of economic reforms is described, before exploring different explanations of Tanzania's success in the chapters 3 to 9. 2 The context of economic restructuring and urban and regional growth 2.1 Introduction After years of structural adjustment, liberalization, poverty eradication policies and programs and donor support, Tanzania has created a healthy macro economic context in which growth can take place. An overview of the liberalization policies in Tanzania is given in annex 1. The Tanzania Private Sector Foundation (TPSF) considers, that now the investments in reforms in Tanzania start paying off. The country has embarked on structural adjustment programs in the 1990s and is currently reaping the fruits. Tanzania belongs to the top of African pro-business reform countries. According to the Financial Times (6-9-2006) it simplified business regulation, improved property rights and made it easier to start an enterprise. This information is based on a World Bank IFC report, which found that the cost to register a new enterprise reduced 40 percent. Through a change of the legislation small investors also receive better legal protection. Importing goods has been easier by using a computerized system. The average number of days to import goods went down from 51 to 39 days. For export the average went from 30 to 24 days. Finally the taxes to be paid for this kind of transactions went down 3 percent. 2.2 Economic success of Tanzania After independence in 1961 president Nyerere soon embarked on socialist policies. The government of Tanzania nationalized and placed all enterprises under the state ownership and management in the mid 1980s. These policies had a negative impact on private sector development in the country. In 1992 the government launched the Parastatal Sector Reform Program (PSRP) to privatize public enterprises. Privatization envisaged to give private sector a frontline role as the engine of economic growth. Ever since, the participation of the private sector has increased and management of the restructured economic activities in the country has improved, particularly in the capital, the city of Dar es Salaam. 1 To strengthen the business and investment environment, consultative dialogue meetings took place involving the government, private sector under the auspices of the Tanzania National Business Council (TNBC). Moreover the government continued to implement the Business Environment Strengthening for Tanzania (BEST) Program. As an example of the economic success of the reform policies, the Tanzania Revenue Authority (TRA) collected 25 million US$ per month in 1995. Ten years later this is 160 million! With about 200 million the government could cover its budget instead of being dependent on foreign donors, who currently finance about a quarter. This effort is even more remarkable, since in general import duties have gone down as a consequence of the Uruguay round of the World Trade Organization (WTO). Currently the Tanzanian customs use only two rates (25 and 10 percent) and are moving to a single rate. Local governments have also done better (Heymans and Kumar, 2002). They collect the development levy, land tax and a cess levy on the value added tax (0.3%), while for national government the income, consumption (VAT of 20%) and import duties are very important. The Ministry of Finance is responsible for tax reform. The tax system has also been made slightly easier, by abolishing a number of taxes or taking them away from local government and by providing an income transfer to local governments instead. The authorities also removed the stamp duty, but it came back under the income tax. The value added tax is a hot issue because the rate is considered too high by the businessmen. They don't see that it is deductible. The shocking figure about taxes is that 82 percent is collected in Dar es Salaam and only 18 percent in the rest of the country! The challenge is to strengthen revenue collection and not to exclude too many cases. The 2005/06 budget estimates that the government will collect TSh. 2.067 trillion from domestic sources, an equivalent of 14.3 percent of Gross Domestic Product (GDP; Citizen,12-7-2005). The role of private and partially foreign investments in booming sectors like mining and tourism is also important. Just like the steep increase of the horticulture sector this is driven by foreign direct investment. Newspapers made a lot of noise about the US$ 470 million investments in the country in 2004, but this is a fraction of what for example the Netherlands invests in Brazil or the west invests in China (US$ 60 billion in China in 2005). It is not so clear why China would invest in Tanzania other than for raw materials. You can not expect them to play the role that Japan has played for many East Asian tigers. Technical education has also been neglected in Tanzania, meaning that companies have to train their own people. Again a factor increasing their costs. The World Population report (UNFPA, 2003) notes that urban areas have become engines for growth in the global economy as centers for diversity and change. Is this happening in Tanzania as well? Dar es Salaam is certainly very different from the rest of 2 the country, since most of the industries of Tanzania are concentrated there. The city has the biggest population and is the biggest market. The port also plays an important role in the concentration process, as will be explained in chapter 4. The creation and development of secondary towns is a strategy to spread development, but does it really work in Tanzania, before the reform (when an equal distribution was the official policy), or ever since because there is some economic growth to distribute now? The issue will be discussed in chapter 5. Agriculture is largely rain fed and contributes about 46% to GDP. The sector is very prone to droughts and only 191.000 hectares of land can be irrigated, which represents one fifth of the total areas which could be irrigated (World Bank, 2006a). The recent irrigation sector master plan aims to double the irrigated area in 10 years. Currently the on-farm water efficiencies are also very low (between 10 and 20%). Besides gold and tanzanite, Tanzania has an emerging horticulture sector and more traditional crops such as sisal, coffee and cotton. Problems are that the current financial structures are completely against agriculture, production takes place far from the coast and the infrastructure and the technology are often rudimentary. However, there is no lack of entrepreneurship in the rural areas, which is proven by the fact that households engaged in off-farm activities are less poor. 2.3 How does growth affect the poor? Good macroeconomic information is available since some ten years in Tanzania. However, little is known about the micro implications of economic growth for the poor living in different cities and regions in the interior. The existing inequalities come out very clear in the Household budget survey 2000/01 (NBS, 2002b). The question is what can be done to spread welfare, rather than trying to stop it from staying in the capital, where a lot of it is generated.. The World Bank (2005) Country economic memorandum provides an in-depth analysis of the Tanzanian economy. However, it deals mainly with what we can be measured in Tanzania. A number of the interesting issues in the development process of a country, region or city are difficult to measure. For example: 1. What explains the dynamics of cities like Dar es Salaam and Arusha? 2. Which factors contribute to poverty reduction in the urban and rural areas? 3. Why did poverty alleviation happen in Dar es Salaam and hardly seems to happen in the rest of the country? 4. What is the role of migration, the urban informal sector, remittances and smuggling to redistribute the positive effects of growth? 5. To what extent has decentralization contributed to urban dynamics and how does it affect poverty? Table 1 Poverty reduction between 1991 and 2001 Variable Dar es Other Rural Comments Salaam urban 3 Poverty From 28 to From 29 to From 41 to National level -3% reduction 91-01 18 % 26 % 39 % Average household is smaller Average hh size 4.8 4.9 5.9 in Dar Working age (15-62.9% 57.0% 50.3% Largest share working age in 64) Dar es Salaam Younger people 25.6% 22.2% 18.6% More in 15-44 range Importance 3.5% 3.2% 2.1% 2.4% of hh budget remittances It is difficult to say whether poverty did or did not decline in the rural areas. There are no figures except two national household surveys. The most recent one may have measured poverty in 2000/1 at the wrong moment (during a small dip in the economy) and hence poverty eradication efforts don't show (NBS, 2002a). One can also argue on methodological grounds (the one dollar a day, or the level of the poverty line) against the current picture. One can also say that rural and urban poverty are different and fixed at too low a level.iv The valuation of subsistence as the biggest problem. The National Bureau of Statistics (NBS) currently used existing prices, but should use the opportunity cost according to him. He also stresses that there are big differences between different rural areas. Finally, a poverty line analysis may neglect the dynamics of poverty, as was the case in China and India (Van Dijk, 2006c), where research has shown that it is often different people who are poor over time. Poverty is then related to a phase in people's life (transitory poverty), or people are growing out of poverty by undertaking certain economic activities, often in first instance activities which would be classified as urban informal sector. A drop in poverty can be seen in table 1, in particular in Dar es Salaam, but also in the other urban areas and even in the rural areas. Sustainability and expansion of food and cash crop production in the Dar es Salaam region appears difficult to be achieved. Some of the existing problems and obstacles facing this sector are: poor farming methods and equipment, non existence of data about farmers and crop yield, great increase of people especially youths, high rate of urban expansion leading to decrease of agricultural land, land degradation due to soil erosion, inadequate knowledge on improving farming techniques by farmers, outbreak of crop pests and diseases affecting plants and crops, and shortage of inputs including better quality seeds and fertilizers URT (2005: 66) decides on the basis of figures of the Ministry of Finance that the disparities between the rural and urban population have increased, like in many other fast growing countries. Poverty is measured in two ways. Although basic needs poverty has declined, the absolute number of poor keeps growing every year, if the trends are analyzed for the period 1991/92 until 2000/01. However, the Economic survey 2004 (URT, 2005) when discussing the issue doesn't mention the role of the urban informal sector, nor does it give the differences between different regions. 4 In Tanzania statistics are often limited to the formal manufacturing and the agricultural sectors, but an important part of what is happening is not showing in the figures because it happens in the informal sector or `informally'. There are estimates of the informal sector going up to 90 percent (if non estate agriculture would also be included). A more realistic figure of 58% is mentioned in several World Bank publications. It means that labor is allocated, investments are made and raw materials are used without really being measured for half of the economy. It involves flows of people and money, not shown directly in the statistics, but known to many people. We have concentrated during this research on getting whatever data is available and on the stories about the urban informal economy and how it affects other Tanzanian cities and regions. These mechanisms are described in more detail in chapter 5 and 6 on the informal sector. The consequences for Dar es Salaam, the Kilimanjaro and Arusha region are analyzed in chapter 7 and 8, before putting Tanzania in the context of the globalization process (chapter 9) and drawing some conclusions and formulating some recommendations (chapter 10 and 11). 2.4 Explanations for the spread of growth in Tanzania There are different explanations possible for the fast or slow growth of certain regions and cities. Growth could be the result of different theoretical explanations 1. Tanzania benefited from economic restructuring and is now ready to become an African tiger. This view will be reviewed in section 2.5. 2. There is no success at all, just special circumstances (section 2.6) 3. Higher levels of regional development is the effect of decentralization. This explanation is discussed in chapter 3: the positive effects of infrastructure built by different levels of government and of the expenditures of local governments. Regional budgets and tax revenues give an indication of how much is spent in different regions. However, there is also direct central government support to regions, for example by setting up offices in specific regions. 4. There is currently not the infrastructure that is necessary. Once the infrastructure is in place, development will spread easily. This explanation will be discussed in chapter 4. 5. The role of the urban informal sector in spreading economic growth may be more important than what would be expected from the available statistics. The lack of finance for small and medium enterprises (SMEs) and micro enterprises in the informal sector could slow the growth of these activities and hence of other cities and regions, where this is the case. The current mechanisms for urban-rural transfers will be reviewed in chapter 5 and the role of the informal sector in chapter 6. 6. The role of private and partially foreign investments in booming sectors like mining and tourism. These sectors can contribute to the growth of regions or cities through the multiplier effects of these expenditures. For example, the companies and projects use a lot of facilities in Dar es Salaam. Also large amounts of development aid are spent and bring about development in those regions, which are lucky to get it. However, the exact links between foreign direct investment (FDI) aid and regional development in Tanzania are difficult to establish. It is difficult to assess where exactly the FDI and development aid is spent. This 5 explanation will be discussed in the chapter on Dar es Salaam (chapter 7) and on the Kilimanjaro and Arusha region (chapter 8). 7. Tanzania will find it difficult to compete in an international or global context, where markets are flooded with cheap Chinese products. This point of view will be studied in chapter 9, before drawing some conclusions in chapter 10 and formulating some recommendations in chapter 11. 2.5 The Tanzanian tiger The Financial Times (3-8-2005) called Tanzania "A rare example of African success". It immediately added that "The country has done most things right but remains mired in poverty". In Tanzania the business climate has greatly improved during the ten year president Benjamin Mkapa was in power and the last three years the economy grew with 6.2 (2002), 5.7 (2003) and 6.7 percent (2004; website Bank of Tanzania/NBS, 2005). However, the Tanzanian population also grew with 2.9 percent per year.v This means that a reasonable economic growth of 6.2 percent during the last three years translates into a per capita growth of only 3.3 percent.vi On top of that economic growth is not distributed equally. Dar es Salaam, Tanzania's capital, clearly functions as a growth center, while some other regions are still stagnating. The country has a strategic view in Development vision 2025. Tanzania wants to be semi-industrious by 2015 and the country has developed a "sustainable industrial development policy", which may be a good basis for such a future. The role of the manufacturing sector should go from 8 to 15 percent of GDP by 2020. More economic growth requires upgrading technology, human resources development and reforms at the local government level. So far most of the reforms concentrating on the government at the national level. There is also an ambitious economic development program dubbed "Tanzania Mini-tiger plan 2020" (The Citizen, 11-7-2005). It is being implemented and aims at fast-tracking the achievements promised in the Development vision 2025. The minister of State in President's office (Planning and privatization) has told Members of Parliament (MPs) in Dodoma in July 2005 that the new program seeks to replicate economic successes of South East Asian countries. It complements the ongoing second phase of the National Poverty Reduction Strategy (NPRS). The Minister conducted a seminar for MPs together with experts from his office and dr. Shoichi Kobayashi of the Japan Development Institute (JDI). According to The Citizen (11-7-2005) the program aims at increasing the national per capita income from US$ 286 to 1000 and at creating 2 to 3 million new jobs by 2020. The Minister noted that the project involves all sectors of the economy and requires quick decisions. That is why it operates under the President's office according to the Minister. It may be somewhat early to call Tanzania an African Tiger, but it is certainly interesting to study the growth potential (World Bank, 2005) and to indicate the factors that may contribute to a better regional distribution of the effects of economic growth. Macroeconomic policies provided the incentives for the development of sectors like mining and tourism. In the urban areas liberalization had already more impact, also on the 6 informal sector, than in the rural areas where the opportunities were not always picked up by local farmers and small entrepreneurs. Government officials too often see the private sector mainly as a milk cow and not as a partner for example in a private public partnership (PPP). In the same way the private sector considers the government as mainly imposing taxes. The government doesn't seem to have an eye for their needs. Just removing a number of distortions for micro, small and medium enterprises will have very positive effects on economic development. However, in the rural areas the cooperatives were removed, but the private sector did not always step in. Development requires institutional development and an elementary level of infrastructure and still many distortions exist in the rural areas. In several regions there is land scarcity and the resulting land problems point to the need for land reform. Although policies are being put in place, there is no market yet for land, because the policies were not there. The current financial structures are completely against agriculture. The agricultural sector is still very dependent on rain and no bank is willing to risk its money in the current situation. Financial experts interviewed do believe a sustained growth of 6 percent for Tanzania is possible and the banks want to contribute to it. The director of one major bank said: "We want to crack the rural-urban thing". They are looking how other countries are doing it. They don't have the deposit base to go into rural lending on a large scale, but are looking for possibilities to decentralize their operations. The Malaysian model for rapid economic development is pushed in Tanzania, which emphasizes the role of different ethnic groups, of nationalism and strong leadership. It comes with a lot authoritarian traits, which are not easily accepted in Tanzania. The culture there is very participatory (the economics of affection called Goran Hyden it in 1980), which means people are not solitary `I go for profit for myself' characters. However, Tanzania is also a multi cultural and multi religious society and that can help development (diversity as a source of urban dynamics: Van Dijk, 2006a). Also the social networks are very important. Many women and young people build on these networks to advance their careers. 2.6 No success, just increased prices of commodities and donor darling Some argue that there is no success, prices of commodities have increased and that means the values have increased, but not the quantities produced, while the government and donors spend a lot of money in Dar es Salaam, which stays there.vii A lot is probably spent in the centre and there is a clear link with how development is managed. Jacobs (1970) concludes that the relation between the city and its hinterland can be very fruitful, if properly managed with the urban manager as the key actor. It is true that the government and donors spend a lot in Dar es Salaam. However, does the fact that the Government and donors spend a lot in Dar es Salaam mean that the other regions get no chance to develop? The centre is driving development. Much depends on the mechanisms to transmit the positive effects of these expenditures and investments. It is also true that the prices of commodities have increased substantially between 2002 and 2007. The following table gives an indication of some of the trends, but volumes have 7 also increased. 2002 was a bad year with low commodity prices. Since then tea, cotton and tobacco went up, but now coffee is going down again. Among traditional exports cotton increased by 52%, tobacco by 48% and cashew nuts by 28.6% in 2004, due to favorable weather conditions and timely availability and proper usage of agricultural inputs. In some cases the purchasing price for the farmers should increase to get more money into the rural areas. Table 2 Contribution of different sectors to export in 2003 and 2004 Year Traditional exports Non traditional Total value exports 2003 US$ 220.5 million US$ 908.7 million US$ 1.129.2 million 2004 US$ 292.3 million US$1,042.6 million The value went up An increase of An increase of by 18.2% between 32.6% 14.7% 2003 and 2004 Share 21.9(19.5% in 2003) 78.1 (80.5% in 100 2003) Source: Economic survey 2004. In the category Non traditional exports in particular the price of gold increased substantially in the world market. This commodity alone contributed 47.2%. In 1998 non traditional exports were only 232.2 million. The year ending June 2006 total exports on an annual basis amounted US$ 1.736 million, representing an 8.9 % improvement over the previous year (Guardian, 26-8-2006). Tourism and gold account for 51% of total exports. The improvement was due to improved prices and export volumes, in particular the volume of cotton and cloves exported increased substantially. Of course the positive effects of the increased commodity prices were partly offset because Tanzania also had to pay more for its imported oil. However, the country is not facing a major balance of payments crisis, although the value of Tanzanian shilling was somewhat slipping during the summer of 2006. The following table gives an impression of local procurement of goods and services by mining companies in Tanzania. It gives an illustration that the current policy of increasing the impact of these industries for the rural areas can be successful. The local expenditures have increased since 1999, but could increase even more in the near future. The figure has quadrupled in 6 years, but the Tanzanian government has recently renegotiated the contract with the owner of the gold mines (Citizen, 23-8-2006: 1-2) to assure a greater impact of the mines in the regions where they work. Table 3 Local procurement of goods and services by mining companies in Tanzania Year Local procurement of goods and services 1999 US$ 36.1 million 2000 104.7 2001 109.5 2002 125.4 8 2003 146.2 2004 133.5 The results are that in the first place the company will pay the government annually US$ 7 million before starting paying corporate tax. Secondly, the company pays each of the relevant district councils an annual US$ 200,000. The company assured the President that they would now buy as many locally made products as will be available. It had been calculated that the firm sourced over seas for an amount of US$ 655.4 million during these six years. The firm will pay 30 percent corporate tax and improve the relations with local small miners. This is an example of maximizing the local effects of investment! 2.7 Conclusions Many experts interviewed don't believe in all these stories of Tanzania becoming a (mini) tiger economy. There is Japanese technical assistance supporting this idea. One of these experts visited the Arusha region and advised the Regional Commissioner (RC) what should be done. His advise is not very different from what the RC could have learned form the World Bank (2005) Country Economic Memorandum (CEM). However, the important thing is that these ideas float and they inspire a certain confidence: we can do it! In the end trust in a country, its government and its economy may be one of the most important factors determining its chances to develop. Tanzania has achieved economic stability. The challenge is now to spread the effects of the resulting economic growth. It will be argued that the urban informal sector plays an important role in this process. The sector generates flows of people, money, goods and services, which bring development to other part s of the country. If a more positive approach to this sector is taken, attention can also be given to the issue of rising its productivity and competitiveness. Labor, entrepreneurship and existing financial mechanisms can be further developed. 3 Decentralization and its impact on urban poverty in Tanzania 3. 1 Introduction: what to expect from decentralization? Tanzania has a history of centralization and decentralization trends. Prior to 1972 all sectoral ministries and the development planning process were centralized. In fact local governments were abolished in the early 1970s and re-introduced in 1982. However, a program for enhancing the effectiveness of local government was only introduced in 1994.viii Decentralization is an important part of political and administration reform, also in Tanzania. According to Lee and Gilbert (1999) in 63 of the 75 developing countries with more than five million inhabitants an active decentralization policy is carried out. The expectations of the political benefits include strengthening of the democratization process and improved management at the local level, but the most important one was the reduction of regional development disparities (particularly in the basic social services) so as to arrest rapid rate of rural-urban migration. 9 Tanzania gives more attention to decentralization. A meeting of regional leaders was held for the second time in 2006 and shows the authorities take decentralization more serious. The president himself addressed the meeting for two hours suggesting a code of conduct, which made it very explicit that a lot is expected from these regional leaders. Many countries are struggling with their central-local relationships. More power is now vested at the local level in Tanzania. The local people need to take the initiative. They should formulate their own priorities. They should make their own participatory plans and prepare programs, develop economic, social and environmental policies and generate more local revenues. One indication is the importance of return migration Decentralization policies usually provide economic opportunities for local governments and entrepreneurs (Van Dijk, 2006a). It is important to create the conditions for the diffusion of growth at the local level. This means the focus is on the enabling role of local government (Helmsing, 2000). Decentralization can help in three ways: a. It provides local people the opportunity to take all kinds of initiatives b. It makes available money for local investments because it generates local taxes. However, too many local tax collection can also work as a disincentive for the private sector c. It allows local governments to develop policies undertake more development activities The claims are generally difficult to back by empirical evidence. Many other factors play simultaneously and there is often a difference between how the rules of the game have been formulated and how the game is played (for Ghana see Laryea, 2006). In this contribution we will focus on the practice and outcomes and in particular on the effects of decentralization on poverty reduction. After a theoretical section, we go into more detail concerning the history of decentralization in Tanzania in section 3. In section 4 the local government reform process is studied, while in section 5 we take a closer look at the responsibilities of the different levels of government, before trying to assess what really happens in section 6. In section 7 an effort is made to test the different theories of decentralization, while finally in section 8 some conclusions are drawn. 3.2. The theoretical framework 3.2.1 Income transfer, services or employment creation Tanzania has undertaken important steps in the field of decentralization. In this paper we want to analyze what the effects have been for poverty in the different regions: to what extent has decentralization contributed to the economic dynamics of the country (growing more than 6% per year during the last three years) and how does it affect poverty in different regions of Tanzania? Poverty alleviation at the local level is possible in three different ways: a. Income transfers or targeted subsidies to local governments b. Developing services and infrastructure also for the poor and tailored to the needs of the poor 10 c. Creating employment opportunities for poor people increasing their purchasing power in this way 3.2.2 Deconcentration, delegation or devolution, monopoly or pluralism? We distinguish two main approaches to analyzing decentralization in developing countries, which can be combined. In the first place the Type-Function Framework (TFF) by Cheema and Rondinelli (eds, 1983) and secondly the Administrative Design Framework (ADF) developed by Cohen and Peterson (1999). Cheema and Rondinelli (eds, 1983) provide the following definition of decentralization: it is "the transfer of authority and responsibility for public functions from the central government to subordinate or quasi-independent government organizations or the private sector". Cheema and Rondinelli (eds, 1983) define five forms of decentralization: political, market, fiscal, spatial and administrative decentralization, which tend to come in certain combinations. The TFF analyses decentralization according to forms and types (deconcentration, delegation and devolution). By this approach, decentralization is classified by forms on the basis of objectives: political, market, fiscal, spatial and administrative. The combination of forms results in certain types of decentralization. The TFF argues that various combinations of the main forms of decentralization, political, administrative and fiscal decentralization result in types of decentralization - deconcentration, delegation and devolution.ix The ADF identifies `states' that describe how concentrated the roles of government still are: 1. Institutional Monopoly, or centralization, where roles are concentrated at the spatial center in an organization or institution 2. Distributed Institutional Monopoly, or decentralization to local level governmental institutions or private sector firms and organization through deconcentration, devolution, and/or delegation, but where roles are distributed spatially, but are still concentrated in one organization or institution 3. Institutional pluralism, or decentralization through deconcentration, devolution, and/or delegation, but where roles are shared by two or more organizations or institutions, which can be at the spatial center, distributed, or a combination of both. are: The ADF suggests that various combinations of roles in service provision result in an institutional monopoly, distributed monopoly or pluralism at centralized or decentralized levels of governance. A combination of these two frameworks provide a conceptual framework for the study (based on Laryea, 2006). It implies that central-local relations are shaped by the forms and types of decentralization. The distribution of roles determines whether service provision for example is mainly through monopoly or plural arrangements. The resulting institutional arrangements influence the performance of local governments in the provision of services. For example where pluralism emerges at the decentralized level of government, it will yield better performance than in the case of 11 distributed monopoly (Laryea, 2006). The underlying idea is that decentralized decision- making can contribute considerably to improving the provision of urban services.x The theoretical framework developed by Cheema and Rondinelli (the Type Function Framework, or TFF) is more structure oriented, while the Administrative Design (ADF) Framework suggested by Cohen and Peterson (1999) emphasizes the basic principles of administration (accountability, effectiveness and efficiency) and focuses more on the roles of different levels of government. 3.2.3 The autonomy of the urban/regional manager Van Dijk (2006a) evaluates decentralization from an urban managers perspective and stresses the importance of being able to formulate and implement at the local level economic, social and environmental policies. On top of that the urban manager needs to be able to generate the financial means necessary to implement the strategy developed in cooperation with the major stakeholders. These frameworks and the implicit theoretical relations will be used to analyze the situation in Tanzania. Van Dijk (2006a) suggests to check whether urban or regional managers can: a. Develop and implement local economic policies? b. Develop and implement local social policies? c. Develop and implement local environmental policies? And, d. Do these lower levels of government have access to the necessary financial means? Only if he or she has some autonomy with respect to these challenges at lower levels of government can an urban or regional manager be fully effective. 3.3. The history of decentralization in Tanzania Through the Decentralization Act of 1972, the Central Government delegated planning and implementation of development programs functions to the regional and Local Government Authorities (LGAs). The envisaged main objective of the decentralized system was to introduce bottom-up planning approach and achieve the following: - To promote economic activities in the regions and accelerate development efforts to the needs of the rural poor, - To reduce interregional development disparities particularly in the distribution of basic social services, and - To redress the imbalances between urban and rural development so as to arrest rapid rate of rural-urban migration. Since 1997/98 the Government devolved more powers to LGAs through reduction of central government presence at the regional level (The Regional Administration Act No.19 of 1997) and provides compatible resources to strengthen and enable the LGAs to assume full responsibility for socio-economic development. The decision to delegate powers and decision-making process to LGAs entails fundamental changes, which 12 demand central government to establish supportive frameworks and programs. In view of this, the government, under its Public Sector Reform Program introduced two major reform programs: the Regional Administration Restructuring Program (RARP), and the Local Government Reform Program (LGRP). Local governments collect the development levy, land tax and a cess levy on the value added tax (0.3%), while for national government the income, consumption (VAT of 20%) and import duties are very important. The Ministry of Finance is responsible for tax reform. The tax system has also been made slightly easier, by abolishing a number of taxes or taking them away from local government and by providing an income transfer to local governments instead. The authorities also removed the stamp duty, but it came back under the income tax.xi The value added tax is a hot issue because the rate is considered too high by the business people. People don't see that it is deductible. However, local governments have done a better job in tax collection (Heymans and Kumar, 2002). 3.4. Description of the roles of different levels of government We distinguish the roles of different institutions and look at their functions and responsibilities before assessing how the system actually functions. Often there is a discrepancy between the system on paper and how it works in practice.xii 3.4.1 Different levels of government Like in most countries Tanzania has the system of trias politica: there is the legislature, the judiciary and the executive branch. Central and Local Governments form the executive part. The executive party of the Central and Local Governments operates at various levels. The village level is a corporate body in Tanzania meaning they can vote by-laws; but the wards, which are just an administrative arrangement, are not. There are two political levels in Tanzania (the elected politicians at the national and the council level) and three administrative levels: the national bureaucracy, the regional and local level bureaucrats). The division between a council and a ward is only an administrative arrangement. 3.4.2 The national level At this level the President who is also the Chief Commander of the Armed Forces heads the State. The Vice President, the Prime Minister, Union Ministers and Ministers, who form the Cabinet, support the President. The Sector Ministries are headed by Ministers and are mainly responsible for policy formulation, planning and implementation of sector development plans. The new poverty program of the government has singled out malaria, tuberculosis and aids as priority diseases to fight. In the framework of the Millennium Development Goals maternal and infant mortality will also get some attention. However, the health systems in the districts are very important and the challenge is how the national level can provide incentives to the local level to implement such programs. 3.4.3 The regional level and below At regional level the Regional Commissioner (RC) is a representative of the President at this level. The Regional Administrative Secretary (RAS), who is the head of the Regional 13 Secretariat (RS), supports the RC. The RC oversees law and order and good governance.xiii The role of RS could be summarized as: - Interpreting national policies and plan guidelines to the various actors based in the region - Coordinating the planning and budgeting exercise in Local Government Authorities (LGAs) within the regions - Supporting LGAs to effectively and efficiently discharge their responsibilities and improving public service delivery. One notes the emphasis on law and order and interpreting national policies to determine their impact for the local level. However, also law and order and good governance are important and the RC is supporting the LGA to discharge their responsibility. Tanzania has rural and urban districts. The District Commissioner (DC) is assisted by the District Administrative Secretary (DAS), who heads the district administration. The DC coordinates and supervises all government functions at the district level which mainly centers on the maintenance of law and order. Before the re-introduction of LGAs, the office of DC was the center of development activities within the district. With the re- establishment of LGAs in 1982 and 1984 the role of the DC's office changed from focusing on development activities to the maintenance of law and order. The Divisions are the lowest organ under central government structure and are headed by Division Secretary. One of his/her duties is to assist the DC in ensuring the maintenance of law and order in the division. A geographical unit is given the status of city or village if they can organize themselves. But becoming a village also means having more autonomy. The regional secretariat is very small and has just a regulatory function. If the region would be strong, that would defeat the idea of decentralization. But the city should be strong. The Local Government is headed by the Minister of State responsible for Regional Administration and Local Government under the President's Office, which was created in year 2000. The major responsibilities of this President's Office with respect to local government are: - To prepare policies on regional development for the regions in general; - Management and administration of local government affairs at the national level, and - To co-ordinate the activities of the Regional Secretariats and Local Government Authorities in the country. There are 98 Rural District Councils (RDCs) and 21 Urban Councils (UC) in Tanzania mainland (2004). The two types of councils do not differ much with respect to functions, organizations and committee structures. The RDCs operates at three levels: District, Ward and Village. The general public elects the members of District and Village Councils after every five years. The RDCs are considered as appropriate focal points for planning, implementation evaluation and monitoring of development programs since local communities (and villages) are represented at this level through their elected 14 representatives. Also it is at this level where priority needs and constraints to development can be identified and solutions be formulated. In addition, sector departments are brought together at this level under the chairmanship of the District Executive Director. The setup is part of a planned process where the centre is trying to stimulate initiative at the regional level, but at the same time tries to control the process from above. There are about 580 Wards in the country (mainland Tanzania). Ward Development Committees (WDC) under the chairmanship of Ward Executive Secretary (WES) assisted by Ward Extension Officers (WEO) are responsible for scrutinizing and consolidating project proposals from villages within their jurisdiction into Ward Development Plan (WDP) before submission to their respective District/Urban Council. Many of the ward committees function properly and they are often the only decision- making bodies in the rural areas. The local government grant system is helping them. They sometimes get directions to undertake certain activities. The Guardian (11-7-2005) mentioned for example that wards were "directed (by the RC) to build secondary schools in Lindi". There are over 8,000 villages in the country. A Village is the lowest level in the hierarchical administrative set up of the LGAs. Each village is supposed to have a village council, whose responsibilities include the formulation of socio-economic plan for the entire village and supervise its implementation. The village plans are submitted to WDC for scrutiny, and consolidation into WDP. 3.5. Local Government Reform Programs Higher regional growth requires upgrading technology, human resources development and reforms at the local government level. So far most of the reforms were concentrating on the government at the national level. The original public reform program included civil service reform, parastatal sector reform and financial and planning system reform. The Civil Service Reform Program included five components: ministerial organization and efficiency reviews, pay reform, personnel control and management, administrative capacity building, retrenchment and redeployment of staff. A sixth component, focusing on local government reform was added in the 1994 Action Plan and was smaller than the other components. The reform process started by 1996 and ten years later we can try to assess some of its effects. Implementation of the RARP has been completed. The LGRP is currently being implemented. The Regional Administration Secretariats (RAS) are expected to have two major roles namely: an administrative role and a development role. The administrative role of regional secretariat encourages the maintenance of peace and order of the region and administration of internal operations of the office. The development role is the main function of the Regional Administration and centers on helping capacity building within LGAs in order to deliver their services, efficiently and effectively. 15 Thus, the Regional administrative secretariat (RS) is expected to provide the following services: - Management support services which focus on areas of Local Government administration and finance, planning and economic analysis, legal matters, auditing and community development; - Economic development supports services that focus on production related activities in agriculture, livestock, commerce, industry and natural resource sectors; - Physical planning and engineering services which focus on support to infrastructure and management activities entailing support for and regulation of technical designs, surveys, civil engineering and land development; and - Social Sector Development support services which focus on development activities in to health, education and social welfare sectors in the regions. Under LGRP, the Government formulated and endorsed the vision of the future of LGAs system in 1996. The features of the vision as elaborated in the Local Government Reform Agenda in 1996-2000 are: - Largely autonomous institutions, with freedom to make policy and operational decisions consistent with government policies without interference by the central government Institutions; - Strong and effective institutions, which possess resources and authority necessary for effective performing their roles and functions that have been mandated; - Democratically governed, whereby their leaders are chosen through a full democratic process; - Fostering participatory development by facilitating the participation of the people in planning and executing their development programs, similarly fostering partnership with civil groups; - Reflecting local demands and conditions that take into account demands of its service by the local people and socio-economic environment prevailing in the area; and - Conducting their activities in a very transparent manner and being accountable to the people. Consequently, the overall goal of Local Government Reform Program is to improve the quality of and access to public service provided through or facilitated by LGAs. There are six components, each of which aim to contribute to the achievement of this goal. The objectives of the components include: - Governance: to establish broad based community awareness of and participation in the reform process and promote principles of democracy, transparency and accountability; activities to improve governance are undertaken with the Zonal Reform Team of the local government; - Restructuring of Local Authorities: to increase the effectiveness of Local Authorities in delivering quality services in a sustainable manner; - Finance: to increase the resources available to Local Authorities and improve the efficiency of their use; - Human resources development: to increase the accountability and efficiency of human resource use at Local Authority level; and 16 - Institutional and Legal Framework: to establish the enabling legislation, which will support the effective implementation of the reform measures. The launch of the Local Government Reform Program (LGRP) in 1999 is a promising sign that the central government is gradually devolving more power to Local Government Authorities (LGAs) and even supplies the necessary finance. The Local Government Reform Program (LGRP) deals with the transformation of the local government system. The purpose is to improve services for the poor. The major areas of the project are education, health, roads, water and human development. The project helps to restructure the organizational structure of local governments and has a special financial support component for local government, which works with grants, not dictated by the center, but based on priorities planned by the people, reflecting their priorities. The project did studies on financing local government (for example, Franzsen et al., 2002 and Franzsen, 2004). The project started with 24 local governments, trying to enhance their capacity to collect more money. Secondly, the project discussed how to share grants from the central government on a fairer basis. A formula has been developed, based on the level of education, health, agricultural and road development and available drinking water. The project wants to make it more transparent engender realistic planning and help local governments to determine their own priorities. It boils down to a systemic reform and the project also helps with revenue enhancement. There is also local government grant system. The Local Government Support Project (LGSP) can provide capital grants and capacity building grants. If a city or local government doesn't qualify at the moment, it may receive technical assistance. On top of this money there is also the Tanzanian Social Action Fund. In 2002 this project was prepared. The project document provides an analysis of the situation and training components to deal with the issues that emerged. The project has a website www.poralg.go.tz and there was a predecessor project, including private sector development in water and electricity. Solid waste is also an activity, which the project supported, just like providing vehicles for emptying the pit latrines. They helped to change the formula for grants. 70 percent is based now on population, 10 percent on area size and the rest on the poverty level. The Local Government Support Project was created to finance decentralization. It receives World bank support and includes other donors.xiv These bilateral donors pay the program and are entitled to sit on the steering committee, meeting every six months. They are folding their programs into budget support. The LGSP can give the grants. In principle the money is distributed equally, on the basis of population, (70%), poverty (20%) and size (area = 10%). For the second half of 2004 50 percent went to the councils and 50 percent to the village level. Besides the LGSP there is also the LGCDG, the Local Government Capital Development Grant System. Several donors are development partners in the project. They participate in the board meetings, which are held on a quarterly basis. In fact there is a Local government technical committee and a Steering committee both meeting quarterly. They 17 advise to help them to make well-informed decisions. It helped and the Steering committee approved the first release of funds in the first half of 2005. The grants of the local government urban sector rehabilitation project are distributed over the country.xv The programs will go on for some time but there are reviews after every three years. The project is currently working under a plan for 2005-08. The project also supported the development of strategic plans at the local level and now the citywide plan is almost out. This was promoted under the Sustainable cities program of the UN Habitat Organization, which started in 1992. These initiatives certainly help to make available finance at the local level for development activities. They contribute to spreading development as found in the municipalities of Dar es Salaam and in two regions: Arusha and Kilimanjaro. 3.6 The evidence: what actually happens in Tanzania How can decentralization stimulate local governments to develop further? Using the criteria formulated above we will not try to evaluate decentralization in Tanzania. The opinions on the effects of decentralization differ. Very few people consider it a major force, if only because it has been in place only for a relatively short time. 3.6.1 The expectations: improved management Many benefits were expected from more decentralization, in particular more democratization, improved management and a reduction of the regional disparities. However, the implementation of decentralization strategy after 1997 was not totally satisfactory (Steffensen et al., 2004 and Mitullah, 2005). Many shortcomings were noted, partly because by then the central government continued to maintain strong presence at regional level. However the process has strengthened the democratization process by introducing mechanisms of accountability and by forcing local governments to compete for funds. Some argue that there are not enough infrastructures outside the capital. Infrastructure delivers inputs to farms, helps to market agricultural produce, distributes raw materials and facilitates both domestic and international trade. There is a direct link between economic development and the growth rate of infrastructure. The rehabilitation of physical infrastructure, particularly the trunk and regional roads under the Integrated Road Program (IRP) during the period improved significantly marketing networks. Recently the government is working on the privatization of the railroads, which hopefully leads to an improved functioning and could also be a big boost for the harbor of Dar es Salaam. We have noted more infrastructure and transport in the rural areas, contributing to their development. A lot of money is spent on improving the infrastructure and transport (World Bank, 2006b). Transport contributes to poverty alleviation by assuring cost-effective transport of goods and people. It also links the country to the global economy and assists neighboring landlocked country to connect to the world. World Bank (2006) provides an overview of the transport infrastructure and its management. 18 The Agriculture sample census 2002/03 (NBS, 2006) shows some improvement in assets in the rural areas. The higher levels of regional economic growth are probably also the effect of decentralization. This explanation is discussed below: the positive effects of infrastructure built by different levels of government and of the expenditures of local governments has contributed to higher growth figures. Regional budgets and tax revenues give an indication of how much is spent in different regions. However, there is also direct central government support to regions, for example by setting up offices in specific regions. An effect of decentralization is that more regional initiatives are taking place as has been shown by a comparison between Arusha and Kilimanjaro districts (chapter 8). Picking two regions in the north concluded that the Arusha region is better managed and booming because of tourism, but also because of agriculture and trade links with Kenya and Uganda. Wheat, coffee and flowers (ten Dutch horticulturalists and at least one Frenchman), plus mining are mentioned, although the latter concerns mainly some small- scale activities. Tourism is an activity, which takes place outside Dar. The increase has had a positive impact on rural demand. Also the three municipalities in Dar es Salaam have been most active in developing economic policies, for example providing space and infrastructure. Other local governments have been able to do something about solid waste collection. Lobo (2006) evaluates the experience of involving the private sector in waste collection in Arusha, which is certainly an example of improved management at the local level. One of the remarkable achievements of Tanzania is the improved accountancy system for decentralized projects. There is a problem that many of the good people are at the central level. However, the Ministry of Finance has trained 400 accountants for rural development projects. 3.6.2 Reducing regional disparities The economic survey 2004 reports about regional inequalities by giving the regional distribution of projects coming through the Tanzania Investment Center. It indicates that Dar es Salaam region led with 262 projects, followed by Arusha 79 projects. Other regions that attracted investors were Kilimanjaro 21 projects; Mwanza-20; Mtwara 9; and Shinyanga 8; Morogoro, Mbeya and Iringa regions, registered 7 projects each; Pwani, Lindi and Mara 5 each; Tanga region registered 4 projects; Dodoma, Kigoma and Tabora 2 projects each; and Kagera, Rukwa and Manyara 1 project each; while Ruvuma and Singida did not get any project. Some registered projects however, have branches in more than one region. The following table gives the figures for the five regions mentioned. The list shows a strong concentration of projects in just a few regions. The national accounts of Tanzania (NBS, 2006) give the regional per capita GDP at current prices from 1992 until 2004 (see table 4). In 1992 the Kagera region had the lowest per capita GDP, while in 2004 Kigoma scores lowest. More importantly Dar es Salaam's per capita income in 1992 is 4.5 times the per capita income of Kagera, while in 2004 its per capita GDP is only three times that of Kigoma, and less than three times that of Kagera, suggesting a more equal income distribution and a substantial growth of per 19 capita GDP in current prices in the different regions, due to decentralization and economic growth. As shown in table 1 above remittances from rural to urban and from urban to rural also contribute to regional development. Table 4 Per capita income data for the five regions (in thousand T.Sh. per year) Region 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Arusha 61 74 96 124 149 177 201 227 252 279 304 294 356 DaresSalaam 142 171 226 267 311 371 435 492 518 556 589 616 634 Kagera reg. 32 39 50 65 79 95 109 123 140 150 161 225 227 Kigoma 32 41 53 68 82 99 113 127 134 155 126 205 206 Kilimanjaro 40 47 59 74 86 101 117 132 139 153 255 307 376 Source: NBS(2006). What do we know about the financial flows to the local governments? How much donor support gets beyond Dar es Salaam? The LGRP has published financial statistics for the period 2000-2004 that give an impression of the importance of grants and own sources of revenue for the Arusha and Dar es Salaam regions, which are reproduced in table 5. Decentralization has certainly contributed to the rapid development of the Arusha region. The table shows that the most important grants to local governments are for education, but the grants can vary substantially over the years. The total of grants is important however. It reaches between four and nine times the own sources of revenues in Arusha, while the total of grants is between two and one and a half times the own source revenues in Dar es Salaam Table 5 Grants and own sources of revenue for the Arusha and Dar es Salaam regions 2001-2004 Arusha and Dar es Salaam Category 2001 2002 2003 2004 2001 2002 2003 2004 Educat bgrant 5961112 8362870 9073650 4727457 8106223 10508917 12292088 5504412 Healthgrant 1234230 1127381 1622756 1071726 2181286 2585499 2697401 1352899 Roadgrant 357298 479867 474594 60407 534870 550636 354549 1327925 Watergrant 103948 96406 144706 125241 140741 292658 184465 115733 Agic.grant 23179 394313 432321 78421 52692 54365 47324 22848 LA b.grant 516785 632272 611336 583298 448451 622851 1387778 345800 Other grant 656517 1332014 1634541 1360326 1371969 5155903 2930478 1076065 Basket funds 417212 4427109 1532410 1762506 1231144 1613886 3011175 699367 Compens 0 0 452970 412704 0 0 1333056 1158558 Tot.grant 9270281 16852232 15979284 10182088 14067376 21424716 24238314 11603607 Dev levy 389568 341927 60048 0 2045401 1995142 859328 0 Prop tax 208392 392318 272161 250970 1514026 1662982 1583749 1197029 Pcesses levies 170366 167106 125009 42242 525751 567769 499471 155035 Service levies 196629 150332 316637 274337 4044272 4519950 5317695 2804136 Land rent 1848 21464 7833 44717 0 0 0 0 License 701619 655730 752647 269037 4014847 4699216 4923033 1339991 Charges 271670 258517 274624 304158 1038080 1590824 1759498 551417 Other 749375 185220 331633 22881 422931 649800 513118 489136 Subtotal 2689466 2172614 2140593 1158342 13605308 15685684 15455892 6536743 Total 11959747 19024846 18119878 11340430 27672683 37110399 39694206 18140350 20 Total resources for the Dar es Salaam region are only two times what the Arusha region can spend. The most important own source revenue in Arusha is the development levy (until 2004) and currently different charges, while the most important own source revenue in Dar es Salaam is the service and other levies, with licenses fees and permits on the second place. Finally one notes that there are a large number of targeted block grants and that there are many own sources of revenue. Instead of continuing with so many, the system may benefit from some more focus. 3.6.3 Democratization, recent new initiatives August 2006 a meeting of regional leaders was held for the second time in 2006 in Arusha. The president delivered an important speech and this event shows the authorities take decentralization more serious. There are advantages and disadvantages to decentralization. One needs an economic basis to justify decentralization, but it is good to make the people responsible through elections. The president addressed the meeting for two hours suggesting a code of conduct, which made it very explicit that a lot is expected from these regional leaders. Before the meeting a large number of regional leaders had been replaced.xvi The good thing is that the politicians now know they will be held accountable for what they have promised. The Arusha meeting is highly relevant and will get a follow-up. 3.7. Testing the different theories about the effects of decentralization 3.7.1 Income transfer, services or employment creation In Tanzania income transfers take place, but at a limited scale. The effects on the income distribution between regions will be analyzed below. The development of infrastructure and services in the regions is also happening although only at a limited scale, while the government's role in creating employment opportunities is even more limited. The following table gives an impression of remittances. Table 6 Remittances as a contribution to total monthly income per household Source/Location Dar es Salaam Other urban Rural Mainland Tanzania Remittances 3,664 3,165 2,719 2,835 As % of total 3.5 3.2 2.1 2.4 income Source: NBS (2002). 3.7.2 Deconcentration, delegation or devolution, monopoly or pluralism? In the literature it is noticed that decentralization sometimes leads to pluralism (for example in different states in India). Alternative Cohen and Peterson (1999) draw attention to another form of decentralization which the call distributed monopoly. The latter refers to a situation where roles are distributed spatially but are still concentrated in one organization or institution (the `decentralized monopolist'). 21 The best way to describe what is actually happening in Tanzania is by looking at the cities and regions we have studied in more detail. OECD (1996) argues that decentralization in Africa in the 1970s meant deconcentration of the national administration and strengthening of the regional administration. In countries with a one- party government power had often been centralized and decentralization was only promoted as a reaction to the economic stagnation of the 1980s, after the oil and debt crises. However, locally elected bodies tend to have limited capacity and funds and were often unable to carry out the functions transferred to them. Central government continues to maintain a strong presence at the regional level. The process of decentralization is to a large extent administrative deconcentration, where the RS have an important role in keeping law and order. At the same time there seems to be legal and sectoral centralization. 3.7.3 The autonomy of the urban/regional manager Economic policies What is their experience with formulating economic policies at the district level? Most experts interviewed noted that this is desirable, but not yet the case. Most regional officials don't seem to consider this as their responsibility yet. However, there are big differences from region to region as far as the formulation of local economic policies and the development of local economic activities is concerned, as can be shown for the Arusha and Kilimanjaro region, where the first region promotes economic activities in a more systematic way and with more success. However, promoting economic activities is not yet a policy at the local level, but supplying more facilities/services provision is. As far as economic policies are concerned, the local governments implement the policies formulated at the national level, but also come up with their own ideas, such as raising their own taxes or creating boarding schools. How to do this is a local decision. It may go as far as deciding what to grow, for example the choice to go for developing horticulture in the Arusha region. In particular local governments in Dar es Salaam have some local economic development policies within the national development policy. Decentralization resulted in more competitive municipalities for example in Dar es Salaam. However, also the example of Arusha doing much better than the Kilimanjaro region is an example of what is possible in the framework of decentralization. Finally it can be expected that decentralization leads to better regional and urban governance if the meetings of regional leaders will be continued. Social and environmental policies Do local governments also have social and environmental policies? No, at the moment the local governments implement social and environmental legislation, but have no local social and environmental policies. However, Tanzania has a tradition of pro-poor rural policies and that may start to pay off (Owen and Teal, 2005). Although the local governments mainly implement national policies, they could pass their own by-laws at the local level. Consequently a drop in poverty could be shown in table 1, in particular in 22 Dar es Salaam, but also in the other urban areas and even in the rural areas. The data on the income distribution suggested that the income distribution is skew but has improved. The LGS project expenditure to some extent compensate for the existing inequalities by spending more in less developed regions. Fiscal decentralization Do the local governments have enough money to spend on development efforts? The LGSP project described makes funds available according to its capacity and the quality of the proposals. If there is more demand they like to be flexible. Expenditure is currently 90 percent of their budgeted funds, so the LGSP never had funding problems as a project. The plans prepared by the local governments are very participatory, meaning the local governments and the population are involved in it and discuss it. They take 3 to 4 months and are then converted in a memorandum of understanding specifying that in this plan period so much money is committed. However, in case of an emergency they also provide assistance. The main instrument is finance and hence the LGSP discusses funding at the village level. This avoids confusion if there are too many donors active at the local level.xvii At the local government level the property tax has a great potential and the private sector need to be sensitized how the system works. Lending by local governments is another option, but commits future generations. The projects mentioned in section 3.5 want to improve service delivery. The impression exists that the fiscal pressure at the local level increased after the re-introduction of local government in Tanzania. It decreased recently, because central government has abolished certain local taxes and now compensates the local level through transfers. For decentralization to succeed it is important that at the local level there is adequate capacity to design and implement good tax systems, properly thought through and to avoid distortions. The current capacity at the local level is often not sufficient in this respect. A lot seems to depend on the people leading the decentralization process and an old and wise man such as the Regional Commissioner in Arusha can do a very good job if he works with ambitious civil society organizations and private sector representatives in the region. Elsewhere a young and ambitious civil servant may help to start things moving. 3.8 Conclusions We argue that decentralization has contributed to local development, because of the flows of goods, services, capital, ideas and people it has generated. There is still some resistance against the implementation of decentralization policies at the national level. Also the accountability in the process can be improved. The pace of harmonizing the roles between the different levels of government in Tanzania was slow. Sectoral ministries are reluctant to give away authority. The Local Government reform program also needs to be marketed more. After 2008 the LGSP project hopefully has created the momentum to do things better at the local government level. The responsibility is now with the local government and 23 empowerment of the local governments should be the result of this project. By then this approach should also be mainstreamed. It should be day-to-day government business. The LGSP helped to build capacity at the local government level. Decentralization is the framework for further development activities and the citizens and politicians should use it as such. A point of critique is the rather arbitrary local tax increases. Decentralization provides the opportunities; the challenge is now to design mechanisms to get capital and entrepreneurial talent to rural areas. For the Tanzanian bureaucracy it is important that the president is in favor of small enterprise development and decentralization. Many officials still see enterprises only as units to be taxed. It was hoped that decentralization would have diminished the rural urban differences. That could have led to less migration and hence solve the problems of big cities like Dar es Salaam. There are however no indications, despite a slight improvement of the income differences, that the number of migrants or their destination has changed over time. Local governments have received more autonomy in the framework of decentralization. Local governments can and will do more, although they are not able to take up all the responsibilities fully yet, if only because of the lack of sufficient local finance. The local government reform component of the reform program, which really started in 1996 and the local government grant system and increased tax collection have provided the local level with more funds. However, firm support for the local government reforms has sometimes been missing. So far decentralization has never been a hot political item and the challenge for local government to take up their own development has rarely been emphasized. 4. Infrastructural bottlenecks for urban economic development? 4.1 Introduction Tanzania is big (883,987 km2) with a widely dispersed population of 34.4 million people in 2002. It has a very low population density of 39 people per km2 (in the same year). There are big differences in altitude, ranging from sea level on the coast to 19,340 feet above sea level. This has made it difficult to develop a transport infrastructure that adequately serves all parts of the country. Tanzania faces structural constraints, which condition its potential for growth and development in the long run. During the 1960s, it was widely accepted that the major growth hindrance in Sub-Saharan Africa (SSA) was capital shortage (Tax, 2000) rather than policy failure. The idea that proper policies were missing dominated in most of SSA countries during 1980s. This situation brought about adoption of various development policies/strategies to mitigate growth hindrance over period of time (see annex 1). To what extent is the lack of infrastructure currently the main bottleneck for Tanzania? 4.2 An overview of the infrastructure In this section we will discuss three types of infrastructure: power/energy, transport networks and the role of the port in Dar es Salaam. Tanzania depends to a large extent on biomass (see box 1). This dependence on biomass, compounded by population growth, 24 more industrial activity, further development of agriculture and increased urbanization, results in extensive deforestation, especially around urban centers, is putting unsustainable pressure on the natural environment energy-resource base. Box 1 Energy Tanzania's energy sources include non-commercial primary energy (fuel wood, charcoal) and commercial energy (petroleum, natural gas, hydroelectricity, coal and some geothermal). Economic growth normally increases energy demand. Tanzania has abundant energy resources of electricity, natural gas, biomass, coal and renewable energy sources. But the energy potential of Tanzania is underdeveloped and the cost of imported fossil fuels is beyond the means of most Tanzanians. As a consequence, people are forced to depend on biomass, particularly fuel wood for cooking and heating, for their energy requirement. Consumption pattern of energy in Tanzania shows that wood-fuels (fuel-wood, charcoal and agricultural residues) account for 92 percent of the final energy consumption; with electricity and petroleum fuels accounting for 0.8 and 7.2 percent of final energy consumption respectively (URT, 2000). Source: Mlingi and Assey (2005). However, strategies to develop energy sources of natural gas (Songo Songo, which is about to be used in Dar es Salaam), coal (Mchuchuma and Liganga), and petroleum exploration (along the Indian Ocean coastal belt) are being implemented. Exploitation of these energy source will undoubtedly facilitate the development of other activities in the economy while ensuring environmental sustainability. The national electricity company has a management contract with a foreign firm. This firm is currently focusing on three issues, besides efficiency increases: a. Extending the network to more people b. Increasing the efficiency of their operations c. Preparing more private sector involvement. Before this the electricity firm was in dire strait, almost going bankrupt. The Swedish SIDA funds the contract and the World Bank manages the contract on their behalf. The foreign firm should turn around the Tanzanian National Electricity Supply Company (TANESCO) financially and make it financially viable. Secondly, they had to turn it around technically, because there was load shedding almost every day. Corruption needed to be eliminated and they were asked to look into privatization options. Since then a number of events (the Enron scandal and power shortages in California) caused that the Tanzanian government is not as eager any more to privatize TANESCO, although the company is still on the list of the PSRC. The focus is now on increasing efficiency and to become financially independent. Interruptions are down from daily to maybe once a month. They went for internal 25 unbundling and separated generation, the network and the distribution system. After a first 2.5 years contract the same foreign company obtained a second contract for 2.5 years. They have formulated new aims. Financial viability is still important and the internal ring fencing will be concluded. The focus is now on accelerated electrification and they continue the technical turnaround. There are performance targets and the contract specifies the bonuses and penalties in case of non-performance. The emphasis is also on capacity building of the current staff and working on an exit strategy for the foreign firm at the end of the five-year period. The foreign firm would like to have a study showing the potential of further developing the coastal region in Tanzania since such a study would quantify future demand. There exists a national grid, but only 8 to 10 percent of the Tanzanian population is covered. The customer base of the company is very small. All big towns have electricity, but not all are in the grid. Sometimes they use diesel generators, which tends to be very costly. The more they produce the more they loose. Fortunately there is also some hydropower, which is relatively cheap. Total employment has gone down somewhat through resigning and retirement. TANESCO went for outsourcing and their construction teams have been retrained to now oversee private construction teams. They have not recruited for ages, but now need some young blood and some specialists. Meter reading is often done by private firms. Generation plants are often partnership arrangements, but these contracts may not always have been negotiated very well. The allowed return on investments was much too high in dollar terms in a number of cases. In the books of TANESCO the foreign firm found all the previous assistance as loans to be paid by the company to the government. These might have been World Bank project loans, but the money was lent on to them at a very high rate of interest. Another problem is that they loose money on every KWH generated in this way, since the fuel costs are four times the price paid for the electricity. In principle TANESCO has an electricity surplus so they are eager to increase the number of connections. They also prefer to use gas as a source of energy, since it it much cheaper and you only need relatively simple gas turbines. In South Africa they can produce at only 3 US$ cents per KWH. Good infrastructure allows to deliver inputs to farms, helps to market agricultural produce, distributes raw materials and facilitates both domestic and international trade. There is a direct link between economic development and the growth rate of infrastructure. The rehabilitation of physical infrastructure, particularly the trunk and regional roads under the Integrated Road Program (IRP) during the period improved significantly marketing networks. Box 2 Road networks 26 Road transport is the main mode of transport supporting the majority of economic activities in Tanzania. Up to year 2000 Tanzania had 85,000 kilometers of road network, and only 5 percent are paved (UTR, 2000). Unpaved roads, which are in good condition, constitute only 14 percent of the total unpaved roads, 25 percent are in fairly good condition, while 61 percent are in poor condition. Road transport is crucial for district, region and zone development compared to other mode of transport in Tanzania. For instance, road transport handles about 70 percent of internal freight traffic, 64 percent of transit cargo and is a major mode of passenger transport. Moreover, the number of commuter buses in the City of Dar es Salaam has increased to 9,872 in 2003 from 5,801 in 2002, while in Arusha municipality the number has decreased to 960 in 2003 from 1118 in 2002 (UTR, 2004). The significant increase in Dar es Salaam was due to increased demand (population increase) while the declining number in the case for Arusha municipality is difficult to explain. Source: Mlingi and Assey (2005). The basic transport networks links all the main production centers with varying degrees of efficiency. Transport plays an important role in the development process of a nation. In the broad sense the infrastructure includes roads, rail connection, rivers and the sea, air connections, different types of communication and water and energy resources. Different boxes try to make up the state of art with respect to different types of infrastructure. Box 3 Rail transport Tanzania has two railway systems: the Tanzania Railway Corporation (TRC) operating about 2,640 km of tracks, and the Tanzania Zambia Railway Authority (TAZARA) jointly owned by Tanzania and Zambia governments with 1,860 kilometers from Dar es Salaam to Kapiri Mposhi in Zambia with 970 kilometers of truck within Tanzania. While TRC connects Dar es Salaam with Kigoma, Mwanza, Arusha and Tanga regions; TAZARA links Dar es Salaam with several regions including Morogoro, Iringa, and Mbeya. Although the two railways have played a crucial role in providing services, the performance of TRC has had the greatest impact on the internal trade system. This is partly because its services cover a more populous part of the country than TAZARA. Both railway systems are affected by a badly maintained network and weak institutions and management. Repair shops are handicapped by lack of trained personnel and spares. Moreover, shortage of rolling stock and locomotives, deterioration of telecommunication networks systems, and accidents have negatively affected performance of these railway systems. For instance, in 2003, TRC transported 1,442,713 tons of freight cargo compared to 1,445,757 tons in 2002, a decrease of 0.2 percent, while in 2003, freight cargo transported through TAZARA declined to 613,693 tons from 677,000 tons in 2002, a decline of 9.4 percent (URT, 2004). 27 Source: Mlingi and Assey (2005). Box 4 Air transport Tanzania has two international airports, at Dar es Salaam and Kilimanjaro, and over 50 airstrips for domestic flights to Zanzibar and all regional headquarters. This mode of transport continued to be provided by a number of companies, majority of which are owned by the private sector. About 10 local private companies provide domestic services in the country, with regular flights to some of neighboring countries like Kenya, Uganda, Rwanda Malawi etc. The number of passengers has increased significantly despite the fact that, majority of these fights are small in size and they face stiff competition in air transport services on local and regional markets. For instance, in 2003, local private airlines carried 579,107 domestic passengers compared to 475,734 in 2002, an increase of 21.7 percent (URT, 2004). Source: Mlingi and Assey (2005). What are the current constraints for further development in Tanzania according to the business sectors? We asked the Confederation of Tanzanian Industries (CTI) to respond on behalf of their members. They like a further improvement of the business environment and the formalization of the informal sector. Access to finance can be improved, just like the existing infrastructure, in particular the port, the railroads and the roads. Improving infrastructure may have similar dynamic effects as formalizing the informal sector, but without causing the distortions that may result from formalization. In Mwanza improved roads have resulted in substantial changes because people can transport their goods and sell them elsewhere. Box 5 Communication The current communication systems (postal, telecommunication, television, radio and internet services) are inadequate to meet domestic and international business demands. Since 1985, when Tanzania adopted the open market economy policy, expansion of these services has increased in terms of numbers particularly in urban areas. For instance, telephones capacity (lines) increased from 2,930 in 2000 to 234,269 in 2003, while mobile phones customers increased from 805,000 in 2002 to 1,298,000 in 2003 (URT, 2004). However, these facilities are poorly equipped in rural areas, while there are virtually no sufficient and efficient communication networks in the rural areas. The adoption of the National Information and Communications Technologies (ICT) policy in 2003 has makes it possible for "enabling sectors" such as telecommunication, broadcasting to work together whereby "enabled sectors" such as education and health can become further empowered through the appropriate development and application of ICT. 28 Source: Mlingi and Assey (2005). The importance of the port of Dar es Salaam is clear and the maritime transport system would also need to be discussed, but this goes beyond the current study (see box 6). Box 6 Port and maritime system There is a port system operated by the Tanzania Harbors Authority (THA) centered in Dar es Salaam with minor ports at Tanga, Mtwara and Mafia Island. The Dar es Salaam port has been rehabilitated and expanded to carter for domestic and neighboring countries namely: Burundi, Malawi, Rwanda, Uganda, Zaire and Zambia. The rehabilitation has increased the capacity to handle cargo from 3.3 million tones in 1985 to 4.9 million tons in 2003. This has been the case because the number of ships handled at Dar es Salaam port increased from 3,894 in 2002 to 4,034 in 2003 (URT, 2004). Moreover, efficiency has increased for example, 22.6 containers were handled per hours in 2003 compared to 20 containers in 2002. Tanzania also has inland water transport on Lakes Victoria, Tanganyika and Nyasa. However, the potential of water transport in the three lakes and along the coast in the Indian Ocean has not been fully exploited. This is due to the poor infrastructure of the ports, inadequate routine maintenance of facilities and the shortage of skilled manpower. Important bottlenecks in the infrastructure are the connections between the port of Dar es Salaam and the hinterland and between the capital and the western and southern parts of the country. There is not really a strategic view what to do with the harbor in relation to its hinterland. THA recently received a Japanese delegation. They looked at the port on its own and not as part of a logistical chain. The result is that the increase in trade through the port cannot be well served by the railroads (there are even plans to move the railroads out of the center of the city). The current capacity of the railways is only 1.5 tons per month from the port! A lot of cargo cannot leave the port. There is a need for an integrated study for the total logistical chain. Box 7 gives the possible terms of reference for such a study. Box 7 Terms of reference for such a study of the potential of the port of Dar es Salaam for the country and the sub-region a. The potential of the Dar es Salaam port for the hinterland should be determined, looking at serving Tanzania and its neighbors b. The inter-modal linkages need to be studied and suggestions should be made how they can be developed c. The physical infrastructure in Dar es Salaam deserves attention: trains (their 29 limited efficiency), roads, etc. d. The bureaucratic environment needs attention, for example the functioning of the customs and how this can be improved e. Optional: look at water front development in Dar es Salaam Potentially this port can serve Eastern Congo, Rwanda, Burundi and Zambia. It can compete with Mombassa or the South African Ports. Mombassa is better placed to serve Uganda and South Africa can also serve Eastern Congo and Zambia. Mombassa is also a natural port, but is in a sense better because its channel is deeper. THA received a 500 million dollar loan from the World Bank. Remaining problems are that they are not yet fully computerized, the technology is not yet very advanced and the port entry needs to be deeper to allow the Panam canal ships to enter. There were some previous port development studies covering the period 1994-2004. In 1996 there was a study on privatization by a Dutch firm, which also only looked at the port and not so much at its role for the hinterland. The development of the Central corridor project is supported by the World Bank in the framework of the Special Development Initiative (SDI). There are some important trends in the shipping sector. The ships going from east to west are getting bigger all the time. The ships that can sail through the Panama Canal (4000 TUE) are often too big for the port of Dar es Salaam. Bigger ships can go through the Suez Canal and don't even pass here anymore. Ships also tend to become deeper and they need to have quays of 8 to 12 meters. The depth of the port entrance is only 10.5 meter, which would need to be increased to 14 meters. They only do transshipments for the Indian Ocean. There is no real industry related to the port. It is just cargo and stocking. The port has a relatively autonomous status with a Board with 10 relatively independent experts, some engineers, some financial people and some civil servants. Plus two members of parliament are on the board. The Managing Director of the Port was nominated by the President of Tanzania, but should normally be appointed by the Board. The Board would send its recommendations to the Minister who has to approve them. The government owns the port, but THA are the landlords on the government's behalf. THA has gone through a period of restructuring or institutional development. They went from a public company to some kind of Private Public Partnership (PPP), which took some time. They opted for a concession contract where a private party has a 10 years contract to run the facilities. They only put in some working capital and some small amounts for investments. There will be a new round of tendering in 10 years time. The Port Authority can take its own investment decisions, but the Port Authority depends very much on the government for the necessary finance. To get the money they developed sector plans, which need to be approved by the Ministry. The Ministry usually comments, but doesn't interfere. In 2005 the Port Authority developed a new company plan. It will focus on accommodating the new shipping requirements. Their container 30 terminal is doing very well. Borrowing is not directly, but through the government. It could be cheaper for them to borrow directly from the market. They never considered issuing bonds because they had this World Bank 500 million dollar loan. 4.3 Conclusions Infrastructure is a necessary, but not a sufficient condition for development. In this chapter we identified several bottlenecks, but also some interesting opportunities and initiatives to improve the current situation. Suggestions were made for a study identifying the future demand for electricity and a study of the development potential of the port of Dar es Salaam. The question now is how the new opportunities created by an improved infrastructure are taken up by the local businessmen in the formal and informal sector (chapter 5 and 6) and how the authorities in the framework of decentralization use these new opportunities (chapter 7 and 8). 5 Urban rural dynamics in Tanzania 5.1 Introduction Tanzania has grown more than 6 percent per year during the last three years. This study started out of concerns about the widening gap between the rural and urban areas. Are there enough linkages between the two sectors and do the urban areas function as an engine of growth, also for the rural areas? Tanzania and in particular Dar es Salaam benefited from economic restructuring in the nineties, but how is the growth spreading to the rural areas? Tanzania intends to achieve, according to Vision 2025, a modern resilient economy capable of achieving regional and international competitiveness. The World Bank Country Economic Memorandum contracted out this study since it did not have much data on the contribution of the informal sector while it may be an important factor explaining this 6% growth in the rural and the urban areas, which concerned the formal and the informal sector. In this chapter a new theoretical framework will be introduced to analyze rural-urban linkages. We also present the evidence with respect to the role of Dar es Salaam for the hinterland. Finally we draw some conclusions how the rural areas can benefit more from the urban-rural flows and formulate some recommendations to promote them. 5.2. Spread in economic growth is more important than reflected in statistics What is the role of the urban informal sector in spreading economic growth is more important than what would be expected from the available statistics. How to explain six percent growth in rural and urban, formal and informal sector? The following figure conceptualize the urban-rural dynamics. Figure 1 31 Urban rural linkages and dynamics Urban Goods Rural system system with its own Ideas absorbing dynamics Money these flows People Services New conceptualization based on flows to and from We see from figure 1: · Flows of people: migrants as physical capital to Dar es Salaam, the migrant comes and returns with skills, ideas, equipment and goods · Flows of goods and services in and out · Flows of money: remittances, loans and savings flow largely through the informal sector Economists usually study rural-urban linkages in terms of `exports' of goods from the rural areas and imports from the urban areas. In this proposal we will use the term informal flows of people, money, ideas, goods and services. This conceptualization goes beyond the traditional registration of goods exported from or imported in the rural areas and beyond the informal sector concept. The idea is shown in figure 2 and the differences are conceptualized in table 7. Figure 2 Flow of people Urban areas & Flow of goods rural areas and services Flow of ideas 32 Flow of money This conceptualization may help us to: a. Discover a new dynamics b. Do justice to ongoing redistribution practices c. Consider an issue in an integrated, rather than a sector (rural or urban sector) way d. Formulate policies, which may help to spread development through a more equal distribution and increased total development. e. Discover mechanisms of innovation, of spreading ideas and initiatives f. Look at the rural urban and the urban rural flows The line of the study on Tanzania was: how can we explain six percent growth in rural and urban, formal and informal sector? By looking at the following flows and their effects: · Flows of people: migrants as physical capital to Dar es Salaam, comes and returns with skills, ideas, equipment and goods · Flows of goods and services in and out · Flows of money: remittances, loans and savings flow through the informal sector This conceptualization emphasizes: a. The two systems cannot be separated b. The relations are much more complex and go in both directions c. They can be very positive for the urban or for the rural areas. d. Hence amore equal distribution of the benefits of development can be achieved. The challenge concerning the relations between a city and its hinterland is to maximize the positive interactions between the rural areas and the towns and vice versa. An example is given in Box 2.1 in Van Dijk (2006a) of how this could be achieved in the case of Kaya in Burkina Faso. Research has shown that regional economic systems with a clear role for their cities are a critical part of the development process (Ohmae, 1996 and Jacobs, 1970). Jacobs (1970) suggests there were cities before there were agricultural communities. Secondly she argues that the cities have a huge impact on the surrounding areas. Then she considers cities to be examples of an import substitution policy to eventually develop its exports. These policies created employment and wealth. Then the city has a much bigger 33 impact on the neighboring rural areas and will eventually stimulate these areas to develop their export to the cities by providing ideas, technology and the necessary inputs. The current research starts with the argument to focus on informal flows in the case of Tanzania. We will provide a number of arguments to study rural-urban relations. Subsequently a number of the relevant questions are asked before providing some evidence. The informal sector plays an important role in the rural-urban relations. There are in Tanzania informal flows of people, money, goods and services (the five flows: Goods- Ideas-Money-People-Services or GIMPS), which spread development over the country. If a more positive approach to these activities is taken, attention can also be given to the issue of raising the productivity and increasing the competitiveness of the cities and regions wherever they are located. The quality of labor, entrepreneurship and existing financial mechanisms can be developed further. Flows are about linking up to global formal/informal value chains! For a country it means that it is no use to put the emphasis on more protection because it will not work. Rather find out where Tanzania can be competitive. A number of relevant questions can be asked with respect to this topic. 5.3 Why call it informal flows? What is the reason to call it informal flows? 1. They may not even be perceived; 2. The flows are often not measured; 3. They tend to fall outside the existing tax or registration system. The use of the term informal is not important however; it mainly has a signaling function: 1. It is something usually not covered in the official statistics; 2. Some of these activities may take place in the informal sector; 3. They may have their own character, which could disappear if formalized. Table 7. The terms informal sector and informal flows compared Differences Urban Informal sector Informal flows between the rural and urban areas Definition Poorly defined concept after Possibility to measure the 30 years of research five flows, if efforts are made Informal meaning Not registered enterprises, Not registered flows or not paying tax or the minimum legal wage Goal of promoting it Local economic Link up with global value development chains for goods, services, ideas, capital or even 34 migration Focus Focus is on the enterprise Focus is not on the enterprise, but on the business environment and value chains What is new? a. Dynamics From a static approach to More dynamic approach b. Scale From local economic Global value chains development to c. Competition From local competition to Global competition An element which is left out in this conceptualization in terms of rural urban and urban rural flows is the impact of the urban areas on the quality of the environment in the rural areas. Research focusing on rural urban linkages can help to identify the `foot prints' of the city (Rees, 1992), the much larger area affected by the pollution produced in the city. Informal flows are different from the urban informal sector (see table 7). The focus is not on the enterprise any more and the goal is no longer local economic development, but developing relations and maximizing their positive impact on the rural and the urban system. In line with business economics a distinction made is between the enterprise and its environment.xviii 5.4 Why study rural urban linkages? There is a long tradition of studying rural urban linkages (for example UNCHS, 1985 and Rondinelli and Ruddle, 1978). The conceptualization has not changed very much ever since, although sometimes different terms are used (for example de-agrarianization by Bryceson and Jamal, eds, 1997). Usually economists would study the issue in terms of forward and backward linkages. It is all about the goods moving in and out of the rural areas, usually at the benefit of the towns. Data are collected for the construction of GDP and regional product figures. Flows are often netted, meaning a neglect of the important backward flows. Even Owuor (2006) starts to define urban linkages in terms of migration studies or urban- rural interaction studies. The relation is analyzed in terms of work or in terms of goods brought from the rural to the urban areas and back. Very little is known about the complex rural-urban interaction because: 1. Often very different disciplines take out different aspects of the issue, from diffusion of cultural ideas a to increased smuggling. However, they don't try to get the whole picture. 2. Traditional data collection serves to produce national level statistics, in which the origin of goods and services is regions and not necessarily rural or urban 3. Rural and urban issues tend to be analyzed separately by different experts, disciplines and development organizations. 35 The distinction rural-urban is not always clear cut. For Dar es Salaam the Begamayo outskirts are peri-urban areas but there is still an infrastructure and a context of connectivity. The rural households are not really rural, more peri-urban. On the other hand there is hardly anyone in Dar es Salaam who has been born in Dar. There are hardly any third generation urbanites. All of them still have projects in the rural areas. They send money and build a house in their village, stimulating this rural-urban dynamics. 5.5 Why do we want to study rural-urban linkages differently? Why do we want to study rural-urban linkages in a different way? Three reasons can be mentioned: 1. To avoid thinking in terms of two separate systems, an urban and a rural, served by two types of experts and functioning more or less isolated from each other. 2. To find out whether the urban system can stimulate the rural and the rural can stimulate the urban and 3. To discover to possibilities of spreading development given in many countries the development expenditures tend to be spent in cities and bring about a process of development over there. In the cities most institutions spending the money are based. 5.6 Relevant questions in the Tanzanian case Using this framework, the following relevant questions can be asked and will be discussed in the following subsections: 1. Do these flows lead to rural development in the case of Tanzania? 2. How can Tanzania benefit more from these flows? 3. Can we increase the positive effects? 4. What is required for more positive effects? 5.6.1 Do these flows lead to rural development? Some evidence The theory would expect positive effects whose measurement is possible. Many ideas flow back to the rural areas, driven by trade and technology and in particular telecom. Increasingly internet and computers also play a role. You can see it happening. It is an example, but in a village you may see a copy shop! But they still have to learn better not to supply tomatoes only three months per year and the same for spinach. Most investment is driven by households, since banks provide no loans to these people. Remittances are very important. The real question is do these flows lead to rural development? What evidence do we have: · Not according to incomes in household surveys of 1990 and 2001 · Not according to remittances, which are more important in the urban areas. However, these remittances are not really measured in the 2001 household survey because they are net and we do not know for which purpose they have been used · But: decentralization leads to better regional and urban governance (example Arusha and Dar es Salaam) 36 · Tanzania has many pro-poor rural policies (Owen and Teal, 2005) · More infrastructure and transport (below and chapter 4) · Improvement in assets in rural areas, the Agricultural survey shows improvements through diversification and new technologies (Agriculture sample census 2002/03; NBS, 2006) · The importance of return migration (NBS, 2002) More infrastructures outside the capital Infrastructure delivers inputs to farms, helps to market agricultural produce, distributes raw materials and facilitates both domestic and international trade. There is a direct link between economic development and the growth rate of infrastructure. The rehabilitation of physical infrastructure, particularly the trunk and regional roads under the Integrated Road Program (IRP) during the period improved significantly marketing networks. Recently the government is working on the privatization of the railroads, which hopefully leads to an improved functioning and could also be a big boost for the harbor of Dar es Salaam. A lot of money is spent on improving the infrastructure and transport (World Bank, 2006b). Transport contributes to poverty alleviation by assuring cost-effective transport of goods and people. It also links the country to the global economy and assists neighboring landlocked country to connect to the world. World Bank (2006b) provides an overview of the transport infrastructure and its management. The lack of electricity of a good railway system and the relative high minimum wages mean that not many foreign investors are interested in setting up a factory in Tanzania. The cost of bringing a container to Rwanda is still higher then shipping one from the port of Dar es Salaam to Japan. 5.6.2 How can Tanzania benefit more from these flows? How can Tanzania benefit more from these flows? Alternative model: use informal sector to further spread development. Decentralization provides the opportunities; the challenge is now to design mechanisms to get capital and entrepreneurial talent to rural areas: · Formalization? Does it kill initiatives? · Credit or loans? Is capital the real bottleneck? · Export promotion: globalization also affects the informal sector and provides opportunities · Other SME policies: training and innovation · Take real problems as points of departure (see Tripp, 1997) These issues will be elaborated in chapter 6. Several donors and for example the Netherlands Embassy are also involved in a number of projects in the field of entrepreneurship development, some of which also have an impact in the rural areas.xix The BEST project with the Tanzania Private Sector Foundation will be taken over by the World Bank, which will be using this structure to develop SMEs. There is 100 million 37 dollar available, which is a strong increase in budget from an organization that currently handles 25 to something like 60 to 70 million dollar per year! At the end of September 2006 there will be a mid term review. To adjust the old laws really took time. These kinds of adjustments are a slow process. The Better regulation unit in the ministry of Planning had to fight for its position and the chief officer has been replaced. So there are not as many concrete results as hoped. So far the building is ready for something bigger with a more concrete output. The World Bank will use this structure for an enterprises development program, but only a part of it is for SME. US$ 40 million will go into developing a proper business environment, dealing with things like registration: land, your company, etc. US$ 10 million is for access to finance and US$ 50 million goes to enterprise development and is mainly matching existing initiatives. There are about three or four serious NGOs in the micro finance sector in Tanzania. The National Micro Finance Bank may be one of the banks to be supported. The World Bank supported project wants to give matching grants to companies that try to be competitive. It will not select markets or take the entrepreneurs by the hand. Rather give them a grant if they want to export, or if they need information about standards for flowers. It would be co-financing to assure that they are really interested as well. For the Tanzanian bureaucracy it is important that the president is in favor of small enterprise development and decentralization. Many officials still see enterprises only as units to be taxed, capitalist bloodsuckers. The advantages of working with the Tanzania Private Sector Foundation is that the organization manages to speak with one voice in the negotiations with the governments. There is a need for more non-Asian black entrepreneurs, but it is all a question of networks, which are more difficult to access for an African entrepreneur. Cultural diversity is generally considered an asset for a country or a city. In Tanzania it is also important to maintain the multicultural character of the society. Even if this means that black entrepreneurs need preferential treatment, if only because entrepreneurs from Europe or Asia have better access to business support services (BSS). 5.6.3 Can we increase the positive effects of these flows? Let us try to stimulate these flows by: · Stimulating entrepreneurship · Promoting innovation in agriculture and in off-farm employment (non agricultural activities) · Developing literacy and professional skills · Reinforcing rural urban relations · Increasing the purchasing power in rural areas Rutasitara (2002: 85-86) mentions as characteristics of the non agricultural activities that these are typical sole proprietorship activities, mostly managed by family members. The activities tend to be intertwined with the agricultural calendar and usually use local 38 inputs. Income may be unstable and hired labor is limited. However, there are low capital requirements to enter into the rural non-agricultural activities and the technology tends to be simple. These activities are easily differentiated by gender, offering employment (and hence income) to the disadvantaged. However, rural households with non-agricultural activities suffer less of poverty (Rutasitara, 2002). He gives characteristics of these activities, which can be considered the origin of many urban informal sector activities and hence we will draw on this study. A lot of the products and services end up in cities and towns. The rural informal sector, or these non-agricultural activities are important examples of rural-urban linkages (Bagachwa, 1999). He notes that a dynamic rural informal sector not only employs 21 percent of the nation's labor force, but also provides estimates of the value added of the sector. He considers the sector (including the urban informal sector) contributes one third of the total official economy's value added. 5.6.4 What is required for more positive effects? This requires change in policies, attitudes and investment: · Policies: for example improve purchasing power through off-farm employment in labor intensive rural public employment schemes to build required infrastructure · Attitude: stimulate decentralization, promote entrepreneurship and facilitate flows · Investments: more in education, health, infrastructure, micro finance and innovation 5.7 More evidence of positive developments in the rural areas Positive developments can be observed in the rural areas. For example the increase in horticulture (stimulated by a number of foreign investors). Secondly the Regional administrative secretariats are now assumed to deliver economic development supports services that focus on production related activities in agriculture, livestock, commerce, industry and natural resource sectors. Of course there is abject poverty in Dar es Salaam and in many of the country's regions (for example the article on Tanga region in Citizen 25-8-06). However, there are also a number of real positive developments. In the first place the new president is really inspiring and still has a good press, even after being almost 9 months in office and despite the drought, the energy crisis and problems with the EAC. I will distinguish some general favorable conditions and some factors that favor in particular the rural areas. 5.7.1 General favorable conditions 1. A coherent vision on developments and trade policies. The plan 2025 was very ambitious but it serves a purpose. Japan is still pushing the Tiger idea or a mini-tiger. It is based on Malaysia because of the ownership, the role of locals and the role of local business men. Tanzania has constraints like infrastructure and insufficient technical education but the intentions are there. With the help of the ESRF the government has developed a private sector development strategy and it will also have a SME policy and have already a trade policy, but there is not always the implementation. The National Strategy for Growth and Reduction of Poverty (NSGRP) is a good 39 oversight of what the country does and wants to achieve. Tanzania really has a strategy for the future, in fact it has several. There is for example the plan 2025, besides the mini tiger plan and De Soto's formalization approach.xx The Tanzanians may have listened to too many people and they are trying to do too many things at the same time. The result has been the higher economic growth and the attractive and more investors-friendly climate. For the trade policy, see MITI (2003). 2. More attention for decentralization (see above chapter 3) 3. More foreign direct investment Tanzania received in 2004 more foreign direct investment than Uganda and Kenya, namely US$ 470, 237 and 46 million. 4. More property tax collected (see box 8) 5. More joint ventures with US and Chinese companies are promoted. Tanzania woes Chinese investors and although these are mainly interested in selling their products and buying raw material in Africa, the last mission made some promises. According to the Guardian (24-8-6) the vice president has pledged government support to investors trying to realize their goals. The Chinese answered that they seek investments in business industry, minerals, agriculture and irrigation and has helped Tanzania with the development of an Export Processing Zone (EPZ). Box 8 More property tax collected Typically Municipalities also feel more pressure from their inhabitants. The City Council for Dar recently decided to put its vision and mission on a sign outside the building. For the Ilala Municipal Council the request to improve social infrastructure on the outskirts of its area did not go unnoticed and even hit the national newspaper The Citizen (21-8- 2006). Property tax collection in Dar es Salaam reached TSh. 2.72 million during the fiscal year 2005/2006, which is a huge increase if compared for example with 1994 when it amounted to only TSh 112 million (Citizen 25-8-06). 6. The government stimulates universities to pay more attention to improving `business skills' of students and staff. Special programs have been developed at Dar Es Salaam University and also several institutes have specialized in providing appropriate technology for these self starters. 7. Several attempts are undertaken to improve market access as well as productivity. A number of activities envisage improving competitiveness. Tanzania's Business Environment Strengthening project (BEST) for example is providing support to private 40 business organizations in Tanzania. They can get grants to advocate for changes in the business environment. Money can also be obtained to allow such organizations to improve their capacity to advocate more effectively for changes in Government policy, regulations and so forth. Box 9 Rural development initiatives mentioned in the newspapers 1. Lack of laws dogs Public private road venture 2. Vodacom's mobile\health clinic provides services in Arusha 3. Lake Victoria environment wins protection 4. Leadership (because of a speech by president Kikweta) 5. Corruption informers must be protected 6. Dodoma to breed beef cattle 7. President: no delegation of duties to regional commissioners 8. Experts quash EAC proposal 5.7.2 Some factors that favor in particular the rural areas Some of the factors that favor in particular the rural areas are: 1. An improved infrastructure (above and chapter 4) 2. Many rural development initiatives (see box 9) 3. Activities to improve the competitiveness of agriculture (see above) 4. Efforts to extend the micro credit network (already mentioned) 5. Indicators showing increased activity in the rural areas These indicators range from the sales of telephone cards to consumption of construction materials. The impression is that in the rural areas a lot of phone cards are sold and everybody is calling. That is a good sign that people are connected. The number of buses would also be a good indicator. All the consumption goods that one finds in the urban areas can also be bought in rural areas. Decentralization has provided more infrastructure. There are local multipliers for all these expenses. 6. A trend to diversification For example the rise of Jatropha biofuels (Van Eijck, 2006) or the surge in horticultural investments. 5.7.3 Factors which could constrain the expected dynamic development Some factors which could constrain the expected dynamic development: 1. Increased corruption 41 2. Continuing to hinder entrepreneurs. There are lots of examples ranging form the gold mines to the small entrepreneurs who have to move their businesses out of the center of Dar es Salaam or Arusha 3. Socialist-oriented land policies (Citizen, 25-8-06) 4. Conflicts between different population groups 5.8 Conclusions and recommendations 5.8.1 How can the rural areas benefit even more from these urban-rural flows? It is possible to think in terms of an alternative approach to the urban rural dynamics. This approach would facilitate these flows and try to maximize their positive effect, just like the government is trying to maximize the effects of a gold mining company for the rural areas where it is working. Stimulating the flows would mean: · No longer local economic development, but linking to global formal/informal value chains! · Not: more protection because it will not work · Rather find out where Tanzania can be competitive · Develop literacy and professional skills · Stimulate entrepreneurship · Promote innovation in agriculture and in off-farm employment (non agricultural activities) · Reinforce rural urban relations · Increase purchasing power in rural areas However, this also requires a change in attitude, policies and investment priorities: · Policies: for example improve purchasing power through off-farm employment in labor intensive rural public employment schemes to build required infrastructure · Attitude: stimulate decentralization, promote entrepreneurship & facilitate flows · Investments: more in education, health, infrastructure, microfinance and innovation. Financial deepening is important. 5.8.2 Policy recommendations 1. Move from an improved business climate to entrepreneurship development in the urban and rural areas, for small entrepreneurs and farmers The recommendation is to support the current reform process and in particular instilling the confidence and skills in small entrepreneurs and farmers and supply them with ideas how to make a living. This requires involvement of the private sector, of NGOs and organizations of producers themselves. 2. Maximize the local effects of investments Finally the question is how can the rural areas benefit even more from these urban-rural flows? The challenge is to maximize the positive interactions between the urban and the rural areas. Sectors like flowers and horticulture, mining and tourism need to be 42 stimulated to maximize the local impact. However, it is not desirable to go against the effects of globalization. Hence the European countries should try to re-activitate the Doha round and turn it in to a real development round, as was promised at the beginning. Secondly, we realized that we have left out certain negative and positive aspects of rural areas in this conceptualization: a. The environmental impact, the footprint approach b. The criminal relations: smuggling, illegal woodcutting, prostitution, etc. c. The recreational aspects of staying in the rural areas, to the extent that this is not measured as a service. Many may prefer to go there for a holiday or retirement. 3. Cities can be engines of development for the rural development process Many business support services, new ideas, technologies and expertise are located in cities. The challenge is to strengthen the spontaneous process of flows between urban and rural and maximize its potential positive effects. Coming back to Jacobs (1970) these cities provide ideas, products and markets to the rural areas and in this way can contribute to their development. It would not be wise to isolate the two systems too much. 6. The role of the informal sector to spread development beyond Dar es Salaam 6.1 Introduction Tanzania has witnessed a fairly good macro economic performance during the last three years (at the average 6.2 percent economic growth per year), but the impact on poverty reduction is limited. Only in Dar es Salaam there was clearly observable poverty reduction (World Bank, 2005). The issue addressed in this study is: why did poverty alleviation happen in Dar es Salaam and hardly in the rest of the country? What are the transmission mechanisms distributing the positive effects of economic growth over different cities and regions? How does this economic growth affect the poor in other cities and rural areas? We will put this issue in the framework of being a global city (London for example) or a peripheral node (Dakar for example). We will look to what extent Dar es Salaam is a global city or a peripheral node with growth potential? Box 10 Global cities 1. Global cities like New York, London and Tokyo are the production sites of global control over the lengthening production chains of which even the industrial home workers in remote rural areas are now part. 2. Globalisation is not just about networks and resulting flows, it is also about places. 3. This concentration is the result of globalisation and can be explained by: a. Financial service providers concentrating around important financial centres. 43 b. The need to have an innovative milieu requires networking among specialized business services t produce new managerial technological solutions. c. A high tech infrastructure is required to assure permanent connectivity d. A specific logistical infrastructure based on the previously mentioned functions. 4. Professional business services are the managers of the decentralization and globalisation process. Source: Sassen (1991). 6.2. Global cities and peripheral nodes Cities are important in the development process. They are competing between each other and globalization is increasing this process. According to Sassen (1991) globalization exposes cities to ideas, goods and services from elsewhere, which challenges the local production of industrial goods and of services, creating global cities The old model of global integration was based on increased exports of goods and services. Developing countries were considered the periphery of a dominant capitalist system residing in Europe, North America and Japan. The dominant metropolitan core controlled the process from within. The original meaning of metropolitan zone of influence meant forming part of the motherland as distinct from its colonies. The term is now used for an ever expanding urban area with some supra regional role in trade or providing services to a larger area.xxi Besides global cities there are peripheral nodes, which may have an important regional function. In Sassen's book global cities go beyond a region or metropolitan area. She points to the role of knowledge, information and business services in the longer existing process of concentration of global production chains in the global cities or metropolitan areas. The characteristics of the global city are presented in the following box. A critique of Sassen her definition of global cities could be that she focuses on one particular type of city, namely the global service centre. On the other hand there are lots of cities in for example China and India that are centres of manufacturing activities and export an important part of their production to other countries, being part of all kinds of global value chains. The governance structure of such value chains (commanding what is happening in the chain) tend to be concentrated in such global manufacturing or services cities. These cities can compete with other aspiring global cities if they are fed by knowledge activities such as research and development, finance, marketing and design activities.xxii However, the dynamics of the current world economic system (Wallerstein, 1989) is that other cities than the traditional metropoles are trying to take over the role of leader of production chains and try to become centers of knowledge and innovation as well. This process of competition offers chances to cities as far apart as Nanjing (Van Dijk, 2006a) and Dar es Salaam. If they don't succeed in becoming global cities in the Sassen sense, they may still be global manufacturing cities (say Shenzhen in China) or may become peripheral nodes in the world capitalist system with a potential important role to play in the future for the national or supraregional area surrounding them. The question remains 44 how such modes transmit development to the rest of the country and what the role of the informal sector is in this respect? 6.3 The context of economic restructuring and urban and regional growth After years of structural adjustment, liberalization, poverty eradication policies and programs and donor support, Tanzania has created a healthy macro economic context in which growth can take place. The Tanzania Private Sector Foundation (TPSF) considers, that now the investments in reforms in Tanzania start paying off. The government of Tanzania nationalized and placed all enterprises under state ownership and management up to the mid of 1980s. This movement had a negative impact on private sector development in the country. In 1992 the government launched the Parastatal Sector Reform Program (PSRP) to privatize public enterprises. Privatization sought to give private sector a frontline role as the engine of economic growth. Following that, the participation of the private sector has increased and management of the restructured economic activities in the country has improved, particularly in the City of Dar es Salaam. In order to strengthening business and investment environment, consultative dialogue meetings took place involving the government, private sector under the auspices of the Tanzania National Business Council (TNBC). Moreover the government continued to implement the Business Environment Strengthening for Tanzania (BEST) Program. 6.4 Role of urban informal sector in spreading development and alleviating poverty In other countries the informal sector is residual in Tanzania it seems to be the other way around. Some say 90 percent of the economy is informal (Tripp, 1997), while the remaining 5% are employed in the formal sector including the government and public co- operation. With such a definition 50 percent would then be in traditional farming. Other authors go as far as stating that the informal sector employs two third of the population in Tanzania (ILO, 2005). It doesn't seem useful to put traditional agricultural activities in the category informal sector. In this paper the term will be reserved for informal income opportunities in cities and non-agricultural activities in the rural areas. They are considered informal since the units involved have no legal status and hence the people working there do not receive the legal minimum wage or benefits from the social security system (Van Dijk, 1987). The difference between the urban informal sector in West Africa in the seventies and the current informal sector in East Africa seems to be that this urban informal sector has become part of the global economy. For example informal tailors no longer compete with modern textile and garment industries in Tanzania, but they have to produce at lower prices than Chinese textile and garment industries and have to come up with more attractive products than the second clothing from Europe, which are almost everywhere available nowadays. ILO (1991) made an effort to estimate employment in the informal sector. The report suggest 2 million people are involved, which would be 2.5 times the number of people with formal wage employment. According to the ILO employment in the informal sector 45 would be growing at 3.5 percent per year. The integrated labour force survey 2000/1 (NBS, 2002d) measured informal sector activities by the number of households involved. It turns out that in Dar es Salaam 62 percent of the households are involved informal sector activities. The percentage is 61 for other urban centres (used to be 42). In the rural areas 27 percent of the households have informal sector activities (used to be 21; URT 2002). In terms of number of people the labour force survey found it would be 1,439,847 people doing it as a main activity and 1,363,010 having informal sector activities as a secondary activity. This suggests that currently a much larger number of people are involved in this sector than according to ILO (1991). Tripp (1997) stresses the importance of women in the urban informal sector. NBS (2002d) provides the figures. In Dar es Salaam and in the rural areas there are slightly more males than females in both the main and secondary activities. In the other urban areas, more females were employed than males in both the main and secondary activities. According to different World Bank sources more than 50 percent of the urban population works in the urban informal sector in Tanzania. However, their activities are not covered in detail in the statistics (what kind of activities and how much does it earn a person). More importantly this sector has become a mechanism to generate and redistribute growth. Hence we should understand better how it functions. The informal sector is the result of and resulting in flows of people, money and goods. These mechanisms of the urban informal sector like migration, remittances and smuggling will be discussed in the following sections. Table 8. How informal flows affect different locations in Tanzania Factor Dar es Salaam Arusha Kilmanjaro Flows of people Migration Strong inward As well in as out of Mainly out of the migration the region region Lifetime migrants 1,319,360 249,971 113,743 2004 Flows of money Loc. tax revenues Increasing Increasing Declining Remittances, In and out but Coming in for the Coming into the investments and constantly in and going out for region to support government increasing the out migration. those left behind, expenditure Totals increase amounts decrease Flow of goods and services Local products and In and out, In and out, from Flows are not as small scale services including from abroad, partially important as before abroad illegal Source for migration: Mbonile (2004). 46 If more than 50 percent of the Tanzanian economy is informal, the current dynamics should also be reflected in the informal sector. The six percent economic growth will also apply to this sector, because a lot of money is spent in the sector and most of the money generated in the sector will be spent informally and bring about multiplier effects. Hence, the question is what determines the further development of the informal sector? Where do the people come from? How do they finance their activities? What is the role of the government in their development? A first impression of some relevant factors is given in table 8. For each flow we try to indicate how it works out in Dar es Salaam, the Arusha and the Kilimanjaro region. Tripp (1997), ILO (1978, 1991 and 2004), Wuyts (1998), Bagachwa (1995 and 1999) and others help to get a picture of this sector and in particular of the investments made and income derived from working in the sector to understand its re-distributive function.xxiii A number of studies deal with rural urban linkages (Bryceson and Jamal, eds, 1997 and ESRF, 2001). Finally some studies are very much policy oriented (ILO, 2004 and Mitullah, 2005). ILO (2004) focuses in particular on the restrictive regulation and how to deregulate to allow more informal sector development. 6.5 Rural informal sector activities As mentioned, rural households with non-agricultural activities suffer less of poverty. Rutasitara (2002) gives characteristics of these activities, which can be considered the origin of many urban informal sector activities and hence we will draw on this study. The importance of non-farm activities is very much linked to the possibility of markets for these products. These products can be marketed more easily than agricultural products, which all come on the market at the same time and require the farmers to deal with only one party, who has a much better understanding of the price. Plus agricultural products face transport and storage cost; it requires more information and more capital. However, it is more difficult to get the message to the rural population that non-farm activities are more rewarding. It also requires knowledge and technologies and some small investments. Micro finance schemes can be very useful (see Gallardo et al., 2005). Table 9 Most important rural informal sector activities in Tanzania Small businesses in rural areas: trade and Manufacturing activities services Trade of: - Agriculture related products, such as Agro processing activities grain, vegetables, fish and meat Blacksmithing - Industrial products Carpentry - Construction related products Construction material: mud bricks, stone - Wood for cooking bricks, wooden planks, etc. Hawking and street vending Farm implements Repair: Tailoring - bicycles Handicrafts and other crafts - cars Pottery 47 Major food crops grown in the Dar es Salaam region are: cassava, sorghum, maize, rice, sweet potatoes, bananas, and varieties of legume. Cash crops include: cashew nut, coconuts, oranges, pineapples, and mangoes. Farming of these crops is done on small scale plots and land tilling is mainly done by using hand equipments. Very few people are using tractors and traditional upgraded technology. 6.6 Different activities in the urban informal sector Table 10 gives an impression of the activities in the informal sector of Dar es Salaam. It should be noted that very different activities in this sector have very different problems. This suggests there is also not a one size fits all solution. The ILO (2005) perspective is getting the informal sector out of the margin into the mainstream. This requires removing a number of bottlenecks and introducing some specific facilities. For example, the ILO considers it very important that there will be a plastic recycling plant in the country soon with the help of UNIDO. Scrap metal has already become valuable, and the hope is that people will recover plastic as well. Currently most of the plastic bottles are compressed and exported to China! Table 10 Most important informal sector activities in Dar es Salaam Small business license holders in Dar es Manufacturing activities Salaam: trade Operating a stall for assorted items Tailors Selling firewood Making buns, rice cakes and pastries Selling buns, rice cakes and pastries Working as a carpenter Vending fruit Working as a cobbler Selling fried chips, chicken, eggs and or Repairing watches barbecued meat Slaughtering animals Vending fish Welding Source: Trip (1997). The informal sector may help families out of poverty, but it is also very much a subsistence sector in the rural areas. The question is how to get these people to invest and to expand their businesses. A push from outside may mean much more for the local economy, than efforts to develop policies for the informal sector in the capital. Such a push could be the training received, a trip to another city or country or financial support from a family member living elsewhere. The informal sector may have its own dynamics, but currently lack of credit facilities or fear for taxation does restrain it (Kweka et al., 2004). 48 6.7 Mechanisms in the urban informal sector: flows of people Because of its high population growth Tanzania has a relatively young population. According to the 1988 census 46 percent of the population was under 15 years of age. The total population increased from 7.5 million in 1948 to 34.5 million in 2002 (URT, 2003b). Migration is common in Tanzania. People are moving form high density areas to less populated areas and people are opting for living in the cities. An important transmission mechanism for skills and income is the labor market and its functioning is influenced by good urban management. One could look at these migrants as physical capital, in the form of human bodies, being brought to the regions with more potential or to urban areas, where the capital will generate a higher rate of return than in the rural areas. There are also international flows of people. The number of people leaving Tanzania is not known, but Mbonile (2004) estimates international migration towards Tanzania to be 236,775 in 2002, based on the 2002 population census. 12 percent of these migrants settle in Dar es Salaam , 2.1 percent in Arusha and 2.3 percent in Kilimanjaro. In total the census reveals 4,671,641 lifetime migrants, or even twice as much if in and out-migration movements are counted separately. He also notes big differences per region. Rural to urban migration played a big role in the urbanization process in Tanzania in the first two decades of the country's independence. The major factor for rural youths to move to urban areas has been the search of a better life especially aspiration for salaried jobs, access to services and social facilities (URT, 2005). A lot of people seem to go directly to Dar es Salaam (Mlingi and Asey, 2005: 71). There is less this two-step approach of going to regional capitals first in Tanzania. For Dar es Salaam City, the number of people born outside Dar es Salaam region was 56.3 percent in 2004. This means that more than half of the city's population was migrants. In the same year Arusha had slightly more than half i.e. 50.7 percent were not born in the localities of the Arusha region. This shows that immigration has had a big impact on the population growth of major urban centres in the country. However, since the 1980s, natural population increase has also become a major component of urban population growth. Table 11 Urbanization trend in Tanzania over 1967-2002 1967 1978 1988 2002 Total % Total % Total % Total % Tanzania 786,567 6.4 2,412,900 13.8 4,247,272 18.3 7,943,561 23.1 Mainland 685,092 5.8 2,257,921 13.3 4,043,684 17.9 7,554,838 22.6 Zanzibar 101,475 28.6 154,979 32.6 203,813 31.8 388,723 39.6 Source: URT (2005). Towns like Dar es Salaam, Mwanza and Tanga, which each had a population of less than 100,000 people in the 1940s and early 1950s had already big population by 1988. Dar es Salaam, for instance, as a region accounted for 5.9 percent of the country's total population in 1988. Tanzania urban population has increased from 6.4 percent in 1967 to 49 23.1 percent in 2002 (table 11). Tanzania has also shown high urban population growth rates for the last three decades compared to other developing countries. Based on the 2002 Population and Housing Census, Dar es Salaam had 2,487,288 people of whom 1,254,853 were males and the rest females. The city's population grew from 25,000 in 1919; 45,000 in 1945; 128,742 in 1957; 272,821 in 1967 and to 843,000 in 1978. The 1988 census recorded the city's population to be 1,360,850 and in 2002 the city population was 2,487,288 million. The growth of informal sector has contributed to rural-urban migration. The major pull factor has been the search for a better life especially aspiration for employment opportunities, access to better services and social facilities. This in turn has put pressure on the utilization of undeveloped and developed facilities in Tanzania's cities. Rapid growth of squatter settlements, peri-urban shantytowns and unsanitary environmental conditions are some of the common problems. There are also growing health and social problems in these cities. 6.8 Informal flows of money Components of the flows of money, one of the mechanisms of the urban informal sector, which can be traced, are remittances, loans and savings. Remittances are a tricky concept, but the amounts involved are substantial. Conceptually remittances are transfers without an exchange, while otherwise goods are exchanged for barter or money. The NBS (2002a: 155) registers the net remittances (received minus provided) but doesn't give a clue how the money is used. Remittances and social networks are the key to survival strategies in the rural and urban areas. It turns out that net remittances make up 3.5 percent of the monthly income per household in Dar es Salaam. In the rural areas the percentage is 2.1 and in other urban centres 3.2. The average for Mainland Tanzania is 2.4 percent. The urban informal sector mobilizes a lot of funds and generates income opportunities for the poor. The current policy trend is formalization, where poor people are given titles for land and other assets (their enterprise), which may make it easier for them to obtain loans. Loans will be discussed below, while savings in the informal sector are unknown because they don't pass through the formal banking system. Again micro finance and saving schemes could make these amounts visible. 6.9 Flows of goods and services Tripp (1997: 164) distinguishes different informal marketing arrangements. Between members of one village or of a nearby village. Secondly, shuttle markets, where the products are taken by bus or truck to a nearby town and then interregional trade where they are sold elsewhere and other goods are brought back. Finally there are open urban markets and export markets, where goods can arrive formally or informally. The Quarterly report of industrial commodities (NBS, 2002b) provides data on industrial production in 1985, 2000 and 2003. However, the location of the production and the destination of the products is not indicated. If this is the case for formal sector products it 50 can hardly be expected that such data are available for informal sector. However, we have learned form several sources that the two sectors compete. The CTI doesn't consider the competition from the informal sector as very damaging for their members. In fact they want the informal entrepreneurs to become members of Confederation of Tanzanian Industries (CTI). For that reason they have diminished the membership fees for these smaller companies. However, they note that the bigger enterprises can also gain access to external finance, which is difficult for the small ones. The flows of goods are sustained through small-scale transportation and often would be called smuggling because the formal requirements are not met. These informal movements of goods and services in Tanzania are not always covered in the official statistics. However, although large quantities of goods and services `exported' are not covered, the information is well known, for example concerning: 1. The type of food going from Tanzania to Kenya 2. The cattle going on hoof, even to Uganda 3. Other raw material: coffee, tea and cashew nuts going out of Tanzania, often illegally 4. Sugar comes in from Malawixxiv 5. However, also hides and precious stones (Tripp, 1997: 87) 6. Textiles, possibly through under invoicing or through Mombasa 7. Chicken from Zambia 8. Prawns to Botswana 9. Makonde carvings and other crafts to the west (Tripp, 1997: 96) 10. All kinds of industrial products from Kenya usually essential goods 11. From Zanzibar they smuggle everything, using small boats. If you look at the statistics of the Island, the inhabitants import much more than they could possibly absorb The government claims it contains smuggling. That is true for certain products (dry batteries and electrical products form South East Asia), but in other areas there are problems. Smuggling is not very important according to some officials we talked to. They consider that they manage to control it. Nobody likes to see it growing. One way to reduce it was reducing the duties in the framework of different trade agreements. The Regional Commissioner in Arusha district also coordinates with partners on the other side of the border. The main category of smuggling is industrial products produced cheaply in Kenya according to government officials. The attitude of officials towards smuggling is that they don't think Tanzanians smuggle so much. These are usually small producers and people on the other side of the border. However, the International Monetary Fund (IMF) published a study showing that 50 billion of Tanzanian shillings per year is lost as government revenues in Tanzania, because of smuggling along the Indian Ocean coastline, according to a newspaper article (Citizen 12-7-2005). 51 6.10 Promoting on the role of the informal sector An alternative model of development is suggested for example by the ILO (1972) and by De Soto (1989). The emphasis would be on developing the informal sector further. These units are not always registered, do not pay all the different taxes or the legal minimum wage, but they certainly make up a local sector, creating a lot of employment and with a lot of entrepreneurial capacity. We should be tapping into it. Most small-scale activities are still producing for the domestic market and some for informal exports. New exceptions are the flowers and certain fish from Lake Victoria such as the Nile Birch and the Tilalupa, which can be exported. The availability of finance is in general not considered the bottleneck in Tanzania. It is available; the issue is to design mechanism (micro finance and guarantee arrangements) to get it into the rural areas and small towns and to reach the farmers and the small entrepreneurs. Different levels of government can also mobilize more money through partnerships with the commercial and non-commercial private sector and with other levels of government. The informal sector plays an important role and has a tremendous potential to spread development. We have to look for those activities that can grow out of the sector when they reach a certain level. Typically the gem stones trade, these stones are bought and sold in certain streets of Arusha, are in the informal sector. There is no viable formal industry and the government should not do it. It happens in this way as well, but could develop further. The growth of the informal sector could be quantified in terms of number of people involved. However, it is more difficult to get estimates of their turnover, income, investments and exports. There are certainly a number of indications that these variables are growing at the same speed or faster than the overall economy. If asked many small entrepreneurs claim a lack of finance is their major problem. The question is whether loans are the real bottleneck. Other problems for micro and small entrepreneurs are a lack of: a. Space b. Infrastructure c. Market outlets d. Skills e. Innovative capacity 6.11 Current policies of the government for the informal sector and SMEs Policies for of the government for the informal sector and SMEs in Tanzania are very much based on what Altenburg (2005) calls `the minimalist approach to private sector development'. De Soto (1989) is an important thinker behind this approach, which according to Altenburg fails to point to the complementary factors, which are important for micro and small enterprise promotion. We will discuss now some current policies and indicate what doesn't get sufficient attention in Tanzania. Current policies of the government for the informal sector and SMEs in Tanzania include: a. Formalization policies b. Credit or loans 52 c. Export promotion d. Other SME policies The current presidential candidate suggests he is very much in favor of local small businesses, so SME development it is a political issue now. From 2004 the small entrepreneurs can get a license fee without paying. Also the taxation system is more smooth and predictable. There have been also lots of policy reforms for the better. 6.12 Formalization process In order to empower informal sector, the government has prepared the Property and Business Formalization Program (PBFP). The Program seeks to facilitate the transformation of informal (extralegal) properties and businesses into formal legal entities. Its aim is to enable Tanzanians to use the formalized resources as collateral in accessing credit from financial institutions and thereby enhance economic growth and ultimately reduce poverty. This requires according to De Soto (1989) title deeds. However, the question is what would be the value of such a deed in a village? Does Tanzania have the legal system to allow financial institutions to sell such property in case of non-repayment of loans? The formalization work has already started in Dar es Salaam. The Business Registration and Licensing Agency (BRELA) has been created. It was established in 1997 and launched at the end of 1999. It provides eight reasons to register as a business (see box 11). The Agency is responsible under the MIT for business facilitation and regulation. Its mission is to regulate and facilitate business operation in conjunction with other partner institutions in Tanzania, to ensure that they are in accordance with sound business and commercial principles. Box 11 BRELA reasons to register your business 1. Giving your business undertaking its legal life 2. Giving you the exclusive right to personally use your company and business name 3. Enabling you to enjoy the pride of being an honest Tanzanian or Foreign business undertaker 4. Giving you the right to enter with confidence into the competitive business arena using your company name and business name corporate identity 5. Granting you exclusive legal right to enjoy the product of your invention or innovation 6. Protecting your intellectual property rights, thereby stimulating the urge for further inventive, innovative and creative activities 7. Enjoying protection of your literary and artistic works against infringement and privacy 8. Affording you the chance to contribute to national economic prosperity 53 The slogan is that BRELA provides business legality in Tanzania. Besides registration they are also granting patents, overseeing copyrights and involved in business and industrial licensing. Mainstreaming informal enterprises into formal sector faces some challenges, which include: inadequate knowledge of the dynamics of the informal sector, and inadequate strategically located areas for allocating space to the informal sector operators. The PBFP program is currently limited to the first two phases, one of Diagnosis. The first phase has been completed. The second phase of Reform design will start when the evaluation of the first phase is completed. In Tanzania there is a lot of talking about formalization of the informal sector. In practice this does not mean much more than abolishing the license fee if your revenue is below 20 million Tanzanian Shilling (TSh). Secondly, small entrepreneurs don't have to renew the license, like in the past. If the revenue is over 20 million, they pay 20,000 TSh, but only once. For tax purposes, if an enterprise doesn't have records, it pays a lump sum. If it keeps records, it pays a percentage. 6.13 Credit or loans Financial experts estimate that in total there is only one billion US dollar in loans in Tanzania. The top five commercial banks probably provide 70 percent of that. The five international banks half of it. These are Standard Chartered, City bank, Barclays, Stanbic and Cooperative and Rural Development Bank (CRDB). The big international banks are approximately good for 200 million $ for Standards, 150 for Barclay, 100 for Stanbic and 50 for the City Bank. Then the local banks NBC 200, CRDB 150 and Exim 80 and another 80. However, these loans don't go to SMEs, let alone the informal sector. The sectoral composition of loans is: large local, 55 percent, multinational firms (without mining, which is self supporting) 20%, SMEs 10 and consumer loans 15 percent.xxv Agricultural gets only 10 percent of this the rest going to trading and manufacturing. The most revealing figure is the growth of the loan portfolio. With an economic growth of 6 percent, credit is growing at 25 percent per year. There is no mechanism in place to provide agricultural credit. This is a reason for some people to argue in favour of an agricultural bank, but the bankers we talked to consider this is too interventionist and the World Bank would be against it. Micro finance is not very much developed in Tanzania (Gallardo et al., 2005). Micro credit can play an important role for that micro and small enterprise development in the future. This is not the way the formal sector banks will go, but there is scope for alternative micro finance institutions in Tanzania to expand their business. Despite its name, the National Micro-finance Bank (NMB) is not very much into lending to small and micro entrepreneurs. Even now that the Dutch cooperative RABO bank has taken a major share in the NMB micro finance for micro and small enterprises will not increase substantially in the near future. The NMB actually requires a regular income of 100 dollar per month to be eligible for loans. 54 Investing by Tanzanian banks in risky business is not encouraged because of the high rate of interest paid on treasury bills (9 to 10 percent and risk free, while inflation is only 4 percent). Micro loans is currently not interesting for the formal sector banks, because of the cost involved in lending to one entrepreneur would be a few hundred dollars. With a saving component it may become interesting in the future. 6.14 Export promotion Several efforts are made to promote exports by small and medium enterprises in Tanzania. The Ministry of Industry and Trade with support from Denmark (DANIDA) focuses on a limited number of medium scale companies, which are willing and able to export and consider new markets (outside the region or Africa). The project helps these enterprises to overcome the barriers to export and to focus on quality requirements, helping them to satisfy international standards. For practical reasons the project focuses on a limited number of cities. The project does provide grants to help Tanzanian companies to export. The Danish consultant to the Ministry of Industry and Trade claims that they don't deal with the informal sector, but she agrees that there is some kind of continuum between informal micro and formal SME. They consider the main bottlenecks in the sector are regulatory and administrative constraints, taxes and compliance cost. Although the administrative system is decentralized, entrepreneurs often have to go to Dar es Salaam for tax purposes or to obtain an import license. 6.15 Other SME policies Innovation is not a big thing in the small enterprise sector in Tanzania. A lot of entrepreneurs are mainly copying what is happening elsewhere. Also incubator centres don't exist yet. Still there are many Tanzanian producers in sectors like cloth, vegetable oil, soap, etc., which could benefit from innovation or incubation centres (Bongenaar and Szirmai, 2003). Van Dijk and Sandee (2003) suggest using the term innovation in the broad sense of introducing new ideas, new products and ways of working or financing or selling the production of micro and small enterprises. The Technology Development and Transfer Centre (TDTC) at the university of Dar es Salaam now produces for example sieving machines for small scale mining (Sunday Observer 7-7-2005) and could be used for developing/adapting technologies for other activities as well. Although most of these informal sector enterprises are small, they trigger local economic development. However, there are export associations in Tanzania, even if these sectoral institutions are not always called export associations. For coffee, horticulture, sisal and mining such associations do exist and could be used. Also Trade Fair sales can help SMEs. The Sunday Observer (17-7-2005) reported that sales during the Dar es Salaam International Trade Fair (DITF) have been twice the record sales of 2004 (TSh. 78 versus 38.9 million in 2004). The ILO promotes developing the urban informal sector in East Africa by promoting subcontracting of solid waste collection and treatment to small enterprises (Mitullah, 2005). Preparing training manuals is part of their approach to the issue. Local 55 organizations develop the material. Other ideas mentioned in ILO (2004) to promote the informal sector are: - A new pact between Central government and small businesses - Improving their access to public procurement - Development of entrepreneurship - More efficient institutional arrangements 6.16 Conclusions on promotion policies The process of formalizing the informal sector, which has started in Tanzania under the influence of De Soto (1989), has some dangers. One is that it will kill the sector, whose secret is that they are not paying all the taxes and satisfying all the rules and regulation developed for the formal sector. Secondly, it is noted that the large number of activities currently undertaken for this sector may stimulate people to actually go to the cities and try to benefit from these programs. Hence it could increase rural urban migration. Finally there will always be an informal sector, even if you make registration easier and paying taxes more transparent and justifiable. New people will start economic activities without bothering to register or to pay taxes. Most informal sector activities don't believe in formalization and hence they want to stay out of the tax and credit web. However, the dynamic part of the informal sector will eventually make the shift. The approach of De Soto, giving everybody titles, is not a sufficient condition for successful small enterprise development. A lot more (education, training, markets, innovation) is necessary to turn the small and micro entrepreneurs into dynamic business(wo)men. Another point of departure for the promotion of the informal sector would be to take the real problems of these small entrepreneurs as the point of departure. In Trip (1997) one finds a number of positive and negative factors, summarized in the following table. Table 12 Positive and negative factors for the urban informal sector Positive Negative Rotating credit societies 117 Licensing restrictions 103 Networking strategies 122-24 Oppressive laws 103 Women taking control over their own lives Harassment 124 117 Battle over grounds 158 The moral economy model 127 One can add: taxes, lack of space, lack of Ending enforcement policies 193 credit, etc. Souce: Tripp (1997, with page numbers in the table). A positive policy environment largely determines the further development of informal sector activities. If their contribution to the economy is recognized and promoted the sector can really help to spread development over the country. Micro and small businesses face a lot of competition and they need a friendly environment to be able to survive and to develop their businesses. The authorities have at least put in place a lot 56 more positive regulatory framework. The current president from Tanzania was very pro- private sector and CTI hopes that this approach will continue with the next president, after the elections in 2005. As one of the CTI officials phrased it: "these small guys also help our members to sell their products for example in the interior". Hopefully in due course a real businesses development support system (BDS) will develop. 6.17 Conclusions on global city or local node Combining Sassen's perspective and Kanbur (2001), who argues that an economy needs to be understood from below, Dar es Salaam is a node in the periphery of the global economy. However, it is a well connected node and it does use its position to help to spread development in that region, although possibly not enough. With its growth potential and in particular through informal flows of people, money, goods and services some development is already spreading to the interior of the country and probably more than the official statistics show! The challenge is to serve a larger hinterland as well. 7 Dar es Salaam, a dynamic capital with three local governments 7.1 Introduction Dar es Salaam counts about 3 million inhabitants and has the highest density in the country (1793 people per km2). The city of Dar-es-Salaam is located between 6.36 and 7 degrees South of Equator and Longitudes 39.0 and 33.33 degrees East of Greenwich. India Ocean bound it in the east and Coast region on the other sides. This city was declared a Township in 1920, and in 1949 it was upgraded to a Municipality under the first appointed British Mayor Mr. Percy Everett. When Tanganyika became independent in 1961, Dar-es-Salaam Municipality was elevated to a City status and continued to be the headquarters of the then Independent Tanganyika and later the United Republic of Tanzania (URT). The City Council was dissolved in 1972, following a decision of the Government of Tanzania to abolish all local governments in favor of a real decentralized system. However, this decentralized system put too much emphasis on rural development. As a result, the urban centers deteriorated so much that the urban local authorities had to be reinstated in 1978. The city has total surface area of 1,393 km2, which is about 0.16 percent of the entire Tanzania Mainland's land area. It has three Municipalities namely: Temeke, Ilala and Kinondoni. Temeke municipality has the largest surface area of 652 square kilometers, followed by Kinondoni with surface area of 531square kilometers, while Ilala is the smallest with surface are of 210 square kilometers. Urbanization was given impetus due to the caravan trade of the Arabs. For instance, during 19th century, Tabora, Mpwawa, and Ujiji came to prominence as urban centers partly because of the westward penetration of Arab trading activities in East Africa. With the coming of Germans in the 1880's, Bagamoyo became their administration center till 1891 when this was moved to Dar es Salaam, then regarded as a `small town'. Urban centers sprung as satellites of large scale farming areas like Mbeya in 1935, and mining sites like Chunya in 1938. Ports like Mtwara in 1949, and Tanga in addition to administrative centers as well as railway stations like Korogwe and Kilosa that were facilitating commerce, also started to grow. 57 These towns and others during the whole period of colonial period grew at a slow rate. Population growth was slow. For example, Dar es Salaam was reported to have a total population of 25,000 people in 1919, and 45,000 people in 1945. The main reason was that urbanization, as a way of life was considered suitable only for the alien population. Restrictions on the internal movement of the indigenous population played another important role in slowing down urbanization. However, rapid urbanization began to be experienced after independence in 1961. Dar es Salaam now counts about ten percent of the population, but generates between 17 and 18 percent of GDP. The data on poverty show that the capital is the only place where some poverty reduction has taken place in the nineties. The City is divided into three ecological zones, namely the upland zone comprising the hilly areas to the west and north of the City, the middle plateau, and the low lands including Msimbazi valley, Jangwani, Mtoni, Africana and Ununio areas. The main natural vegetation includes coastal shrubs, Miombo woodland, coastal swamps and mangrove trees. Dar es Salaam enjoys some vary favorable conditions: · Strong growth of population · High tax revenues, the city generates 82% of all taxes in Tanzania · Its share of GDP is 17.5% while it counts only 13% of the total population · Manufacturing units 1405 out of 3431 · Establishments 9397 out of 28910 · Licenses 18106 out of 81661 · Buses 6600 out of 11279 · 62% of households involved in informal activities In Dar es Salaam you find the big businesses or some very small indigenous business. Mining and tourism are flourishing but informal sector incomes are very low. In purchasing power parity it is slightly better. However, the differences are too big! The success of Dar es Salaam is due to: · Location plus largest part government's services and formal sector, port · Enormous dynamics, decentralization resulted in more competitive municipalities · Increase in total factor productivity 6.2% · Growth exports and 61% households in informal activities · Some of the seven explanations explored in this report 7.2 Activities at different levels of government 7.2.1 Dar es Salaam City Council Dar Es Salaam City has about 110,850 hectares of land with a potential for agriculture, particularly food crops. Out of this 52,000 hectares are in Kinondoni municipality; 45,000 hectares are in Temeke municipality and 13,850 hectares are in Ilala municipality. It is estimated that about 58,278 hectares of land are currently utilized for food and cash crops production (13,600 hectares are in Kinondoni, 33,000 hectares are in Temeke, and 58 11,678 hectares are in Ilala). Each municipality is autonomous and has the same functions. The fields to coordinate at the city level are: 1. Infrastructure: including roads, fire houses, public transport 2. Protocol 3. Other cross-cutting issues The Dar es Salaam City Council was established under Local Government Urban Authorities Act No 6 of 1999. In 2000 the one City Council system for Dar es Salaam was replaced by four local governments, a city wide City Council and three municipalities or local authorities. They are called Ilala (in the middle, where a small part of the port is located), Kinondoni (in the north) and Temeke (in the south where most of the port is located). There is still a City Council made up by elected representatives of these three councils, who are asked to look after cross cutting issues. There is also a regional governance structure, which partially overlaps. The city authorities for example very well manage urban transport, but officially this is a responsibility of regional government. This implies a potential conflict is the region would stick to its guns. What is the role of the City Council? To coordinate things at the citywide level. But there is not much to coordinate. Citywide programs are the fire stations, aerial photography and a rapid bus station as a joint project. They also collect some taxes. In 2005 there will be national parliament, ward level and presidential elections (every five years). Similarly after 2004 the next local government elections will be in 2009. The whole city has one apex council and within it there are three municipal councils of Ilala, Kinondoni and Temeke. The City Council has a Mayor elected by Full Council from among the 20 Councilors who form the City Council. The City Council comprises of 11 elected councilors from three municipalities, 7 Members of parliament, and 2 nominated special seats for women. The Members of the City Council are not elected directly, but indirectly by the three different councils or municipalities. Each one elects 4 representatives. Seven members are members of parliament, plus the mayor makes 20 in total. The council meets three times a year, but the committees meet every 3 months. The committees are answerable to the council. The City director is the secretary of the council and chief executive director of the council. The Mayor and Councilors serve a five-year term, where as Deputy Mayor is normally elected every year from among Councilors. The incumbent can be re elected as many times depending on the will of the councilors. The Dar es Salaam City Council and the Municipalities operate in the same jurisdictional areas, but each of the Municipal Council has been given a jurisdiction area demarcated by an administrative order. The functions of the City Council and the Municipal Councils are provided for in section 7A of the Act No. 8 of 1982. The City Council performs a coordinating role and attends issues that cut across the three municipalities. Some of the major functions of the City Council are to: preparing a coherent citywide framework for the purpose of enhancing sustainable development; promoting co-operation between the City Council and amongst 59 local government authorities within the city area; supporting and facilitate the overall functions and performance of the authorities; and providing peace and security and emergency services such as fire prevention, and control ambulance and auxiliary police. Apart from these functions, the City Council has the responsibility of providing the following services: management of controlled dumpsite; provision of fire and rescue services; public bus service terminal for up-country and neighboring countries; strengthen and improve secondary schools education; food security and health improvement; road services within the city; city bus services and legal services. Main roads go through all three municipalities and coordination is desired. However, licenses are not managed by the City Council but rather by the Dar es Salaam regional government. Similarly, the port, the national airport and the government offices remain a national affair, even if they are located in the three municipalities. There is a Department of urban planning in the City hall where staff is finalizing the strategic plan for the whole city. With assistance from the World Bank some city hall officials are developing a plan for a Dar es Salaam Rapid Transport System, which is a road-based system with public buses (the private ones will continue), but enjoying a separate lane. Street lighting is also something they want to improve at the city level. It is important for safety and will make the city more attractive for tourists. Sources of revenue of the City Council are a cess for the financial sector and the revenue from the upcountry bus terminal. The City hall is also building an abattoir for the city and is currently looking for partners willing to put some money into it. Becoming a clean city is also a cross cutting issue. The City Council deals with finding the dumping site. However, there is no economic policy formulated at the City Council level. They consider this a job of the national government. There are byelaws for environmental policies. They also promote the use of small contractors for construction projects, to create employment. They can give some loans through the Dar es Salaam Community bank to marginal people. The City level government doesn't issue bonds, but the Community bank could do it on their behalf, since they are majority shareholders. Civil servants also benefit from this bank, but councilors don't have access to it. 7.2.2 The Kinondoni Municipality The Municipal director of Kinondoni Municipality works in the District Council Office. The council has standing committees, implements laws and collects revenues. The problem is the property valuation. After some years they now have valuation reports for 7 of the 27 wards. Once valued the owners pays 0.3 percent every year for the building. For the land the central government collects a property tax through the Ministry of lands. Problems of local authorities remain that their staff is under paid, that their tax revenue base has been undermined by abolishing certain taxes (Franzsen, 2004 and Franzsen et al., 2002) and that they don't have access to capital markets yet. There is a reform program since 2000. If the Kinondoni Municipality can meet certain criteria they can benefit from capacity development grants. They did not succeed the first year, but the Municipal director is optimistic for the next round. 60 The Kinondoni municipality can borrow from financial institutions, but it needs the approval of the Ministry of Local Government. As a legal entity they can sue and be sued. They never tried to borrow, because they received loans for road construction. Similarly they have never tried to issue bonds. He expects this to be the case in 10 to 15 years. Kinondoni has set aside land for investment plus it influences the licensing processes. They have created tax incentives and abolished some privatizing of services. All this is participatory and involving the local people. Concerning the informal sector they could organize them for examples in groups. Usually these small entrepreneurs are participating in associations and discuss how they can be helped. Local government can for example provide a site for their businesses. As an example of creating the conditions for development, the Kinondoni Municipality in collaboration with the Swedish International Development Agency (SIDA) is implementing the KICAMP project with a view of developing resources to improve the lives of communities and people residing along the coastal belt. Also, the Kinondoni Municipality is currently constructing a fish market at Msasani as a fish receiving station to supplement the famous fishing station at Magogoni, Kivukoni Front, in the Ilala Municipality. Local governments are not involved in housing schemes. However, there are many slums and local government has to provide the infrastructure. They have software that may help them to upgrade their infrastructure. He is not too happy with City Water (which had private management at the beginning of 2004) since they don't always do what they promised. The problem is to inform each other about the activities that are taking place. They do share information with other councils, plus there are central functions, which are undertaken on behalf of higher level of governments. As part of their social policies they can give loans to poor women from their sources of revenue. There are amounts in the budget for women, youth and economic growth. Tax collection is going up and this means they have money for investment projects. In fact between 30 to 37 percent of their revenues is used for that purpose. Finally they get support from development cooperation, but usually via the national government. For example they draw from national education and health funds. Currently one third of their revenues comes from their own resources and two third from national government and donors. It is also spent on roads and drinking water. In the later case they drill holes with their own equipment. Their vision and mission is improving the level of standards of living of their population. Improved quality of services delivered, provide value for money. They don't provide incentives to industries, but they might provide land. This local government area also houses most universities and institutes for learning. At the national level there are 17 national parties, but in this local government council only the ruling party is represented. 7.2.3 The Temeke Municipality 61 In Temeke the opposition is represented in the council. In total there are 44 councilors, of which 27 represent different wards, 10 are councilors representing special groups, such as women and finally 7 people are member because they are also member of parliament and live in this area. This local government has benefited from an Integrated Coastal Management Project. This project focused on rational use of marine and land based resources, on raising the income level of the fisher population and on conserving marine and land based resources. With SIDA support the first phase is completed and now they are in the second phase (2004-07). With a population of 1.088.867 (Census August 2002) they are in terms of population the biggest local government in the country. The next one counts 700,000 people. The local authorities are constantly involved in constructing schools, dispensaries and hospitals, drilling wells and trying to maintain the roads. They also have a website. There is also an Oysterbay beach redevelopment plan, which has been structured as a PPP. The local government provides the land and the private party has to build and operate it. They also cooperate with other local governments and certainly with the former City Council. The current structure of three local governments works much better. In particular competition, competitive tenders and competing in service delivery have improved things. Solid waste collection is a good example of the positive impact of decentralization and private sector involvement. Although the system can still be improved, the private enterprises, individuals, NGO and CBOs involved are doing a good job. The biggest problem is that so far they did not have a sanitary landfill side. Now 254 acres are available for this purpose. A second problem is that the contractors charge the fee directly to the customers. The local government has no control of contractors and doesn't know what to do with complaints of the population. They also participate in the UN Habitat supported Sustainable cities program. However, besides environmental and social policies, they are not yet involved in economic policies, nor are they trying to gain access the capital market. That would be something for the next 5 years. Working groups are working out different environmental issues at the city level. They benefit from all kinds of programs, such as an African Development Bank (AfDB) health loan, the Road fund board, intensive agricultural development and eco- tourism. The ocean allows fishing and in the local government area all foreign embassies are located. However, they are not really formulating economic policies, rather their focus is on service delivery. 7.2.4 The Iala Municipality The Director of the Municipal Council in Iala when asked whether this local government has a policy with respect to the urban informal sector replied that they make available sheds, which can be rented, but usually small-scale traders occupy these. The local government imposes a market tax, but that is not a big deal. This local government borrowed for the first time 250,000 US$ for buying equipment for resurfacing the road. They discussed issuing bonds with the stock exchange, but it seems to be something for 62 the future. They are active in the environmental field in particular in doing environmental impact analysis and in trying to mitigate negative environmental effects. The council has 32 councilors, of which 21 represent wards. Then 8 are women and 3 are members of parliament. The mayor is elected among them. Do they have a social policy? No, he considers this a national level policy, but urban agriculture is important for them and is usually beneficial for the local population. Iala received 347 million TSh. from the LGSP. It is estimated that livestock farming within the city contribute about 34 percent of the city's requirement (e.g. milk, meat). The remaining percent come mainly from well known regions for livestock farming like Arusha, Dodoma, Shinyanga, and Mwanza. Types of livestock farming in the city include: cows, goats, sheep, chicken, duck, and pigs. Market for the livestock products are within the Municipality. The Municipality and private sector play a greater role in the supply of pesticides and livestock extension. However, lack of proper land use plan has led to poor livestock farming within the city. It is possible to see goats and cows in the streets grazing themselves freely and animal keeping in the residential areas in some areas in the city. This suggests laxity of law enforcers, and lagged behind implementation of land use master plan. Mining activities are carried out under the provision of Mining Act of 1998, which prohibits reconnaissance, prospecting of mining without mineral rights and without a written consent from the relevant authority. The prominent mining activities in the city include: sand, gravel stone or boulders, aggregates, limestone and salt extraction. Participation of city residents in this sector varies significantly depending on type of activity. However, the majority does salt extraction, limestone and coral rocks, which are sold to different consumers in large and small quantities. Such activities offer much of the needed employment opportunities and generate income. The sector reduces unemployment for youth, women and citizens in general. Generally it encourages self- employment. 7.3 Sources of urban growth: issues in the private sector There are three main sources of urban growth for cities in Tanzania, namely: rural to urban migration, natural population increase, and extension of township boundaries. Up to the 1960s the increase in the population of towns was less due to the extension of township boundaries or influx of migrants, but to natural increase. This was the case for Dar es Salaam and Tanga between 1957 and 1967. However, there were substantial boundary changes and expansion in urban areas of Arusha (1959 and 1965), and Mbeya (1957 and 1966). Thus, boundary changes added a substantial population to the total population of these urban areas. Dar es Salaam is now one of the fastest growing cities in Sub Saharan Africa. The city population is estimated to be approximately 3 million in the year 2005, with a growth rate of 4.3% percent. The relatively high population growth rate is due to increased birth rates, immigration rates, and more significantly by transient population.xxvi Of the three Municipalities, Kinondoni had the highest population with a total of 1,083,913 people (43 63 percent) followed by Temeke with 768,451 people (31 percent) and Ilala with 634,924 people (26 percent). Table 13 Dar Es Salaam population by municipality and sex ( 2002) Municipality Male Female Total % of the Growth rate! total 1988-2002 Kinondoni 547,081 536,832 1,083,913 43 4.1 Temeke 387,364 381,087 768,451 31 4.6 Ilala 320,408 314,516 634,924 26 4.6 Total 1,254,853 1,232,435 2,487,288 100 4.3 Source: NBS (2004): Population and housing census 2002. The majority of the city population is aged between fifteen and sixty-four years, which is the working age. However there are few elderly people (only 2% are above 65 years). This implies that life expectancy is relatively low in the city (figure 3). The next chart (figure 4) shows declining levels of population from zero to ten years, which is an indication that there are a good number of babies born but a significant number of them die before the age of ten. The peak age (the age of the majority of the people in the City) is twenty to twenty-nine years, which probably includes students and young migrants born outside Dar es Salaam. Figure 3 City Population Distribution by Age (2002) 2% 33% 65% 0 - 14 15 - 64 65+ Source: NBS (2004): Population and Housing Census in 2002 Each municipality is divided into divisions, which are in turn divided into wards. Wards are divided into villages in the case of rural areas and "Mitaa" in the case of urban areas. Moreover, villages are divided into hamlets, which are the smallest units. Figure 4 Population by age groups 64 400000 350000 300000 n 250000 200000 pulatio Po150000 100000 50000 0 14 5*-9 + 0-4 10* 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75-79 80 Age groups Source: NBS (2004): Population and Housing census 2002. Economic growth is influenced by economic policies or the lack of such policies. The Confederation of Tanzanian Industries (CTI) was asked what they think of the current economic policies in the capital. The CTI is often accused of being a Dares Salaam club. Indeed 70 percent of the members are here, but they also have a presence in Arusha, Mwanza and Tanga. The ten major issues for their members are: Table 14 Dar es Salaam: municipalities, divisions, wards, streets, villages & hamlets Municipality Division Wards Streets Villages Hamlets Ilala 3 22 65 9 37 Temeke 3 24 97 15 62 Kinondoni 5 27 114 14 14 Total 11 73 276 38 113 Source: City Director (2004): Dar es Salaam City Council Profile 2004. Table 15 Growth rates of the urban population in selected countries, 1970-1980 Country Annual Urban Country Annual Urban Growth Rate Growth Rate Tanzania mainland 10.2 DRC-Congo 7.2 Malawi 6.8 Uganda 7.0 Rwanda 5.9 Kenya 6.8 Source: URT (2004b). 1. Economic policies of the government: CTI wants for example the government to abolish cross-subsidies for tariffs of utilities. Already the tariff for electricity went down for industries. In the case of petroleum there was a lot of tax evasion. Now they have sophisticated flow meters at the port and introduced marking. 2. Taxation, which are now streamlined a bit. CTI has two seats on the tax reform committee. Their views are considered. They discuss what the benefits are, what shortfalls in revenue can be expected from economic integration and how to meet 65 those. In the eyes of their members the local governments still think they can impose taxes any moment and anywhere. They want to discuss taxes and by-laws (on parking, on waste collection, pollution, etc.) with them. They also wish to have their employees examined in local government health centers. However, decentralization has led to a change in mindset. Nowadays they will get a response when they ask something from a central government institution. 3. Changes in the labor laws were necessary, but have been achieved. Some labor laws dated from 1964. 4. They complain about the electricity supply. There are strong variations that damage certain machines. 5. Security is an issue, although the situation is not as bad as in Kenya. 6. Competition for their members comes from Kenya and South Africa. However, CTI consider the presence of South Africa in certain sectors (for example in mining and telecom) as very beneficial. For example, prices of mobile operators have come down substantially. 7. The impact of private sector involvement in drinking water supply is not yet felt at the level of their members. 8. Land is also a problem. Particularly in the Central district of the city it is scarce. Foreign investors can get it through the Landbank. 9. CTI considers that the port of Dar es Salaam can play a more important role in the regional economy. This requires an improved infrastructure. The government started already with privatizing some of the facilities in the port. However, the process of privatizing the railroads is very slow. They want to give concessions to two parties, for the tracks and the carriages separately. Also the roads are bad. 10. The urban informal sector is currently not really competing with the manufacturing sector, given their limited capacity and concentration in certain areas. In the seventies and eighties this sector was taboo. They could eventually compete in activities like leather, furniture and foodstuff. There are also no examples of subcontracting to the informal sector in Tanzania so far, although the ideas exist. They have visited Japan and have seen it work. The constitution of the CTI requires you to be a company, which would exclude informal sector enterprises. The Tanzania Chamber of Commerce, Industry and Agriculture (TCCIA) convincingly argued that although the private sector gets a much better treatment, there is still room for improvement. One problem is that an important part of their constituency is in Dar es Salaam. The Chamber wants the tax system to be simplified and would like one shop for licensing. They are consulted on the WTO negotiations and point to the logistical problems of developing enterprises outside Dar es Salaam. According to them there can be no growth without direction. If the master plan for Dar es Salaam mentions the role of the city as a hub, then the issue is still how to enhance its position. How much of the port can be used for the wider region? How can that potential be developed? City marketing is still a new concept in this country. The executive director of the Tanzania Private Sector Foundation (TPSF), the Apex organization of the private sector in Tanzania also gave his opinion about desirable 66 economic policies. The particular nature of this foundation is that it coordinates trade bodies and has the government as a member. They consider their main role to be complementary to what the other associations are doing. Cross cutting issues are corruption, taxation and reforms to create a smooth environment for the business sector. TPSF is relatively young. Issues it deals with are: 1. Protection: going from an inward to an outward orientation. In that context the TPSF argued that investors from other African countries should get the same treatment as those form Europe. 2. Economic empowerment of Tanzanians: how can these businessmen get cheap loans? 3. Develop policies on Small and medium enterprises (SMEs) promotion and in general develop better trade and industrial policies. 4. Changes in the investment act 5. The role of Tanzania in the East African Community (EAC, Uganda, Tanzania and Kenya) and the Southern African Development Community (SADC; seven countries: Angola, Mozambique, Tanzania, Botswana, Lesotho, Namibia, Swaziland; duties will be abolished by 2008): how best to do business with our neighbors? 6. Power and water need to be available to allow growth and roads need to be improved. Mwanza by train takes three days and to drive there you can no go faster than 50 km per hour. The road to Morogoro is good. In 2006 all these roads should be tarred. 7. Economic Partnership Arrangements (EPA) are required by the European Union (EU), but complicated for Tanzania, since they are member of two blocs. They have chosen to do it the negotiations in the framework of SADC and to try to coordinate with the EAC negotiations. The issue is not just tariffs, but also the stringent non-tariff barriers, such as phyto-sanitary requirements and the so-called tracing requirements, which make it necessary to always be able to track a product to its origin and which makes exporting more difficult in developing countries. It is time to do away with all subsidies in agriculture. However, Tanzania expects to really benefit from a further liberalization of international agricultural exports, as is discussed in the current WTO Millennium round. 8. The cost of production are higher in Tanzania because of the price of electricity, which is 12 US$ cents per Kw hour, which is more expensive than in Southeast Asia. 9. The government micro finance bank is not doing as well as local decentralized NGO micro finance schemes. A remarkable statement is that local government is much closer to the private sector than the national government. They listen to the problems of the private sector and give a response. They can provide space and infrastructure, but currently no other incentives. They may be able to deregulate and certainly try to create a more positive environment for the private sector and engage in a dialogue with the Chamber. 67 7.4 The economy of Dar es Salaam In Dar es Salaam City the degree in which the private sector is involved in a certain activities varies. Many people are involved in wholesale and retail trade, followed by hotels and restaurants and the third being manufacturing activities. Construction, transport, storage communication and construction activities are under private sector. However, the share of these activities in the city's GDP does not follow this pattern. Manufacturing sector has the highest share, followed by construction sector. Contribution of Dar es Salaam City's GDP to the National GDP has been fluctuating between 18 percent and 16 percent from 1992 to 2001 (table 16). Compared to other regions Dar Es Salaam has the largest share of the National GDP. Currently, under Private Sector Development Initiatives (PSDI), a lot of these products has to come from SMEs, which are expected to be established, or existing ones need to be revived by using appropriate technology and the skills available in the region. The average per capita GDP of Dar es Salaam had increased from TSh. 267,976 in 1995 to T. Shs. 554,287 in 2001, equivalent to an increase of 107 percent. In 2002 a survey indicated the per capita GDP for Dar es Salaam to be TSh. 584,086 with 35 percent of the population earning an average income of TSh. 387,319 per annum. However, about 875,000 (35 percent) people in Dar es Salaam earn less than one US dollar per day. By comparing Dar es Salaam with other regions Dar es Salaam has the highest per capita GDP followed by Shinyanga and Arusha regions. Table 16 Regional GDP at current prices: Arusha, Kilimanajaro and Dar es Salaam Region/year 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Arusha 94782 118728 155655 214066 265143 328305 383893 446730 510055 582554 663643 Kilimanjaro 54628 68799 91197 119476 147294 182789 217244 252803 274676 310701 348893 Dar es Salaam 232460 291260 402690 498624 605241 758249 916280 1066259 1158513 1282449 1459013 Mainland in 1275916 1607762 2125325 2796640 3452560 4281600 5125311 5977699 6706381 7624616 8618071 total Source: NBS(2003a). However, the cost of living in Dar es Salaam has also been rising day after day despite government interventions, to reduce poverty. The Household Survey 2000/2001 showed that 7.5 percent of Dar es Salaam population was unable to get adequate food (food poverty) compared to 13.6 percent in 1991/92. In the same period and 17.6 percent were unable to get basic needs (basic need poverty) compared to 28.1 percent in 1991/92. This shows remarkable effects of the efforts to reduce poverty, although the levels are still high. In comparison with other regions in Tanzania mainland, Dar es Salaam city has the lowest percent of people that are unable to get basic needs and food . Dar es Salaam and its Municipalities want to have in place sustainable social and economic development through participatory resource mobilization and utilization thus enhancing the quality of social and economic services by using the existing resources and opportunities. The major economic activities in Dar es Salaam include: manufacturing, 68 urban agriculture, tourism, transport and communication, forestry and fishing, mining and quarry, utility services, construction, finance and insurance (City Director, 2004). Around Dar es Salaam there is considerable "urban agriculture', perishables in particular for the urban market (Foeken, et al., 2004). However, it is estimated that urban farming activities in the Dar es Salaam provides the population with 354,657 tons of food (134,060 tons in Kinondoni municipality, 55,597 tons in Ilala municipality and 164,100 tons in Temeke municipality). Cereals and meat tend to come from further away. Some regions have specific climates, which make them very appropriate for specific products. For example the Tanga region in the north is known for its temperate climate that allows it to grow all kind of fruits and vegetables, if appropriately stimulated to do so. Industrialization on a small and large scale is a common feature in Dar es Salaam. Small- scale enterprises concentrates on manufacturing goods for the domestic market and are almost all developed in areas close to residential areas. Medium and large-scale industries are located in the designated industrial areas, the sector includes textiles, chemical, food processing, light manufacturing, aluminum and glass, plastic products, rubber products and cement production. Products of these industries are for the domestic market and for export to neighboring countries like Kenya, Uganda, Zambia, Malawi and Democratic Republic of Congo. Sustainability and expansion of food and cash crop production in the region appears difficult to be achieved. Some of the existing problems and obstacles facing this sector are: poor farming methods and equipment, non existence of data about farmers and crop yield, great increase of people especially youths, high rate of urban expansion leading to decrease of agricultural land, land degradation due to soil erosion, inadequate knowledge of farmers about improved farming techniques, outbreak of crop pests and diseases affecting plants and crops, and shortage of inputs including better quality seeds and fertilizers Being located along the Indian Ocean, fishing is one of the major economic activities in the city. The fish catch are mainly for domestic use since majority of fishing mongers are using traditional fishing equipment like canoes and nets. Lack of modern fishing equipments and processing factories have hindered the expansion of this sector in terms of employment and income generation to the City's residents. Poor fishing practices along the seashore in the city have resulted into environmental pollution and soil erosion. In order to arrest this situation some concrete and sustainable measures are implemented by the respective municipality authorities. Dar es Salaam, through its Municipalities of Ilala, Temeke and Kinondoni has parks and gardens as centers for agro extension services and selling of vegetables and trees seedlings. The city also has good and attractive beaches with a diversity of cultural and archaeological sites with potential for investment attraction. It has a coastal belt of about 100 kilometers occupied by many tourist attractions and hotels. However, Dar es Salaam is yet to promote this sector fully so as to make Ilala, Kinondoni and Temeke coastline one of the favored tourist destinations. 69 In the case of new investment out of 400 registered private projects in the country 231 were registered in Dar es Salaam (URT, 2003c). How does the CTI look at the FDI, which has been around the 250 million US$ during the last three years (URT, 2005)? CTI is happy with it but notes that 90 percent goes into mining. Very few goes into manufacturing, although the manufacturing sector had the highest number of projects (95) followed by construction sector (41) and tourism (37). Furthermore, in order to attract more investors, number of measures has been taken by both central government and the municipalities. An attractive package of investment incentives is offered through the Tanzania Investment Center (TIC), which has to sensitize local and foreign investors and participates in efforts to enhance the business and investment environment in Tanzania. The main incentives include: import duty drawback on raw materials; zero rated VAT on goods manufactured for export; straight line accelerated depreciation allowance on capital goods; right to repatriate 100% of the foreign exchange earned together with profits and capital; fast track to obtain other permits such as residence/work permits, industrial and other trading licenses; permit to employ up-to five foreign expatriates upon fulfilling the requirements; improvement in business licensing and procedures and availability of basic infrastructure including established Export Processing Zones (EPZ). Tanzania is trying to develop export-processing zones (EPZ) but not all are successful. There are three in Dar es Salaam, one was closed down, but the others seem all right. They are not paying taxes and not supposed to dump their products on the local market. There are many bottlenecks for increased private sector development in Dar es Salaam and Tanzania in general, some of which can be categorized under three different headings: setting up of business, financing production and cost of production. They will be discussed in that order. In the past problems relating to initiating a business were more severe to both foreign and local investors. The major obstacles used to be red tape and lack of coordinated facilities. For example, Mjema and Danielsson (1999) estimated that the time used from filling the first application to reception of necessary documents may run into a month or more. However, the establishment of Tanzania Investment Center (TIC) and Business Environment Strengthening for Tanzania (BEST, under the President's Office Planning and Privatization) has reduced drastically the complexity of investment procedures in Tanzania. BEST is sometimes called the `better regulation unit' (Sunday Observer 17-7- 2005), but is applauded for its efforts to minimize the unnecessary red tape in license provision to pretty traders. Their specific focus is on modifying laws, regulations and rules related to the provision of licenses, or more in general those governing business activities in Tanzania. While Tanzania registered as low wage economy due low productivity in agriculture, cost of production are nevertheless relatively high compared to other countries in the region. The high taxes and /or non-transparent tax, followed by inadequate supply of infrastructure (including financial), financing, corruption and inflation has been 70 registered as among the obstacles that contribute to high production cost in Tanzania (Word Bank, 2002). Moreover, price of utilities such as water, telecommunication and electricity is very high compared to other countries in Sub-Sahara. Furthermore Mjema and Danielsson (1999) observed that in Dar es Salaam, services are often of poor quality, with frequent power breaks, shortage of water and sometimes high traffic congestion. Despite high concentration of financial institutions in Dar es Salaam various studies reveal that there are some banking features that seems to be among the obstacles in financing production in private sector. Although many banks quote lending rates for medium and long-term loans, in practice no such loans exist. Lending rates are limited to short term credit, appropriate for financing a running business or as trade credits. According to World Bank (2002) there is high discrepancy between saving and lending rates, lending rates are 10 percent points above saving rates. A similar discrepancy exists between the discount rate and lending rate. This has been due to the fact that it is difficult and costly for banks to assess risks of projects and reliability of prospective borrowers (Mjema and Danielsson, 1999). 7.5 Sources of revenue of Dar es Salaam The City Council and the Municipalities have two major sources of revenue. One is their own sources of revenue, which include: licenses, fees, permits and services and other levies. Second is central government subvention specifically for development projects and person emoluments for teachers and health staffs to fill the council's budget deficit. Table 17 shows actual revenue collected by the city and central government subvention (grants) to the same between 2001 and 2003. During 2001, 2002 and 2003 the City received a total grants of Tshs.312,035, Tshs.1,828,977 and Tsh.1,861,522 equivalent to 9.2, 37.8 and 46.4 percent of the total revenue respectively. This trend suggests declined City's capacity to deliver services by using its own sources of revenue. In the case of the municipalities the existing sources of revenue and the amount of revenue collected from particular source differ very much. Looking at their own sources of revenue, in 2003 Ilala municipality had the highest revenue collection (48.9%) compared to Kinondoni (29.3%) and Temeke (21.7%) municipalities (table 18). This has been mainly attributed to the existence of various industries, port, and shopping centers. Moreover, Kinondoni municipality had the highest share of total grants (46.0%) in 2003 compared to Ilala (29.0%) and Temeke (25.0%). Most grants directed to the municipalities are meant for construction of economic infrastructure. Kinondoni has the most developed infrastructure compared to Ilala and Temeke. Table 17 Sources of revenue of the city Source 2001 2002 2003 Actual Actual Actual 1.Own Sources Development levy 1,777,859 1,786,081 818,915 Property tax 0 0 0 Produce cesses, levies and royalties 0 0 0 Services and other levies 137,455 89,437 77,896029,973 71 Land rent 0 0 0 Licenses, fees and permits 999,659 993,593 832,525 Charges 94,114 101,821 306,516 Others 54,472 40,792 110,881 Sub total local revenues 3,063,557 3,011,723 2,146,733 2. Grants Education block grant 0 0 0 Healthy block grant 0 0 0 Roads block Grant 0 0 0 Water block grant 0 0 0 Agriculture extension grant 0 0 0 Local administration grant 0 0 0 Other government grants 312,035 1,828,997 1,083,082 Basket funds and non Got Grants 0 0 0 Compensation grant for abolished taxes 0 0 0 Subtotal Grants 312,035 1,828,977 1,861,522 Total Grants and Revenues 3,375,592 4,840,720 4,008,255 Source: URP President's Office Regional Administration and Local Government Constrained revenue collection by the municipalities and unpredictable grants support has caused poor implementation of development programs aimed at improving the life of low income and the city's population at large. This suggests the need of expanding tax base of these municipalities, and equally strengthening of local government and private sector partnerships in provisional of services. Table 18 Comparison of own and grants collected and received by municipalities (All figures are in TSh. 000s in 2003) Municipality Own source Grants % of total own % of total sources grants Ilala 6,514,154 6,495,690 48.9 29.0 Kinondoni 3,903,395 10,288,321 29.3 46.0 Temeke 2,891,611 5,592,781 21.7 25.0 Total 13,309,160 22,376,792 100.0 100.0 Source: Revenue statistics for Local Government authorities on Mainland Tanzania, 2004 7.6 Conclusions on Dar es Salaam Dar es Salaam is very different from the rest of the country. Rapid development in this city is largely the result of better macroeconomic policies and a restored confidence in the Tanzanian economy. Such policies affect development directly and indirectly. Better governance helps a lot indirectly. Changes in pricing and taxation may help the agricultural sector directly. Unfortunately decentralization also led to too much taxation in the beginning and it is Central government that has pushed the local governments to put order in their house. They abolished certain taxes and forced the local governments to rationalize tax collection by focusing on a limited number of taxes and complementing their budgets with income transfers. Table 19 summarizes some of the positive and negative factors of Dar es Salaam versus Nairobi. 72 Table 19 Dar es Salaam versus Nairobi Dar es Salaam positive trends Nairobi negative factors Politically stable Decentralization stimulates better urban Political crisis at local government level governance No real decentralization and poor urban Not as strictly regulated as before governance Opportunities for informal sector No policy with respect to informal sector Potential of the harbor Keeping of cattle in the city is not allowed Possibilities for urban agriculture Serious traffic congestion and urban A lot of donor money is available pollution Space for expansion Source: this study and Karanja (2005) for Nairobi. Dar es Salam shows an enormous dynamics. In particular the three municipalities have become very active. Although they don't really formulate their own economic policies, they started recognizing the importance of having economic activities and creating the conditions for such activities, including the informal sector. In this respect the three municipalities can serve as examples for local governments in the interior of the country. All Local Government Authorities (LGAs), which includes the City, Municipality, Town and District Councils, have a similar organizational structure irrespective of differences in population, resource-base and infrastructure. This has been one of the weaknesses of LGAs. However, the LGA can also create local incentives. It would be interesting to have an infrastructure master plan for Dar es Salaam, since there doesn't seem to be a vision. The strategic plan, which is currently prepared at the City Council level (City Director, 2004), will help to develop such a perspective. The City Director (2004) notes in the City profile that there is inadequate knowledge about the dynamics of the informal sector, that the city has inadequate strategically located areas for allocating space to informal sector operators, but at the same time he admits that a robust informal sector will increase employment and improve living conditions of those who are economically marginalized. Reading the profile it seems the capital doesn't yet know how to deal with this sector, which is a pity, given its importance. The overview of the problems of Dar es Salaam and the disjoint way in which the different municipalities deal with their problems shows the importance of a strategic planning exercise for cities and regions. Dar es Salaam is involved in a plan, which will clearly bring out the urgency of certain bottlenecks (City Director, 2004). Dar es Salaam is a peripheral node in a global system. It should benefit more from its location and try to serve a larger area because of the benefits it would bring to its population. Given the functionality of Dar es Salaam as the country's main port, administrative center and 73 center of economic activities the city can play a very important role in the further development of the country,. 8. Regional growth in Tanzania: local-global interactions in a low-income country 8.1 Introduction Cities play an important role in the development process. London, New York and Tokyo are global cities according to Sassen (1991). Dar es Salaam, Tanzania's capital, is probably a peripheral node in such a global system, but the country is at least linked to the global system in this way and we intend to study the role of the Tanzanian capital as a means of transmission of growth to a larger hinterland. This area is potentially ranging from other Tanzanian regions to neighboring countries like Burundi, Rwanda and Eastern Congo. In this contribution these transmission mechanisms, which in an African context are largely informal flows people, ideas, money, goods and services. The effects of these flows will be studied in two regions in the interior by comparing their performance on a number of indicators. The study will explore the sub regional foundation of economic take-off and development of Tanzania. The country is one of the poorest African countries, but at the same time it recorded a very high rate of growth during recent years. After years of restructuring its economy, finally something is happening in Tanzania. Decentralization is also taking off and may be another factor why the economy is picking up. We will focus in particular on the dynamics of two regions and the role of their urban centers. Arusha city, the capital of the dynamic Arusha region will be compared with Moshi, the capital of this Kilimanjaro region. These two regions are both located in the north of the country, on the border with Kenya. However, between these regions (and even within) there exist strong differences as will be shown in the comparison below. Comparing two such regions in a general dynamic context may allow bringing out the factors that promote regional growth and the role of regional capitals in that process. We will first introduce the research and the theoretical framework. Subsequently a dynamic and a stagnating region will be introduced and will be compared on a number of variables to draw some conclusions on the local-global interactions in a low-income African country. 8.2 Research questions The Tanzanian capital Dar es Salaam clearly functions as a growth center, but so far a similar local economic system cannot be found in the interior. What are the factors determining why one region does develop and another remains at the same level and even declines in terms of population and relative contribution to the mainland gross domestic product (GDP in NBS, 2003 and table 20). This contribution deals with what explains the dynamics of two regions in Tanzania and what can be done to achieve a more equal distribution of the fruits of development within one country? What is the role of cities like Moshi and Arusha in the regional productive system? 74 Table 20 Regional GDP at current T. Sh.: Arusha, Kilimanajaro and Dar es Salaam Region/year 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Arusha 118728 155655 214066 265143 328305 383893 446730 510055 582554 663643 Kilimanjaro 68799 91197 119476 147294 182789 217244 252803 274676 310701 348893 Dar es Salaam 291260 402690 498624 605241 758249 916280 1066259 1158513 1282449 1459013 Mainland in 1607762 2125325 2796640 3452560 4281600 5125311 5977699 6706381 7624616 8618071 total Source: NBS(2003). Note one US$ = about 1.108 Tanzanian Shilling. In more general terms the question is how local economic development strategies and economic development and trade strategies affect the potential of two regions in Tanzania and how these regions react to the opportunities provided to them.xxvii Data were collected for the Kilimanjaro and the Arusha region, which both have a potential for tourism and play an important role in regional trade. In both cases the role of a central city (Moshi and Arusha) will be studied as well. 8.3 Theoretical framework for regional development studies Recent literature on regional development stress co-operation, networks and institutions in regional innovation systems (Fornahl and Brenner, eds, 2003). These factors may not yet be so important in rural Tanzania. However, as noted by Scott and Garofoli (2004), regional economic systems with their virtuous circles of cumulative causation are a critical part of the development process. To fully understand the process the influence of critical variables as externalities, increasing returns to scale and competitive advantages accruing to two regional production systems in Tanzania need to be assessed. To check whether Tanzania is a tiger economy, we could have used the framework coming out of the miracle study (World Bank, 1993 and summarized in box 12), used for a recent comparison of India and China (Van Dijk, 2006c). Box 12 Factors explaining the success of the High Potential Asian Economies a. Sound fiscal policy b. Low inflation c. Convertibility of the currency and a stable rate of exchange d. High level of domestic savings e. Heavy investments in education and a higher level of human physical capital f. Keep the economy open to foreign technology g. No government interventions in the economy h. A higher level of productivity growth i. An earlier and steeper decline in fertility However, these are mainly macro economic variables, important at the national level. Comparing regions, we have chosen a number of other indicators, summarized in Box 13, taking as a variable to be explained the difference in the development of regional income 75 and as potential explanatory variables the available resources, population pressure, social (underlying) variables like education and health and indicators of what we call the informal flows of people, goods and capital. Box 13 Suggested variables for the comparison between the two regions 1. Available land 2. Population density 3. Infrastructure (roads and electricity) 4. Health and education 5. Migration 6. The role of the regional capital: Moshi and Arusha city 7. The role of different economic sectors 8. Growth redistributing mechanisms 9. Role of different stakeholders 10. Local development policies 11. The organizing capacity of regions and cities An additional objective of the larger study undertaken for the World Bank is an assessment of the role of urban areas, and in particular of Dar es Salaam, Moshi and Arusha city in the Tanzanian economy.xxviii The data on poverty show that the national capital is the only place where some poverty reduction has taken place since adjustment policies started in the nineties (NBS, 2002a). The comparison between the two regions will be made on a number of factors, which are often considered important in the literature to explain the big difference in per capita income. In particular the following mechanisms expected to contribute to redistribution of growth in Tanzania were studied in more detail: a) Decentralization creating opportunities at the local level b) Redistribution through development aid and infrastructural investments and c) Through the often (informal) flows of people, ideas, capital, goods and services Finally, the concept of organizing capacity of regions and cities will be used. According to this theory the success of urban development projects depends on the organizing capacity of a city or a metropolitan area. Organizing capacity means to be able to anticipate, respond to and cope with changing intra and inter metropolitan/city relations due to internal and external processes of change at the proper spatial scale level. It can be defined as the ability to enlist all actors involved, and with their help generate new ideas and develop and implement a policy designed to respond to fundamental developments and create conditions for sustainable developments and creates conditions for sustainable economic growth (Van den Berg, Braun and Van der Meer, 1997). The elements of organizing capacity are: 76 - Vision on long-term sustainable development (for strategies, programs and projects to hold on to) - Formulation of concrete, measurable objectives - Strategic and coherent thinking and acting - Leadership qualities to manage processes and projects adequately - Creating and supporting strategic networks of relevant partners, needed to develop and implement policies successfully - Creating political and societal support , and - Emphasizing communication strategies both with in the city administration and as external communication (to citizens, companies, public bodies, etc.) Key elements of organizing capacity for community-based projects that are intended to solve city problems are the organizational tools, represented by "T" (the instruments with which to achieve an adequate community based projects) and the project process, represented by "P". Moreover, the contexts, represented by "C", under which cities are operating, are also equally important elements in organizing community-based projects. Context variables (C) Spatial-economic conditions (C1) Administrative organization (C2) The social context (C3) Tools of organizing capacity (T) Vision and strategy (T1) Strategic networks (T2) Leadership (T3) Political support (T4) Societal support (T5) Communication strategy (T6) Project process (P) Project implementation monitoring and output (P1) Evaluation and outcome (P2) The above theoretical framework could be summarized into a table. The assumption is that the success of community-based projects (S) depends on the three important categories of variables (C, T and P). The three main variables (and their components) could be dependent (inversely or directly) or independent to each other. 8.4 A dynamic and a stagnating region Tanzania has 21 semi-autonomous regions and 53 districts. Important products are coffee, tea, sisal, all kinds of nuts (in particular cashew nuts, groundnuts and macadamias) and cotton. Tourism is a strongly growing sector and can be linked much more with local activities. Foreign Direct Investment (FDI) is increasing, however it is limited to mining and tourism and some manufacturing, in particular textiles. Most of 77 these activities are concentrated in and around Dar es Salaam where there is also a considerable "urban agriculture', perishables in particular for the urban market (Foeken, et al., 2004). The Kilimanjaro region is known for the Kilimanjaro Mountain, the highest mountain in Africa and a huge volcano at the foot of which lies Moshi town, the regional capital. There is an international airport for the whole northern region, called Kilimanjaro and built exactly in the middle of the two capitals Arusha and Moshi. The Kilimanjaro mountain is the best known tourist attraction in Tanzania and the border with Kenya was closed for some time, when the Kenyans advertised that the mountain could very well be visited from Nairobi, which is indeed only a four hours drive by car. The Arusha region is facing an expanding market for its goods and services. Europeans go there because they enjoy the climate, in particular for tourism and horticulture. Arusha city is growing fast, due to a large number of international institutions and good connections with Kenya and Uganda. Nairobi is close, while Dar es Salam takes 8 to 10 hours by road. There is still a lot of corruption at the local government level and in the border towns. Purchasing power is much higher in Arusha, also because of Tanzenite (a special blue jewel) and many gemstones are traded in the city. Even agriculture is better in the Arusha region, given the number of new horticulturalists and the green vegetable industry, which is growing fast. Both regions benefit from the location near to Kenya and Uganda. NBS (2002c) gives a socio-economic profile of the Kilimanjaro region. The format is very similar to NBS (1998) and URT (2004), which give a similar profile for the Arusha region. Key data are summarized in the following table. Table 21 Kilimanjaro and Arusha regions compared on key variables Variable Kilimanjaro Arusha Per capita income (T.Sh. 152,004 277,367 2001) Size 13,209 km2 34,515 km2 Population 2002 1,381,149 1,292,973 Population growth 1988-02 4.0% 1.6% Population density 105/km2 37/km2 Number of districts 6 5 Tarmac roads 383 km 337 km Electricity 45% of villages 1.6 & 21.9% of rural and urban households Population per hospital bed 680 1072 Infant mortality(1988) 67/1000 75/1000 Under five mortality (1988) 104/1000 129/1000 Number of secondary schools 112 (1997) 63 (2002) No. of vocational training 66 (1997) 68 (2002) 78 Migration -124,383 +141,724 Source: NBS (1998 and 2002c) and URT (2004). 8.5 Kilimanjaro region Soon after German explorers viewed the ice-capped mountain of over 4000 meters in 1848 Christian missionaries followed to live among the local population. The Chagga, are the indigenous people of the Kilimanjaro region. These developments gave them an early start with development, ahead other areas of the country. The region has plenty of rainfall and fertile volcanic soils. These soils had attracted the Chagga and the mountain slopes from the low-lying plains are heavily settled. NBS (2002c: viii) speaks of "one mass of sub-urban settlements stretching from Rambo in the east to Hai in the west". The Kilimanjaro region suffered from the collapse of the price of coffee and subsequently of the coffee culture and the collapse of many industries, in particular the closing of an important beer brewery. A vicious circle of fewer industries and diminishing purchasing power started. A lot of people of the Kilimanjaro region now do business in Arusha city and travel back to Moshi at night or in the weekend. In socio-economic terms the region is one of the poorest regions of Tanzania, which needs to be explained. To acquire economic activities one needs to have the right people and the infrastructure in place. In the 1950s and 1960s because of the coffee boom, Kilimanjaro was far ahead of Arusha. Plus the region has a strong structure of agriculatural cooperatives, which was abolished in 1976. When also the local government structure was abolished, the region started to decline. In 1984 everything was reestablished, but they could not go back to the good old days. The local government support project (LGSP) now starts to rebuild some of that infrastructure. However, by now every local government council in the district is struggling to be as attractive as possible. The Kilimanjaro region is relatively left behind. It doesn't happen there and with the decline in coffee prices the bottom is out of the local economy. In the small Kilimanjaro region there is a high population density and there is land scarcity. Plus everything is very expensive. The average land available per household is only 0.6 hectares. The money earned in the plantations is not staying in the region. Also the region doesn't catch up on tourism. There are no tourist facilities at a reasonable rate. In Arusha you can stay for less than 10 dollar a night in a clean hotel. That is not possible in Moshi. Hence the tourists go to Arusha. Anyway most of them don't come for Mt. Kilimanjaro but for the wild parks, which are in the Arusha region or the former, larger Arusha region, with Arusha city as its center. Kilimanjaro should have provided cheaper accommodation a long time ago. In the Kilimanjaro region cattle raising is quite traditional. There are some Frisian cows, but not enough. They could have gone for zero grazing, but it has not happened. They don't have farms of 200 to 300 animals, which would be economically more viable. The milk is currently sold to each other and not to the market. 79 Also in the Kilimanjaro region the government tried to formalize activities and killed a number of initiatives. For example the successful coffee cooperatives were formalized to create jobs for government officials, but they are no more successful now. Another example in the field of education can be given. The Kilimanjaro region has many schools, but they are too expensive, because they have all facilities in one place (boarding school type), which make them expensive and mean that private people can not make money from setting up boarding houses. Although the population is very entrepreneurial in the Kilimanjaro region, in particular the Chagga ethnical group, they do their business often outside the region and just send some money back to the people left behind. You can see it because the villages in Kilimanjaro tend to have very niece houses. Investors don't want to invest in the Kilimanjaro region, they invest outside the region. 8.6 The Arusha region The Arusha region used to be 82,428 km2, but a separate region was created . This region is growing fast because of the pleasant weather, the large number of establishments, and the large number of resources: gems, flowers, better quality coffee and a number of other products (wheat). Tourist attractions are concentrated here and the region has a good eco system, although the environment is becoming a problem, in particular due to air pollution. The wild parks are only at 1.5 and 2 hours from Arusha. Even Serengti is only two hours away. They have what the regional commissioner called the eighth world wonder, the Ngorongoro crater. The LGSP project is doing very well in Arusha, except for two councils. If these are doing less this is often a question of political support, or lack of leadership, of local capacities. 8.7 The Kilimanjaro and Arusha regions compared 8.7.1. Available land What strikes immediately in table 21 is that both regions have almost the same number of inhabitants, but Arusha is much bigger in size. Probably the effective land use is even smaller, given the high mountains in the Kilimanjaro region. 8.7.2. Population density The higher population density in the Kilimanjaro region is the result of the limited availability of land. It is almost three times as high as in the Arusha region. Despite this high population density the population growth in the region is also much higher as if the people are insufficiently conscious of the current constraints on the resource system. 8.7.3 Infrastructure (roads and electricity) Some of the salient features of the infrastructure in the two regions, springing from table 21 are the large stretch of tarmac roads in the Kilimanjaro districts. The length is about 12.5 percent more than in the Arusha region, which is almost 2.7 times bigger in surface. 80 The Kilimanjaro region also has the most extensive rural electrification network in Tanzania with 45 percent of the 449 villages already supplied The members of the Confederation of Tanzanian Industrialists (CTI ) we interviewed experience an ongoing improvement of the infrastructure in the Arusha region: roads, electricity and even water. It means goods can now be transported at much lower costs. Compared to five years ago there are now fewer outages and companies are no more complaining as much about power cutoffs. They have put in place some transformers and the peaks are better spread. The CTI representative goes to the power suppliers himself to discuss the problems. Telephone lines are OK, just like the internet. Recently the Arusha District council received 345 million T.Sh. from the fund for local governments, while the Arusha Municipal council got 154 million T.Sh. For comparative purposes: Iala (a district in Dar es Salaam) received 347 and Moshi 215 million T.Sh.. 8.7.4 Health and education Health and education are still better in the Kilimanjaro region as a leftover of a more fortunate past. However, the Arusha region is catching up and infant mortality is for example decreasing rapidly, while the number of schools is also increasing fast. Interestingly the Arusha region has a slightly higher number of vocational training institutions, which helps to prepare young people for further development in the formal or informal sector. 8.7.5 Migration In the Arusha region slightly more than half of the population (50.7%) were not born in the region. This shows that immigration has had a big impact on the population growth of major urban centers in the country. However, since the 1980s, natural population increase has also become a major component of urban population growth. The big difference between the two regions is however the balance of in and out migration. In the declining Kilimanjaro region it is negative, with an annual outflow of almost ten percent of the population. A complete inverse trend is taking place in the Arusha region where the inflow is more than ten percent in the same period. 8.7.6 The role of the regional capitals: Moshi and Arusha city In Moshi there is not much industry left. There is still a matchbox factory and one can find some hotels. However, the capital Moshi is considered too expensive because of the high prices for land and other facilities. The conclusion is that if no initiative is taken, or the wrong decisions are taken, or the right decisions are taken too late a region can become stagnant or even declining for a long period. The Arusha region is facing a growing market. In terms of urbanization we noted a much higher level of urbanization than in the Kilimanjaro region. The increasing number of government services, international organizations and industries will have contributed to this trend. There is a foam factory and there are timber products units in Arusha city. 81 Even if Arusha city is dynamic it doesn't mean its regional hinterland is also moving in the right direction! There exist big differences between the different districts within the Arusha region. Some districts are nomadic with very traditional Masai raising cattle. However, Arusha is also the region of the huge coffee farms and some pockets of mining of Tanzanite. Finally, a growing number of 15 horticulturalists are active and have created an Association with some support from the Netherlands Embassy. The enabling environment for them is less in Tanzania than for example in Ethiopia. Problems are organizing charter flights to export their products. This requires cooperation between them. Agglomerations of economic activities hardly seem to be a factor stimulating the competitive advantage of a region like Arusha. When the tourists opted to go to Arusha city the operators also went there. In fact most tourists don't come to climb the Mount Kilimanjaro, but rather to visit the wild parks, which are usually located in the Arusha region or in parts of the former Arusha region (Serengeti). 8.7.7. The role of different economic sectors The role of different economic sectors in the two regions should also be compared. Both regions share tourism, but in practice more tourists go to Arusha than to Kilimanjaro. Two factors mentioned are one that most people don't come for climbing mount Kilimanjaro and secondly, the latter region has not managed to provide cheaper accommodation for international tourists, which is available in Arusha. In the dynamic tourism sector a distinction can be made between activities, which are reinforcing regional development and activities that are not. What makes a region tick? For Arusha the existing business environment and the many opportunities in this region has contributed to the boom. One notes the physical concentration of activities, which creates what economists call agglomeration economies. Investors work in a stable political context and FDI is flowing into the sector. However, CTA members also think that efforts to formalize informal activities will kill them. Formalizing is considered by many people a just another way to create jobs in the government sector. Economic activities in the Arusha region are extremely lucrative: from mining to tourism and from crossing the border with all kinds of goods to serving the international organizations that have settled there. Kilimanjaro is not that attractive and faces land scarcity since years. Tourism expanded already tremendously and the sector became much more service oriented. Arusha also benefited from the East African Community (EAC) building and the United Nations (UN) Rwanda Tribunal, which employs 600 foreigners and 300 local people. In total the UN has 800 expatriates in Arusha! The safari trip organizing companies have come to Arusha. There are very few in Moshi, but a large number in Arusha now. In general the impression is that the Arusha region is booming because of tourism, but also because of agriculture and trade links with Kenya and Uganda. The wheat, coffee 82 and flowers (mostly Dutch horticulturalists and at least one Frenchman), plus mining is mentioned, although the latter is mainly some small scale activities. 8.7.8 Redistributive mechanisms The redistributive mechanisms mentioned contribute to redistribution of growth in Tanzania and were studied in more detail. In the first place decentralization has certainly contributed to the rapid development of the Arusha region. Local governments can and will do more, although they are not able to take up all the responsibilities fully, if only because of the lack of sufficient local finance. Local governments have received more autonomy in the framework of decentralization. Finance needs to be available and the financial sector seems to be aware of the potential. As one banker interviewed said: "We want to crack the rural-urban thing" and they are looking how other countries are doing it. They don't have the deposit base to go into rural lending on a large scale, but are looking for possibilities to decentralize their operations, which could certainly help regional development. According to the Regional Commissioner in Arusha the success of his region is not linked to decentralization, because decentralization came later and the animals were there already. Still I consider decentralization as helping and brought that subject up with the Regional Administrative Secretary. The Administrative Secretary's task is divided between keeping law and order and promoting economic development. The Regional Commissioner is backstopping, coordinating and advising, but the local governments have to do it. The regional administration is the link between the central and local government, and it seems to inspire the local level in this case. The Administrative Secretary confirmed that decentralization creates the conditions for a more active role of local government and creates checks and balances, by promoting transparency and accountability. The struggle at different levels of government goes on to receive more responsibilities and the necessary funds from the national for the district level to carry them out. However, certain ministries want to hold on to their power. The fact that the departments at the national and at the districts level are not aligned is a problem. They don't correspond one to one. For example at the national level there may be in the ministry a department of roads and one of water and at the local level these two activities may be together. In that case they are not in step, or sometimes not even in contact. The alternative would be to allow the informal sector to develop further. These units are not registered, but they certainly make up a local sector, with a lot of entrepreneurial capacity. The authorities should be tapping into it. A typical example is the gem stones trade, they are bought and sold in certain streets of Arusha, are in the informal sector. There is no viable formal industry and the government should not do it. It happens in this way as well. There are numerous examples of successful small enterprises. Women who used to sell food on the street and now supply hospitals under a catering contract. There are endless examples of smuggling in this region, which can take a lot of different forms. For 83 example up to 60 percent of the meat in Kenya comes from Tanzania. It is organic cattle (no artificial products in the meat) and hence it should command a higher price, just like the coffee, if Tanzania can get it certified. In the same way most of the oil seeds leave Tanzania, but they come back as Kenyan vegetable oil. Finally, the farmers produce very good garlic in this region, but the good quality stuff is smuggled to Kenya. Other examples are sweets and biscuits. In Mwanza the CTA members complain about smuggling these goods through the lake. In the Kilimanjaro region the locally produced matchboxes are facing illegal imports and counterfeits. Smuggling becomes more sophisticated and a lot of it comes through Mombassa and Moshi. 8.7.9 The role of different stakeholders The CTI and its members are active in food processing, for example producing edible oil, in textiles, or in making iron and steel products. As a pressure group they knock on the door of the regional administrator. However, skilled labor is short in the region and the labor productivity is not very high. Investors want work permits to bring in foreigners. That seems to be easier in Uganda and Kenya. However, they have now created a vocational training program locally to improve the skills of the workers through a government agency. For example they train masons and carpenters. All employees should contribute to this effort. The Chambers of Commerce, Trade, Industry and Agriculture are rolling out over the country. They now exist at the district level and even small and medium size enterprises (SMEs) are welcome. The CTI also counts on the SMEs as his future members. Some of their members export to the UK. The conclusion is that the different stakeholders are organized and increasingly focal. 8.7.10 Local development policies Do local governments pursue economic policies? The liberalization policies have allowed the private sector to emerge. Do the local authorities involve the private sector? The answer of the Regional Commissioner in Arusha was: "Yes we try, including the NGOs". The Japanese tiger expert has been here and made some suggestions for the future development of the region. In the Kilimanjaro region the authorities want to start a new economic zone. There is a machine tool factory not working any more, which can be used as the base. They could produce insecticides and pesticides and develop it into an industrial zone. The dynamics of the Arusha region is also the result of the improved business climate there. In the Arusha region the authorities plan to organize a event for potential investors as well. Secondly, Arusha is preparing an investment policy for the region, but most investors want to invest in Arusha district (93 km2) and not in the region as such. Arusha city is almost full, but the authorities try to induce them to come to other districts as well. The Arameru district around Arusha is a logical next location. In the Munduli district there is still land available, but there are fewer other amenities. With other amenities the authorities will attract other investments, but it also works the other way around, once the 84 investment is there the authorities will have to invest in infrastructure. In terms of a positive policy environment, Arusha has a lot to offer. The high tariffs of utilities and of taxes in Tanzania compared to Kenya and Uganda are a problem. Tanzania has the highest rate of VAT (20 percent versus 17 in the case of the other EAC partners) and local taxes are high. The local government is not always very cooperative. However, they CTI works very close with the regional administrator. As a pressure group the CTI knocks on the door of the regional administrator. 8.7.11 Organizing capacity We note that there is certainly a vision on long-term sustainable development (for strategies, programs and projects to hold on to) at the national level. These ideas are now filtering through at the regional level. The Regional Commissioner in the Arusha region seems to have a vision, even if it is mainly that he is responsible for law and order and should stimulate lower levels of government to take up their responsibilities. We do not yet see a culture of formulating concrete, measurable objectives at the local government level. However, strategic and coherent thinking and acting is becoming more and more the norm. In the same way the current system allows local leadership qualities to emerge and to manage the development processes and projects adequately. The functioning of the Chambers of Commerce and CTI as important stakeholders in the region suggest that creating and supporting strategic networks of relevant partners, needed to develop and implement policies successfully. It is taking place more often, in particular in the Arusha region. We also see the creation of political and societal support by major political actors, which helps to create an enabling environment for local economic development.. 8.8 Conclusions on booming and declining regions and cities in Tanzania The two regions are interesting because they are to a large extent comparable: regions in one country, located in the north under similar geographical and climatological circumstances and both bordering with Kenya. Moreover both are attractive for tourism and have a similar size population and have recently been opened up by improved roads leading to the coast and in particular to Dar es Salaam. There are also differences because the Kilimanjaro region is overpopulated and had an earlier boom period. Arusha has received important government investments and has still land available, for example for horticulturalists from abroad interested in the climate and relatively cheap labor. What strikes is that not so much the underlying factors like health and education explain the difference in development between the two regions. In fact both regions are involved in similar economic activities, although Arusha has the more promising agricultural and industrial activities, while Kilimanjaro has suffered some setbacks from the decline of coffee and the closing of some important industries. 85 The different population density between the two regions are largely caused by differences in population increase, which is a result of fertility and mortality changes and factors such as migration. Immigration has been the leading factor for high population density in big cities like Dar es Salaam. Equally, the availability of arable land and its quality is of considerable importance in explaining the internal disparities in population densities between different regions The Arusha region is trying to create the atmosphere for local people to produce more by arranging classrooms, teachers, medicines, etc. The authorities build the administrative offices, the hospitals and dispensaries, but some contribution will come from local government and some from the people themselves. He said that the Kilimanjaro region peaked too early. In principle the region has no proper funds except for what it gets from central government. Factors explaining the differences between the regions can be distinguished in general and specific factors contributing to the growth and decline of the two regions. The latter ones are listed in table 22. Comparing the Arusha and Kilimanjaro regions, the CTI members interviewed considers these two are relatively well off compared to other regions in Tanzania. In fact the Northern part of the country cannot be compared to the west and the south, where there are as niece tourist attractions, but they have not been marketed properly. In terms of explanation of the success of the Arusha region people mention the good climate, the availability of land and the city functions as a hub in the northern part of the country. Concerning the decline of the Kilimanjaro region people mention low coffee prices and more ordinary coffee, while Arusha had better quality coffee, which still commands a good price. Also land scarcity is often mentioned and the lack of medium priced tourist facilities in Kilimanjaro. Table 22. Specific factors explaining the performance of the two regions The Kilimanjaro region The Arusha region 1. Changes in the agricultural sector: 1. The secretariat of the East African decline sisal, coffee and land Union scarcity 2. Seat of the Rwanda war criminals 2. No middle class tourist court accommodations 3. Potential horticulture and available 3. Some industries have been closed land and water down 4. Tourists prefer Arusha 4. Decrease in local demand 5. Increased local demand 86 8.9 Conclusions and policy implications Kilimanjaro has developed its own investment center in Moshi, but so far it has not helped the region very much. As far as economic policies are concerned, the local governments implement the policies formulated at the national level, but also come up with their own ideas, such as raising their own taxes or creating boarding schools. How to do this is a local decision, it goes as far as deciding what to grow, for example this choice to go for the flowers in Arusha (these people are bothered by legal and illegal inspectors). They can also create local incentives. In particular local governments in Dar es Salaam have local economic policies within the national policy of the unitary state. Do local governments have economic, social and environmental policies? No, at the moment local governments implement environmental legislation, but have no local social policies. Again they implement national social policies, but could pass their own byelaws. Financial experts do believe a sustained growth of 6 percent for Tanzania is possible and the banks want to contribute to it. Cities play an important role in the development process, also in Tanzania. Where Dar es Salaam is a peripheral node in the global world system, different often informal transmission mechanisms transmit part of the growth in the capital to regions, which all benefit from these impulses in a different way, depending on their resources, history and human resources. However, the flows of people, ideas, money, goods and services certainly contribute the spreading development in Tanzania. 9. A potentially promising economy is negotiating the terms of globalization 9.1 Introduction Globalization has enormous consequences for developing countries (Visser and van Dijk, 2006). Not only all kinds of products appear in the market, which were never there (in an increasing degree they come from China and India), also in different platforms countries have to decide whether they want to participate in the globalization process and at which conditions or price. The discussions on this topic take place in bilateral or in a multilateral negotiations for trade agreements, in a global, or regional context, or just with a few neighboring countries. It is currently negotiating the terms of globalization in a number of bilateral and multilateral fora, for example in the framework of the Doha Round of the World Trade Organization (WTO; temporarily suspended), with European Union (EU) and the United States (US), in the East African Community (EAC) and Southern African Development Community (SADC). It will be shown that for a country like Tanzania it is not easy to negotiate simultaneously in the framework of the WTO, with the United States (the African Growth and Opportunity Act or AGOA) and Europe (an Economic Partnership Arrangement or EPA) and to be active in a regional context in particular in the EAC and the SADC. Matters have become more complicated because Tanzania is pushed to choose between becoming a full member of the EAC or of SADC (Citizen 22-8-2006: 1-2). The question asked is: how best to do business with your neighbors? 87 9.2 Tanzanian economic policies Tanzania does have a strategy for the future, in fact it has several. The result of the reform process that took place in the nineties has been a higher economic growth and a more investors-friendly climate (IFC, 2006). Tanzania has taken a number of initiatives, such as trying to increase the number of joint ventures with Chinese and with US firms. Sometimes it seems the country follows too many different winds. There is for example the plan 2025, there is the Mini-tiger plan with Japanese technical support and there is De Soto's formalization approach for the Tanzanian informal sector. With the help of the Economic and Social Research Forum (ESRF) the government has a private sector development strategy and it will also have a small and medium enterprises (SME) policy and has already a trade policy. However, these strategies are not always fully implemented. The join SADC or EAC discussion comes in parliament in October 2006. The arguments for choosing SADC used are that Tanzania's roots are more in the Southern African countries and in supporting their struggle against apartheid. In EAC Tanzania is very much afraid of Kenya and the experience with the first EAC has proven them right. Tanzania woes Chinese investors and although these are mainly interested in selling their products and buying raw material in Africa, a mission visiting the country in August 2006 made some promises. According to the Guardian (24-8-2006) the vice president has pledged government support to investors trying to realize their goals. The Chinese answered that they seek investments in industry, minerals, agriculture and irrigation and China has helped Tanzania with the development of an Export Processing Zone (EPZ). It is not immediately clear why China would invest in Tanzania other than for assuring its supply of raw materials. You cannot expect them to play the role that Japan has played for many East Asian tigers. Technical education has for example been neglected, meaning that companies have to train their own people. This is a factor increasing the costs for a foreign investor. Tanzania also gives more attention to decentralization. A meeting of regional leaders was held for the second time in 2006 and shows the authorities take decentralization more serious. The president himself addressed the meeting for two hours suggesting a code of conduct, which made it very explicit that a lot is expected from these regional leaders. In 2003 the Ministry of Industry and Trade of Tanzania formulated a national trade policy for a competitive economy and export-led growth (MIT, 2003). It starts with making clear that trade plays a `pivotal role' in achieving an ambitious future and that it needs to be stimulated. Import substitution and autarky are currently terms of the past. Tanzania wants to benefit in particular from the positive effects of international capital flows, for example through joint ventures with American and Chinese companies. MIT (2003) is quite a comprehensive document and makes clear that the country really wants to go for competition, regional integration and multilateral trading, but as will be shown the current situation has become more complicated because of a breakdown of the WTO negotiations, pressure to finish the inter-regional schemes, in particular the EPA agreements with the EU and the suggestion made to choose between SADC and EAC. In 88 the meanwhile the EAC has started an ambitious program in the direction of a union, skipping a number of stages, which are normally in between creating a free trade area and full economic integration (see box 14). This rush to unification of the EAC countries risks according to certain people, "to dilute our credits and to share chaos, crime, tribalism, more corruption, civil unrest, etc" (letter to the Citizen, 25-8-2006). Box 14 Different levels of economic integration A free trade area, an area with no visible trade restrictions; A custom union, which is the same plus a common external trade policy; An international commodity market: the same but also no invisible trade restrictions; Common market: the same plus a free movement of factors of production; Monetary union: the same plus a common currency; Economic union: the same plus a common economic policy. Source: Nielsen et al. (1991). MIT (2003) attaches a pivotal role to trade and doesn't discuss autarky, import substitution and protection, but rather argues in favor of globalization and further regional and global integration. The EAC is implementing a custom union a higher form of integration on the way to an East African Union. A custom union has a common external policy and tariffs and that is exactly the problem. If Uganda and Kenya decide on a different tariff in the framework of the Common Market for Eastern and Southern Africa (COMESA, where these two countries participate in and negotiate through with the EU about EPA) Tanzania will have to accept it. Secondly, the two blocks negotiating with the EU don't agree yet on the priority sectors. That discussion may need more time. In SADC they have chosen fisheries, which may not be a priority for COMESA. In 2004 Tanzania received more foreign direct investment than Uganda and Kenya (US$ 470, 237 and 46 million). 2002 was a bad year with low commodity prices. Since then tea, cotton and tobacco went up, but now coffee is going down again. As an exporter of raw materials Tanzania very much depends on world market prices. The National Strategy for Growth and Reduction of Poverty (NSGRP), which was produced in 2002, is a good oversight of what the country does and wants to achieve. A lot of money is spent on improving the infrastructure and transport. So far Tanzania's growth seems to continue, although the economy has suffered from the drought and electricity cuts. Normally 70 percent of the country's electricity consumption would come from hydropower, but that is not working with the current level of the reservoirs. In 2006 in the second half of the year load shedding became every other day in Dar es Salaam, which is very disturbing for small businesses like internet cafes and car repair shops. It may have a negative effect on economic growth. On the other hand tourism, gold and minerals keep increasing in price or volume contributing to Tanzania's high economic growth and good export performance. 89 9.3 Trade and the role of China Export has increased by 18.2 percent in 2004. The most important export products are classified by URT (2005) as non-traditional and traditional. The first category makes up 78.1 and the latter 21.9 percent. Products are coffee, cashew nuts, cloves, cotton, sisal, tea and tobacco. Non-traditional exports concerns gold, minerals, manufactured goods, fish and fish products, horticultural products and other products. When a close look is taken on the trade statistics, it turns out that the real dynamics for Tanzania is in the trade with India and China. Their shares in Tanzania's trade are increasing rapidly over the last years. It is difficult for Tanzania to compete with the top league, for example South Africa, which is considered as very advanced. The impact of China's rapidly increased textile exports after the termination of the Multi Fiber Agreement (MFA) has been devastating in a number of African countries. In Tanzania it is very limited, because the textile sector had already problems for a long time and is still in a restructuring process. Textiles from Southeast Asia have hit Tanzanian textiles and garments, already before the end of the MFA. It started in the nineties when the country opened up its economy. All shops in Dar es Salaam are filled up with Southeast Asian textiles and Tanzania is mainly producing specific types of textile and clothing for the local market. The impression exists that a lot of these products are dumped by China and Dubai, but the government doesn't want to follow up on this information. Good shirts are sold for less than 1500 T Sh. Some trade experts think that Tanzania should ask the WTO to set up a joint committee to investigate the dumping. Subsequently the country can then try to revive its own textile and garments sector. 9.4 Frameworks in which Tanzania negotiates the conditions for globalization Tanzania is a member of the WTO. It participates in a custom union with Kenya and Uganda (the East African Community) and is negotiating a new partnership with the European Union. The results determine the global context in which Tanzania is functioning. Kweka and Booth (2004) explain why there are linkages between trade and poverty in Tanzania. They also emphasize the role of policy in assisting the development of these linkages. The EAC is a custom union, which still allows Uganda and Tanzania to impose tariffs for some time (five years as of 31-12-2004). URT (2005: 29) notes that in January 2004 the EAC Council of Ministers approved an East African Industrial Development Strategy. They also ratified protocols on sustainable development of Lake Victoria Basin and a protocol on cooperation in road transport. Tanzania gets new opportunities to develop activities in these frameworks. Transport not only contributes to poverty alleviation by assuring cost-effective transport of goods and people. It also links the country to the global economy and assists neighboring landlocked country to connect to the world.xxix There is some optimism over the EAC. Previously Kenya mainly benefited from foreign investment and exported its products to its partners. Almost all multinationals had their 90 headquarters in that country. Now the other member countries have some time before all import duties go to zero (5 years, starting in 2005). The manufacturing sector hopes to be competitive by that time, but rightly fears South Africa. The Southern African Development Community (SADC) includes seven countries: Angola, Mozambique, Tanzania, Botswana, Lesotho, Namibia, Swaziland; duties will be abolished by 2008. It will be no surprise that South Africa calls the slots in this forum. The EPA with the European Union needs to be concluded before the end of 2007. Many of the Lomé countries (ACP) fear that the European Union (EU) will offer all developing countries tariff zero. 9.5 Current ongoing trade negotiations The WTO Doha round has been interrupted when the major powers could not reach an agreement. Hence the chance that the Doha round will fail is there and the number of bilateral regional agreements is increasing and will probably result in more drastic changes in trade conditions. Members of the employers association (Confederation of Tanzanian Industries) are often informed about the results of the Doha Trade Round of the WTO and the Confederation of Tanzanian Industries (CTI) is involved at the national level. However, the government takes the lead. He has not yet seen the Diagnostic Trade Integration Study (World Bank et al., 2005). It is important that the different actors are involved in the process of shaping the external environment for Tanzania. The country has gone through a period of rapid change and risks severe competition from more developed developing countries. It is imortan6t that it looks for and finds the activities where it has a competitive edge. Tanzania recently did away with all subsidies in agriculture. The country expected to really benefit from a further liberalization of international agricultural exports, as discussed in the WTO's Doha round. However, the EU and the US did not want to go far enough in the eyes of a number of other countries. There is a lot of skepticism with respect to the EPAs, which are often considered as a way to defend Europe's interests in its former colonies! Tanzania's capacity to negotiate directly with the EU is limited. In the negotiations about EPA all clusters should finish in July 2007, then there are three months for finalizing the agreement. The negotiations concern one theme at the time. They have finished Sanitary and Phytosanitary regulation and rules of origin and started January 2006 on non-agricultural commodities, which will require probably more time. The last theme is services. Organizations like the CTI follow the negotiations, but most of their members have no big trade with the EU. They don't see a threat in such an agreement, but the CTI informs them that EU products are coming to Tanzania. Since Tanzania can already export everything for free to Europe the EPA will not help much. Although Tanzania is member of EAC and SADC, the country has chosen to do the negotiations with the EU in the framework of SADC and to try to coordinate with the EAC partners negotiating in the framework of COMESA. The issue is not just tariffs, but 91 also the stringent non-tariff barriers, such as phyto-sanitary requirements and the so- called tracing requirements, which make it necessary to always be able to track a product to its origin and which makes exporting more difficult in developing countries. As far as the current trade negotiations are concerned they are afraid of the common external tariff since Kenya and Uganda are negotiating in the Common Market for Eastern and Southern Africa (COMESA) framework and Tanzania in SADC. The tariffs are actually quite low, about 25 percent. In the case of EPA a real problem with the EU is the stringent rules of origin and Sanitary and Phytosanitary Standards (SPS) rules. Tanzania wants the EU to allow them some accumulation, for example if they import fabric from India and make garments here. The EU only allows a 60 percent value increase. The other bone of contention is subsidies. In agriculture we have a competitive advantage and hence the EU should lower the tariffs. The Tanzania government is currently discussing market access and export development in the framework of the EAC. The Export Processing Zone of Kenya is more advanced than what Tanzania and Uganda have and they are working on harmonization. Also the SPS in agriculture have been worked out as a follow up on the original negotiations. The negotiations in the framework of the EAC concerns the list of products, which cannot immediately be liberalized. The Ugandans have 130 products as exceptions and Tanzania only 25. According to Tanzanian sources Kenya has now started with non-tariff barriers for products in which Tanzania has a competitive advantage. Trade specialists regret that the negotiations with the EU are not taking place in the framework of the East African Community. This would started when Kenya and Uganda decided in 2003 to go for negotiating in the framework of COMESA and Tanzania did not want to follow and turned to SADC. This may have been an error. SADC is comprised of Southern African countries, which have different interests. This means there will be no preferences for Tanzania, which are against South Africa's interest. Hence Tanzania is left totally alone. The general opinion is that smuggling undermines the EAC. However, if tariffs are removed or lowered even more and when it is made easier to trade and to register the goods officially one may see the former smugglers going through the border legally. Also programs to facilitate border control and to stimulate cross border trade may increase. She doesn't believe in export-led development in Tanzania in the short run. The textile is a typical example. It is mainly for the local market. The focus is on local tissues and products, which face very little international competition. In the same spirit there is a considerable chance that the US will try to further develop bilateral trade relations, since the WTO negotiations have come to a standstill. Tanzania missed the opportunity to benefit from the American program providing free access to the US market for a number of products under AGOA, which has benefited in particular textiles and clothing exports in a number of African countries. URT (2005: 40) remarks that Tanzania is yet to benefit from this opportunity since exports to the US market has increased only marginally from US$ 11.3 to 13.7 million between 2003 and 2004. The exports concerned besides textiles: minerals, forest products and raw and processed 92 agricultural products. Imports from the US increased from US$ 69.7 to 78.1 million between 2003 and 2004. As far as the EAC is concerned Tanzania made its partners clear what for Tanzania sensitive products are. For these 25 products liberalization will be more slowly. In fact the reasons behind these 25 will be the same in the framework of EPA. But EPA also has assistance provisions. They will put resources to make our laboratories up to international standards and Tanzania wants its share of this money. Finally, SADC is an alternative, but the larger picture is that Tanzania is or has been also active, besides SADC and EAC in the Southern African Customs Union (SACU) and COMESA. The journal This Way (26-8-2006) rightly concludes that: "once COMESA and SADC become custom unions, countries with multiple memberships will ultimately have to opt for membership in one custom union only". 9.6 The dilemma EAC or SADC When at the beginning of 2006 an EU official said that Tanzania had to choose between EAC or SADC he received a reprimand. During a recent meeting of SADC some members phrased the same question and Tanzania will now have to give an answer. In its trade strategy it assumes it can follow both tracks, but both groupings have changed gear and want to go much further than originally announced. The EAC want complete unification of the countries involved and SADC intends to reach the next stage of integration in a relatively short period. In August 2006 the 12th meeting of the EAC Council of ministers has taken place in Tanzania. According to the Deputy Secretary General of EAC the community faces a big challenge of meeting the people's expectations. Already the launch of the EAC Custom Union had heightened expectations of increased productivity and trade in the region. People expected wealth creation and an improvement of people's standards of living. Negotiations concerned the admission of Rwanda and Burundi to the Community and establishing a protocol on the free movement of persons, labor and services (Citizen, 25- 8-2006). SADC will carry out a plan of transforming the block into a common market. The timetable is a free market by 2008, a custom union by 2010 and a common market by 2015. SADC held its annual meeting mid 2006, where some tough decisions needed to be taken. A few years ago, in Kampala, it was decided that Tanzania would go for SADC as a framework to negotiate an EPA with the EU. The EAC has a good relation with COMESA and signed a memorandum of understanding. Tanzania accepts what Kenya and Uganda negotiate in that framework. Tanzania has attended most of the meetings. Tanzania would like to do the same with SADC, but faces the domination of South Africa in the Southern African region, through the Southern African Customs Union (SACU). Unfortunately the Ministry of Industry and Trade (MIT) in Tanzania has few trade experts left to do all these negotiations. 93 For the SMEs choosing SADC or EAC is not an important issue. They are interested in the Tiger model. Tanzania can do it, but it is again a question of reaching more efficiency at different levels: the national, the regional and the local. Tanzania has to improve its business climate further: the fact that there is no water in the reservoirs to generate electricity is not a sufficient argument not to have electricity! The country hired an South African consortium to run TANESCO, but they changed the contract. For example the company is not allowed to increase the prices! The government also embarks on strange projects, such as housing for the middle income groups. That you could leave to the market. The private sector sees more opportunities to do business in EAC and COMESA and fear South Africa in SADC. 9.7 Consequences To play a role in the global economy it will be important that Tanzania becomes internationally competitive. A number of activities envisage improving competitiveness. Tanzania's Business Environment Strengthening project (BEST) for example is providing support to private business organizations in Tanzania. The entrepreneurs can get grants to advocate for changes in the business environment which are important to their members or the sector they represent. Money can also be obtained to allow such organizations to improve their capacity to advocate more effectively for changes in Government policy, regulations and so forth. Is the urban informal sector competing with the legal exports and imports? The people working there have been harassed for so long that they hardly believe `formalization' programs as suggested by De Soto. Plus they see the cleaning up programs continue in Dar es Salaam and Arusha. However, the informal sector contributes to the dynamics of these regions. The formal sector uses containers, the informal bicycles. It means their capacity is limited. It is may be 70 versus 30 percent as far a border crossing is concerned. What is needed are better conditions for competition: water and electricity, developing the local labor force and solving a number of other issues. At the EAC they discuss the non-tariff barriers a lot. Tanzania suspects that Kenya throws up some of those because Uganda and Tanzania have already tariff zero in certain products. Politically the presidents are pushing for a federation of the three countries and they would have to facilitate that! 9.8 Conclusions The Tanzanian economy functions in a global context, which it can influence only to a small extent through negotiations in different trade fora. A relatively small country has to play simultaneously at least five matches with few specialists in the public sector and no clear strategy. The trade policy document from 2003 is already three years old and already dated. One can choose between two attitudes. In the first place trade is available and just needs some checking. Also concerning the informal trade this approach would be important. Secondly, one can try to impose tariffs and control exports and imports. In practice this is not an easy task. 94 Tanzania is in a very interesting stage in its development. If it continues with its open and problem solving approach it can become one of the successful African economies, which should also affect its poor positively. How this works exactly is difficult to assess, since the data are not readily available about the mechanisms by which growth reaches the poor. The key question is where can Tanzania be competitive? Certainly in tourism, but also in horticulture seems to be a promising sector. Don't forget that Kenya earns more foreign exchange through vegetables and flowers than from tourism (Financial Times, 26-9- 2005). Tanzania finds it difficult to compete in an international context and its markets are flooded with cheap products from Kenya, China and South Africa. In a more liberalized world economy a lot of specialized agricultural products could also be exported and the country has an interest to anticipate such developments, which can be expected in the framework of the Doha round and the current negotiations with the EU. The lack of electricity of a good railway system and the relative high minimum wages mean that not many foreign investors are interested in setting up a factory in Tanzania. The cost of bringing a container from the port of Dar es Salaam to Rwanda is still higher than shipping one to Japan. Tanzania has to really do its best to negotiate favorable tade deals for the country. 10 Conclusions Tanzania is in a very interesting stage in its development. So far the economic growth seems to continue, although the economy has suffered from the drought and electricity cuts. Normally 70 percent of the energy would come from hydropower, but that is not working with the current level of the reservoirs. Load shedding is every other day. This is very disturbing for small businesses like internet cafes. It could have a negative effect on growth. On the other hand tourism, gold and minerals keep increasing in price or volume. It may be too early to call Tanzania an African Tiger, but using the label increases the confidence of investors. If Tanzania continues with its open and problem solving approach it can become one of the successful African economies, which should also affect its poor. How this works exactly is difficult to assess, since the data are not readily available about the mechanisms by which growth reaches the poor. However, we started with five possible explanations, and will now summarize the evidence. Tanzania certainly benefited from economic restructuring and may eventually become an African tiger. More important is the effect of decentralization discussed in chapter 3, 7 and 8. It means people at the local level have more freedom to act and possibilities to mobilize finance for their development activities. We noted the positive effects of infrastructure built by different levels of government and of the expenditures of local governments. There is also direct central government support to regions and Dar es Salaam and Arusha have certainly benefited form it. Still there is currently not the infrastructure necessary for accelerated development (chapter 4), but the informal sector plays an important role in redistributing the growth as 95 shown in chapter 5 and chapter 6. Private and partially foreign investments take place and lead to dynamic sectors like mining and tourism, which contribute to the growth of the regions or cities where they are located and through the multiplier effects of these expenditures. Also large amounts of aid money are spent in certain regions and bring about development in those regions. We argued that these are redistributive mechanisms that work and improve the situation of people living outside Dar es Salaam, although it may not always be shown in the official statistics. Despite a lack of data, the description and partial analysis of the informal flows of people, money and goods and services has shown the importance of these flows for people in cities and regions. Dar es Salaam is currently the dynamic center of the country, linking it to other parts of the world economy. Hence the challenge is to develop that role and at the same time use the potential of this city for the development of other cities and regions, by stimulating the informal flows as redistributing mechanisms. As a conclusion it seems that a number of the factors work together in regional and urban development. Such development also happens because the population has more confidence in the current government and that government is trying to do a number of things differently. If the current positive environment can be maintained more investment can be expected and the informal flows of people, money and goods and services will assure a certain dynamics and a spread of the benefits of development throughout the country. In the future we can think in terms of an urban hierarchy, where different processing activities take place at different geographical levels. Also tourism is an important sector to develop further and to help development spread over the country. However, despite all efforts it is estimated that Kenya still receives about one million tourists per year, while Tanzania only gets about 300,000 per year. 11 Some recommendations The old thinking was that development would filter through and eventually reach the poor, but in the nineties policy makers became aware of the need for complementary policies, such as social development funds (ESRF, 2001) and more attention for specific groups. Micro finance programs, social development funds and public employment schemes are all examples of such additional policies. From China we have also learned that development may start with educating the poor and providing the necessary health care and infrastructure. Recommendations can be formulated for the decentralized local and regional governments, which have been consciously created by the national government. The challenge is to use the different levels of government optimally. Local and regional governments could play a more pro-active role in the field of local economic development policies. Like in China local governments could compete with each other to attract foreign direct investments, or local investments, which would contribute to employment, local revenues and development in general. 96 Secondly, other sources of income for local governments, such as loans or issuing bonds should be considered. Van Dijk (2000) describes how in India local governments were connected to the capital market. Given the interest of private financial institutions to become more active in the rural areas and the scope for more micro financial activities in the rural areas, a potential could be unlocked. In the third place, a more dynamic strategy can be developed to further develop the capital and to market the city and in particular its port as an entry point for Eastern Africa. Besides considering the role of the port we should look at the connecting infrastructure in the national and supra-regional context. The port and infrastructure could serve Congo, Burundi, Rwanda and Zambia, if the connecting infrastructure is improved and a study is carried out what is necessary to become a real competitor of Mombassa, the ports of Mozambique or Durban. Solutions may also be increasing the technology level and skill intensity in agriculture. In the field of urban management (Van Dijk, 2005b) it is important to go beyond the typical town planners approach of indicating which activities should take place where, but failing to tell how this should be achieved. Urban managers would discuss with the stakeholders what each one could contribute and how the plans can be financed and implemented. There is a need to look at the possible densification in Dar es Salaam. The current vast area means investments in infrastructure will be very costly and people spend a lot of time and money on traveling. The urban informal sector is very much a local sector in Tanzania. However, local products will still make up the majority of the expenses of the Tanzanian population. Most of the money is spent on food, housing and clothing, which are using mainly local products, except for the imported garments and secondhand cloth. Hence it makes sense to develop the sector in more systematic way as recommended for example by ILO (2004). The question is where can Tanzania be competitive? Certainly in tourism, but also in horticulture seems to be a promising sector. Don't forget that Kenya earns more foreign exchange through vegetables and flowers than from tourism (Financial Times, 26-9- 2005). Tanzania finds it difficult to compete in an international context and its markets are flooded with cheap products from Kenya, China and South Africa. In a more liberalized world economy a lot of specialized agricultural products could also be exported and the country has an interest to anticipate such developments, which can be expected in the framework of the Doha round and the current negotiations with the EU. 12 References Altenburg, T. (2005): A critical assessment of the minimalist approach to private sector development. Bonn: EADI Conference September, 23 pages. Bagachwa, M.S.D. (1995): The informal sector under adjustment in Tanzania. Kilmanjaro: Workshop organized by ERB of the University of Dar es Salaam, 18 pages. Bagachwa, M.S.D. (1999): The rural informal sector in Tanzania. In: Bryceson and Jamal (eds, 1999). 97 Berg, L. van den, E. Braun, and J. van der Meer (1997): Metropolitan organizing capacity. Aldershot: Ashgate. Bongenaar, B. and A. Szirmai (2003): The role of a research and development institute in the development and diffusion of technology (SIDO in Tanzania). In Van Dijk and Sandee (eds, 2003). Bryceson, B.F. and V. Jamal (eds, 1999): Farewell to farms, De-agrarisation and employment in Africa. Aldershot: Ashgate. Castells, M.(1997), The information age: Economy, society and culture. Oxford: Blackwell. Cheema, G.S., and D.A. Rondinelli (eds, 1983): Decentralization and development: Policy implementation in developing countries. Beverly Hills: Sage. Cohen, J.M. and S.B. Peterson (1999): Administrative decentralization: Strategies for developing countries. Bloomfield: Kumarian Press. CIDA, SIDA, DFID and RNE (2003): Tanzania pro-poor financial sector deepening program, Strategy paper. Dar es Salaam, 19 November, 48 pages. CIDA, SIDA, DFID and RNE (2004): Introduction to the program for Business Environment Strengthening For Tanzania (BEST), 107 pages. City Director (2004): Dar es Salaam City profile, Dar es Salaam City Council, 66 pages. CTI (2004a): Achievements. CTI (2004b): Secretariat. Danielsson, A. and G. Mjema (2000): Tanzania 1999, Obstacle for the private sector, p. 21-51. Demombynes, G. and J.G. Hoogeveen (2004): Growth, inequality and simulated poverty paths for Tanzania, 1992-2002. Washington: World Bank Policy Research Working Paper 3432, October, 36 pages. De Soto, H. (1989): The other path. New York: Harper and Row. Dijk, M.P. van (1987): Burkina Faso, Le secteur informel de Ouagadougou. Paris: L'Harmattan. Dijk , M.P. van (2000): Municipalities' access to (inter-)national capital markets for financing urban infrastructure. In: K. Singh and B. Ta'i (eds) Financing and pricing of urban infrastructure. New Delhi: New Age International, pp. 157-179. Dijk, M.P. van (2006a): Managing cities in developing countries, the theory and practice of urban management. Beijing: Renmin university press and Cheltenham: Edward Elgar. Dijk, M.P. van (2006b): Innovativeness and competitiveness in Nanjing: Overlapping governance structures stimulating an emerging IT cluster. Rotterdam: Erasmus University, HOVO (forthcoming). Dijk, M.P. van (2006c): Different effects of globalization for workers and poor in China and India, Comparing countries, clusters and ICT clusters? In: Tijdschrift voor Economische en Sociale Geografie, dossier Globalization and labor, Vol. 97, No. 5, pp. 501-513. Dijk, M.P. van and Sandee (eds, 2003): Innovation in small enterprises in the Third world. London: Edward Elgar. Egziabher , T.G. and A.H.J. Helmsing (eds, 2005): Local economic development in Africa, Enterprises, communities and local government. Maastricht: Shaker. ESRF (2001): An assessment of the impact of macro and sector policies on poverty reduction initiatives. Dar es Salaam: ESRF, 63 pages. 98 ESRF (2004): Quarterly economic review, Vol. 7, Issue I, January-March. Foeken, D., M. Sofer and M.M. Lozi (2004): Urban agriculture in Tanzania. Leiden: Africa Studies Centre. Footer, M.E. (2005): An institutional and normative analysis of the WTO. Rotterdam: Erasmus University. Fornahl, D. and T. Brenner (eds, 2003): Co-operation, networks and institutions in regional innovation systems. Cheltenham: Edward Elgar Franzsen, R.C.D., D. Johnstone, T. Johnstone and W.J. McCluskey (2002): `Evaluation of the Rating Valuation Consultancy Services and Property Tax Administration in Dar es Salaam and Eight Project Towns in Tanzania', Report for the Government of Tanzania and the World Bank (June 2002). Franzen R.C.D. (2004): The current two-tier system of local government in Dar es Salaam and the functional and financial division between these tiers. Report for the Government of Tanzania and the World Bank, 81 pages. Gallardo, J., K. Ouattara and W.F. Steel (2005): Microfinance regulation: Lessons from Benin, Ghana and Tanzania. In: Savings and development, Vol. XXIX, No. 1, pp. 85-96. Helmsing, A.H.J. (2000): Decentralization, enablement and local government in low income countries. Utrecht: University, Inaugural address. Helmsing, A.H.J. (2005): Local economic development, in: Tegegene and Helmsing (eds, 2005). Henderson, V. (2000): How urban concentration affects economic growth. Washington: World Bank. Heymans, C. and Krish Kumar (2002): The scope for fiscal and institutional restructuring in Dar es Salaam as a component of the Tanzania LGSP. World Bank, draft March-May, 63 pages. Hoogeveen, J.G. (2004): Reaching the MDG/NSGRP poverty objectives. Contribution to the CEM, 9 pages. Hyden, G. (1980): Beyond Ujamaa in Tanzania, Underdevelopment and an uncaptured peasantry. London: Heinemann. IFC (2006): An international investment climate study. Washington: World Bank. ILO (1972): Employment, income and equality: A strategy for increasing productive employment in Kenya. Geneva: ILO. ILO (1978): Towards self-reliance, Development, employment and equity issues in Tanzania. Addis Ababa: JASPA. ILO (1991): Tanzania meeting the employment challenge. Addis Ababa: JASPA. ILO (1993): The state of the informal sector in Tanzania. Addis Ababa: JASPA. ILO (2004): Informal sector road map study for Tanzania, Dar es Salaam, September, 85 pages, with UNDP and UNIDO. ILO (2005): Local employment in the informal sector in Tanzania, draft report. Dar es Salaam, April, 126 pages. Jacobs, J. (1970), The economy of cities. New York: Vintage. Kanbur, R. (2001): Economic policy, distribution and poverty: the nature of disagreements. In: World Development, 29, pp. 1083-1094. Karanja, A. (2005): Solid waste management in Nairobi, Actors, institutional arrangements and contributions to sustainable development. Maastricht: Shaker. 99 Kulumba, S.M. and C. P. B. Mkai (2004): Urbanization in Tanzania, chapter 13 in: URT (2002), pages 230-296. Kweka, J. and D. Booth (2004): Precedence for trade and poverty linkage in Tanzania. Dar es Salaam: ESRF, 5 pages. Kweka, J., O. Mashindano, D. Rweyemamu and V. Leyaro (2004): Supply factors affecting rural credit in Tanzania. Dar es Salaam: ESRF. Laryea-Adjei G. Q. M. (2006): Central-local relations in the provision of basic services, Provision of water and santiation services in Ghana. Rotterdam: Erasmus Uinversity PhD. Lee, K.S. en R. Gilbert (1999): Developing towns and cities: Lessons from Brazil and the Philipines. Washington: World Bank. Lindeboom, W., A. Makbel and I. Ruyobia (2006), Smallholder characteristics asessed by household welfare differences. Dar es Salaam: REPOA and NBS. Local Government Capital Development Grant System (2004): Planning guidelines for village and Mitaa, October, 30 pages plus annexes. Local Government Capital Development Grant System (2005): Manual for the assessment of councils against minimum access conditions and performance measurement criteria. Dare es Salaam: The President's Office, November, 24 pages. Mbonile, M. J. (2004): Migration, chapter 7. In: URT (2002), pp. 117-132. MIT (2003), National trade policy, Background papers, Trade policy for a competitive economy and export-led growth. Dar es Salaam: Ministry of Industry and Trade. Mitullah, W.V. (2005): Tapping opportunities in decentralized governance and informal activities for urban development in East African Countries, Dar es Salaam: ILO. Mlingi, C.E. and J. S. Assey (2005): The consequences of urbanization in Tanzania, A case study of Dar es Salaam and Arusha municipalities. Dar es Salaam: draft report consultants UNESCO-IHE, 94 pages. Mwase, N. (1995): Road and bridge maintenance strategies in Eastern and Southern Africa: Opportunities and challenges. In: International Journal of Transport Economics, Vol. XXII, No. 1, February, pp. 65-84. Naschold, F. and A. Fozzard (2002): How, when and why does poverty get budget? Poverty reduction strategy and public expenditure in Tanzania. London: ODI, WP 165.. NBS (1999): Arusha region, Socio-economic profile. Dar es Salaam: President's Office Planning and Privatization. NBS (1998): Arusha region, a socio-economic profile NBS (2002a): Household budget survey 2000/01. NBS (2002b): Industrial commodities, Quarterly report, September 13 pages. NBS (2002c): Kilimanjaro region, Socio-economic profile. Dar es Salaam: President's Office Planning and Privatization. February, 237 pages. NBS (2002d): Integrated labor force survey 2000/01, Analytical report. NBS (2003a): National accounts of Tanzania 1992-2002. NBS (2003b): Annual survey of industrial production, Vol. III Directory of industries 1999 and 2000. Dar es Salaam. NBS (2004): Population and Housing Census in 2002. Dar es Salaam. NBS (2006): Agriculture sample census 2002/03. Dar es Salaam: National Bureau of Statistics, Census team. 100 Ngoi, G., M. K. Misanga and H. A. Khamis (2004): Economic characteristics of the population. In: URT (2002), pp. 48-64. Nielsen et al. (1991): Economic integration. NSGRP (2002 ): National Strategy for Growth and Reduction of Poverty. Dar es Salaam. Ohmae, K. (1996), The end of the nation state, The rise of regional economies. London: Harper Collins. OECD (1996): Evaluation of programs promoting participatory development and good governance, Synthesis report. Paris: OECD DAC Expert group on aid and evaluation. Owen, T. and F. Teal, (2005): Paths Out of Poverty in Ghana and Tanzania in the 1990s, Oxford: Centre for the Study of African Economies, University of Oxford. March. Owuor, S. O. (2006), Bridging the urban-rural divide, Multi-spatial livelihoods in Nakuru town, Kenya. Amsterdam: UVA, PhD. Parker, A.N. (1995): Decentralization, the way forward for rural development. Washington: World Bank, Policy Research Working Paper 1475. Rees, W.E. (1992), Ecological footprint and appropriated carrying capacity: what urban economics leaves out. In: Environment and Urbanization, 4, 2, pp. 121­31. REPOA (2004a): What is the Tanzania governance notice board? 4 pages. Dar es Salaam. REPOA (2004b): Trust in public finance: Citizens' views on taxation by local authorities in Tanzania. 4 pages. Dar es Salaam. REPOA (2004c): Domestic water supply: The need for a big push. 4 pages. Dar es Salaam. Rondinelli, D.A. and K. Ruddle (1978), Urbanization and rural development, A spatial policy for equitable growth. New York: Praeger. Rutasitara, L. (2002): Economic policy and rural poverty in Tanzania: A survey of three regions. Dar es Salaam: REPOA. R&AWG (2002): Poverty and human development report 2002. Dar es Salaam: Research and Analysis Working group. R&AWG (2003): Poverty and human development report 2003. Dar es Salaam: Research and Analysis Working group. Sassen, S. (1991): Global cities. New York: Princeton Scott, A.J. and G. Garofoli (2004): The regional question in economic development. Background document for the Bellaggio workshop. SNV (2004): Strengthening local governance, Finding quality advisory approaches. Nairobi: Netherlands development organization, 94 pages. Steffensen, J., P. Tidemand and E. Mwaimpopo (2004): A comparative analysis of decentralization in Kenya, Tanzania and Uganda. NCG, Tax, S. (2000). The Social Consequences of Structural Adjustment Reforms in Tanzania. The Tanzanet-Journal, 1(1): 4-18 [http://www.Tanzanet.org/vol1-No1.html]. THA (2004): Annual report 2002/2003. Dar es Salaam: Tanzania Harbors Authority. THA (2004): Ports Handbook 2004/5. Dar es Salaam: Tanzania Harbors Authority. Tripp, A. M. (1997): Changing the rules, The politics of liberalization and the urban informal economy in Tanzania. Berkeley: University of California Press. UNCHS (1985), The role of small and intermediate settlements in national development. Nairobi: UN. UNFP (2003): World Population report 2003. New York: UN. URT (2002): Population and housing census, Analytical report. Dar es Salaam: NBS. 101 URT (2003): Economic survey 2002. Dar es Salaam: NBS. URT (2004): Profile Arusha, draft 138 pages. Dar es Salaam: NBS. URT (2004): Revenue statistics for local government authorities on mainland Tanzania, 20 pages. Dar es Salaam: NBS. URT (2004): Annual establishments survey Dar es Salaam: NBS. URT (2004): Medium term plan and budget June 2005-08, June 107 pages. Dar es Salaam: NBS. URT (2005): Economic survey 2004, draft, 121 pages. Dar es Salaam: NBS. Van Eijck, J. (2006), Transition towards Jatropha biofuels in Tanzania? An analysis with strategic niche management. Eindhoven: TU. Visser, E.J. and M.P. van Dijk (2006): Economic globalization and workers: introduction, In: Tijdschrift voor Economische en Sociale Geografie, dossier Globalization and labor, Vol. 97, No. 5, pp. 461-467. Wallerstein (1989): The world economic system. New York: Academic Press. Wangwe, S. (2005): Third phase and its national legacy. Dar es Salaam: Conference at the University, July, 14 pages. Water for Africa (2004): Dar es Salaam Community managed water and sanitation program, Case study. World Bank (1993): The East Asian miracle. Washington: World Bank, New York: Oxford. World Bank (2002): Tanzania at the turn of the century. Washington: IBRD. World Bank (2004a): Project appraisal document for the Local government support project. Washington: IBRD. World Bank (2004b): Investment climate assessment: Improving enterprise performance and growth in Tanzania. 97 pages. World Bank (2005): Tanzania, Country economic memorandum and poverty assessment, preliminary draft, June 30, 184 pages. World Bank (2006a), Water resources management. Dar es Salaam: World Bank office. World Bank (2006b), Transport and poverty. Dar es Salaam: World Bank office. World Bank (2006c), Regional and global integration. Dar es Salaam: World Bank office. World Bank, MIT, WTO and UNCTAD (2005): Diagnostic Trade Integration Study, Volume I. Dar es Salaam. WTO (2005): Trade policy review, Tanzania, February 2000, 14 pages internet www.wto.org/english/ tratop_e/tpr_e/tp128_e.htm. Wuyts, M. (1998): Informal economy, wage goods and the changing patterns of accumulation. Dar es Salaam: ESRF. Annex 1 Political developments and changing economic policies in Tanzania Since independence 1961 up to 1990 the head of state in Tanzania was elected under the single political party. Julius Nyerere's resignation as the chairman of leading political party the CCM in August 1990 changed the Tanzanian political agenda (Kahama, 1999). From 1995 onwards the country has held multiparty elections. The current president Benjamin William Mkapa (also CCM) has won two consecutive elections. Any president in power has a maximum of ten years in the office if selected. Multiparty elections have broadened democracy and good governance in Tanzania. 102 Democratic governance is a precondition for sustained pro-poor economic growth. The government has embarked upon on ambitious and far-reaching reform programs since 1972. Recent ones being the introduction of multi-party political system, the establishment of the Financial Sector Reform Program (FSRP), the Parastatal Sector Reform Program (PSRG), the Civil Service Reform Program (CSRP), the creation of the Tanzania Revenue Authority (TRA) and the introduction of the Value Added Tax. Others were the introduction of the Local Government Reform Program (LGRP) and the concept of good governance. Between 1973 and 1976 intensive implementation of villagesation program was undertaken. This program was accompanied by the increased use of by-laws to regulate agricultural practices and other aspects of rural life. For instance, in 1973 new agricultural producer price policies were introduced and in 1976, the government banned the cooperative unions. The activities under jurisdictions of cooperative unions were shifted to the established agricultural and marketing system in which each registered village acted as primary society while crop authorities were given sole responsibility for crop purchase, processing and sales. In 1991, the government launched the Financial Sector Reform Program (FSRP). The financial sector inter-mediation is one among the prerequisite for economic growth and development since it facilitates efficient mobilization and channeling of savings into investment. History shows that during early 1970s, the organized financial system in Tanzania was narrow and comprised only 12 institutions, besides the central bank. Consequently, the Banking and Financial Institutions Act 1991 opened up the financial sector to private enterprise. The Act paved the way for the establishment of private domestic and foreign banks. To a large extent, the Act also removed the barriers to competition and ensured that market forces restore efficiency in the mobilization and allocation of domestic resource. These changes increased the number of financial institutions drastically. Currently the number of commercial banks and non-banks operating in Tanzania are 23 and 11 respectively. In 1992 the government launched the Parastatal Sector Reform Program (PSRP) as part of sector reform initiatives. Historically much of the public sector in Tanzania was established after the Arusha Declaration. The sector was expanded with a view to enabling the government to control the modern economy, promote growth and provide financial resources to the government in the form of surpluses. However, the expected benefits were not realized as planned due to, inadequate capital, lack of foreign exchange, inappropriate technologies, lack of competent technological skills, and government interference in management. As a consequence, the main objective of PSRP was to privatize all the ailing public enterprises that were draining government resources through subsidies. Moreover, the reform entails divestiture of most of the state enterprises, liquidation of those deemed unviable and fully commercialization and the introduction of large element of competition for those that remain under public ownership. 103 The available statistics show that the PSRP has achieved remarkable progress since its inception. For example, up to year 2003, a total of 289 enterprises were privatized and 328 non-core assets were sold (URT 2004). Furthermore, it has been observed that, a substantial number of these enterprises have been refurbished and are currently operating efficiently, providing goods and services to the public satisfactorily and making significant contribution to the economy. In 1994 the Civil Service Reform Program (CSRP), was adopted, with the prime objectives of re-defining the roles of government machinery. In a nutshell, the CSRP entailed a review and restructuring of various ministries/departments of the government together with a redefinition of their functions, retrenchment and redeployment of staff, review and reform of the remuneration system, and development of the human resources in the civil service to improve performance. Remarkable achievements have already been realized under CSRP. For instance, it was found that the reform has already articulated a short-term action program for improving the personnel information system and ensuring effective control on personnel numbers (Mtatifikolo, 1994). In additional, updating of the computerized personnel records system; a comprehensive review of staffs training need; and implementation of personnel control policies are some of the ongoing activities under CSRP. In 1996 Tanzania Revenue Authority (TRA) was established as part of fairly comprehensive tax reforms that started since 1985. The Tanzania Revenue Authority (TRA) has focused on revenue agenda of simplifying tax structure and streamlining transparency in revenue collection. Within the same period cash budgeting was introduced. This enforced fiscal discipline by binding government to a balanced domestic budget and limiting growth in the monetary base. Whether cash budgeting has positive or negative effect upon economic growth is debatable. For instance, it has been observed that cash budgeting eliminate deficits but at the expense of greatly reduced levels of expenditures and predictability of resources flows. In 1998 the government introduced Value Added Tax (VAT) that forms one of the tax policy reform components. The VAT system replaced the sales tax and is a consumption tax charged on all taxable goods and services. It is a multistage tax levied at each stage of production and distribution up to the retail stage. For the international merchandise trade, it is levied on taxable imports from, persons whether registered or not registered for VAT. It has been argued that VAT system has many benefits compared to sales tax. Despite the benefits envisaged to VAT it has been observed that for the benefits to be realized it requires among other things institutional support and political commitment. In the case of Tanzania it appears that the two pre-requisite requirements for prosperity of VAT are gradually improving. For instance, World Bank (2001) observed that the introduction of VAT in Tanzania has helped to remove the overall distortionary character of the tax regime and raise its buoyancy. In 1999 the Local Government Reform Program (LGRP) was launched. One of its main objectives is to foster governance by enabling local authorities to find local solutions to local problems, involving local people more in the policy process, and by making local 104 authorities more directly accountable to their people. The LGRP plans show that when fully implemented, city, municipal, town and district councils will be democratically elected and fully responsible for service delivery in education, health, water, roads, agriculture; social development; and maintenance of law and order. Under this reform, central government will gradually move away its implementation responsibilities and play the role of auditor. This shift in function has already resulted in a reduction in regional staff of central government from 700 to 80 per region (Naschold and Fozzard, 2002). In 1999 good governance philosophy was adopted and as a result it has become one of the important components in the government machinery. Governance is the manner in which a government or state governs its territory and the people which it judicially controls. It encompasses the state's institutional and structural arrangements, decision- making processes, the implementation capacity, and the relationship between government officials and the public. Among the attributes of good governance include accountability based on the ideal of popular sovereignty and public choice; popular participation in the decision-making process based on political and social pluralism. It has been observed that in Tanzania (recent years), governance has come under severe-pressure because of creeping corruption, poor accountability, overloaded and inefficient legal system ambivalence in sanctifying the fundamental human rights and erosion of meritocracy in public service. Following this situation in November 1999 the national Anti-Corruption Strategy and Action Plan (ACSAP) was adopted. In order to mitigate corruption the government has of recent attempted to address some of the measures to strengthen good governance. Some of key measures undertaken include introduction of multi-party political system, establishment of the TRA to administer revenue collection, and the establishment of a Parastatal Sector Reform Commission (PSRC) to manage the privatization process. Others were the introduction of the Integrated Financial Management System (IFMS), regional sub-treasurer, and restructuring the government machinery through a Civil Service Reform Commission (CSRC). The government continued to concentrate on infrastructure, health and education, macroeconomic management and the provisional of information, forecasts, advice, research and development, in an effort to provide the most conducive atmosphere for private enterprise expansion and for working closely with the private sector (in partnership). Also the government has jurisdictions intervention to deal with monopolies and situations where social costs are high (such as with environmental problems of pollution and soil conservation), where the distribution of income is socially unacceptable (such as vulnerable groups-sick, handicapped, the aged, women and children), and dealing with external economic and diplomatic relations. From 1961 to date various sources show that Tanzania's experience with economic policies varies significantly. During the first two decades of independence (1961-1970s), Three-Year Plan (TYP) and Five-Year Plans (FYPs) guided the economic policy and planning. Moreover in 1972 the Annual Development Plans (ADEPs) were introduced in 105 order to strengthening FYPs. However, in late 1970s FYPs were shelved and emergency plans/strategies were implemented during 1980-1989. In 1992 and 1997 Rolling Plan and Forward Budget (PFB) and Medium Term Expenditure Framework (MTEF) were adopted respectively. Basically since independence to date the economic policies so far implemented can be grouped into five broad categories as follow. Post-independent period (1961-66) This was the period where the majorities of economic activities were privately owned except the main utilities such as water, electricity, post and telecommunication. During the period various strategies were adopted, but among the growth strategies specified were the transformation and improvement approaches. Generally, emphasis was placed on modernizing the economy and achieving structural transformation. As a result, investments were directed at developing economic infrastructure, expanding the capacity of the manufacturing sector, and expanding output from sectors, which were producing primary commodities such as cotton, coffee, sisal, tobacco etc. Credit provision, market outlets, and cash incentives to farmers were among of strategies adopted. Pre-economic crisis period (1967-79) This was the period where several clear articulations of policies under the auspices of Arusha Declaration were witnessed. However, the major economic policy thrust was less reliance on the external economy with the state took control of the commanding heights of the economy. In order to achieve the above objectives the main growth strategies specified were self-reliance, rural development, decentralization, villagisation, basic industry strategy and small-scale industry strategy. Also much greater emphasis was placed on long-term oriented investment activities with long gestation periods and widely diffused returns. The constructions of TAZARA railway line, TAZAMA pipeline and Tanzania highway in 1974 were among of the long-term investments. Economic crisis period (1980-85) During 1980-85 the economic policies adopted were dominated by responses to emerging resources gaps. The most prominent economic policies were envisaged in the National Economic Survival Program (NESP), and Structural Adjustment Program (SAP). The NESP was a short-term; crisis alleviation approach based on the internal mobilization and utilization of resources. The noted prominent strategies under NESP were the reviving of various export commodities, and controlling recurrent expenditure for Ministries, Departments, Agencies (MDAs) and Regions. SAP on the other hand was more articulated than NESP and included plans, which addressed strategies to improve foreign exchange earnings. For instance, the numbers of import controls were lifted and `'own- fund'' imports were instituted, plus trade liberalization and devaluation of shilling by almost 40 percent in 1984. In addition, imports duties and sales tax were reduced considerably. Economic reform period (1986-95) This was the period when formal agreement was made by Tanzania government with the International Monetary Fund (IMF) and the World Bank (WB). The economic policies adopted over the period were envisaged into Economic Recovery Program (ERP I & II) 106 and Economic Social Adjustment program (ESAP). To a large extent ERP-1&II program was a continuation and intensification of strategies introduced during SAP. In spite of been influenced by donors ideas, its design and intentions were mainly focused on market oriented economic reforms, trade liberalization, privatization and reduction of government interventions. Among of the major strategies employed includes, exchange rates and interest rates adjustment, tight fiscal and monetary policies, decontrol of prices, and liberalization of domestic and foreign trade. On the other hand, ESAP was adopted with a concern to correct the adverse impact of ERP I and II on the vulnerable groups in the society. Post economic reform period (1996 to date) Starting from 1996 to date, the major focus of economic policy has been the consolidation of previous recovery efforts. Among the prominent strategies under ESAP includes fiscal management and structural institutional issues, deregulating investments, divesting public enterprises, establishing free resources and product markets, and restructuring of the financial sector. In additional, various governance and public administration, private sector development and public expenditure reforms have continued to be reviewed and implemented simultaneously by the government. My suggestion that local governments should not just focus on increasing revenues, but should think what to do with the money and how to attract private finance was met with three remarks: 1. It seems the MOF is already thinking about introducing infrastructure bonds, but it is not yet mentioned for the local authorities. The local authorities later told me that this is something they can do in five or ten years time! 2. The local authorities are facing the problem that a number of their sources of revenue have been abolished by the national authorities! 3. Finally, local governments had been abolished in Tanzania for a number of years and it takes some time to really get them back on the map. However, it would be good to point to the important role they can play. Moreover, there has been a specific focus on poverty reduction as a key component of the development agenda in Tanzania under Poverty Reduction Strategy (PRS) and the current one the National Strategy for Growth and Poverty Reduction (NSGPR). This suggests a major point of departure from the earlier practice where poverty reduction was assumed to be an automatic result of growth and good economic management. The objectives of ESAP are also strongly backed by the Highly Indebted Poor Countries (HIPC) initiatives of channeling debt relief funds to priority sectors and activities for poverty reduction. However, it appears that if the PRS cum NSGPR and HIPC initiatives have to be effective, major constraints within and among all sectors should be well addressed and articulated. The government has continued to implement with notable success, various economic policies and reforms in order to achieved stated targets as envisaged into the National Development Vision 2025. Apart from various sector's policies, the government continued to put in place a conducive environment for private sector development. The 107 current National Strategy for Growth and Poverty Reduction (NSGRP) also emphasizes the need to achieve and sustaining broad-based and equitable growth. The strategy focuses on scaling up investments towards modernizing small, medium and large-scale agriculture for increased productivity and profitability, promoting off-farm activities including small and medium size enterprises with particular emphasis on agro- processing. Other policies include a National Trade Policy and a small and Medium enterprise development policy, private sector strategy etc. Notes i The president addressed the meeting for two hours suggesting a code of conduct, which made it very explicit that a lot is expected from these regional leaders. Before the meeting a large number of regional leaders had been replaced. It is quite remarkable that the president has retired or kicked out something like 31 regional commissioners, because they did not do what they were supposed to do! Many were not clean and corruption gets a lot of attention. ii The plans for one local tax has not yet been accepted. iiiPrepared with the help of the Economic and Social Research Forum (ESRF). iv We should stresses the limitations of the poverty is one dollar a day concept. v Between 1988 and 2002 according to NBS(2002c). vi Population growth is not an issue addressed in Tanzania, given the role of the Catholic Church and Islam in the country and the current importance of AIDS. vii One could take any supply chains and find out where the money is staying! Besides looking at the income one could also look at the expenditure side: where is the money spent? viiiSteffensen (et al., 2004) focused on the effects on local service delivery and which factors constrain the development of a more effective and efficient functioning system of decentralization. ix Parker (1995) proposes normative criteria for analyzing development outcomes of decentralization, including: (i) effectiveness, that is, providing minimum standards of service delivery cost- effectively, and targeted toward disadvantaged groups; (ii) responsiveness of decentralized institutions to the demands of local communities, at the same time as meeting the aims of broader public policy; and (iii) sustainability, as indicated by political stability, fiscal adequacy and institutional flexibility. x A hypothesis which can not yet be tested is that the approach to decentralization would show that when pluralism emerges at decentralized levels of governance it yields better results than in case of distributed monopoly. xi The plans for one local tax has not yet been accepted. xii Functions that can be carried out by lower levels of government are planning, fiscal policies and revenue generation, budget generation and expenditure management, staffing, program and project implementation, information management and operations and maintenance. xiiiFor the Netherlands development cooperation good governance is also measured by efficiency of ministries and their `administrative capacity'. In the education sector up to 50 percent of the means provided could not be traced. This doesn't have to be corruption it can also be this lack of `administrative capacity'. Enrolment has substantially improved, but the quality of education is still worrying. In particular the quality of education for young girls is very poor. xiv For example, the Netherlands is heavily involved in decentralization and local government reform. Together with other donors it plans to participate in a new water project. In that case Dutch development cooperation money will go to the local governments who have to spend it. xv Publications explain how the councils are assessed (LGCDGS, 2005) and provide the Planning guidelines for village and Mitaa (LGCDGS, 2004). xvi It is quite remarkable that the president has retired or kicked out something like 31 regional commissioners, because they did not do what they were supposed to do! Many were not clean and corruption gets a lot of attention. xvii Besides donor money, local governments can also use the Tanzanian Social Action Fund (TASAF). xviiiOther authors give a different interpretation local economic development (Helmsing, 2005 in Egziabher and Helmsing, eds, 2005). 108 xix For example in VEETA, MPT MSM , UDEC University of Dar es Salaam with Donald Alomi (see article in newspaper), Zumba, with university of Amsterdam and via the PSOM project. The Netherlands Embassy also participates in BEST and the Financial deepening program xx President Mkapa had brought in De Soto in his last year and the Norwegians financed him. They are as non EU not in Best and now the business formalization approach is overlapping! xxi Cosmopolitan refers to the life style that tends to come with such metropolitan areas. xxiiThe movement of the headquarter of Philips electronic company from Eindhoven where the company started to Amsterdam in the nineties perfectly illustrates this trend to be where knowledge activities such as research and development, finance, marketing and design activities are concentrated. xxiii Data on small enterprises in Tanzania were generated under the World Bank initiated REP studies. Later the University of Oxford has collected additional data. xxiv The Guardian (16-7-2005) reported that the TRA impounded sugar worth 6 million TSh. coming from Malawi in July 2005. xxv There is a loan guarantee project, which the banks use, but the amount for SMEs in the project is only 2 million dollar, which bankers consider a joke. 200 to 300 million would be needed! xxvi In term of ethnicity groups, Zaramo and a few other tribes especially Ndengereko and Kwere originally dominated the City. However, due to urbanization many people of different ethnicity and origins have immigrated to the city in big numbers. This has caused the undefined cultural change. xxvii The latter developed in the framework of the European Union Economic Partnership Agreements or EPAs and by Tanzania's participation in the Doha or Millennium Development round. xxviiiDar es Salaam counts about ten percent of the population, but generates between 17 and 18 percent of GDP. xxix World Bank (2006b) provides an overview of the transport infrastructure and its management. 109 Growth, Inequality and Simulated Poverty Paths for Tanzania, 1992-2002 Gabriel Demombynes and Johannes G. Hoogeveen* Abstract ­ Although Tanzania experienced relatively rapid growth in per capita GDP in the 1995- 2001 period, household budget survey (HBS) data shows only a modest and statistically insignificant decline in poverty between 1992 and 2001. To assess the likely trajectory of poverty rates over the course of the period, changes in poverty are simulated using unit-record HBS data and national accounts growth rates under varying assumptions for growth rates and inequality changes. To this end the projection approach of Datt and Walker (2002) is used along with an extension that is better suited to taking into account distributional changes observed between the two household surveys. The simulations suggest that following increases in poverty during the economic slowdown of the early 1990s, recent growth in Tanzania has brought a decline in poverty, particularly in urban areas. Unless recent growth is sustained, the country will not meet its 2015 Millennium Development Goal (MDG). Poverty reduction is on track in urban areas, but reaching the MDG target for bringing down poverty in rural areas, where most Tanzanians live, requires sustaining high growth in rural output per capita. * Demombynes is a consultant to the World Bank. Hoogeveen is with the World Bank. Please send correspondence to both gabriel@demog.berkeley.edu and jhoogeveen@worldbank.org. World Bank Policy Research Working Paper 3432, October 2004 The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the view of the World Bank, its Executive Directors, or the countries they represent. Policy Research Working Papers are available online at http://econ.worldbank.org. This paper was written in preparation for the Tanzania Country Economic Memorandum. Robert Utz was instrumental in starting the paper and his support and feedback are gratefully acknowledged. The authors thank Emmanuel Mungunasi for collecting the data on regional growth rates and Trudy Owens and Patrick Ward for assistance with the consumption aggregates for 1991/92 and 2000/01. Helpful comments and suggestions were also provided by Louise Fox, Tim Harris, David Sahn, Erik Torbecke, Stephen Younger, participants at the PADI workshop in Dar es Salaam in February 2004, and various members of the Research and Analysis Working Group. 1 1 Introduction Following a period of economic stagnation in the early 1990s, Tanzania has seen sustained gains in per capita output since 1995. The 2000/01 Household Budget Survey (HBS) offers the opportunity to assess how economic growth at the national level has impacted household consumption and poverty levels since the 1991/92 HBS. At the national level, the two surveys show only a small and statistically insignificant decline in the headcount rate. This fact, combined with the relatively rapid growth of recent years, has raised concerns that recent economic growth may not be substantially reducing poverty. The survey data, however, provides only two snapshots in time and fails to represent the full evolution of poverty over the course of the intervening nine years. The primary purpose of this paper is to assess the likely trajectory of poverty rates over the full span of the period between the surveys. This is done by applying macroeconomic growth data to the micro-level household survey data. Changes in consumption are simulated year-by-year using unit-record survey data, under varying assumptions for growth rates and inequality changes. The growth data is drawn from national accounts statistics. The analysis follows the procedure outlined in Datt and Walker (2002) and Datt et al. (2003). The paper also implements an extension to the Datt and Walker method. The original procedure was designed to project poverty rates forward from a single survey. When the task is to estimate the trajectory of poverty rates during the period between two surveys, as with Tanzania for the 1990s, an alternative method can be used. The method involves scaling national accounts growth rates for multiple parts of the distribution. Unlike the Datt and Walker method, this approach guarantees that the simulated distribution will closely match the distribution in the final survey year. The paper is structured as follows. Section 2 describes the data sources, key findings from the Tanzania Household Budget Survey report, and issues surrounding the data. Section 3 presents some poverty and growth diagnostics: a decomposition of growth and inequality, a sectoral decomposition of poverty changes, and growth incidence curves. Section 4 outlines the simulation methodology--both the Datt and Walker method and the alternative approach. Section 5 presents the simulation analysis. The section investigates different assumptions regarding the data and the method of simulation. Our preferred scenario is presented in Section 5.7 after which Section 5.8 considers the main assumption underlying the Datt-Walker approach: that economic growth translates into higher consumption for all population groups. This is done by investigating whether since 1994, and for given GDP growth, inequality changed rather than that poverty dropped. Section 5.9 presents MDG projections, and Section 6 concludes. 2 Data Sources 2.1 Household Budget Surveys The primary sources of poverty data for Tanzania are the Household Budget Surveys conducted in 1991/92 and 2000/01. Basic poverty statistics calculated from the survey data are shown in Table 1. In Tanzania1 as a whole, in rural areas, and in urban areas other than the capital, the household survey data shows a statistically insignificant growth in consumption per adult equivalent of 6.4%, with small and insignificant drops in the fraction living in poverty, and no 1When we refer in this paper to Tanzania as a whole we only refer to mainland Tanzania, i.e. the United Republic of Tanzania, excluding Zanzibar. 2 change in inequality. In Dar es Salaam, per capita consumption increased by 43%, headcount poverty dropped by a third, and inequality as measured by the Gini coefficient jumped from 0.30 to 0.34. All the changes in Dar are significant at the 95% confidence level. Table 1: Consumption Data from Household Budget Survey Data Other Mainland Dar es urban Rural Tanzania Salaam areas areas % Below Basic Needs Pov. Line 1991/92 38.6 28.1 28.7 40.8 (2.1) (2.8) (5.0) (2.4) 2000/01 35.6 17.6 25.8 38.6 (1.6) (2.7) (2.2) (2.0) Consumption per Adult Equivalent Mean, 1991/92 10223 11161 12445 9820 (290) (410) (842) (326) Mean, 2000/01 10880 15944 13536 10060 (250) (779) (487) (273) Ratio (00/01) to (91/92) 1.064 1.429 1.088 1.024 Gini Coefficient 1991/92 0.33 0.30 0.34 0.33 (0.01) (0.01) (0.02) (0.02) 2000/01 0.34 0.34 0.35 0.32 (0.01) (0.01) (0.01) (0.01) Notes: All figures shown are as calculated from Household Budget Survey data. Standard errors are given in parentheses. All figures were calculated from household-level data on a per adult equivalent basis, with weights calculated by multiplying household size by household sampling weights. 1991/92 figures were converted to 2000/01 Shillings by multiplying by 2.611811, which is the ratio of the poverty lines used to calculate the poverty levels, using nominal values in the official poverty report. The poverty statistics calculated for this paper differ slightly from those in the official published report, Household Budget Survey 2000/01 (United Republic of Tanzania, National Bureau of Statistics, 2002). These are reproduced in Appendix Table 1. While the headcount rate figures are identical to those that were calculated for this paper, the published mean consumption figures differ slightly from those calculated for Table 1. This is partially because the published figures were calculated on a per capita (rather than per adult equivalent) basis.2 2Also, for this analysis, the 1991/92 survey figures were converted to 2001/02 Shilling figures by multiplying by 2.611811, the ratio of the (Basic Needs) poverty lines from the two years. In the official figures, a price adjustment of 2.49 was used. The official figures and the analysis here also employ regional price adjustments. This combination of price adjustments may have been implemented differently for the official mean consumption figures than is done here. The report indicates explicitly that the mean consumption figures were calculated on a per capita basis, rather than the 3 There are three caveats associated with the figures for Dar es Salaam. First, the HBS report says "Note that there is some evidence that consumption expenditure was under-reported in Dar es Salaam in the 1991/92 HBS. This would mean that poverty levels may in fact have been slightly lower in 1991/92 and the decline smaller .... Although [an earlier report] attempted to adjust for this under-reporting in the 1991/92 data, this was not repeated in this analysis as it was difficult to assess its accuracy" (p. 80, footnote 21). A second concern stems from the fact that both the 1991/92 and 2000/01 surveys used a sampling frame based on the 1988 national population census. Between 1988 and 2002, the population of Dar es Salaam grew extremely rapidly, at an annual growth rate of 4.4 percent, for a cumulative increase of 84 percent. Given such a high rate of population growth, it is possible that by the time of the second HBS survey, the true geographic distribution of the population differed substantially from that in the census-based sampling frame. In particular, it is likely that new settlements were created in areas of the city that were not populated or only sparsely populated in 1988. Consequently, households in such areas might not have been included in the survey sampling frame, or included with only a very low sampling probability. If households in new settlements are poorer than average households in the city then the drift of the population from the 1988 sampling frame biases the 2000/01 poverty figures downward from their true values. Both this phenomenon and the possible underreporting of consumption in the 1991/92 HBS imply that the drop in poverty in Dar es Salaam may not have been as steep as it appears in the survey data.3 Finally, a related weighting issue has implications for the national poverty figures. Because the 2000/01 survey sampling frame was based on the 1988 census, the sampling weights understate the relative weight of areas like Dar es Salaam that experienced rapid population growth. The overall effect of this phenomenon is small. Poverty calculations done after reweighting the survey data to reflect the regional population distribution in the 2002 census show estimated national poverty incidence to be 35.3%, versus 35.6% using the original weights. This difference can be attributed entirely to the greater weight accorded Dar es Salaam with the revised weights. Using population estimates from the 2002 census, 7.4% of the population lives in Dar es Salaam as opposed to 5.8% when population numbers from the 1988 census are used. The difference between the estimates is not due to changes in poverty incidence within the three strata. Reweighting has no effect on poverty estimates for Dar es Salaam and rural areas as a whole, and in other urban areas it increases the estimate negligibly, from 25.8% to 25.9%. To maintain comparability with the official poverty statistics, in the remainder of the paper we calculate poverty rates using the official (non-reweighted) sampling weights. Only in section 3.2, table 4, where we decompose changes in poverty into changes by sector and population shifts across sectors, do we make use of reweighted survey weights. 2.2 National Accounts The other main source of economic information for Tanzania is the national accounts information, tabulated in the Economic Survey 2002 (United Republic of Tanzania, Office of the President, 2002). Key data drawn from the report is shown in Table 2. adult equivalent basis used for the poverty figures. It is also possible that a slightly different consumption aggregate than that employed here was used to calculate mean consumption. 3It would be possible to examine the sampling frame issue in more detail by comparing the distributions of household characteristics in the 2002 census and the 2000/01 survey for Dar es Salaam. If the distributions match closely, drift in the sampling frame is unlikely to have had a large effect on the poverty figures. If the distributions do not match, the problem could be addressed by reweighting the survey data to match the distribution of characteristics in the census data. The issue of reweighting for Dar es Salaam will be addressed as part of the ongoing poverty mapping project. 4 Several points are evident from the national account data. First, during the period between the household surveys (taken as 1993-2001), cumulative growth in per capita GDP was only 6.1%, despite the sustained gains during the final years of the period. This is because the country faced substantial declines in per capita output at the beginning of the period. Second, output in the urban sector grew slightly more rapidly than in the rural sector over the full span of the period, and at the end of the period, urban growth substantially outpaced rural growth. Third, year-to-year growth rates in household consumption per capita differed from those for GDP per capita, but overall growth in household per capita consumption for the period was just slightly greater, at 8.3%. Table 2: Growth Data from National Accounts Per Capita Rural Growth Growth Per Capita Hhold. Rate of Per (monetary Urban Cons. Implicit Per GDP Capita + non- Growth (all as GDP Capita Growth GDP monetary non- Fraction Price Hhold. Year Rate Growth agriculture) agriculture) of GDP Deflator Cons. 1990 6.2% 3.3% 0.90 0.62 1991 2.8% -0.1% 0.7% -0.9% 0.90 0.79 -1.0% 1992 1.8% -1.1% -1.7% -0.5% 0.89 1 -2.0% 1993 0.4% -2.5% 0.2% -5.0% 0.90 1.26 -1.3% 1994 1.4% -1.5% -0.8% -2.2% 0.91 1.64 -0.4% 1995 3.6% 0.7% 2.9% -1.6% 0.91 2.08 0.3% 1996 4.2% 1.3% 1.0% 1.6% 0.91 2.46 1.4% 1997 3.3% 0.4% -0.5% 1.3% 0.93 2.96 2.5% 1998 4.0% 1.1% -1.0% 3.2% 0.96 3.40 4.6% 1999 4.7% 1.8% 1.2% 2.4% 0.95 3.79 0.8% 2000 4.9% 2.0% 0.5% 3.4% 0.91 4.05 -2.6% 2001 5.7% 2.8% 2.6% 3.1% 0.91 4.36 2.9% 2002 6.2% 3.3% 2.1% 4.4% 0.87 4.65 -0.9% 1993-2001: Average 3.6% 0.7% 0.7% 0.7% 0.9% growth rate Cumulative 37.1% 6.1% 6.2% 6.0% 8.3% change Source: Economic Survey 2002, United Republic of Tanzania Notes: This paper examines growth between the years 1992 and 2001, which is calculated as the cumulative effect of growth rates for the years 1993-2001. GDP growth figures are taken from Table 4A, p.16. Per capita GDP growth rates were calculated by subtracting 2.9%, the average population growth rate between the 1988 and 2002 censuses. Household consumption as a fraction of GDP was calculated from figures in Table 2B, using GDP at factor cost prices. The implicit GDP price deflator was calculated by dividing nominal GDP at factor cost prices (Table 2B) by GDP in 1992 prices (Table 3). Separate rural and urban per capita growth rates were calculated using the growth rate and economic composition figures by sector and the national rate of population increase (this latter assumption is relaxed in Section 5.4). Because rural non-agricultural output is not separately identified in the national accounts, the rural GDP growth rate figures are based on the agriculture sector alone (monetary and non-monetary) and do not include growth in the non-agricultural sector. A substantial fraction of income in the rural area is earned from non-farm sources. According to the HBS report (Table 9.2), the principal sources of rural income are: 60.4% from agriculture, 17.8% from non-farm self employment, 8.3% from employment, and 12.8% from transfers and other receipts. During the period of study, growth in the non- agricultural sector exceeded growth in the agricultural sector but only by 0.1% point so that rural growth may be only slightly underestimated. 5 3 Poverty and Growth Diagnostics 3.1 Decomposition Analysis of Growth and Inequality A useful way to take a first look at the impact of growth on poverty is by decomposing the change in the headcount rate. In general, a change in poverty can be attributed to the interaction of two processes -- growth in mean consumption and a change in consumption inequality. The formal decomposition proposed by Datt and Ravallion (1992) attributes changes in poverty to a growth effect, an inequality effect, and an interaction effect (the residual). Results from this decomposition applied to the Tanzania data are shown in Table 3. The decomposition shows that overall growth reduced poverty in the country as whole by 4.6 percentage points. The slight increase in inequality, however, increased poverty by 1.1 percentage points. The same general pattern holds for rural areas and urban areas other than Dar es Salaam. The results are more notable for Dar, where the substantial decline in poverty was attributable to the city's economic growth. Holding inequality constant, growth reduced poverty by 16.3 percentage points. The growth-induced decline in poverty was partially countered by increasing inequality, which drove up poverty by 9.8 percentage points. Table 3: Growth and Inequality Poverty Decomposition Other Mainland Dar es urban Rural Tanzania Salaam areas areas Poverty Rate in 1991/92 38.6 28.1 28.7 40.8 Poverty Rate in 2000/01 35.6 17.6 25.8 38.6 Change in Poverty -3.0 -10.5 -2.9 -2.2 Breakdown in levels Growth Component -4.6 -16.3 -4.4 -2.6 Redistribution Component 1.1 9.8 1.5 -0.4 Residual 0.6 -4.0 0.0 0.9 Breakdown in percentages Growth Component 155% 156% 154% 119% Redistribution Component -35% -93% -53% 20% Residual -20% 38% -1% -39% Notes: Decompositions were calculated using the approach of Datt and Ravallion (1992). The analysis shown here uses 1991/92 as the base year for the decomposition. 3.2 Sectoral Decomposition of Changes in Poverty Another way to break down the overall change in poverty is by considering the contribution of changes in poverty in each sector to the national change. Following Huppi and Ravallion (1991), the national change in the headcount rate can be attributed to a combination of changes within each sector and shifts in population between sectors. There is also a small portion of the change that is due to an interaction effect between intra-sector changes and population shifts. Results from a decomposition for Tanzania are shown in Table 4. A somewhat surprising point is that while poverty declined much more in Dar es Salaam (from 28.1 to 17.6 percent), most of the small drop in the national poverty rate was due to a modest decline in poverty in rural areas. 6 This is due to the fact that only a small fraction of the population lives in Dar.4 Consequently, even the large drop in poverty in Dar made only a small dent in the national poverty rate, while the small drop in the rural poverty rate had a relatively large effect on the national poverty rate. The decomposition also shows that just 11.6 percent of the decline in the headcount at the national level is attributable to the shift of the population from poorer rural areas to wealthier urban areas, chiefly Dar es Salaam. Most of this shift is probably due to migration. The small size of this "population-shift effect" demonstrates that rural-urban migration cannot substantially reduce poverty in Tanzania in the short run. Because the population is overwhelmingly rural, even the rapid shift of population to wealthier Dar, which saw its population grow at a 4.4 percent annual rate versus 2.9 percent for the country as a whole, did not substantially reduce the national poverty rate. Table 4: Sectoral Decomposition of the Change in Poverty Contribution to Change in National Headcount Rate Population Percentage Share in Absolute of Total 1991/92 Change Change Dar es Salaam 5.35 -0.56 17.08 Other Urban Areas 12.6 -0.34 10.49 Rural Areas 82.06 -1.82 55.41 Total Intra-sector Change -2.72 82.98 Population-Shift Effect -0.38 11.60 Interaction Effect -0.18 5.41 Change in Poverty -3.28 100 The sectoral decomposition was calculated using both the 1991/92 and 2000/01 HBS data. For this analysis, in order to ensure that the decomposition properly accounted for changes in population shares, the 2000/01 data was reweighted to reflect the regional population distribution reported in the 2002 national census. As a result, the decline in the headcount rate shown here, 3.28, is slightly greater than the decline of 3.0 shown in Table 1 (from 38.6 to 35.6), based on calculations using the original weights. Altogether, 55 percent of poverty reduction nationally was attributable to poverty declines in rural areas, and 39 percent was due to population shifts and declines in urban areas, with the remainder accounted for by the interaction effect. What does this suggest about the potential sources of poverty reduction for Tanzania in the future? Because both urban poverty reduction and urban-rural migration were already rapid during the 1990s, it is very unlikely that either can be an increased source of poverty reduction in the near future. This is especially the case because the headcount rate in Dar is now low enough, at 17.6%, that the potential for reducing it further is limited. Consequently, accelerating the rate of decline of poverty can only be accomplished by increasing poverty reduction in rural areas. While migration and urban growth can play secondary 4 Table 4 shows 5.35 of the population percent living in Dar. This figure is based on the weights in the 1991/92 HBS. Census data, which provides more definitive population information, show that the fraction living in Dar was 5.8 percent in 1988 and 7.4 percent in 2002. 7 roles, much more than 55 percent of poverty reduction will have to come from rural growth if Tanzania is to make substantial progress in bringing down poverty for the nation as a whole. 3.3 Growth Incidence Curves The impact of growth on poverty can be examined graphically through a growth incidence curve (GIC), which illustrates the distribution of growth. A GIC is a plot of the growth rate for each quantile of the distribution of per adult equivalent consumption. Growth incidence curves for the nation as a whole and the three strata are shown in Figure 1. The curves were constructed based on the two household surveys, and the annual growth rates reflect average changes over the 1993-2001 period. The growth incidence curves were generated by splitting the distributions into 15 quantiles or bins. Although the vertical scale is the same across the different curves, the horizontal scale plots percentiles within each sector, so points at the same percentile level for different sectors correspond to different levels of consumption. The vertical line in the graph indicates the poverty headcount index in 1992, while the horizontal line denotes the mean overall growth rate. We consider growth "absolutely pro-poor" if the mean growth rate for the poor is greater than zero and "relatively pro-poor" if, in addition, the mean growth rate for the poor is at least as large as the growth rate in the overall mean. Hence, "absolute pro-poor growth" only requires that the poor be better off on average in absolute terms, while "relative pro-poor growth" requires the distributional shifts to be pro-poor as well. In terms of Tanzania as a whole, the growth incidence curve lies entirely above zero, implying that consumption increased at every point in the distribution. In other words, growth for the country as a whole was pro-poor in absolute terms. Likewise, growth was absolutely pro-poor in Dar es Salaam and rural areas. Although mean growth rates were modest everywhere except Dar, growth did improve consumption for rich and poor alike. Only in the distribution for the other urban areas stratum were there absolute declines, in the poorest percentiles. In Dar es Salaam and other urban areas mean growth rates for the poor were below growth in the mean for the stratum as a whole and growth was not relatively pro-poor. In rural areas, however, growth was relatively pro-poor as the modest amount of growth that did take place was concentrated among the poorest. In rural areas the rate of growth among the poor averaged 0.5% per year, versus mean growth of 0.3% for rural areas overall. This mixed picture for rural and urban areas means that for the nation as a whole the mean growth rate for those who were poor in 1991/92 is 0.6%, just below the growth rate in the mean of 0.7%. Another way to view the curves is to recognize that a GIC that is upward sloping implies that the poor benefit from growth less than the non-poor, leading to an increase in inequality. The curves for Tanzania show that urban areas, which experienced the greatest gains, also saw increases in inequality. It should be noted that given the size of the confidence intervals on the curves, small differences between growth rates shown on the curve are not statistically significant. Additionally, for rural areas, other urban areas, and the nation as a whole, we cannot reject the hypothesis that the curves are flat, i.e. that the impact of growth is evenly distributed. The greater rate of growth for non-poor households in urban areas, however, is statistically significant. In summary, the curves show that growth has had widely different distributional impacts by sector. As the large majority of people live in rural areas, the national growth incidence curve resembles the growth incidence curve for rural areas, showing equally distributed, limited, growth with some relative gains for the poorest. Figure 1 8 Growth Incidence Curves Growth Incidence Curve: Nation as a Whole Growth Incidence Curve: Dar es Salaam 6 Growth incidence curve 6 ntelaviuqEtludArep 95% confidence bounds 5 5 Growth rate in mean 4 3 ntelaviuqEtludArep 4 3 noi pt 2 noi pt 2 um um 1 1 Cons Cons ni ni htworG 0 -1 htworG 0 -1 % % Growth incidence curve ualnnA -2 ualnnA -2 95% confidence bounds Growth rate in mean -3 -3 0 20 40 60 80 100 0 20 40 60 80 100 Percentiles of Household Consumption per Adult Equivalent Percentiles of Household Consumption per Adult Equivalent Growth Incidence Curve: Other Urban Areas Growth Incidence Curve: Rural Areas 6 Growth incidence curve 6 Growth incidence curve ntelaviuqEtludArep 95% confidence bounds 5 ntelav 95% confidence bounds 5 Growth rate in mean Growth rate in mean 4 3 quiEtludArep 4 3 noi pt 2 noitp 2 um mu 1 Cons onsC 1 ni ni htworG 0 -1 htworG 0 -1 % ualnnA -2 %launnA -2 -3 -3 0 20 40 60 80 100 0 20 40 60 80 100 Percentiles of Household Consumption per Adult Equivalent Percentiles of Household Consumption per Adult Equivalent 9 4 Poverty Simulation Methodology 4.1 Basic Approach This section briefly sketches the approach used in the main part of the analysis. A fuller treatment is given in Appendix 1 of Datt and Walker (2002) and in Datt et al. (2003). The general approach is to simulate changes in consumption by applying annual GDP per capita growth rates to unit-record household survey data. For this application, consumption is defined in per adult equivalent terms. Allowance can be made for varying population and GDP growth rates at the sectoral/regional level. The simplest version assumes that growth is distribution neutral, i.e. that inequality is unchanged. Alternatively, we can incorporate estimates of changes in inequality levels in the simulation. The discussion presented here is done going forwards from the 1991/92 survey data. It is also possible to project the simulations backwards from the 2000/01 survey. Except for one case included in the appendix, the simulations conducted for this paper are done going forwards. Notation Define the following terms for a sample of n households: · ci is per adult equivalent consumption for sample household i in year t, where t=0 ,t represents the survey year · wi is the individual weight for household i in year t. These weights are the product of the ,t household sampling weights (the inverse of the probability of selection of the household in the survey) and household size. · the sector (or, alternatively, region) of household i in the survey year is given by si . This is ,t assumed fixed over time, but we do allow for population shifts across sectors. · gt is the real GDP growth rate for the sector of household i in year t. Si · t is the population growth rate for the sector of household i in year t. Si Basic projections The basic form of the projections is to calculate per adult equivalent consumption recursively, (1) ci,t = ci,t-1(1+ gt -t ) , Si Si while adjusting the weights for population changes year-by-year: (2) wi,t = wi,t-1(1+t ) . Si 10 Note that in the consumption equation, the growth rate of consumption per adult equivalent is implicitly approximated as being equal to the growth rate of GDP per capita. 5 Inequality adjustment The assumption of distribution-neutral growth can be relaxed by adjusting consumption for each household within each sector year-by-year. This adjustment is made after the growth projection for each year. Take the percentage change in the Gini coefficient in sector Si in year t as Gt . The S adjusted level of per capita consumption for household i is then (3) ci,t ADJ = ci,t - Gt (cSi -ci ). S ,t This produces a proportional shift in the sectoral Lorenz curve by adjusting consumption for each household relative to its deviation from the sector-specific mean. Note that the mean here is necessarily the weighted mean, calculated with the weights given by{wi }. This adjustment ,t effectively redistributes consumption from households below the mean to those above the mean (for an increase in the Gini), while leaving mean consumption by sector constant. It would be possible to undertake alternative redistribution procedures that would achieve the same outcome in terms of changes to the Gini and the mean. 4.2 Two Survey Approach The Datt and Walker approach outlined above was designed to project changes in poverty based on national growth data and a single household survey, in the absence of data from multiple household surveys. For the analysis constructed in this paper, the problem is somewhat different, as the goal is to understand changes in the distribution during the period between two household surveys. While in the Datt and Walker general case the distribution of consumption is known only at the beginning of the simulated period, for Tanzania in the 1990s the full distribution is known both at the beginning and the end of the period.6 An extension of the Datt and Walker method can be used to force the distribution at the end of the simulation to closely match that of the survey data. Specifically, extend the notation from above to reference each household's quantile, with n the number of fractions in which the distribution is broken down. For instance if n equals five, the distribution would broken down in quintiles. · gt Si ,Qiis the real GDP growth rate for the quantile and sector of household i in year t. · t Si ,Qiis the population growth rate for the quantile and sector of household i in year t. These quantiles are defined over individuals by household consumption per adult equivalent and can be in terms of the full national distribution, or in terms of each sector's distribution. The simulation is carried out using these growth rates: 5In fact, the 2002 Census report shows that children made up a slightly smaller percentage of the total population in 2002 as they did in 1988. This means that the growth rate of consumption per adult equivalent was slightly larger than the growth rate of consumption per capita. 6More precisely, estimates of the distribution are known from the household surveys. 11 (4) ci = ci ,t ,t -1(1+ gt Si ,Qi-t ). Si Because these year-by-year sector-quantile growth rates are not directly observed, estimates must be used. What is observed (via survey-based estimates) is each sector-quantile's cumulative growth in mean household consumption per adult equivalent between the two surveys. This is simply (5) GS,Q = cT S,Q -1, cT0S,Q where ct S,Q is the mean of household consumption per adult equivalent in year t, and T0 and T are the initial and final years, respectively. One way to estimate year-by-year sector-quantile growth rates would be to assume constant growth over time within each sector-quantile. The preferred approach, in the spirit of Datt and Walker, is to scale the national accounts year-by-year growth rates by the sector-quantile growth rates. To see how these can be calculated, note that cumulative growth in GDP per capita in the national accounts is simply the product of the series of annual growth rates: T (6) (1+ G*) = ( 1+ gt -t ) . * * t=T0 +1 An asterisk is used here to refer to overall growth versus growth in particular sector-quantile. Note that the first year for which the growth rate is applied is not the initial year but the following year. Multiply both sides of that equation by (1+G S, Q ): T (7) (1+ GS,Q)(1+ G*) = (1+ GS,Q) ( 1+ gt -t ) * * t=T0 +1 We carry out a series of algebraic manipulations to find the series of scaled growth rates by sector and quantile: (1+ GS,Q) T (8) 1+ GS,Q = * * (1+ G*) ( 1+ gt -t ) t=T0 +1 (9) T 1(T-T0 ) (10) 1+ GS,Q = t=T0 +1(1(+GS,Q) * * 1+ G*) (1+ gt -t ) T (11) 1+ GS,Q = t=T0 +11+ (1(+GS,Q) 1(T-T0 ) 1+ G*) (1+ gt -t ) -1 * * Equation (11), with a structure parallel to that of equation (6), implies a series of year-by-year growth rates for sector S and quantile Q that results in cumulative growth equal to GS . The effect ,Q 12 is to scale the year-by-year growth rates from the national accounts for each sector-quantile such that they cumulatively produce the growth observed in the survey for that sector-quantile. When year-by-year growth rates calculated this way by sector and quantile are applied in the simulation, starting with the initial year survey data, they should produce a distribution in the final year which closely matches the final year survey data. In these simulations, each quantile's mean household consumption moves proportionally with the GDP per capita in the national accounts. This procedure is the natural extension of the inequality adjustment Datt and Walker apply in one version of their method. In the approach outlined here, rather than adjust the distribution to generate a specified change in a single inequality measure, we set the growth rates quantile-by-quantile to produce a particular distribution in the final year.7 This procedure is approximately (though not exactly) equivalent to setting the growth rates for individual quantiles by shifting the national growth rate pattern up or down. To illustrate this, in Figure 2 below, the true growth rates for Tanzania from the national accounts are plotted along with growth rates calculated using the procedure for two hypothetical quantiles with cumulative growth of 1% and 40%. Recall from Table 2 that cumulative GDP per capita growth 1993-2001 in Tanzania was 6.1%. Figure 2 Simulated Growth Rates for Hypothetical Quantiles Using Two Survey Method 7.0% r pe 6.0% noitp 5.0% m 4.0% nsuoC tne 3.0% d val 2.0% Per Capita GDP Growth (6.1%) hol uiqEt Quantile w/ 1% Cum. Growth 1.0% Quantile w/ 40% Cum. Growth useoH uldA 0.0% ni -1.0% ge anhC -2.0% % -3.0% -4.0% 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Year Note that the initial and final years of the simulation will match the "true" distributions in the survey data, and consequently the simulation's cumulative change in mean consumption will match the change implied by the survey data. If cumulative growth in GDP per capita differs from 7The "two survey" approach has a natural relationship to growth incidence curves: the simulations scale growth for each quantile to match the average growth rates that are plotted in the GICs. 13 cumulative consumption growth in the survey data, the simulated year-by-year changes in mean consumption will not be equal to the year-by-year GDP per capita changes. In Tanzania, the cumulative GDP per capita growth (6.1%) is close to growth in consumption per adult equivalent in the survey (6.5%). As a result, the difference between the GDP per capita changes and simulated mean consumption changes in each year is small. In general, use of a large number of quantiles will cause the simulated distribution in the final year to more closely match that in the final survey year data. However, when using weighted data, complications may arise that force the use of a smaller number of quantiles. The simulation requires that at least one household be associated uniquely with each quantile. Because the quantiles are calculated by individual, with weighted data a household may be classified as being in two adjacent quantiles, even though the number of households exceeds the number of quantiles. If this is the case, a smaller number of quantiles should be used. For the analysis in this paper, 50 quantiles are used. 4.3 Reconciling National Accounts and Household Survey Data The approach taken in this paper is to apply per capita GDP growth rates from the national accounts data to unit-record household survey data to simulate changes in household per capita consumption. This approach assumes some correspondence between the micro-level household data and the macro national accounts data. Because the two sets of data are drawn from entirely different sources and are designed to capture different phenomena, it is to be expected that the match-up between the two sources of data is rough at best. In general, it is often difficult or impossible to reconcile differences between the national accounts and household survey data. As Ravallion (2003) says in a review of comparisons between national accounts (NAS) and survey data, "It is evident that when the levels or growth rates from these two data sources differ, there can be no presumption that the NAS is right and the surveys are wrong, or vice versa, since they are not really measuring the same thing, and both are prone to errors." In terms of growth rates in Tanzania for the period between the household surveys (1993- 2001), there is surprisingly good correspondence between the HBS data and the national accounts. The cumulative GDP per capita growth of 6.1% is very close to the 6.5% growth in consumption per adult equivalent in the household survey. This close match between the general picture painted by the micro and macro data provides support for the validity of the simulation exercise undertaken for this paper. It is also possible to calculate the growth rate of per capita consumption directly from the national accounts data. Figures from these calculations are shown in the last column of Table 2 and are plotted in Figure 3 below, along with the GDP per capita growth rate figures. At 8.3%, the cumulative change in per capita household consumption is only slightly larger than the cumulative change in per capita GDP. Mathematically, this is because household consumption changed just slightly, from 90% to 91% of GDP, between 1993 and 2001. While details on the particular methodology used to develop the Tanzania national accounts data are not available, typically consumption figures in national accounts statistics are estimated as a residual in the national income calculation. As a consequence, there is great uncertainty in the estimates, particularly in year-by-year fluctuations. For this reason, growth rates in GDP per capita are used in this paper, except in one variation of the simulation analysis. 14 Figure 3 Growth Rates of GDP per Capita and Household Consumption per Capita in Tanzania, National Accounts Data 5.0% 4.0% 3.0% etaR 2.0% GDP per Capita th 1.0% Hhold Cons. per Capita (National Account Data) ow Gr 0.0% -1.0% -2.0% -3.0% 1988 1990 1992 1994 1996 1998 2000 2002 2004 Year 4.4 Sources of Error in the Simulation Analysis Several sources of error enter the simulation analysis. First, there is the usual sampling error associated with the survey data. Second, there is uncertainty associated with the national accounts growth estimates. Third, there is drift between the national accounts GDP levels and household consumption levels, i.e. the growth rates of the two may not be equal. Fourth, the true year-by-year changes in the distribution of consumption differ from the changes assumed in the simulations. Fifth, the uniform application of national population growth rates to the rural and urban sectors appears inconsistent and sector specific growth rates might be preferable. Of all of these, only sampling error is readily quantifiable, and the associated standard error terms are calculated for all simulated poverty rates. It is important to recognize, however, that these standard errors are only an extreme lower bound for the true error. The fifth source of error, finally, is considered in Section 5.4. 5 Simulation Analysis The simulation analysis is conducted under a variety of scenarios, with varying assumptions, first using the Datt and Walker approach and then using the two survey approach. The intent is both to explore the sensitivity of the results to the underlying assumptions applied, and to identify variations which most closely match the change in the distribution of consumption observed between the two surveys. 15 Note that for all simulations, the 1991/92 HBS data is taken as corresponding to calendar year 1992, and likewise the 2000/01 data is taken as being from 2001. Consequently, the projections apply the annual growth rates starting in 1993, the year following the first survey. The simulations are carried forward one additional year past the second survey, to 2002. An alternative choice would be to assign the two surveys to calendar years 1991 and 2000, respectively, and apply the growth rates from 1992 onwards. Several versions of the analysis (not shown) were run with this assumption, and results were similar. 5.1 Simulation with Uniform National Growth Rates The simplest version of the analysis is to apply the national GDP per capita growth rates, as shown in Table 2, uniformly to all households. Headcount rates from this simulation are shown in Appendix Table A2 and plotted in Figure 4 below. Poverty levels for the two years corresponding to the surveys, 1992 and 2001, are shown in italics. Headcount rates from the 2000/01 HBS are also appended to the end of table for comparison purposes. The simulated poverty rates for 2001 match the survey data surprisingly well. The survey estimate of 0.357 is well within the confidence interval of the simulated value, 0.343. The simulated headcount rates for "other urban areas" and rural areas also match up well with the survey estimates. This simplest form of the simulation, however, does not reproduce the steeper decline in poverty in Dar es Salaam. Figure 4 Simulated changes in poverty using national growth rates 5 .4 neilyt .4 er 5 pov .3 ow bel .3 onit 5 acrF .2 .2 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Year Mainland Tanzania Dar es Salaam Other urban areas Rural areas The overall picture shows the inverse of the growth pattern in Tanzania in the decade: declining GDP per capita in the early 90s drove a steep rise in poverty, while the country's improved growth trajectory in more recent years has achieved a steady decline in poverty. With additional versions of the simulation, we can assess the sensitivity of this general pattern to the underlying assumptions. Note that applying uniform growth rates to all households does not imply identical 16 rates of poverty reduction across rural and urban areas. The simulated rates of poverty reduction are determined by both the growth rate and the distribution of consumption within each of the sectors. 5.2 Simulation Using National Accounts Household Consumption Data Results from a simulation applying growth rates for household consumption per capita (shown in the last column of Table 2) are displayed in Appendix Table A3 and Figure 5. While the overall pattern is unchanged, applying these growth rates instead of those for GDP per capita implies that rather than dropping fairly steadily since the mid-90s, poverty dropped steeply 1996-98 and has since remained flat. This is due to the fact that, according to the national accounts data, household consumption as a fraction of GDP grew during the mid-90s and then fell at the end of the decade, offsetting the effect of rapid growth. Figure 5 Simulated changes in poverty using hhold exp per capita growth rates from national accounts 5 .4 neilyt .4 er 5 pov .3 ow bel .3 onit 5 acrF .2 .2 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Year Mainland Tanzania Dar es Salaam Other urban areas Rural areas 17 5.3 Simulation with Separate Urban/Rural Growth Rates Appendix Table A4 and Figure 6 show results from a simulation applying separate urban and rural growth rates, and the national population growth rate of 2.9%. The separate growth rates were calculated from the data in the Economic Survey 2002 by grouping changes in output by sector; the rural growth rate was calculated as the growth rate of the agricultural sector (monetary and non- monetary), and the total growth of other sectors was taken as the urban growth rate. The simulated 2001 headcount rates are lower across the board than in the national growth scenario, and in all cases the simulated values fall within the confidence intervals of the estimates from the 2000/01 HBS. While the overall pattern is broadly the same as in the first scenario, this simulation shows a much steeper rise and fall for poverty in Dar es Salaam, and a flatter poverty trajectory in rural areas and the nation as a whole. Figure 6 Simulated changes in poverty using separate urban and rural growth rates 5 .4 neilyt .4 er 5 pov .3 ow bel .3 onit 5 acrF .2 .2 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Year Mainland Tanzania Dar es Salaam Other urban areas Rural areas 5.4 Simulation with Alternative Urban/Rural Population Growth Rates The sectoral (urban/rural) GDP per capita growth rates employed in the previous section are the same as those displayed in Table 2. These were calculated from sectoral GDP figures, assuming a uniform national population growth rate of 2.9%. This population growth rate corresponds to the change in national population between the 1988 and 2002 censuses. Alternatively, we can use sectoral population figures to calculate the per capita growth rates. Annual population growth averaged 4.6% in urban areas and 2.5% in rural areas. Appendix Table A5 and Figure 7 below display results from a simulation employing these alternative growth rate figures. This simulation fairs much worse than others in terms of matching the true 2002 poverty rates. It shows the rural headcount rate declining to 0.318, far below the true 18 value of 0.387. For Dar es Salaam, rather than declining to 0.176, the simulation shows poverty rising to more than twice that level, 0.358. It is unclear why using sectoral population growth rates produces such implausible results. We suspect this is related to population growth assumptions underlying the national accounts data. However, without a detailed understanding of how the national accounts data were constructed, we are unable to say conclusively. For purposes of this paper, we focus on the sectoral per capita growth rates calculated with the national population growth rate, on the basis that they provide a much better match to the 2001 survey data. Figure 7 Simulated changes in poverty using alternative urban and rural growth rates 5 .4 neilyt .4 er 5 pov .3 ow bel .3 onit 5 acrF .2 .2 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Year Mainland Tanzania Dar es Salaam Other urban areas Rural areas 5.5 Simulation with Regional Growth Rates Another alternative is to apply growth data at the regional level, using annual GDP data by region. This data, in current prices, was converted to real values using the GDP deflator in Table 2, and the annual levels were used to calculate regional GDP growth rates for each year. Separately, regional population growth rates were calculated using the regional totals from the 1988 and 2002 national censuses. These regional GDP and population growth rate estimates are shown in Appendix Table A6. Regional GDP per capita growth rates were then calculated year-by-year by subtracting the population growth rates from the GDP growth rates.8 Population changes were also taken into account to adjust the weights used in the calculation of poverty rates. Results from this simulation are shown in Appendix Table A7 and in Figure 8 below. The resulting simulation closely tracks the results using only rural/urban growth rates, with the exception 8While the regional GDP changes are different values year-by-year, the population growth rate by region is assumed constant over time. 19 of Dar es Salaam, which is its own region in the regional data. The simulation shows a large increase in poverty in Dar es Salaam, contrary to the decline observed in the survey data. Figure 8 Simulated changes in poverty using growth rates by region 5 .4 neilyt .4 er 5 pov .3 ow bel .3 onit 5 acrF .2 .2 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Year Mainland Tanzania Dar es Salaam Other urban areas Rural areas The simulated poverty increase in Dar es Salaam is the product of a steep decline in GDP per capita in the official figures, which is not compatible with the 43% increase in mean per capita consumption seen in the HBS for the capital. While details of the methodology behind the regional growth data are not available, it appears likely that the regional figures were calculated using approximations from national data. Indeed, the official figures imply that GDP growth rates for Dar are identical with those of 17 other regions (out of 19) in 1999 and with 11 other regions in 2000. Such approximations are unlikely to offer an accurate picture in a rapidly changing region. In Dar, population growth was extremely rapid; figures calculated from the 1988 and 2002 censuses show that population grew at a 4.4% annual rate. As a result, it is likely that the approximations used to calculate regional GDP underestimate the growth of output per capita in Dar. A reasonable alternative is to substitute the overall urban growth rate (used in the simulation described in section 5.3). Appendix Table A8 and Figure 9 below shows results from simulations using regional growth rates, but substituting overall urban growth rates for Dar. 20 Figure 9 Simulated changes in poverty using growth rates by region, but substituting urban rate for Dar 5 .4 neilyt .4 er 5 pov .3 ow bel .3 onit 5 acrF .2 .2 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Year Mainland Tanzania Dar es Salaam Other urban areas Rural areas 5.6 Simulation with Regional Growth Rates and Inequality Adjustment The simulations presented thus far assume that growth was distribution neutral.9 According to the survey data, inequality jumped up in Dar es Salaam, while remaining essentially unchanged in other urban areas, rural areas, and the country as a whole. The survey data shows that the Gini coefficient for Dar increased from 0.30 to 0.34 between 1992 and 2001. In this simulation, the inequality adjustment proposed by Datt and Walker (2002) and in Datt et al. (2003) and described in the methodology section is applied by sector (Dar es Salaam, other urban, and rural) in a simulation, using region-specific growth rates. The overall urban growth rate is again substituted for the Dar-specific growth rate. A constant inequality adjustment is applied such that the Gini coefficient in the simulation increases from 0.30 to 0.34 for Dar es Salaam. Results from this simulation are presented in Appendix Table A10 and Figure 10 below. While the general path of the evolution of poverty is similar, the net drop in poverty in Dar is smaller than that observed in the survey data. 9More strictly, the simulation with separate urban-rural growth rates entails the assumption of distribution-neutral growth within each sector, while the simulation with separate regional growth rates used the assumption that growth within each region was distribution-neutral. 21 Figure 10 Simulated changes in poverty using regional rates with inequality adjustment 5 .4 neilyt .4 er 5 pov .3 ow bel .3 onit 5 acrF .2 .2 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Year Mainland Tanzania Dar es Salaam Other urban areas Rural areas 5.7 Simulation using "Two Survey" Approach Finally, simulations were carried out using the "two survey" approach. This was done in two ways. In the first, the national distribution of consumption was used, and national GDP per capita growth rates from the national accounts were applied. In the second, changes in the distributions were simulated separately for each of the three strata: Dar es Salaam, other urban, and rural areas. For this second approach, urban sector national accounts growth rates were used for Dar and other urban areas, while rural sector national accounts growth rates were used for the rural areas stratum. Results from these two simulations are shown in Appendix Tables A11 and A12, and Figures 11 and 12 below. As expected, the simulated values for 2001 closely match the estimates from the 2000/01 HBS.10 These simulations show broadly the same patterns as the simulations using the Datt and Walker approach: rising or flat poverty in the initial years, followed by small declines. Unlike the Datt and Walker simulation, these are able to capture the divergence between the rate of declines in poverty in Dar es Salaam and the rest of the country. Unlike the simulation using the national distribution and national growth rates (Figure 11), the simulation using within stratum distributions and urban/rural growth rates (Figure 12) shows small increases in poverty in 1997 and 1998. This is due to the fact that per capita rural output declined in those two years, even while per capita output for the country as a whole increased. 10They do not match exactly because the simulated changes are based on an approximation of the true distributions using 50 quantiles. The maximum number of quantiles used was limited due to issues associated with weighting, as discussed in Section 3.2. 22 Figure 11 Simulated changes in poverty 2 survey approach, using national distribution 5 .4 neilyt .45 er .3 pov .3 ow 5 bel .2 onit .2 acrF 5.1 .1 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Year Mainland Tanzania Dar es Salaam Other urban areas Rural areas Figure 12 Simulated changes in poverty 2 survey approach, using within stratum distribution 5 .4 neilyt .45 er .3 pov .3 ow 5 bel .2 onit .2 acrF 5.1 .1 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Year Mainland Tanzania Dar es Salaam Other urban areas Rural areas 23 5.8 Did Poverty Decline? The simulations suggest that poverty declined in Tanzania after 1994. This decline is attributable to the key assumption underlying the simulation approach, that positive per capita GDP growth translates into consumption increases for all population groups (and consequently a reduction in poverty). The appropriateness of this assumption can be scrutinized by investigating the implications of the possibility that poverty did not decline. Specifically, if poverty did not decline between 1995 and 2001, while per capita GDP grew at the levels shown in the national accounts, what would the level of inequality had to have been in 1994? The assumption that poverty did not decline would require that all the growth was concentrated among the non-poor, meaning that inequality increased and was therefore lower in 1994 than in 2001. We consider this scenario by simulating changes backwards from 2001 to 1994, using the 2000/01 HBS data. In this simulation, we set out to keep poverty constant, while adjusting for changes in mean consumption.11 This can be done by holding consumption constant for those in poverty in 2001 (so that the fraction of poor remains unchanged) and restricting all growth to the non-poor. We also assume that consumption grew at the same rate among those households that experienced some growth. Note that growth in consumption going backward in time corresponds to reducing consumption. Now consider a household whose consumption in 2001 is just above the poverty line. If we were to reduce its consumption as we move backwards in time, the household would fall below the poverty line and poverty in 1994 would not be identical to poverty in 2001. Through a series of simulation exercises, we find that to keep the level of poverty constant between 2001 and 1994 while adjusting for changes in mean consumption, we must restrict the growth adjustment to those in the top 53% of the consumption distribution in 2001. With the scenario so defined, we can consider what inequality would have been in 1994 if poverty did not decline between 1994 and 2001. Under this scenario, the Gini coefficient for 1994 would have been 0.31, compared to 0.34 in 2001 and 0.33 in 1992. The difference between 0.31 and the values in both 1992 and 2001 is statistically significant. It is possible but unlikely that inequality would have declined from 0.33 to 0.31 in just two years. This means that the scenario considered--that poverty did not decline at all 1995-2001--is unlikely to have taken place. This provides some further support for the hypothesis that poverty rates followed the inverse-U pattern shown in the simulations.12 5.9 Poverty Incidence and the Millennium Development Goals Finally, we consider how declines in poverty compare with declines that would be needed to meet the Millennium Development Goal target, interpreted here as a reduction by half from 1992 levels in both rural and urban poverty by 2015. Figure 13 shows the simulated poverty trajectories using the two-survey approach based on the stratum-level distributions of consumption. 11Specifically, the national accounts data shown in Figure 2 imply that per capita GDP in 1994 was equivalent to 90.3 percent of per capita GDP in 2001. We assume, as earlier, that growth in per adult equivalent consumption was equal to growth in per capita GDP. 12It is possible to construct scenarios where poverty did not decline 1995-2001 but inequality also remained unchanged. Specifically, this would be the case if growth took place only among middle-income households, while consumption remained constant for both the poor and the rich. 24 The figure shows that recent growth has brought the urban poverty rate approximately on track to achieve the MDG by 2015. By the interpretation used here, the MDG target urban poverty rate for 2001 was 22.9 percent, while the estimate of the actual value from the 2000/01 HBS survey was 23.3 percent.13 Figure 13 Poverty incidence and the MDG goals simulated poverty based on two-survey approach, stratum distribution 5 .4 neilyt .45 er .3 pov .3 ow 5 bel .2 onit .2 acrF 5.1 .1 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 Year All urban areas (including DSM) Rural areas MDG, urban target MDG, rural target The rural poverty rate remains substantially above the path necessary to achieve the MDG target. The MDG target for rural areas is attainable, but it will require sustaining growth at or above the rates achieved in 2001 and 2002. The rural GDP per capita growth rate was 2.6 percent in 2001 and 2.1 percent in 2002. A simulation forward from the 2000/01 HBS data, assuming distribution- neutral growth, implies that the rural poverty reduction target will be met if a rural GDP growth rate of 2.3 percent per capita is maintained through 2015. The estimate of 2.3 percent is a lower bound, assuming no increase in inequality. If future growth is accompanied by increases in inequality, a growth rate greater than 2.3 percent in GDP per capita will be needed to achieve the MDG target. Assuming rural population growth net of migration continues at the rate of 2.5 percent annually and inequality does not increase (the national growth rate 1988-2002), rural GDP growth will need to reach 4.8 percent to achieve the MDG target. Similarly, a simulation done for the entire country on the basis of a single national growth rate shows that annual GDP growth for the country as a whole will need to reach 5.1 percent to achieve the MDG target, assuming no increase in inequality and no change in the population growth rate. Figure 14 shows the GDP growth rates that would be needed under alternative population growth rates and inequality scenarios. The numbers underlying the figure were calculated from simulations assuming increases in the Gini coefficient in the form outlined in section 4.1. Because 13Note that the urban poverty estimate for 2002 which is below the MDG path is from a one-year extension of the simulation, rather than from the HBS survey directly. 25 the same Gini could be associated with a variety of poverty rates, the particular correspondence here should be taken as suggestive of the general relationship between the Gini, population growth, and GDP growth. The figure illustrates that higher population growth rates or increases in inequality would require a GDP growth rate of above 5.1 percent to achieve the MDG target. Likewise, a lower GDP growth rate could be sufficient to achieve the MDG target if inequality declines or population growth slows.14 Figure 14 GDP Growth Rates Needed to Meet 2015 Millennium Development Goal Target for Poverty Reduction in Tanzania Under Various Inequality and Population Growth Rate Scenarios 0.41 0.40 Annual Population Annual Population Growth Rate Annual Population 0.39 Growth Rate Remains at 1988- Growth Rate Slows to 2% 2002 Rate of 2.9% Increases to 4% 15 0.38 20 in 0.37 ent 0.36 cii 0.35 effoC The horizontal dotted line 0.34 indicates the Gini coefficient in 2002. The Gini0.33 verticle dotted lines indicate required GDP 0.32 growth rates assuming no change in inequality by 2015. 0.31 0.30 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 Annual % GDP Growth Rate to Meet MDG Poverty Target 14The effect of changes in inequality on the GDP per capita growth rate needed to achieve a particular poverty rate is described by a simple formula. We assume that growth consists of a combination of distribution-neutral growth and a mean-neutral change in inequality of the form outlined in section 4. Define the following notation: n is the number of years between the final and initial years,GF is the Gini coefficient in the final year, G0 is the Gini coefficient in the initial year , cpovline is the poverty line, c0 is mean consumption in the initial year, and cX is the consumption in the initial year of the household at percentile X, where X is the target headcount. The annual per capita GDP growth rate 1 / n cpovline necessary to achieve X is the following: -1. cX (GF -G0)(c0 -cX ) - G0 26 6 Conclusions The starting point for the analysis in this paper was the observation that despite rapid growth in per capita GDP in the late 1990s, survey data shows that the drop in poverty nationally between 1992 and 2001 was small. A plausible explanation for this pair of facts is that poverty first increased during the period of economic stagnation that ended in 1995 and only declined once rapid growth was achieved in the second half of the decade. Consequently, recent growth has reduced poverty, even though the change observed between the two surveys shows only a small net drop in the fraction who are poor. The simulated poverty trajectories in this paper support this view. Under a variety of scenarios, the simulations imply that poverty rates have followed an inverted U-pattern, increasing to over 40% or higher in 1994 and then dropping, down to below 36% in the 2000/01 survey. Variations of the analysis which take into account separate urban-rural growth rates show small increases in rural poverty in 1997 and 1998, when rural per capita income declined while urban per capita output grew, followed by declining rural poverty. We employ both the Datt-Walker macro-micro projection technique and a modified "two survey" version of their method. In the Datt-Walker analysis there are instances where growth patterns diverged substantially from the general pattern. These are the simulations using Dar-specific growth data, aggregate household consumption estimates from the national accounts, and alternative urban/rural population growth rates. In all three cases, it is likely that the results are the consequence of rough approximations in the national accounts data. In most cases, relative to the survey estimates, the simulations with the Datt and Walker method overestimate drops in poverty in rural areas and the nation as a whole, while underestimating the drop in poverty in the capital. This reflects the particular distributional incidence of growth (in the survey data) which is illustrated by the growth incidence curves. Our modified "two survey" method provides a close match between the simulation and the final year survey data, irrespective of the particular distributional incidence of growth. Consequently, we take the two versions of the analysis using this method (Figures 12 and 13) as our preferred simulated poverty trajectories. We also presented some supporting evidence that suggests it is very likely that poverty declined since 1994. The alternative hypothesis--that poverty has not declined--would require that inequality dropped significantly from 1992 to 1994. It is important to recognize that the precision and accuracy of the presented results is only as good as the underlying data. Uncertainty is associated with the macro data, the micro data, and the assumptions that go into the simulation analysis. Only the sampling error arising from the household survey can be readily quantified, and the standard errors on the survey-based estimates provide an extreme lower bound on the standard errors on the simulated poverty rates. This suggests that the year-to-year pattern of changes in poverty may differ substantially from what is implied by the simulations. Nonetheless, the analysis presented here provides the best estimates of poverty rate trajectories in Tanzania with available data. In terms of the Millennium Development Goal of cutting poverty in half by 2015, growth at the end of the last decade has put Tanzania roughly on the path necessary to meet the goal for urban areas. This is despite the fact that the poverty-reducing impact of growth has been partly offset by increasing inequality, particularly within Dar es Salaam. Rural areas, where the large majority of Tanzanians live, are lagging behind their MDG target. Reaching the MDG goal for poverty reduction by 2015 will require sustaining and improving upon the rural growth rates achieved in recent years. Given that only a tiny minority of the population lives in the capital, even rapid urban growth will have only a small impact on poverty in 27 the nation as a whole. Tanzania needs strong growth in rural output per capita to make a substantial dent in the national poverty rate and attain the national poverty MDG. This requires sustained growth in rural GDP, a reduction in the rate of population increase or a combination of both. 7 References Datt, G., K. Ramadas, D. van der Mensbrugghe, T. Walker, and Q. Wodon [2003], "Predicting the effect of aggregate growth on poverty," in F. Bourguignon and L. Pereira da Silva (eds), The Impact of Economic Policies on Poverty and Income Distribution: Evaluation Techniques and Tools, The World Bank and Oxford University Press. Datt, G. and T. Walker [2002], Povstat 2.12, A Poverty Projection Toolkit, User's Manual, The World Bank, October 14, 2002, mimeo. Datt, G. and Ravallion, M. [1992]. "Growth and Redistribution Components of Changes in Poverty: A Decomposition with applications to Brazil and China in 1980s", Journal of Development Economics 38: 275-295 Huppi, M. and Ravallion, M. [1991] "The Sectoral Structure of Poverty During an Adjustment Period. Evidence for Indonesian the Mid-1980s", World Development 19: 1653-1678 Ravallion, M. and Chen, S. [2002] "Measuring pro-poor growth", The World Bank. Ravallion, M. [2003], "Measuring Aggregate Welfare in Developing Countries: How Well do National Accounts and Surveys Agree?" Review of Economic Statistics, 85(3): 645-652. United Republic of Tanzania, President's Office [2002], Economic Survey 2002. United Republic of Tanzania, National Bureau of Statistics [2002], Household Budget Survey 2000/01. 28 Appendix Simulation with Regional Poverty Rates Backwards from 2001 Data The analysis presented thus far involves simulating changes in poverty forwards from the 1991/92 survey data. Although Datt and Walker do not discuss the possibility in their work, it is also possible to project changes backwards from the 2000/01 data. In terms of the notation used in Section 4, this involves recursive calculations for household consumption and the weights as follows: ci,t (12) ci,t-1 = (1+ gt -t ) , Si Si and wi,t (13) wi,t-1 = (1+t ) . Si For Tanzania, the backwards analysis has one advantage over the forward analysis. Because the 2000/01 HBS is representative at the regional level, regional poverty rates going back in time can be simulated. The earlier HBS is only representative at the stratum level, so it cannot be used to produce region-level estimates. Year-by-year simulated headcount rates at the regional level are shown in Appendix Table A9, and headcount trajectories are plotted for a sample of regions in A1. The table also shows the standard errors on the 2001 estimates, which are calculated from the actual survey data. Overall, most but not all regions show a pattern mirroring that of changes in the national poverty rate: increases in poverty in the first part of the decade followed by declines in recent years. These regional poverty rates should be interpreted with great care. In addition to the caveats already mentioned, the standard errors on the regional predictions are large because the regional sample sizes are small. As discussed in the previous section, growth estimates for some regions in certain years (in particular 1999 and 2000) appear to have been obtained by very rough approximations based on national growth rates. In Mtwara, simulated poverty incidence drops at an implausible rate, from 64% in 1997 to 40% in 1999. As there is no baseline with which to compare the results of the regional predictions, it is very difficult to assess their accuracy. There is one region for which this can be done, Dar es Salaam. The backwards simulation implies that the headcount rate for Dar was 10.0% in 1992, while according to the 1991/92 HBS survey it was 28.1% Given the wide divergence between poverty incidence based on regional growth data and the HBS data , the regional results should be taken as broadly suggestive rather than indicative of particular regional trends. 29 Figure A1 Simulated Regional Poverty Rates for Selected Regions 0.80 0.70 et 0.60 Ra yrt 0.50 Lindi Mwanza ve Mara Pot 0.40 Mtwara Rukwa un Morogoro coda 0.30 Mbeya He 0.20 0.10 0.00 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Year 30 Appendix Tables Appendix Table A1: Published Poverty Statistics Other Mainland Dar es urban Rural Tanzania Salaam areas areas % Below Basic Needs Pov. Line 1991/92 38.6 28.1 28.7 40.8 2000/01 35.7 17.6 25.8 38.7 Consumption per Capita Mean, 1991/92 8686 14896 12733 7661 Mean, 2000/01 10120 21949 14377 8538 Ratio (00/01) to (91/92) 1.17 1.47 1.13 1.11 Gini Coefficient 1991/92 0.34 0.30 0.35 0.33 2000/01 0.35 0.36 0.36 0.33 Source: Household Budget Survey 2000/01, United Republic of Tanzania National Bureau of Statistics Notes: All figures shown are as published in the HBS report. Poverty lines were calculated on a per adult equivalent basis.1991/92 per capita consumption are given in 2000/01 prices, calculated using not the consumer price index but a Fisher Ideal price index calculated using price data in the HBS itself. The Fisher index implies that an average consumption basket increased in price by a factor of 2.49 between the surveys. Appendix Table A2: Simulated Headcount Rates, Using Uniform Nation Growth Rate Year Mainland Tanzania Dar es Salaam Other urban areas Rural areas 1992 (HBS) 0.386 0.281 0.287 0.408 (0.021) (0.028) (0.050) (0.024) 1993 0.403 0.297 0.297 0.427 1994 0.416 0.313 0.299 0.441 1995 0.409 0.301 0.298 0.433 1996 0.399 0.292 0.294 0.422 1997 0.397 0.290 0.289 0.420 1998 0.388 0.283 0.288 0.410 1999 0.369 0.268 0.282 0.390 2000 0.355 0.258 0.262 0.376 2001 0.343 0.243 0.257 0.363 2002 0.316 0.220 0.231 0.335 2001 (HBS) 0.357 0.176 0.258 0.387 (0.016) (0.027) (0.022) (0.020) Distributions were simulated going forwards, taking the initial distribution from the 1991/92 HBS survey data. Estimates from the 2000/01 HBS are shown for comparison purposes. Estimates from Household Budget Surveys are shown with standard errors in parentheses. 31 Appendix Table A3: Simulated Headcount Rates, Using National Growth Rate, Household Consumption Growth Rates from National Accounts Year Mainland Tanzania Dar es Salaam Other urban areas Rural areas 1992 (HBS) 0.386 0.281 0.287 0.408 (0.021) (0.028) (0.050) (0.024) 1993 0.393 0.284 0.288 0.416 1994 0.397 0.290 0.294 0.420 1995 0.394 0.284 0.288 0.417 1996 0.386 0.281 0.287 0.408 1997 0.361 0.261 0.266 0.382 1998 0.334 0.235 0.246 0.355 1999 0.327 0.225 0.245 0.347 2000 0.347 0.247 0.259 0.367 2001 0.324 0.224 0.245 0.343 2002 0.334 0.235 0.246 0.355 2001 (HBS) 0.357 0.176 0.258 0.387 (0.016) (0.027) (0.022) (0.020) Distributions were simulated going forwards, taking the initial distribution from the 1991/92 HBS survey data. Estimates from the 2000/01 HBS are shown for comparison purposes. Estimates from Household Budget Surveys are shown with standard errors in parentheses. Appendix Table A4: Simulated Headcount Rates, Using Separate Urban/Rural Growth Rates Year Mainland Tanzania Dar es Salaam Other urban areas Rural areas 1992 (HBS) 0.386 0.281 0.287 0.408 (0.021) (0.028) (0.050) (0.024) 1993 0.386 0.327 0.307 0.402 1994 0.396 0.347 0.319 0.411 1995 0.370 0.330 0.308 0.383 1996 0.363 0.313 0.302 0.375 1997 0.365 0.300 0.297 0.379 1998 0.367 0.279 0.285 0.385 1999 0.358 0.261 0.265 0.379 2000 0.353 0.243 0.257 0.375 2001 0.339 0.220 0.234 0.363 2002 0.319 0.206 0.219 0.342 2001 (HBS) 0.357 0.176 0.258 0.387 (0.016) (0.027) (0.022) (0.020) Distributions were simulated going forwards, taking the initial distribution from the 1991/92 HBS survey data. Estimates from the 2000/01 HBS are shown for comparison purposes. Estimates from Household Budget Surveys are shown with standard errors in parentheses. 32 Appendix Table A5: Simulated Headcount Rates, Using Alternative Urban/Rural Growth Rates Year Mainland Tanzania Dar es Salaam Other urban areas Rural areas 1992 (HBS) 0.386 0.281 0.287 0.408 (0.021) (0.028) (0.050) (0.024) 1993 0.385 0.341 0.317 0.398 1994 0.391 0.370 0.329 0.402 1995 0.374 0.390 0.367 0.375 1996 0.369 0.390 0.367 0.368 1997 0.370 0.391 0.374 0.368 1998 0.370 0.386 0.363 0.370 1999 0.361 0.382 0.336 0.364 2000 0.355 0.370 0.329 0.358 2001 0.335 0.358 0.327 0.334 2002 0.319 0.341 0.317 0.318 2001 (HBS) 0.357 0.176 0.258 0.387 (0.016) (0.027) (0.022) (0.020) Distributions were simulated going forwards, taking the initial distribution from the 1991/92 HBS survey data. Estimates from the 2000/01 HBS are shown for comparison purposes. Estimates from Household Budget Surveys are shown with standard errors in parentheses. The per capita growth rates underyling this simulation were calculated on the basis of separate urban/rural population growth rates. 33 Appendix Table A6: Regional Real GDP and Population Growth Rate Estimates Regional Population Regional Real GDP Growth Rates Growth Rates Region 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 1988-2002 Dodoma 0.011 0.014 0.032 0.041 0.039 0.026 0.045 0.016 0.084 0.070 0.023 Arusha -0.002 0.006 0.082 0.045 0.030 0.017 0.045 0.068 0.061 0.070 0.040 Kilimanjaro 0.003 0.017 0.031 0.041 0.033 0.034 0.045 0.016 0.051 0.055 0.016 Tanga -0.004 0.019 0.026 0.041 0.031 0.039 0.045 0.081 0.110 0.055 0.018 Morogoro -0.006 0.009 0.038 0.046 0.025 0.040 0.045 0.016 0.138 0.060 0.026 Pwani 0.014 0.038 0.012 0.030 0.043 0.030 0.045 0.016 0.023 0.045 0.024 Dar es Salaam -0.002 0.061 -0.025 0.024 0.042 0.051 0.045 0.016 0.028 0.069 0.044 Lindi 0.013 0.014 0.039 0.041 0.033 0.027 0.045 0.016 0.024 0.045 0.015 Mtwara 0.011 0.008 0.047 0.043 0.030 0.334 0.108 0.003 0.080 0.047 0.017 Ruvuma 0.010 -0.004 0.060 0.048 0.026 0.025 0.045 0.015 -0.073 0.050 0.026 Iringa -0.004 -0.003 0.052 0.050 0.022 0.036 0.045 0.016 0.096 0.065 0.016 Mbeya 0.006 -0.006 0.059 0.049 0.025 0.028 0.045 0.016 0.098 0.065 0.024 Singida 0.010 0.005 0.042 0.046 0.035 0.026 0.045 0.016 -0.032 0.046 0.023 Tabora 0.010 0.006 0.043 0.045 0.034 0.026 0.045 0.016 -0.010 0.050 0.037 Rukwa 0.008 -0.018 0.077 0.054 0.020 0.024 0.045 0.016 -0.046 0.048 0.036 Kigoma 0.015 0.017 0.036 0.037 0.037 0.025 0.045 0.016 0.104 0.050 0.049 Shinyanga 0.006 -0.014 0.066 0.053 0.026 -0.041 0.045 0.131 0.051 0.070 0.034 Kagera 0.011 0.010 0.040 0.043 0.034 0.027 0.045 0.092 0.028 0.052 0.032 Mwanza 0.005 -0.008 0.059 0.050 0.026 0.098 0.045 0.165 0.184 0.079 0.033 Mara 0.010 0.019 0.026 0.039 0.040 0.029 0.045 0.150 -0.042 0.050 0.027 Regional real GDP growth rate estimates were calculated based on nominal GDP level data. Nominal values were converted to real values using an implicit GDP price deflator calculated using data in Economic Survey 2002 . The GDP price deflator was calculated by dividing nominal GDP at factor cost prices (Table 2B) by GDP in 1992 prices (Table 3). Regional population growth rates are average annual rates, calculated on the basis of regional totals in the 1988 and 2002 national censuses. 34 Appendix Table A7: Simulated Headcount Rates, Using Regional Growth Rates Year Mainland Tanzania Dar es Salaam Other urban areas Rural areas 1992 (HBS) 0.386 0.281 0.287 0.408 (0.021) (0.028) (0.050) (0.024) 1993 0.400 0.325 0.296 0.421 1994 0.419 0.300 0.306 0.444 1995 0.407 0.364 0.297 0.427 1996 0.392 0.378 0.281 0.410 1997 0.389 0.382 0.281 0.406 1998 0.380 0.374 0.264 0.398 1999 0.369 0.374 0.258 0.385 2000 0.350 0.391 0.257 0.361 2001 0.333 0.404 0.233 0.343 2002 0.323 0.386 0.228 0.333 2001 (HBS) 0.357 0.176 0.258 0.387 (0.016) (0.027) (0.022) (0.020) Distributions were simulated going forwards, taking the initial distribution from the 1991/92 HBS survey data. Estimates from the 2000/01 HBS are shown for comparison purposes. Estimates from Household Budget Surveys are shown with standard errors in parentheses. Appendix Table A8: Simulated Headcount Rates, Using Regional Growth Rates, Urban Rate for Dar es Salaam Year Mainland Tanzania Dar es Salaam Other urban areas Rural areas 1992 (HBS) 0.386 0.281 0.287 0.408 (0.021) (0.028) (0.050) (0.024) 1993 0.400 0.327 0.296 0.421 1994 0.421 0.347 0.306 0.444 1995 0.405 0.330 0.297 0.427 1996 0.389 0.313 0.281 0.410 1997 0.384 0.300 0.281 0.406 1998 0.374 0.279 0.264 0.398 1999 0.362 0.261 0.258 0.385 2000 0.341 0.243 0.257 0.361 2001 0.322 0.220 0.233 0.343 2002 0.312 0.206 0.228 0.333 2001 (HBS) 0.357 0.176 0.258 0.387 (0.016) (0.027) (0.022) (0.020) Distributions were simulated going forwards, taking the initial distribution from the 1991/92 HBS survey data. Estimates from the 2000/01 HBS are shown for comparison purposes. Estimates from Household Budget Surveys are shown with standard errors in parentheses. 35 Appendix Table A9: Simulated Headcount Rates by Region, Based on Backwards Simulation from 2001 Region 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 (HBS) Std. Error (2001) Dodoma 0.42 0.43 0.44 0.43 0.42 0.40 0.40 0.36 0.37 0.34 (0.06) Arusha 0.39 0.40 0.41 0.39 0.39 0.40 0.40 0.40 0.39 0.39 (0.07) Kilimanjaro 0.41 0.41 0.41 0.41 0.40 0.38 0.36 0.34 0.34 0.31 (0.06) Tanga 0.54 0.55 0.55 0.54 0.53 0.51 0.50 0.47 0.41 0.35 (0.06) Morogoro 0.40 0.43 0.44 0.44 0.43 0.43 0.40 0.35 0.39 0.29 (0.03) Pwani 0.48 0.48 0.47 0.48 0.48 0.46 0.46 0.45 0.46 0.46 (0.08) Dar es Salaam 0.10 0.12 0.11 0.13 0.17 0.17 0.15 0.14 0.18 0.18 (0.03) Lindi 0.60 0.60 0.60 0.60 0.59 0.58 0.58 0.56 0.53 0.53 (0.14) Mtwara 0.69 0.69 0.69 0.68 0.65 0.64 0.48 0.40 0.42 0.38 (0.04) Ruvuma 0.38 0.38 0.39 0.38 0.38 0.38 0.38 0.37 0.37 0.41 (0.08) Iringa 0.44 0.46 0.47 0.45 0.43 0.43 0.41 0.38 0.38 0.29 (0.05) Mbeya 0.26 0.26 0.27 0.26 0.25 0.25 0.25 0.23 0.23 0.21 (0.05) Singida 0.53 0.55 0.58 0.55 0.52 0.50 0.50 0.49 0.49 0.55 (0.05) Tabora 0.19 0.20 0.21 0.21 0.20 0.21 0.21 0.21 0.24 0.26 (0.04) Rukwa 0.20 0.21 0.25 0.21 0.21 0.21 0.22 0.21 0.24 0.31 (0.04) Kigoma 0.28 0.29 0.35 0.36 0.36 0.37 0.39 0.40 0.42 0.38 (0.04) Shinyanga 0.42 0.46 0.53 0.47 0.46 0.47 0.57 0.55 0.45 0.42 (0.07) Kagera 0.31 0.32 0.32 0.32 0.32 0.32 0.32 0.31 0.29 0.29 (0.09) Mwanza 0.64 0.66 0.69 0.68 0.66 0.68 0.63 0.62 0.55 0.48 (0.06) Mara 0.49 0.51 0.52 0.52 0.51 0.51 0.49 0.48 0.44 0.46 (0.08) Distributions were simulated going backwards taking the final distribution from the 2000/01 HBS survey data. Simulations are based on regional GDP growth estimates calculated from regional GDP level estimates and regional population growth rates. Regional average population growth rates were calculated using 1988 and 2002 census counts. Appendix Table A10: Simulated Headcount Rates, Using Regional Growth Rates, Adjusting for Inequality Using Datt-Walker Method Year Mainland Tanzania Dar es Salaam Other urban areas Rural areas 1992 (HBS) 0.386 0.281 0.287 0.408 (0.021) (0.028) (0.050) (0.024) 1993 0.401 0.330 0.296 0.421 1994 0.421 0.353 0.306 0.444 1995 0.406 0.348 0.297 0.427 1996 0.390 0.341 0.281 0.410 1997 0.387 0.338 0.281 0.406 1998 0.376 0.313 0.264 0.398 1999 0.364 0.300 0.258 0.385 2000 0.343 0.283 0.257 0.361 2001 0.325 0.270 0.233 0.343 2002 0.315 0.250 0.228 0.333 2001 (HBS) 0.357 0.176 0.258 0.387 (0.016) (0.027) (0.022) (0.020) Distributions were simulated going forwards, taking the initial distribution from the 1991/92 HBS survey data. Estimates from the 2000/01 HBS are shown for comparison purposes. Estimates from Household Budget Surveys are shown with standard errors in parentheses. 36 Appendix Table A11: Simulated Headcount Rates, Two Survey Approach, Using National Distribution Year Mainland Tanzania Dar es Salaam Other urban areas Rural areas 1992 (HBS) 0.386 0.281 0.287 0.408 (0.021) (0.028) (0.050) (0.024) 1993 0.405 0.283 0.297 0.429 1994 0.419 0.283 0.302 0.446 1995 0.414 0.270 0.299 0.441 1996 0.405 0.250 0.297 0.432 1997 0.404 0.247 0.297 0.430 1998 0.397 0.225 0.288 0.425 1999 0.389 0.211 0.287 0.416 2000 0.375 0.188 0.281 0.401 2001 0.358 0.175 0.258 0.385 2002 0.339 0.140 0.242 0.367 2001 (HBS) 0.357 0.176 0.258 0.387 (0.016) (0.027) (0.022) (0.020) Distributions were simulated going forwards, taking the initial distribution from the 1991/92 HBS survey data. Estimates from the 2000/01 HBS are shown for comparison purposes. Estimates from Household Budget Surveys are shown with standard errors in parentheses. Appendix Table A12: Simulated Headcount Rates, Two Survey Approach, Using Within Stratum Distribution Year Mainland Tanzania Dar es Salaam Other urban areas Rural areas 1992 (HBS) 0.386 0.281 0.287 0.408 (0.021) (0.028) (0.050) (0.024) 1993 0.390 0.313 0.308 0.408 1994 0.400 0.325 0.326 0.416 1995 0.380 0.293 0.319 0.396 1996 0.374 0.280 0.316 0.388 1997 0.382 0.265 0.308 0.401 1998 0.386 0.243 0.297 0.409 1999 0.377 0.214 0.288 0.401 2000 0.374 0.189 0.281 0.401 2001 0.358 0.181 0.259 0.385 2002 0.343 0.138 0.242 0.372 2001 (HBS) 0.357 0.176 0.258 0.387 (0.016) (0.027) (0.022) (0.020) Distributions were simulated going forwards, taking the initial distribution from the 1991/92 HBS survey data. Estimates from the 2000/01 HBS are shown for comparison purposes. Estimates from Household Budget Surveys are shown with standard errors in parentheses. SMALLHOLDER GROUND WATER IRRIGATION IN TANZANIA Shiva S. Makki, IJsbrand de Jong, and Henry Mahoo The purpose of this note is to explore the potential for smallholder ground water irrigation systems to complement the ongoing river basin based irrigation projects. We would examine various ground water irrigation alternatives, including open/dug wells, borehole tubewells, and rainwater harvesting, for their economic and environmental viability in Tanzania. The motivation for this note comes largely from the experiences in several Asian countries where ground water irrigation made a significant contribution towards poverty alleviation, rural employment, and food security. In Bangladesh, for example, ground water represents nearly 70% of total water used in irrigation (FAO Aquastat). According to a 2002 survey, there were a million tube wells in the country. Similarly, ground water represents over half of irrigation water use in India, a third in Pakistan, and about 20% in China. In many of these countries, the irrigated land accounts for about a third of total cultivated land, but the irrigated agriculture accounts for up to 75% of food production. Among Sub-Saharan African countries, only Nigeria, South Africa, and Zimbabwe have successfully expanded ground water irrigation systems to support smallholders. In Nigeria, for example, the National Fadama Development Project, involving tubewell-pump package to small farmers, successfully boosted farm incomes and contributed to significant reduction in poverty among the project participants (World Bank, 2002). It is in this context we argue that small scale ground water irrigation systems could have a wide-ranging and profound impact on the lives of poor people who depend on agriculture for livelihood and help towards transforming the Tanzanian agricultural sector from subsistence activity to a more income generating activity.1 1. Irrigation ­ A Necessary Input for Enhancing Farm Productivity, Stabilizing Food Production, and Commercializing Agriculture In recent years, the demand for food has grown rapidly in Tanzania due to increases in population, which is growing at 2.5% annually, and per capita income, which is growing at 4% annually. If this trend continues, Tanzania will have to depend on food imports or food aid to meet its food demand. For example, a recent study estimates that without increases in rice productivity, Tanzania may need up to $70 million in foreign exchange to overcome the rice shortage in the country by 2017 (JICA, 2004). The study also estimates that irrigation can increase farm incomes by 3 to 4 times in a single crop season. 1A combination of high rates of investment in crop research, infrastructure and market development, and appropriate policy support fuelled green revolution in Asia. But major increases in food production occurred in irrigated areas. There is also a strong positive link among investment in irrigation, poverty alleviation, and food security. In India, for example, 69 percent of people in non-irrigated areas are poor, while only 26 percent of people in irrigated areas are considered poor. Agricultural productivity growth in Tanzania has not kept pace with demand for food. The agricultural productivity remains low because of its dependence on fluctuating and erratic rainfall, the use of inferior quality technology, and inefficient agricultural support services, including extension services, access to markets, access to credit, storage facilities, and food processing. Until recently, growth of food production came primarily from increases in area cultivated, but as good land becomes less available, the country will be forced to increase crop and livestock yields. Both rainfed and irrigated agriculture will need to be intensified, but irrigated agriculture has a higher potential for greater productivity. 1.1 There is significant potential to expand irrigated agriculture Of the total 44 million hectares suitable for agricultural production in Tanzania, only 10 million ha is under cultivation and out of this only 200,000 ha is irrigated. This represents mere 2 percent of total cultivated area in the country. The estimated irrigation potential is up to 2 million hectares. As much as 75 percent of irrigated land is farmed by smallholders in some 600 small-scale irrigation schemes, typically using small diversions and furrows in the highland areas, as well as small diversions for paddy production in the lowland areas. In addition, there are substantial areas where smallholder farmers practice traditional systems of flood recession or water harvesting for paddy production. 1.2 Crop yields are higher under irrigation The roles played by irrigation are that of improving the quality of output, reducing risks associated with production and hence stabilizing supply of outputs. These factors manifest themselves in higher agricultural productivity, increased outputs and higher rural incomes. For example, in the two world bank supported smallholder irrigation development project areas, rice and maize yields and total production have increased significantly under irrigation (Table 1.1). Average rice yields has more than doubled from 1.98 t/ha to 5.27 t/ha in Pangani, and from 1.46 t/ha to 4.06 t/ha in Rufiji. Average maize yields have also increased from 1.06 t/ha to 4.86 t/ha in Pangani, and from 1.06 t/ha to 4.86 t/ha in Rufiji. Table 1.1 Crop Yields (tones/hectare) in the RBMSIRP Project Area Crop Yields (tonnes/hactare) Corp Pangami Rafiji Rainfed Irrigated Rainfed Irrigated Rice 1.98 5.27 1.46 4.06 Maize 1.06 4.86 1.06 3.30 Tomatoes 2.00 4.00 2.00 3.00 Onions 2.50 4.00 2.50 3.50 Source: ICR River Basin Management and Smallholder Irrigation Improvement Project, 2004 2 1.3 Irrigation improves returns to labor and land In addition to yield improvement, irrigation schemes, have enabled farmers to have stable crop production from year to year and has provided the conditions suitable for economic use of agricultural inputs such as high yielding varieties, fertilizers and herbicides. Access to water could provide a more favorable environment for crop and livestock intensification, which is crucial to increasing returns to their only assets ­ labor and land Historically, farmers in Tanzania were reluctant to invest in costly inputs such as improved seeds, fertilizers, and pesticides under dryland conditions. Irrigation could make the use of better quality inputs economically viable and increase returns to farm labor and land. For example, evidence for maize and beans suggests that the gross margin or net revenue is nearly 4 times larger under irrigated agriculture compared to production under dryland agriculture. Table 1.2 Economics of Irrigation in Tanzania, 1997 Maize Rice Beans Dryland Irrigated Dryland Irrigated Dryland Irrigated Yield (kg/ha) 1,100 2,550 2,975 12,350 400 980 Revenue ($/ha) 165 383 120 200 256 627 Gross Margin ($/ha) 56 205 28 174 116 413 Source: FAO, 1997 for Maize and Beans; WG2 TF report for Rice 1.4 Irrigation facilitates the shift from subsistence to commercial agriculture Stable production systems are necessary for the commercialization of agriculture sector. In a country where droughts and floods are frequent, irrigation holds the key for stabilizing agricultural production. Investments in irrigation is, therefore, central to the development of Tanzania's agricultural sector and its transformation from subsistence farming system to commercial, profitable production systems. Stable water supply would promote a more flexible production regime which would reduce price variability and, thus, income variability of poor households. Irrigation water would reduce production risks and uncertainties and enable farmers to raise high-value cash crops, such as rice, sugarcane, vegetables, flowers, and fruits. It also improves the capacity of farmers to respond to market requirements through reliable and diversified supplies. This could reduce the vulnerability of poor households and sustain a more stable economic growth in the rural areas. Fresh fruits, vegetables, and cut flowers offer major growth opportunities for smallholders under irrigation. FAO data shows that the exports of fruits and vegetables increased nearly 3 times between 1990/91 and 2002/03. The share of fruits and vegetables in total agriculture exports nearly doubled during the same period. To a large extent this was facilitated by improvements in irrigation. 3 1.5 Full benefits from irrigation depend on supporting infrastructure High value agricultural and horticultural products will serve as a driving force for boosting small farm incomes on a broad scale. However, shifting from subsistence crops to high value crops require a several critical inputs including improved/quality seeds, fertilizers, pesticides, management know-how, information and financial services and, most importantly, a stable supply of irrigation water. Such a change may also require developing stable linkages to output markets, both domestic and foreign. It may also require strengthening and realigning the agricultural research and extension systems to make them more responsive to the needs of farmers. Irrigation also possesses the unique feature of being developed indigenously as opposed to the other inputs (e.g. fertilizers, pesticides, machinery) which can be physically transported (through imports) from countries. Given the existing state of irrigation technology, a country-specific, irrigation development strategy has to be implemented in order to reap the benefits of irrigation in agricultural production and ultimately, poverty reduction. 2. History of Irrigation Development in Tanzania From the mid 1960s to the mid 1980s, a period marked by strong central government policies, irrigation development was focused on capital-intensive full-control irrigation scheme development on parastatal farms. This continued to the early 1990s during which time the traditional smallholder sector received a meager share of total irrigation investment. Rice is by far the most important crop irrigated in Tanzania, but sugarcane is also irrigated. Traditional irrigation schemes that use water harvesting and simple diversion structures, account for the bulk of the irrigated rice area. In addition, there are traditional schemes that have been upgraded, new smallholder schemes, irrigated parastatal farms and a few private sector irrigated farms. Irrigation is now seen as an important aspect of Government's agricultural development strategy to improve food security, increase farmers' productivity and incomes, and to produce higher valued crops such as flowers, fruits and vegetables. 2.1 Donor Supported Irrigation Projects A large number of projects are underway to develop irrigation infrastructure in Tanzania. In assisting the government, donors are playing a critical role in the overall development of the irrigation subsector. Tanzania has identified nine river basins for irrigation water and hydropower generation. In addition, there are also projects involving shallow surface water and ground water. The river basin surface water irrigation represent nearly 90% of the irrigated area in Tanzania. In terms of monetary support, however, the support to irrigation sub-sector has been very limited compared to other sectors of the economy. 4 · The World Bank-funded River Basin Management and Smallholder Irrigation Improvement Project (RBMSIIP), which started in 1996, covers the Pangani and Rufiji River Basins. The project aims to: (i) strengthen Government's capacity for water basin management; (ii) address environmental concerns; (iii) support the downsizing of the irrigation support function of the MAC through privatization; and (iv) rehabilitate and upgrade traditional smallholder irrigation schemes. RBMSIIP was the first IDA investment in small holder irrigation in Tanzania. The project spearheaded the review of irrigation policy and laid the foundation for strengthening the institutional and legal framework for river water management. This project also created much enthusiasm among the stakeholders to participate in various water management interventions. The irrigation component of the project contributed significantly in increasing crop yields and farm incomes. 5317 farm families, covering 5.059 hectares, benefited from the project. Crop yields increased by 2 to 4 times, while farm incomes increased by 3 to 3.5 times in the project area. The water use efficiency also increased substantially. The overall Economic Rate of Return (ERR) for the project is estimated to be 16%, while the ERR to small holder irrigation improvements was estimated to be 10%. The project also had a substantial impact on institutional development. Tanzania now has a national water policy that promotes sustainable use of water resources. · The IFAD-funded Smallholder Development Project for Marginal Areas (SDPMA) contained a smallholder irrigation development component, as does the Mara Region Farmers' Initiative Project which will bring up to 1,200 ha of irrigation to smallholders through the construction of small earth bunds for harvesting rainwater. Despite several institutional problems, the project benefited a significant number of farmers, both in terms of productivity increase and income gain through the production of high value cash crops such as paddy and vegetables. The crop yields increased by up to50% in some areas. The increased crop yields, combined with low construction costs achieved in the SDPMA, made the small holder irrigation economically attractive and financially viable in Tanzania. The project demonstrates that the financial viability is achieved by establishing a cost recovery system in which the beneficiaries pay full operation and maintenance costs, and a part of capital costs of the irrigation system. · USAID assisted in the construction of the Bahi scheme (20 ha) and the small Kintinku scheme during 1980-82 period. These small projects paved the way for further development of simple water harvesting for the supplementary irrigation of rice in marginal areas. · The UNDP/FAO-funded Institutional Support to Irrigation Development (ISID) project developed guidelines on possible privatization of Irrigation Department activities and the sale of irrigated parastatal farms during 1987-94 period. · The Development Association of the Netherlands (SNV), has been supporting small- scale irrigation development in Dodoma Region through the Small-Scale Irrigation Development Program (SSIPDO), and more recently in Kilimanjaro and Arusha 5 Regions through the Traditional Irrigation Improvement Program (TIP). This program will be made into an NGO and will concentrate on sustainable land use in river basins, on the empowerment of water user associations (WUAs), and on organizational activities at district and micro catchment levels. · ADB in association with UNCDF has supported two high-cost irrigation schemes through the Smallholder Irrigated Rice Project (Kitivo and Murampuli schemes) in Tanga Region and are continuing to develop the Madibira Smallholder Irrigation Project in Mbeya Region. · Other agencies involved in irrigation development are JICA, which is constructing the high-cost Lower Mushi Smallholder Irrigation Project and preparing a new scheme near Bagamoyo. · Three experimental smallholder irrigation development projects are being supported by UNDP in Zanzibar, Kilimanjaro and in Mbeya, and the Italian Government is funding the Humbolo irrigation scheme in Dodoma Region. · The DANIDA-financed Agricultural Sector Support Program includes a smallholder irrigation improvement component, focusing on improving the management of water resources and crop husbandry practices in existing smallholder irrigation schemes in Mbeya, Iringa and Morogoro Regions. · The FAO Special Program for Food Security (SPFS) started in Tanzania in 1995 in Dodoma and Morogoro Regions. It covers 16 villages and is aimed at increasing production and productivity of the two main staple crops, rice and maize, through the participatory demonstration of improved farming technologies, especially crop intensification. Part of the program includes a water control component which covers the demonstration of low-cost methods of irrigation rehabilitation and improved systems of water management. · NGOs have become increasingly important in irrigation development. International NGOs with projects in horticulture or irrigation, include CARE-Tanzania, SNV, TechnoServe (USA), CONCERN, the Irish Foundation for Co-operative Development (IFCD) and World Vision. National NGOs include the Presidential Trust Fund (PTF), the Tanzania Youth Development Employment Foundation, the Arusha Diocese Development Office, and the Mission for the Needy. · Participatory Agricultural Development and Empowerment Project (PADEP). Support to demand driven community based sub-projects including watershed management and improvement of traditional irrigation schemes. The development of irrigation systems in Tanzania have been slow due to little investments to develop irrigation infrastructure, the lack of affordable of irrigation technology and equipment, and the lack of other production (e.g. HYVs) and post harvest technologies (e.g. processing, storage) that could make investments in irrigation worthwhile. Studies have also 6 shown that lack of local capacity to manufacture and service irrigation equipment is also a major hindrance small scale irrigation development in the country. A recently established government working group on irrigation in its first report on investment needs for irrigation, identified following constraints to development of the irrigation sub-sector in Tanzania: · The absence of a national irrigation policy and strategy to guide the direction of all styles of irrigation; · Intensifying competition for water in key river basins; · Limited financial resources available for smallholder irrigation development; · Limited access to markets by and availability of credits to smallholders; · Limited capacity, both within the private and public sector, to promote smallholder irrigation development; · Limited choices of irrigation technologies and crops available to smallholders, in particular those technologies that are affordable to the poor; and · Inadequate extension services in the area of high value crop production under irrigation. Table 2.1 Major Existing/Recently Completed Irrigation Projects in Tanzania Project or Program Donor Remarks Agriculture Sector Program Support DANIDA Progressing well, still under implementation. II (ASPS) Operating in Mbeya and Iringa regions covering schemes namely, Utengule Usongwe (Mbeya Rural), Naming'ongo (Mbozi), Nyanzwa and Irindi (Kilolo). Private Agriculture Sector Support DANIDA On going (PASS) Special Program in Food Security FAO/ ADB On going in selected districts (SPFS) Participatory Irrigation Development IFAD, Implementation continuing until 2006 in the Program (PIDP) Ireland, WFP marginal areas (Dodoma, Singida, Tabora, Manyara and Shinyanga). Performing well and drawing on experiences of previous projects. Participatory Agricultural WB On going. Involved in very small demand driven Development and Empowerment irrigation schemes. Project (PADEP) National Irrigation Master Plan. JICA Finalizing past support and nearing completion. Lower Moshi, Ndugu & Mwega JICA Completed, schemes are performing well except Projects Lower Moshi due to water crisis. River Basin Management and WB Good, all schemes Completed in June 2004 and Smallholder Irrigation Improvement performing well. Preparation underway for possible (RBMSIIP) phase II project. LAMP - Soil and Water SIDA Completed Conservation Source: ASDP Working Group 2 Report on Irrigation development in Tanzania 7 3. Smallholder Ground Water Irrigation A suitably designed ground water irrigation system could reduce reliance on large bodies of water, including rivers and lakes, and promote a more sustainable use of locally sourced and managed irrigations systems. Since the surface water availability varies with the rainfall, open wells and borehole or tube-wells can be constructed to spread the availability of water throughout the growing season and also facilitate intensification of agriculture, including crop and livestock production. Compared to large surface irrigation schemes whose design is driven by topography and hydraulics, ground water development is often much more amenable to poverty targeting and is generally less capital intensive. Smallholder ground water irrigation, perhaps, provides the best opportunity yet for enhancing the livelihoods of the poor in Tanzania. 3.1 Ground Water Availability The total renewable water resources in Tanzania is estimated to be around 80 cubic km/year, of which 30 cubic km/year are ground water resources (FAO, 2004). Ground water availability is mainly defined by geology and climate. Several hydrological studies commissioned by the World Bank, FAO, and UNDP indicate that the water table varies significantly across the country, but there are many potential locations where water tables are shallow and water yield is quite significant (Hydrological Map of Tanzania, 1990). Some of high ground water potential areas identified by FAO include: · Makutupora in Dodoma region and Ruvu basin in coast region · Sanya-Hale plain in the Pangani basin · Arusha and the Karoo Sandstone in Tanga region · Fault zones around Kilimanjaro · Parts of Morogoro, Iringa, and Mbeya, Mtwara, and Lindi Volcanic areas of northern and southern Tanzania as well as the sedimentary coastal basins are potential groundwater resource areas. Boreholes drilled in the volcanic areas have yields up to 800 cubic meters per hour and those in sedimentary coastal areas yield about 50 cubic meters per hour. An hectare of rice-paddy requires about ____ cubic meters of water during the entire season. Groundwater is a major supplement for surface water for many parts of the country and is a vital source of water in semi-arid water scarce areas. Tanzania's groundwater is an important resource locally for rural, urban, industrial and irrigation supply. Cumulatively, the lakes, wetlands and aquifers provide huge natural storage capacity. The country also has 2.7 million hectares of wetlands (Usangu, Malagarasi, Ugalla, Kirua Swamp and others on rivers Suiwe, Mara, etc). The country's water resources offer enormous opportunity for poverty reduction in rural areas and prospects for economic growth nationwide. 8 3.2 Ground Water Use: Abstraction and Distribution · Open wells. Ground water is extracted from open wells or borehole wells, both with varying capacity. Open wells are dug by providing a stone wall or concrete ring to a suitable depth. Over a wide part of the country the ground water development has concentrated mainly on shallow open wells for domestic purposes. The ministry of agriculture in Tanzania estimates that the cost of digging a well is about $3,000 (in 1996/97). · Borehole tubewells. Shallow tubewells can be drilled by hand with simple soil auger-type tools, by power rotary drilling, or with a drilling method called "jetting" or "washboarding." Boreholes of about 6 inches diameter rig are drilled to a depth of 6 to 30 meters with suitable casing or strainer pipes. · Rainwater harvesting. Since the country receives reasonably good rainfall during two seasons a year, it is not unreasonable to devise methods to harvest rainwater. Runoff rainwater can be stored behind bunds or tanks and can used for irrigation later. In areas with vast terrain and gentle slope, it is possible to construct small and medium tanks to collect and store during heavy rains for supplemental irrigation during rainy season and full irrigation during the dry season. These tanks can also help recharge aquifers, which feed into open or tube wells. The average cost of a tank to store about 30,000 cubic meter of water is about $4,000 which will irrigate about 2.5 hectares of paddy or 5 hectares of other crops such as fruits and vegetables. Irrigation equipment The choice of pump sets depend on various criteria including size of flow, depth of water, and availability of electricity. There are several manufacturers of these machines around the world, but most accessible ones are in India and South Africa. The cost of these machines vary across regions. A typical submersible pump costs about $750 in India. The cost of drilling a 10 meter well with 6 inch rig costs about $800 in India. Some of the pumps are run by electricity, while some are by diesel. · Mechanical pumps: Submersible pump sets (4 inches and 6 inches size); Centrifugal Deep well turbine Propeller pumps · Manual pumps: Treadle pumps (manual pump, costs about $50 to $100 per equipment). MoneyMaker pumps (manual pump, costs about $12 to $25 per equipment). Hand pumps 9 · Application equipments: Drip irrigation Surface/gravity flows Sprinkler irrigation Cost of irrigation systems The cost of irrigation systems in Tanzania is currently very high and smallholders cannot afford to invest in them. For example, the cost of digging an open well is about $3,000 to $5,000. In addition, irrigation equipment (pumps and pipes) also add to the cost (up to $5,000). In recent years, drilling of borewells for irrigation is becoming popular because of the ease with which they can be drilled. Borewells are more efficient ­ both in terms of economics and success rates ­ compared to open dugwells. However, borewells are still more expensive in Tanzania compared to South Asia (Table 3.1). Table 3.1 Costs of installing and maintaining irrigation systems Irrigation Costs Tanzania West Bengal Tamil Nadu Kenya Open wells digging $3,000-$5,000 Tubewell Drilling/rig installation $600 to $750 $300 to $500 Rainwater harvesting* $4,000-$5,000 Pumpsets, etc. $2,000-$5,000 $600 to $750 $400 to $600 Diesel/Petrol pump $500 - $5,000 Annual Maintanance/well $100 to $200 Treadle Pumps $50 - $100 $15 to $25 MoneyMaker Pumps $12 to $25 Drip Irrigation Systems $20 - $200 Irrigation Costs per hectare Surface Water $7,000 to $10,000 Ground Water $1,500 to $2,500 *tank to store about 30,000 m3 of water Cost of drilling and pump decreases with demand In South Asia, rapid groundwater development has supported a booming pump industry, which is characterized by both economies of scale and intense competition. As a result, South Asia's rural poor have benefited from low costs of drilling and pumps sets. In Tanzania, in contrast, pump irrigation development is so slow and limited that costs of drilling and pumps are high and quite beyond the reach of smallholders (Table 3.1). Researchers from UK's Cranfield university found that "In Africa the cost of a borehole 10 drilled by a truck-mounted rig can be extremely high costing as much as 10­20 times the cost of the drilling and pump in Asia. High unit costs mean that too few wells are drilled and communities and farmers remain dependent on international aid programs for this form of infrastructure development" (Carter 1999). In Tanzania, the groundwater potential remains almost untapped. In Nigeria, the cost of constructing shallow tubewells was reduced by about two-thirds between 19983 and 1990, with a commensurate increased returns on tube well investments. 3.3 Ground Water Management Tube wells and open wells are easy to manage and can make use of manual pumps. Each well could potentially irrigate up to 5 hectares. These types of irrigation also avoids some of the management challenges posed by previous large surface irrigation schemes in Tanzania. The maintenance of the pump/well is local and farmers have direct stake in its upkeep and usage. Individual or a group of farmers could invest in these types of wells. Efficient distribution methods such as hand or treadle pumps and drip irrigation could be used to lift and distribute water (since pumping water involves direct costs to farmers, manual or electricity, it would promote efficient use of water). These new irrigation options would boost peri-urban agriculture and small and medium sized farmers. Well ownership and management · Private ownership: The well owners enjoy absolute rights to use water in their wells in any manner and in any quantity they like. Mostly used by large farmers. · Co-operative ownership: In an irrigation project in West Bengal, India, local leaders took the initiatives to help small and medium size farmers with irrigation. A cluster of tubewells were established with financial assistance from the World Bank. The project benefited 300 farmers covering an area of 432 acres. The project was managed by a cooperative society. Cost and benefits of the project are below: Table 3.2 Cost of Smallholder Irrigation Project, West Bengal, India Items Cost/Number Total cost of the project ($): $53,185 Annual maintance cost ($) $3,248 Number of Tubewells 30 Command Area of the Project (Acres): 432 Number of Beneficiaries 300 Water user fee ($/cropping season/acre) $48 Cost per well ($) $1,773 Cost per Acre ($) $123 Source: West Bengal Irrigation Project, World Bank, 1994-95 11 · Government ownership: Another alternative would be for the local government to manage the irrigation project ­ government could dig wells and install irrigation pumps; farmers are assessed a fee per acre; beneficiaries are identified by the local government committee appointed by elected officials. · The comparative appraisal of the utilization of irrigation systems under different types of management reveals that the performance is better under co-operative management than the other two types. Under cooperative management water is allocated with the active participants of the farmers who have similar needs and clear boundaries. They had a sense of ownership and responsibility to maintain and operate the irrigation structure. Watershed management and recharge of ground water Pumping from shallow aquifers by many wells, in an attempt to irrigate a larger area than possible given the average groundwater recharge, can reduce the water supply of neighboring tubewells, lower water tables, diminish economic returns, and degrade water quality. Policies to calibrate the spread of wells to the groundwater recharge potential are critical to sustainability. It is normally the responsibility of local governments to monitor and evaluate groundwater and develop rules for groundwater abstraction. The simplest regulation consists of promoting and enforcing a socially accepted compact for minimum spacing and maximum pump capacity, rather than a full-fledged groundwater use rights and obligations system separated from land ownership rights. 3.4 Institutional Framework Irrigation infrastructure is necessary to boost the agricultural production, but may not be sufficient to make agriculture profitable. In order for the smallholder ground water irrigation to be economically viable and environmentally sustainable, in addition to improving irrigation infrastructure, the country will need to deal with the whole host of constraints facing agriculture sector. Effective institutions are vital for the success and sustainability of ground water irrigation systems. To promote expansion of smallholder irrigation, poor farmers must have access to cost- effective irrigation technologies that provide a rapid return on investment, a reliable and quality supply of improved crop varieties and other inputs, land tenure rights, and markets to absorb increased production. Public and private investments in the assessment of the supply and quality of groundwater, in the regulation of ground water abstraction, and in technical support to farmers are vital to allow ground water irrigation to go on reducing poverty. This may require institutional intervention designed to provide technical know-how, support agriculture research and extension, and develop markets and infrastructure (Table 3.3). 12 Table 3.3 Institutional Framework for Sustainable Development of Smallholder Irrigation Systems Required Conditions Targets Technical Self- - Capacity building of Irrigation staff, LGA Staff and Extension Workers reliance - Raising of technical knowledge of farmers on O & M and water management - Application of appropriate irrigation development level - Execution of environmental conservation Financial Self- - Improvement of government financial situation by reform of taxation system reliance - Strengthening and support of micro-finance to farmers - Enlargement of opportunity on private sector investment in irrigation development Institutional/ - Definition on roles and responsibility of Irrigation Section, LGA and IOs under Organizational decentralization Strengthening - Strengthening/reform of Irrigation Section, ZIUs and LGAs - Legal framework strengthening for IOs (legal status, land tenure, water right, ownership and responsibility of irrigation infrastructure) - Institutional strengthening for raising technical capability (extension services and training) - Institutional strengthening for raising financial capability (collection of water fee and O & M cost, micro-finance) - Promotion and support program of private sector (creation of attractive climate for investment, tax incentives for BOT, long and stable security of tenure) Source: ASDP Working Group 2 Report on Irrigation development in Tanzania Improving technical know-how · Agricultural research and extension is crucial in generating and disseminating appropriate technologies for smallholders who adopt intensive production systems with irrigation. These technologies include not only low cost irrigation equipment and effective water management methods, but also other technologies such as improved crop varieties and animal breeds, and fertilizers, that can bring significant returns to investments in irrigation. Similarly, rural finance, post-harvest technologies, and marketing are all critical to make the investments in irrigation is economically viable. · Improved and cost effective water-lifting technologies with a higher discharge than traditional methods may facilitate expansion of irrigated agriculture. Currently, both drilling and equipment costs are high in Tanzania. The irrigation equipment could be obtained from Asia at a relatively lower cost compared to the current import costs from Europe and South Africa. The comparable equipment in Asia would cost only a third of the currently imported prices (FAO, 1997). It is also difficult to get local servicing of irrigation equipment in Tanzania as local capacity is limited. 13 · Surveying, drilling test wells, water sampling, and water level monitoring are useful to build a database and track long-term trends. The rational management of the groundwater resource is difficult without a basic understanding of the distribution and yields of aquifers, and their vulnerability to pollution and overdraft. These monitoring activities are usually a government responsibility, but local authorities or communities can carry out some of the work. · Training and extension are critical to facilitate good installation, operation, and maintenance of tubewells and for the development of local capacity for maintaining and repairing wells and pumping equipment. To optimize the benefits of tubewell investments, extension and training will be needed for irrigation water management, improved agricultural technology, and marketing systems. Strengthening of internal capacity is a key ingredient for the success of smallholder irrigation systems. · To make the smallholder ground water irrigation sustainable, there is a need for simultaneous program to undertake watershed management and ground water recharge. Identifying watershed area may require public sector or community role as individual small holder may not find such an activity feasible. Ground water recharge may also involve large area and several communities. Improving access to land and land tenure · Land ownership may also represent a major constraint: lack of proper land titles may limit access to credit, prevent investments on irrigation and land improvements. · Land tenure systems are a potential impediment to the expansion of high value agriculture and horticulture, since if farmers do not own the land they will be reluctant to invest in permanent irrigation structures such as tubewells and sprinkler sets. · Conflicting land use can also be a problem, especially when dry season irrigation encroaches on lands traditionally used by herders. · Private investments in ground water irrigation requires changes in many land institutions including land leasing, renting, and enforcement of rights. Without recognized ownership rights, investors would not make investments that may take many years to recover. Improving input and output markets Irrigation should be complemented by developments in input and output markets. In order to have an impact on growth and poverty reduction irrigation infrastructure need to be integrated with an entire range of complementary measures including better access to markets, processing, storage, exports, inputs, and extension service. 14 · Irrigation is only useful for income generation in so far as they enable small farm holders to take advantage of market opportunities; The income generating potential of irrigation is directly related to the degree to which small farmers are integrated with input and output markets · Fertilizer, seed, and pesticides need to be adequately supplied by the commercial sector on the basis of market prices. If not, farmers often produce their own seed leading to reduction in yields as seed quality deteriorates. Vegetable seeds adapted to the prevailing environmental conditions are an important input that is frequently lacking. Subsidized inputs, typically fertilizer, have been shown to make these items less accessible to smallholders, because these benefits tend to be captured by larger farmers, creating scarcity. · Market outlets for irrigated production are imperative for a successful smallholder irrigation sub-sector. Proximity to markets or reliable transportation linkages must be present, particularly since horticultural products are perishable. Price cycles often accompany horticultural production, which may require additional investment in value-added production (such as drying) or better storage to smooth out supply. Of course, access to market information is also important. 3.5 Policy Framework to Develop and Manage Water Resources Over the past decade, the Government of Tanzania (GoT) has recognized the imperative to manage its water resources on a multi-sectoral and river basin basis. Through support of four Bank projects (USRP, RWSP, RBMSIIP and DAWASA), the GoT has made significant strides to lay the foundations of sound water resources management and to promote investments in water resources development in general and urban and rural water supply in particular. A new water policy, formulated in 2002, embraces integrated water resources management and development as a key framework to promote sustainable economic growth and poverty alleviation. Figure 3.1 highlights conceptually the linkage of the various water using sub- sectors (such as domestic supply, irrigation, hydropower, environment and other uses) with water resources management. A new water sector development strategy (with support from GTZ), legislation and institutional framework are presently being finalized. It is important to also integrate the water sector strategy with the various water relate projects and programs. The water sector strategy therefore needs to be a broad strategy for addressing water issues economy-wide. Currently, irrigation services are provided by the ministry of agriculture and food security through their Zonal offices that link with the Districts and Regions. Ministry of water and livestock development also plays an important role in the assessment of water resources and its allocation to alternative uses. 15 Figure 3.1 Policy Framework for Water Resources Management Source: CWRAS Aid Memoire, 2004 Policy reforms are required to reflect the needs of irrigation sub-sector: · Simplify administrative regulations and restrictions on marketing and trading of irrigation equipments; · Simplify the procedures (reduce import duty, provide import subsidy, etc) for importation of irrigation equipments ­ drilling machines, pumps, etc.; · Fund irrigation investments (micro finance, joint venture financing) · Additional reforms that have an impact on the performance of irrigated agriculture include trade liberalization, reforms of crop marketing boards, reforms in export and import tariffs. 3.6 Public-Private Partnership · Public-private division of responsibility is important in formulating policy for ground water development. Open well or tubewell investments are a private good that should be the responsibility of the beneficiaries of the investment. The public sector role should generally be limited to establishing a conducive policy and institutional environment for investment. Direct subsidies for tubewell drilling and operation are best avoided unless there is a compelling poverty reduction argument for the subsidies. One-off matching grants may be useful in situations of great poverty and poorly functioning financial markets. · Sharing of investment costs between private and public sector through public-private partnerships (PPP) would reduce the often heavy burden on the public budget. 16 Performance of irrigation would also benefit from the higher efficiency with which private operators are able to provide goods and services. Investments from private smallholder farmers would, thirdly, foster a sense of ownership and lay the foundations for sustainability. · In PPP, private funds complement government funds, land improvement tax, credit, and water charges. The comparative importance of the resources depends on the specific circumstances of each case. Models for private investment include direct investment in production (including investments by smallholder farmers), provision of irrigation services (including O&M of infrastructure), leasing of irrigation technologies, and infrastructure development. · Examples of successful PPP in irrigation are rare. Constraints are related to the weak legal framework, and the perception that agriculture is uncompetitive, poorly productive and risky. In addition, agriculture is often considered unattractive in comparison with other options, even if the macro framework for the economies were corrected. A favorable environment and support measures should be created where private investors feel comfortable about rates of return, and a steady cash flow comparatively early on in the process. The challenge is to identify the key building blocks of this environment. · Low-cost productive technologies must be available to smallholders in terms of both location and price and must correspond to their needs. A variety of ways of providing smallholders with access to these technologies have been used. For example, these may range from importation of treadle pumps to local manufacturing. In order for market-driven sustainable development to occur, all parties in the supply chain must make a profit. In the case of treadle pumps, tubewells, and improved piped water distribution systems, the manufacturers, installers, and the gardeners all benefit. 3.7 Water User Associations · Water user associations (WUA) in areas where groundwater irrigation predominates are valuable for organizing hardware and infrastructure maintenance. WUAs could hold community water rights and overview water use among its members. · For effective water resource management, aquifer management organizations under the umbrella of the river basin committee or authority is probably the best way forward on groundwater resource management. · Many WUAs are comparatively new and still weak. They require significant investments in capacity strengthening and in defining roles and responsibilities. Transferring irrigation management responsibility from government to water user associations demands the provision of sufficient support, and a comprehensively prepared capacity building program. 17 · Water management has to be structured so that governments are responsible for the main infrastructure, and farmers are responsible of local bodies, such as WUAs and distributary boards. · WUAs provide the mechanism for registering the demands and rights of all users in a particular catchment. Under the new water policy, they comprise legally constituted bodies drawing their membership from the various water users in a particular locality. They link in with individual users or Irrigation Organizations who represent their members. The role of Catchment water user organizations is to manage allocation of water resources at local level and the equitable allocation of these resources during problems such as droughts. They also serve to assist in overcoming local disputes. In the areas where WUAs have been formed, they have been found to be extremely useful in finding solutions to water management issues. 18 4. Conclusions and Policy Implications 5. Recommendations 6. References 19 TANZANIA: GROWTH , EXPORTS AND EMPLOYMENT IN THE MANUFACTURING SECTOR Vandana Chandra, Pooja Kacker and Ying Li PRMED Input for the CEM 2005 TANZANIA'S MANUFACTURING SECTOR: GROWTH, EXPORTS AND EMPLOYMENT INTRODUCTION Remarkably, over the past forty years, the share of Tanzania's manufacturing sector in its economy has remained virtually unchanged at around 8 percent of GDP (Table 1). For most of the first 30 years, government's experiment with state control and planning capped the share of the private sector in the economy at 12 percent. However, since the late 1980s, even after slow return to a market-based system, manufacturing's role in the economy is unchanged: the vision of structural change that could modernize this rural economy through diversification and industrialization remains distant. From a relatively small base, growth in manufacturing has picked up from about 2 percent per annum during 1990-1998 to 6 percent in 1999-03 (Table 2). In the last two years (2002-03), it rose to 8 percent p.a.. However, as growth in the other sectors has also accelerated, manufacturing's relative contribution to GDP, economic diversification and modern job creation is small. The overwhelming dominance of primary production, especially in the rural sector continue to challenge policy makers to think harder about what more needs to be done to unleash Tanzania's manufacturing potential to move closer to the vision of a modern economy. A vibrant manufacturing sector is critical for the Tanzanian economy for several reasons. First, as international experience has shown, manufacturing has been an important source of growth in several small countries, even in sub-Saharan Africa (see Boz ---- on Lesotho, Mauritius etc.), leading the way to a modern economy envisioned in the latest PRSP. Typically, a growing manufacturing sector triggers the development of ancillary activities and better-paying jobs vital for a modern economy. Second, a large manufacturing sector is necessary for export diversification to reduce Tanzania's dependence on external shocks caused by weather cycles and terms of trade fluctuations. The latter routinely affect its export earnings tied to a basket comprised primarily of primary commodities such as gold, and traditional agricultural products (coffee, cashews, tea etc..). Besides adding uncertainty to foreign exchange earnings, the adverse effects of terms of trade shocks also spillover to household incomes, especially of the rural poor. Weather shocks, such as droughts, frequently impact agricultural production that accounts for over 50 percent of GDP and the majority of livelihoods in this predominantly rural economy. As manufacturing is considerably less vulnerable to such external shocks, it is valuable as a potential source of growth, exports and jobs that can generate higher and more sustainable incomes. Why Tanzania's manufacturing sector has not yet developed into such a source of growth and what can be done to accelerate this process is the subject of this paper. 2 `Tanzania at the Turn of the Century' (CEM, 2001) examined the country's economic performance during 1960s ­ 1999 and concluded that the Government of Tanzania's (GoT's) intensification of macroeconomic policy reforms between 1995-1999, established a stable macroeconomic environment, especially low inflation, conducive for investment and growth. As GoT, also implemented structural reforms in parallel, many gross economic distortions that earlier constrained growth were eliminated effectively. The strengthening of private investment incentives was expected to boost private investment and establish a strong manufacturing sector. Since the late 1990s, FDI flows increased and were related largely with mining and privatization. The share or growth of private investment or FDI in manufacturing is not known. Overall, the share of private investment (and indeed total investment in GDP) has plummeted from over 20 percent in the mid-1990s to 12 percent in recent years. While it is true that manufacturing's performance improved with growth rising to 8 percent in 2002-03, it is more difficult to discern if this accelerated performance rests on a sound foundation for growth and defines a trend. It is unclear whether the sector's performance is driven by steady investment and employment creation, or whether much of the high growth is the outcome of once-off efficiency gains associated with the reforms in the late 1990s. It is too early to tell. Recent reports (CEM 2001 and RPED 2004) compared Tanzania's manufacturing performance with its immediate neighbors, Kenya and Uganda, but this does not shed light on country-specific factors. Several key issues remain unanswered: · Besides low inflation and macroeconomic stability, what are the determinants of manufacturing growth in Tanzania? Recently the GoT implemented reforms in the financial and telecommunications sectors. How have these affected the manufacturing sector? · What are the main obstacles to growth in manufacturing since 1999? What are the main policy levers that GoT can use to sustain growth in the sector? Can they be prioritized? · If a small open economy like Tanzania is able to produce competitively, it should be able to grow rapidly by exploiting exports as a key source of growth. What is the relationship between growth and exports in the manufacturing sector? · The PRSP noted that the benefits of macroeconomic reforms have not trickled down sufficiently to the poor. Expansion of the Tanzania's manufacturing sector has the potential to directly create jobs that fetch wages/incomes higher than those generated in the rural or informal sectors. Is manufacturing a source of vibrant employment growth in Tanzania? Meaningful answers to most of these questions require a microeconomic framework that delves below the macro to assess the impact of reform on firms. This paper adds to the existing knowledge-base on Tanzania by grappling with these questions using RPED firm surveys for the 1999 - 2003 period. By using 1993 as a reference point, it assesses where Tanzania's manufacturing sector stands more than 10 years after it initiated serious economic reforms. 3 The organization of this paper is as follows. Section 2 provides a brief description of manufacturing activity in Tanzania in the 1990s, but especially since 1999 when the broad macro reforms were in place. Sections 3, 4 and 5 address the issues listed in the previous section in the context of investment and growth, exports and employment in the sector. The starting point of each section is issues identified by the CEM of 2001. The discussion begins from the broad macro trends, especially since 1999. This is followed by micro evidence. International comparisons are also provided. The performance of the sector is evaluated in terms of pertinent reforms, especially those since the late 1990s, and their impact. Section 6 concludes with policy implications and recommendations. 1. BACKGROUND RECENT TRENDS AND DEVELOPMENTS IN THE MANUFACTURING SECTOR At 8 percent, the share of manufacturing in Tanzania's GDP is presently small. It has also shrunk in the past 10 years. In 1985, it was over 9 percent of GDP (Table 1, chart 1). In the past few decades, agriculture has remained the key driver of overall growth, and services, especially construction, have also grown fast. Between 1964 and the late 1980s, when the government imposed a ceiling of 12 percent on the share of private investment, the latter was restricted to a narrow range of activities such as construction, transport, agricultural equipment and miscellaneous manufacturing. Today, while the size of the manufacturing sector remains relatively small, its structure is fairly diversified (see Table ---). Production is concentrated in Dar es Salaam in three types of firms: agro-processing and food (beer, spirits and cigarettes), textiles and other light industry such as furniture, and heavy industry producing metals (aluminum, iron sheets), cement, paints and plastics. Industrial capacity utilization increased to 7 hours a day in 1998 from just 4 hours a day in 1995 (BOS). Recent growth has occurred mostly in consumer goods such as food products and beer, edible oils and detergents that have attracted new investment. Growth in heavy industry such as chemicals (fertilizers and paints) and metals has declined. What remains remarkably unchanged and a source of concern about the nature of manufacturing growth since 1999 is the disconnect between it and private investment. Several developments explain the persistence of this disconnect over the past three decades. In the 1970s- late 1980s, it was explained by productivity losses and disincentives for private investment associated with government's socialist polices. The spillover from these was blamed for the disconnect in the early 1990s. During the 1990s, its perpetuation was attributed to declining public investment ensuing with the reduction in fiscal deficits in the early to mid-1990s (CEM 2001). Public investment was considered complementary to private investment. The reforms of the late 1990s were expected to promote a business friendly environment and re-establish the natural link between investment and growth. This did not happen. 4 Public investment increased from about 3 percent in 1999 to almost 8 percent of GDP in 2003, but aggregate private investment did not pick up correspondingly. Instead, from a share of 18 percent of GDP in 1992, it plummeted after 1996, settling at 11 percent of GDP in 2000-03. In manufacturing, private investment declined even more steeply. From 35 percent of total private investment in 1992, it fell to about 25 percent in 2000-02 (Table ...). During 1997- 2001, following the range of fiscal incentives for investors, mining was the main beneficiary of private investment growth of over 16 percent per annum (EIU, 2003). Recent reforms have attracted FDI flows in telecommunications as the industry was privatized, and into ports, airports, utilities and the financial sector (EIU 2004/05 page 26 ­ 27). These trends suggest that, at least in manufacturing, the recent acceleration in growth is related not with a corresponding rise in private investment, but is a combination of the improved quality of private investment and increased utilization of existing capacity. It is consistent with the channeling of private investment flows into mining and construction. At a micro level, an analysis of firm surveys for 2000-02 also show that the association between private investment and growth in manufacturing remains elusive. · If the data is to be believed, the sustainability of the recent growth without corresponding investments in manufacturing is questionable, particularly in the medium to longer term. In the short term, reforms that further reduce regulatory obstacles and corruption will continue to yield significant efficiency gains. Payoffs from the latter can probably support the prevailing growth rates in manufacturing over the next 2-4 years. However, without a significant and sustained increase in private investment, this trend is unlikely to be sustained for too long. RECENT TRENDS IN MANUFACTURING EMPLOYMENT In 2001, out of a labor force of approximately 18 million, 16 million were employed. Of these, Tanzania's non-agricultural private sector across all sectors employed only 4.5 percent. Manufacturing's contribution to total employment was even smaller at 245,000 or 1.5 percent of the employed. As a share of non-agricultural private employment (industry and services), manufacturing accounted for about a third. The pattern of labor demand reflected the technical skill-intensive nature of manufacturing. Of the total employed in all industries, manufacturing absorbed about 34 percent of all craft and related-skills workers, 32 percent of all machine and plant operators and 11 percent of all clerks. The manufacture of goods such as textiles and garments or furniture and non- metallic products does not occur in formal manufacturing alone. For example, only 19 percent of the manufactures of wearing apparel, spinning, weaving and finishing, and 35 percent of furniture making and the production of non-metallic products occurs in formal firms that employ paid workers; the remainder is produced by the self-employed (ILFS 2001). 5 The CEM (2001) notes that in addition to higher growth, a wider sharing of the benefits of growth was deemed equally vital for poverty reduction. This implied the creation of gainful employment opportunities for the poor, given the enormous pressure created by a high rate of population growth (3 percent per annum) and a dependency ratio of 50 percent .1 Manufacturing has the potential to contribute to better employment creation as it generates better paying jobs relative to agricultural earnings that typically fall below the poverty line. For example, in 2001, the median2 monthly manufacturing income was TZ Sh. 40,000 compared to 8000 in agriculture, 30000 in mining, 36000 in construction and 20000 in trade. The industrial average was TZ Sh. 30,000. Those who were self- employed in manufacturing earned TZ Sh. 21,429 per month. For policymakers interested in sustained poverty reduction through job creation, promoting growth in manufacturing has a clear attraction. Whether this has indeed transpired cannot be confirmed by aggregate data. Using firm survey analysis, we seek to answer this issue as well as what constrains employment creation in Tanzania's manufacturing sector. LONGER TERM VISION FOR GROWTH AND OBSTACLES SKILLS In 2001, the longer term goal of a modern, export-led economy was based on several factors favorable for Tanzanian firms in manufacturing. Among these were easy access to three sea ports that were operating with greater efficiency; expectations of improved power supply from increases in hydro and thermal sources; potential sources of minerals and coal, sustained reduction in policy and administrative barriers, especially in international trade and a stable macroeconomic environment. The imperatives for scaling up growth in the medium term rested on an increased investment rate of 25 percent of GDP. This was to be achieved through better quality and lower-cost infrastructure services to be improved through private-public partnerships. In addition to 1) infrastructure, the other critical obstacles to investment were 2) a weak institutional and legal framework regarding property rights, relatively high business risks, 3) low level of skills and educational attainment that hampered acquisition of knowledge, use of superior technologies and FDI and hence, limited productivity growth, 4) the HIV/AIDS epidemic eroding the productive, prime-age cohort of 15 ­ 59 in the labor force, 5) a weak and primitive system of financial intermediation, and 6) a governance system that was not sufficiently transparent and accountable. (CEM 2001). 1Overall, between 1990/91 ­ 2000/01, the labor force increased by 58 percent. The urban labor force grew by 14.3 percent compared to a decline in the rural labor force of 3 percent (ILFS). 2The comparable means were TZ Sh. 103,407 in manufacturing; 15234 in agriculture, 76277 in mining, 49693 in construction, 31,301 in trade. The industrial mean was TZ Sh. 49,954. (ILFS 2001, Tables 9.4 and 9.5) 6 In 2004, with overall investment stagnant at 20 percent of GDP, GoT was engaged in a range of reforms in the area of tax and the financial sector, improvement of public enterprise performance especially in energy, combating corruption, and strengthening the legal and institutional environment for private sector growth (IMF, July 2004). POST-1999 REFORMS RELEVANT FOR REDUCING THE OBSTACLES TO GROWTH IN MANUFACTURING Macroeconomic stability, exchange rate and trade reforms. The post-1995-96 reforms helped to establish macroeconomic stability, lower inflation to single digits (about 5 percent in 2004), and led to a steadily depreciating and liberalized exchange rate for trade-related transactions. The continued depreciation of the real exchange rate approaching its equilibrium levels in 2003-04 provided favorable conditions for manufacturing export growth. Additional benefits were anticipated from further trade liberalization through tariff cuts and the establishment of a customs union between Tanzania, Kenya and Uganda (EAC). By 2009, the latter is likely to eliminate remaining internal barriers to trade (IMF, 2004) and should boost manufactured exports further. Structural reforms. Several structural reforms - civil service, privatization, trade liberalization - reigned in fiscal discipline, reduced distortions, improved economic efficiency and speeded up manufacturing growth to 6 percent per annum during 1999- 2003. Apparently, starting in 2005, the next critical area of reform is customs administration which through reductions in the transactions costs of trading will bestow potential productivity gains for the sector (IMF, 2004). While the scope for continued efficiency gains is significant in the near future, without additional investment, growth in the sector may not be sustainable. The challenge for policy makers is to explore if, in addition to efficiency-improving reforms, they can do anything more to improve productivity and attract investment in the sector. Financial sector reform. Between 1967 ­ 1991, the state controlled the commercial banks in Tanzania. In recent years, private participation has steadily increased. In 2003, there were 20 banks, including most international banks that operate in Africa. To reduce the continued hurdles to private credit for business investments, in 2001, GoT's other planned reforms included: privatization of the Cooperative and Rural Development Bank, institution of commercial courts to legally enforce loan repayments and improve bank lending to businesses. Although the effect of these reforms on investors in manufacturing is not known, they had several potentially beneficial effects for the sector. Overall, the ratio of non-performing loans to capital fell from 22 percent in 2002 to 9 percent in 2003 (IMF, 2004). The reduction in interest rate spreads from 18.4 percent in 1997 to 11.4 percent in 2003 led to a doubling of bank credit to the private sector. The latter rose from 3.5 percent of GDP in 1997 to 7.6 percent in 2003. In 2000, for many manufacturing firms, at 14 percent the interest rate spread was a deterrent to investment (CEM 2001). To promote economic development, GoT designed several non-bank financial institutions, merchant banks and the Tanzania Investment Bank designed. In 1998, the Dar es Salaam stock exchange (DSE) was established to attract investors; since 2002, 7 foreigners have also been permitted to invest in the DSE. At present, the listed companies include Tanzanian Breweries, Tanzanian Oxygen, Tanzanian Tea Packers, Tanzanian Cigarette Company and Tanga Cement. Five stockbrokers have been licensed to operate but the market remains constrained because of limited stocks and low liquidity. The sale of GoT's majority stake in the National Commercial Bank (NCB) is planned for mid-2005. Combined with the recent establishment of a credit bureau to screen borrowers, modernization of the land registry to enable use of land as a collateral, and plans for restructuring of the state-owned People's Bank of Zanzibar, Tanzania Postal Bank and Tanzania Investment Bank, the efficiency of the financial sector is expected to increase, improving GoT's efforts to increase access to credit (IMF 2004). Infrastructure - electricity. In 2001, Tanzania had a large and unexploited potential for power generation, with a relatively well developed but costly hydroelectric system, by African standards. The source of the cost inefficiencies were related with a weak distribution system and low revenue collection. 85 percent of the hydroelectricity was supplied by the National Grid; the remainder was thermal. The power company, TANESCO, was financial unviable, depending on GoT for large operating transfers. With the anticipated coming online of the Songo Songo gas reserves which would enable the switch from diesel to gas based power, the costs of electricity were expected to decline. The coming on line of the Kihansi hydropower project was further expected to boost power supply from 350 MW to 530 MW (CEM 2001). In 2004, GoT indicated that transfers to TANESCO would decline in 2005/06 due to ongoing financial management reforms, and cost reductions facilitated by the switch from diesel to gas (IMF 2004). In the medium term, GoT would continue to prepare the company for eventual sale. Infrastructure reforms. o In 1998, GoT established a Road Fund; around 2001, it established an independent road agency to manage and maintain the roads network (CEM 2001). o In 2000, the anticipated privatization of telecommunications, water transport and power were to improve supply and lower costs (CEM 2001). In 2001, only about 70 percent of the installed water capacity was fully used. 68 percent of the urban areas were covered but 52 percent of the urban coverage was disrupted by technical and commercial losses. By 2002, water sector reforms were expected to significantly improve coverage and reliability. [Did these happen?] Combating corruption. In 2005, GoT expects to legislate a new anti-corruption law (IMF 2004). Improving the business environment. The launch of the Investors' Roundtable (IRT) has enabled dialogue between GoT and the private sector, focusing attention on critical issues such as the revised Land Act and a new business licensing system (IMF 2004). 8 Human capital. Human capital that provides a near-guarantee for better paying employment in a poor country like Tanzania, displayed disappointing trends. In addition to a large share of primary school dropouts3, in 2001, Tanzania had the lowest rates of secondary and higher education achievements in sub-Saharan Africa. In 2001, the gross secondary school enrollment ratio was only 5 percent, far lower than Sub-Saharan Africa's 27 percent and East Asia's 69 percent (update). In 1990, the share of its population with at least some formal post-primary education was lower (3.2 percent) than at independence in 1964 (5 percent). In spite of universal primary education, failure of post-primary education to keep pace with population growth, emigration of skilled labor, and the high cost of secondary and tertiary education relative to income levels in Tanzania are some of the reasons for weak human capital accumulation. The HIV/AIDS epidemic is further eroding the productive and skilled segment of the labor force. In 2001, the cost of treating one AIDS patient was equivalent to educating 9 primary school pupils. As a large and fast growing manufacturing sector can be critical in achieving sustainable economic growth and in creating better paying jobs in Tanzania, using the micro- evidence, we need to resolve at least three issues for policy makers in sections 3 - 5: 1) if private fixed investment has been declining, what explains manufacturing growth rates of 8 percent per annum in 2002-2003? Is the recent spurt of growth reflecting the efficiency gains associated with private investment climate reforms? How long will this last? 2) Is the delayed response from private investors due to factors that policy makers should have done/can do something about? This could imply accelerating outstanding reforms or provision of critical public goods such as physical infrastructure, labor skills that rapidly eroded during 1995-2000 as government implemented stabilization reforms? 3) In light of a dampened private investment and employment response in manufacturing, do we have evidence that makes a case for re-thinking the growth strategy for manufacturing in Tanzania? Is there a compelling case for additional actions that policy makers can take to reign in private investment and productivity growth? 2. TANZANIA'S MANUFCTURING SECTOR - A MICROECONOMIC PERSPECTIVE STRUCTURE The 2003 RPED firm survey for Tanzania covers the period 2000-2002 and comprises of a sample of 276 manufacturing firms in 8 sectors ­ food and agro-industry, chemicals and paints construction materials, metals, furniture and wood products, paper, printing and publishing, plastics and textiles, garments and leather products. The sample can also be disaggregated by firm size measured by the number of employees: micro (1- 5), small (6 3In 1998, 70 percent of the primary school pupils failed to achieve a passing grade (CEM 2001). 9 ­ 29), medium (30 ­ 99) and large (100 plus). The distribution of the firms across sectors and sizes is shown below. 10 Year Size 1 ­ Size 2 - Size 3 - Size 4 - All Micro Small Medium Large firms 1- 5 empl. 6 ­ 29 empl. 30 ­ 99 100 plus empl. empl. Number of firms and 2002 16 (6.1%) 109 (41.3%) 72 (27.2%) 67 (25.3%) 276 share in brackets (100%) Sales growth p.a. 2001 -3.5 4.0 14.5 17.9 8.8 (%) Sales growth p.a. 2002 2.2 8.5 10.5 20.9 11.4 (%) Sectoral share in Sales growth in Sales growth in sample (%) 2001 2002 Agro industry 29.3 8 18 Chemicals 9.8 14 13 Construction 3.9 12 7 Metals 10.5 0 21 Furniture 23.5 6 3.5 Paper, printing, publishing 9.1 10.0 10 Plastics 2.5 12.0 16.6 Textiles and garments 11.2 18.5 17.0 All firms 100 (number =276) 8.8 11.4 Non-exporters Exporters Non-exporters Exporters Sales growth Median median mean mean in 2001 8% 24% 66% 91% in 2002 11% 23% 41% 49% 71 percent of the firms are located either in the capital Dar es Salaam (39 percent) or other cities/towns with a population of over 1 million (33 percent). Most of the firms in these larger urban centers are in heavy manufacturing such as plastics (71 percent of the industry), construction materials (64 percent of the industry), and chemicals and paints, and metals (almost 50 percent of the industry). Towns with smaller populations of 250,000 ­ 1 million house about 18 percent of all firms4 and even smaller towns with populations of less than 250,000 house the remaining 11 percent of all firms. More labor intensive-industries are located mostly in smaller towns with populations of 1 million or less. About 72 percent of food and agro-industry processing agricultural produce, 66 percent of the furniture and wood products industry and 65 percent of the textiles and garments industry is located outside the larger cities. 4All numbers are derived from the sample (representative?). 11 The average age (median) of a manufacturing firm is about 13 years. Only 7 out of a sample of 276 firms are older than 50 years. The oldest firm, aged 82 is engaged in producing food and agro-processing. The firms survey is a good instrument for understanding private manufacturing sector development in Tanzania. Except for 2 percent that are publicly held and 1 percent owned by the GoT, the remainder are privately owned. Only 16 percent of all firms were previously state-owned; of these, 75 percent were privatized after 1996 when GoT accelerated the privatization process. 40 percent of the privatizations occurred in 1997- 98, 8 firms were privatized after 1999. One fifth of Tanzania's manufacturing firms have some degree of foreign ownership. Presently, the average share of foreign equity is 72 percent. Only about 5 percent of the firms are 100 percent foreign-owned. The majority of the firms' shareholders are Tanzanian nationals (60 percent). The shares of non-African nationals is 14 percent, Kenyans, Ugandans and other Africans between 2 ­3 percent each. The ethnicity of the principal owners is predominantly African (44%), followed by Asians (26 percent), Europeans (96 percent) and Lebanese (4 percent). Start-up capital for about 72 percent of the firms came from the owner's savings or internal funds from some other source. Bank loans financed the start up of 15 percent and equity or sale of stock only 10 percent of the firms. Family and friends provided startup capital for about 6 percent of the firms. Money lenders and informal sources are less important. PERFORMANCE OF MANUFACTURING FIRMS Aggregate growth in GDP per annum during 1999-03 was 5.4 percent, while that in manufacturing was about 6 percent (Table 5). Firm survey data for value added in manufacturing is unreliable; however, trends in sales growth during 2001-02 are consistent with aggregate growth trends.5 Firm sales grew at 11 percent6 in 2002 and 9 percent in 2001 but the variance was large, ranging from (-) 26 to (-) 98 percent for the bottom quartile to over 66 percent for 5 percent of the firms that grew the fastest. At least a quarter of the firms in both years grew at rates of ­9 to ­ 99 percent per annum. Are large firms growing faster? Between 2000 ­ 2002, growth in sales increased with the size of the firm7 measured by the number of employees, a trend that was also prevalent in Tanzania in the early-mid 1990s. Large firms and exporters out-performed others with sales growth at about 19 percent in 2002. Sales in micro firms grew at only 2 percent. In 5A reason why aggregate trends were probably slightly lower was that the aggregates also cover a large number of micro firms that were under-represented in the sample. 6Due to large variations, the measure of central tendency is typical the median in this paper, unless otherwise stated. 7All results reported here are statistically significant as shown by linear and non-parametric tests of statistical significant (Pearson's, Spearman's, Kruskal Wallis etc.) 12 general, between 2001 and 2002, except in medium firms, growth accelerated across all firm sizes. Are firms in some sectors growing faster than others? Growth across manufacturing sub- sectors was not uniform. The fastest growing sectors were metals (20%), agro-industry (18%), plastics (17%) and textiles/garments (16%). Furniture sales grew at only 3 percent. To some extent, developments in other sectors of the economy reinforced these trends. Agriculture, which drives the largest sub-sector (agro-industry within manufacturing grew at about 5.5 % during 2001-03. Growth in mining and quarrying sectors that drive growth in metals, picked up in 2002-03. DETERMINANTS OF GROWTH For the period 2000 ­ 02, the key factors in Tanzania driving firm growth, defined as growth in sales, are age, growth in investment, growth in exports, and use of newer technologies (Table ---). Firm size8, sector or firm location (large cities or smaller towns) in general do not affect growth9. Nor do other firm characteristics such as the nationality or ethnicity of the firm owner seem to matter. Older firms grew more slowly than younger ones. Firms that experienced higher export growth also grew faster; for each percentage point increase in export growth, sales grew by 0.87 percent. This suggests that for globally competitive firms, the ability to export is a critical source of growth. It relaxes the demand constraint imposed by the size of Tanzania's domestic market on firms who are unable to export. Firms with access to newer technologies measured by the percentage of the workforce that used computers, grew faster too. For a one percent increase in the computer users in the workforce, a clear proxy for technical skills in the labor force, sales grew by 0.5 percent. While aggregate trends for 1960 - 1999s show a disconnect between growth (proxied by sales here) and investment (CEM 2001), in contrast firm data suggests that during 2000- 02, there was a positive, albeit weak, relationship between growth in sales and investment. An explanation for the appearance of this positive but weak link in the post- 1999 reform period could be that the reforms are starting to appeal to investors in the manufacturing sector, even though slowly. While this analysis explains what drives manufacturing growth in Tanzania, it does not answer at least two related issues. One, why are the large variations in sales growth not associated with sector- or size-class- specific effects? And, two, can we explain whether the growth-investment relationship applies uniformly across all firms or it is peculiar to certain firm characteristics? A quantile analysis of the determinants of growth in sales examined the variance in growth by focusing on the fastest (75th ) and slowest (25th) growing segments of the firm distribution. While it did not answer the first question very well, i.e., whether certain 8These findings contradict the traditional theory of firm growth (Evans (??), Nugent and Nabli (??)) which argues that firm growth is a simple function of size and age. 9One explanation why firm size or sector do not seem to matter is that within each of these categories, the variance in sales growth is large, and neutralizes the effect of a higher average growth rate. 13 determinants of growth affect faster and slower growing firms differently, it confirmed that export growth, technology and the firm's age are robust determinants across all firms (Table --). For the fastest growing firms in the 75th quantile, location in larger cities and towns, relative to firms located in towns with under 50,000 population negatively affected growth. Location did not seem to affect growth in any other quantile. Investment growth did not explain the variance in growth in the tails of the distribution. Table: Determinants of firm growth (sales) in Tanzania ­ 2000-02 ­ a quantile analysis Quantiles All firms 25% 50% 75% Constant 14.7 *** 1.33 13.8*** 69.52** (3.64) (5.22) (3.2) (20.07) Age of firm -0.62*** -0.53* -0.29** -0.55*** (0.65) (0.27) (0.13) (0.16) % of computer users 0.46*** 0.29* 0.25* 0.33** (0.16) (0.17) (0.17) (0.17) Invest/asset02 0.19+ 0.07 -0.09 0.31 (0.13) (0.14) (0.06) (0.23) Export growth `02 0.86*** 0.69*** 0.68*** 0.63** (0.14) (0.16) (0.08) (0.30) Location: default is towns with a population less than 50,000 Loc : capital city 12.52 -37.3** (28.37) (20.18) Loc : city pop.> 1 miln. 20.54 -40.8** (27.96) (18.4) Loc: town pop. 250,000 -4.58 -45.04** ­ 1 mln. (31.96) (19.7) Loc: town pop. 50,000 -0.96 -54.09*** ­ 250,000 (42.64) (20.9) Notes: *** significant at the 1% ** significant at 5% * significant at 15% + significant at 16% Standard errors are shown in brackets, bootstrapping was applied. The second issue related with the weak link between investment and growth is partially unraveled by firms' age. An econometric investigation shows that firms aged over 25 years had lower investment rates, lower rates of technological pick-up as given by the use of computers among the workforce and grew more slowly than their younger counterparts (Table ---). For example, sales in firms aged 25 or more grew at 8 percent or less compared to rates nearly twice these in younger firms. Firms aged 10 years or less grew at 19 percent. Growth of younger firms in the below 25 age bracket is strongly related with investment growth rates; for a one percent increase in investment, sales grew by almost 0.43 percent. In contrast, for firms over 35 or 45, investment growth was not associated with firm growth. Interestingly, export growth does not seem to be sensitive 14 to firm age and was robust across age categories, but the use of computers is. For firms under 15 years, a one percent increase in computer use in the workforce increases growth by 0.64 percent. The contribution of computer usage halves with firm age but was insignificant for firms under 10 years. Table: Link between firm age, investment and growth in Tanzania's manufacturing sector ­ 2000-02 LessThan LessThan LessThan LessThan LessThan LessThan 10yrs 15 yrs 20 yrs 25 yrs 35 yrs 45 yrs Intercept 9.96** 6.8** 8.9*** 8.09*** 8.38*** 6.4** % of compusers 0.04 0.64*** 0.48** 0.38** 0.39** 0.31** in workforce invnass02 0.57** 0.39** 0.38** 0.43** 0.13 0.18 gexp02 0.69*** 0.64*** 0.66*** 0.68*** 0.71*** 0.73*** Number o f 68 91 110 124 147 162 firms tested *** 1% ** 2 ­ 10% Issues for policy analysis: the determinants of growth in Tanzania's manufacturing sector are firm age, growth in investment, exports and the level of technical skills in the workforce. There is little that policy makers can do about firm age. Hence, the three key policy issues we now turn to are: 1) What are the determinants of investment in manufacturing and what can policy makers do about it? 2) What are the determinants of the ability to export and propel export growth in manufacturing and what can policy makers do about it? 3) What can policy makers do to promote technical skills accumulation in the workforce in Tanzania's manufacturing? Policies that support the adoption of newer technologies such as those that enable the use of computers among the force promote productivity and growth. 3. INVESTMENT GROWTH IN TANZANIA'S MANUFACTURING SECTOR 1999 seems to mark a breakpoint in Tanzania's recent investment trends, reversing the relationship between growth and investment. As noted earlier, growth in investment and GDP in manufacturing and the overall economy were unrelated during the 1960s ­ mid- late 1990s. From 30 percent in 1992-98, aggregate manufacturing investment measured as a share of total private investment dropped to 23 percent in 1999 (Table 15, Economic Survey 2002). This was consistent with the drop in total private investment from 18 15 percent of GDP in 1992 to 11 percent in 1999-03. In absolute 1992 prices, aggregate private investment in 1992-2003 was still below (TZ Sh. 240,679) its 1992 level (TZ Sh. 243,527).10 In spite of the consistent drop in private investment, growth in manufacturing and overall GDP picked-up significantly. In 2000, there was an upward level-shift in manufacturing investment as a share of the total private investment. The latter increased to 28 percent in 2000 and has since held at that level. As this step increase in manufacturing investment in the last 3 ­ 4 years was complemented with higher growth in manufacturing, and indeed the overall economy, it probably reflects an improved economic environment stemming from the reforms, especially in the late 1990s. In spite of the emergence of this positive albeit belated nexus between investment and growth, the level of private investment in Tanzania remains too low to be a driving force of improved expansion in manufacturing in the future. Micro-evidence from firm survey data validates that on average, investment growth stagnated during 2001- 02. When measured by the change in investment as a share of firm assets during 2000-02, investment grew by 0.2 percent per annum in 2001 and 2002. About half of all manufacturing firms invested in 2001-02 while the other half did not. These results also apply to other measures, such as, a change in investment as a share of the firms' replacement or resale value of firm assets In general, investment growth in younger firms aged 20 or less increased with firm age. Firms less than 10 years invested almost 10 times more than the average. Investment growth in micro and small firms was nearly zero; in medium firms about 0.5 percent and in large firms it was 2.1 percent per annum. The range of investment growth for firms in the top ten percentile in each size class was large, varying from 19 ­ 119 percent. Investors did not reveal a particular preference for any sector. Agro-industry, chemicals, construction, paper and plastics experienced investment growth of 1 ­ 3 percent per annum. Investment growth among exporters was 1.6 percent per annum in 2002. DETERMINANTS OF INVESTMENT GROWTH IN TANZANIA As in any other country, investment decisions in Tanzania are affected by a variety of intricately linked factors. There are over100 distinct factors that affect investor decisions. As most of these factors are also strongly correlated with each, they create empirical complications. 11 For example, the firms that are large as also likely exporters, less 10With the exception of 1994, private investment levels remained below the 1992 level (Table 7, Economic Survey). 11As the number of observations is far fewer than the number of explanatory variables, there is a high degree of multi-colinearity and few degrees of freedom. These econometric problems are relatively common in firm survey data relative to household or labor force surveys that have a larger number of observations. The fact that there are a limited number of observations and a large number of explanatory variables in the dataset makes it difficult to model a clean and comprehensive investment story in one step. There is a high level of intra-group and inter-group multicollinearity which diminishes the predictive power of a single multivariate investment model. A variety of innovative econometric approaches are used to grapple with this problem and arrive at empirically rigorous results. 16 financially constrained, and have or can afford better infrastructure, better technology etc.. In the absence of knowing their relative importance, they can easily result in a laundry list of policy recommendations for increasing investment in Tanzania's manufacturing sector. This approach is not very useful for policy purposes as it makes it difficult to identify policy priorities required for the proper sequencing of reforms and informed policy decisions. A muti-pronged methodology was used to resolve this issue. As a first order of priority, the relative importance of various policy reforms is examined. To maintain analytical tractability and empirical rigor, the factors affecting investment decisions are organized into 7 ­ 8 different groups of determinants which are related to a broad area of policy reform such as financial, infrastructure, technology, labor, demand, etc.. Within each group, there are a large number of factors which differentially impact overall investment. For example, within the financial group, the policy implications of issues related to start- finance are quite different from those related to access to working or investment capital which are quite distinct from how the level of the interest rate affects investors, and so on. Hence, the second order of priority is to zoom in on the most binding constraints within each policy reform. This helps to identify which areas of the financial sector, to continue the example, need more attention and can contribute more to investment growth. Last, in keeping with tradition, the full investment story is pulled together. Because of the econometric complications, an effort is made to keep the model simple with at most one or two key variables from each policy area. IDENTIFYING POLICY PRIORITIES FOR HIGHER INVESTMENT AND GROWTH Time and resource constraints make it difficult for policy makers to focus on and implement a large number of policy reforms at the same time. In the first round, the relative importance of each determinant-group that influences investment decisions in the manufacturing sector was assessed econometrically (Table ---). Firm demographics defined by age, size, sector, location and nationality of the owners explained about 13 percent of investment growth across all firms 2000-02, and 28 percent in younger firms aged 20 or less.12 In general, across firms, investment growth declined with age and increased with firm size. The location of firms in towns with a with a population of a million or less had a negative impact on larger firms but did not seem to affect smaller firms. Relative to textiles and garments, firms engaged in metals and metals producing sectors had lower investment growth. Interestingly, across all firms, owners' with ethnic origins from Lebanon contributed positively to investment growth. In smaller firms, while owners whose nationality was Ugandan made a positive contribution to investment growth, those of Tanzanian, Kenyan or non-African nationalities had the opposite impact. 12As the age of manufacturing firms significantly affected the nexus between investment and sales growth in recent years (Table ---), the determinants of investment were analyzed for the cluster of all manufacturing firms as well as those aged 20 years or less. 17 Table ---: Indicative reforms priorities for increasing investment in manufacturing Rank Firms for under 20 Rank firms or less for all 20 or All firms yrs. firms less R2 R2 Demographics ­ age, size, sector, nationality 1 of owner, location etc. 0.13 0.27 2 Demand 0.059 0.150 4 3 3 Financial group 0.286 0.220 1 1 Assets & technology, 4 equipment, capital intensity 0.117 0.176 2 2 5 Infrastructure group 0.113 0.216 3 1 6 Investment climate group 0.046 0.147 5 4 7 Labor skills, productivity, HIV etc. 0.112 0.228 3 1 Source: Authors' estimates based on econometric investigation Table ­ indicates the relative importance of the 6 key policy-related determinants of investment in manufacturing pertaining to a particular area of policy reform. Firm demographics is an exception as policy makers can do little about most of the characteristics, at least not directly. Within each reform area, there are as many as 15 ­ 20 different factors that explain investment growth. The broad determinant groupings are: 1) market demand measured by sales in the domestic market to different customers such a state owned enterprises, large firms, MNCs, or in export markets which generally not constrained by the size of the domestic market; 2) finance as given by access to credit for start-up, investment and working capital, and interest rates etc.; 3) assets and technology reflecting the age of the equipment, access to superior technologies, extent of computerization, capital intensity etc.. 4) infrastructure as noted by the paucity of power, water, transport, telecommunications, and the extent of substitution between private and public infrastructure and its costs etc..; 5) quality of the investment climate as measured by the level of corruption, bribery, confidence in the judicial system, etc.. 6) human capital ­ skills and education of workers, managers, HIV-related costs and the level of productivity. Special attention was paid to younger firms aged 20 or less as these displayed significantly higher investment and sales growth in recent years and were the main drivers of growth in Tanzania's manufacturing sector. The top 4 priority policy reforms 18 which affect older and younger firms were related with the financial sector, technology, infrastructure and human capital. For younger firms, the next priority area was the market demand constraint. Access to finance for start-up, investment and working capital was severe across the board. Foreign firms had a slight advantage either because of parent companies easily financed investment or because they had easier access to bank loans. Most domestic firms were severely constrained in terms of access; those that had access found the interest rates too costly to finance investment. Good infrastructure and good labor skills are equally critical in explaining investment growth in manufacturing firms in Tanzania. The nature of the determinants and their significance levels underscore the complementarity between human capital and physical capital. For example, while transport and utilities (such as electricity and water) fall under infrastructure, physical capital also encompasses superior technologies of production reflected in the type and age of the machinery, use of modern information technologies, level of computerization etc.. But all these, in turn, are critically dependent upon human capital embodied in the ability of workers to operate such technologies and learn how to use superior technologies of production. While market demand is ranked 3 or 4, the data makes a rigorous case for incentives for export promotion. There are inherent spillovers between export-ability and investment in the quality of physical and human capital, including better technologies. Apart from enlarging the size of the market, exporting imposes on firms global competitiveness standards that encourage the choice of technology associated with higher labor productivity, dynamic learning processes and constant technological upgrading to compete in the global marketplace. INFRASTRUCTURE13 - A CONSTRAINT TO INVESTMENT GROWTH Within the infrastructure group, the four main constraints are related with electricity, water, road transport and others factors such as telecommunications, internet connectivity, waste disposal facilities etc.. Overall, infrastructure-related factors explain about 15 percent of investment growth; for younger firms aged 20 or less with the fastest investment growth, this statistic rises to 23 percent, mostly because of the exacerbated impact of water and transport-related constraints (Table __). In order of importance, the top three problem areas are (1) irregular electric supply leading to investments in the private provision of power for those who can afford it. (2) production losses related to poor transportation, frequently resulting in private provision of transportation. (3) Internet connectivity for all firms boosts investment growth. In younger firms, investment in water infrastructure to compensate for the irregular supply of public water has a particularly negative effect on investment growth (Table ---). 13See RPED Investment Climate Assessment, 2004 for details on the firm surveys. 19 Table: Relationship between investment growth and infrastructure ­ identifying the binding constraints. All firms Aged 20 or less Sig. Sig. Coeff. Level Coeff. Level Intercept 41.7 *** 25.60 ** Demographic group - Age -0.43 *** - Small 3.40 * 6.62 *** - Medium 4.30 ** 10.51 *** - Large 8.4 *** Metals and related products -3.80 * Infrastructure group Electricity Own generator dummy -30.72 *** -18.15 ** Generator age* -0.76 *** -0.41 * Water Share of water provided privately -0.04 * -0.06 ** Share of water from public resources Own well dummy Own water infrastructure -4.95 *** Other Infrastructure Transportation Own road dummy 5.71 ** 8.53 *** Own waste disposal dummy 2.79 * 3.72 ** Own freight transport dummy -3.45 * -5.49 *** Production losses due to poor public transport 0.23 *** 0.12 * Percentage of average cargo lost -0.21 * -0.11 Telecommunications costs 0.000007 * 8.6E-06 ** Internet access dummy 4.04 ** Located in an industrial estate -4.85 *** -2.66 * Number of observations 231 171 R- squared 0.15 0.23 Source: Authors estimates *** = sig, level 1 - 5% ** = sig. Level 6 - 15% * - sig. Level 16 - 30 % Poor public infrastructure has driven many firms to invest in private infrastructure in Tanzania. Apart from efficiency considerations related with the loss of scale economies when each individual small or large firm installs a private generator, for the private investor the cost of installing private infrastructure is an additional cost of doing business in Tanzania. Every Shilling invested in a private generator or water source or road is a Shilling of forgone investment in expanding production and contributing to growth and employment. 20 About 62 percent of all firms are located in special industrial estates,14 presumably because of some public policy that either allocates industrial sites within these designated areas to firms or attracts them there for some other reason. In any case, firms located within these estates do not appear to benefit from better infrastructure. Power outages are frequent and impose high financial costs on firms. Measured in terms of a single spell lasting 8 hours (the equivalent of one working shift), power outages occur on average at least 18 working days per year. Some firms experience disruptions for as many as 56 working days per year. As many as 54 percent of the firms have installed private power generators at costs ranging from Tz.Sh. 172,700 ­ 812,500 per employee.15 Most generators are about 7 ­ 8 years old. Generators lose efficiency with age and have to be replaced. In sum, the recurrent costs of installing and maintaining private generators in the face of power outages inflict high costs on investors, constraining investment growth, as evident from the large and significant coefficient on the electricity determinants in table --. Power outages and fluctuations also damage production equipment. Industrial water is available to manufacturing firms from a variety of sources including public. Due to irregular supply, about 40 percent of the industrial water is obtained from private sources. While this affects investment growth adversely, as Table ­ shows, water is relatively less constrictive for investors than power and transport. Its negative impact on investment growth is directly related to two factors: (i) the larger the share of privately provided water, the more negative the effect on investment; and (ii) private investment in water infrastructure has a particularly negative impact on investment growth in younger firms (Table --). Tanzanian firms obtain 58 percent of their industrial water from the local municipality, 28 percent from a private well, 10 percent from private vendors and the remainder from communal sources. Private provision of water entails additional costs of construction to channel water from its source to factories. About 34 percent of the firms own a well, 31 have invested in their own water infrastructure and about 12 percent share a water source with the local community. The data does not permit comparisons of the costs of water from private and public sources, but as Table ­ shows, private investment in water infrastructure substitutes for direct investment that could lead to firm growth. On average, firms paid water bills ranging from Tz. Sh 71000- 32700 per employee in 2002 for the use of industrial water from all sources. The fact that a large proportion of firms have not invested as heavily in private industrial water as power should not be mistaken with water not being a constraint but rather its being a serious deterrent to investment growth in water-intensive sectors like textiles and garments which are labor intensive or agro-industry and food which are potentially high-growth activities in Tanzania. For a country with a rich rural hinterland that produces a variety of agricultural produce as input for agro-industry, the shortage of reliable freight transport by road, rail and air is a direct constraint to growth. About 41 percent of the firms lose anywhere between 1 ­ 50 percent of their cargo in transit due to spoilage, breakage etc., adversely affecting investment growth. About 15 ­ 20 percent of all firms lose as much as 2- 3 percent of production and sales respectively due to poor transport facilities. In extreme cases, as 14Presumably, the industrial estates provide better or less costly infrastructure and possibly, tax breaks. 15The higher range applies to firms in the upper 10 percent of the distribution. 21 much as 80 percent of total sales are lost. Empirically, transport-related lost production or sales, and the private provision of roads seem to affect investment growth positively. One explanation for this could be that existing investors have factored in transport-related losses in their investment decisions. After incurring the fixed cost of investment in a private road, the latter is an asset which has a positive effect on future investment growth. The same does not apply to infrastructure that requires large but recurrent investments. Investment in generators, water infrastructure, and freight transport that need regular upgrading or replacement negatively affect investment growth in the firm. At least 10 percent of the firms invest in private roads; 18 percent have invested in freight transport to cart goods back and forth. About 26 percent of firms have invested in transportation for workers to compensate for lack of good public transport to and from the workplace. Another area with problematic infrastructure is telecommunications. Only 49 percent of the firms have internet connectivity, but power outage-related spillovers diminish the ability of these firms from accessing information when they want. Firms pay about Tz. Sh. 11428 per employee per year in telecommunications bills. While this may be seem a meager amount, the opportunity cost of regular and easy access to information technology is yet another factor that constrains Tanzanian firms from linking better with global markets in pursuit of faster growth. Key policy issues: power, transport., especially freight, roads, telecommunications (internet); for younger firms, also water. CAPITAL/BANK CREDIT ­ A CONSTRAINT TO INVESTMENT GROWTH Constraints related with the financing of investment in Tanzania are numerous. The degree to which investment financing is a constraint is largely determined by the nationality, ethnicity of the investors and firm owners, the share of foreign equity in the firm, or the firm's ability to export. Most of these are demographic characteristics that policy makers cannot do much about. Such factors allow investors to transcend constraints like access to bank finance and its price, the interest rate otherwise imposed by the financial sector on domestically owned firms. As the financial sector has undergone several important reforms in the recent past, in this section, we focus on access to bank loans as a source of startup, investment and working capital. The role of interest rates is discussed descriptively as the data is bad. In 2001-02, the banking system played a relatively negligible role in meeting the financial needs of the manufacturing sector. Our analysis shows that (a) only about 20 percent of the domestically owned firms, mainly non-exporters, use bank loans for investment and operational purposes. Exporters or foreign owned firms relied more on private foreign finance, especially for investment purpose. (b) Exporters rely more on bank loans as a source of working capital. (c) Availability of bank finance had a positive and significant effect on investment growth (table___). Two key policy issues arise. First, increasing the level of accessibility to bank finance for domestic firms by widening and deepening of the banking system is necessary to facilitate investment growth in the manufacturing sector. Greater coverage and new products for different needs are needed. 22 Second, a decline in interest rates would facilitate investment growth by reducing the cost of working capital for exporters and the overall cost of financing and debt for all firms. Start-up finance. Firm ownership drives with access to sources of investment and working capital. For examples, firms that have access to bank loans for start-up also have access to bank loans for investment and working capital (Annex 8). Among non- exporters, about 77 percent of the equity was held by the domestic private sector; about 16 percent by foreign owners and less than 5 percent by government. For these largely domestically-owned firms, 7 percent of the startup capital came from informal sources and 10 percent from the banking system. In contrast, exporters obtained nearly twice as much equity (31%) from foreign firms who also facilitated access to capital for investment and working finance. Most exporters relied less on debt capital - only 5 percent of their start-up capital was loan-financed relative to 10 percent for non- exporters. Table __ shows that there was a positive and significant relationship between bank loans and investment growth in 2001/02, especially among younger firms aged 20 or less. However, for start-up, relative to bank loans, the importance of equity or private capital is almost five times more important. And the latter is limited by personal constraints. This makes clear that focusing on increasing firms access to bank loans as a source of start-up can go a long way towards increasing investment growth in Tanzania's manufacturing sector. Annex 8 clearly demonstrates that this applies across all types of firms.16 Bank loan financed investment and working capital, and interest rates. Although relatively small, the importance of bank loan financed firm capital is critical for manufacturing firm investment and growth. Although it does not emerge in the multivariate analysis presented in Table __ below due to severe multicollinearity problems, there is a strong, significant and positive relationship between investment growth and access to working and investment capital from banks (see Annex 8), and their price or interest rate. As the two types of capital are essential for firm operations and growth, both are considered here. Firms who have access to bank loans for investment capital also have access for working capital and vice versa, and are sensitive to interest rates. Only 20 percent of the domestic firms have access to bank loans for investment and working capital. For this small proportion, bank loans financed about 14 percent of total investment and 10 percent of all working capital. The balance of investment capital is limited to retained earnings. The remaining domestic or non-exporting firms who represent over 80 percent of all manufacturing firms (Note, exporters comprise less than 20 percent of all firms) do not have access to bank capital. Loan applications carry a high transactions cost and the negatively relationship between applications and investment growth indicates that most of those who applied, were rejected and could not invest. The small proportion of forms who do have access to bank loans are sensitive to the interest rate for investment purposes but indifferent with respect to working capital. The policy implication of this is that 16Due to severe multicollinearity, this does not show up in the multivariate in Table __- but is clear from Annex 8 that examines this relationship rigorously at the bivariate level. 23 1) a reduction in interest rates can boost investment growth for firms who have access. 2) Additionally, extending access to 80 percent of the domestic firms who are largely non-exporters will also help to increase investment in manufacturing. Exporters use bank loans to finance working capital more than investment capital (for which they rely more on their foreign partners/parents or retained earnings), but are also sensitive to the level of the interest rate (Annex 8). Over 25 percent of their working capital relative to only 10 percent for non-exporters is financed by bank loans. The crux is that even if investment capital is relatively cheaper from their foreign parents/partners, expensive working capital can deter investment growth. Exporters use fewer bank loans to reduce the overall cost of debt-financed investment in spite of considerably lower interest rates. In 2001-02, exporters paid interest rates of about 10 percent compared over 16 percent for firms that were more dependent on the banking system. Nevertheless, in comparison to all time low world interest rates, 10 percent was relatively high and deterred bank financed investment for expansion. The story of exporters reinforces the case for a lowering of interest rates to facilitate cheaper working capital for exporters who are the key drivers of investment and growth in Tanzania's manufacturing sector. Exporters or firms with links to foreign firms also enjoy several exclusive benefits that facilitate working capital, especially at a short notice, through overdraft privileges with domestic banks. Nearly 49 percent of the exporters had overdraft privileges with the banking system in 2002 compared to only 28 percent for non-exporters. Table __ indicates that for younger firms, such a privilege is particularly conducive for investment growth. Other non-bank factors that affect investment growth are: 1) firm profits - the larger the share reinvested, the higher growth in investment (table __). 2) Being a `priorsoe' applies to a firm that was formerly state-owned, and is measured by the number of years since its privatization. So, the longer a firm has been privatized, the more closely it would mimic a private firm that was never an SOE. The negative sign makes sense ­ firms that were only recently privatized experienced large inflows of investment capital but the older ones are similar to the majority of private firms whose investment levels were flat. 3) There is a significant and positive relationship between investment growth and the share of foreign ownership in the firm (Table --- shows a). A higher share of foreign ownership facilitates private foreign financing besides fetching several other financial benefits associated with foreign ownership. 24 Table: Relationship between investment growth and finance for investment ­ identifying the binding constraints All firms Aged 20 or less Sig. Sig. Coeff. Level Coeff. Level Intercept 4.83 ** Demographic group - Age 0.16578 * -0.455 *** - Small 5.433 *** 2.8577 - Medium 5.586 ** - Large 3.7 * *** Metals and related products -5.67 ** Owner's ethnicity - Lebanese 12.33 *** 13.519 *** Owner's nationality Ugandana Sources of start-up finance Owner's saving /internal fund Equity, sale of stock 0.347 *** 0.165 *** Bank loan 0.06 * 0.099 *** Family or friends Share of private domestic ownership Share of private foreign ownership 0.040 * Prior SOE -0.302 *** Overdraft facility dummy 3.325 * Applied a bank loan dummy -4.23 ** -4.385 ** Share of working capital financed: Private resouces Bank loans Public resouces/funds Share of investment capital financed: Private resouces Bank loans Public resouces/funds Share of profits reinvested 0.050 *** 0.06 *** Interest rate on bank loans Number of observation 151 116 R-squared 0.439 0.327 Source: Authors estimates *** = sig, level 1 - 5% ** = sig. Level 6 - 15% * - sig. Level 16 - 30 % HUMAN CAPITAL AND INVESTMENT GROWTH Analysis of the contribution of human capital to growth, especially in countries with a largely unskilled, and less educated labor force presents the problem of a sample bias ­ by definition, all existing firms have self-selected to work with the available workforce 25 while potential employers who could create better paying jobs by employing skilled workers have chosen to invest elsewhere. It is therefore not surprising that the contribution of human capital to investment growth in Tanzania, relative to that of other determinants, is understated in the analysis that was conducted (Table ---). TABLE: RELATIONSHIP BETWEEN INVESTMENT GROWTH AND HUMAN CAPITAL ­ IDENTIFYING THE BINDING CONSTRAINTS. All firms Aged 20 or less Sig. Sig. Coeff. Level Coeff. Level Intercept -4.69 6.27 * Age of firm -0.033 -0.542 *** Males less than 30 0.00127 -0.0586 * Females 30 - 45 years 0.270 ** 0.211 ** Effect of only education of employees Primary education - complete 0.118 *** 0.112 *** Secondary and vocational education Tertiary education or a diploma Graduate and post-graduate education Worker training 3.03 * 2.228 Effect of HIV in skilled workers -0.017 -0.00096 Effect of HIV in unskilled workers 0.008 -0.00622 Productivity (sales revenue) per non-mg worker -4.760-E8 * -4.2E-08 ** Manager's education 1.20 * 0.02539 Manager's prev. export exp -3.25 -0.277 Manager's work experience Number of observations 152 103 R-squared 0.093 0.177 Source: Authors estimates *** = sig, level 1 - 5% ** = sig. Level 6 - 15% * - sig. Level 16 - 30 % The key human capital determinants of investment in Tanzania from the self-selected sample of firms are age and education levels of workers, worker training in addition to education, productivity of the non-managerial workforce, education and experience of managers, and the adverse impact of HIV on firm profits. Severe multicollinearity problems complicate the modeling of these highly correlated factors in the multivariate model in Table.... In spite of this problem some factors such as workers age, education, and skills training emerge as being significant and positive but others like productivity have the wrong sign and yet others are masked by econometric problems and are insignificant. In any case, Table __ shows that among variables with policy relevance, primary education, worker training, and managers education affect investment growth positively. Worker productivity measured as sales per non-managerial worker should have a positive sign but appears with a negative sign and a coefficient that is almost zero. Closer examination through bivariate analysis shows that worker productivity is 26 positively associated with investment growth, as would be expected. Before dismissing the remainder of the human capital determinants as being unimportant for investment growth because they fail to appear in Table ___, further econometric scrutiny was conducted and is reported below. About 44 percent of the manufacturing workforce had completed primary education, another 25 percent had secondary education, 12 percent had some vocational education and 7 percent had tertiary or a diploma. In firms with higher investment growth such as exporters, the proportion of graduates or post-graduates was twice as high at 10 percent compared to 5 percent in firms that sell domestically. Worker training at the firm level was also more prevalent in exporting firms (71 percent of the workforce had received some type of formal training compared to only 47 percent in non-exporters). These descriptive statistics suggest a positive relationship between investment growth and higher education, especially graduate and post-graduate and formal training in the manufacturing sector. Rigorous bivariate investigation shows that there is a strong, positive and significant relationship between investment growth and primary, secondary, vocational, diploma (technical) and graduate education in Tanzania's manufacturing sector. The same also applies to worker training (Annex 9). Primary education is linearly related with investment while other types of education and training have a non-linear relationship. Moreover, worker training at the level of the firm is strongly associated with the share of the workforce that has secondary, graduate, diploma or other types of technical or tertiary education ­ a fact that may explain why firms in Tanzania or other sub-Saharan African countries invest less in training compared to firms elsewhere (East Asia for example). The presence of a large workforce with mostly primary education is hardly the suitable input for investment in firm level training which is costly. This finding has serious policy implications: 1) If emphasis on education is not extended beyond primary and secondary levels, firms in manufacturing will not invest in training and investment growth will continue to remain low. 2) Low skills or no skills are also constrains labor productivity which requires workers with superior skills to operate new technologies or learn better processes. 3) Tanzania's manufacturing sector will either not attract or only attract investors shopping for low-skilled workers. The type of low-wage employment that the paucity of skills creates is inconsistent with the country's vision of moving workers out of poverty. To move out of poverty through manufacturing requires moving into higher value added production but the latter needs more skilled workers. 4) More importantly, Tanzania's low labor cost advantage is not permanent. With increasing integration into global markets, its `low skills--low cost' workers will have to compete with `high skills ­low cost' workers from other competitors, especially those in Asia. Early signs of this are clear from the stagnant investment rates in manufacturing, a further testimony to the sector's unsustainable growth, the recent 8 percent notwithstanding. 27 Unfortunately, data problems do not permit a meaningful analysis of the impact of HIV on firms investment rates. But from other sources of information, it is apparent that the pandemic is inflicting severe costs on firms through decreased labor productivity. TECHNOLOGY, QUALITY OF FIRM EQUIPMENT AND INVESTMENT GROWTH An earlier section of this paper noted the strong relationship between overall growth in manufacturing and the use of better technologies of production in relation to other firms in the sector. A good example is the proportion of computer-users in the workforce which indicatives the level of worker skills and the sophistication of the machinery and equipment in use (i.e., automated processes with computerized controls as opposed to manually driven machines and implements). Table: Relationship between investment growth and technology and capital stock ­ identifying the binding constraints All firms Aged 20 or less Sig. Sig. Coeff. Level Coeff. Level Intercept Demographics 1.9 - Age -0.0319 -0.527 *** - Small 5.70 *** 7.75 ** - Medium 6.734 *** 9.44 ** - Large 4.031 * 4.83 Metals and related products -4.451 ** -6.54 ** Technology parameters: Share of machinery less that 10 years old 0.055 *** 0.029 * Assets employment ratio (K/l) -4.66E-8 *** -3.9E-08 *** Capacity utilization -0.038 * -0.004 Investment in technology 5.41 *** 4.24 ** Share of computer users 0.04 0.0012 Investment in R&D -0.157 0.82 Number of observations 241 161 R-squared 0.265 0.17 Source: Authors estimates *** = sig, level 1 - 5% ** = sig. Level 6 - 15% * - sig. Level 16 - 30 % 28 For manufacturing firms, technology and the quality of assets are the second most important determinants of investment growth (Table --- above). This makes sense as the quality of machinery or equipment and the technology in use jointly affect the level of productivity and profits. The three key policy related determinants of investment in Tanzania's manufacturing sector are (i) the proportion of a firm's machinery that is 10 or less years old ­ the larger the share, the better the performance, and the greater the impact on investment growth; (ii) the capital assets to employment ratios reflecting the capital intensity of production ­ the higher this ratio, the more investors need to invest in capital to obtain the same unit of output. Not surprisingly, this has a negative impact on investment growth; and (iii) the proportion of firms that invest in the use of new technology. This factor contributes the most to investment growth and is strongly related with the proportion of computer users in the firm, as well as the level of investment in R&D to develop or acquire new technologies (table --- above). The last two variables are not significant due to severe multicollinearity in the model. Another factor such as the level of capacity utilization (that averages about 60 percent across the sector) should also have a positive correlation with investment but appears with a negative sign and is insignificant. This is probably due to multicollinearity in the model (Table__). At the bivariate level, there is a robust relationship between the two. Further analysis reveals that capacity utilization and the age of the firm's machinery that is less than 10 years old are strongly correlated (the source of the multicollinearity in the model!). This makes sense. o About 48 percent of all firms including 65 percent of all exporters had invested in new technology during 1998 ­ 2000. Micro firms made hardly any investment while between 61 ­ 77 percent of the medium and large firms invested in new technology. Agro industry, chemicals and plastics were the leading sectors in 29 which over 60 percent of the firms invested in new technologies. The key source was technology licensed from a foreign owned company ­ on average about 37 percent of the large firms used this source while the numbers got smaller for smaller firms. o Across all firms, only 5 percent of the workforce could use computers but this number applied mainly to medium and large firms for whom the average was 8 and 15 percent respectively. Interestingly, relative to exporters, more workers in non-exporting firms use computers. Among sectors, textiles and garments and wood and furniture has the least proportion of computer-users. o About 22 percent of all firms invested in R&D to develop new technologies. The numbers rose with firm size. Micro firms did not invest at all, but as many as 43 percent of the large firms invested in R&D to develop new technologies. About 30 percent of exporters compared to 19 percent of the non-exporters invested, indicating that export-orientation introduces incentives for firms to maintain their globally competitive edge. The average spending on R&D varied from Tz. Sh. 0.3 mln in small firms to 5.0 mln in large firms. Relative to an average of Tz. Sh. 3.9 mln for all sectors, firms in agro industry and metals invested an average of Tz. Sh. 10.0 mln each, while the outlier was the furniture and wood sector in which firms invested as much as Tz.Sh. 300.0 mln.. o On average, 12 percent of the firms were ISO certified with the highest proportion among large firms (37 percent) and exporters (24 percent). Firms manufacturing plastics, textiles, construction materials and agro products had higher levels of ISO certification than others. These descriptive statistics suggest that nearly half of Tanzania's firms recognize the benefits of new technology and have invested in it to some extent, acquiring it from various sources. This is a positive trend. However, if the outcomes are measured by the proportion of computer users in the workforce and the level of ISO certification, the latter signaling global competitiveness, these firms have a long way to go. This point is underscored by the level of ISO certification among exporters (43 percent). Conspicuous is the fact that the technology wave has nearly bypassed smaller firms who neither invested in, or use modern technology in production. Given the critical importance of this source for overall firm growth, as well as investment growth, it seems worthwhile for policy makers to take on the challenge of figuring out what ,if anything, can be done by government to promote the adaptation and adoption of superior technologies at the firms level. In this context, the experience of other developing countries, outside the Africa box may be instructive (see new study on the How to of Technological Change for Faster Growth, forthcoming.). Policy makers need to explicitly focus on the following issues to boost investment growth in non-traditional ways through the use of better technologies: 30 1) given that capacity utilization is constrained by the age of the machinery and that younger machines also had a significant and positive impact on investment growth, are there any tax incentives that policy makers can offer firms to facilitate the replacement of outdated technologies to improve productivity and growth? If implemented efficiently, such a policy should more than pay for itself through increased sales and tax revenues. 2) As the capital intensity of production negatively impacts investment growth, it may be worthwhile for policymakers to think of policies that can improve labor productivity to compensate for the substitution of capital for labor. This requires a multi-pronged approach across the domain of human capital improving policies that can accelerate the production of technical skills, and the level of computer literacy. The larger public goods' aspect of these goods requires coordinated public action to increase investment growth. 3) Given the role of new technologies and the relationship between the latter and larger firms and exporters on the one hand, and the higher rates of investment growth in these groups on the other, policy makers may want to consider what types of interventions other developing countries have used to promote better technologies to in turn, promote export growth and overall manufacturing growth. As it stands, these efforts are far too small and ad hoc. It is not clear whether the investment in R&D for better technology is productive and whether research at the individual firm level is the most efficient. Most successful countries have exploited the late comer advantage to rapidly accelerate use of new technologies with efficient support from public policy. There are numerous unexplored options available for Tanzanian policy makers to pursue this hitherto neglected source of investment and overall growth. MARKET DEMAND AND INVESTMENT GROWTH Lack of adequate demand for goods manufactured in Tanzania is the fourth most important constraint to investment growth for all firms. For younger firms, this constraint is ranked number three. In light of the relative inward-orientation of manufacturing firms, this finding is neither surprising, nor at odds with the state of manufacturing firms in most of Sub-Saharan Africa. Only about 27 percent of the firms are able to export some amount of their output and no more than 18 percent at least 10 percent of it. Hence, Tanzania's manufacturing firms are constrained by the size of the domestic market that is further limited by low purchasing power, a manifestation of the country's low income status. Even exporting firms are heavily dependent on the domestic market where they sell an average of 80 percent of their output. TABLE: RELATIONSHIP BETWEEN INVESTMENT GROWTH AND DEMAND ­ IDENTIFYING THE BINDING CONSTRAINTS. All firms Aged 20 or less 31 Sig. Coeff. Level Coeff. Sig. Level Demographics - Age 0.009 -0.37 *** - Small -5.398 * -7.76 ** - Medium 3.415 * 2.83 * - Large 2.783 2.54 Metals and related products -5.088 ** -5.31 * Export growth 0.088 *** 0.10 *** Proportion of output sold domestically 0.05 *** 0.10 *** Sales to MNCs 0.021 -0.01 Sales to SOEs -0.17 * -0.14 Sales to large domestic firms 0.0847 ** -0.05 Number of observations 182 119 R-squared 0.19 0.317 Source: Authors estimates *** = sig, level 1 - 5% ** = sig. Level 6 - 15% * - sig. Level 16 - 30 % Other than continued macroeconomic stability, especially a favorable exchange rate, there are few direct policy levers available to raise aggregate demand in the domestic market. However, policies that can accelerate export growth are more likely to affect investment rates. As in other small-open economies, investment growth in Tanzania's manufacturing sector is nearly twice as responsive to higher export growth (as opposed to the share of output exported) than the level of domestic demand (measured by the share of output sold domestically, Table ---). Sales to sate-owned enterprises hamper investment growth. In contrast, sales to large domestic firms increase investment by almost as much as export growth. This is hardly surprising given that larger firms and exporters share many common characteristics. Among younger firms, investment growth is equally and strongly responsive to export growth and sales in the domestic market. The policy implications for raising aggregate demand to increase investment growth are clear: given that domestic demand will continue to remain limited by households' level of income/purchasing power, policies that can accelerate manufacturing export growth are the strongest policy levers available to decision makers. In addition to a competitive exchange rate, such a strategy requires a focused approach to improving the global, as opposed to the regional competitiveness of Tanzania's manufacturing firms. Among various options available are ones that can raise productivity through the use of superior technologies, aggressive marketing reforms to help firms penetrate new markets and a strengthening of policies that can spur nascent exports. The fact that the majority of firms selling in the domestic market perceived competition from cheaper imports (not lack of customers) as a threat to growth underscores this point. 32 INVESTMENT CLIMATE AND INVESTMENT GROWTH Adequate data to assess the overall quality of Tanzania's investment climate given by the time taken to obtain various clearances to operate in the country, and the official and un- official costs (bribes) of doing business is not available from the survey. In most cases, the number of firms that responded to this section is too few. As the investment climate is an important factor in investor decisions, two approaches are used to work with the deficient data. First, to give policymakers an idea of the systemic inefficiencies associated with the time costs (delays) and financial costs associated with the clearances needed to manufacture in Tanzania, the information provided by firms, albeit sparse, is presented descriptively. This provide direct pointers for further streamlining efficiency- improving procedures to reduce the general costs of doing business in the country. Second, to econometrically estimate the nexus between investment growth and the quality of the investment climate, we give government departments responsible for providing clearances the maximum benefit of the doubt by treating firms that did not provide answers as having no problems, i.e. they do not have to wait for too long, the fee associated with each procedure is not too high, they do not pay any bribes etc., i.e. all missing values are treated as being equal to zero. It may be noted that as this approach grossly understates the problem: for a balanced view, any judgments regarding the business climate should jointly consider the descriptive and econometric evidence. 1. The data on delays in customs clearance is good. 81 percent of all exporters responded. For customs clearances for exports and imports, the normal waiting time is 20 days (average); the longest time it took during the previous year was, on average, 36 days. For a few firms, the delays were as long as 97-150 days. 2. The delays in obtaining inspection certificates seem reasonable for non-exporters but not exporters who have to wait more than twice as long. About 40 percent of the latter reported that it took them 12 days (maximum of 72) to obtain clearances from 6 different inspectorates (tax, labor and social security, fire safety, sanitation, policy, environment etc). 71% of the exporters who responded said it took government an average of 27days to give them clearances (one firm reported a maximum of 410 days). Except for micro, all other firms pay bribes to obtain inspection clearances. 3. On average, it takes about 7 days for a firm to obtain each of the following connections: water (33% of firms), import licenses (33% of firms) and operating licenses (77% of firms). While the 7 day period seems fairly reasonable, there are two issues that suggest that improvements could reduce these time costs: one, unless, government departments coordinate the service delivery connections, it take a firm as long as 5 weeks to get all connections and become operational; and two, averages mask the enormous delays incurred by some firms. The maximum time taken for any of these connections varied between 180 ­ 360 days, sufficient to drive the potential firm out of the market! The bribes paid to obtain these connections on average ranged from Tz. Sh. 35,000 for a water connection to Tz. Sh. 150,000 for an import license. 33 4. Obtaining a telephone connection takes as long as 14 days and an electric connection takes about 30 days (33% reported). While the responses were lower for bribes paid, the numbers range from Tz. Sh. 50,000 for a telephone line to Tz. Sh. 60,000 for an electricity connection. 5. Registering or re-registering a firm, obtaining a basic permit, construction permit, or any other permit on average takes between 7 ­ 10 days each (about 30 ­ 77 percent of firms answered each question). The problem is with firms, typically medium and large ones, who often have to wait as much as 180 days. Construction permits take the longest. For one firm, the wait was as long as 998 days. The bribes paid for such permits are correspondingly high ranging from Tz. Sh. 50,000 ­ 180,000 with the outliers in the range of Tz. Sh. 0.5 ­ 2.4 million. 84% of all firms hire a company/agent to assist with registration. After adjusting the data for missing values, the time costs associated with delays in obtaining clearances and permits seem to affect investment growth more negatively than the financial costs of either bribes or fees17 (Table ---). The number of days taken to register or re-register a firm seem to affect investment growth in all firms more than the delays associated with inspections or obtaining a service connections, although all are equally significant. For younger firms, none of the indicators of the cost of doing business in Table one seem to affect investment growth adversely. In fact, the relationships are often positive though insignificant. Perhaps, one reason for these odd relationships is that the costs of doing business have been internalized by incumbent firms who treat them as fixed costs; hence, the rate of investment in manufacturing is not responsive to them.18 This hypothesis is also complicated by the fact that the rate of investment growth was almost stagnant during 2000-2002. Whether evaluated descriptively or technically, the policy implications of the costs of doing business in Tanzania are significant and serious. These costs not only constrain investment growth in incumbent firms but likely deter potential firms from entering the country's manufacturing sector. This may be one explanation why investors have not rushed to invest and expand Tanzania's manufacturing sector in several decades, as evident from its meager share of 8 percent of GDP in the past 3 ­4 decades. 17For all the reasons listed earlier, these results must be evaluated in conjunction with the descriptive evidence in the preceding paragraph. 18This argument is also validated at the bivariate level when most of these relationships are positive or insignificant with respect to investment growth. These results apply with or without adjustments for missing values. 34 TABLE: RELATIONSHIP BETWEEN INVESTMENT GROWTH AND THE BUSINESS ENVIRONMENT ­ IDENTIFYING THE BINDING CONSTRAINTS. All firms Aged 20 or less Sig. Coeff. Sig. LevelCoeff. Level Intercept 1.78 Age -0.039 -0.5 *** Small firms 6.45 *** 10.72 *** Medium firms 8.58 *** 13.22 *** Large firms 7.25 *** 10.22 *** Metals and related products -5.46 ** -4.76 * Dummy for number of bribes paid 2.02 * Average number of days to register/re-register firm -0.10 *** 0.076 *** Number of days to clear customs (X&M) -0.014 Number of days required for inspection -0.06 *** -0.003 Total value of bribes paid to get service connections Tz. Sh. Number of days required to get service connections -0.059 *** -0.008 Total official cost to register/re- register Tz. Sh. Total value of bribes paid to register firm Tz. Sh. -0.00001 Total cost of inspections Tz. Sh. Total days of delay for all clearances to do business in country 0.056 *** 0.086 *** Total cost of obtaining all clearances to operate in country Tz.Sh. 4E-06 Confident that judicial system will protect property rights Member of business association- dummy Number of observations 256 173 R-squared 0.23 0.313 Source: Authors estimates *** = sig, level 1 - 5% ** = sig. Level 6 - 15% * - sig. Level 16 - 30 % 35 OVERALL DETERMINANTS OF INVESTMENT GROWTH As evident from the preceding discussion, since each policy area covers many policy issues, the policy-specific analyses helped to identify and rank policy levers that have the greatest impact on investment growth. For example, In the context of infrastructure, the priorities are: electricity, roads, freight transport and access to the internet- these affect investment growth disproportionately more than other services. For younger firms, access to industrial water is an additional problem. While the establishment of industrial estates has attracted many manufacturing firms, they have failed to deliver the benefits they were supposed to. (Help: what was the purpose of the industrial estates? ) In the context of the financial sector, availability of bank finance for start-up capital, particularly for younger firms, is critical. Recent financial sector reforms and lower interest rates notwithstanding, the insignificance of bank loans for investment and working capital is evidence of the negligible role played by banks in Tanzania's manufacturing sector. In the context of human capital, the critical policy priorities for raising investment are: higher rates of completion in primary education, formal training for the workforce, manager's education, and policies that can raise worker productivity. With respect to primary education, this is a biased result, as the absence of a critical mass of more educated workers has led incumbent investors to self-select investment in low value-added activities that generate low skills-low wage jobs. For higher worker productivity, growth and better quality jobs, policy makers need to consider longer term investments in technical education. In the context of the use of superior technologies that can raise worker productivity, public policies that can assist firms to acquire/replace new machinery/equipment, invest in new technologies are key priorities. In the context of reducing the costs of doing business, policies that can reduce the delays caused in registering/re-registering firms, customs clearances, inspections and service connections are key. In the context of market demand, policies that can promote more and faster export growth are key. Table ­ presents the combined effect of the determinants of investment growth in Tanzania during 2000-02. Most of the results are consistent with the policy-specific investigation of the earlier sections.19 Investment grew least in micro firms and the fastest in medium sized firms. No particular sector mattered but investment growth was sensitive to location. Firms that were formerly state-owned (SOEs) dampened investment growth (coeff ­0.31). The more recent the privatization leading to a large boost of new investment, the smaller the new investment. The lumpy nature of investment validates this result, e.g. firms that invested in 2000 did not invest again in 2001. 19In several cases, the signs are the opposite of those in the policy specific analysis (formal worker training for example) or the coefficient is not significant (electricity generator dummy or age of the generator). These are attributable to the technical problems (severe multicollinearity discussed earlier). 36 TABLE: THE DETERMINANCTS OF INVESTMENT GROWTH IN TANZANIA - identifying the binding constraints All firms Aged 20 or less Coeff. Sig. Level Coeff. Sig. Level Intercept Age 0.32 *** -0.29 * Location in industrial estate -2.80 -0.667 Location in small town (pop. < 50,000) -8.12 * -5.07 * Micro firm -7.88 * -9.397 ** Medium firm 4.10 * Export growth rate 0.038 0.11 *** Share of output sold domestically 0.025 0.119 *** Sale to large domestic firms 0.116 ** Share of foreign firm equity 0.016 Investment capital - bank -0.06 * -0.037 Working capital - private 0.06 ** -0.019 Prior SOE -0.31 ** Age of equipment < 10 yrs 0.09 *** 0.044 ** Capital-labor ratio -1.02E-11 *** -3.4E-12 * Workers education - primary complete 0.08 *** 0.052 ** Formal worker training provided -4.72 ** -0.118 Manager's education -1.439 ** Investment in new technology dummy 8.95 *** 5.95 *** Age of the electricity generator -0.14 Own generator -4.21 1.98 Cost of telecommunications 0 0.00002 * Invested in private road 3.41 8.725 *** In transit cargo losses -0.48 ** Provide own freight transport -4.15 * -3.12 Transport-related production losses 0.35 *** Dummy for number of times bribes are paid -4.08 ** Total bribes paid to do business in country 0.00009 * No. of days to clear customs (X& M) 0.147 ** Averagenumber of days to register/re-register firm 0.248 *** Confident that judicial system will protect property rights -0.972 -0.736 Number of observations 142 107 R-squared 0.47 0.52 *** = sig, level 1 - 5% ** = sig. Level 6 - 15% * - sig. Level 16 - 30 % Table ­ highlights the importance of several policies levers, including those listed earlier in this section. Firm location has a strong impact on investment growth. Firms in larger 37 cities/towns invested more than those in smaller towns with populations less than 50,000 (coefficient of ­8.12), suggesting that there is a case for urban agglomeration-related economies (in the form of better infrastructure, labor force, markets etc.). Firm location in industrial estates had a strong but negative effect (coefficient - -2.80) on investment growth. It appears, that the latter were created for some purpose but have ceased to serve it anymore (tax breaks? Better infrastructure?). Policy issue : Infrastructure is among the most important determinants of investment growth in manufacturing but it appears difficult for government to improve the quality and quantity of infrastructure economy-wide, at least in the medium term. Is there a case for developing a few urban agglomerations (would the industrial estates perform this function??) where government can target larger quantity (as in more electric and water connections, roads etc.), and better quality industrial infrastructure to attract investors? As most firms already finance most services to some extent, some cost recovery should be possible. Lessons from export processing zones could be applied more widely in this case. The strongest contribution (coefficient of 8.95) to investment growth is associated with investments in new technologies that help firms to remain globally competitive and raise productivity. As was noted earlier, use of superior technologies contains several aspects of public goods if it is to be promoted to develop the manufacturing sector as opposed to a firm, typically foreign-owned in Tanzania, who can afford it. Use of newer and younger machinery, typically less than 10 years contributed positively to investment growth (coefficient of 0.09), raised capacity utilization, and productivity. 1. Policy issue : Higher skills among managers and workers (technical and computers for automated processes), timely and current industry information, means of accessing such information in an affordable manner (i.e. access to and use of computers), and the adaptation of known international technologies to local conditions are all aspects of technological capability building that have used the enabling hand of public policy in most countries to propel technological change in the private sector. The same factors also raise worker productivity that attracts investment to Tanzania. Some technological upgrading can occur through the use of newer, imported technologies in the form of machines and equipment. While this is a clear private good, given that the bulk of the firms are indigenous, policies that can assist firms to upgrade old/outdated technologies Closely related to location was the negative contribution of industrial infrastructure for manufacturing firms. Consider the combined effect of electricity-related problems, goods lost in transit due to poor transport, especially freight, and the costs firms have to incur when they invest in roads is aggregated in table ---. 2. Policy implication: The aggregated effect of infrastructure deficiencies on investment growth is probably the single most important negative factor driving the low investment growth rates in the manufacturing sector. The policy priorities are clear: electricity, roads, freight transport and access to the internet. The positive contribution of a workforce that has completed primary education (coefficient of 0.08) is more than offset by the negative effects of firm expenditures in the 38 formal training of workers (coefficient of ­4.72). One way to rationalize this outcome in the context of Tanzania's labor market characterized by a paucity of skills other than primary is that although the availability of trained workers enhances investment, for firms training imposes additional costs that are significant and high. At a broader level, it also deters investment in high value production due to the public goods aspect of training in an economy where skills have a high premium and lead to negative externalities. Policy issue : In addition to primary education, public policy must focus on technical skills development and training to attract investment in high-value production. Bank financed investment capital (coefficient of ­0.06) reduced investment growth , probably because high interest rates made investment costly. In comparison, privately financing had an equally significant and positive effect (coefficient of 0.06) on investment growth. Besides addressing the direct financial constraint for investment or working capital, there are several positive externalities associated with a financial sector that can reach out to more and smaller domestic firms at affordable interest rates. For example, access to superior technologies, private costs of infrastructure financing, investment in new machinery all require easy and timely access to affordable capital. Policy issue : Access to commercial funds for investment and working capital, and lower interest rates are necessary to facilitate investment growth in manufacturing ­ apart from direct investment in expansion of facilities, this should also enable firms to upgrade technology, purchase newer machinery and equipment, invest in private infrastructure and critical skills. If not addressed, investment growth will remain dependent on foreign firms who have access to parent company finance, or the availability of retained earnings. For reasons discussed in the preceding section, the investment climate indicators are not significant in the overall investment model in table --- . Nevertheless, closer inspection in the preceding section suggests that delays in obtaining licenses and permits are more important than the financial costs. 1. Policy implication: Policies that reduce the overall time costs of obtaining licenses and permits, preferably by training government officials in operating a one stop shop for all investors, will go a long way in reducing the costs for incumbent firms and likely attract new investors who presently opt for other locations with lower time costs. Further, polices that can instill more confidence in the judicial system to protect private property rights are needed. Longer term leases (100 years?) to attract fixed investment in land and buildings is one way to go. Commercial courts that can speedily resolve disputes is another. Technical problems similar to those in the case of doing business resulted in masking the strong and positive effect of export growth on investment growth. Of course, as only 18 percent of the firms are exporters, being able to sell more domestically is critical to spurring investment but there are no policy levers for this. 2. Policy implication: To lift the demand constraint on investment growth in Tanzania's manufacturing sector, there is no substitute to public policies that can encourage firms to export more and faster. This issue is examined next. 39 PERCEPTIONS OF INVESTORS Table: Perceptions of firms - Percentage of Enterprises rating problems as major or very severe constraints on enterprise operations and growth in Tanzania in 2003 80 73 59 58 60 56 51 48 43 40 31 31 27 25 25 25 24 23 20 20 12 12 0 y n s ns o tem ns erta cen tation dis l sys ulatio catio Cost ofTax regu unc li ss to Lanpracticspor y ness s of workers ve r d rates ing ation ptio ncingstabi lity inty sing rde es Tax Elec Financ inistr tricit u lation Corr adm Acce Tran Lega r reg muni Acceroec ss to finaonic intoms om Cus egu lator usi com B e, theft andSkill petiti Labo com Tele Mac R Crim Anti- Source: Tanzania Investment Climate Assessment, 2003 Results from the 2003 RPED Survey Report list the rankings that firms assigned to the main obstacles to doing business in 2003 (Table ---). As in previous years, taxes and their administration were in the top five constraints.20 While the corporate taxes in Tanzania are comparable to Kenya or Uganda, the payroll taxes in Tanzania are nearly twice as high, suggesting that the tax burden may indeed be a problem for manufacturing firms (Table ---). Compared to the econometric analysis in the earlier section, the top five RPED rankings consistent with our results apply to (i) electricity: our analysis shows that in addition to electricity, transport, especially freight, and roads also affect investment adversely. (ii) interest rates and access to bank finance: for those who have access to bank finance, high interest are indeed a problem, but for the large majority, access to bank capital is a serious constraint to expansion. It is difficult to validate the high rank assigned to corruption by firms as most did not respond to the level of bribes they pay. 20There is no data in the survey to confirm this but a comparison with neighboring countries indicates that at 30% or higher, taxes are among the top five constraints in Uganda and Kenya. 40 Table: Tax rates applicable to manufacturing firms in Tanzania, Ghana, Kenya and Uganda. Government Tanzania Ghana Kenya Uganda Corporate tax on profits (%) 30.0 35.0 32.5 30.0 Standard applied value added tax rate (%) 10.0 10.0 16.0 17.0 Payroll tax contribution by employee (%) 10.0 5.0 5.0 5.0 Payroll tax contribution by employer (%) 10.0 12.5 5.0 5.0 Time spent with officials negotiating/obtaining licenses, permits etc. (%) 14.4 12.55 10.47 14.77 Source: Africa Competitiveness Report 2000/2001, World Economic Forum 4. EXPORTS AND GROWTH IN MANUFACTURING Manufacturing's contribution to Tanzania's exports has improved since 1999. Its share in non-gold exports (that were about 52 percent of total exports in 2003) increased to 17 percent in 2003 (US$ 100 million) from 6 percent in 1999, just exceeding the average share of 16 percent during 1991-98 (Table --). Overall, Tanzania's non-gold21 export performance was dominated by agricultural exports, especially non-traditional goods like cut flowers and fish products whose share rose from 19 to 46 percent during 1999 and 2003. 21Gold and other mineral exports accounted for 48 percent of total merchandise exports in 2003. In 1991- 98, their share was 8 percent of total merchandise exports. 41 Share of total exports - gold and non-gold (%) 1985-90 1991-98 1999 2000 2001 2002 2003 1999-03 TOTAL 100 100 100 100 100 100 100 100 of which gold 0 8 13 27 39 43 48 37 of which non-gold exports 100 92 87 73 61 57 52 63 Share of non-gold exports 100 100 100 100 100 100 100 Traditional exports Coffee 21 16 17 12 7 8 12 Cotton 19 6 8 7 6 8 7 Sisal 1 2 1 1 1 1 1 Tea 6 5 7 6 6 4 6 Tobacco 6 9 8 8 11 7 8 Cashewnuts 11 21 17 12 9 7 13 Cloves 0 4 2 3 1 2 2 SUBTOTAL 64 64 60 49 40 37 49 Non-Traditional exports Petroleum products 2 0 0 0 0 0 0 Manufactured goods 16 6 9 12 13 17 12 Others 19 29 31 39 48 46 39 SUBTOTAL 36 36 40 51 60 63 51 Source: Economic survey, Table 18 and 19 While manufactured export growth could not have occurred without a favorable trend in the real exchange rate in the 1990s (Figures --& --) , an equally if not more important factor seems to have been wide ranging reforms that generated large efficiency gains and led to accelerated production in manufacturing and boosted the competitiveness of the country's manufacturing sector. 42 Tanzania: Exports and the Effective Exchange rate, 1991=2003 1200.00 160.0 140.0 1000.00 Effective Exchange rate 120.0 niolli 800.00 100.0 m$SU 600.00 Total Exports 80.0 60.0 400.00 40.0 200.00 20.0 Manufactured Exports 0.00 0.0 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Source : Economic Survey Tab 19 and The IMF (Eff. Exch. Rate, annual average 1990 =100) Traditional and Non-Traditional Exports (US $ Million) 1000 900 800 niolli 700 600 m$ 500 SU 400 300 200 100 0 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Petroleum Products Minerals Manufactured Goods Other Exports Traditional Exports Source : Tab 19 Economic Survey 43 SIZE OF THE MARKET, EXPORT ABILITY AND OPPORTUNITY FOR GROWTH Tanzania's low per capita income levels constrain the size of the internal market for its manufacturing firms. For firms who are competitive, exporting relaxes the local demand constraint, unleashing the potential for growth. Micro evidence validates this hypothesis (Table ). Only about 18 percent of the manufacturing firms export some proportion of their production. Defining exporters as those who export at least 10% of their production, unravels a large part of the growth story. Median22 growth in firms that are exporters was about 23-24 percent in 2001-02 compared to 8 ­ 11 percent per annum for non-exporters. Table : Relationship between export growth and overall growth in manufacturing Non-exporters Exporters Non-exporters Exporters Sales growth Median Median Mean Mean in 2001 8% 24% 11% 114% in 2002 11% 19% 61% 25% Investment rate 1.63% 6.9% 0.15% 6.36% 2002 Source: Firm survey data The nexus between the manufactured supply response and manufactured exports is strong in at least one way: as a small open economy, ceteris paribus (especially the exchange rate), in the short term Tanzanian exporters can sell all that they can produce. This implies that if the domestic constraints to production can be relaxed, exporting can produce some rapid short term gains, as evident from the double digit growth experienced in recent years (Table ---). In the medium to longer term, the ability to export will be determined, as in any other country, by the global competitiveness of manufacturing firms. Table: Performance of the manufacturing sector 1980 - 2003 1980 1985 1990 1995 1999 2000 2001 2002 2003 Exports of manufactures (Cur. US$ mln.) 90.60 32.80 72.60 109.20 21.50 34.20 56.20 65.90 73.00 Ratio of Exports to Production (%) 11.47 9.10 20.09 31.30 3.68 5.48 8.72 9.98 10.58 Growth of manu exports % -34.29 -0.91 3.13 41.82 -27.85 59.07 64.33 17.26 10.77 Growth of manuf VA % 2.08 5.28 12.93 1.81 7.00 3.20 2.44 4.56 Source: Regional database(SIMA) At least 50 percent of the exporters sold 71 percent of their output to global markets either directly or indirectly. Of the other half, 25 percent of the firms exported 20 percent of output while the remainder exported nearly 100. Clearly, reinforcing the ability to produce and export is key to growth in Tanzania's manufacturing firms. What are the determinants of exporting and what can policy makers do about them? We turn to these issues next. 22Given the high degree of variability, medians are more appropriate measure than means. 44 DETERMINANTS OF EXPORTING Is exporting a function of size?While exporting is more pervasive among larger firms, for those smaller firms who are competitive, exporting accounts for a large share of their business. In 2001-02, about 11% of small firms, 33 percent of medium sized firms and 55% of large firms were exporters. The amount of total production exported was not related to firm size, suggesting that if a firm is globally competitive, it can export most of what it produces. Smaller firms exported more than medium sized firms. The median proportion exported was about 56% in small firms and 82-85 percent in large firms. Among medium sized firms, the proportion exported tumbled from 48% in 2001 to 30 % in 2001. Table : Share of firms that are exporters, by sector % of firms that exported in past 5 years Agro-processing 46.2 Chemicals and Paints 10.3 Construction materials 1.3 Metals 7.7 Furniture/Wood 12.8 Paper, Printing, Publishing 3.8 Plastics 3.8 Textile/Garment and Leather prod. 14.1 Total 100.0 Does location or sector matter? In general, firm growth is not affected by whether the firm is located in Dar el Salaam or some other part of the country. However, for exporters, locations outside Dar el Salaam seem more conducive to higher growth. 73 percent of all exporters are located outside Dar. Perhaps locations outside Dar play a facilitating role in providing easy access to local inputs such as agricultural produce for Tanzania's fastest growing export industry, i.e. agro-industry with the largest proportion of exporters (Table ...). Does access to finance matter? In most exporting firms, about 61% of the equity is private and domestically owned; about 31% is foreign owned, and 3 % is government owned compared to non-exporters among whom the share of domestic private equity is 75% and foreign equity is 16%. Among exporters, own resources and equity capital were the main sources of start-up compared to non-exporters who had twice the access to bank capital for start-up (11%). However, exporters enjoy access to commercial loans through 45 overdraft privileges (49% compared to only 28% of the non-exporters); lower interest rates, and greater access to bank financed working capital (25% compared to only 10% of the non-exporters). Demand for skilled labor matters. In contrast to non-exporters, exporting firms demand higher education levels (graduate, technical and vocational) and higher skills. For example the skilled to unskilled worker ratio in exporting firms is 3.8 compared to 2.3 in non-exporters. Exporters also pay a premium for higher skills: for example, the average premium for managers is about 20%, for professionals about 37%, for technically skilled workers about 19%. Unskilled workers in exporting firms, on average, earn 6% less while non-production workers earn the same. More exporting firms (71%) invest in formal training than non-exporters (47%). A larger proportion of managers in firms that export are foreign, with more experience, especially in the export business. The proportion of workers with computer skills is on average, five times higher than in non-exporters (only 2% of the workforce can use computers relative to 10% in exporters). Infrastructure matters. Table: Status of infrastructure in manufacturing firms % of % of non- exporters exporters Location in industrial estate 66.67 60.56 Firm has its own generator 82.69 47.22 Firm has its own water well 54.90 30.00 Firm has installed its own water infrastructure 49.02 27.75 Firm shares communal water source 35.42 6.83 Firm has invested in own road 28.85 6.02 Firm provided worker transport 38.46 22.22 Firm has invested in own freight transport 34.62 14.35 Firm has invested in own waste disposal facility 55.77 36.11 firm has invested in other infrastructure 5.77 3.7 Firm has internet connectivity 80.77 42.13 46 Table --- provides a clear indication of why the production response of exporters, measured by either sales or investment growth, is many-fold that of non-exporters. In spite of overly long waiting times for clearances and registration permits, and the associated official and unofficial costs, exporters have a clear comparative advantage ­ in the absence of publicly provided infrastructure and basic services, they have more resources to finance their own infrastructure - they simply have more and better infrastructure to work with. For most of the key types of infrastructure, there are at least twice as many if not more exporting firms than there are non-exporters. 47 The amount exported to each global destination matters. Two main regional markets account for about 70% of Tanzanian exports. Western Europe and the regional African market, dominated by Kenya and Uganda absorb equal 35% (mean) of the total export value (table ---). A fairly large number of exporters are involved in the process - about 65 - 67% of the firms sell nearly 35% of the exports to these two main destinations, suggesting that at least of Tanzanian exporters are globally as competitive in Western Europe as they are in the African market where competition from low income Asian exporters such as China is rising. Considering that there were only about 52 firms in the sample that exported more than 10% of their output, it is encouraging that exports to developed country markets are not dominated by one or two large firms, as is often the case in many low income countries. Clearly, if the critical inputs can be put in place, there is scope for more entrants in the export of manufactures. Table: Main destinations and share of output exported to each % Value No. of firms % of firms exported to that export to that export to each region this this Mean Median destination destination West Europe 34.77 10 30 66.7 East Europe & Central Asia 3.88 0 7 15.6 North America (USA & Canada) 5.90 0 8 17.8 North Africa/ Middle East 7.50 0 12 26.7 Other Asia 9.10 0 13 28.9 Kenya 17.90 0 29 64.4 Uganda 3.08 0 12 26.7 Other Africa 15.44 0 25 55.6 Others 2.33 0 6 13.3 Source: Firm Survey, 2003 48 Which government programs promote export growth? Table --- shows that of approximately 75 firms that export a positive amount of their output, less than 25% each use the 13 or so use the export incentives/programs. By far the most popular programs are retention of export proceeds in a foreign country (typically the export destination) and the bonded warehouse or similar scheme. In order of importance, in the second rank are the customs duty drawback, profit tax exemption and duty certificates used by 14 ­ 16 percent of the exporters. Of the remaining, the export development fund (EDF), export processing zone (EPZ), export adjustment fund scheme and foreign inputs facility are clear cases for rationalization or closure in their present form, given their low utilization rates. Table: Usefulness of existing export promotion programs No. of firms that No. of firms that use the % fo firms that responded incentive use the incentive Manufacture-in-Bond Scheme 73 5 6.8 Customs Duty Drawback 73 12 16.4 Duty suspension on imported inputs 75 4 5.3 Bonded warehouse or similar scheme 75 16 21.3 Profit tax exemption 73 10 13.7 Export Credit Guarantee e.g. Nexim 74 4 5.4 Export Development Fund (EDF) 74 2 2.7 Retention of export proceeds in foreign country 73 18 24.7 Export Processing Zone 74 2 2.7 Export Adjustment Fund Scheme 73 0 0.0 Foreign Inputs Facility 75 2 2.7 Duty Certificates 74 11 15.0 Other 57 3 5.3 Source: Firm Survey 2003 Table ---- presents the key determinants of manufactured exports defined by the proportion of output exported by firms. Relative to garments and textiles, firms in agro processing contributed positively to exports but those in chemicals, furniture and wood, and paper and printing each reduced the proportion exported by almost equal amounts. Large firms, especially those owned by individuals whose nationality was Kenyan were the main drivers of export growth, while firms with Ugandan owners performed poorly, reducing exports. In terms of ethnicity, firms with owners of African origin contributed negatively to growth. The relationship between export growth and firms with Asian owners was not significant. These results reflect the entrepreneurial skills and export experience of firm owners and make sense ­ unlike most sub-Saharan African countries, Kenya has the longest tradition of private enterprise; unfortunately, in the past decades, an adverse domestic climate has encouraged most of its entrepreneurs to cross the border 49 and operate in neighboring countries such as Tanzania where the investment climate is presently more conducive to exporting. In addition to the owners experience, firms as organizations that have benefited from prior learning to export to global markets in the past 5 years are able to export more. Similarly, given the complexity of exporting to increasingly challenging global markets, new entrants into the Tanzanian market (but not necessarily new to the global market) also seem to export more. They probably come equipped with the knowledge that exporters need, and do not need much lead time to learn the skills. 50 TABLE: DETERMINANTS OF EXPORT GROWTH ­ IDENTIFYING THE BINDING CONSTRAINTS. Coeff. Sig. Level Intercept 7.83 ** Age -0.19 *** Sector - agro proc 5.16 ** Sector - chemicals -6.13 * Sector - furniture/wood -4.73 * Sector - paper/printing -7.35 ** Size - large 10.63 *** Owner - Kenyan 26.8 *** Owner - Ugandan 12.98 * Ethnic origin - Asian -3.32 * Ethnic origin - African -5.16 ** Machinery's age < 5 yrs. 0.03 % of workers that can use computers 0.22 *** Share of skilled workers in workforce -0.05 * Share of non-prod. workers in workforce 0.07 * share of workers with grad.&post-grad.edu. 0.22 *** Years when firm started exporting 0.76 *** Exporting experience in past 5 years? 7.15 ** Bank loan financed start-up costs -0.09 ** Access to bank financed working capital 0.06 * Access to bank financed investment capital -0.05 * Time taken to clear customs at border -0.26 *** Share of exports to specific destination Share of output exported to Western Europe 0.36 *** Share of output exported to Eastern Europe& Central Asia 1.03 *** Share of output exported to the U.S. 0.48 *** Share of output exported to other Asian countries 0.41 *** Share of output exported to Africa -0.11 ** Trade promotion programs Manufacture-in-Bond Scheme -38.06 *** Duty suspension on imported inputs -32.66 *** Bonded warehouse or similar scheme 12.75 ** Export Credit Guarantee e.g. Nexim -3.37 Export Development Fund (EDF) -51.3 *** Retention of export proceeds in foreign country 14.77 *** Export Processing Zone 33.31 *** Duty Certificates 9.38 ** Number of observations 240 R-squared 0.71 *** = sig, level 1 - 5%; ** = sig. Level 6 - 15%; * = sig. Level 16 - 30 %; 51 The availability of workers with graduate and post graduate education and computer literacy enhances exports. This is natural as more manufacturing processes, especially in heavy industry such as chemicals, paper, metals, machinery etc are becoming increasingly computerized. However, the premium on skilled wages given the paucity of skills in Tanzania imposes additional costs on exporting firms relative to non-exporters, and reduces export growth. While Tanzania's exporters have greater and easier access to the banking system, with the exception of working capital that promotes exports, bank-financed investment and start-up capital are costly and dampen export growth. For technical reasons, infrastructure related variables were excluded from the export growth model but Table .... provides a clear guide to how critical this factor is in releasing the constraints to export growth in Tanzania. Among the bureaucratic hurdles, the time taken to clear customs is an unambiguous deterrent to export growth. There seems to be discrimination against exporters who have to, on average, wait as long as 20 days compared to only 7 days for non-exporters to obtain customs clearance for exports and imports. Exports destined to SADC or the local regional markets in Kenya and Uganda do not grow as fast as those outside of Africa, especially to Western Europe, Eastern Europe, the US, and other Asian countries where exporters can sell larger amounts without running into the market size constraints that characterize low income African markets, and where competition from other regional African exporters selling similar products is high. Among the export promotion programs that help to raise exports are the bonded warehouse scheme, policies that permit the retention of foreign exchange earnings in another country, EPZs and duty certificates. While the others make sense, given their usage among exporters, the significance of EPZ presently used by only about 3% of firms is curious. Further investigation into what is special about the EPZs in Tanzania is needed to inform this issue. EXPORT GROWTH AND COMPETITIVENESS ­ WHAT DOES IT MEAN AND WHAT WILL IT TAKE TO ACCELERATE EXPORT GROWTH? To expand exports, Tanzania's incumbent exporters in manufacturing need to export larger amounts, and more manufacturing firms need to start exporting. Compared to its neighbors in Sub-Saharan Africa and other global competitors who are low wage ­ high skills producers competing with Tanzanian firms in most markets, there is sufficient room for expansion, especially as the size and purchasing power in the domestic market remain limited. Table C1 and figure F1 provide international comparisons to support this case. Not only do its Asian competitors such as India, Bangladesh and China export a significantly larger proportion of the manufactured output in spite of much larger 52 domestic markets with higher purchasing power, the share of manufacturing in their economies is also significantly larger and has grown over time. Figure F1: Percentage of firms exporting at least 20 percent of production 60 50 production of 40 30 >20% 20 10 exporting 0 firms % NigeriaEthiopia via nda l bia h Boli bique China zam Eritrea Uga Tanzania IndiaNicaragua KenyaZam Senega Morocco Mo Banglades Source: Figure 2 in Eifert, Gelb and Ramachandran, 2005. Table C1: Selected Economic Indicators, 2000­2002 GNI per Ag Investment (FDI), Manufacturing, Country Trade Manufacturing %GDP capita, $ %GDP %GDP %GDP (growth) % merchandise exports Eritrea 160 111 21 39 (5.3) 8 (5.4) - Ethiopia 100 49 52 18 (1.2) 7 (5.0) 9.8 Nigeria 290 81 35 20 (2.4) 4 (3.7) 0.2 Kenya 360 57 19 14 (0.4) 13 (1.0) 22 Mozambique 210 79 23 40 (8.6) 13 (9.2) 7.5 Senegal 480 38 18 18 (1.3) 13 (7.3) 37 Tanzania 280 71 45 17 (3.7) 8 (5.9) 18 Uganda 250 40 31 20 (2.6) 10 (2.9) 6.5 Zambia 330 75 22 18 (2.9) 11 (4.5) 17 China 940 52 15 37 (3.7) 38 (8.7) 88 Bangladesh 380 33 23 23 (0.3) 16 (5.6) 92 India 480 31 23 22 (0.6) 15 (5.6) 77 Algeria 1,720 61 10 23 (1.6) 8 (-1.0) 2.3 Morocco 1,190 66 16 25 (4.2) 17 (4.0) 64 Bolivia 900 49 15 16 (9.3) 15 (1.9) 17* Nicaragua 720 73 18 29 (5.0) 14 (1.2) 13 *was 40% in 1999, but oil & gas production has reduced it rapidly since 53 Source: Table 4 of Eifert, Gelb and Ramachandran, 2005. 54 In addition to the determinants of exports as shown in table ---, higher export growth in any country today is critically dependent upon the global competitiveness of the export sector. In the medium to longer term, the ability to simply produce manufactured output that is not absorbed by the domestic demand does not necessarily imply that it can be sold in the global market place. This is particularly true of countries like Tanzania who are relatively small players in an increasingly complex global marketplace with no obvious or established edge in any manufacturing activity. Several factors define global competitiveness today. 1. To accelerate exports of manufactures, Tanzanian firms need to remain competitive with not just their regional counterparts (Kenya, Uganda or SADC) but especially the low income exporters in these regional markets. This issue is sensitive to the lower purchasing power of the Sub-Saharan African market that is more amenable to high skill-low wage producers like China who benefit from higher investment inflows attracted by higher labor productivity. The issue applies equally to global markets in Europe and Asia which are destinations for Tanzanian exports. Table C2: Net value-added per worker and capital per worker, median by firm size VA / L, $ K / L, $ country Micro small medium large very large Micro small medium large very large Eritrea 2,900 2,450 5,450 2,000 1,600 14,750 17,700 52,050 52,650 14,500 Ethiopia 600 550 750 1,050 650 950 2,450 3,750 4,600 4,400 Nigeria .. 1,400 1,500 2,850 3,100 .. 17,200 12,850 24,900 19,150 Kenya .. 1,750 3,100 6,300 2,300 12,600 6,800 11,700 10,000 6,800 Mozambique 350 1,250 2,800 2,200 .. 2,700 6,200 5,600 12,250 .. Senegal 6,150 7,500 17,100 15,600 14,500 9,450 6,900 11,300 11,950 1,000 Tanzania 1,350 1,850 4,200 3,400 6,800 1,050 5,900 4,750 13,250 13,150 Uganda 925 1,000 1,600 4,800 950 800 1,550 4,700 8,850 1,050 Zambia .. 800 950 1,250 2,500 .. 9,650 14,000 6,700 13,750 Bangladesh 950 1,300 1,650 1,200 1,150 400 1,450 1,650 800 1,150 India 1,850 3,500 3,200 3,800 5,750 1,500 1,500 1,850 2,850 6,000 Bolivia 1,300 1,700 3,050 2,450 7,350 1,200 2,150 4,700 5,350 11,050 Nicaragua 2,100 2,450 3,650 10,200 3,700 950 1,300 1,250 5,700 1,350 China 1,850 2,350 4,150 3,850 4,250 1,400 850 1,250 1,400 4,200 Morocco 4,100 4,850 3,600 5,250 5,750 4,350 2,150 2,550 Source: Table 7 of Eifert, Gelb and Ramachandran, 2005. It would be myopic to compete with neighboring Kenya or Uganda when the competitor may be China in the home and neighboring markets. In a nutshell, to attract more investment for higher export growth, Tanzanian firms need to raise the productivity of the labor force and the returns to capital (profitability) to global levels in the exporting sector/industry. Table C2 shows that while the value added per worker in dollar terms in large and very large firms in Tanzania is 55 competitive with its Asian counterparts, the capital intensity of production is significantly higher in Tanzania. For each unit of labor, investors in Tanzania must invest more than 6 times as much as they invest in India or China and nearly 12 times what they would invest in Bangladesh. The labor intensity of production in Asia is clearly many times higher due to higher skills and increased production. Besides higher employment creation, this also raises the returns to capital (see Figure F2. For global competitiveness, firms need both higher labor value added and higher returns to capital. The latter is zero in Tanzania relative to returns of 5 ­ 9 percent in Kenya and Uganda, and 50 percent or more in Asia which is attracting investors and exporting significantly more manufactures. Figure F2: Median profit margin & median return on capital 0.60 0.50 0.40 0.30 0.20 0.10 0.00 -0.10 -0.20 EritreaEthiopiaNige ria ZambTanzaniambiqueUganda ia KenyaSenegaMo l rocc olivia raguagladesh China o B India za Nica Mo Ban profit margin return on capital *profit margin = (profits / sales ­ 1); return on capital = (profits / capital ­ 1) SOURCE: FIGURE 3 IN EIFERT, GELB AND RAMACHANDRAN, 2005 2. Tanzania's manufactured exports fall into three categories: heavy industry (chemical, paints, metal products, paper, plastics etc.) based on its rich minerals base; agro-processing based on its rich natural resource base, and light manufactures (textiles and garments, furniture and wood products) that seem to be declining because of competition from other lower wage producers, especially after the expiration of the MFA. Firms producing these three categories of manufactures need different strategies to remain globally competitive and expand exports. As heavy industry relies on locally available natural resources, public policies to support its expansion and raise productivity should be of high priority. Agro-processing is also natural resource based but of a different type. The 56 processing of local non-traditional produce such as fish, fruits and vegetables or even cashew is relatively new to Tanzanian firms and must comply with complex phyto-sanitary standards to gain entry in developed country markets. While difficult, these exports also provide new opportunities for Tanzanian firms as they are unlikely to run into market size constraints in developed countries. The scope for growth is also enhanced by prospects of moving up the value chain. For example, with better technologies and higher technological capability, fish processing firms could graduate from simple fillets to higher value semi-prepared fish foods for the EU, presently their single largest destination for fish products. and quality and quantity of infrastructure as almost all other intermediate inputs can be imported. Expansion of light manufactures is more difficult as it entails competition with large MNC-managed global supply chains with a high level of technological sophistication and keen cost competition from low-wage ­ hgh skills producers. Without entry into a global segments of these chains, there is little scope to expand exports for small producers like Tanzania. All three strategies require technological upgrading to raise productivity, attract new investment, reduce costs and expand exports. The bottom line on competitiveness implies improving labor productivity through technical skills and technological upgrading to achieve and continually maintain global standards. Two types of strategies are needed to achieve this. 1. First, public policies that (1) facilitate cost reductions by minimizing wastage ­ examples are the provision of more and higher quality infrastructure (reducing waste due to poor freight transport; improve capacity utilization through regular power supply, more roads etc.) so firms can re-allocate private resources towards investment in firm expansion instead of the private provision of public goods. (2) public policies that enable the private sector to upgrade equipment, acquire better technologies, improve firm level learning and human capital accumulation, and reduce the entry costs of exporting. Figure F3 displays the business environment related losses in Tanzania as measured by power outages, crime, shipment losses and delivery delays are about 4% of sales relative to less than 2 percent in Uganda and even Ethiopia and under 2% in Tanzania's Asian counterparts, India, china and Bangladesh. Clearly, there is plenty of room for public policies that can contain wastage and reduce costs, as evident from the empirical analysis and these international comparisons. Another example of this is evident from Figure F4 that illustrates that the indirect costs of production, associated with private provision of infrastructure or poorly provided public infrastructure are almost 25% in Tanzania relative to 7 ­10 % in Asia, and 18% in Uganda. Table C3 underscores the same point using the example of privately financed generators that crowd out private investment in firm expansion when firms need to substitute for public goods such as power, a critical input of production. China emerges as a unique example of a developing country that until two decades ago was low income but today fares remarkably in power provision compared to even India and other sub-Saharan countries. This is part of 57 the explanation for China's rapid growth miracle that today challenges every low income exporter. 58 Figure F3: Business Environment -related losses, % of sales, average of 5th to 95th percentiles of sample 6 5 sales 4 of % 3 2 losses, 1 0 bad a gua pia que India - Calcuta da NigeriaTan a Chinadera Senegal zani Ethio EritreaUgan ZambiaKeny India - Hy Ban gladeshNicara Mozambi power outages crime, shipment losses, delivery delays Source: Figure 4 in Eifert, Gelb and Ramachandran, 2005 Figure F4: Cost structures: % of total costs, average mozambique zambia eritrea tanzania kenya ethiopia nigeria uganda bolivia morocco india senegal bangladesh nicaragua china 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 indirect labor capital inputs Source: Figure 5 in Eifert, Gelb and Ramachandran, 2005 59 Table C3: Share of Firms Owning Generator by Size Country Micro Small / Medium Large / Very Large Bangladesh 0.28 0.53 0.88 Bolivia 0.13 0.11 0.31 China 0.0 0.14 0.38 Eritrea 0.38 0.43 0.63 Ethiopia 0.03 0.23 0.43 India 0.23 0.76 0.91 Kenya 0.46 0.67 0.89 Morocco 0.14 0.15 0.28 Mozambique 0.20 0.23 0.63 Nicaragua 0.06 0.29 0.81 Nigeria 0.83 0.96 0.99 Senegal 0.23 0.19 0.16 Tanzania 0.18 0.60 0.89 Uganda 0.04 0.44 0.87 Zambia 0.30 0.28 0.61 Source: Table 12 of Eifert, Gelb and Ramachandran, 2005. Table C4: Share of Firms with Access to Loans/Overdrafts by Size Micro Small / Medium Large / Very Large Bangladesh 0.36 0.74 0.81 Bolivia 0.75 0.76 0.84 China 0.19 0.70 0.84 Eritrea 0.25 0.62 0.75 Ethiopia 0.29 0.66 0.88 India 0.42 0.72 0.88 Kenya 0.17 0.51 0.25 Mozambique 0.32 0.36 0.46 Nicaragua 0.53 0.67 0.72 Nigeria 0.33 0.64 0.89 Senegal 0.1 0.42 0.67 Tanzania 0.08 0.37 0.66 Uganda 0.07 0.35 0.72 Zambia 0.1 0.54 0.80 Source: Table 13 of Eifert, Gelb and Ramachandran, 2005. 60 Table C4 indicates an example of the facilitative role of the state in Tanzania relative to its global competitors in the arena of the financial sector. Compared to Tanzania, firms across almost all size classes - micro, SMME and large firms - in Asia and most other sub-Saharan African countries had greater access to bank finance, and over draft facilities for working and investment capital purposes. 2. The second type of strategy entails keeping up with external and global developments that define the size of the global market for a country's exports. This requires a dynamic approach and partnerships with global firms through incentives. Maintaining the competitive edge in even the packaging of cut flowers, vanilla or fish processing today is an externally driven challenge that is most efficiently achieved with external players as is happening in the cut flower industry where European firms supply the latest technologies to flower growers in Tanzania and Kenya. For example, moving up the agro processing value chain in these industries cannot be achieved by existing domestic export promotion programs or policies to attract FDI alone. More creative and comprehensive approaches are warranted. 5. EMPLOYMENT As Table ­ indicates, in 2001, manufacturing share of total employment in Tanzania was a meager 1.5 percent (245,000 jobs). Table : Distribution of the Currently Employed Labor Force by Industry (Standard Definition) 2000/01 Main Industry Number Total for all Sex Manufacturing industries Male 161,699 8,351,291 Female 83,750 8,563,513 Total 245,449 16,914,804 Percentage Male 1.9 100.0 Female 1.0 100.0 Total 1.5 100.0 Source: Intergrated Labour Force Survey, 2000/01 - Table 3.4 61 Manufacturing's contribution to production-related jobs, i.e., craft and related workers and plant and machine operators in the country was significant ­ it accounted for a third of total jobs in each category. It also employed 10 percent of Tanzania clerical workers and 10 percent of the professional staff (mangers, professionals, technical staff ). In contrast to agriculture, manufacturing created only 2 percent of the country's elementary jobs, clearly establishing that policies that are conducive to growth in manufacturing also favor the creation of better paying jobs (Table --). The policy challenge is of course more complex: in addition to investment increasing policies, policy makers must also address the creation of skills that can be employed by firms. Table: Distribution of the Currently Employed population by Main Occupation (Standard Definition) 2000/01 Main Industry Percentage Total for all Occupation Manufacturing industries Legislators, Administrators & Managers 2.9 100.0 Professionals 4.0 100.0 Technicians & Associate Professionals 2.7 100.0 Clerks 10.8 100.0 Service & Shop Workers 1.2 100.0 Agriculture & Fisheries Workers 0.0 100.0 Craft & Related Workers 33.8 100.0 Plant & Machine Operators & Assemblers 32.1 100.0 Elementary Occupations 2.1 100.0 Total 1.5 100.0 Source: Intergrated Labour Force Survey, 2000/01 - Table 3.5 Survey data shows that during 2000-01 and 2001-02, median employment growth was zero; the mean was about 7 percent. Firms in the bottom 25% experienced significantly negative growth rates, while the opposite was true for firms in the top 25 percent of the distribution. On average, these offsetting effects yielded zero growth. Towns with populations of less than 50,000 experienced negative rates of about 2 percent. Among sectors, plastics was the only one with positive average rates of 5 percent. Employment growth among exporters was no different from that among non-exporters. But exporters employed workers with higher education, especially those with diplomas, graduate, post graduate and technical degrees and paid premia for higher skills ranging 62 from 19% for skilled employees to 37 percent for professionals. They also employed more female workers (12 percent of firms) compared to non-exporters (5%), and invested more in formal worker training (70%) compared to only 47% among non-exporters. DETERMINANTS OF EMPLOYMENT GROWTH Table --- illustrates the determinants of employment growth in the manufacturing sector. Relative to textiles and garments, firms in metals and plastics create more jobs compared to agro processing or chemicals. Relative to smaller towns, firms located in cities with a population of over 1 million help to raise employment. Table : Determinants of employment growth in manufacturing firms Sig. Coeff. Level Intercept 27.6 * Age 0.21 ** Sector - agro proc -7.9 ** Sector - chemicals -9.5 ** Sector - Metals 17.4 *** Sector - plastics 25.7 *** Location in large city(pop. Over 1 ml.) 5.14 * Age of male workers < 30 yrs. 0.2 *** Age of female workers < 30 yrs. -0.08 ** Worker with grad. & post-grad educ 0.2 * Managers education 2.39 *** Log of skilled wages -3.96 *** % of exports in total output -0.12 ** Overall sales growth 0.15 *** Number of observations 158 R-squared 0.24 *** = sig, level 1 - 5%; ** = sig. Level 6 - 15%; * = sig. Level 16 - 30 %; Workers characteristics or the supply side of the labor market also affects job creation. The availability of younger males aged 30 or less contributes to job creation but the same characteristics in females has a negative effect. Higher education, especially at the graduate or post graduate level, i.e. technical level, as well as higher education among managers both affect employment positively. But the wage levels of skilled Tanzanian workers are a deterrent to employment growth. Labor demand is significantly and positively driven by overall sales growth but is negatively affected by the proportion of output that is exported. A possible reason for this odd result is that exports levels and employment growth were negatively related in 2000 ­ 02. 63 Policy issue: Export growth has a positive and significant effect on overall firm growth and investment, but a negative, albeit weak, effect on employment growth in Tanzania's manufacturing sector. This suggests that exporters in Tanzania are investing in less labor-intensive activities. The policy implications are clear: labor productivity at prevailing wage levels is too low to attract investors in labor- intensive export sectors. What can policy makers do to raise labor productivity? Improve the supply and quality of technical skills, adopt policies that enable firms to acquire better technologies that raise labor productivity and promote firm learning. 6. REFORM PRIORTIES FOR ACCELERATING GROWTH In the first section, we established from existing official data that while sales manufacturing sector has grown in the impressive range of 6 ­ 8 percent per annum since 1999, this trend has not been matched by corresponding growth in investment in manufacturing firms, either from macro or firm level data. In fact, investment growth was nearly nil during 2000-02. In later sections, we found that employment growth in manufacturing was also nil for the same period. Similarly, manufacturing exports grew at double digit rates, but unlike non-exporters, exporting firms also experienced higher investment growth. However, while manufactured exports contributed to overall economic growth, they did not contribute to employment growth. Three hypotheses seem are validated by our analysis: 1. recent growth in Tanzania's manufacturing sector seems more of an outcome of efficiency improving reforms in the late 1990s than an outcome of sustained investment in manufacturing in the recent past. However, in contrast to the past three decades since the 1960s, there was a link, albeit weak, between growth in manufacturing and investment in the 2000-02 period. This suggests that unless investment is somehow boosted, these impressive growth rates will not be sustained in the medium term. The link suggests that growth in investment, ceteris paribus, will support overall growth. 2. recent growth in manufacturing is strongly related to the export growth of manufactures ­ exporters experienced significantly higher growth rates than non- exporters. Policies that promote exports are good for growth. 3. whatever is good for growth is not good for employment in 80 percent of the manufacturing firms selling mainly in the domestic market. They did not add to employment growth. Among exporters, investment growth is higher, but is not being channeled into labor-intensive production. This implies that capital as embodied in machinery and equipment is preferable to labor skills. The bottom line: machines raise profits more than workers. Labor productivity needs to be raised to establish or re-establish (?) the link between growth and employment across exporters and non-exporters. 64 Our analysis in sections 3 ­ 5 shed light on the determinants of growth: investment and growth. In-depth analysis of each of these key determinants established the top policy priorities that can propel investment and export growth. International comparisons were also provided to reinforce the hypotheses. We use some innovative techniques to prioritize the main areas for reform, given that policymakers are already overburdened with a laundry list of things to do. In section 6, we examined the determinants of employment to understand what drives it and explore why it is disconnected from growth and investment. Each section flagged key policy issues. In every section, two common themes emerged. There was no running away from the same recurring reality: · for faster growth, more Tanzanian manufacturing firms will have to start exporting outside Africa where demand is high. For this, policymakers need to 1. facilitate greater investment, and 2. use some innovative interventions that can raise productivity, improve the global competitiveness of all firms, and hook them to global supply chains. Is it critical to appreciate that while removing the obstacles to investment is necessary, this in itself does not guarantee that Tanzanian firms, especially indigenously owned, will become globally competitive and start exporting significantly more. Catching-up with global markets is beset with complex public goods issues that cannot be achieved by the private sector alone ­ the enabling, development-hand of the public sector is needed to accomplish public-private partnerships to achieve this. A discrete and new strategy is needed to address this in addition to (i). This approach needs to be carefully orchestrated to include all firms, not just foreign owned ones, as it is believed that potential investors are deterred by occasionally voiced views that companies should remain Tanzania- owned. International experience shows that technological catch-up necessary for rapid export growth in manufacturing cannot ride on the shoulders of foreign capital alone. Policy Recommendations 1. Policies that can facilitate investment growth are well known to Tanzanian policy makers who have been pursuing them for a while now. Our analysis helps to identify the binding constraints in infrastructure, financial sector, human capital, business environment etc. to enable even more focused interventions given the paucity of time, financial resources and capacity. These are listed at the end of each discussion in sections 3 ­ 5. They are reviewed briefly below in the context of ongoing reforms to the extent we are aware of them. ........to be repeated here. 65 2. The use of innovative interventions to facilitate an increase in labor productivity, improve global competitiveness and faster export growth of manufactures is an approach that is being recommended as a discrete new strategy. This could be adopted using clever policies and programs to put the private sector in the driver's seat by enabling firms to participate in productivity-enhancing activities such as rapid technical training of the workforce as opposed to waiting for the schooling system to churn out a meager number of skills, adapt and adopt newer technologies that can improve value added, scale up production (for example, from one shift a day to three or two) and most of all, enable firms to catch-up to global standards, especially those required to connect to global supply chains in light and heavy manufactures, and comply with international standards in non-traditional sectors such a agro-processing in which Tanzania has a clear comparative advantage. The challenge is one of learning how to scale up nascent or newly emerging sectors for export markets. International experience shows that this was best done through constructive public-private partnerships using the measuring rod of export growth to pre-empt the hazards of rent-seeking or putting the public sector back into what private firms are doing; production. For more on this strategy, policy makers may wish to refer to some recent research emerging on "The How to of technological change for faster growth," (Draft, World Bank, 2005). 66 CONSTRAINTS TO TECHNOLOGY ACCESS IN TANZANIAN HORTICULTURE A CASE STUDY OF BARRIERS TO THE INTRODUCTION OF IMPROVED SEED AND PEST CONTROL TECHNOLOGIES The World Bank June 2006 This background note has been written by Annabella Skof as an input to the Tanzania CEM. Henry F. Gordon provided valuable inputs for this report. I. Introductory remarks This background note analyses the constraints to technology access in the Tanzanian horticultural sector. Access to technology in horticulture is essential to foster growth and enhance productivity in the existing horticultural sector as well as to attract further foreign investment. While access to technology has a clear impact, it is not the only condition for innovation, growth and investment. The study will however not discuss constraints to these ends. The scope of this paper is limited to barriers to the introduction of improved varieties of seeds and pest control technologies. II. Background (1) Overview of Tanzanian horticulture The Diagnostic Trade Integration Study for Tanzania explicitly identifies floriculture, vegetables, fruits and horticultural seeds as important emerging export crops of Tanzania. While the share of exports in horticulture and floriculture is comparatively small (about 1.1% of total exports or around US$ 12million in 2003), it is one of the few exports that has increased its share in total exports since the mid 1990s. A wide range of horticultural crops can be grown in Tanzania given existing diversity of agro-climatic zones. However, despite high production potential in many parts of the country, horticulture is currently only relatively well developed in the northern regions (Arusha, Kilimanjaro and Tanga) and the southern highlands (Mbeya and Iringa). With increased regional trade, Tanzania is more integrated with neighbors. Kenya is particularly important as a transit point to European markets as well as a source of final demand. Kenya still enjoys a regional first-mover-advantage in the field of horticultural industry due to its 30-year headstart in the development of infrastructure and skills essential to horticulture export. It is therefore a center of horticulture technology and skills for Tanzania. (2) Private marketing chains and private sector development in horticulture Floriculture, vegetables and the horticulture seeds sub-sectors are dominated by a few large, mostly foreign owned/managed companies producing for export. The major export destination for products of these sub-sectors is the EU market. The cut flower industry is highly concentrated in the hands of seven large companies due to high capital investments needed. All flower exporters have fully adopted the quality standards required by the main retail outlets. In particular in the seed multiplying business, the potential is large and not fully exploited with only two companies multiplying horticultural seeds. Conditions are very favorable due to cheap labor, climate and low freight costs. The two large companies in the vegetable sub-sector are contracting outgrowers as a way ton increase their export volume and to be able to fulfill export orders. However, most outgrowers are larger farms. There are initiatives to involve small- and medium- scale farmers, but these require investments to facilitate training, technical assistance (to June 2006 Draft 1 comply with high export standards) and organization of farmers. Nevertheless, there is great potential to increase such initiatives, provided that the industry expands further. In the fruit sector, only a small percentage of the production is exported to the EU markets. Most of the fruit is for domestic consumption or regional export markets in Southern and Eastern Africa. Tanzania enjoys a significant trade in regional fruit and vegetable exports to Kenya. Regional exports have the advantage that capital investments required are much lower. In order to facilitate the process of the public-private dialogue, the Government created the Tanzania Seed Trade Association (TASTA) and the Tanzania Horticultural Association (TAHA). Cooperation between exporters has been improving as a result. In December 2005, a stakeholder meeting with representatives from Ministry of Agriculture, Food Security and Cooperatives (MAFSC), Ministry of Industry and Trade, exporters and small-scale produces has been held. A committee consisting of each one representative of TAHA, TASTA, small-scale farmers, SOKOINE University and MAFSC has been formed. (3) Legal and regulatory framework In Tanzania, pesticides legislation has been in place since 1979, when the Tropical Pesticides Research Institute Act was enacted. Later, the TPRI Act was incorporated into the Plant Protection Act of 1997 and its accompanying Regulations of 1999, operationalized in 2001. The Plant Protection Act stipulates that all pesticides to be imported and used in Tanzania have to be registered with the registrar of pesticides. According to the Plant Protection Act, the Office of the Registrar of Pesticides is part of the Ministry of Agriculture, Food Security and Cooperatives and the registrar is appointed and employed by MAFSC. Yet the registrar is based at TPRI in Arusha, and for all practical purposes, answers to the Director of TPRI. Therefore, the registration of pesticides has never been completely divorced from TPRI, as it was originally intended. However, revenues of registration of pesticides used in agriculture are collected by the MAFSC (unlike for registration of pesticides for human use, which is financially beneficial for TPRI). Pesticides have to be tested by a scientist from a public institution. Tests are carried out in field trials of the product for three seasons. The scientist submits a written report to the Registrar. The field report and the registration dossier have to be approved by the Pesticides Approval and Registration Technical Subcommittee (PARTS) and the National Plant Protection Advisory Committee (NPPAC) in order to register a pesticide. Biological control agents have to be registered with the Biological Control Agents Registration and Control (BCARC). BCARC is part of MAFSC with its seat in Dar es Salaam. However, the main office is with Kibaha Research Institute. The Biological Control Agents Sub-committee (BCAS) is responsible for approval of the registration dossiers. Until date, the Tanzania Official Seed Certification Agency (TOSCA) is responsible for the regulation of the agricultural seed sector, testing of imported seeds, quality control of seeds for national use as well as for export and the inspection of fields under seed cultivation. The Seed Act of 1973 has been replaced with the Seed Act of 2003, which harmonized legislation within EAC. A regulation is currently been drafted. June 2006 Draft 2 The Protection of the New Plant Varieties (Plant Breeders' Rights) Act was enacted in 2002. Breeders who wish to protect their improved varieties can apply for Plant Breeders' Rights (PBRs), which are granted for 20 years if the variety passed novelty, distinctiveness, uniformity and stability tests. The Plant Breeders' Rights Registry is part to the MAFSC. The Registry can buy tests either from the registration authority of another country, or from the Tanzania Official Seed Certification Institute (TOSCI), which carries out tests. III. Barriers to introduction of technologies Overall, there are no import restrictions on most inputs (for example machinery, seeds, agro-chemicals) for the agricultural sector as a whole. Consequently, there are no tariffs or quotas for imported seeds and pesticides for use in horticulture. In both cases, only a cess fee of 1.2% of the invoice value has to be paid to acquire an importation permit. However, there are issues that constrain the access to improved seed and pesticide technologies. (1) Pest Control Technologies Chemical pesticides (insecticides, fungicides, herbicides) Pesticides are of high importance in horticulture, in particular due to strict international quality standards for exports (such as EUREPGAP). In general, improved pesticides are not only more effective, but also less dangerous with regards to health and environmental hazards. In addition, there are more and more biological pesticides on the market. All pesticides to be imported and used in Tanzania must be registered by the Registrar of Pesticides at TPRI. The structure of the Office of the Registrar is unclear and not transparent to other stakeholders, since the registrar is appointed and employed by MAFSC, but based at TPRI in Arusha, and for all practical purposes, answers to the Director of TPRI. Revenues from registration of pesticides used for agriculture go to MAFSC; however the Registrar's budget is allocated by the Treasury through TPRI. This structure complicates and slows processes for the Offices of the Registrar. However, TPRI cannot influence the registration of pesticides for use in agriculture. In total, registration of pesticides costs between US$ 4,050 to 7,550, depending on field test fees ($ 2,000 ­ 5,000) and on the registration category. Pesticides should be tested in field trails for three seasons. The Committees that need to approve registration usually interpret a season as one year, even though for many horticultural commodities (e.g. tomatoes) several growing season can occur in one year. Hence, registration could be done faster in many cases. Furthermore, the Committees [who sits there? Potential conflict of interest?] only meet once a year. The size of Tanzania's horticulture is, in particular in comparison to Kenya, small. Typically in horticulture, farmers use a large variety of pesticides, but only a small quantity of each product. Kenya's horticultural sector is already very developed and sizeable; hence quantities of imports are larger. Therefore, the cost of registration is relatively lower in Kenya for importers of agro-chemicals. As a consequence, many June 2006 Draft 3 pesticides used in Tanzania's horticulture sector are smuggled into the country. The situation has improved a lot in the past decade, however, there are still a number of products in use that are not yet registered in Tanzania. In 2005, the list of registered pesticides grew by a total of 392 pesticides in Tanzania, compared to 256 in 1991.1 An additional problem is that already registered pesticides are often limited to certain crops although they could be used for other crops too. While the big producers in the horticultural sector buy pesticides from importers, small- scale farmers rely on retail shops and so-called stockists. In December 2005, there were 725 pesticides wholesale and retail shops registered in the official pesticides business register, along with 34 pesticides importers. A recent survey by the MAFSC found that most pesticide dealers and handlers sell and handle unregistered, expired, repacked, unlabelled or poorly labeled pesticides. 100% of the inspected pesticides dealers showed `somewhat serious abnormalities' in handling pesticides, with 79% showing `very serious abnormalities'.2 Most stakeholders identified two main problems: First, in Tanzania compared to Kenya, there is still a backlog in registered pesticides urgently needed in the horticultural sector. Secondly, the regulation as well as the interpretation of the law and current practices, make the registration of new products slow and costly. Biological control agents In addition to the cited problems with the access to improved agro-chemicals, there are numerous reports that it is even more cumbersome and time-consuming to register biologicals. All biological control agents considered for use in Tanzania have to be registered. Importation of biological control agents is only allowed if (i) they obtain an import permit and are imported in accordance with the permit, (ii) they are imported through points of entry for high risk material and subjected to quarantine procedures, and (iii) the collections in authoritative identified voucher specimens of the biological control agents and targeted pests are made available. Import permits can only be obtained if the biological is registered with BCARC. In addition to the conditions contained in the import permit, the inspector in charge may add further terms and conditions. Importers need to ensure that persons involved in the distribution are adequately trained, as well as to make information regarding safety and environmental impact publicly available. The current practices of registration of biological control agents thus are time-consuming and cumbersome. To tackle these issues concerning pest control technologies, there are initiatives taking place. DfID funds the project `Harmonisation of African Phytosanitary Legislation (HAPL)' led by the Natural Resources Institute, University of Greenwich. The aim is to facilitate harmonization with international laws (WTO and FAO rules). As part of HAPL, the Tanzanian pesticide control legislation is reviewed. The Government of Netherlands has taken the initiative for a Partnership for Market Access with the objective to improve access of horticultural products to the markets of 1Akhabuhaya, Jonathan (2005): "Needs for pesticide safety outreach programmes in developing countries: a Tanzanian example" in: African Newsletter for Occupational Health and Safety, Vol. 15, no. 3, December 2005. 2 Ibid. In total, 52 pesticides shops in the Southern Highlands Region and the Central Regions were inspected. June 2006 Draft 4 Europe and other industrialized countries by enhancing cooperation between public and private stakeholders. The Netherlands provide funding for TAHA and facilitated two Stakeholders' Consultative Workshops to identify problems and propose projects. In March 2006 writeshops for project concepts have been held. One project is to develop a market oriented strategy for pesticides in Tanzania. (2) Seeds The issues at stake in accessing improved varieties of seeds are different to those in pesticides. Registration of varieties of horticultural seeds is voluntarily. There are also no serious impediments to importing new varieties. However, there is room for improvements. Often, shipments of seeds are held for one week, sometimes longer at the port, for the purpose of inspections. Furthermore, if the invoice value of the shipment exceeds US$ 5,000 a Cotecna inspection is required, which consumes a minimum of one month in time. Therefore, many importers import in many, small shipments, rather than few, larger shipments. In addition to the cess fee of 1.2% of the invoice value, the cost of importation of seeds includes TSh. 5,500 for a plant importation permit (PIP) and TSh. 12,000 for IDF. The majority of seeds used by commercial vegetable farms is imported. Bringing in or developing in country improved varieties is however not enough to ensure innovation. Transfer and dissemination of knowledge for using these varieties is also key. In Tanzania, only big horticultural farms buy seeds directly from importers. So-called stockists provide seeds for smallholder farmers. Very often, stockists un- and repack seeds, and/or store seeds inappropriately, which can severely impact the quality. Inspections to ensure quality are rare. Moreover, in most cases, smallholder farmers are unaware of improved varieties and the benefits of these. Most smallholder farmer use farmer-saved seeds or cheaper traditional varieties. The quality of farmer-saved seeds is in many cases deteriorated (low disease tolerance). Hybrid seeds are considerably more expensive, but have much higher yields, which would increase the farmers' income. Another issue in access to technology in Tanzanian horticulture industry is the protection of intellectual property rights. Recently, in 2002, Tanzania adopted the Plant Breeders' Rights Act, which enables application for plant breeders' rights (PBRs) with the registrar in the MAFSC. To date, there have been no applications for PBRs for horticultural varieties. There is a lack of awareness of the possibility to protect improved varieties in Tanzania by the private sector both locally and internationally. There is also a strong feeling that the enforcement of PBRs is going to be weak. Many producers feel that their rights are not sufficiently protected, which has negative impacts on investment and innovation. For this reason, the private sector (TAHA) requests the Tanzanian government to join UPOV (International Union for the Protection of New Varieties of Plants), to create an investment incentive to attract further foreign companies to engage in Tanzanian horticulture. The potential additional benefits of a UPOV membership have not yet been assessed, but might not be very high given that the Tanzanian Plant Breeders' Rights Act is based on UPOV guidelines. The cost of acquiring PBRs is not very high and harmonized within EAC: US$ 200 application fee plus US$ 600 for technical evaluation. In addition, there is a fee of US$ 200 payable every year for the durance of PBRs (20 years). June 2006 Draft 5 IV. Preliminary recommendations Pesticides: Streamline legislation After termination of the review by NRI, the legislation should be streamlined according to recommendations in order to close loopholes and make the process of registration more transparent. There need to be clear institutional arrangements regarding TPRI and the registrar. Harmonization within EAC Working groups on pesticides exist in EAC. Reciprocal recognition of registration within EAC should be the outcome of this process. Recognize Kenyan registration Provided that reciprocal recognition among EAC countries will take time, Tanzania should automatically register pesticides that are already registered in Kenya. Quality control of and training for stockists Stockists should be frequently inspected to ensure quality of pesticides. Training needs to be provided at the same time. Seeds: Quality control of stockists Stockists should be frequently inspected to ensure quality of seeds, as well as pesticides. This would also help to undermine the circulation of counterfeit products. Training Stockists and smallholder farmers should receive training to increase awareness of the benefits of hybrid seeds and the use of improved varieties. Public-private partnerships are an option, given the interest seed importers have in increasing sales of their products. Dissemination of information of PBRs Given the domestic as well as international lack of awareness that Tanzania now has PBRs in place, efforts to inform stakeholders are essential. Efforts should be increased considerably, in particular on an international level in order to attract foreign direct investment and to stimulate innovation. June 2006 Draft 6 List of persons consulted Name Organisation/Company Position Location Jacob Wiersma Royal Netherlands Embassy Second Secretary Dar es Salaam Aart Dekkers Royal Netherlands Embassy Policy Advisor Dar es Salaam Audaz Ministry of Agriculture, Food Head of Legal Dar es Salaam Rutabanzibwa Security and Cooperatives Unit (MAFSC) Geofrey Kirenga MAFSC Assistant Dar es Salaam Director, Crop Development Patrick S. N. MAFSC Registrar, Plant Dar es Salaam Ngwediagi Breeders Rights Mr. Lumbandia MAFSC Head of Seed Dar es Salaam Unit Jonathan Tropical Pesticides Research Registrar, Arusha Akhabuhaya Institute (TPRI) Pesticides Silvest N. Samali HORTI Tengeru (Horticulture Director Tengeru, Research Institute) Arusha Bob Shuma Tanzania Seed Trade Executive Arusha Association (TASTA) Director Colman Ngalo Tanzania Horticultural Executive Arusha Association (TAHA) Director Tjerk Scheltema Arusha Cuttings (TAHA) General Manager Arusha Mr. Jerome Kiliflora (TAHA) Arusha Heikki Niskala Mount Meru Flowers Project Manager Arusha Adrian Moss TRIACHEM Arusha List of persons consulted by telephone Name Organisation/Company Position Mike Chamber Gomba Estate Jan Beukema Enza Zaden Manager June 2006 Draft 7 FOSTERING INNOVATION, PRODUCTIVITY, AND TECHNOLOGICAL CHANGE TANZANIA IN THE KNOWLEDGE ECONOMY Anuja Utz* Abstract Sustained growth that capitalizes on Tanzania's recent macroeconomic stability and structural re- forms will increasingly depend on the economy's capacity for innovation--that is, the capacity to produce of a wider array of goods and services, accelerate the pace of technological change, and integrate with the global economy. Enhancing this capacity will require investment in human re- source development, strengthening of the innovation environment, and strengthening of Tanzania's information and communication technology (ICT) infrastructure. This paper argues that the quality of education, particularly post-primary education, is a crucial element of the capacity to innovate. The paper clarifies that innovation in the Tanzanian context refers to products that previously have not been produced successfully in Tanzania and the adoption of technologies and processes that are new to the country. Innovation is the path to economic diversification and moving up the value chain. Finally, while Tanzania has been able to benefit from the ICT revolution, more work is needed to review and modernize telecommunication policies and regulations to generate fair com- petition and reduce high communication and operational costs. Improvement in these areas could have important payoffs for Tanzania. World Bank Institute Washington, D.C. *The author is the Senior Operations Officer in the Knowledge for Development Program at the World Bank Institute. Copyright © 2006 The International Bank for Reconstruction and Development/The World Bank 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. The World Bank enjoys copyright under protocol 2 of the Universal Copyright Convention. This material may nonetheless be copied for research, educational, or scholarly purposes only in the member countries of The World Bank. Material in this series is subject to revision. The findings, interpretations, and conclusions expressed in this document are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or the members of its Board of Executive Directors or the countries they represent. Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy Anuja Utz 2006. 64 pages. Stock No. 37258 Contents Introduction ..................................................................................................................................... 1 1. Recent Economic Developments in Tanzania ........................................................................ 2 2. What is the Knowledge Economy? ........................................................................................ 7 3. Benchmarking Tanzania's Overall Knowledge Readiness .................................................... 9 4. Tanzania's Performance on the Four Pillars of the Knowledge Economy .......................... 16 5. Conclusions .............................................................................................42 Annex 1: Tanzania's Knowledge Economy Scorecards ............................................................... 47 Annex 2: Costs of Doing Business in Tanzania ............................................................................ 48 Annex 3: International Assessments ............................................................................................. 50 References and Bibliography ........................................................................................................ 53 iii List of Boxes, Figures, and Tables Box 1: Tourism in Tanzania ......................................................................................................................... 3 Box 2: Gold in Tanzania............................................................................................................................... 4 Box 3: Strengthening Supply Chains for Fish in Tanzania........................................................................... 4 Box 4: The Cut-Flower Industry in Tanzania............................................................................................... 6 Box 5: Growing Importance of Knowledge--Global Trends....................................................................... 8 Box 6: National Public Attitude Surveys on Democracy, Markets, and Civil Society in Africa ............... 18 Box 7: Human Capital Flight from Africa.................................................................................................. 28 Box 8: Mobile Phones in Tanzania............................................................................................................. 35 Box 9: SME Development in Tanzania--Key Initiatives........................................................................... 39 Box 10: The Mauritian Miracle--Can Others Emulate It?......................................................................... 41 Box 11: Strengthening Policy Making in African Countries: The Role of Policy Institutes and Think Tanks ....................................................................................................................................... 43 Figure 1: Knowledge Makes the Difference between Poverty and Wealth.................................................. 7 Figure 2: Global Map--Knowledge Economy Index................................................................................. 13 Figure 3: Cross-Country Comparison, 1995 and Most Recent.................................................................. 14 Figure 4: Tanzania's Performance, 1995 and Most Recent....................................................................... 14 Figure 5: Governance Comparisons ........................................................................................................... 17 Figure 6: ICRG Risk Rating, 1984­2003 ................................................................................................... 20 Figure 7: Adult Literacy Rates, 1960­2000................................................................................................ 22 Figure 8: Average Years of Schooling, 1970­2002.................................................................................... 23 Figure 9: Reading Scores ................................................................................................................ 25 Figure 10: Mathematics Scores................................................................................................................... 25 Figure 11: Marginal Social Returns Per Year of Education Based on the Integrated Labor Force Survey............................................................................................................................ 26 Figure 12. Predicted Earnings in the Manufacturing Sector Based on Firm Surveys................................ 27 Figure 13: Projected Shortfall of Health Care Workers in Tanzania.......................................................... 30 Figure 14: ICT Infrastructure: Telephones, Personal Computers and Internet Hosts................................. 32 Figure A.1: Tanzania's Knowledge Economy Scorecards ......................................................................... 47 iv Acronyms and Abbreviations AERC African Economic Research Consortium AIST African Institute for Science and Technology COSTECH Tanzania Commission for Science and Technology CDTT Centre for the Development and Transfer of Technology EIU Economist Intelligence Unit EPZs export processing zones UNESCO United Nations Educational, Scientific and Cultural Organization FDI foreign direct investment GER gross enrollment ratio IPI Institute of Production and Innovations K4D Knowledge for Development KE knowledge economy KEI Knowledge Economy Index MSI Dutch-German consortium NER net enrollment ratio OECD Organisation for Economic Co-operation and Development PEDP Primary Education Development Program R&D research and development RPED Regional Program on Enterprise Development SSA sub-Saharan Africa TARP Tanzania Agricultural Research Project TCRA Tanzania Communications Regulatory Authority TTCL Tanzanian Telecommunications Company TVET technical/vocational education and training UNECA United Nations Economic Commission for Africa v Acknowledgments This report was developed as an input to the Tanzania Country Economic Memorandum (CEM), cur- rently under preparation by the Africa Region of the World Bank. It has been reviewed by colleagues at the World Bank and IFC, including Mavis Ampah, James Keith Hinchliffe, and Robert Utz. Jean- Eric Aubert, Derek Chen, and Yevgeny Kuznetsov of the Knowledge for Development (K4D) pro- gram provided valuable inputs for this report. The views expressed in this paper do not represent those of the World Bank. vii Introduction The application of knowledge, as manifested in areas such as entrepreneurship and innovation, re- search and development and in people's education and skills levels is now recognized to be one of the key sources of growth and competitiveness in the global economy. But many developing countries are failing to tap the vast and growing stock of global knowledge because of their limited awareness, poor economic incentive regimes, and weak institutions. By building on their strengths and by carefully planning appropriate investments in human capital, effective institutions, relevant technologies, and innovative and competitive enterprises, developing countries can benefit from the knowledge revolu- tion and make the transition to a knowledge economy. The knowledge economy (KE) is often taken to mean cutting-edge scientific discoveries, high-tech enterprises such as semiconductor fabrication, or information and communications technology (ICT). However, the term can describe more broadly the creation of any new knowledge and the use of exist- ing knowledge to do things better. Developing countries can benefit by tapping and transferring knowledge from other countries and by applying that knowledge to suit local needs. But lessons from developed countries need to be properly contextualized to the situation of developing countries. This paper examines how Tanzania can use knowledge and technological advances to foster innova- tion productivity, and competitiveness in all sectors of the economy. The paper highlights Tanzania's position on four pillars of the knowledge economy that are considered to be critical for countries to make more effective use of knowledge for their overall economic and social development. These pil- lars are as follows: · An economic and institutional regime that provides incentives for the efficient use of existing knowledge, creation of new knowledge, and the flourishing of entrepreneurship · An educated and skilled population that can create, share, and use knowledge well · A dynamic information infrastructure that can facilitate the effective communication, dis- semination, and processing of information · An efficient innovation system of firms, science/research centers, universities, think tanks, consultants and other organizations that can tap into the growing stock of global knowledge, assimilate and adapt it to local needs, and create new knowledge A country's economic and institutional regime describes the framework within which society and economy work--in other words, the "rules of the game," both formal and informal. And the basic en- abler of any country's transition to the KE is an education system that encourages learning and the exploration of new knowledge. The information infrastructure and ICT provide mechanisms to ex- change knowledge and to reduce knowledge transaction costs. The innovation system drives techno- logical change. Effective use of knowledge in any country requires appropriate policies, institutions, investments and coordination across the above four pillars, though not necessarily all at the same time. Organization of the Paper This paper is organized as follows: Section 1 provides a brief background discussion of recent eco- nomic developments in Tanzania. Section 2 looks at the concept of the KE and highlights some global trends. Section 3 benchmarks Tanzania's knowledge readiness. Section 4 looks at Tanzania's perform- ance on the four critical pillars--economic and institutional regime, education, innovation, and ICT. Section 5 presents some conclusions. 1 2 Anuja Utz 1. Recent Economic Developments in Tanzania Since the mid-1990s, political stability and sound fiscal and monetary policies have allowed Tanzania to make substantial progress in macroeconomic stabilization and structural reforms. Macroeconomic stability in turn has laid the foundation for economic growth and a wide range of structural reforms, including privatization of state-owned enterprises, liberalization of the agriculture sector, efforts to improve the business environment, and strengthening of public expenditure management. As a result, Tanzania's GDP has grown at a rate of more than 5 percent over the last five years, and reached 5.7 percent in 2003 and 6.7 percent in 2004. Table 1: Sectoral Contribution to GDP (in percent) Sectors 2003 2004 Agriculture 46.7 46.4 Industry 18.4 19.1 Mining and quarrying 3.0 3.2 Manufacturing 8.6 8.8 Electricity and water 1.6 1.6 Construction 5.2 5.5 Services 39.3 39.2 Trade, hotels, and restaurants 16.8 17.0 Transport and communications 5.4 5.4 Financial and business services 9.9 9.7 Public administration and other services 7.2 7.1 Minus financial services (indirectly measured) ­4.6 ­4.6 Total GDP 100 100 Source: United Republic of Tanzania (2005). Sectoral characteristics of recent economic trends in Tanzania are shown in Table 1 and discussed be- low. · Agriculture accounted for 46.4 percent of GDP in 2004 and average growth in the agriculture sector has been about 5 percent in recent years. Enhancing agricultural sector performance is critical for growth and poverty reduction because it is the largest sector of the Tanzanian economy. It comprises about 70 percent of employment, mostly in subsistence farming and smallholder cash cropping, and a large share of foreign exchange earnings. · The industrial sector accounted for 19.1 percent of GDP in 2004. Rapid growth was because of strong performances in manufacturing (see next bullet point), mining and quarrying, and construction. Mining has been among the fastest growing subsectors of the Tanzanian economy in recent years. Mining and quarrying output grew at 15.6 percent in 2004, accounting for 3.2 percent of GDP. Growth rate in construction was 11 percent in both 2003 and 2004, and the contribution of this sector to GDP was 5.5 percent in 2004 as compared to 5.2 percent in 2003.Electricity and water grew at 4.7 percent in 2004, contributing to 1.6 percent of GDP. · The performance of the manufacturing subsector has also been steady. Growth in manufactur- ing was 8.6 percent for both 2003 and 2004. The share of manufacturing in GDP increased from 8.6 percent in 2003 to 8.8 percent in 2004. According to the Organisation for Economic Co-operation and Development (OECD), in 2002 this subsector was the top recipient of for- eign and local investment, and 103 manufacturing projects were approved by the Tanzania In- vestment Centre (TIC), compared with 82 in 2001. Of the approved projects, 35 were by local Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 3 investors, 42 were by foreign investors, and 26 were joint ventures (OECD Development Cen- tre 2004). · In the service sector, the trade, hotels, and restaurants subsector grew quickly from 6.5 percent in 2003 to 8 percent in 2004, mainly because of the strong growth in tourism (Box 1). The contribution of this subsector to GDP increased from 16.8 percent in 2003 to 17 percent in 2004. Growth in the transportation and communications subsector increased from 5 percent in 2003 to 6.2 percent in 2004 because of increased investment in mobile telecommunications, the construction and rehabilitation of airports and roads, and increased investment in transpor- tation agencies. The contribution of transportation and communications to GDP in 2004 re- mained the same as in 2003, at 5.4 percent. Financial and business services grew at around 4.5 percent in 2004 (the same as 2003), contributing to 9.7 percent of GDP. The public admini- stration and other services subsector also grew at around 4.5 percent in 2004, contributing 7.1 percent to GDP. Box 1: Tourism in Tanzania According to the Economist Intelligence Unit (EIU), the development of Tanzania as a multicenter tourism des- tination has considerable potential. A growing number of international travel companies are offering tours com- bining safaris in the country's national parks and game and forest reserves with trips to the "Spice Island" of Zanzibar. According to the Directorate of Tourism, the number of tourists rose from 201,744 in 1992 (earning US$120 million in foreign exchange) to a peak of 628,188 in 1999 (earning US$733 million). Numbers fell for a few years after that because of concerns about terrorism in East Africa. However, after some time with no new major incidents, tourist figures recovered from the low of 501,688 in 2000 to 576,000 in 2003, earning a total of US$731 million. There has been considerable investment in the tourism industry in recent years, partly helped by the govern- ment's privatization policy. According to the Tourism Confederation of Tanzania, US$400 million has been in- vested in 160 projects since 1990, of which 130 were new ventures. This has increased the quality and number of hotels and rooms available. Nevertheless, although prospects for increasing the number of visitors are good, there is a risk that in the short term the tourism industry will run up against capacity constraints, which, coupled with Tanzania's poor infrastructure, could hinder the rapid development of the industry. Source: EIU (2004b). A significant change appears to be taking root in the composition of Tanzania's exports. The share of traditional goods in total exports has been declining, from about 40 percent in 1997 to 22 percent in 2002. The decline reflects both the sharp fall in agricultural commodity prices in world markets and stagnant volumes of exports. But some interesting pockets of vitality also are emerging in the Tanza- nian economy, such as gold, fish, and flowers. · The value of merchandise exports rose in 2004 to about $1.335 billion from $1.129 billion in 2003 (and increase of 18.2 percent), mainly because of Tanzania's nontraditional exports, es- pecially gold (Box 2). Overall, traditional exports (coffee, cotton, sisal, tea, tobacco, cashews, and cloves) contributed 21.9 percent and nontraditional exports (mainly gold) contributed 78.1 percent of total merchandise exports. 4 Anuja Utz Box 2: Gold in Tanzania Gold production has expanded in Tanzania because of the increased participation of foreign and local investors with substantial capital and modern technology. Commercial gold production in Tanzania is dominated by five mines: the Bulyanhulu and Pangea mines, which are owned by a combination of Barrick Gold (Canada) and the Khama mining company (in which Barrick is a major shareholder); Geita gold mine, owned by a Ghana- ian/South Africa company, AngloGoldAshanti; Golden Pride Mine, owned by Australia's Resolute Mining Ltd.; and the North Mara mine, owned by Placer Dome (Canada). Together, these mines have boosted output from only 4,890 kilograms of gold in 1998 to 45,299 kilograms in 2003. This makes Tanzania the fourth-largest gold producer in Sub-Saharan Africa, just behind Mali (which produced 52,000 kilograms in 2003) and well behind Ghana and South Africa. Reform of the mineral law has also boosted investment in the gold mining sector. Source: EIU (2004b) · Another success story has been the sharp increase in exports of fish and fish products in recent years. In 2004, however, the export value of fish and fish products decreased by 8.8 percent to $124.2 million from $136.2 million in 2003, because of unfavorable weather changes, among other factors. Thus, there is concern that both the gold and fishing industries are reach- ing the limits of expansion of natural resource extraction and therefore of future growth, coupled with concern about the industries' environmental impact. A recent study has examined the supply chains for fish in Tanzania and highlighted several ways in which this sector could be strengthened (Box 3). Box 3: Strengthening Supply Chains for Fish in Tanzania Two different supply chain systems for fish coexist in Tanzania, both of which support lake- and ocean-to-market commercial operations. The two chains serve two distinct markets: one international and one domestic. The interna- tional export-oriented supply chain system is sophisticated, entails low transaction costs, is well invested, and is quite close to best international practice. The domestic supply chain is rudimentary in technique and technology, is poorly organized, entails high transaction costs, and incurs substantial risk for fishermen and boat operators. The amount of investment in the domestic supply chain, moreover, is very small compared to the level of investment in the parallel export chain. The same primary products move through both channels. However, the way in which they are proc- essed and marketed is quite different. A substantial portion of Tanzania's fish move in a fresh-product form, some move to market in a fast-frozen form, and still others are processed and distributed in a salted, smoked, or cooked form. Distinct distribution channels serve these two markets, and the channels vary in sophistication from "near World Class" to rudimentary, and from "best in class" food safety practices to "hit or miss" practices that are neither controlled nor regulated. Moreover, the service industries that have developed on the periphery of these two supply chains have also developed along separate tracks. Little crossover service provision and even less common use of assets (such as in cold chain warehousing, information systems, and so on) takes place beyond the point at which primary fish products are landed, sorted, and sold. In the supply chain that supports the domestic market, transactions are executed directly between fishermen and a relatively small number of retailers, wholesalers, buyer's agents, and even consumers who buy at fish-landing sites and pay on a cash-and-carry basis. Intermediaries in the domestic market are highly specialized by type of fish and by end market. They typically resell fish in specific markets where they believe that their superior knowledge and access to local commercial networks can realize above-market returns. In many parts of Tanzania, fish marketing is limited to the local village in which fishermen are domiciled or to neighboring communities within walking distance. Little trading takes place in support of long-distance commerce. As a result no integrated market exists in Tanzania for fish. Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 5 The supply chain that supports export fish processing and international distribution is the most developed. The pri- vate sector companies that have invested in this chain are mostly locally based and operate not only in Tanzania but also in Uganda and sometimes in Kenya. Most Nile perch continues to be purchased by the major fish processors, as are most of the prawns, and because of sharp increases in demand for both fishes in overseas export markets, prices have risen slightly and demand continues to exceed supply. Thus, this supply chain is more intensely managed. Ex- port processors purchase their inputs through specialized agents who typically provide fuel, equipment, and credit to fishermen who are affiliated with them, who are trained, and whose fishing processes are certified. The export chains are integrated by the large export-processing companies. Inventory moves transparently through these chains. Prices are set for extended periods and are based on published guidelines that the industrial processors set collectively and that normally entail a significant premium above domestic market prices. Quality control standards are rigorously enforced; as a result a substantial volume of fish rejected by export processors finds its way back into the domestic market. Recommendations for strengthening the fish supply chain in Tanzania, especially the domestic supply chain, include building stronger linkages within the domestic channels; developing the two major wholesale fish markets of Banda Beach Market in Dar es Salaam and Kirumba Market in Mwanza; and investing in transport, storage, and cold-chain handling. Furthermore, the domestic supply chain could gain new markets by following the product-sourcing re- quirements of the new supermarket chains that have recently arrived in Tanzania; similarly, the domestic chain or could follow the seafood-procurement requirements of fast food outlets. Note that meeting higher product require- ments can also lead to improved food safety standards. Another potential leverage point is the land- ing/market/processing clusters and commercial focal points around fishing communities. The domestic fishing sector would also benefit from such improvements as external markets for specialized products and services, equipment leasing, logistics management, transport, market information, and storage and banking. These improvements can be accomplished by defining service specifications and then outsourcing the desired services to qualified providers, by joint venturing the development of new services with qualified providers, or by demonstrating the commercial viabil- ity of new service launches through feasibility studies or business plans. Finally, new market network services need to be developed to include issues such as third-party inventory management and temperature-controlled storage, cash management, price discovery and data dissemination, insurance, and transport delivery. Source: Adapted from Kopicki (2005). · The cut-flower flower industry in Tanzania has also experienced notable growth since its in- ception in 1987. In particular, the switch to greenhouses has resulted in productivity gains and profitability, as well as the entry of investors into the industry. The Netherlands is the main market for Tanzania's flowers with a take of 90 percent; the remaining 10 percent goes to Germany, Norway, England, and Sweden. Box 4 highlights some factors that have been re- sponsible for the success of the cut-flower industry in Tanzania. 6 Anuja Utz Box 4: The Cut-Flower Industry in Tanzania The success of the cut-flower industry in Tanzania, beginning in the mid-1990s, has been facilitated by a combina- tion of factors that provide lessons for developing other export commodities in the country. Cut-flowers are a nontraditional, high-value commodity with access to a rapidly expanding international market. Initial producers in Tanzania were able to tap international linkages and gain access to overseas sources of technology and capital. A key facilitating factor was that foreign firms with essential know-how were per- mitted to come to the country and provide the crucial ingredients that allowed the industry to take off and become competitive in a world market dominated by producers from developed countries. Growth in interregional linkages involving technology transfers and the physical movement of people fur- ther contributed to the development of cut-flower exports. Water, land, and abundant cheap labor were readily available. In contrast to most other agricultural (traditional) crops in Tanzania, there has been very little government involvement in the floricultural industry; the government has neither given special consideration to the cut- flower industry nor provided intensive services to it. Instead, the cut-flower industry represents an interest- ing case of a private sector enterprise that organizes itself and makes its own arrangements in order to meet international standards. Source: Semboja, Mbelwa, and Bonaventura (2000). Foreign direct investment (FDI) in Tanzania has seen its ups and downs over the last few years. Spurred by investments in mining, FDI surged in the late 1990 to reach highs of $517 million in 1999 (or 6 percent of GDP) and $463 in 2000 (5 percent of GDP). Subsequently, FDI fell to about $327 mil- lion in 2001 (3 percent of GDP) and further declined to $240 million in 2002 (2 percent of GDP).1 The IMF estimates that Tanzania received FDI inflows of $478 million in 2003­04 and an estimated $495 million in 2004­05 (IMF 2005). Donors play an important role in the Tanzanian economy. The country is heavily dependent on donor assistance, which finances more than 40 percent of Tanzania's budget. During the past few years, the government in collaboration with donor community has developed a range of strategy papers and pol- icy initiatives to guide its development agenda and poverty reduction efforts, including the Tanzania Development Vision 2025, National Poverty Eradication Strategy (NPES), Tanzania Assistance Strat- egy (TAS), Poverty Reduction Strategy Paper (PRSP), Public Expenditure Review (PER), and Me- dium-Term Expenditure Framework (MTEF), among others. Sustaining and accelerating economic growth in Tanzania will require greater attention to the acquisi- tion and use of new knowledge to increase productivity in all sectors of the economy. In particular, Tanzania needs to develop strategies to use existing and new knowledge to improve performance in traditional sectors, exploit opportunities for leapfrogging, and develop competitive new sectors. These developments require Tanzania to assess its current global standing and learn from the experiences of other countries. A priority in the battle against poverty will be the creation of a greater range of sus- tainable income generation and employment activities, including more export opportunities. More ef- fective sharing and use of knowledge could contribute significantly to the creation of new economic activities by, for example, increasing farm productivity, identifying of new markets for farm products, creating new enterprises based on traditional craft industries, and diversifying of rural economies. With its rich endowments of natural resources and with government committed to achieving higher levels of growth and development, Tanzania is well placed to benefit from the global KE, as discussed below. 1 Source: World Bank internal database. Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 7 2. What is the Knowledge Economy? Knowledge has always been at the core of any country's development process. Figure 1 illustrates this concept for the cases of Ghana and the Republic of Korea. Nearly 40 years ago the per capita income of Korea and Ghana were the same. However, since that time Korea has increased its per capita in- come by a factor of 8.9 in real terms, mainly because of more effective use of policy and technical knowledge, while Ghana has decreased almost by 0.1. Figure 1: Knowledge Makes the Difference between Poverty and Wealth Thousands of constant 1995 U.S. dollars 14 Repubic of Korea 12 10 Difference attributed to 8 knowledge 6 Difference 4 due to 2 physical Ghana and human 0 capital 1960 1965 1970 1975 1980 1985 1990 1995 2000 Source: World Bank, K4D program. Today, we are in the midst of a knowledge revolution that is being spurred by the twin forces of glob- alization and technological advances. This knowledge revolution manifests itself in many different ways: there are closer links between science and technology; innovation is more important for eco- nomic growth and competitiveness; education and life-long learning are increasingly important; and more investment is being made in intangibles (such as research and development (R&D), software, and education) than in fixed capital. And, of course, there is the ICT explosion that brings with it worldwide interdependency and connectivity. These trends have led to increased globalization and competition: in 1990, trade represented 38 percent of world GDP; in 2001 it represented 57 percent (Box 5). The knowledge revolution is creating a constant state of restructuring at the global, country, sector, and firm levels. While this raises tremendous possibilities for enhancing growth and competitiveness, it also carries risks that countries, firms, organizations, and individuals will not keep up with the rapid changes. Consequently, countries' competitiveness depends more than ever on their ability to access, adapt, and utilize knowledge for development. The term "knowledge economy" (KE) was coined to reflect the increasing importance of knowledge for economic development. Despite the hype that surrounds this concept, the idea of a KE is not new. Knowledge and its use has always been a critical ingredient of economic success. However, recent times have seen its importance increase. In fact, knowledge has become the key driver of economic competitiveness and success, and has added massive value to economic production through increases in productivity. The application of new technologies and new ideas--both in the form of new inven- tions and new applications of existing knowledge--has brought revolutionary change to virtually all markets and sectors. 8 Anuja Utz Box 5: Growing Importance of Knowledge--Global Trends The OECD uses the term, "knowledge economy," (KE) to draw attention to the importance of knowledge in all economic activities. The definition has evolved a focus on manufacturing industries that make intensive use of technology to include services that are also heavily knowledge based. The KE now accounts, on average, for roughly half of nongovernment economic activity in the OECD. Some features of the KE are as follows: · More investment and trade in intangibles. Investment in intangibles has been skyrocketing. In OECD countries, public investment in intangibles (such as education, R&D, and software) has now reached almost the same level as that for machinery and equipment--8.6 percent of GDP compared to 9.0 per- cent. This estimate almost certainly is low because it does not include private investment in education, public and private investment in skills training, or investment in design, marketing, advertising, brand development, engineering, publishing, and the arts. · More emphasis on education and training. In OECD countries the proportion of adults with at least an upper secondary education rose from 61 percent to 66 percent and with tertiary education from 21 per- cent to 24 percent from 1998 to 2003. In high-income OECD countries, the percentage of the labor force with tertiary education increased from roughly 21 to 27 percent from 1989 to 2001. Developed countries have also been improving the skills of their labor force through extensive continuing educa- tion both in universities and in firms. Developing countries lag in this respect as their educational at- tainments are low and their workforces need retraining to be competitive in the knowledge economy. · More FDI. FDI, one of the key agents of globalization, affects areas such as technology, management, access to markets, and access to finance, labor, and natural resources. FDI inflows increased 16 times between 1982 and 2002 and FDI's share of world gross fixed capital formation increased from 2.4 per- cent to 10.3 percent. · More R&D. Of global R&D, 88 percent is undertaken by high-income countries, with 31 percent of global R&D centered in one country, the United States. Multinational companies are now carrying out R&D in foreign countries and are establishing more strategic alliances--even mergers and acquisi- tions--to collaborate on technology and acquire technological assets. For example, R&D expenditures of foreign affiliates worldwide climbed from an estimated US$30 billion in 1993 to US$67 billion in 2002, with more R&D going to developing countries. But at the same time, only a small number of de- veloping countries and economies in transition are internalizing R&D, with developing countries in Asia being the preferred destination. Also on the rise is the number of international collaborations in patenting and technical publications. For example, from 1999 to 2001, 6.7 percent of all patents filed at the European Patent Office were the result of international collaborative research and 15.4 percent were owned or co-owned by a foreign resident. Source: UNCTAD (2005b); OECD (2005a, 2005b); World Bank internal databases. In short, a KE is an economy that creates, disseminates, and uses knowledge to enhance its growth and competitiveness. Contrary to some beliefs, it is not necessarily about high tech or IT. The application of new techniques to subsistence farming can increase yields significantly, or the use of information and logistical services can allow traditional craft sectors to serve much wider markets than before. These are both examples of the KE in action. A successful KE requires a strong and robust economic and institutional regime, a well educated and skilled population, an efficient innovation system, and a dynamic information infrastructure. Section 3 benchmarks Tanzania's overall readiness for the KE, or its "knowledge readiness." This concept should be distinguished from "e-readiness," which is mainly concerned with a country's physical ICT infrastructure and the skills of the population to utilize this infrastructure. Knowledge readiness, on the other hand, has a much wider scope and includes other systems such as education, wide-ranging vocational skills training, and business research and innovation, as well as progress in ICT. It is about the ability of a country to create, access, share, and apply knowledge across a wide range of sectors, whether or not "e-technologies" are used. Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 9 3. Benchmarking Tanzania's Overall Knowledge Readiness This section looks at three sources that benchmark countries' overall receptiveness to knowledge and innovation. The first is the World Economic Forum's Global Competitiveness Report 2005­2006; the second is the World Economic Forum's Africa Competitiveness Report 2004; and the third is the World Bank's Knowledge Assessment Methodology (KAM). In these three sources, Tanzania is com- pared mostly with neighboring countries such as Kenya and Uganda, which share many aspects of economic structure and environment. For example, these three countries have revived the East Africa Community, set up processes to oversee the new community, and established commissions for in- creased cooperation in areas such as joint railways operations and the harmonization of revenue col- lection methods. World Economic Forum's Global Competitiveness Report 2005­2006 The recently released Global Competitiveness Report 2005­2006 by the World Economic Forum pro- vides information for 117 countries. The main aim of the report is to enhance understanding of the key factors that determine economic growth and to explain why some countries are much more successful than others at raising income levels and opportunities and in joining the upper ranks of international competitiveness. The report includes two indices: · The Growth Competitiveness Index (GCI), which aims specifically to gauge the ability of the world's economies to achieve sustained economic growth over the medium to long term. The GCI identifies three "pillars" in the evolution of growth in a country: the quality of the macro- economic environment, the state of the country's public institutions, and the level of its tech- nological readiness. The index uses a combination of hard data, such as budget deficits, the level of Internet access in schools, and survey data, and examines such areas as judicial inde- pendence, the prevalence of institutionalized corruption, and the extent of inefficient govern- ment intervention in the economy. These various pieces are brought together under several "sub-indices," each capturing a different aspect of the growth process and aggregated to give an overall competitiveness "score" for a country. · The Business Competitiveness Index (BCI) emphasizes a range of company-specific factors that are conducive to improved efficiency and productivity at the micro level. The BCI is a complement to the medium-term, macroeconomic approach of the GCI, and evaluates the un- derlying microeconomic conditions defining the current sustainable level of productivity in each of the countries covered. The underlying concept is that whereas macroeconomic and in- stitutional factors are critical for national competitiveness, they are necessary but not sufficient factors for creating wealth (which is created at the micro level by companies). In the 2005­2006 report, Finland is number one in the GCI rankings for the third consecutive year. The country is very well managed at the macroeconomic level and it also scores very high in measures that assess the quality of its public institutions. Furthermore, the private sector shows a high proclivity for adopting new technologies and nurturing a culture of innovation. The United States, like last year, is ranked second: its overall technological supremacy is partly offset by a weaker performance in areas that capture the quality of its public institutions and the stability of the macroeconomic environment. While most of the countries of sub-Saharan African are not very competitive, the region does have a number of relative success stories, including South Africa (42nd), Botswana (48th), and Mauritius (52nd). 10 Anuja Utz The survey shows that Tanzania improved its growth competitiveness and therefore its rank on the GCI from 82nd place last year (out of 104 countries) to the 71st this year (out of 117 countries), while Kenya and Uganda both fell in the rankings, from 78th spot last year to 92nd this year for Kenya, and from 79th spot last year to 87th place this year for Uganda. According to the 2005­2006 report, among low-income countries, Ghana, Tanzania, and Pakistan have made the largest improvements in com- petitiveness. Ghana has benefited especially from improved public schools and less corruption, and Tanzania and Pakistan have both reported better labor-employer relations. The sub-indexes that comprise the 2005­2006 GCI showed the following trends: · Tanzania maintained its ranking from last year in the Macroeconomic Environment Index at the 72nd spot, surpassing Uganda at the 88th spot, and increasing its lead over Kenya which is placed 106th. · Tanzania rose in the Public Institutions Index from 88th position last year to the 60th spot this year. Kenya comes in at 94th place this year and Uganda at 95th. · Tanzania dropped slightly in the Technology Index from 84th spot last year to 86th this year. In the current rankings, Kenya and Uganda claimed the 71st and 82nd spots, respectively. The 2005­2006 global BCI ranking showed the following trends: · Tanzania improved its position from 90th place last year to 82nd this year, while Kenya cap- tured the 68th place (down from 63rd last year) and Uganda fell from the 71st spot last year to 85th this year. Tanzania gained some ground in both sub-indices that comprise the BCI-- company operations and strategy and the quality of the national business environment. World Economic Forum's Africa Competitiveness Report 2004 This report, also developed by the World Economic Forum in 2004, highlights the prospects for growth, obstacles to improving competitiveness, and the need to accelerate the pace of economic change in 25 selected African economies.2 It also uses the GCI, and shows that the following trends: · Tanzania ranks 9th out of 25 countries on the overall GCI index, surpassing Uganda, which is ranked 14th, and Kenya, which is placed 15th. The most competitive economy among the 25 countries is Botswana, followed by Tunisia, South Africa, and Mauritius (Table 2). Table 2: Africa Growth Competitiveness Index 2004 Country Rank Score 1. Botswana 4.56 2. Tunisia 4.49 3. South Africa 4.37 4. Mauritius 4.12 5. Namibia 3.99 6. Gambia, The 3.93 7. Egypt 3.84 8. Morocco 3.77 9. Tanzania 3.49 10. Ghana 3.46 Source: WEF 2004. 2These include: Algeria, Angora, Botswana, Cameroon, Chad, Egypt, Ethiopia, The Gambia, Ghana, Kenya, Madagascar, Malawi, Mali, Mauritius, Morocco, Mozambique, Namibia, Nigeria, Senegal, South Africa, Tanza- nia, Tunisia, Uganda, Zambia, and Zimbabwe. Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 11 The rankings on the three sub-indexes that comprise the GCI reveal the following: · On the Public Institutions Index, Tanzania is ranked 9th again, ahead of Uganda at 18th spot, and Kenya at 21st. Botswana once again tops the list, followed by Tunisia, Malawi, and The Gambia. · On the Macroeconomic Environment Index, Uganda is ranked 12th and Tanzania placed 14th, followed by Kenya at 15th. The top four spots go to Botswana, Tunisia, South Africa, and Mo- rocco. · The Technology Index places Tanzania in the 12th spot among 25 countries, where it lags be- hind Kenya, which is ranked 8th, and Uganda, which is ranked 10th. South Africa claims the top rank in this index, followed by Mauritius, Tunisia, and Botswana. The 2004 report also provides national competitiveness balance sheets for all 25 countries. Tanzania's balance sheet makes note of the following competitive advantages: · In the area of macroeconomic environment, Tanzania's advantages include low recession ex- pectations, public trust of politicians, and low levels of distorting government subsidies and diversion of public funds. · In the case of Tanzania's public institutions, the country has advantages in the areas of judicial independence and lack of favoritism in decisions of government officials. · Tanzania also rates well in terms of technology variables, such as prevalence of foreign tech- nology licensing, FDI and technology transfer, company spending on R&D, univer- sity/industry research collaboration, and firm-level technology absorption. According to the report, the country also has had success in ICT promotion and in government prioritization of ICT. · Other indicators where Tanzania exhibits competitive advantages include wage equality of women in the workplace, government effectiveness in reducing poverty, reliance on profes- sional management, and railroad infrastructure development. On the other hand, the report also refers to several competitive disadvantages: · In the macroeconomic environment, the national savings rate, interest rate spread, real ex- change rate, country credit rating, access to credit, and inflation are cited as competitive dis- advantages. · Public institutions in Tanzania need to address weaknesses in tax collection, public utilities, and exports and imports. Organized crime is a problem and property rights need to be further secured. · In terms of technology, Tanzania ranks low in many ICT-type indicators including telephone lines, cellular phones, personal computers, Internet use, Internet hosts, Internet access in schools, quality of competition in the ISP sector, and laws relating to ICT. Tanzania does not rank highly in terms of technological sophistication or utility patents. In addition, it needs to do more to boost tertiary enrollments. · Other areas where Tanzania exhibits competitive disadvantages include the business costs of diseases such as malaria, HIV/AIDS, or tuberculosis; infrastructure (such as air transport, telephone, and quality of electricity supply); private sector employment of women; and the extent of bureaucratic red tape.3 The national competitiveness balance sheet thus provides some insights about the relative importance of various elements relating to the macroeconomic environment, governance, the state Tanzania's pub- lic institutions, and its level of technological readiness and sophistication. These have a bearing on 3For more information, see the Tanzania country table, WEF (2004). 12 Anuja Utz Tanzania's ability to be competitive and to sustain high growth rates in the long term. Even though the country shows some strengths, the balance sheet shows several areas where policy reforms are vitally needed if the country is to fulfil its potential. Knowledge Assessment Methodology (KAM) The third methodology that is used to look at Tanzania's overall knowledge readiness is the World Bank Institute's Knowledge Assessment Methodology (KAM).4 The KAM dataset includes 80 quanti- tative and qualitative variables and 128 economies that help to benchmark how an economy compares with its neighbors, competitors, or other countries it wishes to emulate on the four pillars of the KE. KAM helps to identify problems and opportunities that a particular country faces in making effective use of knowledge for development and where it may need to focus policy attention or future invest- ments. Using the KAM, the following benchmarking exercise compares Tanzania with the Africa region (comprising 25 countries),5 with Kenya and Uganda, and with leading economies in Africa such as Botswana and South Africa that have made significant strides in harnessing knowledge to improve growth and competitiveness. Tanzania and the World The following figures paint a preliminary picture of Tanzania's overall readiness or preparedness for the knowledge economy. Figure 2 introduces the Knowledge Economy Index (KEI),6 which is the av- erage of the performance scores of a country or region in the four pillars of the knowledge economy (economic and institutional regime, education, innovation, and ICT). The KEI shows that many Afri- can countries are clustered at the bottom third of the distribution on the global knowledge economy map, suggesting that these countries could do much more to harness knowledge for their overall eco- nomic and social development. 4The analysis is based on the 2004 KAM. The latest KAM is available at: http://www.worldbank.org/kam. 5The following 25 countries from sub-Saharan Africa are included in the KAM database: Angola, Benin, Bot- swana, Burkina Faso, Cameroon, Côte d'Ivoire, Eritrea, Ethiopia, Ghana, Kenya, Madagascar, Malawi, Maurita- nia, Mauritius, Mozambique, Namibia, Nigeria, Senegal, Sierra Leone, South Africa, Sudan, Tanzania, Uganda, Zambia, and Zimbabwe. 6For more information, see the KAM (www.worldbank.org/kam), where the following three variables are cho- sen as proxies for each of the four pillars that constitute the Knowledge Economy Index (KEI): · Economic and institutional regime: tariff and non-tariff barriers, regulatory quality, and rule of law · Education and human resources: adult literacy rate (percent, age 15 and above), secondary enrolment, and tertiary enrollment · Innovation system: researchers in R&D, patent applications granted by the United States Patent and Trademark Office, and scientific and technical journal articles (weighted by per million population). · Information infrastructure: telephone per 1,000 persons, computers per 1,000 persons, and Internet us- ers. Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 13 Figure 2: Global Map--Knowledge Economy Index Source: KAM (www.worldbank.org/kam). Tanzania is placed at the lower end of the global KE map (with a KEI of 1.67), but has improved from 1995 when it had a KEI of 1.03.7 It matches the performance of Uganda (with a KEI of 1.67), but lags behind Kenya (with a KEI of 2.34), which has shown some gains in the KE. Tanzania is also behind the African regional average (KEI of 1.91), and has a ways to go to match the performance of other leading countries in Africa such as Botswana (KEI of 4.59) and South Africa (KEI of 4.89). These countries are all placed in the middle of the global KE map, suggesting that they have been successful at creating, adapting, and using knowledge for overall development. Disaggregating Tanzania's Performance It is also possible to see how Tanzania has performed on the four individual pillars that comprise the KEI (economic and institutional regime, education, innovation, and ICT). Figure 3 compares the performance of Tanzania with the African region, with neighbors, as well as with Botswana and South Africa. The two bars represent a country's aggregate KEI score for the most recent year for which data are available and for 1995. The color bands in each bar (see figure legend) disaggregate the KEI score by pillar. Figure 4 shows the KEI and Knowledge Index (KI)8 for Tanza- nia, as well as its performance on all four pillars between 1995 and the most recent period for which data are available. 7Countries above the 45 degree line have improved their recent KEI rank relative to their position in 1995 (or closest available date in the mid-1990s), and vice versa for countries below the line. 8The KI (Knowledge Index) is the simple average of the performance of a region or country in three KE pillars (education, innovation, and ICT). 14 Anuja Utz Figure 3: Cross-Country Comparison, 1995 and Most Recent Period for which Data Are Available Source: KAM (www.worldbank.org/kam). Figure 4: Tanzania's Performance, 1995 and Most Recent Period for which Data Are Available Source: KAM (www.worldbank.org/kam). Note: The upper bar for each variable represents the most recent period for which data are available. The lower bar represents data for 1995. Figures 3 and 4 show that between 1995 and most recent period for which data are available, Tanzania has improved its overall knowledge readiness, which is reflected by better rankings in the KEI and KI. In particular: · Noticeable improvement is evident in Tanzania's economic and institutional regime because of improvements in its regulatory regime and the rule of law. This improvement is borne out by Tanzania's sustained strong performance on many governance indicators (see Section 4). · The country shows progress in innovation because of increasing numbers of patents and arti- cles in scientific and technical journal. · A small positive change is evident in education because of increasing adult literacy rates and tertiary enrollments. · In addition, Tanzania has strengthened its information infrastructure by increasing telephony (fixed and mobile phones), Internet use, and personal computer penetration. Tanzania also has a relatively free press, and there are a range of daily and weekly Kiswahili and English lan- guage newspapers.9 9There are other ways to highlight Tanzania's performance in the KE. Individual KE scorecards are presented in Annex 1, where Tanzania is compared to the world (128 countries) as well as to the African Region (comprising 25 countries). Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 15 But Figures 3 and 4 also show that even though Tanzania has made progress, the change is not as impressive relative to what is happening in other countries. For example, Figure 3 shows that be- tween 1995 and the most recent period for which data are available, Tanzania has made a substan- tial improvement in its overall knowledge readiness as evidenced by positive changes in the KEI for the economic incentive regime, ICT, and especially innovation. Uganda has also made strides in improving its economic incentive regime and innovation, and Kenya has improved both its innova- tive capacity and its information infrastructure over the past decade or so. But overall, Kenya's per- formance surpasses that of the African region, while Tanzania's does not. Figure 3 also shows that Botswana and Malaysia have slightly improved their performance since 1995, while South Africa and Thailand have not. Thus, this relative comparison shows that even though Tanzania has made progress, world-wide progress in the variables that are used to track knowledge and innovation- related performance is much stronger. Section 4 looks at Tanzania's performance on each of the four KE pillars and identifies some chal- lenges and opportunities related to using knowledge for increasing growth and national competitive- ness. 16 Anuja Utz 4. Tanzania's Performance on the Four Pillars of the Knowledge Economy Pillar 1: Economic and Institutional Regime Taking advantage of the potential offered by the knowledge revolution requires a flexible society and economy, able to cope with the need for constant change. Such flexibility requires that countries have an effective regime of economic incentives as well as institutions that can promote and facilitate the redeployment of resources from less efficient to more efficient uses. A well-functioning economic and institutional regime allows organizations, people, and institutions to adjust to changing opportunities and demands in flexible and innovative ways. Without strong economic incentives and an institutional regime that deploy resources productively, it is possible to have a strong educational base and a highly developed R&D infrastructure but still to miss out on the full benefits of these knowledge-related achievements. Key elements of the economic and institutional regime include a financial system that mobilizes and allocates capital to its most productive uses; flexible labor markets including support for up-skilling of the labor force; an appropriate legal and regulatory system and strong rule of law that supports entre- preneurship; effective safety nets to facilitate adjustment to constant restructuring; a legal system that can cope with the many demands emanating from such restructuring and redeployment; and effective, transparent, and accountable government. This section looks at Tanzania's institutional regime from the viewpoints of governance and the costs of doing business. Governance Governance can be broadly defined as the set of traditions and institutions by which authority in a country is exercised. This includes the process by which governments are selected, monitored, and replaced; the capacity of the government to effectively formulate and implement sound policies; and the respect of citizens and the state for the institutions that govern economic and social interactions. In recent years many measures have been developed to capture the governance dimensions of a coun- try. The World Bank Institute has developed scorecards that trace six areas of governance for 209 countries from 1996 to 2004. Figure 5a compares governance indicators in Tanzania from 1998 and 2004; Figure 5b compares 2004 governance indicators for Tanzania with the rest of sub-Saharan Afri- can.10 10The governance indicators presented in Figures 5a and 5b reflect the statistical compilation of responses on the quality of governance. These statistics derive from a large number of enterprise, citizen, and expert survey re- spondents in industrial and developing countries, as reported by a number of survey institutes, think tanks, non- governmental organizations, and international organizations. Countries' relative positions on these indicators are subject to margins of error that are clearly indicated. Consequently, precise country rankings should not be in- ferred from this data. The 2004 confidence range at 90 percent is depicted by the dotted lines. Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 17 Figure 5: Governance Comparisons a. Tanzania: 1998 versus 2004 b. Regional Comparison, 2004 Figure 5a shows that between 1998 and 2004, Tanzania improved its rankings on three of the six vari- ables: voice and accountability and government effectiveness improved by a small margin, while con- trol of corruption showed much more improvement. This is not surprising, because in 1999 Tanzania adopted a comprehensive National Anti-Corruption Strategy and Action Plan. The plan is focused on advancing institutional reforms; increasing transparency and accountability in government business; increasing transparency and accountability in the participation of civil society and private sector; and monitoring progress on areas such as the rule of law, legal framework, financial management, pro- curement, and so on. In order to ensure transparency and accountability in the use of funds, Tanzania also adopted the Integrated Financial Management System (IFMS) to automate all accounting activi- ties. In fiscal 1998­99 ten pilot Ministries, Department, and Executive Agencies (MDAs) were auto- mated; by fiscal 1999­00, all MDAs were on board and payments were being centrally processed (Mwanza 2002). Moving on to the regional governance comparison, Figure 5b for 2004 shows that Tanzania led sub-Saharan Africa on all six governance dimensions. Afrobarometer--an organization that conducts surveys in Africa--has undertaken national sample surveys regarding the attitudes of citizens in 15 selected African countries (including Tanzania) about their national economies, democracy, good governance, markets, civil society, and other aspects of development.11 Through its surveys, Afrobarometer seeks to answer questions related to the economic, political, and cultural shifts in these economies. The 2002­2003 surveys for all of the15 countries on average show that the following: · Economically, the present mood in Africa is somber, but people are optimistic about the fu- ture. · Culturally, Africans value equality but also express an emergent individualism. · Politically, they continue to prefer democracy and reject authoritarian rule. 11See http://www.afrobarometer.org/. The Afrobarometer surveys are repeated on a regular cycle, and because the instrument asks a standard set of questions, countries are systematically compared and trends are tracked over time. Survey results are designed to enable Africans and interested outsiders to educate themselves about public opinion in sub-Saharan Africa and to influence policy makers accordingly. 18 Anuja Utz · In terms of institutional performance, they view the management of the national economy in a moderately positive light. Box 6 provides some highlights from this survey for Botswana, Kenya, Tanzania, and Uganda. Box 6: National Public Attitude Surveys on Democracy, Markets, and Civil Society in Africa The Afrobarometer Round 2 (2002­2003) surveys show the following interesting results for Botswana, Kenya, Tanzania, and Uganda: · Tanzanians (53 percent) and Ugandans (57 percent) are more likely to opt for a "free market economy" than a "government-run economy," as compared to Botswana (24 percent) and Kenya (43 percent). · Most people call for the protection of property rights under a rule of law. Ugandans are most insistent that "the government must abide by the law in acquiring any property, including paying the owner" (95 percent); this result is not surprising given their history. Uganda is followed by Kenya (87 percent) and Tanzania and Botswana at 83 percent. · Kenyans (67 percent) and Tanzanians (57 percent) seem to be satisfied with reform outcomes and "the gov- ernment's reduced role in the economy"--more so than their counterparts in Botswana (43 percent) and Uganda (49 percent). · Support for democracy ran strong in Kenya (80 percent) in 2002, undoubtedly reflecting the euphoria of a peaceful transfer of power following the December 2002 election. People in Botswana and Uganda are also supporting democracy by a large margin (75 percent). The Botswana result is no doubt fueled by the multi- party system, which has been gradually putting down roots over time. The Tanzanian population's support for democracy is at 65 percent. · Africans also embrace free and fair elections. Popular support for open elections is highest in countries that have recently transited to democracy (Kenya, 89 percent). Support for open elections is also high in Uganda (83 percent), a country in which political competition is presently allowed among individual candidates but not between political parties. Botswana and Tanzania both come in at 76 percent. · As for whether "the government can solve ... all or most ... of the country's problems," a more pragmatic view prevails in Tanzania (45 percent) and Uganda (48 percent), compared to Botswana (58 percent) and Kenya (64 percent). · Many countries view the management of the national economy in a positive light, ranging from 83 percent in Kenya to 68 percent in Tanzania to about 60 percent in both Botswana and Uganda. Source: Afrobarometer 2004. Costs of Doing Business in Tanzania Several constraints to private sector investment and the overall investment climate have been identi- fied in Tanzania. Taxation, cost and availability of finance, availability of electricity, and corruption are seen by a majority of surveyed firms as a major or very severe constraint to private sector devel- opment in Tanzania. The World Economic Forum's Africa Competitiveness Report 2004 cites the fol- lowing as being the most problematic factors for doing business in Tanzania (from most problematic to least): inadequate supply of infrastructure, access to financing, tax rates, corruption, tax regulations, inefficient government bureaucracy, inadequately educated workforce, crime and theft, poor work ethic in national labor force, policy instability, foreign currency regulations, inflation, restrictive labor regulations, and government instability/coups.12 These are important entry points for improving Tan- zania's overall business climate and thereby fostering increased economic activity and employment opportunities. The recently released World Bank (2006) report titled Doing Business in 2006: Creating Jobs pro- vides important information on the scope and manner of regulations that enhance or constrain business 12World Bank 2004c; World Economic Forum 2004 (Tanzania Country Profile). Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 19 activity in a large sample of countries, including Tanzania. The report presents indicators on the ease of doing business, such as starting a business, dealing with licenses, hiring and firing workers, regis- tering property, getting credit, protecting investors, paying taxes, trading across borders, enforcing contracts, and closing a business--all critical ingredients of an effective investment climate. The indi- cators are used to analyze economic and social outcomes, such as productivity, investment, informal- ity, corruption, unemployment, and poverty, and identify what reforms have worked, where, and why. In the 2006 report, New Zealand tops the list of economies for ease of doing business. Among the top African countries are Mauritius, which is ranked 23rd, South Africa, which is ranked 28th and Namibia, which is placed 33rd. Annex 2 provides a detailed snapshot of the business climate in Tanzania as compared to neighbors, competitors, and countries it would like to emulate, such as Botswana, Kenya, Uganda, South Africa, Malaysia, and Thailand. The annex reveals several specific regulations and policies that discourage investment, productivity, and growth in Tanzania. Below are some examples: · Tanzanian entrepreneurs can expect to go through 13 steps to launch a business over 35 days on average, at a cost equal to 161.3 percent of gross national income (GNI) per capita. Kenya and Uganda require 13 and 17 steps respectively, over 54 and 36 days, at a cost equal to 48.2 percent and 117.8 percent of GNI respectively. Botswana requires only 11 steps but far more days (108) at a lower cost (10.9 percent of GNI). For the two Asian comparators, Malaysia and Thailand, it requires both less time and money to start a business--30 days and 20.9 per- cent of GNI for Malaysia, and 33 days but just 6.1 percent of GNI for Thailand. · Regulations in Tanzania are rather rigid when it comes to the hiring and firing of workers. Tanzania's overall Rigidity of Employment Index is 69, higher than in Kenya (28), Uganda (13), Botswana (30), and South Africa (52). Malaysia comes in at 10 and Thailand at 18. · In Tanzania, it takes 61 days to register property, less than in Botswana (69 days) and Kenya (73 days) but more than in Uganda (48 days) and South Africa (23 days). In Thailand, it takes just 2 days to register property, whereas in Malaysia it takes much longer--143 days. · Regarding credit, Tanzania has a score of 5 on the Strength of Legal Rights Index13 as does Uganda, Thailand, and South Africa, which shows weaknesses in all these countries. Kenya and Malaysia have a score of 8 and Botswana has a score of 9, which indicates better sharing of credit information and better protection of the legal rights of borrowers and lenders. · The degree to which investors are protected through disclosure of ownership and financial in- formation is measured by the Extent of Disclosure Index. This index14 shows that Tanzania with a score of 3 and Kenya with a score of 4 need much improvement in this area. Uganda has a score of 7, followed by Botswana South Africa, both with a score of 8. Malaysia and Thailand get perfect scores of 10. · It takes on average 248 hours to pay taxes in Tanzania, as compared to 52 hours in Thailand, 140 in Botswana, 237 in Uganda and 372 in Kenya. · Enforcing contracts is a major problem in most countries surveyed. It takes on average 240 days to resolve a payment dispute in Tanzania, compared to 209 days in Uganda and almost a year (360 days) in Kenya. In Botswana and South Africa too, resolving a payment dispute takes a long time--154 days on average for Botswana and 277 days for South Africa. The comparator Asian countries take even longer--300 days in Malaysia and well over a year (390 days) in Thailand. 13Higher scores indicate that laws are better designed to expand access to credit. 14This index ranges from 0 to 10, with higher values indicating more disclosure. 20 Anuja Utz · While it takes three years to go through bankruptcy proceedings in Tanzania, it takes 2.7 years in Thailand, 2.2 years in Botswana, Malaysia, and Uganda, 2 years in South Africa, and a high 4.5 years in Kenya. The Composite International Country Risk Guide (ICRG)15 is another broad indicator that gives some idea of how investors view the investment climate in a particular country. The ICRG ratings for Bot- swana, Kenya, Tanzania, and Uganda are presented in Figure 6. It shows that Tanzania's risk rating has been declining since 1997, which is paradoxical given that the decline coincides with a growth period in Tanzania's economy. The reasons for the fall in Tanzania's risk rating should be further in- vestigated. Figure 6: ICRG Risk Rating, 1984­2003 90 80 70 gn 60 rati 50 40 CRGI 30 20 10 0 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Year Uganda Tanzania Kenya Botswana Source: World Bank internal database. The indicators discussed above reveal some strengths and weaknesses of the business environments of Tanzania and other countries. The weaknesses need to be addressed in order to improve the countries' competitiveness in an increasingly integrated and knowledge-based global economy. To its credit, in recent years the government of Tanzania has made significant progress in enhancing the business envi- ronment, especially through the implementation of a program called Business Environment Strength- ening in Tanzania (BEST). Another positive government action was the establishment of the Tanzania Investment Centre (TIC) in 1997 as a one-stop center for promoting investment. The Banking and Fi- nancial Institutions Act was amended in 2003 to empower the Bank of Tanzania to regulate and super- vise activities of all savings and credit associations and to facilitate the provision of long-term finance to the productive sector, for example, by allowing housing finance companies. The Tanzania Bankers Association has also improved access to credit through the establishment of a credit bureau. The Land Act of 1999 has been amended to accelerate land surveys and to modernize the land registry to enable the commercialization of land leases and to facilitate their use as collateral for bank loans. The Income Tax Act of 2004 is also considered more conducive to private sector growth than previous income tax laws. Further initiatives are underway to formalize personal property rights.16 These and other policy measures and initiatives will help improve Tanzania's overall investment climate and make it more conducive to promoting investment in all sectors of the economy. 15The ICRG risk rating is an overall index, ranging from 0 to 100 (highest risk to lowest), based on 22 compo- nents of risk grouped into three categories (political, financial, and economic). 16OECD (2004). For more information, go to: www.oecd.org/dev/africanoutlook. Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 21 Pillar 2: Education and Human Resources Education is the fundamental enabler of the KE. Well-educated and skilled people are key for creating, sharing, disseminating, and using knowledge effectively. Basic education provides the foundation for lifelong learning and increases peoples' capacity to assimilate and use information. Secondary and tertiary education can develop core skills, including technical skills, which encourage the mind of creative and critical thinking needed for problem-solving and innovation. Higher education in engi- neering and scientific areas is needed to monitor technological trends, assess what is relevant for a firm or the economy, and use new technologies. The production of new knowledge and its adaptation to a particular economic setting is generally associated with higher-level teaching and research. Creating a culture of continuous learning and openness to new ideas is critical for a knowledge-based economy. A lifelong learning system is a system which encompasses learning throughout the life cycle (from early childhood to retirement) and includes formal training (schools, training institutions, uni- versities); nonformal learning (on-the-job and household training), and informal learning (skills learned from family members or people in the community). The basic elements of a system of lifelong learning are comprehensiveness, the ability to acquire new basic skills (acting autonomously, using tools interactively, and functioning in socially heterogeneous groups), multiple pathways to learning and problem solving, and multiple providers of education. Developing countries such as Tanzania face many challenges in developing systems of education. These challenges include providing universal access to basic education; increasing access to and pro- viders of secondary education; strengthening tertiary education, which is generally weak; improving the linkages between formal and informal education systems and the labor market; and raising the quality of learning. Tanzania's economy today is largely market oriented and it has in place many elements required for private sector-led growth. However, it does not have a sound base of adequately qualified and trained workers, which is essential for rapid economic growth and effective diversification of its production and export bases. Literacy is one of the keys to creating a sound workforce. Figure 7 shows that in 2001 Tanzania's adult literacy rate (77 percent) was higher than that of Uganda (69 percent), lower than that of Kenya (84 percent), Botswana (79 percent), and South Africa (86 percent). 22 Anuja Utz Figure 7: Adult Literacy Rates, 1960­2000 85 Mauritius 75 South Africa Kenya 65 t Malaysia Botswana Percen 55 Uganda Tanzania 45 Ghana Tanzania Ghana 35 Kenya Uganda Botswana Mauritius Malaysia South Africa 25 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 Year Source: World Bank internal database. According to Cohen and Soto, in 2000 (Figure 8), Tanzania's average years of schooling (3.4) was higher than that of Uganda (3.22), but lower than in Kenya (5.08), and far from that in South Africa (7.22). Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 23 Figure 8: Average Years of Schooling, 1970­2002 9 Tanzania Ghana Kenya Uganda Mauritius Malaysia Malaysia 8 South Africa 7 Mauritius g 6 Kenya oolinh Scfo 5 South Africa Ghana earsY 4 Tanzania 3 Uganda 2 1 1960 1970 1980 1990 2000 Years Source: Cohen and Soto (2001). In recent years the Tanzanian government has recognized the need to raise educational levels in the population as a necessary condition for enhancing economic growth. The government's focus on in- vestment in primary and secondary education, if sustained, will accelerate increases in literacy and in the average years of schooling for the medium to long term. The general education system in Tanzania comprises seven years of primary education, four years of lower secondary, and two years of upper secondary. Appropriate programs for primary and secondary education have been put in place to enhance access and increase the quality of education. Key im- provements to date have included the abolition of primary school fees in 2001, significant increases in budgetary funding for primary education, and the implementation of the Primary Education Develop- ment Program (PEDP). Under the PEDP Tanzania's gross enrollment ratios (GERs) for primary edu- cation increased from 78 percent in 2000 to 106 percent in 2004. The net enrollment ratio (NER) in- creased from 59 percent in 2000 to 91 percent in 2004.Girls represent 49 percent of the total enroll- ment. A key challenge for the government is to focus more on improving the quality of primary education. In terms of inputs, availability of textbooks has improved. On average, before launching the PEDP, one book on each subject was shared by eight students (a ratio of 1:8). In 2003, the book/pupil ratio im- proved to 1:4, and the government's target is to reach a ratio of 1:1 by 2006. Teachers' knowledge and mastery of the curriculum has also improved through the preservice and inservice teacher training in- terventions. The proportion of Grade A teachers increased from 46 percent in 1999 to 58 percent in 24 Anuja Utz 2004. However, greater numbers of qualified teachers are still needed (World Bank 2005b). The PEDP program has strengthened education institutional capacity and management, as indicated by the en- hanced capacity of the Ministry of Education and Culture to provide policy and guidelines and moni- tor education delivery. The PEDP has led to decentralized primary education, and has also strength- ened community participation and school-level management and accountability. Resource availability and utilization has also been improved, as measured by increased nonsalary expenditures in the pri- mary school budget. Tanzania has one of the lowest net enrollment ratios for secondary education in sub-Saharan Africa. Only about 9 percent of the relevant age group attends secondary education compared with an average of 27 percent for sub-Saharan Africa in year 2000, including about 11 percent at lower secondary and less than 2 percent at upper secondary. Only 22 percent of primary school leavers in Tanzania have a chance to continue their education at the secondary level, compared with 50 percent in Uganda in 2001. Secondary enrollment ratios are low for all population groups, but especially for low-income youth and students in rural areas. Few government schools have been established and inadequate in- centives exist to provide nongovernmental schools in rural communities because households are un- able to pay the fees (World Bank 2004e). There are three main challenges facing Tanzania in secondary education: increasing access, raising quality, and reducing costs. In order to support reforms in secondary education, the government has launched the Secondary Education Development Program (SEDP), which has among its aims the fol- lowing: increasing the proportion of the relevant age group completing lower and upper secondary education; expanding enrollments with equity; improving learning outcomes of students, especially girls; and enabling the public administration to manage secondary education more effectively. In order to expand enrollments with equity the SEDP includes measures to make more efficient use of re- sources, provide development grants to schools and communities (mainly in underserved areas), ex- pand the teacher supply, lower household costs for secondary education and expand the scholarship program for students from poor families, and enhance the partnership with the nongovernmental sec- tor. The program for quality improvement includes reforms of curricula and examinations, provision of textbooks and teaching materials through capitation grants to schools, and quality improvements in preservice teacher training together with the establishment of a system for professional inservice teacher development. SEDP also includes institutional reforms and capacity building at central, region, district, and school levels for more efficient operation of the secondary education system. The Southern and Eastern Africa Consortium for Monitoring Educational Quality (SACMEQ) moni- tors and evaluates the quality of education in selected southern and east African countries.17 The SACMEQ II Project (2000­2003) has been completed in 14 countries: Botswana, Kenya, Lesotho, Malawi, Mauritius, Mozambique, Namibia, Seychelles, South Africa, Swaziland, Tanzania (Mainland), Tanzania (Zanzibar), Uganda, and Zambia. The results of SACMEQ II show high reading and math score results for primary schools in Tanzania compared to other countries. In reading, Tan- zania is placed third of 14 countries, behind Seychelles and Kenya (Figure 9), while on math scores, Tanzania is in fifth place, behind Mauritius, Kenya, Seychelles, and Mozambique (Figure 10). 17For more information, go to: http://www.sacmeq.org/. Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 25 Figure 9: Reading Scores Figure 10: Mathematics Scores SEYCHELLES 582 MAURITIUS 584.6 KENYA 546.5 KENYA 563.3 TANZANIA 545.9 SEYCHELLES 554.3 MAURITIUS 536.4 M OZAM BIQUE 530 SWAZILAND 529.6 TANZANIA 522.4 BOTSWANA 521.1 SWAZILAND 516.5 M OZAMBIQUE 516.7 BOTSWANA 512.9 SOUTH AFRICA 492.3 UGANDA 506.3 UGANDA 482.4 SOUTH AFRICA 486.1 ZANZIBAR 478.2 ZANZIBAR 478.1 LESOTHO 451.2 LESOTHO 447.2 NAMIBIA 448.8 ZAM BIA 435.2 ZAMBIA 440.1 MALAWI 432.9 M ALAWI 428.9 NAM IBIA 430.9 0 100 200 300 400 500 600 700 0 100 200 300 400 500 600 700 Tanzania should also try to integrate some of the new international practices into its education system. These practices include learning from the experience of countries that have participated in interna- tional assessments such as the Trends in International Mathematics and Science Study (TIMSS). In 2007, the TIMSS will include some countries from SSA such as Botswana, Djibouti, Ghana, and South Africa. In addition, even though assessments such as the OECD's Programme for International Student Assessment (PISA) and the International Adult Literacy Survey (IALS)18 mostly pertain to developed countries, they do provide valuable insights for developing countries on the types of skills that are needed by students to effectively participate in the knowledge economy. (See Annex 3 for in- formation on these assessments). Tanzania's performance in tertiary education is very weak. Tertiary GERs stood at 0.94 percent in 2002 compared to 3.24 percent for Uganda in 2002 and 2.9 percent for Kenya in 2001, whereas Bot- swana and South Africa had tertiary GERs of 4.69 and 15.05 percent, respectively, in 2002.19 The 1990s saw the inception of the Institutional Transformation Program at Tanzania's University of Dar Es Salaam (UDSM), which led to a surge in enrollment from 3,164 in 1993­94 to 6,846 in 2000­ 01 and 14,221 in 2003­04. Female students comprised about 30 percent of total student population in 2003­04 (Cooksey, Levey, and Mkunde 2001). In April 2001 an Education Fund was established in Tanzania to sponsor the higher education of children from very poor families (EIU 2004b).20 A num- ber of departments at UDSM have changed their course offerings and moved in the direction of de- mand-driven courses. Extensive computerization of the campus has been completed in recent years. UDSM is considered by many to be in the top echelon of African universities. In fact, the World Uni- versities Ranking 2005 by Internet Lab (Aduda 2005) placed UDSM 13th among the top 100 African universities and 2,576th in the world. The University of Nairobi was ranked 24th in Africa and 4,385th in the world, while Uganda's Makerere University was placed 18th in Africa and 3,504th globally. In the last 10 years, private universities have emerged in Tanzania. Today the country has nine private universities, mostly of denominational nature and small in size, which award diplomas in areas such as financial and business management, wildlife management, community development, social welfare and cooperatives, and transport and media operations (ESRF 2002). 18 TIMSS 2003: See http://timss.bc.edu/timss2003i/conference_IR.html; PISA 2003: See http://www.pisa.oecd.org/pages/0,2987,en_32252351_32235731_1_1_1_1_1,00.html; IALS: See http://www1.oecd.org/publications/e-book/8100051e.pdf. 19 These statistics should be viewed cautiously, because the countries' systems of post-secondary education dif- fer. Tanzania has many students enrolled in post-secondary nonuniversity courses; perhaps these are not counted in the official statistics. 26 Anuja Utz Analysis of the Integrated Labor Force Survey data (Figure 11) suggest that marginal returns to educa- tion are around 8 percent for every year of primary education, but increase sharply for additional years of lower secondary education. This data provides strong support for government's investment in ex- panding primary education and the current focus on secondary education. Figure 11: Marginal Social Returns Per Year of Education Based on the Integrated Labor Force Survey 25 20 15 Percent10 5 0 Primary Lower Upper University Secondary Secondary Education level Source: Authors' calculation based on World Bank (2004e). Analysis suggests that marginal returns to education in the manufacturing sector increase significantly for higher education (Soderbom et al. 2004) (see Figure 12). Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 27 Figure 12. Predicted Earnings in the Manufacturing Sector Based on Firm Surveys Source: Soderbom et al. 2004. Note: The different lines (years 1-4) represent data from different RPED surveys in which workers have been split into old and young cohorts. The Regional Program on Enterprise Development (RPED), Africa Region, World Bank, is an ongoing research project with the overall purpose of generating business knowl- edge and policy advice useful to private sector manufacturing development in sub-Saharan Africa. These estimates do not take into account the cost of education, which is typically significantly higher for tertiary education. The difference in the profile of marginal social returns to education for workers in the manufacturing sector and the overall labor force may suggest that at the tertiary level there are certain degree programs that are well rewarded by the manufacturing sector, but that many other de- gree programs result in lower-paying jobs. A logical policy response would be to shift the supply of higher education to those programs that seem to be in demand by the manufacturing sector. The differ- ence in the profiles of social returns to education also suggests that while a limited supply of workers with relevant tertiary education is a constraint for manufacturing, for other sectors the limited supply of Tanzanians with secondary skills may be even more of a constraint. The government has been the major financier of technical/vocational education and training (TVET) in Tanzania, with assistance from donors. But the TVET system faces several problems, including ineffi- cient resource utilization, inequitable distribution of educational opportunities, poor labor market link- ages, and a lack of coordination between donors and the government. The unsustainable costs of train- ing appear to be caused by low capacity utilization and low student-to-faculty ratios. Also, the distri- bution of education opportunities is inequitable and is biased towards primary schools enrolling stu- dents from wealthier backgrounds. Recognizing that the TVET system had failed to produce graduates who were suited for the labor market, policy changes were introduced in 1996 that emphasized the government's continued responsibility in the provision and financing of more and better basic educa- tion. This emphasis was coupled with a reduction in untargeted subsidies through increased cost shar- ing, liberalization of private education and training at all levels, and decentralization of authority. The Vocational Training and Education Authority (VETA) that was set up in 1994 is working to ensure that training is responsive to the labor market (Gill and Dar 1998). 28 Anuja Utz Brain Drain Africa is a capital-scarce region and loss of this limited resource is widely considered detrimental to sustained growth and development. There is an analogous problem with human capital. Weakness in human capital and particularly skill deficiency is a drag on investment and growth in Africa. Progress in overcoming shortages of skilled and trained manpower seems to be disappointingly slow, despite substantial resources devoted by governments and donors during the last three decades. This defi- ciency persists at a time when Africa is losing a very significant proportion of its skilled and profes- sional manpower to other markets and is increasingly depending on expatriates for many vital func- tions. Although African immigration to the developed world is comparatively low, a high proportion of its migrants are highly skilled professionals. For example, it has been estimated that for some Afri- can countries more than 30 percent of their highly skilled professionals are lost to the OECD coun- tries. Nearly 88 percent of adults who emigrate from Africa to the United States have a high school education or higher. There are more African scientists and engineers working in the United States than there are in Africa. The UNECA estimates that between 1960 and 1989, some 127,000 highly quali- fied African professionals left the continent (Tebeje 2005). The emigration of technically skilled peo- ple has left SOME 20,000 scientists and engineers in Africa servicing 600 million people. Box 7 dis- cusses some ways of dealing with human capital flight from Africa to the developed world (Ndulu 2001). Tanzania is no stranger to the brain drain phenomenon. The most vulnerable occupations at the na- tional level include medicine, accountancy, law, engineering, and science-based occupations. Data from two premier institutions of higher learning provide a proxy of this phenomenon. Out of a teach- ing staff of about 861, about 149 or 17.3 percent left UDSM between1990 and March 2002. The ma- jority of those who left were from the Faculty of Arts and Social Sciences (38), followed by Medicine (17), Engineering (13), Law (11), Science (10), and Commerce (9). The ranks at which most of them exited were Lecturer and Senior Lecturer. The same was the case at the Sokoine University of Agricul- ture. Of a staff of 239 people, about 50 or 21 percent, left in the same period (1990­2002). Again the majority that left were either Lecturers or Senior Lecturers (ESRF 2002). Tanzania is also facing a massive loss of skills in the field of medicine, especially of doctors and sci- entists. Low salaries for doctors are the principle reason for the brain drain. The doctors that remain seek higher wages in private hospitals in large urban centers, leading to a lack of doctors in some of the country's district hospitals. In a bid to increase the number of health professionals in the country, the Tanzanian government has recently promised to cover all training costs for medical students in both public and private universities (Balile 2003). Box 7: Human Capital Flight from Africa The African brain drain is manifested predominantly as a growing emigration of high-skill workers to exploit opportunities abroad. The region is losing its talent because it cannot remunerate, preserve, and utilize this talent effectively. Low productivity in the economy and sluggish job growth are largely behind the low rates of return to investment in tertiary education. Coupled with this are risks of professional atrophy because of lack of interac- tion with peers at the professional frontier. In the absence of a vibrant private sector to absorb talent, demand for it will remain low and rates of returns in the local market are likely to stagnate at the current low levels. Political risks and civil unrest militate against retention of skilled workers trained locally or repatriation of those trained abroad. The problem is set to intensify in the foreseeable future because of the intensification of the globalization process and reduced cost of movement. While individual countries may take measures to reverse human capital flight, a collective initiative including enhanced labor mobility within the region would probably increase retention of skilled personnel. Such efforts could be enhanced by expanding growth opportunities at home and easier interaction with peers through profes- Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 29 sional networks. Dealing with human capital flight from Africa requires actions on many fronts, including the following: · Raise the demand for knowledge at home through private sector growth and professionalization of the public sector. The ongoing transition towards a private sector-led, market-oriented economy in many African countries, if durable, bodes well for shifting the incentive structure in favor of scarce talent. Second-generation civil service reforms that aim to reintroduce meritocracy in public service also can have positive effects on the demand for skilled and professional employees. A direct effect will be a rise in the demand for professionals, higher-quality education at entry into public service, and further train- ing on the job. An indirect effect will be the beneficial impact of improved governance on private in- vestment. · Adapt the tertiary education system to the challenge of containing human capital flight. This requires a more market-oriented approach for the delivery of tertiary education. Although encouraging changes are taking place at private institutions, the bigger challenge is for traditional public-funded universities to be both cost efficient and responsive to the changing needs of their clientele The following steps can help meet the challenge of efficient, responsive tertiary education: Sharpen the responsiveness of traditional universities in Africa to the rapidly changing needs of so- ciety and subject their programs to global pressures for excellence. Makerere University in Uganda offers an innovative example: it is transitioning from a traditional, elitist, predominantly publicly financed institution to a more market-oriented institution funded largely by private resources. Rationalize the tertiary education system by opening more private universities and other private ter- tiary institutions. Private institutions are often responsive to the increasing demands for nonconven- tional and nonacademic courses that address local needs and employment opportunities. Increase quality control and independent certification in the tertiary education system. This in- cludes progressive movement towards applying international standards in the certification process. International certification standards would help create a local pool of world-class talent, which in turn would encourage and sustain the relocation of multinational corporation to the region and would raise returns to tertiary graduates. Encourage regional and subregional collaboration, for example by supporting and sustaining the emerging regional and subregional research and training networks. Regional networks are a rational and cost-effective way to support the capacity building of national institutions and also provide a platform for professional peer review and support. Increase the cost-effectiveness of delivery of tertiary education. Cost effectiveness is necessary if tertiary institutions are to survive in the midst of strong local and world-wide competition. · Adopt strategies for reversing brain drain, including assisted permanent return, recruitment of African talent in the Diaspora, and recruitment of qualified African emigrants in technical assistance schemes. In terms of involving Africans in the Diaspora, it is important to design schemes that allow skilled Afri- cans working abroad to contribute to the development of their home countries without giving up the higher wages and better living standards that their current permanent residence affords. This approach will also enable the large number of African intellectuals and intellectual communities abroad to par- ticipate more actively in strengthening the capacity for quality education and training in the region. Recently, scientists and professionals in the African Diaspora and their counterparts back home teamed up to establish a world-class African Institute for Science and Technology (AIST). AIST is rooted in strong public-private and industry-academy partnerships, which should be conducive to long- term sustainability. AIST is a global effort to foster Sub-Saharan Africa's economic growth and diversi- fication, industrial development, and employment creation through the promotion of excellence in sci- ence, engineering, and their applications based on competitive processes, transparency, and accountabil- ity. The institute which will form the nucleus of the Nelson Mandela Foundation for Knowledge Build- ing and the Advancement of Science and Technology. Source: Ndulu (2001); Aderinokun (2005). 30 Anuja Utz Threat of HIV/AIDS HIV/AIDS poses significant threat to human resource development in Tanzania. It is estimated that 3 million people including children currently are infected with the disease. It is established that the widespread HIV/AIDS epidemic has substantial slowed economic growth and productivity. Previous studies that focused on Africa calculated the annual loss of GDP growth to range between a modest 0.3 percent to a high of 1.5 percent. But a new report argues that the costs are likely to be much higher. Previous estimates overlooked the impact of HIV/AIDS on children if one or both parents die. Countries facing an HIV/AIDS epidemic on the same scale as South Africa, for example, could see family incomes cut in half and could face economic collapse within several generations is swift action is not taken (Bell, Gersbach, and Devarajan 2003). Tanzania therefore needs to continue taking steps to curb the scourge of HIV/AIDS. A recent article by Hazlewood and Prakash (2005) reiterates the fact that Tanzania faces an acute shortage of health care workers. Low pay, poor working conditions, and limited training programs contribute to the problem, which is exacerbated by the rising burden of treating HIV/AIDS patients. Hazlewood and Prakash estimate that Tanzania will have to find nearly 10,000 more workers to ad- dress the rising needs of HIV/AIDS patients and three times that number to meet the Millennium De- velopment Goals (Figure 13). Improving the productivity of individual health care workers and of the health care system as a whole could increase Tanzania's capacity by two-thirds, even without hiring additional staff. Some improvements, such as telephones and motorbikes for better communication, would be relatively easy to make; others such as better management of the flow of patients and the implementation of planning and accounting tools would require more investment and training. New investment and training would require health organizations and the government increase the capacity of their training programs by at least half to ensure a sustainable workforce. Also, they would need to aggressively recruit trained staff to alleviate the immediate shortages. Figure 13: Projected Shortfall of Health Care Workers in Tanzania Source: Hazlewood and Prakash (2005). Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 31 Pillar 3: Information and Communication Technologies (ICT) Rapid advances in ICT are dramatically affecting economic and social activities, as well as the acqui- sition, creation, dissemination, and use of knowledge. Advances in ICT also are affecting the way in which manufacturers, service providers, and governments are organized and how they perform their functions. As knowledge becomes an increasingly important element of competitiveness, the use of ICT is reducing barriers of cost, time, and space to economic transactions, allowing the mass produc- tion of customized goods and services and compensating for limited factors of production. The pervasive and global ICT revolution is disrupting all kinds of relationships, helping build new types of organizations, widening the knowledge and productivity gap, and posing serious risks for the unprepared. Some countries are concerned about being left behind (the "digital divide") and others are inspired by the opportunities to leapfrog and participate in the new knowledge industries. To avoid being left behind, countries must harness the full potential of ICT and use it to improve education and innovation, public sector management, private sector competitiveness, and capacity building. Research shows strong linkages between ICT and growth. Compelling evidence supports the observa- tion that strengthening the telecommunications infrastructure and service is pivotal in promoting trade and economic growth. It is estimated, for example, that a 10 percent decrease in the bilateral price of phone calls is associated with an 8 percent increase in bilateral trade (Fink, Mattoo, and Neagu 2002). A developing country which between 1996 and 2003 put into service an average of 10 additional mo- bile phones per 100 inhabitants would have 0.59 percent higher per capita GDP growth than an other- wise identical country (Vodafone 2005). In Africa, significant evidence exists to suggest that if tele- phone growth rate were10 percent instead of 5 percent (and growth in electricity generation was 6 per- cent instead of 2 percent), the increase in Africa's growth rate would be at least 0.9 percent higher (Es- tache 2005). The information infrastructure in a country consists of telecommunications networks, strategic infor- mation systems, policy and legal frameworks affecting system deployment, and skilled human re- sources needed to develop and use information systems. Tanzania, Kenya, and Uganda are all at a very nascent stage of ICT application and use. It is not surprising that all three countries lag Botswana, South Africa, and Malaysia by a huge margin, as shown in Figure 14. The figure also shows that Mau- ritius has been doing better than South Africa and Botswana on telephony and personal computer penetration and is close to the level of Malaysia. In the case of Internet hosts, South Africa and Malay- sia are the undisputed leaders. 32 Anuja Utz Figure 14: ICT Infrastructure: Telephones, Personal Computers and Internet Hosts Telephones (Mainlines and Mobile Phones) Per 1,000 persons Tanzania Ghana Kenya Uganda Malaysia 500 Botswana Mauritius Malaysia South Africa Mauritius 400 300 South Africa Botswana 200 Tanzania 100 0 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 Year (Figure 14 continues on next page) Per 1,000 persons Personal Computers 140 Tanzania Ghana Kenya Uganda Malaysia Botswana Mauritius 120 Malaysia South Africa 100 Mauritius 80 South Africa 60 40 Botswana Tanzania 20 Kenya 0 1990 1992 1994 1996 1998 2000 2002 Year Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 33 Per 10,000 persons Internet Hosts Tanzania Ghana 40 Kenya Uganda Botswana Mauritius 35 Malaysia South Africa South Africa 30 25 Malaysia 20 Mauritius 15 10 Botswana Tanzania 5 0 1994 1995 1996 1997 1998 1999 2000 Year Source: World Bank internal database. (Figure 14 concluded) A 2005 report by the United Nations Conference on Trade and Development (UNCTAD) provides some insights into the international digital divide and evaluates ICT development using a range of in- dicators to benchmark connectivity, access, ICT policy, and overall ICT diffusion in 165 countries. In the benchmarking analysis, OECD countries continue to dominate the upper rankings, while South Asian and African countries occupy the lower half of the rankings. The more developed African coun- tries enter the rankings relatively early, with Mauritius in 52nd place and South Africa in 66th Place. Botswana ranks 80th, Kenya ranks115th, Tanzania ranks 135th, and Uganda ranks 154th. These rankings indicate that many SSA countries have a considerable way to go in terms of ICT connectivity and dif- fusion (UNCTAD 2005a). However, in recent years Tanzania has been making some progress in ICT. In 2003, it published a cross-sectoral National ICT Policy (www.moct.go.tz/ict) that relates ICT to relevant sectors of the economy such as education, manufacturing, health, and tourism. It was developed in response to the poor harmonization of initiatives that has led to random adoption of different systems and standards, unnecessary duplication of effort, and waste of scarce resources, especially through the loss of poten- tial synergies. The policy notes that weak ICT infrastructure and the lack of adequately trained and skilled personnel are the main barriers to increased adoption of ICT in Tanzania. The policy is de- signed to correct these weaknesses and is a broad-based national strategy that addresses Tanzania's developmental agenda, and calls for the creation of appropriate institutional arrangements to ensure that all stakeholders can rise to the challenge of implementing the ICT policy. It is worth mentioning that stakeholder discussions on ICT policy and related issues were held mainly through an e-mail list or e-think tank comprising representatives from government, private sector, and civil society. 34 Anuja Utz The policy and legal and regulatory framework for Tanzania's telecom sector encourages private sec- tor participation.21 This sector is regulated by the Tanzania Communications Regulatory Authority (TCRA, formerly, TCC). The performance of Tanzanian Telecommunications Company (TTCL) has improved considerably since February 2001, when a Dutch-German consortium, MSI, took a 35 per- cent stake in the company. The remaining shares were allocated to local financial institutions (14 per- cent), international financial institutions (10 percent), and TTCL employees (5 percent). The govern- ment retained a 36 percent stake. At present TTCL has around 250,000 operational lines (EIU 2004b). Sector liberalization and the privatization of TTCL has had a significant impact on market dynamics, particularly in the supply of telecommunication services. Market revenues grew from US$143 million in 1998 to US$389 in 2003 and the compound annual growth rate rose to 19 percent in 2003. Overall teledensity grew from 0.3 in 1998 to 2.57 in 2003. The mobile market has grown almost 21 percent annually since liberalization and there has been growth in competition, new products, and services. Benchmarking of Tanzania against other countries in the same region shows the need for further sector reform. Despite the competition, tariffs remain high and teledensity in Tanzania is one of the lowest in the Southern African Development Community, in part because of the poor interconnection frame- work, the lack of regulatory independence, and other issues such as lack of infrastructure sharing. In general, the country's postal and telecommunications services are weak and provision of telephone lines (fixed lines) has been meager. An inadequate regulatory framework persists, and competition has been hampered by various issues such as inadequate interconnection agreements/directives, high fees and royalties levied by the TCRA, and the absence or nontransparency of regulatory oversight. However, progress has been made under the 1997 National Telecommunications Policy, and this trend is expected to continue. Significant liberalization has also taken place in various segments and private operators are now providing basic, mobile, data, paging, Internet, payphone, and other value-added services. For example, the mobile telephone market in Tanzania has a number of operators, is fully competitive, and is growing rapidly. New mobile operators have committed significant financial re- sources to the development of a state-of-the-art telecommunications infrastructure and the recruitment of mobile subscribers. There are four providers operating under 15-year licences: MIC Tanzania, Zan- zibar Telecoms, Vodacom, and Celtel. There are now more mobile subscriptions than fixed lines; this cross-over point was reached in 2000 (just five years after the first mobile was sold). Table 3 shows that overall teledensity--mainlines plus mobile phones--grew to 24 phones per 1,000 inhabitants between 1996 and 2003. Tanzania's mainline and mobile phone penetration is higher than that in Uganda. 21For example, Draka Comteq, which consists of 11 companies in Denmark, Finland, France, Germany, Nether- lands, Norway, the United Kingdom, the United States, and Singapore, has recently won an approximately US$30 million turnkey project in Tanzania. The project involves the supply and installation of two long- distance links, covering some 2,300 kilometers of optical fiber cable and 500 kilometers of optical power ground wire. (Source: EIU 2004a.) Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 35 Table 3: ICT Indicators for Tanzania and Comparators 1996­2003 2002 2000 2003 Mobile Personal Internet Internet Mainlines Radios Televisions telephones computers hosts users (per 1,000) (per 1,000) (per 1,000) (per 1,000) (per 1,000) (per 10,000) (per 1,000) Botswana 87 241 150 44 40.7 13.99 50* Kenya 10 42 221 26 6.4 0.32 200* South Africa 107 304 336 177 72.6 41.94 2,890* Tanzania 5 19 406 4 4.18 0.16 240 Uganda 2 16 122 18 3.32 0.07 125 Sub-Saharan 11.90 3.10 6,233 15 37 198 69 Africa average Source: World Bank (2004a and 2005a). * denotes data for 2001. There are also several studies documenting the improvement in prices received by farmers as a result of better access to telephony in general and mobile phones in particular in developing countries in Asia and Africa. One example is the case of fishermen in India who use mobile phones to get informa- tion about prices at different ports before deciding where to land their catch. This example was con- firmed in a study done by Vodafone of fishermen on Mafia Island, off the Tanzanian coast (Vodafone 2005) as well as on the island of Zanzibar (Box 8). Anecdotal evidence also exists that traders in Dar es Salaam now can place orders with producers of bananas throughout the country, thus linking de- mand and supply in real time and enhancing the efficiency of markets. Box 8: Mobile Phones in Tanzania Mobile use in Tanzania and across Africa is gaining momentum and users continue to show creativity in using the devices. Some 97 percent of Tanzanians say they can access a mobile phone and it is interesting how those phones are being used. On the island of Zanzibar, for example, where fishing is one of the mainstays of the economy, many fishermen now carry mobile phones while they are at sea, and they use them to check market prices. If there are too many fish in Zanzibar, they go to Dar es Salaam to get better prices. Phones also serve another even more vital use--allowing fishermen in trouble to call for assistance. Mobile phone call centers have sprung up all over Tanzania. Most people do not actually own phones, so this is how many people communicate. Others have developed even simpler businesses based around mobiles, such as reselling their air time or sending and receiving text messages. Mobile phones seem to have created a new sector of the economy and some now wonder if the emphasis on the Internet when considering the digital divide was wrongheaded. On Zanzibar, the latest product by the island's cell phone operator, Zantel, is a fixed-style phone for homes that will not be connected by copper wire or even fiber optics. Zantel is thinking about providing service through CDMA wireless technology. This is a very useful and effective technology that is much faster than copper wir- ing. Zantel has recently been granted a license to extend service nationwide. To extend their network across the country will cost just half a billion dollars. Even without expansion, Zantel's subscriber base has been growing 25 percent per year. Source: Hancock (2005). 36 Anuja Utz The 2005 Vodafone report includes research on the social and economic impacts of mobiles. This re- search was compiled from the results of surveys on the use of mobiles carried out in rural communities in South Africa and Tanzania; surveys of small businesses in Egypt and South Africa were also con- sulted. The community surveys assess factors affecting mobile use and the range of potential impacts in relatively poor rural African communities. The surveys suggest that mobile telephony is frequently accessed by the poorest people, thanks in part to widespread sharing. The surveys suggest that gender, age, education, and even the absence of electricity do not present insurmountable barriers to access. Individuals surveyed in rural communities highlighted savings in travel time and costs and easier communication with family and friends, in addition to access to business information and easier job searches. A majority of small businesses reported increased sales and profits, time savings, and greater efficiency. The surveys also present important conclusions about social capital or the strength of social networks and contacts in the rural communities. Mobile phone ownership was positively linked to life satisfaction and a willingness to help others. A majority of respondents said owning a mobile had im- proved their relationship with family members living elsewhere. Thus, according to three articles in the Economist (2005a, 2005b; 2005c), mobile phones have tre- mendous potential to help boost entrepreneurship and economic activity. The phones provide an alter- native to bad roads and unreliable postal services, widen farmers' access to markets, and allow swift and secure transfers of money. But the primary obstacle to wider adoption is the cost of handsets. In the developed world, phones typically cost around $200 (though most users pay less thanks to subsi- dies from network operators) or less than 1 percent of the average income per person. In the develop- ing world, a $50 handset would account for 14 percent of the annual income of someone earning $1 a day. So the first step in promoting the adoption of mobile phones is to reduce the cost of the handsets. Several such schemes are under way. Motorola, for example, has agreed to supply up to 6 million handsets for less than $40 each (see Economist 2005b). Industry observers believe cheaper handsets could expand the market by as many as 150 million new subscribers a year. As well as boosting eco- nomic development in poor countries, this will help to close the "digital divide" between the commu- nications-rich and communications-poor. Tanzania has comprehensive Internet service, including 3 licensed data-service providers and 21 Inter- net service providers. Most Internet users access the Internet from urban Internet cafes--the Internet is not much present in rural areas. Tanzania in 2003 had almost double the number of Internet users (240,000) as Uganda (125,000). The government has also developed a fairly comprehensive national Website (http://www.tanzania.go.tz/), which provides considerable background information on the economy and the political structure of the country; it hopes that the site will help to raise the country's international profile and attract foreign investment. In addition to the national Website, a number of ministries, state institutions, and embassies have their own sites. In terms of developing human resources in IT, training centers that focus on the development of ICT knowledge workers are only now beginning to emerge. For example, the Soft Tech Training Center, established in 1993, is committed to the development of local expertise through ICT skills enhance- ment. The government has also initiated plans to encourage Tanzanians to develop content that is rele- vant to local interests, and Tanzania has implemented several ICT applications relevant to its national objectives. Examples of such applications include an information system to strengthen the capacity of wildlife institutions and a computerized case-flow management system that has facilitated an increase in transparency and professionalism in the judiciary system (Accenture, Markle Foundation, and UNDP, 2001). Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 37 Pillar 4: Innovation System The innovation system in any country consists of the institutions, rules, and procedures that affect how it acquires, creates, disseminates, and uses knowledge. Innovation in a developing country concerns not just the domestic development of frontier-based knowledge. It also encompasses the application and use of existing knowledge in the local context, and requires a climate favorable to entrepreneurs. Such a climate is free from bureaucratic, regulatory, and other obstacles and fosters interaction be- tween the local and global business world, as well as between global and local sources of knowledge. These knowledge sources include traditional establishments such as universities and public laborato- ries, as well as users, think tanks, industrial milieu, indigenous communities, and so on. Innovation systems comprise the interactions among these entities and they are more or less complex and sophis- ticated depending on the level of a country's development. As an economy develops, the role played by domestic research capabilities naturally tends to increase, whereas in low-income countries innova- tion is much dependant on foreign technology and its adaptation to the local context. Innovation in the context of developing countries thus needs to be understood in a broad way: it con- sists of the design, development, and diffusion of a technology (or a practice) which is new for the society concerned. Innovation needs and opportunities can be discussed at three main levels: the first level is the local improvements through adoption of available technologies; the second level relates to the development of competitive industries through adaptation of technologies (initially produced in or by developed countries); and the third level is developing brand new innovations of global signifi- cance. Innovation policies in developing countries must primarily deal with the building of an appro- priate technical culture and the establishment of incentives to stimulate cooperation and entrepreneur- ship. Innovation often begins with demonstration projects that aim to creating trust among the key ac- tors and to facilitate reforms. It is a long-term process and generally takes a decade or so to begin bear- ing fruits in terms of job and wealth creation or other benefits. There are various ways that countries can strengthen their innovation systems. An important channel is tapping into global knowledge through trade, foreign investment, and technology transfer. Countries need to create and adapt knowledge, and a primary way of doing so is to harness knowledge from spe- cialized research institutions to meet the productive needs of the economy. Along with knowledge creation, knowledge dissemination can lead to the growth of more efficient enterprises and the growth of technical services and information such as agricultural extension services. Tanzania and other countries in sub-Saharan Africa have a long way to go in terms of technology crea- tion. Their progress or lack of can be measured by variables such as patents granted to residents, re- ceipts of royalties and license fees, the diffusion of recent innovations through channels such as the Internet, and high- and medium-technology exports. Countries also need to strengthen their human skill base; one measure of skills strength is the gross tertiary science enrollment ratio. Tanzania spends about 0.2 percent of GDP on R&D. This figure is not significantly different from av- erage R&D expenditure in sub-Saharan Africa, estimated at 0.5 percent of GDP (UNESCO Institute for Statistics 2001). R&D activities in the agricultural sector have had some success in developing high-yielding and disease-resistant varieties of crops, as well as in creating technologies for increasing crop productivity through soil and water management.22 22 The Bank's Second Tanzania Agricultural Research Project (TARP) aims at increasing efficiency and produc- tivity of crops and livestock production systems with sustainable use of natural resources, focusing on the needs of the small-holder sector. The project consists of three components. First, strengthen the agricultural research system. This will result in more effective and efficient agricultural research institutions and will also produce an 38 Anuja Utz In terms of outputs of the innovation system, from 1988 to 2001 the number of scientific articles pub- lished worldwide grew by 40 percent. Africa failed to keep pace with this growth, and its publication counts actually declined by 12 percent in absolute terms. In 1988, Africa accounted for 1.26 percent of all scientific publications; by 2001 its share was only 0.76 percent (Juma 2005). A review of Tanzania's policies shows a high commitment to science and technology in development, but it needs to develop effective public-private partnerships. In the past, Tanzania had no deliberate strategies or plans for appropriate selection, acquisition, and transfer of technology or for effective integration of imported technologies with local capacity for R&D. However, efforts have been put into place over the years to make sure that the speed of technology transfer is effective and sustainable in the economy. Tanzania enacted the first National Science and Technology policy in 1985, which was subsequently revised in 1995. The major thrust of this policy is to establish relative priorities and pro- grams to generate new knowledge and to determine strategies for science and technology development in Tanzania. Tanzania also established the Tanzania Commission for Science and Technology (COSTECH) in 1986 and the Centre for the Development and Transfer of Technology (CDTT) in 1994 in an effort to institute workable mechanisms for coordination of capacity building efforts, adop- tion of new technologies, strengthening R&D, and facilitating information exchange and extension services.23 While these are laudable initiatives, the reality is that these institutions, especially the CDTT, lack adequate resources, infrastructure, equipment, and trained personnel to respond to the increased needs of the local entrepreneurial society. Their support from the government is decreasing, which has led COSTECH and CDTT to undertaking reforms to become independent executive agencies. The agen- cies are also look for profitable new lines of businesses that could enable them to meet their personnel and operational expenses. This trend is worrying as it could hinder these institutions from conducting unremunerated research that could be beneficial to the society as a whole. In Tanzania there are currently about 63 R&D institutions. This includes institutions that study agri- culture (including livestock and forestry) (28), industry (11), medical (11), and wildlife and fisheries (4), as well as universities and other higher learning institutions (9) (ESRF 2002). Most of these are government institutions that conduct scientific research and design and manufacture machinery and equipment for agriculture, as well as researching appropriate technologies for rural, small, and me- dium-scale industrial enterprises. However, these government institutions do not provide good incen- tives for R&D; as a result, few researchers pursue R&D activities. effective arrangement for donor coordination and for coordinating regional and other collaborative research pro- grams. Second, improve the organization and management of agricultural research. This will provide demand- driven, client oriented research, encourage efforts to introduce participatory techniques, and create guidelines for small-farmer oriented and gender-sensitive research. Third, increase human resource development. This will produce more research personnel for needed disciplines and will also strengthen financial management and in- formation services. The Implementation Completion Report for this project, currently under development, should provide insight on the success in achieving the above objectives. 23For more information, see the Science and Technology Section of the Tanzania Country Profile on Tanzania's national Website: http://www.tanzania.go.tz/science_technology.html. According to this Website, Tanzania's science and technology infrastructure includes education infrastructure and R&D institutions, such as the Uni- versity of Dar es Salaam, Sokoine University of Agriculture, University College of Lands and Architectural stud- ies, Muhimbili University College of Health Science, Rwegalulira Water Resources Institute, National College of Mbeya, Arusha, and Dar es Salaam Institute of Technology. Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 39 A recent paper on innovation in Tanzania (Goedhuys 2005) sheds light on the various sources of firm learning, investment, and collaboration and their relative importance for product innovation. The paper uses using firm level data on learning and product innovation in Tanzanian manufacturing and com- mercial farming. The results indicate that larger and foreign-owned firms invest significantly more in human and physical capital than do local microenterprises or small-, and medium-size enterprises (SMEs) and they are better connected to the Internet. Their ways of upgrading technology also reveal a better financial endowment. SMEs on the other hand report more intensive collaboration with other local firms on product development, marketing, and on the input market and upgrade technology. The collaboration takes place through in-house activities, imitation, and cooperation with suppliers and universities. Through collaboration, SMEs are able to offset the scale disadvantages they face in com- peting for market information and for new machinery and specialized labor necessary for product in- novation in imperfect markets. Box 9 provides information on key initiatives in Tanzania aimed at financing SMEs. Box 9: SME Development in Tanzania--Key Initiatives There are three key initiatives in Tanzania aimed at financing small and medium sized enterprises (SMEs): the Small Industries Development Organization (SIDO), the National Microfinance Bank (NMB), and the newly established SME Credit Guarantee Scheme. SIDO remains the main government arm for promoting SMEs in Tanzania. Some of the key measures employed by SIDO include: the construction of 16 industrial estates with more than 140 sheds at regional headquarters; the establishment of 10 training-cum-production centers that offer simple rural-based technologies; the introduction of lease-purchase programs through which more than 2,000 entrepreneurs were assisted with machines and working tools; and the creation of feasibility studies. In collabo- ration with other stakeholders, SIDO supported the establishment of SME associations to empower the private sector. The Tanzania Food Processors Association (TAFOPA) and the Tanzania Small Industries organization (TASISO) are probably the two most prominent SME associations. In April 2003, the government of Tanzania adopted, under the leadership of the Ministry of Industry and Trade, a comprehensive SME Development Policy. The SME Development Policy focuses on three main areas: (i) the creation of a supportive business environment; (ii) development of financial and nonfinancial services; and (iii) creation of an appropriate institutional infrastructure. The SME Development Policy takes into account the spe- cial constraints and opportunities faced by SMEs and aims to strengthen institutions that will address these con- straints. Source: OECD Development Centre (2005). In 2001, the UNDP's Human Development Report introduced the Technology Achievement Index (TAI), which tries to captures how well a country is creating and diffusing technology and building a human skill base, and thus reflecting capacity to participate in the technological innovations of the network age. This composite index is not a measure of which country is leading in global technology development, but focuses on how well the country as a whole is participating in creating and using technology. A good example is the comparative TAI rank of the United States, global technology powerhouse, and Finland. The United States has far more inventions and Internet hosts in total than does Finland, but does not rank as highly in the TAI because in Finland the Internet is more widely diffused and more is being done to develop a technological skill base throughout the population. The TAI thus recognizes that technological achievements are a large and complex issue. However, it should be stressed that the TAI is constructed using indicators, not direct measures, of a country's achievements in four dimensions. It provides a rough summary--not a comprehensive 40 Anuja Utz measure--of a society's technological achievements. Many aspects of technology creation and diffu- sion and human skill levels are hard to quantify. And even if they could be quantified, a lack of reli- able data makes it impossible to fully reflect them. For example, important technological innovations occur in the informal sector and in indigenous knowledge systems. But these are not recorded and cannot be quantified in the TAI. Even though the TAI results are somewhat dated, they show three trends: a map of great disparities among countries, diversity and dynamism in technological progress among developing countries, and a map of technology hubs superimposed on countries at different levels of development. Two details of the trends are particularly noteworthy: · The map of great disparities shows a group of four countries with TAI values ranging from 0.744 for Finland to 0.066 for Mozambique. These countries can be considered leaders, poten- tial leaders, dynamic adopters, or marginalized. · Tanzania and Kenya are both listed as marginalized (below 0.20), which means that technol- ogy diffusion and skill building have a long way to go in these countries, and that large parts of the population have not benefited from the diffusion of technology.24 Mauritius provides an interesting example for African countries interested in diversifying their economies to become more productive and achieve faster growth. Traditionally, the strength of the Mauritian economy (the "Mauritian Miracle") has rested on three pillars: sugar, export processing zones (EPZs), and tourism (Box 10). In the short to medium term these sectors will continue to grow, but Mauritius has realized the need to diversify and has embarked on an ambitious program to find new drivers for economic growth. The country is reengineering itself to move up the value chain and build a KE based on financial services and ICT. Measures are also being taken to modernize the sugar and EPZ sectors through higher productivity, better technology, and greater capitalization. If Tanzania can learn from the successes and failures of other countries, it may be able to "leapfrog" its past economic performance. 24The World Bank has also recently undertaken work to unravel the "how to" of technological change and un- derstand why some developing countries are able to catch up with developed countries while others lag and chart low-growth trajectories. Catching up involves accessing, learning, and applying the technological standards of developed countries; growing faster, generating better-paying jobs; and reducing poverty. The Bank work in- cludes a series of studies that cover the extremes of the industrial spectrum: wheat and maize in India; salmon and wine in Chile, Nile perch in Uganda, oil palm in Malaysia, cut flowers in Kenya, medium-tech electronics in Malaysia, high-tech electronics in Taiwan, and software exports from India. The industries have been chosen based on exceptional comparative performance in the last decade, large contributions to overall growth, and for evidence that technological change played an important role in their success. For more information, go to: http://www-wbweb.worldbank.org/prem/prmep/economicpolicy/20050216.html. Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 41 Box 10: The Mauritian Miracle--Can Others Emulate It? In 1961 James Meade, the 1977 recipient of the Nobel Prize in economics, made a dire prognosis on Mauritius's de- velopment prospects. According to Meade, the country's heavy economic dependence on one crop (sugar), vulner- ability to trade shocks, rapid population growth, and ethnic tensions were all against the country's economic success. In addition, Mauritius had a tropical climate and was farther from world markets than most African countries. Meade predicted that the outlook for peaceful development was poor, but history proved him wrong. It was partly openness to FDI and trade that brought about Mauritius' success. Local investors were heavily involved in the export process- ing zone (EPZ). In 1984 only 12 percent of employment in the EPZ was accounted for by wholly foreign-owned firms. Mauritian nationals owned about 50 percent of the total equity of EPZ firms. And contrary to popular belief, Mauritius had a highly restrictive trade regime: exports were relatively open but imports were closed. What, then, led to the Mauritian miracle? Mauritius ensured that the export sector was not adversely affected by the restrictive trade regime. Policy makers used several instruments to segregate the export sector from the import sector: Duty-free access was provided to all imported inputs, ensuring that the export sector was competitive in world markets. Tax incentives were provided to firms operating in the EPZ. The labor market for the export sector was effectively segmented from the rest of the economy. Employers in the EPZ had great flexibility in discharging workers and paying overtime. A large proportion of workers in the EPZ were women, whose minimum wages were lower than those of the rest of the labor force. In the 1980s wages in the EPZ were 36­40 percent lower than in the rest of the economy. Sound macroeconomic policies also helped the EPZ. Between 1973 and 2000 annual inflation averaged 7.8 percent, fiscal deficits were sustainable, and the exchange rate maintained the country's export competitiveness. The rule of law prevailed and property rights were respected. The economy has grown at 5­6 percent a year for the past 20 years and unemployment fell to 1.5 percent by 1990. Mauritius was a poor country with per capita income of just $260 (in nominal terms) at the beginning of the 1960s. In 2005, Mauritian per capita income was about $3,800. Two other factors facilitated the export-led growth. First, Mauritius enjoyed preferential access to U.S. and EU mar- kets (for sugar and textiles and clothing). Second, ethnic diversity helped attract investment from the Far East and ensured participatory politics that in turn established a stable political environment. Can other countries in Africa emulate Mauritius's success? It may not be easy. Preferential trade margins for African countries are eroding in an increasingly globalized world. Nevertheless, Mauritius's good macroeconomic environ- ment and political stability--conditions that ensured success despite selective trade policies--are necessary ingredi- ents for any economic success story. Source: UN Economic Commission for Africa (2003). 42 Anuja Utz 5. Conclusions The above analysis points to several ways that Tanzania could strengthen each of the four pillars of the KE. The challenge is to build on the progress made so far and to take advantage of the opportunities offered; continue to build the conditions for more effective creation, dissemination, and use of knowl- edge, both domestic and global;, and take practical steps, driven by knowledge initiatives, to stimulate new forms of income and employment generation. The following provides a brief summary of some of the issues that need to be tackled in each of the four pillars. Economic and Institutional Regime: Though substantial progress has been made in this domain, the following actions are recommended: · Further strengthen the overall regulatory framework, increase government effectiveness, and control corruption in Tanzania. As mentioned, the government has adopted a comprehensive National Anti-Corruption Strategy and Action Plan, which focuses on a variety of measures of measures to increase transparency and accountability. · Improve the overall investment climate. High cost and limited availability of infrastructure services, barriers in human resources and the labor market, and red tape in the public sector are viewed as principal constraints to achieving even higher growth rates. There is a need to understand the strengths and weaknesses of Tanzania's national business environment in order to reveal obstacles to its competitiveness. · Enhance the current competition policy framework in order to attract more and better FDI. This includes eliminating barriers to entry that continue to deter foreign investment and limit- ing the contestability of most domestic markets (including administrative barriers, taxes, and trade policies), resolving competition issues in a few strategic sectors that have benefited from FDI over the past few years, and strengthening existing legal and institutional frameworks.25 · Proactively disseminate the research and analysis performed in academic institutions to gov- ernment and nongovernment players. Disseminating this research requires strengthening pol- icy-making resources and institutions in Tanzania. Currently, ESRF and Research on Poverty Alleviation (REPOA) provide such analysis, but are also constrained in financial and human resource terms in their ability to deliver timely policy recommendations. In addition, these in- stitutes also are being used by donors to conduct research on a wide range of areas, which re- duces their ability to provide timely information to government on issues that are important for policy making. Box 11 discusses the role of policy institutes and think tanks in strengthen- ing policy making in African countries. · Encourage the donor community in Tanzania to integrate their support through a coordinated system to avoid overlaps and duplication; and help build the capacity of local Tanzanian pro- ject teams in various aspects of project management. 25For more information, see International Finance Corporation/World Bank (2002). Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 43 Box 11: Strengthening Policy Making in African Countries: The Role of Policy Institutes and Think Tanks There is no doubt that most African countries urgently need to strengthen their policy-making resources. Outside of government, capacity for policy analysis resides in universities, academically oriented research institutes, and, more recently, national policy institutes. A 1994 study conducted by the African Economic Research Consortium (AERC) in eight sub-Saharan African countries found that most government and nongovernmental organizations did not possess their own in-house capacity for policy analysis, relying instead on the institutions mentioned above. A more worrying finding from the study, perhaps, is the fact that research and analysis performed in aca- demic institutions is rarely available to either government or nongovernment users because of poor dissemina- tion. However, the study did find that the barriers between the two groups are coming down because researchers have gained credibility by producing more policy-oriented research and presenting their recommendations in practical formats. Thus, a key question is how policy think tanks can develop into effective and sustainable insti- tutions that can provide leadership in policy analysis and development management. National policy institutes have emerged during the past five years funded by regional and international donors, governments, the private sector, and by revenue generated from undertaking specific tasks for clients. As knowl- edge intermediaries at the local level, the national policy institutes play a particularly useful function in enhanc- ing access to global and local knowledge. At the same time, they help shape the agenda of knowledge producers through feedback. The most successful institutes have built strong credibility with both users and suppliers of knowledge and play important roles as honest brokers in dialogues involving competing interest groups. The institutes also respond to requests for policy analysis by mobilizing local talent to carry out necessary research. In many cases, they operate knowledge access systems (including Websites) for users and researchers. By main- taining a forward-looking research agenda, national policy institutes seek to proactively influence the policy agenda. Intermediation efforts are not confined to national levels only. Regional or subregional policy research networks are also emerging and across several disciplines and specializations. The most prominent example is a network of economists engaged in policy-oriented macroeconomic research, which is coordinated and supported by AERC. Other examples include a network of environmental economists (East Africa Environmental Economics Network), one of social scientists (Council for Development of Social Science Research in Africa), and a fran- cophone network dealing with industrial policy. Policy seminars organized by these networks can serve to bridge the gap of suspicion between academics and policy makers. Not only do the networks provide a showcase for local talent in policy analysis, but they also foster important dialogue on policy issues. Joint initiatives between public and private researchers have also proved to be an effective way of cultivating close relations. Another study looked at the experience of think tanks in Africa and derived the following tentative conclusions: 1. Policy think tanks in Africa that have mushroomed since the early 1990s have played a greater role in the policy process than is often acknowledged. They emerged because of demand on the supply side. These think tanks need to be encouraged to present African perspectives on development from both the demand and supply side. 2. There is no single model of a think tank or for relations between think tanks, funders, and governments. Think tanks must adapt their strategies and products to particular national or regional environments. It is important to develop a brand identity for the institution through the development of core products and to have a balance between research and consultancy. If think tanks could specialize on a subregional or even regional basis, networking could be more complementary and less competitive. 3. Think tanks should put in place proper governance structures and invest in capacity building and ex- change of best practices; this could be accomplished through training in policy analysis, regional semi- nars on think tank management, and so forth. 4. It is important to develop linkages that allow think tanks to cover their core costs. The relationships among universities, policy research think tanks, and policy dialogue and lobbyist think tanks needs to 44 Anuja Utz be revisited with a view to tapping more effectively avenues of complementarity. Greater emphasis is needed on creating twinning arrangements with think tanks from the North, Asia, and Latin America, particularly in areas such as training, distance learning, and other appropriate capacity-building activi- ties. Global institutions such as the World Bank Institute and the Global Development Network should strengthen their institutional links with policy think tanks in Africa based on more institutionalized, continuous, and sustainable interactions as opposed to the current one-off or single activity-based net- working. 5. Policy think tanks produce both public goods and private goods. Activities that produce private goods should be developed along with markets for such goods. Activities such as publishing periodic eco- nomic reports, analyzing data collected by the government, and publications on trends of interest to business could be sold to the market. The surplus generated through selling private goods can be used to finance the production of public goods. 6. Governments should be encouraged to develop tax laws and other laws governing the establishment and development of think tanks to encourage their establishment and development. 7. The demand side of policy research needs to be deepened. For instance, an exchange of visits and fel- lowships between policy researchers and policy makers and other practitioners would enhance familiar- ity with each others' positions and promote less formal interactions. Encouraging policy dialogue and debates and toning down identification of think tanks with narrow interest groups or specific political parties could contribute to deepening demand. 8. In general, in the current situation, there is no reliable domestic resource base which can ensure the sus- tainability of policy think tanks in Africa. Dependence on donor resources is inevitable for some time. The challenge is to design mechanisms to ensure that the activities of policy think tanks are insulated from frequent donor policy changes and agendas. Source: Ndulu (2001); Wangwe (2003). Education: Key challenges facing Tanzania in this domain include the following: · Sustain and improve the quality of education as enrollments increase by recruiting teachers, constructing classrooms, increasing preservice teacher training, and providing subsidies for purchase of teaching and learning materials. · Ramp up secondary education, including improving its quality and relevance to needs of the economy. · Strengthen the governance and administration of the country's three public universities, in terms of financial sustainability, up-to-date content, and teacher training. · Use the potential of distance education to expand access to education services while at the same time improving equity. The Open University of Tanzania offers degree programs by cor- respondence and through regional centers. The costs are also low, as the state covers tuition. The university still suffers from low enrollments (about 10,000 students), partly because of lack of content and partnerships with international academic institutions that could provide online degree programs. Combining distance education modalities with extended face-to-face interactions at Tanzania's other public universities may be one way to boost enrollments and increase access to higher education. · Reform teaching methods and the curriculum at all levels to include skills and competencies for the KE (including communication skills, problem-solving skills, creativity, and teamwork). · Harmonize technical education offered in secondary schools with that offered in the current technical colleges, and then link up with the proposed Zonal and Regional colleges and insti- tutes. These institutes and colleges should offer differentiated products to meet the needs of the whole variety of enterprises in Tanzania, which includes mining, fisheries, major cash and food crops, external trade, and metal industries. Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 45 · Increase the interface between industry and education and offer differentiated curricula that better meet the new skills required by changing markets and technologies. · Devise strategies to proactively deal with problems of skills loss through brain drain. · Meet the challenge of HIV/AIDS, which poses substantial risks for human resource develop- ment. ICT: In order to strengthen its information infrastructure, Tanzania should take the following steps (or strengthen existing measures): · Attract investment for the telecom area from both local and foreign entities. · Review and modernize telecom policies and regulations to generate fair competition and reduce high communication and operational costs. · Finalize and adopt the proposed Electronic Communications Bill, which is key to defining the ground rules for sector development (including in rural areas). · Implement the new converged licensing framework, which will ensure further liberalization of the market. · Build capacity to undertake such reforms, including through the establishment of systems and processes to review the performance of the regulatory institutions. For example, given the great demands and expectations placed on the national regulator (TCRA) by telecom sector re- forms, the Swedish government through the Swedish international Development Cooperation Agency (SIDA) is assisting TCRA in creating capacity to meet its existing and future chal- lenges. Also, TCRA is also learning from SIDA's experiences in operating in a more competi- tive market. · Support the development of the rural telecommunications infrastructure, for example, by uni- versal access schemes. Rural areas mostly lack telecom services or provide only limited access in areas adjacent to main towns and on major trunk roads. Local communications content should also be developed (in Kiswahili). · Enhance technical and business-related skills development among the population using ICT and technical institutes and vocational centers. For example, the UDSM is offering IT training in its computer center. · Continue to use global experiences in the telecom sector. In many areas of telecom reform, Tanzania has benefited by adopting best practices from both developed and developing coun- tries. The functions and roles of the national regulator (TCRA) is the best example in this case. Further benefits from global experience and best practices will depend on the capacity of TCRA and other institutions to learn from the experiences of other countries. Innovation: In order to encourage innovation, it is important for Tanzania to take the following steps: · Improve the regulatory environment, which currently includes a multiplicity of taxes and con- flicting laws that impede innovations in the business sector. · Devote adequate financial resources to encouraging innovation, as through fiscal incentives such as research grants, tax incentives, and venture capital. · Strengthen institutions like the COSTECH, CDTT, VETA, and the Institute of Production and Innovations (IPI) of the UDSM. These institutions have been established to support R&D and innovation activities in Tanzania, fund investors' discoveries and innovations, and organize forums and technological exhibitions. However, many of these entities are weakly linked to the productive sector of Tanzania, mainly because of the lack of financial and other technical resources needed to reinforce their capacities. 46 Anuja Utz · Develop adequate technical support, in the form of standards and technical assistance, for both private and public sector productive and innovative activities, and create a conducive envi- ronment for business start-ups. For example, the banking system should provide loans for new business start-ups. · Promote entrepreneurship and support technical and business-related skills development by expanding opportunities for technical and managerial training. Such promotion includes strengthening primary, secondary, vocational (polytechnics), and tertiary education. In particu- lar, science and technology at the university and polytechnic level should be strengthened in order to improve the quality and quantity of human resources. · Increase sharing of information among innovative industries in both local and foreign markets. Existing industrial associations like the Tanzania Chamber of Commerce and Industry lack re- sources to collect and disseminate industry news on R&D innovations, quality standards, prices, and so forth. There are only ad hoc linkages between local and foreign industrial asso- ciations and government institutions such as the Board of external trade and embassies. Such weaknesses limit innovation in Tanzania. · Tap into the growing stock of global knowledge and attract more and diversified sources of FDI, which is currently mainly concentrated in the mining sector. · Disseminate lessons of local innovation efforts and continue promoting indigenous knowledge initiatives.26 · Enhance the transfers of knowledge and technology, not only in Tanzania, but also in sub- Saharan Africa.27 · Conduct comprehensive, nationwide innovation and R&D surveys to establish concrete factors that either facilitate or hinder innovative activities. The outcome of the surveys can help put in place concrete innovation policies and strategies. These conclusions provide a starting point for further conversation with Tanzanian stakeholders to identify policies that can help Tanzania make more effective use of knowledge for its overall eco- nomic and social development. Through such policies, Tanzania can improve its competitiveness and growth prospects. 26For more information on indigenous knowledge initiatives, go to: http://www.worldbank.org/afr/ik/what.htm. This site offers three case studies on the themes of agriculture, education and health, nutrition, and population: "Rural Seed Fairs Southern Tanzania--Why southern zone rural seed fairs?"; "Communicating Local Farming Knowledge"; and "Traditional Medicine in Tanga Today--The Ancient and Modern Worlds Meet." 27As mentioned earlier, the Nelson Mandela Foundation for Knowledge Building and Advancement of Science and Technology in Sub-Saharan Africa is playing a leadership and catalytic role in establishing the African Insti- tute of Science and Technology (AIST) to promote research in Africa. The Foundation will have scientific cen- ters of excellence across the four regions of the continent. For more information, go to: http://www.nmiscience.org/aist.html. Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 47 Annex 1: Tanzania's Knowledge Economy Scorecards The following two scorecards show the performance of Tanzania between 1995 and the most recent period for which data are available on the four pillars of the knowledge economy (KE), as well as on two variables relating to performance: GDP growth and the human development index (HDI). Figure A.1a compares Tanzania with the world. On the four pillars, the scorecard reveals the same informa- tion as in Figures 3 and 4 above. Figure A.1: Tanzania's Knowledge Economy Scorecards a. Compared to the World b. Compared to the Africa Region Source: Knowledge Assessment Methodology 2004 (KAM: www.worldbank.org/kam). Note that discontinuous lines indicate unavailable data. Note: Figure A.1a shows the relative performance of Tanzania as compared to 128 countries that have been in- cluded in the KAM. Figure A.1b shows the relative performance of Tanzania as compared to 25 countries that represent the Africa Region in the KAM. When compared to the Africa region, a more positive picture begins to emerge for Tanzania. Not sur- prisingly, Figure A.1b shows that that Tanzania has improved on its economic and institutional per- formance; on its information infrastructure (especially telephones, both fixed and mobile), in line with the strong uptake and demand for mobile telephony that is spreading through the African continent; and on its Internet usage. 48 Anuja Utz Annex 2: Costs of Doing Business in Tanzania Indicator Tanzania Malaysia South Thailand Botswana Kenya Uganda Africa Ease of doing business (rank) 140 21 28 20 40 68 72 Starting a business Procedures (number) 13 9 9 8 11 13 17 Time (days) 35 30 38 33 108 54 36 Cost ( percent of income 161.3 20.9 8.6 6.1 10.9 48.2 117.8 per capita) Minimum capital (percent 6 0 0 0 0 0 0 of income per capita) Dealing with licenses Procedures (number) 26 25 18 9 42 11 19 Time (days) 313 226 176 147 160 170 155 Cost (percent of income per capita) 4110.2 82.7 38 17.3 298.8 40 861.8 Hiring and firing of workers Difficulty of hiring index (0­100) 67 0 56 33 11 33 0 Rigidity of hours index (0­100) 80 20 40 20 40 20 20 Difficulty of firing index (0­100) 60 10 60 0 40 30 20 Rigidity of employment index (0­100) 69 10 52 18 30 28 13 Hiring cost (percent of salary) 16 13 3 5 0 5 10 Firing cost (percent of salary) 38 65 38 47 19 47 12 Registering property Procedures (number) 12 4 6 2 6 8 8 Time (days) 61 143 23 2 69 73 48 Cost (percent of property value) 12.2 2.3 11 6.3 5.1 4.1 5.1 Getting credit Strength of legal rights index (0­100) 5 8 5 5 9 8 5 Depth of credit informa- tion index (0­6) 0 6 5 4 5 5 0 Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 49 Indicator Tanzania Malaysia South Thailand Botswana Kenya Uganda Africa Public registry coverage (percent of adults) 0 33.7 0 0 0 0 0 Private bureau coverage (percent of adults) 0 ... 63.4 18.4 30.8 0.1 0 Protecting investors Extent of disclosure index (0­10) 3 10 8 10 8 4 7 Extent of director liability index (0­10) 3 9 8 2 2 2 5 Ease of shareholder suits index (0­10) 0 7 8 6 3 10 4 Strength of investor pro- tection index (0­10) 2 8.7 8 6 4.3 5.3 5.3 Paying taxes Payments (number) 48 28 32 44 24 17 31 Time (hours per year) 248 .. 350 52 140 372 237 Total tax payable (percent of gross profit) 51.3 11.6 43.8 29.2 52.9 68.2 42.9 Trading across borders Documents for export (number) 7 6 5 9 6 8 13 Signatures for export (number) 10 3 7 10 7 15 18 Time for export (days) 30 20 31 23 37 45 58 Documents for import (number) 13 12 9 14 9 13 17 Signatures for import (number) 16 5 9 10 10 20 27 Time for import (days) 51 22 34 25 42 62 73 Enforcing contracts Procedures (number) 21 31 26 26 26 25 15 Time (days) 242 300 277 390 154 360 209 Cost (percent of debt) 35.3 20.2 11.5 13.4 24.8 41.3 22.3 Closing a business Time (years) 3 2 2 3 2 5 2 Cost (percent of estate) 22 15 18 36 15 22 30 Recovery rate (cent on the dollar) 22.3 38.3 33.9 43.9 54.4 15 39.8 Source: World Bank (2006). 50 Anuja Utz Annex 3: International Assessments As countries move into the knowledge economy, policy makers are increasingly concerned about the role knowledge and skills can play in enhancing productivity growth and innovation. Performing suc- cessfully in the knowledge economy increasingly requires new skills. These include technical skills such as literacy, foreign languages, mathematics, science, problem solving, and analysis; interpersonal skills such as teamwork, leadership, and communication; and the ability to learn independently. The following annex reviews three tools for assessing knowledge and skills: TIMSS, and two international surveys on functional skills pioneered by the OECD, PISA and IALS. Trends in International Mathematics and Science Study (TIMSS) TIMSS is conducted every four years at the fourth- and eighth-grade levels. The test results give countries an opportunity to examine the effectiveness of their educational policies and practices in light of achievement worldwide. Forty-nine countries participated in the 2003 TIMSS (TIMES 2003); previous assessments were undertaken in 1995 and 1999. Major find- ings from the 2003 TIMSS include the following: · In mathematics, Asian countries outperformed other participants and Singapore was the top per- forming country at both the fourth- and eighth-grade levels. At the eighth grade, other top per- formers included the Republic of Korea, and Hong Kong and Taiwan (China). At the fourth grade, the highest achievers were Hong Kong (China), Japan, and Taiwan (China). · In science, at the eighth grade, Singapore and Taiwan (China) had the highest performance, and Korea and Hong Kong (China) also did very well. At the fourth grade, Singapore again had the best academic achievement, followed by Taiwan (China), Japan, Hong Kong (China), and the United Kingdom. · The contexts for learning math and science are important determinants of performance. For exam- ple, the home context can help foster higher achievement, particularly when parents are highly educated, when the language of instruction is spoken at home, when there are many books in the home, and when a computer is frequently used. · Providing students the opportunity to learn the content assessed is fundamental. The content needs to be delivered in the classroom and in an effective way. · Students with higher achievement attended schools with environments that fostered positive cli- mates for learning; such a climate had fewer students from disadvantaged homes and allowed teachers and students to feel safe. Investing in sound comparative information for education is worthwhile, and countries such as India can profit from participating in TIMSS in the following ways: · TIMSS helps to put education in the public eye by creating newsworthy stories that benchmark countries' performance. · TIMSS focuses attention on strengths and weaknesses and highlights areas where action is needed. Many countries have used previous TIMSS studies to kick-start debate on school achievement and develop reform initiatives. For example, South Africa used the previous study to probe reasons for its low performance and to tease out the interaction between home language and language of in- struction. Jordan has highlighted the importance of teacher attitude and motivation, and so on. Canada and Australia have also made changes based on TIMSS results. · TIMSS helps to improve the provision of education by highlighting it and furthering understand- ing of key aspects of teaching and learning. Policy makers can make better decisions and allocate resources more effectively. Teachers likewise gain through a deeper understanding of how differ- Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 51 ent students learn and what inhibits their learning. They can, moreover, benefit from new instruc- tional materials and techniques that aim to maximizing learning. OECD's Programme for International Student Assessment (PISA) PISA attempts to measure student achievement in reading, mathematical, and scientific literacy--that is, not merely mastery of the school curriculum but the knowledge and skills needed for full participa- tion in society. By assessing 15-year-old children who are near the end of their compulsory education, PISA provides an indication of the overall performance of school systems. PISA thus enables coun- tries to monitor regularly and predictably their progress in meeting key learning objectives. PISA 2003 is the second assessment (PISA 2003); the first survey was in 2000. All 30 OECD coun- tries participated in the 2003 PISA, as did 11 partner countries and regions including Brazil, Hong Kong (China), Indonesia, Latvia, Liechtenstein, Macao (China), Russia, Serbia and Montenegro, Thai- land, Tunisia, and Uruguay. The focus of the 2003 PISA assessment was on mathematics. Students were divided into six proficiency levels: the small minority who could perform the most complex tasks was ranked at Level 6; those who could only perform very simple tasks were at Level 1. Students un- able to complete these tasks were "below Level 1." Findings of the assessment included the following: · Only 4 percent of students in the combined OECD area, but more than 8 percent in Belgium, Ja- pan, Korea, and Hong Kong (China), can perform the highly complex tasks required to reach Level 6. · About a third of OECD students can perform relatively difficult tasks at Levels 4, 5, or 6 but over 49 percent of students in Finland, Korea, and Hong Kong (China) can perform at least at Level 4. · About three-quarters of OECD students can perform mathematical tasks at Level 2. · Schools with a positive climate, effective policies and practices, and sufficient resources perform better. Other key features of PISA include the following: · PISA has an innovative definition of "literacy," which is the capacity of students to apply knowl- edge and skills in key subject areas and to analyze, reason, and communicate effectively as they pose, solve, and interpret problems in a variety of situations. · PISA assessment allows countries to monitor regularly their progress in meeting key learning ob- jectives. · PISA measures students' performance in the context of their backgrounds and schools, which al- lows exploration of some of the main features associated with educational success. · PISA has great geographic coverage that includes the 49 countries that have participated already in a PISA assessment and the 11 additional countries that will join the 2006 PISA assessment. These countries comprise one-third of the world population and almost nine-tenths of the world's GDP. · PISA has relevance to lifelong learning; it not only assesses students' curricular and cross- curricular competencies but also asks them to report on their motivation to learn, their beliefs about themselves, and their learning strategies. The International Adult Literacy Survey (IALS) Globalization, technological change, and organizational development are shaping the supply of and increasing the demand for better literacy skills. Literacy in the Information Age, the report from the International Adult Literacy Survey (IALS) (OECD 2000), presents evidence on the nature and magnitude of the literacy gaps faced by 20 OECD countries. Main results include the following: 52 Anuja Utz · In 1428 out of 20 countries, at least 15 percent of adults have only the most rudimentary literacy skills, making it difficult for them to cope with the rising skill demands of the information age. · In six countries less than 15 percent of adults reach the lowest level of literacy skills (Denmark, Finland, Germany, Netherlands, Norway, and Sweden), but even in the country with the highest score on the test (Sweden), 8 percent of the adult population encounters a severe literacy deficit in everyday life and at work. · Low skills are found not just among marginalized groups but in significant proportions of the adult populations in all countries surveyed. Hence, even the most economically advanced societies have a literacy skills deficit. IALS findings can provide insights for policy makers addressing lifelong learning, society, and the labor market. In addition: · IALS offers insights into the factors that influence the development of adult skills in various set- tings, at home and at work. The survey confirms the importance of skills for the effective function- ing of labor markets and for the economic success. · In terms of literacy skills and features of the labor market, in all countries higher levels of literacy skills in the workforce are associated with larger proportions of knowledge jobs in the economy. Literacy skills increase the probability of being in a white-collar, high-skilled position and reduce the chance of being unemployed. In most OECD countries, low skills are associated with a higher incidence of long-term as opposed to short-term unemployment. The benefits accruing to im- proved literacy skills are much higher for workers with tertiary education than for those with sec- ondary education. · As regards literacy, earnings, and wage differentials, educational attainment is the most important determinant of earnings in most countries. There are, however, large differences between countries in how much their labor markets reward education and how much they pay for skills and experi- ence. Labor market rewards associated with education, skills, and experience are amplified or at- tenuated by the relative conditions of supply and demand. 28The 14 countries are Australia, Belgium (Flanders), Canada, Chile, the Czech Republic, Hungary, Ireland, New Zealand, Poland, Portugal, Slovenia, Switzerland, the United Kingdom, and the United States. Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 53 References and Bibliography Accenture, Markle Foundation, and UNDP. 2001. "Creating a Development Dynamic: Final Report of the Digital Opportunity Initiative." Available at: http://www.opt-init.org/framework/pages/contents.html. Aderinokun, Kunle. 2005. "Africans in Diaspora Set Up N600bn Institute." This Day, January 10. La- gos. Aduda, David. 2005. "Top Kenyan Universities Lag Behind Neighbours." The Daily Nation, October 9. Nairobi. Afrobarometer (Michael Bratton, Carolyn Logan, Wonbin Cho, and Paloma Bauer, compilers). 2004. "Afrobarometer Round 2: Compendium of Comparative Results from a 15-Country Survey." Working Paper No. 34. March Available at: http://www.afrobarometer.org/. Balile, Deodatus. 2003. "Tanzania Pledges Support for Science Training." Science and Development Network (SciDev.net). June 19. Available at: http://www.scidev.net/dossiers/index.cfm?fuseaction=dossierReadItem&type=1&itemid=871 &language=1&dossier=10. Bell, Clive, Hans Gersbach, and Shanta Devarajan. 2003. "The Long-Run Economic Costs of AIDS: Theory and an Application to South Africa." Washington, D.C.: The World Bank. Bloom, David, David Canning, and Kevin Chan. 2005. "Higher Education and Economic Develop- ment in Africa." Unpublished manuscript. September 20. Cambridge: Harvard University. Chandra, Vandana, and Erik Tallroth. 2004."Understanding the `How To' of Technological Change For Growth." Draft Concept Note, PREMEP-ESSD Partnership. April 19. Washington, D.C.: The World Bank. Cohen, Daniel, and Marcelo Soto. 2001. "Growth and Human Capital: Good Data, Good Results." Technical Papers No. 179. September. Washington, D.C.: OECD Development Centre. Cooksey, Brian, Lisbeth Levey, and Daniel Mkude. 2001. Higher Education in Tanzania: A Case Study. Carnegie Corporation of New York, The Ford Foundation, the John D. and Catherine T. MacArthur Foundation, and the Rockefeller Foundation. Available at: www.foundation- partnership.org/pubs/tanzania/index.php. The Economist. 2005a. "Less is More." July 7. ------. 2005b. "Calling for an End to Poverty." July 7. ------. 2005c. "Calling Across the Divide." July 7. The Economist Intelligence Unit (EIU). 2003. "Tanzania Country Profile 2003." Available at: http://db.eiu.com/. ------. 2004a. "Business Africa." September 1. Available at: http://db.eiu.com/. 54 Anuja Utz ------. 2004b. Tanzania Country Profile 2004. Available at: http://db.eiu.com/. Economic and Social Research Foundation (ESRF). 2002. Background Note on Tanzania for the World Bank Institute. Developed for the Policy Workshop on Knowledge for Development. Dar es Salaam, Tanzania: ESRF. Estache, Antonio. 2005. "What Do We Know About Sub-Saharan Africa's Infrastructure and the Im- pact of its 1990s Reforms?" Draft, May 24. Washington, D.C.: The World Bank Fink, Carsten, Aaditya Mattoo, and Ileana Cristina Neagu. 2002. "Assessing the Role of Communica- tions Costs in International Trade." Working Paper #2929. Washington, D.C.: The World Bank. Gill, Indermit, and Amit Dar. 1998. "Vocational Education and Training in Tanzania: Finance and Relevance Isues in Tansition." Country Study Summary 19781. October. Washington, D.C.: The World Bank. Goedhuys, Micheline. 2005. "Learning, Product Innovation and Firm Heterogeneity in Tanzania." Discussion paper #2005-7. October. Maastricht, The Netherlands: United Nations University- INTECH. Hancock, Simon. 2005. "Mobile Phones Boom in Tanzania." BBC Click Online, July 22. Available at: http://news.bbc.co.uk/2/hi/programmes/click_online/4706437.stm. Hazlewood, Judith G., and Srividya Prakash. 2005. "Tanzania's Health Care Crisis." The McKinsey Quarterly 3. International Finance Corporation/World Bank. 2002. "Tanzania: Foreign Direct Investment and Competition Policy: Issues and Recommendations." Foreign Investment Advisory Service. February. Washington, D.C. International Monetary Fund (IMF). 2005. "United Republic of Tanzania: Fourth Review Under the Three-Year Arrangement Under the Poverty Reduction and Growth Facility and Request for Waiver of Performance Criteria." July 18. Washington, D.C. Juma, Calestous, ed. 2005. "Going For Growth: Science, Technology and Innovation in Africa." Lon- don: The Smith Institute. Kaufmann, D., A. Kraay, and M. Mastruzzi. 2005. "Governance Matters IV: Governance Indicators for 1996­2004." Washington, D.C. The World Bank. Kopicki, Ron. 2005. "Supply Chain Development in Tanzania." Unpublished draft report for the Tanza- nia Country CEM. Washington, D.C.: The World Bank. Markus Haacker, ed. 2004. "The Macroeconomics of HIV/AIDS." Washington, D.C.: International Monetary Fund. Mwanza, Joel. 2002. "Implementation of Integrated Financial Management System in Tanzania." New York: United Nations. Presentation available at: http://unpan1.un.org/intradoc/groups/public/documents/CAFRAD/UNPAN006453.pdf. Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy 55 Ndulu, Benno J. 2001. "Human Capital Flight: Stratification, Globalization and the Challenges to Ter- tiary Education in Africa," Draft. Washington, D.C.: The World Bank. OECD. 2000. Literacy in the Information Age: Final Report of the International Adult Literacy Sur- vey. Washington, D.C.: OECD. Available at: http://www1.oecd.org/publications/e- book/8100051e.pdf. ------. 2005a. "Education at a Glance." Paris: OECD. ------. 2005b. "Science, Technology and Industry Scoreboard 2005." Paris: OECD. OECD Development Centre. 2004. African Economic Outlook 2003/2004. Paris: OECD. ------. 2005. African Economic Outlook 2004/2005. Paris: OECD. Programme for International Student Assessment (PISA). 2003. Available at: http://www.pisa.oecd.org/. Semboja, Haji Hatibu Haji, Rhoda Mbelwa, and Charles Bonaventura. 2005. "The Cut-Flower Indus- try in Tanzania." Working Paper 152. March. Geneva: International Labour Office. Available at: http://www.ilo.org/public/english/dialogue/sector/papers/tanzflow/. Soderbom, Mans, Francis Teal, Anthony Wambugu, and Godius Kahyarar. 2004. "The Dynamics of Returns to Education in Kenyan and Tanzanian Manufacturing." CSAE WPS/2003-17. Janu- ary. Centre for the Study of African Economies (CSAE), University of Oxford. Tebeje, Ainalem. 2005. "Brain Drain and Capacity Building in Africa." IDRC Reports, February 22. Available at: http://www.idrc.ca/en/ev-71249-201-1-DO_TOPIC.html. Trends in International Mathematics and Science Study. 2003. Available at: http://timss.bc.edu/timss2003.html. UNESCO Institute for Statistics. 2001. "The State of Science and Technology in the World, 1996- 1997." Montreal, Canada. United Nations Conference on Trade and Development (UNCTAD). 2005a. "The Digital Divide: ICT Development Indices 2004." New York and Geneva: UNCTAD. Available at: http://www.unctad.org/en/docs/iteipc20054_en.pdf. ------. 2005b. "Transnational Corporations and the Internationalization of R&D." In World Invest- ment Report 2005. New York and Geneva: UNCTAD. United Nations Economic Commission for Africa. 2003. "Economic Report on Africa 2003." Addis Ababa, Ethiopia. United Republic of Tanzania, 2002. "Tourism in Tanzania: Investment for Growth and Diver- sification." September. Dar es Salaam. ------. 2005. The Economic Survey 2004. June. Dar es Salaam. 56 Anuja Utz Vodafone. 2005. "Africa: The Impact of Mobile Phones." The Vodafone Policy Paper Series, Number 2, March. Available at: http://www.vodafone.com. Wangwe, Sam. 2003. "Policy Think Tanks In Africa: Experience, Lessons, Challenges and Way For- ward." Visiting Fellow. July 15. Washington, D.C.: The World Bank. World Bank. Knowledge Assessment Methodology. Available at: http://info.worldbank.org/etools/kam2005/. ------. 1997. Tanzania Agricultural Research Project--Volume 1, Project Appraisal Document. No- vember 25. Washington, D.C. ------. 2004a. African Development Indicators 2004. Washington, D.C. ------. 2004b. "Doing Business: Removing Obstacles to Growth." Washington, D.C. ------. 2004c. "Investment Climate Assessment for Tanzania." Washington, D.C. ------. 2004d. Local Pathways to Global Development: Marking Five Years of the World Bank In- digenous Knowledge for Development Program. Washington, D.C.: The World Bank. ------. 2004e. "Tanzania: Secondary Education Development Program (SEDP)." Report No. 27631. May 14. Washington, D.C. ------. 2005a. African Development Indicators 2005. Washington, D.C. ------. 2005b. "ICR on Primary Education Development Program in Tanzania." May 17. Washing- ton, D.C. ------. 2005c. "Primary Education Development Program in Tanzania, ICR." May 17. Washington, D.C. ------. 2006. Doing Business in 2006: Creating Jobs. Washington, D.C. World Economic Forum. 2004. Africa Competitiveness Report 2004. Geneva. Available at: http://www.weforum.org/. ------. 2005. Global Competitiveness Report 2004­2005. Geneva. Available at: http://www.weforum.org/. ------. 2006 Global Competitiveness Report 2005­2006. Geneva. Available at: http://www.weforum.org/. ImprovIng CompetItIveness The Role of Information In tanzanIa and Communication Technologies www.infoDev.org An infoDev Study IMPROVING COMPETITIVENESS IN TANZANIA October 2005 The Role of Information and Communication Technologies Draft Report The final version of the full report will be available online at www.infoDev.org. For additional information about this study, please contact Seth Ayers, infoDev, email: sayers@worldbank.org or Tel: 1.202.473.4868. This report was prepared by OTF Group, Inc., a consulting firm that focuses on Country Competitiveness projects in emerging economies around the world. The project team was composed of Jason Bauer, Vicky Obst, Camila Rodriguez and Marcela Escobari. OTF Group was assisted by Leopold Rweyemamu of DataCom Africa. The content of this document does not necessarily represent the views of infoDev, and is the sole responsibility of the authors. TABLE OF CONTENTS 1 EXECUTIVE SUMMARY 1 2 TANZANIA TODAY 7 2.1 A Starting Point 7 2.2 GDP Analysis 7 2.3 Sector Economic Performance and Trade Statistics 8 2.3.1 Methodology and Theory 10 2.3.2 Broad Economic Performance and Trade Statistics 11 2.3.3 Regional Trade Statistics 11 3 TANZANIA'S TRANSITION TO ECONOMIC GROWTH AND A COMPETITIVE ECONOMY 15 3.1 Prosperity: Governments Enable It; Firms Create It 16 3.2 The Evolution from Basic o Complex Products 16 3.2.1 The Role of SMEs 18 3.2.2 Competitiveness and Poverty Alleviation 18 3.3 Creating a Competitiveness Mindset 19 4 ICT AND BUILDING COMPETITIVENESS 21 4.1 Access, Ability and Benefit Model 21 4.2 Access: Mixed Results 23 4.2.1 Limited Access to Telecom and Internet Services 23 4.2.2 Infrastructure Constraints 26 4.2.3 The Urban/Rural Digital Divide 28 4.3 Low Level of Ability 30 4.3.1 Basic Educational Deficiencies 30 4.3.2 Incorporating ICT into the Education System 31 4.3.3 Tertiary Enrollment 31 4.3.4 Training: Not Yet Driven By Private Sector Needs 32 i 5 MIGRATING TO MORE SOPHISTICATED PRODUCTS AND SERVICES 33 5.1 Two Paths: Key Sectors and SMEs 33 5.2 Agribusiness 34 5.2.1 Coffee 35 5.2.2 Cashews 36 5.2.3 Targeted ICT Interventions 37 5.3 Minerals and Mining 39 5.3.1 Gemstone Industry Outlook 40 5.3.2 Gemstone Lapidary 41 5.3.3 Targeted ICT Interventions 42 5.4 Services 43 5.4.1 Tourism 43 5.4.2 ICT in Banking and e-Commerce 49 5.4.3 Future Development of Services Sector: Business Process Outsourcing 50 5.5 Building Competitive SMEs 51 5.5.1 Lack of SME Adoption 51 5.5.2 Potential for SMEs to Build Competitive Advantage Through ICT 54 5.6 Strengthening the Private Sector 56 5.6.1 Incubation 56 5.6.2 Business Development Services 56 5.6.3 An ICT Alliance 57 6 THE ROLE OF GOVERNMENT: BUILDING AND CAPITALIZING ON MOMENTUM 61 6.1 National ICT Policy 61 6.2 Regulatory Regime Reform 62 6.3 E-government 64 7 CONCLUSION 68 ii . Contents 8 RECOMMENDATIONS 69 8.1 Develop and Execute Competitive Strategies in Key Clusters 69 8.2 Facilitate Access 70 8.3 Improve Ability 72 8.4 Strengthen SMEs through Incubation and BDS 73 8.5 Create an ICT Alliance 74 8.6 Upgrade Public Sector and Enabling Environment 75 9 APPENDICES 77 9.1 Economic and Trade Statistics 77 9.2 ICT Assessment-2003 ITU Indicators 83 10 BIBLIOGRAPHY 88 iii BOXES Box 1 The University Computing Centre Ltd.: "One Student, One Computer" 32 Box 2 AKSCG­KILICAFE 36 Box 3 Olam International 37 Box 4 Embedding Technology in a Coffee Bean 38 Box 5 The Gemological & Jewelry Vocational Training Centre 41 Box 6 Ebene Cyber City­Mauritius 51 Box 7 Tanzania Development Gateway (TDG) SME Website Pilot Project 54 Box 8 Improving Logistics & Making Distance Irrelevant: Issam International Ltd. 55 Box 9 The Parliamentary On-Line Information System 65 Box 10 GOT Increases Efficiencies in Human Resource Management 65 FIGURES Figure 1 GDP per Capita 8 Figure 2 GDP by Sector (Mainland) 9 Figure 3 GDP by Sector (Zanzibar) 9 Figure 4 Trade Statistics Framework 10 Figure 5 Tanzania Trade Statistics by Broad Cluster 12 Figure 6 The Evolution of a Developing Economy 15 Figure 7 Declining Basic Commodity Prices, 1845­1999 17 Figure 8 Tanzania Exports 17 Figure 9 The Virtuous Cycle 17 Figure 10 Labor Force vs. GDP Composition 2001 (%) 19 Figure 11 Keys to ICT Usage 21 Figure 12 Five Levels of ICT Usage 22 Figure 13 Teledensity of Fixed vs. Mobile Telephony (1997­2002) 24 Figure 14 Compound Annual Growth Rate of Fixed vs. Mobile Telephony (1997­2002) 24 Figure 15 Internet Users per 100 Inhabitants (1997­2002) 25 Figure 16 Intra-Regional and US Rates 26 Figure 17 Dial-up and High Speed Internet Monthly Costs 26 Figure 18 The EASSY Project 27 Figure 19 National Fiber and Microwave Networks 28 Figure 20 Frequency of ICT Usage within Rural Areas 29 iv . Contents Figure 21 Perceived Impact of Mobile Telephony on Financial Capital among Rural Populations 29 Figure 22 School Enrollment Indicators 2002 30 Figure 23 World Coffee Price Index 35 Figure 24 Leading Cashew Producing Countries 37 Figure 25 Tanzanite Value Chain and Potential ICT Improvements 42 Figure 26 Customer Portrait(TM) of the Traveler to East Africa 44 Figure 27 Inbound Tourism Arrivals 44 Figure 28 Average Length of Stay 45 Figure 29 In-country Tourism Expenditure 45 Figure 30 Spend per Day 46 Figure 31 Tourist Experience and Potential ICT Improvements 47 Figure 32 Outbound Calls to the U.S. 51 Figure 33 Means of Communication Utilized by SMEs 52 Figure 34 Annual Total Investment in ICT 52 Figure 35 Computer Usage 53 Figure 36 BDS Services 57 Figure 37 TCRA Converged Licensing Framework 63 Figure 38 E-Government Stage vs. Capacity, Selected African Countries 2001 66 Figure 39 Tanzania Top 5 Exports by Country Export Value 79 Figure 40 Tanzania Exports by Country and World Share, 2003 79 Figure 41 Distribution of Tanzanian Imports 80 Figure 42 Tanzania's Trade Balance 81 Figure 43 Tanzania's Current Stage and Broad Cluster State 82 TABLES Table 1 Tanzania: Destination of Exports and Imports, 2000­2003 (% of total exports) 13 Table 2 Estimated Backbone Investment Costs 28 Table 3 Annual Private Rates of Return to Education and Training 32 Table 4 Selected Exports Compound Annual Growth 1998­2003 34 Table 5 Minerals Production 40 Table 6 Costs of Website Design and Hosting 50 v Table 7 ICT Alliance Structure 58 Table 8 Productive Sector ICT Strategies 62 Table 9 Tanzania Exports 77 Table 10 Marketed Production of Major Export Commodities (Zanzibar) 77 Table 11 Inbound Tourism Arrivals 77 Table 12 Average Length of Stay 78 Table 13 In-country Tourism Expenditure 78 Table 14 Spend per Day 78 Table 15 Top 50 Tanzanian Industries (1­50) by Export Value, 2003 83 Table 16 Main Telephone Lines 85 Table 17 Local Telephone Network 2002 85 Table 18 Teleaccessibility 2002 86 Table 19 Telephone Tariffs 2002 86 Table 20 Mobile Cellular Subscribers 2002 87 Table 21 Information Technology 2002 87 vi . Contents EXECUTIVE among other benefits. But this tool will only be use- ful to firms and industries committed to pursuing SUMMARY innovative, competitive strategies. In order for ICT to contribute to the economy's expansion, this study's fundamental premise is that Tanzania's lead- ers must also commit to building a stronger founda- tion for economic competitiveness. Introduction Driving along the stretch of road that connects the The study begins by dissecting Tanzania's current eco- city of Arusha to the coffee growers and association nomic situation and exploring the barriers to growth offices in Moshi, Tanzania's contrasts are visibly and competitiveness, followed by a broader discussion striking. Looking out at the lush lowlands against of growth and competitiveness as models for revealing the backdrop of the mountainous Kilimanjaro a nation's unique development path. Applying this region, the magnificence of the country's ecological model to Tanzania, it focuses on a direction for the resources is overwhelming. This vast area is home country, and finally, how this direction can inform to a wealth of mineral resources, including newly future action. A perspective on the most urgent choic- exploited gold deposits and precious stones such as es facing Tanzania today is captured in a recommen- the rare Tanzanite gem. The region's fertile soil pro- dations section at the end of the study. duces a rich Arabica coffee sold in specialty stores in the US and Europe. But a closer look reveals a Tanzanian Growth and dilapidated and poorly maintained infrastructure. Competitiveness Farmers in the field use manual hoes or livestock to To ensure that future generations enjoy greater tend to their crops. It is a stark contrast, one that opportunities and increased prosperity, Tanzania captures the country's unrealized potential. must transform the way it competes. Tanzania's his- toric model has been to compete on its abundant Tanzania faces enormous challenges in reaching this basic natural resources or factor inputs and cheap potential, even relative to its African neighbors. labor. This pattern is both clear and understand- GDP per capita in 2004 was just US$ 322, 28% able. Its agricultural sector has benefited from lower than the average for low income countries.1 favorable climatic conditions and rich soil. Its Already one of the poorest countries in the world, tourist industry has capitalized on such natural Tanzania experienced negligible growth over the last endowments as the Serengeti and Mount decade. GDP per capita compound annual growth Kilimanjaro. Its mining sector has extracted pre- from 1990 to 2003 was just 1.0%.2 Dramatic cious minerals and gemstones, exporting large export growth in mining and minerals, driven by amounts of Tanzania's natural wealth in its raw market forces outside Tanzania's control, masks dis- form, without capturing much of the value. appointing results in the agricultural sector, which contributes much higher proportions of GDP and is Countries following such a model are competing on responsible for the livelihood of more than 80% of comparative advantage. Comparative advantage in the population. Significant economic growth is nec- the age of globalization and innovation is no longer essary to improve the wealth of the average an effective platform for increasing prosperity. In Tanzanian, and yet this goal remains elusive. today's global economy, these types of resources are more plentiful or less important than in the past, The aim of this study is to define the appropriate reducing their value and undermining the wealth of role of Information and Communications societies built on comparative advantage economies. Technologies (ICT) in elevating the growth and competitiveness of Tanzania, and to identify actions It is clear that this model is no longer working for that will facilitate this role. Used effectively, ICT Tanzania. Despite favorable conditions, most of the can be a powerful economic tool, helping firms and population is engaged in subsistence farming on industries reach new markets, reduce communication small plots, and the country's few cash crops such and coordination costs, increase transparency-- as coffee and cashews have lost significant value 1 World Development indicators, World Bank 2004 2 Ibid Executive Summary . 1 over the last decade. As commodity prices have fallen, In targeting key sector initiatives, Tanzania's public most producers have been caught in a cycle of and private sector leaders must choose industries declining quality and quantity. Exports in cashews that have the revenue and employment potential have fallen by 50% CAGR in the last five years, to help the country reach its national objectives. and coffee exports fell by 8% over the same period. Ideally, the data and analysis in this report­which Tanzanian tourism offerings have also been com- focuses on the coffee, cashews, minerals and moditized, easily replaced by regional competition. tourism sectors­can be used as a foundation to Visitor arrival numbers and spend-per-day are formalize support for priority sectors and test falling, even as these indicators improve in the rest hypotheses about how to upgrade these industries. of the region. Similar commodity dynamics in the mining industry, where most of the value is accrued All four of these sectors show exciting growth poten- outside the country, confirm this trend: competing tial. In agribusinesses, the economy's largest sector, exclusively on comparative advantage will ensure the government of Tanzania (GOT) has already the poverty of the average Tanzanian. recognized the need to move toward value added products and specialty markets, but lacks informed The most successful countries in today's global national strategies for the major agricultural sub sec- economy invest in developing competitive tors. Specialty coffee exports by the growers' associa- advantage, in creating wealth by exporting complex tion, KILICAFE, have proven that coffee can be products and services created by highly skilled peo- upgraded to reap premium prices.3 The opportunity ple. A nation's ability to build and sustain these is similar in cashews, where currently only 10% of advantages­a nation's competitiveness­is a key driver Tanzania's production is processed in country. of growth and prosperity. In order to create and Investing in vertically integrating this industry distribute new wealth, Tanzania must build sectors would create over 30,000 direct jobs, and increase and firms that can innovate in response to market the average price per ton from a current US$580 to demand, and begin the transition from a subsis- US$2,934 for processed nuts.4 tence economy to an export economy. The country must focus on creating an environment in which The mining sector is another potential economic the number of competitive firms, both export- bright spot. Production grew at an astounding com- oriented and domestic, can expand quickly. pound annual growth rate of 56.1% from 1999 to 20035, even though margins are small due to the This is not a simple task in an economy such as concentration of raw minerals. Impending regulatory Tanzania, where exports are currently only 12% of changes forcing forward integration into lapidary GDP. For Tanzania to achieve accelerated rates of and jewelry may have positive effects, but only if the economic growth, two sets of objectives must be industry is sufficiently prepared for these changes. pursued simultaneously. The first is to develop the competitiveness of Tanzania's key industries. The Finally, tourism is a key industry with unrealized second is to develop support programs and initia- value to contribute to the Tanzanian economy. tives that strengthen Tanzanian SMEs and the Tanzania has a strong consumer brand and average Tanzanian private sector overall. This will create the length of stay has grown from 7.7 days in 1999 to 11 kind of virtuous cycle the country needs to trans- days in 2003, due to the stabilization of Zanzibar.6 form its economy, a path followed by many for- If Tanzania continues to invest in upgrading its com- merly developing economies: key industries `pull' petitiveness in tourism, the industry's multiplier the rest of the economy, while business develop- effects could touch every citizen in the country. But ment services and an improved enabling environ- this process will require a shift in strategy to compete ment `push' the capacity of the private sector to in today's global tourism game--to create the kinds of respond to new opportunities. complex, unique experiences for which international 3 One of the promising signs comes from the members of the Association of Kilimanjaro Specialty Coffee Growers (AKSCG), also known under the trade name of KILICAFE. This example will be detailed in the study, but they have concentrated on increasing the quality and value of their coffee, and have sold 38% of their coffee as specialty grade. KILICAFE was also the first recipient of a direct-export license for coffee and, as such, was able to directly access premium buyers in the US and Europe. These customers are sophisticated and demanding, but are willing to pay a premium from a supplier who can con- sistently provide a very high quality coffee. 4 Ibid 5 Bank of Tanzania, Economic Bulletin 6 World Tourism Organization 2 . Improving competitiveness in Tanzania tourists will pay a premium. It will depend on deeper Connecting with Customers. Website marketing knowledge of customer preferences, an increasing of specialty coffee producers is helping to build ability to create customized offerings, and a greater strong relationships with importers and roasters investment in operational efficiency. by showing the coffee's origins and its benefit to local communities. Information, Communications and Competitiveness Increasing Efficiency. Tourism companies would As Tanzanian firms begin improving their strategies benefit greatly from using ICT to network with and operations, ICT play a vital role. Technology financial institutions and allow for credit card cannot offer a miracle cure for dormant industries transactions, a major input into tourists' choice of and businesses, but without the ability to consume hotels, tour operators and other major purchases. information and communicate--both internally and externally­Tanzania cannot continue to Creating Differentiation. In the gemstone indus- strengthen its position in the global economy. In try, investments in computer-aided manufactur- other words, in a competitiveness model of growth, ing (CAD CAM) products are being explored, investments in ICT must be guided by strategy. A which could be leveraged to create unique, dis- major part of this study, therefore, is devoted to tinctly Tanzanian designs, enhancing differentia- exploring how Tanzania can develop more competi- tion and increasing the value of jewelry exports. tive industries and firms, followed by the specific role that ICT can play in that transition. These illustrations of the potential value of ICT in improving the competitiveness of Tanzanian firms The value of ICT grows as firms adopt more competi- capture only one part of the equation, the role of tive strategies and vertically integrate into value-added firm-level strategy and operations. These upgrades offerings. As a result, ICT needs can vary significantly cannot happen, however, without an effective across firms and industries, often driven by an organi- enabling environment­one that makes ICT relevant, zation's place in the value chain. While coffee growers reliable and affordable. The GOT and multilateral might benefit from mobile communications to make donor organizations must partner with the private price comparisons, for example, exporters will need sector to create an environment where innovative e-commerce capabilities to transact with sophisticated Tanzanian businesses can use ICT as a tool for buyers abroad. The early adopters in Tanzania and the building competitiveness and capturing market experience of successfully developing economies show opportunities. The following is an overview of the a path where ICT­broadly defined--can become a report's recommendations for building this type of ubiquitous economic tool, customized to the needs environment. and sophistication of a particular user. Many SMEs, however, remain unconvinced of the value of ICT. Only 28.2% of SMEs in one survey used a computer-based medium such as email, and RECOMMENDATIONS 76% of SMEs make no annual investment in ICT.7 This is a rational trend. The benefits of SME usage 1. Strengthen SMEs through Incubation of ICT have been ambiguous, at best. But as tacti- cal support helps these firms to upgrade their oper- and Business Development Services ations and strategies, this dynamic will shift. Below (BDS) are recent examples of this trend in Tanzania, as One approach to jumpstarting the capacity and well as some promising opportunities: competitiveness of SMEs is to create additional support mechanisms such as BDS and incubation Improving Logistics. The growth of "traceability" sys- services. These services can help SMEs target better, tems in agribusiness and mining reflect a response often niche markets, improve productivity, and to consumer demand for knowing the source and serve more sophisticated customers. By providing treatment of food products, coffee and gems. hands-on technical assistance, service providers can 7 Kijo-Ringo, Natujwa Daniel, "Impact of Investment in and Utilization of Information and Communication Technologies on Market Extension: Overview of Small and Medium Enterprises in Tanzania", University of Dar es Salaam, November 2004 Executive Summary . 3 help firms work through the complex process of cross-cutting competence across ministries would understanding markets, planning for multiple sce- allow the government to be more responsive to the narios, and capturing their firm's potential worth in needs of different sectors. the specific structure and language of a business plan that can be used to facilitate financing. The recommendations also stress the need for forums that foster public-private dialogue and give One challenge for Tanzania is that access to BDS the private sector a seat at the table in policy mak- offerings has been concentrated in Dar es Salaam, ing. Issues like e-commerce and facilitating credit provided in an uncoordinated manner, and card transactions rank high among the Tanzanian focused on either very basic micro-enterprises or private sector's priorities. These needs illustrate a large companies8. The development of a national fundamental truth about successful economies: a BDS network that serves as a one-stop-shop for strong competitive environment for the private the SME sector would help to address some of sector rests on an informed and responsive public these gaps. A special focus on agribusiness firms sector. and rural regions could have a significant impact on the productivity of the largest and poorest There are some relatively quick potential wins. By sectors of the population. sharing best practices in the region and abroad, Tanzania's government could increase the speed of A greater investment in business incubators would its transition from paper to electronic systems and also help entrepreneurs to survive the risky start-up centralize back office systems. The effective provi- phase by providing a range of services, from hands- sion of citizen services will become increasingly on management/technical assistance and access to dependent on the government's ability to digitize its finance, to support services and infrastructure such processes. as office space and communication facilities. These incubators would also serve as powerful, high-profile 3. Create an ICT Alliance mechanisms for supporting technology-based firms The ICT sector itself is nascent, but it will grow as and nurturing innovation and entrepreneurship, domestic firms pursuing competitive strategies gen- both urgently needed in Tanzania. erate local demand for ICT solutions. This will shift the focus of the sector away from large government 2. Increase the Efficiency and or donor contracts and towards providing cus- Responsiveness of the Public Sector tomized solutions to innovative Tanzanian firms. Creating a more efficient, more responsive public One way to encourage this transition is to create an sector is another important aspect of building an "ICT Alliance" between ICT service providers and enabling environment. While the onus of the buyers in order to communicate more effectively and responsibility falls on the private sector to make promote `solution-based selling', where providers their businesses more successful, governments and understand their customers' problems at an intimate their donor partners can help limit or eliminate level. These working groups allow ICT providers barriers to the effective use of ICT (such as high (hardware, software and training), as well as govern- telecomm prices and import tariffs on ICT goods), ment as the regulator, to engage ICT stakeholders and can use ICT themselves to facilitate transactions in a structured manner. As part of the Alliance (e-procurement, customs and ports' logistic, etc). Partnership, private sector advocacy and awareness can also be elevated by funding workshops in which ICT coordination within the government needs to SMEs explore ICT solutions. be reinforced to exploit potential synergies and make ICT policy more focused and relevant to the Like any sector poised for growth, the ICT sector rest of the economy. As a tool for strengthening the will require investment capital. This can be a major economy as a whole, ICT should be represented stumbling block for local firms as they can be across ministries. Moving ICT policy from the viewed by financial institutions as riskier invest- Ministry of Communications and Transport to a ments than more established industries. The forums 8 Stevenson & Annette St-Onge. "Support for Growth-oriented Women Entrepreneurs in Tanzania". International Labor Organization. 2005. http://www.afdb.org/pls/portal/docs/PAGE/ADB_ADMIN_PG/DOCUMENTS/PRIVATE_SECTOR_OPERATIONS/ILO-TANZANIA_19.01.2005.PDF 4 . Improving competitiveness in Tanzania created by the Alliance, could also help to bring current education levels prevent them from engag- ICT firms, donors, government and financial insti- ing in more complex transactions. Less than 1% of tutions together to discuss the unique challenges the population attends tertiary education, mostly that ICT firms face. Incubation of SMEs in the due to the low returns from such investment.11 The ICT sector will help with the specific training needs government needs to address this reality, and begin of this group such as basic writing skills in English, to incorporate ICT into its education objectives. customer service skills and training of trainers. Secondary schools present a good opportunity to introduce students to the value of ICT. 4. Build the Technical Platform for Growth Improving ability in the private sector requires a different approach, and should be focused on A critical part of developing Tanzania's global com- enhancing training in applied ICT business skills. petitiveness will be building the actual platform One mechanism would be to strengthen the con- for that growth. The limited use of ICT among nection between academia and the private sector. Tanzanians today reveals the enormity of this barri- However, the first step must be to conduct a er. Less than 2% of the population has access to needs assessment of the private sector, particularly mobile telephony, measured by handset ownership, SMEs, to understand the specific needs of and fixed lines are used by a dismal 1%.9 The Tanzanian businesses that must inform the design country's computing story is similar. Tanzania has of appropriate ICT training programs. These one of the lowest usage rates in the region in terms trainings could ultimately be implemented by of Internet hosts and computers. Internet density is academia, the private sector and/or members of the particularly low at 0.23 users per 100 inhabitants.10 ICT Alliance. The broadening of access to ICT can be accom- plished by improving international connectivity, building out the national backbone infrastructure 6. Develop and Execute Competitive and addressing the digital divide between urban Strategies in Key Sectors and rural areas. Tanzania's biggest challenge to achieving sustained growth and competitiveness will be the dynamic, non- The high cost of connectivity is due primarily to linear process of upgrading its products and services. the use of satellite rather than cable to connect Given the reality of limited resources, minimal growth internationally. There are massive projects under- and significant barriers to global competitiveness, way to connect more of the country with fiber- Tanzania must be prepared not only to create a plat- optic cable, both through the mainland and to the form for all businesses, but also to focus additional international underwater backbone. These projects resources on those key sectors of the economy that will make domestic connectivity more affordable will create the largest short-term gains in productivity and reliable. Ensuring the success of these initia- and employment. Focusing in a structured way on tives and limiting duplication of efforts should be a the most promising sectors will increase Tanzania's priority. The Tanzania Communications Regulatory chances of gaining the expertise and income to rein- Authority (TCRA) will play a critical role in oper- vest in growing the rest of the economy. ating these networks, and this institution must also be able to enforce a fair playing field for telecom and data services providers so that new entrants can generate lower prices for consumers. CONCLUSION 5. Invest in Human Capital While access to appropriate technology and connec- The ultimate transformation of Tanzania's economy tivity are key elements of the growth equation, the will require more than strategic vision and resources. capacity of individual entrepreneurs and citizens to It will also require leadership. Building firm-level benefit from these tools is also critical. Tanzanians' competitiveness across sectors is a formidable task, 9 ITU 2003 10 Ibid 11 World Development Indicators, World Bank 2004 12 World Development Indicators, World Bank 2004 Executive Summary . 5 even in a stable macroeconomic environment, resources for managing successful change processes, which Tanzania has succeeded in fostering. Among including creativity, energy and courage. With these other critical tasks, public and private sector leaders ingredients in place, the magnificence of the road will need to work together to develop complex from Arusha to Moshi will become the only striking national sector strategies that can guide the micro- feature of the journey. Its current contrasts, symbols level strategies of hundreds of firms. It will require of unrealized prosperity, will be a distant memory cultivating the less tangible, but no less critical for Tanzanians. 6 . Improving competitiveness in Tanzania 2 TANZANIA economic performance is lagging by both global and regional standards, with GDP per capita stand- TODAY ing at US$ 322 in 2004.12 But there is reason for cautious optimism. The country's macro indicators show a positive trend, and growth has been steady over the past five years. GDP climbed from US$ 9.1 billion in 1999 to US$ 11.1 billion in 2003. Tanzania also experienced growth in the total value of its exports, from US$ 663.3 million in 2000 to US$ 1,235.1 million in 2004, while inflation 2.1 A STARTING POINT dropped during this same period from 5.9% to 5.0%. Stripping out the dramatic growth in miner- The challenges to economic growth in Tanzania are als from total exports, however, reveals a compound as vast as the country itself, which encompasses annual growth rate of exports from 2001 to 2003 945,000 sq. km. Like many lower-income African of just 4.6%.13 At the same time, the country countries, Tanzania is grappling with low levels of experienced growth in its foreign reserves, from education, poor infrastructure, limited healthcare, US$ 974.2 million to US$ 2,080 million.14 and economic reliance on commodity-based natural resources. On Artadi & Martin's "Global These positive economic indicators coupled with Competitiveness Index," which classifies countries recent political stability create an opportunity for into three stage based on means of production and Tanzania to embrace change, beginning its transfor- ability to attract foreign direct investment, Tanzania mation from a nation focused primarily on subsis- scores on the low end of Stage 1 countries with a tence farming and commodity exporting into one score of 3.12.* To benchmark this number, Angola that produces complex products and services that ranks the lowest at 2.55 while the U.S. ranks the command premiums on world markets. With the highest at 5.21. Tanzania's score is well below Sub- revenues that these activities generate, prosperity then Saharan African countries that will be used for com- comes into focus as a realistic goal. If these resources parison throughout the study, including South are invested in education, healthcare, financial and Africa (4.08), Mauritius (3.86), Uganda (3.50), physical infrastructure, then future generations of Kenya (3.37) and Zambia (3.25). By both regional Tanzanians will have access to increased purchasing and global standards, the snapshot of Tanzanian power and a higher standard of living. development is stark. In addition to some of the more common challenges, Tanzania is facing a complex transition from a socialist to democratic form of governance. Political stability in 2.2 GDP ANALYSIS Zanzibar has been elusive, leading to lower investor confidence. The historically strong role of govern- In mapping Tanzania's transition to a more complex ment in economic planning has led to a low level of and prosperous economy, analysis of GDP and private sector sophistication, as well as a lack of public export data can offer useful insight into the coun- and private sector cooperation in achieving economic try's path. GDP growth rates do not suggest an development goals. The shift to a market-based econ- optimistic outlook, given the low starting base of omy will require the private sector to become the Tanzania's per capita income. Tanzania has experi- engine of growth, and this will be a shift that will enced modest GDP growth from 1999 to 2003 take time and effort in Tanzania. with year-on-year growth staying above 3.6% and peaking at 6.2% from 2001 to 2002.15 Over the The impact of these issues is clear in the recent same time period, the population increased from performance of the economy. Tanzania's relative 32.9 million to 35.9 million.16 GDP growth did *World Economic Forum Global Competitiveness Report, World Bank, 2005­2006, available at http://www.weforum.org/en/initiatives/gcp/Global%20Competitiveness%20Report/index.htm 13 Bank of Tanzania, Economic Bulletin 14 The Economist Intelligence Unit, Country Report: Tanzania. London, UK. February 2005 15 World Development Indicators, World Bank 2004 16 Ibid Executive Summary . 7 FIGURE 1. GDP per Capita $4,500 $4,000 $ $3,500 $3,000 $2,500 International $2,000 2000 $1,500 $1,000 Constant $500 $0 1999 2000 2001 2002 2003 Tanzania Uganda Kenya Mauritius South Africa Zambia CAGR 3.3% CAGR 2.2% CAGR - 0.9% CAGR 2.8% CAGR 1.0% CAGR 1.9% Source: World Bank World Development Indicators not keep pace, therefore, generating a GDP per 7.2% in 2003. Mining's share of GDP, in spite of its Capita compound annual growth rate of just 3.3%, ascendancy in exports, is still low, growing from from US$ 262.40 in 1999 to US$ 308.70 in 1.4% in 1999 to 1.9% in 2003. Some estimates 2003.17 Although this growth seems positive, a show mining eventually growing to 10% of GDP.18 longer view reveals that GDP per capita growth has Although this contribution is small, mining remains been muted with a compound annual growth rate a leader in exports, with gold accounting for US$ of 1.0% from 1990 to 2003. 442 million in 2002.19 Compared to the other African countries reviewed As expected, Zanzibar relies less on agriculture as a in this report, Tanzania's 2003 GDP per capita is driver of GDP than does the mainland, reference the second lowest, see Figure 1, with Mauritius the Figure 3. Instead, Wholesale & Retail Trade, highest at US$ 4,161 in 2003. Encouragingly, Restaurants and Hotels hold the largest percentage Tanzania's CAGR from 1999 to 2003 is the highest. of GDP at 32.7% in 2002, although this number has consistently declined since from a peak of The three primary components of GDP on the 40.1% in 1999. This decline mirrors revenue dete- mainland are Agriculture, Forestry & Fishing; Trade rioration in the mainland's tourism sector, and will & Tourism and; Financial & Business Services, as be explored further in the section on Tourism. depicted in Figure 2. Agriculture, Forestry & Fishing has remained stable at around 45% from 1999 to 2003, while Trade & Tourism has gradually declined 2.3 Sector Economic Performance from 12.4% to 11.8%, and Financial & Business and Trade Statistics Services has steadily increased from 13.7% to 14.3% A key component of building the competitiveness over the same time period. Manufacturing, a GOT of developing economies--in which, by definition, priority, consistently dropped from 7.5% in 2000 to needs are significant and resources limited--is the 17 Ibid 18 Mwalyasi, Raphael B.B., "Impact Assessment and the Mining Industry: Perspectives from Tanzania, April 2004 19 The Economist Intelligence Unit, Country Report: Tanzania. London, UK. February 2005 8 . Improving competitiveness in Tanzania FIGURE 2. GDP by sector (Mainland) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 1999 2000 2001 2002 2003 Agriculture, Forestry & Fishing Mining & Quarrying Manufacturing Construction Electricity & Water Trade & Tourism Transport & Communications Financial & Business Services Public Adminstration source: Planning Commission, Economic Survey 2003 FIGURE 3. GDP by sector (Zanzibar) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 1998 1999 2000 2001 2002 Agriculture, Forestry & Fishing Mining & Quarrying Manufacturing & Handicrafts Construction Electricity & Water Wholesale & Retail Trade, Rest. & Hotels Transport & Communications Finance & Insurance Public Adminstration source: Planning Commission, Economic Survey 2003 Executive Summary . 9 ability to prioritize investment in industries based these industries are resource-based, with the exception on their relative potential to be globally competi- of semiconductors/computers, and competition is tive. To that end, this section builds on GDP statis- mostly based on costs (i.e. gold, timber, oil, memory tics and analyzes Tanzania's trade patterns in relation chips). to comparative and competitive advantage models of competition. The analysis of trade statistics is Across the middle row are broad end-use sectors explained and, finally, the data for Tanzania is involving industrial or supporting functions. These presented. clusters are centers of complex operations and conduits of innovation and upgrading, characterized by com- This analysis is built on the idea that increasing petition based on differentiation as much as costs Tanzania's competitiveness is linked directly to the (i.e. power generators, motor vehicles, and scientific ability of Tanzanian firms to increase exports. By instruments). Finally, the bottom row contains end- exporting, Tanzanian firms will expand their mar- use sectors mostly associated with final consumption kets beyond their own economy, increasing for- goods and services. In these sectors, competition is eign exchange earnings and driving a net increase mostly driven by product differentiation (i.e. break- in the population's purchasing power and standard fast cereals, clothes, furniture). of living. The link between these three horizontal bands and 2.3.1 Methodology and Theory productivity upgrading is relatively straightforward. To illustrate trade patterns and export flows, trade Economies generally begin the upgrading process charts are calculated for Tanzania over the from initial positions at the top (upstream indus- 1992­2002 period (see the Appendices for complete tries, usually extractive natural resource-intensive trade charts). These trade flows are presented in a industries) or the bottom (final consumption goods chart separated by 3 broad cluster bands that reflect and services) bands of the trade chart. Resource-rich consumption patterns in the economy. As illustrated countries typically begin with upstream industries, in Figure 4, the first broad cluster band is depicted and gradually develop competitive industries in the in the top row and shows upstream-sectors charac- mid-band (industrial and supporting goods) or the terized by industries whose primary products are services sector by investing the rents extracted from inputs into products in other industries. Most of their resource exports into developing human and FIGURE 4. Trade statistics framework "Broad Cluster" Definition Examples · Primarily used as inputs into other · Gold 40% Upstream industries · Timber Industries · Competition primarily based on cost · Oil · Memory chips · Centers of complex operations · Power generators · Conduits of innovation and · Motor vehicles Industrial and 10% upgrading Supporting · PABXs Functions · Competition on differentiation as · TV tubes much as cost · Goods destined for final · Breakfast cereals consumption · Television sets 50% · Competition traditionally on Final · Jewelry diamonds differentiated products and services Consumption · Clothes · Furniture 10 . Improving competitiveness in Tanzania knowledge capital that allows them to make this from increased tax revenue, but much of the transition. This process of transformation allows value of these minerals is being captured by for- countries to lay the foundations of a complex indus- eign companies who own the mines and process trial core or a thriving services sector that relies on the minerals outside of Tanzania. skilled human capital. 20 High export concentration in commodity products. Tanzania's export base consists primarily of tra- Resource-poor nations, in contrast, typically start ditional crops that add very limited processing exporting labor-intensive final consumption goods and value to their products (i.e. coffee, tea, cot- derived from basic agricultural products or simple ton, cashews, fishing and tobacco), though manufacturing. As these products are commodi- trending down relatively due to the increase in tized and price competition becomes fierce, coun- extraction upstream industries. Sectors located tries should start focusing on a number of niche in this band provide a starting point for earning markets and, in parallel, invest in human and foreign exchange. However, Tanzania's current knowledge capital to develop a competitive service model of competition hinges on comparative sector. Tanzania's current situation, where the small advantages. Exporting unprocessed crops, such level of exports is divided among minerals and final as the high volume of cashew shipped to India consumption goods and where no cluster now exist for value addition, prevents Tanzania from cap- in the middle band, helps to create the need for a turing more than a small percentage of the clear national growth and competitiveness strategy. cashew's potential value. Through forward inte- gration and investment in higher value products, 2.3.2 Broad Economic Performance Tanzanian firms can increase export revenue. and Trade Statistics This transition is beginning to happen in the coffee industry, where a shift towards premium A deeper analysis of trade statistics allows us to coffee, branded products and a focus on serving delineate the broad economic and export patterns more demanding, higher-paying customer seg- that emerge across Tanzania's existing manufactur- ments rewards coffee farmers and processors ing clusters. The analysis also provides insight into with increased profits. the evolution of an economy towards higher levels of productivity and value-added products. As a starting point, Tanzania has a trade balance of US$ 2.3.3 Regional Trade Statistics ­608.7 million.21 Its export receipts are very small Tanzania currently participates in two major relative to the overall economy, equaling just regional trade schemes: the East African 12.7% of GDP.22 Community (EAC) and the Southern Africa Development Community (SADC). The EAC There are two broad export patterns emerging in partnership was built in 2000 between Tanzania, Tanzania. See Figure 5. Both of these patterns are Uganda and Kenya, with the intent of furthering alarming and underscore the urgency for Tanzania regional integration. However, intra EAC trade is to evolve its economy. still low. Both Uganda and Kenya accounted for less than 9% of Tanzania's exports in 2003, with Increasing reliance on extractive, upstream indus- Ugandan trade making up less than 1% of this tries. Tanzania increasingly exports resources total.23 Imports to Tanzania from its EAC partners almost directly from the ground with very limited accounted for less than 6% of total imports in the processing or value added. The increase in this same year. band is due to mining exports which have increased dramatically over recent years due to Likewise, trade with other African nations outside a liberalization of policy by the GOT. Foreign of the EAC has been meager. Trade with South investment has poured into the country as firms Africa, Tanzania's most important trade partner in capitalize on Tanzania's rich mineral deposits, SADC, offers a revealing example. In 2003, especially gold. The government had benefited Tanzania was exporting slightly more than 3% of 20 Porter, Michael, 1990 21 Bank of Tanzania, Economic Bulletin 22 The Economist Intelligence Unit, Country Report: Tanzania. London, UK. February 2005 23 IMF Country Report No.04/284. P. 69, 72 Executive Summary . 11 12 . Improving FIGURE 5. Tanzania trade statistics by broad cluster competitiveness Forest Petroleum/ Semiconductors/ Materials/Metals Upstream Industries Products Chemicals Computers 1998 2001 2003 43% 43% 44% 44% in 34% 34% 34% 34% Tanzania 10% 10% 7% 7% 1% 0% 1% 2% 0% 0% 0% 0% 0% Power Generation Industrial and Multiple Business Transportation Office Telecommunications Defense and Distribution Supporting Functions 4% 5% 5% 0% 0% 0% 0% 1% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 1% 1% 2% 2% Entertainment/ Final Consumption Food/Beverages Housing/Household Textiles/Apparel Health Care Personal Leisure Goods and Services 85% 85% 65% 65% 57% 57% 54% 54% 42% 42% 35% 35% 11% 11% 9% 9% 12% 12% 8% 11% 11% 8% 3% 3% 2% 0% 0% 0% 1% 0% 0% TABLE 1. Tanzania: destination of exports and imports, 2000­2003 (% of total exports) Exports Imports 2000 2001 2002 2003 2000 2001 2002 2003 European Union 54.4 56.1 53.1 56.7 22.6 23.1 22.0 19.7 United Kingdom 22.0 18.3 18.1 33.9 7.0 7.4 5.7 5.0 Germany 9.9 5.0 3.1 2.7 3.4 3.9 3.6 3.2 Netherlands 6.9 6.7 6.0 6.0 1.9 1.8 1.6 1.5 Belgium 2.8 1.4 2.4 3.1 1.0 1.3 1.4 1.5 Italy 1.3 1.0 2.7 2.1 2.3 2.9 2.7 1.8 Other EU 11.4 23.7 20.8 8.9 7.0 5.8 7.0 6.7 United States 2.3 2.0 1.5 1.0 3.9 3.8 5.5 3.2 Japan 5.1 8.9 10.8 7.8 9.3 8.7 8.3 7.8 India 14.8 10.7 7.2 6.2 5.8 5.1 6.4 7.7 China (inc.Hong Kong) 1.4 0.1 1.3 1.2 4.8 4.1 5.2 5.8 Singapore 1.0 0.8 0.4 1.2 0.4 0.5 0.3 0.8 Kenya 4.8 5.0 4.0 7.0 6.1 5.6 5.7 5.3 Uganda 1.3 0.7 0.6 1.0 0.4 0.7 0.2 0.4 South Africa 1.8 1.1 1.9 3.3 11.5 11.8 11.3 14.1 Zambia 0.7 0.7 2.0 1.6 0.2 0.1 0.3 0.1 Zimbabwe 4.3 0.1 0.2 0.4 0.3 0.2 0.1 0.1 Other 8.1 13.9 17.0 12.6 34.8 36.6 34.7 35.0 Total 100 100 100 100 100 100 100 100 Source: IMF Country Report 2004 (from Tanzanian Authorities) its total exports to South Africa, while imports were efforts to further this process continue to face close to 14%.24 Comparing these numbers to the obstacles. The first problem is the homogeneity magnitude of inter regional trade with the EU and of products, especially between EAC countries. other non-African countries--exports to the EU Tanzania, Kenya and Uganda produce similar amounted to 56.7% of total exports, while imports goods particularly primary agricultural goods, were 19.7% of total imports in 2003--it becomes including coffee, tea, cotton and fish. This lack of clear that progress in this realm has been restrained. product differentiation makes trade less attractive See Table 1. between neighbors, and is further evidence of the need to produce higher-quality products that are Although intra-regional trade and integration are not only competitive within the region, but also on seen as potential drivers of growth for Tanzania, world markets. In pursuit of this goal, one approach 24 Ibid. P.69, 72 Executive Summary . 13 would be a regional effort to increase competitive- Kenya (another EAC member) reported delays in ness in world markets, which is more likely to suc- imports and exports of 7 and 4 days respectively, in ceed than promoting each country individually in Tanzania, enterprises that engaged in foreign trade the market. said delays were on average 14 days for imports and 7 days for exports. Another factor that affects not only the extent of regional trade, but also trade in world markets is the The GDP and export data confirm the diagnosis: existence of non-tariff barriers. Exporting and Tanzania must be prepared to take deliberate importing from Tanzania is hampered by, among actions to change its economic course in order to other things (i) health certificates for agricultural achieve broad growth and increased competitive- commodities; (ii) bureaucratic export procedures; ness. The following section explores a way for (iii) time consuming customs clearance; and (iv) lack Tanzania to begin this process. This section is of transportation and logistical infrastructure.25 focused on the theory of competitiveness, grounded Although these are many of the challenges faced by in Tanzania's specific experience. Its goal is to out- other African nations in furthering their trade ties line a development path that Tanzania can use as a with world markets, the situation in Tanzania is template going forward. It will highlight the role of slightly bleaker. Two recent World Bank reports on competitive firms, including SMEs, that are domestic, the country's investment climate place Tanzania on regionally and internationally focused, and will the lower end of the scale in terms of ports and cus- show how adopting this model can increase the toms delays.26 For instance, while companies in prosperity of the average Tanzanian citizen. 25 Refer to "Doing Business in 2005 and Investment Climate Assessment: Improving Enterprise Performance and Growth in Tanzania" 2004. World Bank Publications. 26 Ibid 14 . Improving competitiveness in Tanzania 3 TANZANIA'S from paternalism and plays the role of partner to the government in the success of the economy. TRANSITION Second, historians will note that the building blocks of the economy­the tools that enable the TO ECONOMIC country to prosper - become more complex. No longer will mineral deposits, cheap labor, or good GROWTH AND roads define the economic opportunities of Tanzania. In the future, the quality of the legal sys- tem, the quality of skilled employees, and the tech- A COMPETITIVE nology that allows Tanzania to communicate with customers and suppliers around the world will ECONOMY drive prosperity. Lastly, and driven by the first two changes, histori- ans will see that the products that Tanzania creates have fundamentally changed. Whereas in the past The transition from an economy that creates com- Tanzania produced commodity products as cheaply modity, low value products to one which responds as possible for sale to the highest bidder, in the to market conditions and innovates to produce future, products will embed unique knowledge high value, complex goods and services is an intri- about specific customer needs and preferences for cate and dynamic process. However, if we were to which discerning buyers will pay a premium. imagine that in 50 years, Ph.D. candidates in eco- nomics at the University of Dar es Salaam will Figure 6, shows this potential transition. This path study the evolution of Tanzania from a Heavily is not deterministic; in fact, most countries have Indebted Poor Country (HIPC) state to a prosper- not made such a transition and other paths have ous and upper-middle income country, they will been followed. However, this is the path that most see three significant changes. closely aligns with Tanzania's current possibilities and future prospects. In this section, we introduce First, they will note the changing roles of the govern- these concepts and highlight the primary areas on ment and private sector, as the government moves which Tanzania should focus to create the condi- from player to referee, and the private sector emerges tions for this complex change process. FIGURE 6. The evolution of a developing economy Private Private Sector Leadership Build a platform for core processes · Choosing Sector customers and and products markets Economic · Defining distribution channels Sell · Investing in employees Leadership products · Stable macroeconomic Government environment Exploit raw · Rule of law materials · Allocation of resources and cheap Government goods Leadership Raw Materials Intermediate Sophisticated or Cheap Labor Goods Products Ability of the Economy to Sustain Complex Relationships Physical Capital Social Capital Developing countries must invest rents from selling physical capital in creating the higher forms of social capital required to develop and sustain a complex economy Executive Summary . 15 3.1 PROSPERITY: increasing investment in higher forms of capital, have GOVERNMENTS ENABLE IT; the ability to create a more complex, secure and sus- tainable economy. Ideally, the government begins to FIRMS CREATE IT craft policy and regulation that creates an enabling environment for business, offering tactical support The experience of developed economies such as the without impeding competition. Without this transi- United States, Canada and the nations of Western tion of leadership to guide the growth of the econo- Europe maps to clear trends in the interactions of my, the forces of paternalism and protectionism can market dynamics and economic players over time. undermine the economy's ability to move beyond Their pattern informs this model and offers one simple exports to complex goods--goods that can possible trajectory for development. Increasing sustain an economy without decapitalizing natural globalization and the attendant shift towards a resources or exploiting low-cost labor. knowledge-based economy have made manufactur- ing-led growth, typified by the experience of This pattern shows up clearly in Tanzania's develop- Singapore and Asian "tiger" countries, reliant on an ment, which followed a socialist model from the abundance of low-cost labor. A country that seeks mid-1960s to the mid-1990s. During this period to develop within these current dynamics must the agricultural sector was dominated by collec- capitalize on its existing natural and comparative tives, as in other socialist countries. The GOT has advantages, but must also use competitive strategy made some of the necessary steps over the past ten to innovate, create products of greater value, and years to provide the private sector with a stable remain flexible enough to respond to changes in the macroeconomic environment, including low rates global marketplace. This model can serve as a guide of inflation, appropriate regulatory regimes, and for Tanzania, which possesses a wealth of resources, transparent governance. The government's current but which has failed to successfully use its endow- challenge is to ensure that a vibrant private sector ments to develop its economy in a significant way. develops over time, to which it can gradually hand over economic leadership. In applying this model of economic development to Tanzania, the first area that must be addressed is economic leadership. The roles of economic players generally evolve and become more clearly delineated as the countries and economies develop. In an econ- 3.2 THE EVOLUTION FROM omy's nascent stages, when limited capital is focused BASIC TO COMPLEX entirely on natural resources, infrastructure, and hard currency, the government often plays a com- PRODUCTS manding role in allocating limited resources to create a stable macroeconomic environment, establish a Developing countries that compete on comparative transparent rule of law, and lay the groundwork for a advantage, relying on basic factor advantages such as private sector. As a result, the government also ends cheap labor and sub-soil assets, tend to export com- up being the most important economic actor, push- modity products. Since competition in these exports ing certain industries and even managing state- is fierce and the products are largely undifferentiat- owned enterprises. In this phase, the key objectives ed, countries and firms typically compete on price: are peace and stability. The Tanzanian government is the lowest-cost producer becomes the market leader. for the most part still in this nascent phase, although This dynamic is the principal cause of the steady trying to move away from its dominant role. decline in real prices of most major commodities over the last 150 years, see Figure 7. In order to As competitive industries emerge and the private keep costs low, producers choose not to invest in sector becomes stronger and less fragmented, the human capital; in contrast, they often pay workers government should begin handing off economic lead- as little as they can. As a result, competition in basic ership to the private sector. At this point, the govern- products can increase poverty and lead countries ment's role needs to shift from being a nurturing par- into a downward spiral of wealth destruction. ent to being an impartial referee. Figure 6 highlights Sole reliance on comparative advantage is no longer how the changing roles of leadership, along with the a viable path to prosperity. 16 . Improving competitiveness in Tanzania FIGURE 7. Declining basic commodity prices, 1845­1999 Tanzania's trade statistics clearly place the economy Why have some countries succeeded at moving on the low end of the "Ability of the Economy to beyond their agricultural roots to produce complex Sustain Complex Relationships" axis, as Tanzania goods and services, while others have stagnated in competes primarily on comparative advantage, tra- an agricultural export economy--or worse, slid ditionally relying on basic natural resources, abun- into a subsistence farming economy? To put it in dant factor inputs and cheap labor. Figure 8 shows stark terms, countries have a choice between strategy that Tanzania's exports in agricultural products have and poverty. When firms and industry clusters have stagnated, partly due to the relatively little value strategies, they can embed their products with added to them. Mining is an anomaly, which has unique insights about customer needs that will be grown due to investments from foreign companies rewarded by the market, thereby lifting them out of exporting unprocessed minerals. the commodity trap. Countries with strategies earn the choice to participate in the Virtuous Cycle described in Figure 9, taking the rents they capture from their natural resources and low-wage labor FIGURE 8. Tanzania exports and investing those rents in building social capital 600 500 FIGURE 9. The virtuous cycle 400 How Does Wealth Actually Get Created? FOB 300 million, Wealth for the Economic Growth Social Equity Nation US$ 200 100 Capacity to Export Investment in Higher Complex Products Forms of Capital 0 1999 2000 2001 2002 2003 Minerals Cashew Nuts Coffee Sustainability Productivity Manufactured Goods Tobacco Cotton Innovation Tea Executive Summary . 17 in the form of educational institutions and also allow local customers to enjoy higher disposable improved governance. Over time, these institutions incomes and their demand for higher quality goods can support a higher skilled workforce that can and services increases, impacting local retailers and produce more complex goods, allowing the country other providers of consumption goods. Local SMEs to migrate from exploitation to innovation. will benefit from this dynamic if they keep their offerings up to the standards of these increasingly The central economic goal of any nation is to demanding consumers. Thus, as exporting firms attain a high and rising standard of living for all of grow local support SMEs and those throughout the its citizens. This goal is measured by increased per rest of the economy will benefit also. As discussed capita income and high-paying, satisfying jobs later, identifying attractive markets, support services, for a large proportion of those who can work. training and access to financing for all of these dif- Consequently, the pursuit of the central economic ferent types of firms can increase the velocity and goal is driven by productivity, which is based on impact of the cycle. the nation's stock of the seven forms of capital: cul- tural, human, knowledge, institutional, financial, man-made, and natural endowments. Increasingly, 3.2.2 Competitiveness and Poverty growing economies rely heavily on forms of capital Alleviation that can encourage innovation in human capital. In the past, because people were not the primary Tanzania must invest in its knowledge base to sus- source of a nation's wealth, societies had to make tain a more complex economy. tradeoffs between economic growth and social equity, with government playing a large role in determining 3.2.1 The Role of SMEs the distribution of wealth. Because highly productive Firms­not governments--create prosperity through and well-compensated people are at the center of the a combination of strategic choices and operational new model of wealth creation, competitive advantage efficiency. Competitive firms make better choices societies can look forward to making economic about which customers to serve and which goods growth and social equity complementary, with the and services to offer them. They also become more private sector naturally working as the primary mech- efficient in how they combine people and other anism for allocating wealth and alleviating poverty. resources to deliver value to customers. As firms become increasingly competitive, not only do they The greatest benefit to the economy of Tanzania will pay higher wages, but they must also invest in the come from home-based companies that understand skills of their employees, which drives an increasing this dynamic, particularly SMEs. A company's home standard of living for the general population. The base is the place where strategy is determined, where majority of businesses in Tanzania and many other core product and process technologies are created and countries are SMEs, which means that SMEs must maintained, where essential skills reside, and where also be able to increase productivity and execute some piece of sophisticated production takes place. It competitive strategies. is the place where the most productive jobs are typi- cally based, and where the beneficial spillovers to By seizing opportunities in both export and domes- other parts of the economy are the greatest tic markets, SMEs play a key role in increasing the purchasing power of the local population. To serve With the emergence of competitive firms and demanding customers in export markets, firms can industries, not only will demand for higher-skilled only prosper through the sale of quality goods and labor increase, generating an increase in wages, but services developed by well-paid, creative employees. there will also be more jobs and economic opportu- The same holds true for SMEs that supply these nity. Increasing employment levels is critical in exporting firms. They must also perform at the Tanzania, with an unemployment rate of 12.9% in highest quality level in supplying components or 2000/1, reflecting a recent doubling in the number delivering services. This dynamic creates the need of unemployed persons, from 405,722 in 1990/91 for business skills and innovation. Increased exports to 912,772 in 2000/01.27 27 Integrated Labour Force Survey 2000/2001, Ministry of Labour, Youth Development and Sports and the National Bureau of Statistics, The United Republic of Tanzania 18 . Improving competitiveness in Tanzania FIGURE 10. Labor force vs. GDP composition 2001 (%) Electricity & 100% Water 90% Financial & Business Services 80% Mining & 70% Quarrying 60% Transport & 50% Communications 40% Construction 30% Manufacturing 20% Other Services 10% Agriculture, Forestry & Fishing 0% % GDP % Total Employment Increasing employment in the same sectors where sectors to spur innovation and create sustained Tanzania has been competing for the past few growth. In doing so, support should be extended to decades, however, will not be enough. The current the SMEs in these sectors that show early signs of economic composition of the country and its labor success. These firms, in particular, will be in a posi- force allocation are major obstacles to creating the tion to create economic opportunities for a new, economic transformation that Tanzania needs. As more highly skilled labor force. depicted in Figure 10, more than 80% of the coun- try's labor force is employed in the agricultural sec- tor, which accounts for 45% of total GDP. The manufacturing sector, in contrast, employs a bit more than 1% of the labor force, but generates 3.3 CREATING A more than 7% of GDP. COMPETITIVENESS MINDSET Two things emerge from this analysis: first, it high- Broad economic change cannot happen in the lights a highly inefficient agricultural sector based absence of broad commitment. People, sectors, on subsistence farming that is in urgent need of nations, and societies have relied on comparative change before people can participate in more pro- advantage for centuries, if not longer. For many, ductive employment. For this change process to the principles of competitiveness are at best foreign, begin, Tanzania must launch a serious campaign to and at worst, threatening to traditional business engage SMEs in the agribusiness sector in a process thinking and strategy. To complete the transforma- of product upgrading. Current agricultural prac- tion to a truly competitive economy, individuals tices are more labor-based than capital-based in must learn to think competitively, understanding comparison to other sectors. Moving from subsis- the principles that drive global competition and tence to cash crops will require the adoption of increasing their receptivity to innovation and new techniques and technologies that will improve change. But this is not an easy shift; in fact, creat- yields. This shift would represent value accruing ing a culture of competitiveness can be the most not only to the Tanzanian economy, but also to the difficult part of the journey. majority of Tanzania's labor force. This change process typically begins with entrepre- The second key finding from this analysis is that as neurs, firms, industries and government officials the agribusiness sector moves up the value chain, committing to the idea that competitiveness is a Tanzania must begin looking to more productive path to sustainable growth. As they test competitive Executive Summary . 19 strategies in their businesses and policies, gradually diffuse a better understanding of competitiveness. the evolutionary process begins for the country's These campaigns often focus on celebrating entre- economy. Making mental models explicit, which are preneurs and innovators, disseminating principles the beliefs, customs and values of a society, is a of competitiveness along with concrete examples of helpful starting point for government leaders. successful businesses executing competitive strate- Culture is such a significant component of econom- gies. These types of campaigns, supplemented with ic development that without understanding the training forums on competitive strategy, with key beliefs and values that inform the actions of eco- players from the public and private sector can also nomic players, change can be virtually impossible. prove helpful. As businesses and government embrace competitiveness, communicating the A communications campaign to engage economic impact of successful, data-driven strategies helps to actors on their ideas and assumptions can help to create an environment that is receptive to change. 20 . Improving competitiveness in Tanzania 4 ICT AND adoption of tools and services. Achieving success requires the tactical usage of ICT to inform strategy BUILDING and improve productivity. If funds are invested in ICT to drive adoption without demonstrable returns in productivity and efficiency, countries will COMPETITIVENESS fall further behind, and future investments in ICT will be harder to justify in light of a negative track- record. When seeking to invest ICT resources in a promising industry or firm opportunity, three inputs to success must be considered, illustrated ICT has a critical role to play in the transition to a further in Figure 11: more developed economy. As an economy shifts Access: Does the recipient organization have the from industries focused on exporting raw materials infrastructure to use the proposed ICT? Factors to more complex, value-added industries, demand that affect access are availability of services such for ICT products and services grows. ICT improves as electricity, telephone, cellular and satellite. firms' ability to communicate with customers, facil- Ability: Do the intended beneficiaries have the itating upgrades in their competitive positioning. It skill set and resources to leverage the technology? can also help to improve logistics, increase trans- Willingness of the organization to pay for ICT parency and make geographic distance irrelevant in services is an ability factor, as is the level of edu- transactions. ICT is an essential tool in the evolu- cation and technology training of an organiza- tion of a developing economy, but its value is tion's employees. In Tanzania, ability is highly dependent on the growth of competitive firms and stratified. Primary schooling is on par with other industries. The benefits of ICT usage for some of Sub-Saharan African countries, but there is still a Tanzania's key industries will be explored in Section lag in secondary enrollment and skilled IT labor. 5, while this section will focus on the central plat- Most of the decision makers at the top of organi- form issues necessary to realize those benefits. zations--both public and private­remain uncon- vinced that significant change is required to benefit from these technologies. 4.1 ACCESS, ABILITY AND Benefit: Does the project increase the competitive- ness of the economic actors, even after accounting BENEFIT MODEL for project costs? Measuring the impact of ICT is not easy. In developed countries, ICT invest- The promise of ICT to improve competitiveness ment has been linked clearly to growth in Total cannot be realized simply through a country's Factor Productivity.28 In developing countries, FIGURE 11. Keys to ICT usage Physical access to ICT · Electricity Access · "Dial Tone" ­ Fixed-line Adoption ­ Mobile Individuals ability to use ICT Value of using ICT · Affordability · Productivity · Education · Differentiation/Innovation ­ Primary/Secondary Government · Quality of Life ­ Tertiary Usage ­ Specialized Training Ability Benefit 28 The World Bank's 2004 "Contribution of Information and Communications Technologies to Growth" by Zhen-Wei Qiang, Pitt A. and Ayers S., analyzed the impact of technology in Total Factor Productivity in developed economies. Total Factor Productivity is defined as the output growth minus the growth rates of capital and labor stocks, weighted by their contributions to output. This study has proven the positive results many Western Economies (especially the US) have accrued from investment in ICT. Executive Summary . 21 FIGURE 12. Five levels of ICT usage Non-networked Networked ICT Intensive ICT Non-ICT users CT Only users ICT users users users Enterprises that Enterprises that Enterprises with Enterprises with Enterprises using make no use of make no use of one or more stand-alone two or more computers and computers, but computers on their computer(s) that internally- have no immediate have access to ­ premises, but with lack internal networked access to and make regular no network networking, but computers that also telecommunication use of ­ connections; they which have an have email/Internet services telecommunication will have access to external connectivity. services; telecommunication email/Internet telephone, fax, services. connection cellular this methodology is difficult to use because infor- understand their true competitive position mation is not readily available, and ICT use may and adapt to market forces. be too limited to effectively track its influence. However, some indicators can be used to measure Each of the three keys to ICT adoption and usage ICT's impact on competitiveness, at both a firm is necessary for ICT initiatives to succeed, and they and institutional level: must be considered simultaneously. This frame- 1. Productivity Improvement: Has the use of work is most useful in highlighting the limits of ICT increased efficiency and/or reduced looking at access and ability metrics in the costs? ICT can be a fundamental tool in absolute. Literacy rates or number of internet users improving productivity by facilitating logisti- are negligible inputs into policy prescription if they cal improvements through faster and more are not tied to the potential benefits. efficient communication along the value chain. It can also make distance irrelevant, These benefits can vary greatly--as illustrated by the particularly for digitized products and services. five levels of usage model--depending on the industry, In a service economy, the ability to provide the position of the user in the value chain, the type of faster services in a more customized way is a product and service offered, etc. Not every firm can norm of competition, and ICT acts as the use or benefit from the same level of ICT adoption. platform. Improving transparency through This is the fundamental difference between ICT in the universal availability of information can developing countries, and ICT in developed countries, also reduce the cost of doing business and where much of the basic technology is ubiquitous in strengthen institutions. both business and personal settings. In fact, the specific 2. Creating Differentiation: Has the firm/insti- type of ICT usage varies not only based on a firm's tution used technology to enter new niche sophistication, but also on its role in the value chain. markets, innovate and differentiate its products The optimal intervention for increased usage and pen- or processes? Has ICT allowed it to provide a etration of ICT throughout an industry's value chain differentiated product to new and more can be characterized according to the Heeks & profitable global clients? The ability to com- Duncombe's* model. See Figure 12. municate with customers, faster and more inti- mately, allows companies to customize their These five levels are not necessarily linear. A firm or offerings to achieve a competitive edge. ICT industry's ICT usage could "leap-frog" over intermedi- also makes information about global competi- ary levels depending on its needs and capabilities. For tors and clients available, helping firms to example, as connectivity improves through satellite * Adapted from "Information, ICTs and Small Enterprise Findings From Botswana", Richard Duncombe and Richard Heeks, Institute for Development Policy and Management, University of Manchester, Manchester, UK, 1999 22 . Improving competitiveness in Tanzania and backbone networking, a firm formerly in the for Tanzania compared with other African countries. "CT Only Users" category could move directly into The Telecommunications sector is also analyzed as "Networked ICT Users". Yet, every firm and every it is a key determinant in the private sector's ability link in an industry's value chain will have distinct ICT to afford and access connectivity. needs. These levels of usage can be helpful in consid- ering current and potential ICT usage and penetra- On the dimension of access, Tanzania is in a precari- tion, and should be weighed simultaneously with ous situation. Fixed and mobile teledensities are some issues of access, capacity and potential value. These of the lowest in Africa (see Figure 13), even with the issues are explored further in the following section. increasing use of mobiles compensating somewhat for the slow growth in fixed-line telephony. From 1997- This model explores an essential truth about ICT: 2002 fixed lines grew at a compounded annual rate although usage correlates highly with improved effi- of 9%, but there was still less than one telephone user ciency and productivity, there is no single level of ICT per 100 inhabitants in the country. Compared to usage that is appropriate to all institutions in a society. Mauritius or South Africa, which have more than 10 For a rural farmer, simply gaining access to a tele- users per 100 inhabitants over the same period, access phone may allow a significant leap in productivity. to fixed telephony in Tanzania seems to benefit just a Consider the time and money the farmer could save privileged few. On a relatively positive note, the com- by phoning an agronomist for advice on the proper pound annual growth rate in mobile telephony has herbicides to combat a fungus­he saves the time it reached over 100% during the 1997-2002 period, takes to travel into town, plus his transport costs. resulting in overall mobile teledensity of almost 2%. Tanzania still runs short on telephone access in rela- At the same time, a bank serving the entire nation tion to other countries in the region, yet most of the may well demand state-of-the-art ICT solutions to projected growth in teledensity is expected to come tie together regional branches with real-time infor- from mobile telephony. mation. In the same way that a bank cannot be most efficient using only telephones, the rural The slow growth in fixed telephony is a result of the farmer would not (and could not) use the manage- telecommunications monopoly and the country's ment information systems to be a better farmer. poor infrastructure capacity. The Tanzania Post and Accurate measurement of the needs and capabilities Telecommunications Corporation (TPTC) used to of the ICT end users is necessary for ICT to create be the exclusive telecoms and postal services real economic value in Tanzania. provider up until 1993, when the state monopoly was dissolved and replaced by three separate entities: Just as an economic diagnostic should be done before the Tanzania Telecommunications Company identifying barriers to national growth and competi- Limited (TTCL), the Tanzania Postal Corporation tiveness, a comparable process is useful in assessing (TPC) and the Tanzania Communications the ICT landscape. The following section offers an Commission (TCC) in charge of overseeing regula- analysis of the current state of ICT in Tanzania, with tion in both sectors. TTCL was privatized in 2001 particular focus on Tanzanians' access to ICT and and granted a four-year exclusivity period to meet their ability to realize its value. This data will lay the installation targets of at least 800,100 lines by the groundwork for an evaluation of the current and end of this term in February 2005. The sharehold- potential ICT benefits for the private sector. ing structure was 35% strategic investor (MSI/Detecom), 14% international financial insti- tutions, 10% local financial institutions, 5% TTCL employees and 36% GOT. MSI/Detecom was awarded the 35% share at a cost of US$ 120 million, 4.2 ACCESS: MIXED RESULTS of which US$ 60 million was paid and the balance currently disputed in the courts. To date, the rollout 4.2.1 Limited Access to Telecom of 800,100 has not been met. Furthermore, there and Internet Services have been complaints surrounding the quality of fixed line telephony. A 2001 report by the Swedish Low SME and private sector ICT adoption in International Development Agency (SIDA) Tanzania can in some part be attributed to low highlighted the fact that almost 30% of telephone access. In this section, we review the state of access Executive Summary . 23 FIGURE 13. Teledensity of fixed vs. mobile telephony (1997­2002) Uganda Kenya Mauritius South Africa Tanzania Zambia LowerIncome Lower Middle Income UpperMiddleIncome High Income 0 10 20 30 40 50 60 70 Percent Fixed lines per 100 inhabitants Mobile subscribers per 100 inhabitants Source: ITU 2003 FIGURE 14. Compound annual growth rate of fixed vs.mobile telephony (1997­2002) Uganda Kenya Mauritius South Africa Tanzania Zambia Lower Income Lower Middle Income Upper Middle Income High Income -10 40 90 140 190 Percent Fixed line CAGR (%) Mobile line CAGR (%) Source: ITU 2003 24 . Improving competitiveness in Tanzania FIGURE 15. Internet users per 100 inhabitants (1997­2002) Uganda Kenya Mauritius South Africa Tanzania Zambia Lower Income Lower Middle Income Upper Middle Income High Income 0 5 10 15 20 25 30 35 40 45 Percent lines were not working, and that rural networks Internet Service Providers in Tanzania, although posed persistent quality problems.29 only 12 are active.33 Despite and, in part, due to the delays and difficulties While these penetration rates demonstrate that with fixed telephony, there has been a dramatic surge Tanzania is lagging the region and world in terms in cellular subscribers, as people increasingly choose of ICT access, this gap has been somewhat alleviat- mobile phones as their primary telephone line. This ed by a rise in mobile telecenters and a boom in is particularly true for prepaid mobile services, which cyber cafes throughout Tanzania, primarily in are considered a cheaper and more flexible option, urban centers. These centers and cafes provide a without the commitments and connection delays remedy for citizens who cannot afford a telephone, that a fixed line requires (fixed monthly payment computer or internet subscription. whether you use the line or not, credit qualifications and connection charges, etc). As of May, 2005, there A more comprehensive analysis of ICT costs in are four licensed GSM mobile operators, including: Tanzania reaches a similar conclusion. Although MIC Tanzania Limited (September 2001); Vodacom access to telecommunications and Internet services Communications Limited (December 1999); has improved slightly, telecom monopolies and small Zanzibar Telecom Limited (January 1997); Celtel and dispersed markets drive up the cost of services (July 2001). All operators have 15-year renewable and curb widespread adoption. Outbound calls are licenses.30 There were an estimated 250,000 mobile still higher priced than inbound calls34, especially phone subscribers in 2001, with 90% of those using when compared to countries that have more devel- prepaid cards to make calls.31 The percentage of the oped telecom networks and that have taken greater population who now report access to a mobile strides towards liberalization (Mauritius, South phone, one belonging to a family member, or part of Africa, and the United States). For instance, a one- a telecenter, is approximately 97%.32 minute call from South Africa to Tanzania costs US$0.49, compared to almost double that price In terms of internet hosts and computers, Tanzania (US$0.95) is the same call originates in Tanzania. has one of the lowest usage rates in the region. A more telling example comes from comparing the Internet density is particularly low at 0.23 users per costs of calling to the US from Tanzania and Uganda. 100 inhabitants. Currently, there are 21 licensed While the cost of making a one-minute call from the 29 Miller Esselaar and Associates "A Country ICT Survey for Tanzania", Swedish International Development Cooperation Agency (SIDA), November 2001 30 Tanzania Communications and Regulatory Authority 31 Miller Esselaar and Associates "A Country ICT Survey for Tanzania", Swedish International Development Cooperation Agency (SIDA), November 2001 32 Hancock, Simon. "Mobile Phones Boom in Tanzania," BBC News. Available at http://news.bbc.co.uk/2/hi/programmes/click_online/4706437.stm. 33 Tanzania Communications and Regulatory Authority 34 International Inbound and Outbound phone rate graph refers to the following: residential, daytime rates. Executive Summary . 25 FIGURE 16. Intra-regional and US rates Intra-Regional Inbound and Inbound & Outbound Calls Outbound Phone Rates from the US Zambia Zambia S. Africa S. Africa Mauritius Mauritius Kenya Kenya Uganda Uganda Tanzania 0 0.5 1 1.5 0.00 0.50 1.00 1.50 (US$/Minute) (US$/Minute) Inbound Calls (To Tanzania, daytime) Outbound Calls (To NYC, US daytime) Outbound Calls (From Tanzania, daytime) Inbound Calls (From NYC, US daytime) US to these countries is the same (US$0.50), a call from Uganda to the US stands at US$0.70, but the FIGURE 17. Dial-up and high speed same call from Tanzania is US$0.95. internet monthly costs On the other hand, the cost of Internet access is not as high as in other countries with the same income level. Zambia For instance, the cost of 20 hr dial-up access in Tanzania is US$ 117, compared to US$ 170 in Kenya. South Africa and Mauritius, however, offer the same S. Africa service for as low as US$33.33 and US$14.80, respec- tively, see Figure 17. High-speed connection is as high Mauritius as US$1,500 per month in places like Zambia35, while Tanzania offers more modest rates (US$ 300/month) that are somewhat higher than in Mauritius (US$ Kenya 213) and South Africa (US$ 102). Notwithstanding, these figures point to an Africa-wide problem in terms of high speed Internet access, since comparable service Tanzania costs US$ 40­50 in the United States. 0 500 1000 1500 2000 Cost/month (USD) 4.2.2 Infrastructure Constraints Dial-up (20 hrs) USD 4.2.2.1 International Connectivity High-Speed (per/month) USD International connectivity in East Africa is scarce, costly, and inconsistent in quality. Countries in west- only gains access to international content and com- ern, southern, and northern Africa are connected to munication systems via satellites. Relative to connec- submarine fiber optic cable systems that provide both tion via fiber optic cables, broadband connection via intra-regional access and access to the other conti- satellites is very expensive. Satellites also have limited nents. East Africa has no such connection, and thus, it bandwidth capacity and experience transmission delays. 35 Zambia and Kenya have exorbitant high-speed Internet costs due to the mode of connection (mainly connect via satellite) and the monopoly from Telkom Kenya and Zambia Telecom. 26 . Improving competitiveness in Tanzania FIGURE 18. The EASSY project Furthermore, some African carriers pay a tremendous project would help to decrease the cost of connec- amount in switching fees, translating into hundreds of tivity within Tanzania, and would also facilitate millions of dollars annually to switch intra-African rapid access to non-locally hosted website content. traffic through foreign carriers.36 It should be pointed out that although the EASSY project has gathered attention; a fiber optic link The EASSY project, which is currently under discus- for international connectivity need not only come sion, aims to provide high-capacity submarine fiber from underwater. A South African firm has begun optic cables to increase international connectivity to invest in providing this link from South Africa between Africa and the international community. to other African countries overland. Figure 18 illustrates the proposed connection in black, connecting the SAFE cable in southern Africa to the 4.2.2.2 The Backbone SEA ME WE cable in northern Africa. The EASSY Tanzania's existing fiber optic networks are owned project would involve the construction of a 9,900 km and operated by separate, uncoordinated institu- submarine fiber optic cable system that would link the tions, namely TANESCO, TAZARA, TRC and East African seaboard from Durban, South Africa SONGAS. Each institution has historically utilized through Mozambique, Madagascar, Tanzania, Kenya, independent network and development plans. and finally culminating in a connection to Djibouti.37 However, growing demand for fiber optic connec- tivity to facilitate high capacity, quality delivery of EASSY would help Tanzania meet the growing voice, data, and image services and applications has demand for broadband, connectivity by ISPs, data forced the Ministry of Communications and service providers, broadcasters, and VOIP providers. Transport and key telecom industry stakeholders to It would also bolster inter-Africa trade by making assess the national ICT backbone infrastructure regional communication easier and less expensive. and the potential for cross-sector coordination. A preliminary feasibility study estimates that the project would cost US$ 200 million, including a Figure 19 illustrates the existing and "ideal" fiber optic US$ 170 million for system supply and US$ 30 and microwave networks within the backbone infra- million for project management. structure. This ideal network is based on work done by The Ministry of Communication and Transport, International connectivity is a critical issue in which conducted a report on the "Status of the improving access to ICT in Tanzania. The EASSY National ICT Infrastructure Backbone" in 2004. 36 EASSY Secretariat Office, Telkom Kenya. Available at http://www.eassy.org/index.htm. 37 Ministry of Communications & Transport. "Technical Report on Feasibility Study for Implementation of the National ICT Backbone Infrastructure," June 2005. Executive Summary . 27 FIGURE 19. National fiber and microwave networks Existing Ideal This was followed by an exhaustive technical feasibility evident in Tanzania, where the inadequacy and unco- study in 2005. According to this research, the project ordinated nature of the current infrastructure back- would require the construction of 6,997 km of fiber bone is a clear obstacle to drawing commercial optic cables and approximately 3,475 km of links. providers into rural areas, leaving these communities Once the networks have been joined and the necessary with limited or no access to ICT requiring fiber optic linkages built, the backbone would provide fiber optic connections. capacity for lease and use by different operators. The penetration rates of fixed phone lines and inter- The costs of building the necessary fiber optic net users are largely skewed by the number of people cables and links in existing networks were estimated who do not have access in rural areas. In considering to be US$169 million. These costs, compiled in how to improve access and ability in rural areas, it is Table 2, are substantial, but the potential benefits important to think about optimizing existing ICT in backbone infrastructure investment and coordi- according to their ability to increase income or nation are great. The GOT should explore ways to reduce cost among rural populations. In a recent sur- spur investment in this project. vey within rural communities, respondents reported the following current uses of ICT by frequency and 4.2.3 The Urban/Rural Digital Divide the relative impact of one of the most prominent Much has been written about the "digital divide" ICT tools, the mobile phone, on their financial capi- between urban and rural areas. This issue is particularly tal. See Figure 20 and Figure 21.38 It is clear that ICT in the form of mobile telephones TABLE 2. Estimated backbone have emerged as an import saver of cost and time in investment costs rural areas. Experts express concern, however, that this "leapfrog" effect of wireless may come at the Description Sub total (US$) expense of fiber optic build-out and its potential for Optical Fiber Cable Installation $106,467,659 more reliable data transfer, increasing the existing rural-urban divide. This pitfall has been tempered in Transmission Equipment Installation $35,742,135 part by the presence of Multipurpose Community Power Supply System Installation $12,255,063 Telecentres, such as the site in Sengerema, which provide shared ICT resources to rural populations. Civil Work And Others $15,067,000 Grand Total $169,531,857 The Sengerema Multipurpose Community Telecentre (MCT) began as a pilot project in 2000, sponsored by 38 Department for International Development (DFID) KaR Project 8347: The Economic Impact of Telecommunications on Rural Livelihoods and Poverty Reduction: A Study of Rural Communities in India (Gujarat), Mozambique and Tanzania, June 2005 28 . Improving competitiveness in Tanzania FIGURE 20. Frequency of ICT usage within rural areas Source: DFID, June 2005 the Commission on Science and Technology The impact of the MCT on quality of life in (COSTECH) and UNESCO. Currently, the MCT serves Sengerema has been important. In addition to Sengerena's population of 500,000 with 25 computers, providing services that enable commerce and educa- internet connectivity, and a broadcast radio. The launch tion--for example, providing farmers information of the MTC required an investment of US$400,000 via CD-R on increasing crop yields and rural stu- of donor money. The MCT began as a telecentre, dents photocopies of national examinations--the offering telephone access, and has since expanded its MCT has been vital in disseminating public health range of services to include radio, photocopying and information. When the MCT opened in Sengerema, internet access. The demand for these services is per- fewer than 50% of babies were vaccinated. The haps best illustrated by tracking the MTC's growth in vaccines were free of cost at local clinics, but new revenue. Its initial monthly revenue in 2001 was mothers did not have adequate information regard- US$300/month, and now stands at US$3,000/month. ing the need for vaccines or their cost. As a result, FIGURE 21. Perceived impact of mobile telephony on financial capital among rural populations Source: DFID, June 2005 Executive Summary . 29 this service was underutilized and failed to impact infant mortality rates. The MCT began broadcasting FIGURE 22. School enrollment radio announcements reminding mothers to vacci- indicators 2002 nate their babies, and within eight weeks the vacci- nation rate among newborn babies approached 100%. 3.2 Uganda* 14.2 89.0 Another important area to consider when assessing 2.9 Kenya* 24.5 ICT access in Tanzania is local content. Ensuring a 66.5 suitable amount of local content will increase access 15.2 Mauritius 70.6 for a large segment of the population that may not 90.4 read English. For instance, Microsoft sees a market 15.0 South Africa 65.5 89.0 for its software among the roughly 100 million 0.9 Swahili speakers in East Africa, and the company is Tanzania* 5.0 68.8 now working to incorporate Swahili into Microsoft 2.4 Zambia* 22.8 Windows, Microsoft Office and other popular 68.4 programs. The same is true for Google, which has 0 20 40 60 80 100 Percent launched www.google.co.ke, offering a Kenyan School enrollment, tertiary (% gross) version in Swahili of the popular search engine. A School enrollment, secondary (% net) major key to access is devising strategies and software School enrollment, primary (% net) to put local languages on the screen, which will increase the value and consumptions of information Source: World Development Indicators, World Bank 2004. technology. This is particularly relevant for rural Note*: Secondary encollment rates for Tanzania are from areas, which are often more limited by language. 2001 http://devdata.worldbook.org/genderstats. Tertiary rates for Kenya and Zambia are from WDI 2000. Uganda ter- Projects aimed at increasing awareness will need to tiary levels from 2001 and primary levels from The World ensure that they endorse programs in Swahili, partic- Bank Education Notes April 2002 with data form 2000. ularly websites for domestic commerce. While these primary education figures are cause for concern, secondary enrollment is still the greatest educational weakness in Tanzania. About 22% of 4.3 LOW LEVEL OF ABILITY primary students have a chance to pursue secondary education, which translates into 5% net secondary enrollment of the relevant age group, compared 4.3.1 Basic Educational Deficiencies with an average of nearly 30% across Sub-Saharan In addition to minimal access, limited ability to use Africa.39 The secondary school system in Tanzania ICT is also responsible for the lack of ICT penetra- faces a number of challenges, including demands to tion in Tanzania. Ability refers to the willingness and increase access (especially for low-income youths preparedness of individuals to use ICT, a direct func- and students in rural areas), improve quality and tion of their level of education and technology training. reduce costs.40 During the 1980s and early 1990s, Tanzania faces significant ability barriers. While recent increased private sector participation and communi- reforms in primary education have lifted enrollment ty involvement led to the rapid growth of non- rates, secondary school enrollment and tertiary educa- government secondary schools as a way to cope tion are among the lowest in Sub-Saharan Africa-- with some of the excess demand.41 Currently, it has just 5% and 1% respectively. See Figure 22. been estimated that close to 40% of secondary stu- dents are enrolled by private providers. The primary education reforms implemented dur- ing the 1990s have had a positive effect. Although The recent increase in primary education, however, will Tanzanian rates are on par with other Sub-Saharan create mounting pressures from the state to provide countries (Kenya and Zambia), they still lag places in secondary schools, expanding access beyond a Botswana, Mauritius and South Africa, which have small group of privileged youth. The picture becomes upwards of 80% net primary enrollments. even more complicated when the urgent need for 39 Ibid 40 Ibid 41 Ibid 30 . Improving competitiveness in Tanzania Tanzania to transform its economy is included; institutions in maintaining systems and software on Tanzania is being crippled by its inability to develop the multiple computers. Schools, in this case, use a central human capital it will need to drive this transformation. server or thin client system to operate several machines. This allows all software, computing power and data 4.3.2 Incorporating ICT into storage to take place on a central server, reducing power the Education System usage and allowing for low bandwidth connections. If the pilot project is effective, it would provide a cost- Efforts are being made to incorporate ICT tools into effective means to introduce students to ICT. secondary schools in Tanzania, but the results have been mixed. The Distance Learning and Education These two initiatives are laudable, but face two Services (DILES) project was created to distribute problems. The first is leveraging the internet in secondary school information to students via the places where connectivity is not available or reli- internet. The initiative was developed with an eye able, and thus much information requires paper for towards reaching students in rural areas, where dissemination. The second is the rationalization of books are expensive and good teachers are hard to ICT investment by the public sector. The issue of find. The project was fully funded by the connectivity needs to be solved before it can be International Institute for Communication and leveraged to improve learning and create incentives Development (IICD) and began operations in 2000. for people to invest in computer and ICT skills. Students who cannot afford textbooks can download curricula and syllabi from the website. The syllabi are 4.3.3 Tertiary Enrollment based on the national syllabus, so they correspond Enrollment ratios in primary and secondary educa- with the teachings the students are receiving in the tion are an indicator of the future quality of human classroom. There are also twelve years of solved capital. Naturally, low enrollment rates in second- national examination papers. Students can order hard ary education contribute to the observed skill gap copies at less than cost; US$ 2.00 for a comprehen- in higher education in Tanzania. The country has sive mathematics syllabus for example and the solved one of the lowest attainments in tertiary schooling national exams at US$ 0.20 per exam. The rationale in the region, with less than 1% of the population was that without this service students would have dif- attending university. Although this is common in ficulty accessing these documents. Sub-Saharan Africa, countries such as South Africa and Mauritius have managed to enroll more than Though DILES was never solely a commercial ven- 15% of the population in tertiary education. ture, as downloads of syllabi have always been free, DILES did hope to eventually commercialize the One indicator of the attractiveness of investing in edu- operation through selling access licenses and CDs on cation in Tanzania, is the estimated rate of return on an offline system. The plan of creating a business education and training. These are presented in Table 3 based on a market of poor rural students, however, and are categorized by type of education and gender. As has not come to fruition. In very remote settings, stu- this table indicates, private rates of return increase with dents do not have access to the internet, and printing educational attainment, and become more relevant for costs are prohibitive. DILES has had to change its university and vocational and technical training. For strategy for reaching the students and is now making instance, wages are approximately 7% higher for some- hard copies of all materials and using donor funds to one completing secondary education than for someone distribute their database of information to more than just completing primary schooling. Likewise, a universi- 1,000 schools and teacher training colleges. ty degree and vocational and technical training increase earnings, on average by 9% and 19.4% respectively. Another initiative aimed at incorporating ICT into sec- However, a recent World Bank publication points to ondary schools is the Thin Client Computers project the fact that given the relative scarcity of post-primary implemented by the GOT in collaboration with SIDA. education in Tanzania, the reported returns to educa- The project is being piloted in urban settings, and is tion appear low relative to international standards.42 aimed at reducing the cost burden of many educational This may reveal the low relevance of post-primary 42 World Bank 2002 Executive Summary . 31 TABLE 3. Annual private rates of return to education and training (percent) Education Level Training Vocational & On the Group Primary Secondary University Technical Job All 3.6 6.9 9.0 19.4 35.2 Male 1.9 6.6 9.9 17.8 33.0 Female 10.8 9.0 11.4 20.2 35.0 Source: World Bank 1997 education to the requirements of Tanzania's economy.43 Although some commendable efforts have been made This dynamic reinforces the urgency of upgrading the to make academic and training programs more rele- economic model for Tanzania. vant to the needs of the labor market and to incorpo- rate IT in the education system, wider ICT literacy will only become a reality when both individuals and 4.3.4 Training: Not Yet Driven By organizations use ICT in their daily lives and when Private Sector Needs appropriate advanced training becomes available to In fact, the general perception of business leaders is those entering (and already in) the workforce. that the disconnect between academic and training institutions and the skill set demanded by the private Workforce ability is coupled with access as prerequi- sector is one of the main obstacles to the develop- sites to productive ICT usage, but they are mean- ment and utilization of ICT. A local Tanzanian busi- ingless if users do not recognize tangible benefits of nessman, the head of a successful group of companies ICT. Investments must be rationalized against that includes several ICT firms, confirmed this asser- potential ICT advantages or competing uses of time tion. In response to this deficiency, he is assisting in and capital. The benefits component of the AAB drafting a proposal for a study to determine the model is explored in the following section. In this appropriate ICT training needs of the private sector. section we investigate the current and potential role By first identifying the needs of Tanzanian business of ICT in certain sectors: Agribusiness, Mining & in ICT, training programs can be developed that Minerals, Services and SMEs in general. In each of address the specific requirements of businesses. The the three economic sectors, we also provide a sector University of Dar es Salaam has taken a proactive audit and offer preliminary recommendations to approach to this issue. improve each sector's competitiveness. BOX 1 The University Computing Centre Ltd.: "one student, one computer" The University Computing Centre, Ltd is the primary provider of ICT training courses within Tanzania. It was created as a private company within the University of Dar Es Salaam in 2001, so that it could concentrate on service provision, revenue generation, and offer competitive compensation to its professors and staff. The UCC has five locations throughout the country and is concentrated in four lines of business: 1) ICT training, 2) Software development, 3) Licensed ISP, and 4) Hardware engineering and maintenance. ICT is its primary revenue gen- erator, and the UCC offers both basic courses in Microsoft applications and ISDL certification, as well as tailor-made courses driven by the government and private sector's ICT needs. Its "one student, one computer" motto and training model have been very successful, and has forced competitors to invest in equipment and hardware. The UCC illustrates both the importance of responding to the demand needs and customer preferences within the ICT market, as well as the potential return on investment of education and training. By customizing their offerings, UCC has grown its client base and generated US$ 2.5 million in 2004, with an expected US$ 3.5 million in 2005. Additionally, the increased earnings potential of UCC graduates is significant. An estimated 100% of UCC students who complete the CISCO training course, have a job at the time of graduation. The course costs $600, and starting salaries upon graduation are between $250 and $300 per month. Source: OTF Group Interview 43 Ibid 32 . Improving competitiveness in Tanzania 5 MIGRATING Leaders in developing countries such as Tanzania live and work in environments where literally TO MORE everything needs urgent attention. Despite such a context, the need to focus on a few industries in order to drive sustainable economic transforma- SOPHISTICATED tion is paramount because--by definition-- human, financial, and institutional resources are PRODUCTS AND scarce in developing economies. Focusing these resources on ensuring that two or three local SERVICES industries get the support they need to compete successfully and generate income and employment is essential. In the case of Tanzania, where many local industries present real income generation and employment opportunities, the need to com- bine support for a few nascent industries with a focused and targeted investment promotion effort, 5.1 TWO PATHS: KEY is crucial. Neglecting to see this reality and diverg- ing time, effort, and resources on too many indus- SECTORS AND SMES tries may be today's greatest mistake in economic development. At US$ 322 per capita income in 2004, Tanzania is one of the poorest countries in the world. If Today's successfully developing economies have Tanzanians were to enjoy middle income status followed a similar process. They became world or of, say US$ 1,000 per capita income by 2013, regional leaders in a few industries. They generate Tanzania's GDP would need to grow from US$ income, expertise, and respect in a first set of prior- 11.1 billion to US$ 43.0 billion in 10 years. ity industries, and then invested their surpluses, incoming investment, and leadership capacity in To achieve such accelerated rates of economic other strategic industries. Mauritius does well when growth, two sets of economic objectives must be its tourism, textile, and sugar industries do well. pursued. First, Tanzania must invest in developing Today, its position in the textile and sugar indus- the competitiveness of key, high-potential indus- tries is being compromised, but thanks to years of tries. Second, Tanzania must develop support pro- success in its core industries, Mauritius has the grams and initiatives that broadly strengthen SMEs resources and credibility to establish strong posi- and the private sector overall. In essence, the dou- tions in other areas such as offshore banking and ble impact of key industries `pulling' the rest of the ICT services. Other successful formerly `developing economy, while business development services are economies' have followed similar paths, including `pushing' the capacity of the private sector to Korea, Singapore, Tunisia and Malaysia. respond to more localized opportunities is the kind Tanzania--as well as most other African of virtuous mechanism that Tanzania needs to drive economies--could learn a tremendous amount by a significant expansion of its economy. studying their experiences. It is important to clarify the first objective in light In parallel with competitiveness-building programs of the many failed industrial strategies throughout for key local industries, Tanzanian leaders should the developing world. This is not a process where also launch initiatives to strengthen the private sec- a few government officials pick economic winners tor overall with a particular focus on SMEs, the and losers. Competitiveness rests on a productive segment that now needs the most support. Private public/private sector dialogue and, ultimately, on sector and enterprise development relies heavily private sector leadership of the economy with on the competitiveness of key local industries. In informed government support. This objective is Africa, as in many other parts of the world, three centered on making necessary tradeoffs, tradeoffs primary categories of local enterprises exist. The that can only be made effectively within this type first category directly drives the main industries of of collaborative environment. the economy. These would include hotels, lodges, Executive Summary . 33 or airlines in a tourism-dependent country. The SMEs mentioned above can benefit from ICT. second category supports larger businesses operat- Finally models are discussed that can provide SMEs ing in key local industries; these could be spare-part with the support required to drive the economy. manufacturers or leather seat suppliers to a strong local automotive cluster. The third category refers to SMEs that indirectly benefit from the increased purchasing power that successful local industries 5.2 AGRIBUSINESS are generating. These would include local retail and grocery stores in a prosperous mining town or To be clear, this section is labeled "Agribusiness" neighborhood restaurants and bars in a thriving and not "Agriculture" to reflect the role of agricul- coffee export town. ture as inputs into more sophisticated products. This perspective highlights the need for strong By building strong key industries, therefore, linkages between the agricultural inputs, the manu- Tanzanian leaders will also be driving private sector facturing processing centers and final markets. and enterprise development directly or indirectly. In Increasing agricultural productivity goes hand in order to achieve such diffused impact, however, hand with increasing productivity in manufactur- Tanzanian leaders will also have to strengthen the ing. Success in the agribusiness sub-sectors ability of local SMEs to exploit the opportunity that described in this section will rely on increased this type of growth presents. One mechanism is to value-added processing/manufacturing. develop a stronger platform of institutions that can provide business development and incubation support In general, agribusiness exports now rely on very services. ICT are critical to these types of initiatives, limited processing of the primary agricultural as well, in two primary ways: 1) the dissemination of goods. Raw materials such as coffee, cashews, assistance and information to a large number of geo- tobacco and tea are exported to international mar- graphically dispersed SMEs and 2) facilitating the kets with very little value added to them. Total upgrade in SME efficiency and strategic capacity. exports in these major sectors have also decreased, as shown in Table 4, partly due to global trends ICT's role in the migration to more sophisticated in commodity prices. Agriculture, Fishing and products and services, therefore, is vital to both key Forestry constitute a significant portion of sectors and SMEs. But these technologies will only Tanzania's economy, responsible for 40­50% of be effective once competitive strategies are in place. GDP. Thus, strategies that increase the value of The following sections begin with a discussion of these sectors' products could have a particularly each sector's strategic challenges and opportunities. broad impact on the prosperity of Tanzanians. Then, within this context, targeted ICT interven- tions are proposed as part of the process of sector An overarching agriculture strategy, however, is not upgrading to differentiated products. ICT interven- what is required. Specific strategies are needed for tions are discussed first at a firm or value chain level, according to the Heeks & Duncombe* Five Levels of ICT Usage model, and then at an industry TABLE 4. Selected exports compound level, addressing the systemic ICT needs and oppor- annual growth 1998­2003 tunities. In most cases, recommendations center on upgrading to the next level of value creation. This Product CAGR should be interpreted as the next step in a long Cashews 47.7% process, and not as a final destination. In identify- ing ever higher value markets, and developing ever Coffee 8.2% more specialized processing abilities, increasing lev- Tobacco 1.2% els of ICT usage will continue to be fundamental. Following these sections will be a broader discussion Tea 0.01% of the SME environment grounded in ICT adop- Cotton 10.3% tion data and an analysis on how both types of * Adapted from "Information, ICTs and Small Enterprise Findings From Botswana", Richard Duncombe and Richard Heeks, Institute for Development Policy and Management, University of Manchester, Manchester, UK, 1999 34 . Improving competitiveness in Tanzania each commodity within the agricultural sector. An informed industry strategy, targeting a specific FIGURE 23. World coffee price index consumer market, can create sustainable advan- tages for agricultural products. Analysis and possi- 2.00 Annual Average C-Price NY ble strategies for coffee and cashews are listed 1.80 below. These provide a starting point for 1.60 1.40 Tanzanians to make and test their own hypotheses 1.20 regarding which markets to target and how best to 1.00 execute distinct strategies. In each of these strate- US$/lb 0.80 gies, ICT can play a role in enhancing productivity 0.60 and facilitating the shift towards more competitive 0.40 0.20 products. 0.00 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 5.2.1 Coffee Coffee, a major component of Tanzanian exports, Source: International Coffee Organization is a sector that can have significant impact on the rural population. The coffee industry in Tanzania provides employment to approximately 400,000 TCB controls the auction in Moshi, cups coffee to families, as 95% of production is grown by small ensure quality and conducts farmer extension efforts. holders.44 Regrettably, it is also a classic example of The coffee industry has taken steps towards access- a commodity sector that has undergone dramatic ing this higher priced specialty coffee market, which upheaval in recent years, as seen by the price include the investment in primary processing sta- volatility in Figure 23. A slight shift in supply, pri- tions, a restructuring of marketing polices to diversify marily driven by Vietnam and Brazil, caused the channels from the state-run auction and the success- market price for commodity coffee to fall spectacu- ful targeting of specialty buyers. larly.45 The effect in Tanzania was a drop in export receipts from US$ 76.6 million in 1999 to US$ Primary processing stations, also known as pulperies 35.3 million in 2002.46 In commodity coffee, or washing stations, are required to treat and triage Tanzania will never be a market-setting country; it coffee cherries in order to produce a consistently will only be reactive, subject to ever decreasing, high-quality green coffee bean. TCB estimates that volatile market prices. there are currently approximately 150 of these pro- cessing stations in Tanzania and has a national goal Fortunately Tanzania recognizes this fact and is taking of reaching 500 stations in the coming years. several positive steps to compete more effectively. In Investment funds are being channeled by the GOT the coffee market, a sustainable non-reactive strategy through the local governments within coffee is to improve quality so that sales are not dependent producing districts. on world commodity prices, but focus on higher quality coffee markets, including specialty coffee. TCB recently agreed to allow the direct export of Customers in these markets are sophisticated and premium, specialty coffee instead of the tradi- demanding, but are willing to pay a premium from a tional channel through state-run auction. This supplier who can consistently provide a very high was an important decision as it allowed producers quality coffee. to market their coffee directly to specialty coffee buyers in the US and Europe. The purchasing National strategy in the coffee sector is coordinated dynamics of the specialty coffee market are very by the Tanzania Coffee Board (TCB). Broadly, the different from those of commodity-grade coffee. TCB's strategy is to increase both production and Coffee in the specialty market is commonly quality, but importantly not to increase production referred to as "relationship coffee," highlighting at the expense of quality. Among its responsibilities, the importance of strong relationships between 44 Baffes, John "Tanzania's Coffee Sector: Constraints and Challenges in a Global Environment" The World Bank, Africa Region Working Paper no. 56, June 2003 45 Coffee prices have recovered in late 2003 and early 2004, but this is due to short term conditions such as the effects of drought in Vietnam the tsunami in Indonesia and lower production in Brazil. 46 Bank of Tanzania, Economic Bulletin Executive Summary . 35 the buyers and sellers/producers. In these rela- Vietnam as the largest producer of cashews in the tionships, buyers require consistently high quality world. Vietnam drove global production to increase coffee that is supplied reliably over time. nearly 70% from 1998 to 200349, see Figure 24. Allowing coffee to be sold directly to these buyers Historically Tanzania's cashew sector has been a gives producers the opportunity to nurture these leading foreign exchange earner, but the sector has relationships. plummeted from its number one ranking in export receipts in 1998. Export receipts have fallen dra- The Association of Kilimanjaro Specialty Coffee matically by 47.7% from 1998 to 2003 from US$ Growers (AKSCG), also known under the trade 101.9 million to US$ 39.4 million.50 Tanzania, name of KILICAFE, was the first recipient of a which was fourth in global production in 1998 and direct-export license for coffee and, as such, was delivered 7.5% of the global supply, has fallen to able to control the sales process of their product. the sixth largest producer providing 4.8% of the They concentrated on increasing the quality and global supply. value of their coffee, and soon sold 38% of their coffee as specialty grade, in contrast to 2% in prior Though reliable price data is difficult to obtain for years. 47 raw cashew nuts, as there is no international indica- tive price for cashews, it is clear that prices in Tanzania have fallen. The decline in export BOX 2 AKSCG - KILICAFE receipts in Tanzania was experienced over a period when production in Tanzania fluctuated marginally (93,200 tons in 1998 to 121,900 tons in 2001 and AKSCG was formed in 2001 by 11 farmer groups in Northern Tanzania, with the facilitation of TechnoServe, a U.S.- based, back to 100,000 tons in 2003).51 The Tanzanian non-profit development organization that assists entrepreneurs and cashew industry must change the way it competes other stakeholders in building competitive businesses and indus- tries that directly benefit poor, rural communities in Africa and in the international cashew market or it will gener- Latin America. The purpose of AKSCG is to locally produce and ate less and less profit from this important sector. internationally market specialty coffee. The association provides In short, Tanzania can no longer afford to send its its members with a range of services, including quality and man- agement training, provision of credit facilities, and marketing sup- unprocessed cashew nuts to India for processing. port. KILICAFE has grown in a short period to become Tanzania's largest coffee farmer association with 80 member farmer groups representing more than 7,000 smallholder farmers In spite of these recent disappointing results, the from three coffee growing regions. cashew sector presents opportunities for Tanzania to capture value through processing and increasing KILICAFE has been successful in both production and marketing. Production of clean, or washed coffee, has grown significantly productivity. Only 10% of Tanzania's production is since inception as shown by the graph below. Building on the currently processed in country, with the balance increase in product quality, marketing efforts have led to much higher prices for members of the organization. In the 2002/3 processed in India.52 According to experts quoted season, members in the southern Tanzania region sold coffee at in the East African newspaper, a viable processing prices 66% higher than non-members in the region, while the northern group's prices averaged 40% higher. Specialty buyers in industry in the country could create 30,000 direct the 2004/5 season include VolCafe, Gepa Fairtrade, Lister & jobs and generate US$ 40 million in incremental Beisler, Peet's Coffee & Tea and Starbucks Coffee. processing revenues annually.53 Unprocessed cashew Source: OTF Group interview, www.KILICAFE.com nuts from Tanzania currently trade at an average f.o.b price of US$ 580 per ton in the world mar- ket, compared with an average of US$ 2,934 for a ton of processed cashew nuts.54 5.2.2 Cashews The world seems to be eating more cashews, given Tanzania has a factory infrastructure, albeit dated, that cashews took first place in world nut crop pro- that it can leverage to enact a processing value- duction in 2003.48 But Tanzania is being left addition strategy. The GOT is taking some appro- behind as the market shifted with the entrance of priate first steps in upgrading and privatizing the 47 USAID Tanzania Data Sheet 2004 48 http://uga.edu/fruit/cashew.htm 49 FAOSTAT 50 Bank of Tanzania, Economic Bulletin 51 FAOSTAT 52 Wakabi, Wairagala "African Cashews to Indian Factories: How the Continent Exports its Profits and Jobs", The East African 20 December 20004 53 Ibid 54 Southern Africa Documentation and Cooperation Centre, Business Times article, "Tanzania loses billions in raw cashew exports", 22 August 2003 36 . Improving competitiveness in Tanzania FIGURE 24. Leading cashew producing countries 35 Vietnam 30 25 India output 20 world 15 of Nigeria % 10 Brazil Indonesia 5 Tanzania 0 1998 1999 2000 2001 2002 2003 Year Source: FAOSTAT dormant cashew processing factories that have been owned by the Cashew Board of Tanzania (CBT). BOX 3 Olam International These factories range from those built in the 1960s to early 1990s. One reason that these factories were Olam is a global supply chain manager of agricultural products and food ingredients, operating in over 35 origin countries and abandoned was that the type of technology used 50 destination markets. Olam focuses on fourteen products, caused too much breakage. Processing cashew nuts including cocoa, coffee, cashew, peanuts, sesame, rice and teak wood. Within the cashew industry, Olam is the single largest involves breaking the cashew nut to access the ker- integrated global player and has been operating in Tanzania nel, the edible portion. Kernels that are broken since 1994. during the breaking process are less valuable than Olam's processing capacity in Tanzania is 10,000 tons per year, unbroken kernels. The loss of value per break is which it satisfies with local supply. Its factory is certified for organ- approximately 50%. ic cashews with traceability to over 1,500 individual farms in five villages. Olam provides the farmers with inputs who sell at desig- nated buying centers. Each farmer registers with Olam, and the There are two dominant technologies employed to cashew nuts from farmers are kept in discrete lots at the factory. break cashew nuts. One is a capital intensive, Buyers of processed cashew nuts from the EU and USA depend on accurate traceability information provided by Olam and its certify- mechanized process that is primarily used in Brazil ing agency, which keeps a database of Olam's product. and was installed in the older Tanzania processing Source: OTF Group interview, www.olamonline.com centers. The second is a much more labor intensive procedure, but leads to less kernel breakage, result- ing in a higher value product. Newer processing centers in Tanzania use this methodology; one firm that does is Olam International. This technology cashews from Tanzania before they export them to seems to be the most appropriate for Tanzania as it western markets. The upgrade of technology, how- not only results in a higher value product, but is ever, will require investment in skills training for also labor intensive, creating much needed employ- workers in the processing centers. ment. If the old factories are refitted with this technology it would be possible to process almost 5.2.3 Targeted ICT Interventions all of the country's production given the current The current role of ICT in the local coffee and production levels and installed capacity. If processors cashew industries is limited, although there are in Tanzania are as efficient and productive as some opportunities in production and marketing if Indian factories, they can be more cost competitive. Tanzania continues to support the movement into Indian importers now have to cover transport costs high quality markets for coffee and value added (approximately US$ 100/ton) to ship unprocessed processing for cashews. Some of the ICT solutions Executive Summary . 37 suggested for coffee and cashews may cross over to for example, highlights traceability of its Tanzanian other industries. Applicability will depend on the factory in corporate marketing. Generally, traceabil- dynamics of domestic production and the interna- ity refers to "the ability to document a history of tional market. the origin, participants, steps and handling involved in the production of a food or feed product".55 The need for traceability has been driven by consumer Production demand. Consumers are increasingly interested in As mentioned earlier, processing centers are required the source of their food, as evidenced by the spike in order to produce specialty coffee. Though ICT in consumption of certified organic agricultural systems are not necessary in the operation of the products. More sophisticated consumers, distribu- pulperies themselves, ICT can add value before the tors and regulatory bodies also want information pulperies are even constructed. Tanzania's neighbor, on agricultural inputs such as fertilizer and pesti- Rwanda, has followed a similar quality framework for cides, as well as processing and transformation pro- developing a competitive coffee industry. Rwanda is cedures. Incidents such as Mad Cow disease high- using ICT to strategically locate coffee washing light the need for traceability in order to prevent health hazards from spreading across geographical BOX 4 Embedding technology in a regions. The global debate on genetically modified coffee bean foods (GMF) is another driver in the proliferation of traceability systems. In an effort to assist in the implementation of the national strategy, a geographical information system (GIS) effort was initiated ICT can play a major role in each of these demand among The University of Rwanda, The Ministry of Agriculture, the components of traceability. Electronic tracking sys- Coffee Board and the PEARL project. The GIS has two primary objectives: first, to ensure the strategic placement of the new tems for products are currently in use, and with the washing stations near coffee producers, water sources, electricity appropriate software, much product data can be and effective means of transportation. It is important when deter- captured and transmitted to ensure that members mining a site for a coffee washing station that it be located near a critical mass of coffee farmers as coffee cherries should be of the product supply chain can communicate for- processed within a few hours of harvest to achieve a high quality wards and backwards. GIS systems such as the one product. The processing operations also use a large amount of water that must be of a high quality. Lake and river water are Rwanda built, as well as global positioning systems not appropriate, so it is critical to be located near a quality (GPS), allow data to be accessed remotely if it is source. Finally, to keep power costs low, washing stations should be near a reliable power supply, preferably the national grid not tracked through other means. rather than expensive generators. The second objective is to use soil and weather data to customize fertilizers and plan the replacement of older coffee trees with new Marketing varieties. Though a small country, Rwanda's mountainous topog- The role of ICT in the marketing strategy of a firm raphy and intensive farming practices lead to a wide variation in weather patterns and soil composition. To optimize yields, appro- depends heavily on the market dynamics and priate fertilizers should be applied to certain areas. As the sophistication of the industry. For example, players Coffee Board and the University of Rwanda test and develop new varieties of coffee trees, the soil and weather data will determine in the local fishing industry in Zanzibar use their where certain types should be planted. cell phones to compare market prices in Zanzibar and the mainland to determine where they should Source: OTF Group interview, www.cgisnur.org sell their catch.56 The daily harvesting and almost immediate perishability of their product requires stations throughout the country. As the GOT sup- these producers to know market prices instantly to ports the construction of washing stations through capture the benefits of unfilled market demand. funding to coffee producing districts, Tanzania could benefit from a similar sighting approach. ICT can be instrumental in website marketing in more sophisticated industry strategies, though not Traceability has become an increasingly important necessarily for direct sales to export markets. For issue for global agribusiness and will increasingly example, in the specialty coffee market it is impor- apply to both coffee and cashews--note that Olam, tant for producers to build strong relationships with 55 USDA Advisory Committee "Global Traceability and Labeling Requirements for Agricultural Biotechnology-Derived Products: Impacts and Implications for the United States 56 Click Online, Tanzania Mobile, BBC World, July 26, 2005 38 . Improving competitiveness in Tanzania importers and roasters. These importers and roasters ing model, since it would potentially involve more are not inclined to purchase supply over a website, players. Under the scope of work of the East African but end use consumers of higher end coffees do Fine Coffees Association (EAFCA), there might be place a premium on knowing where their coffee an opportunity to link the auctions of these coun- comes from and whom their purchase benefits. In tries via the internet. Doing so would increase market an effort to better inform consumers, coffee produc- efficiencies across East Africa, allowing producers in ers in Tanzania could develop websites with local one country to supply product to buyers in another information on the benefits their coffee has provided country in which there may be a shortage, either to communities. Coffee drinkers could see photos locally or beyond African borders. of the producers, coffee washing stations that process the coffee, discover the region in which the ICT can also play a role in assisting Tanzanian agricul- coffee is grown and learn how the coffee strategy tural industries and firms in conducting market has increased the prosperity of coffee farmers and research. The shift to processing cashews, for example, the people employed by the coffee washing stations. should be informed by the identification of demands in markets that Tanzania has not historically targeted. The specialty coffee market is moving towards This also applies to new entrants into the specialty adopting an appellation framework similar to that coffee market, as well as any agricultural sector that is of the wine industry. Coffee connoisseurs can looking to move up the value chain or target new detect flavor variances from relatively small geo- buyers. As the market segments that could be served graphic regions. This is due to differences in soil may vary, so will their preferences, which should composition, rainfall, plant varieties and other cli- determine how the industry organizes its activities. matic or input factors. To satisfy these very Since retailers from developed countries are most demanding customers, specialty coffee marketers interested in traceability of products, long-term rela- will increasingly be identifying the unique charac- tionships, and product range from their suppliers,57 teristics of coffee growing areas and highlighting for example, effectively serving these markets means this information when communicating with con- that Tanzanian suppliers must be able to meet these sumers. The GIS and traceability systems illustrated criteria. If they do not, other countries will. in the Production section can be utilized by coffee growers to ensure that this differentiating informa- tion is provided to buyers. Though a window of opportunity has been opened 5.3 MINERALS AND MINING for specialty coffee producers in Tanzania, the vast majority of coffee sales still take place in the state- Tanzania is endowed with vast mineral wealth: gold, run auction in Moshi. Even for the progressive base metals, diamonds, ferrous minerals and a wide KILICAFE, the primary channel for coffee sales variety of gemstones, including tanzanite, a stone only continues to be the state auction, which accounts found in Tanzania. The nation's exports are increas- for 90% of sales. The auction process is relatively ingly dominated by mining, growing from US$ 73 simple. Buyers are sent samples of coffee that are million in 1999 to more than US$ 540 million in cataloged and displayed on an electronic board. 2003, a remarkable compound annual growth rate of Buyers make bids by pressing buttons, which 49.2%. This is more striking when compared to the increase their bid price by US$ 1.00, 0.60 or 0.40. compound annual growth of 4.6% for all other exports combined over the same time period. At this time there is no internet-based auction pro- cedure and constructing one for use solely in The mining and minerals sector experienced a boom Tanzania may not make economic sense. An elec- in growth in the 1980's when the State Mining tronic-based auction that linked nearby countries Company (STAMICO) monopoly was broken up, that also use auctions such as Kenya and Ethiopia and any Tanzanian was allowed to register a claim could lead to a more efficient and competitive pric- and sell minerals.58 The minerals sector then benefited 57 OTF Group research 58 Economic and Social Research Foundation and International Business Initiatives Corp., Tanzania's Precious Minerals Boom: Issues in Mining and Marketing. January 2001. Executive Summary . 39 TABLE 5. Minerals production Selected Minerals (tons thousands) 1999 2000 2001 2002 2003 Gold 4,890 15,060 30,088 34,065 45,299 Diamonds (carats) 235 354 254 213 237 Gemstones 95 151 97 113 n/a Source: Planning Commission -Economic Survey, Bank of Tanzania - Economic Bulletin & Economic Survey 2002 from legal reform and the adoption of a revised min- ing policy in 1997 that included domestic and export migrate to more complex and valuable products, tax relief and investment incentives. This revival has thereby increasing its global competitiveness. The helped Tanzania become the fourth largest gold pro- next section explores this potential. While gemstones ducer in Sub-Saharan Africa, behind South Africa, account for a smaller percentage of exports than gold, Ghana and Mali.59 Gold has been the main driver of the value of this sector is likely to be greatly underes- mineral exports, as evidenced by Table 5. Gold timated due to smuggling, and holds significant exports in 2003 were US$ 499 million and produc- opportunity due to imminent regulatory changes. tion grew at an astounding compound annual growth rate of 56.1% from 1999 to 2003.60 5.3.1 Gemstone Industry Outlook Tanzania possesses a rich portfolio of gemstones, Though the government's role in jumpstarting the including tanzanite, ruby, sapphire, rhodolite, mining sector is a positive example for other sectors, emerald, amethyst, chrysoprase, peridot and tour- the sector's spectacular performance in terms of maline. Tanzanite is of particular interest, as it is exports and production must be tempered against unique to Tanzania and of greater value than most its impact on increasing the prosperity of the aver- gemstones. In 2001, tanzanite export receipts only age citizen. The GOT has certainly benefited from totaled $16 million. This number greatly underesti- the increase in mining with tax revenues from the mates the value of the tanzanite industry, as ram- sector growing from US$ 4.5 million in 1997 to pant smuggling is alleged to account for 90% of US$ 51.7 million in 2002.61 In order for mining to production export.62 Abolishing high export taxes benefit the general population, however, the indus- could help to greatly reduce this amount. try must move up market and include new products and services with greater value added to them. This figure of $16 million also grossly underesti- Countries only benefit from extractive industries mates the potential value of the gemstone industry. such as mining for the fixed period of time equal to The majority of gemstone exports consist of uncut the country's reserves, leaving ghost towns through- stones, absent lapidary or jewelry creation. Due to out the countryside when the wells or in this case the long history of illicit trade in gemstones, it is mineral veins run dry. The increase of Tanzania's difficult to estimate the total market value or mining receipts has also overshadowed the urgency potential. However, experts speculate that currently of effectively growing other economic sectors. at least 60% of Tanzania's gemstones exit illegally overland, mainly through Kenya, for lapidary and In thinking about strategic reinvestment of mining export.63 Figures such as these confirm that revenues, the gemstones sector is an area within the Tanzania's gemstone strategy only captures a small Mining & Minerals industry that has the potential to percentage of end products' value. 59 The Economist Intelligence Unit, Country Report: Tanzania. London, UK. February 2005 60 The Economist Intelligence Unit, Country Report: Tanzania. London, UK. February 2005 61 Mwalyosi, Raphael B.B., "Impact Assessment and the Mining Industry: Perspectives from Tanzania, April 2004 62 Block, Robert and Daniel Pearl. "Underground Trade. Much Smuggled Gem Called Tanzanite Helps Bin Laden Supporters." The Wall Street Journal, November 16, 2001 63 Economic and Social Research Foundation 40 . Improving competitiveness in Tanzania The current dynamics present tremendous oppor- Box 5 for a further explanation of this potential tunities for forward integration. Stable diamond, dynamic. tanzanite and other gemstone extraction could be coupled with the gold mining to build an interna- The present Tanzanian lapidary sector is extremely tionally competitive lapidary and jewelry industry. small relative to the size of the industry, and is Unlike gold mining, gemstones in Tanzania are focused on cutting larger gemstones (1 gram or primarily sourced by small-scale operations that greater). The majority of lapidary takes place in are domestically owned, so the benefits of success- Arusha, with stones then exported or sold on the fully moving up the value chain in lapidary and local market. Although the compensation for skilled jewelry would have a significant impact on the lapidary is competitive, there are few workers local population. trained in the lapidary process, as local demand is limited. With the regulatory change in 2006, the 5.3.2 Gemstone Lapidary local market may grow exponentially. This would The first step in migrating towards higher value help to fuel demand for skilled labor and legislate products is investment in the lapidary industry. an opportunity for Tanzania to forward integrate in More than 98% of gemstones exported through the the value chain, capturing--rather than exporting-- formal economy are shipped in uncut form to over- the wealth created in the lapidary process. seas lapidary and polishing centers. India receives over 50% of these exports. This massive exporta- tion of rough stones is not driven by a lack of local expertise or in a desire to capitalize on low-cost labor overseas. This trend is directly attributable to BOX 5 The Gemological & Jewelry the regulatory environment. The current Value- Vocational Training Centre Added Tax on gemstones is 20%. Thus, it is advan- tageous from a tax perspective to ship uncut stones The Gemological & Jewelry Vocational Training Centre (Centre), located in Arusha, is Tanzania's current hub of gemstone lapidary of undetermined and presumably low value, rather and training. The Centre offers a three-month course in lapidary, than export cut, higher value stones.64 This VAT during which students are first taught to cut the low value stones. Subsequently, those who acquire adequate skills, are trained to policy is partially responsible for the lack of devel- cut tanzanite. The average size of stones cut in the Centre is 1 opment of a sophisticated local lapidary sector, as gram. The majority of lapidary performed by the Centre is com- there is little demand for such expertise prior to missioned by dealers who buy stones from the miners, and then once cut, sell them to jewelers. export. This regulatory environment also creates disincentives for transparency and compliance, In an average year, the Centre trains 150 students, who pay US$ 350 each to participate in a 3-month course, after which they can which contribute to gemstone smuggling. command a salary of US$ 150 ­ 500 per month. The return on investment in training is substantial, but the industry has historically outsourced much of the lapidary process to India and Sri Lanka to A law due to take effect in January 2006 will ban avoid heavy tariffs. However, with the change in legislation, as of exports of rough gemstones, a move that will effec- January 1, 2006, market demand will increase exponentially, cre- tively legislate forward integration. This regulatory ating tremendous need for labor skilled in the lapidary process. change can act as a catalyst in stimulating significant This presents a prime opportunity for the GOT to encourage for- growth in domestic lapidary. A hint of caution must ward integration in the mining sector by providing strategic support to the lapidary sector. The Gemological & Jewelry Vocational also be mentioned in regards to this legislation. Training Centre is well positioned to assume a leadership role as Though the government is well intentioned and has the lapidary sector expands domestically next year, and to play a an important role in jumpstarting a nascent industry, significant role in training and facilitating the forward integration process in the mining industry. However, substantial human and laws such as these can have negative consequences. financial resources will be required to expand lapidary capacity For instance, although the lapidary industry may and accommodate growing demand. Unless the GOT supports industry leaders and associations such as this Centre and TAMIDA benefit, it is also possible that the amount of rough in rapidly improving the expertise and size of its lapidary labor gemstones smuggled out of the country may increase, force, the potential of this sector will be squandered, further encour- aging elicit export of gemstones for lapidary outside of Tanzania. as firms are not able to adequate prepare for this for- ward integration, and even fear bankruptcy as their Source: OTF Group Interview main business exporting uncut stones disappears. See 64 Yager, Thomas R. "The Mineral Industry of Tanzania," U.S. Geological Survey Minerals Yearbook, 2003 Executive Summary . 41 FIGURE 25. Tanzanite value chain and potential ICT improvements Gemstone Value Chain Mining Lapidary Jewelry ICT Interventions E-procurement for sourcing Electronic recordkeeping for cost CAD CAM software to enhance equipment & spare parts reduction in accounting jewelry design Satellite communication between Tracking system to trace origin of ICT-enabled training courses to mine and HQ: mobile fax and gemstones and prevent boycott rapidly increase the pool of telephony skilled jewelers Electronic database networking to Electronic materials to facilitate share technical, maintenance and dissemination of training E-commerce to reach new management specialists materials customers and markets Source: The OTF Group. 5.3.3 Targeted ICT Interventions from officially sanctioned sources. This traceability As we see in Figure 25, ICT could help to transi- system centers on the licensing of all tanzanite min- tion Tanzania to a larger scale lapidary and jewelry ers and dealers, as well as the issuance of "certificates industry that captures greater product value. of origin" to verify that each piece of tanzanite comes from legal sources. Traceability Systems Improving traceability and transparency is one area Tanzanite, one of Tanzania's most precious gem- in which ICT could enhance the gemstone sector. stones, recently came under fire when in 2002 The Electronic databases and cataloging could ease the Wall Street Journal alleged that it was being smug- burden of accounting for all transactions and has- gled illegally to Dubai and other locations to ten the bureaucratic process. finance Al-Qaeda's terrorist network. While these allegations were never verified, the collateral dam- age caused by this allegation was devastating to the Computer Aided Design / Computer Aided tanzanite industry. Not only were the country and Manufacturing (CAD CAM Software) gemstone's images tarnished, but Tiffany and Co., Jewelry industries in countries such as Sri Lanka are Zale Corp. and QVC Inc. all decided to boycott taking advantage of CAD CAM models to effi- purchase of the precious stone. The boycotts have ciently produce jewelry for domestic consumption subsequently been dropped, but the need for and export. As Tanzania considers forward integra- greater transparency and accountability in the form tion in the gemstone industry, large-scale training of a traceability system became clear. of jewelers could be accomplished through the use of CAD CAM products such as Gemvision. In response to these allegations and recognizing the Gemvision allows users ranging from the smallest potential for future incrimination, the Tanzanian retailer to the largest manufacturer to create new, government, the U.S. State Department and indus- innovative jewelry designs. The software features a try leaders convened in 2002 to assess the situation jeweler-friendly interface and step-by-step menus and develop possible remedies. The output of this that allow users to design unique jewelry through a meeting was the Tucson Tanzanite Protocols, a doc- quick and simple process. The CAM component ument that commits Tanzanite miners and dealers generates a dimensionally accurate wax model ready in Tanzania and abroad to adopt a system of written for casting the design. Gemvision retails at warranties to guarantee that all tanzanite comes US$6,500 for its core Matrix software, and an 42 . Improving competitiveness in Tanzania additional US$24,995 for the CAM equipment. introduced because of the tourism industry. When The software and equipment could be owned and done poorly, however, tourism exploits cheap labor, shared among cluster-level groups within the sector. provides little or no economic multiplier to the This investment would transform the design and larger economy, and deteriorates the environment. innovation process. Processes that used to take years of training and weeks of labor could be Tanzania's tourism sector is still recovering from the accomplished in a matter of hours. Additionally, effects of the terrorist bombing at the U.S. embassy CAD CAM could be leveraged to create unique, in August 1998, and in comparison with Kenya, distinctly Tanzanian designs that incorporate the South Africa, Mauritius, Zambia and Uganda, its rare tanzanite stone. This differentiation would progress is halting. While each of these countries increase the value of jewelry exports, and could also experienced an increase in arrivals from 1999 to be marketed in conjunction with the local tourism 2003, see Figure 27, Tanzania's arrivals declined industry. from 627,000 to 576,000, a compound annual growth rate of -1.7%. The same is true of country CAD CAM development is an area that could be tourism expenditures, which grew in all of the fruitful for Tanzania's ICT industry, particularly in comparison countries, but declined in Tanzania. tandem with increasing competitiveness in the jew- See Figure 29. elry sector. The GOT could sponsor trainings that would offer gemstone dealers and jewelry makers The one positive indicator for Tanzania is that aver- exposure to the potential of CAD CAM. Once its age stay has grown from 7.7 days in 1999 to 11 days impact on productivity and profit margins is in 2003. This is most likely due to tour operators understood, CAD CAM software demand would adding days in Zanzibar to their travel packages. increase within the sector. This is a powerful exam- This is a sign of progress, although there is a clear ple of the kind of tandem growth that will take risk of Tanzania falling into competition for price- place between sectors pursuing competitive strate- sensitive tourists. The decrease in tourism receipts gies and demand for complex, sophisticated ICT and increase in length of stay have led to a decrease products. in spend per day, from US$ 96.73 in 1999 to only US$ 71.65 in 2003. During this same period, all of Following through on this model of forward inte- the comparison countries grew in spend per day. gration and increased competitiveness, however, Kenya, the recent poster child for a decaying tourism will require close coordination between the public model, has closed the gap with Tanzania in terms of and private sector. Investment and commitment spend per day. In 1999 Kenya's gap with Tanzania in will be required by both parties for activities such spend per day was 45%. This figure has been as training domestic artisans, financing the pur- reduced to only 8.5% in 2003. Tanzania's tourism chase of appropriate technology and the in-depth sector has tremendous potential, as demonstrated by primary market research required to identify under- positive consumer perceptions in Figure 26. But it served market needs and compete in the interna- has been underperforming due to lack of industry tional marketplace. coordination and the absence of a sound competitive strategy. A critical piece of creating and managing competi- tive strategy is the ability to incorporate sophisticat- 5.4 SERVICES ed customer preferences into differentiated offer- ings. This is particularly true in the tourism indus- 5.4.1 Tourism try, where demanding customers are willing to pay Tourism is a significant and promising sector of a significant premium for unique, memorable Tanzania's economy. When done successfully, the experiences that do not involve worrying about multiplier effects of tourism touch every citizen in logistical details. the country: directly through jobs or business opportunities, indirectly through additional pur- 5.4.1.1 Tour Operators chasing power of the local population, or through The number of foreign visitors to Tanzania's new services and experiences available to citizens national parks jumped from 367,022 in 1999 to Executive Summary . 43 FIGURE 26. Customer portrait(tm) of the traveler to East Africa Is very safe Has great wildlife Tanzania 5.5 Tanzania 6.2 Kenya 5.1 Kenya 6.1 Uganda 4.7 Uganda 4.9 Rwanda 4.0 Rwanda 4.4 1 2 3 4 5 6 7 1 2 3 4 5 6 7 Is very distinctive among African destinations Is a place to rest, relax and relieve stress Tanzania 5.0 Tanzania 5.4 Kenya 4.9 Kenya 5.0 Uganda 4.4 Uganda 4.5 Rwanda 3.7 Rwanda 3.9 1 2 3 4 5 6 7 1 2 3 4 5 6 7 Source: OTF Group Tourism Visitor Quantitative Survey 2002 n=225 500,266 in 2004. This number is significant, not organized tour, the most popular of which is a five- only as a dramatic increase, but as one which day, four-night visit to the Serengeti, with an aver- occurred despite a general decline in arrivals. The age cost of US$250/day.65 Based on interviews con- majority of these visits take place as part of an ducted in Tanzania, this rapid increase is most likely FIGURE 27. Inbound tourism arrivals 7,000 6,000 5,000 4,000 3,000 Thousands 2,000 1,000 0 1999 2000 2001 2002 2003 Tanzania Uganda Kenya Mauritius South Africa Zambia Source: World Tourism Organization 65 OTF Group interviews: Takims Holidays, Leopard Tours, Multi-Choice Tours, Hima Tours 44 . Improving competitiveness in Tanzania FIGURE 28. Average length of stay 12.0 Mauritius 10.0 Kenya Zambia 8.0 Tanzania 6.0 Days 4.0 2.0 0.0 1999 2000 2001 2002 2003 Source: World Tourism Organization a direct reflection of the trend towards "packaging" ease and sense of security, books an all-inclusive trip tour experiences, an area that could be promising, through an agent in their home country. Then, but often falls prey to commoditization and price once in Tanzania, all of the accommodation, logisti- competition. cal and entertainment needs are arranged by a Tanzanian-based tour operator on the ground. The general structure of tourists' experience when There is often a huge disconnect between the cus- they travel to Tanzania is as follows: the tourist is tomers and operators until they arrive in Tanzania. visiting Africa for the first time and, motivated by However, the tour operators are a key component of FIGURE 29. In-country tourism expenditure 6,000 5,000 4,000 millions 3,000 US$ 2,000 1,000 0 1999 2000 2001 2002 2003 Tanzania Uganda Kenya Mauritius South Africa Zambia Source: World Tourism Organization Executive Summary . 45 FIGURE 30. Spend per day 140 120 Mauritius 100 Tanzania 80 US$ 60 Uganda 40 Zambia 20 0 1999 2000 2001 2002 2003 the tourism industry, as often they are the primary transactions hinders tour operators and other play- interface with visitors once they reach Tanzania. ers in the tourism industry from greatly increasing spend per day. Tanzania's tour operators are positioned in such a way that it is difficult for them to implement and sustain sound competitive strategies. Most tour 5.4.1.2 Targeted ICT Interventions operators act as business-to-business ground opera- There are several areas in which ICT could play a tors, focusing their marketing efforts and relation- greater role within the tourism industry, particularly ship management on independent agents with regarding customer knowledge, communication, whom they work, mainly in Europe, the Americas, and industry-level branding. Several of the poten- Asia and Australia. Tour operators pay up to US$ tial ICT interventions in the tourism value chain 3,000 to participate in an overseas trade fair, at are illustrated at the industry levels of customer which they meet agents and advertise their services. procurement and in-country experience in Figure 31. The tour operators interviewed did not consider For the purpose of this study, a tourist's process of which market segments or customers they were tar- taking a vacation can be reduced to four primary geting; they leave that component to the agents, steps. The tourist begins by planning a vacation who also set the end price, a figure that is often and considering the various options. The next step unknown to the tour operators. is focused on selection. Based on the options and offerings, which destination and range of offerings There are several areas in need of improvement for does he or she select? The third part is the in-country Tanzania's tourism sector to increase its competi- experience, which encompasses all local activities tiveness. Operators' reliance on intermediaries, due and travel. Finally, post-trip, the tourist processes to limited communication and information capa- the high points and low points of the trip, which bilities, leaves them with very little bargaining inform planning the next trip. power and diminished margins. Limited opera- tional capacity makes it difficult for them to offer customized experiences and forces more generic Broad Use of ICT packages. This leaves operators with little ability to From an operational standpoint, limited use of ICT cater to segments of tourist with higher potential restricts tour operators' ability to target customer margins. With few processes for information shar- segments, move into new markets and improve their ing and communication, operators cannot offer operational efficiency. Without these capabilities, complex product offerings that command a premi- operators are in a suboptimal position vis-à-vis their um. Finally, limited ability to facilitate credit card agents, with little bargaining power and minimal 46 . Improving competitiveness in Tanzania FIGURE 31. Tourist experience and potential ICT improvements Stages of a Tourist's Experience The post-trip experience and memories will inform the planning of subsequent trips. In-Country Planning Selection Post-Trip Experience ICT Interventions ·Web presence to ·Use of cost effective ·Improve information · Post-trip survey attract new customers technologies such as sharing and deployed to refine and build name VOIP to improve communication product features and recognition the speed and reduce between ground overalloffering the cost of customer operators. Improve ·Decide which · Maintain communication logistics to enhance customers to target communication with the quality of the based on market ·Provide complex customers, and experience. knowledge and offerings which are leverage them for explicit customer differentiated from ·Enable electronic referrals and new preferences competitors', and for payment via credit customer procurment. which customers are and debit cards. willing to pay a premium ability to determine price and margins. While most monitor customer behavior and preferences and companies involved in tourism do have websites react to them seamlessly, ICT are at the crux of and email, these platforms for communication are these exchanges. The ability of operational man- underutilized. Incorporating electronic databases agers to incorporate this type of data and react to it and optimizing current means of communication immediately, instead of operating on intuition, will would increase Tanzania's tour operators' ability to provide both new and established companies the control their own pricing and margins, as well as necessary edge to regain a competitive position. greater ability to negotiate with agents. One prime example of enhanced competitiveness through ICT in the tourism industry is Leopard Customer Knowledge and Communication Tours. Leopard is one of the biggest ground tour An important way that Tanzanian firms can leverage operators in Tanzania, functioning with a business- ICT is to increase customer knowledge and commu- to-business model where clients are secured nication. Firms compete and win by embedding through independent agents located worldwide. unique insights about customer needs and prefer- Over a decade ago, Leopard began using email to ences into the products they produce and the chan- correspond with its agents, suppliers, and direct nels they engage to sell those products. These clients overseas. Initially, managers were very reluc- insights, which are a fundamental component of tant to embrace this new technology, and still relied competitive firms and clusters, can only be captured on fax and telephone as the primary means of con- through interaction with and feedback from cus- tact with foreign clients. tomers. The closer a firm is to its end customers, the easier it is to understand those customer needs and Over time, however, as the staff became more to control the processes which create those products. familiar with email and electronic communica- tion, the fax and phone became largely irrelevant. ICT provide a quick and cost effective way to Leopard quickly realized that this transition to maintain communication with potential and previ- email-based correspondence led to a cost reduc- ous customers. From online surveys of customer tion of over 50%. Its cost for email is now only preferences, to shared systems that allow firms to US$ 2,800 per year. This ICT improvement Executive Summary . 47 encouraged Leopard to investigate other means of Tanzania's "Authentic Africa" is a nature-lover's embracing technology to cut its costs while still destination with a wide array of activities and maintaining close contact with its customers and offerings, friendly people and cultures, the unique contacts. It is now in the process of setting up offerings of Zanzibar and Kilimanjaro, and a range VOIP phone lines. They expect the cost reduc- of differentiated safari experiences. This type of tions to be important, as after the US$ 3,000 brand and channel-building also allows smaller installation, the cost of international and domestic players to leverage ICT without needing to invest calls will be reduced to the cost of their ISP sub- in the creation and maintenance of an independ- scription (~US$ 120/month). ent website and marketing efforts. This is particu- larly true for small guesthouses and tour operators, which can use this channel to gain new customers. Electronic Payment: Credit Card Payment The ability to accept credit cards is still rare in While these ICT interventions do have the poten- many segments of Tanzania's tourism sector. From tial to enhance competitiveness, it is important to lodges and hotels to retail businesses serving the recognize the parts of the industry structure for tourism sector and the tour operators described which ICT is no substitute for human interaction. above, very few businesses are capable of accepting Websites can be a valuable tool in terms of provid- and processing credit card payments. This impedi- ing additional information to customers and build- ment has a direct impact at an industry level, guid- ing the tour operator's brand, but they hold two ing how tourists choose hotels, tour operators, and important limitations in terms of driving transac- where they choose to make purchases. This issue tion volume. The primary limitation is that most also has a macro effect on spend per day, as tourists tourists coming to Africa want the legitimacy and are limited in their ability to make purchases by the credibility gained through meeting with an agent amount of cash they are carrying. and initiating the transaction in their home coun- try. Booking a trip via a website, particularly if the operator lacks certification, poses significant uncer- Country Web Presence and Branding tainty and risks in the eyes of most travelers. The When tourists consider their destination options, second area of difficulty is the inefficiency of web they generally think in terms of national and communication. While, most tour operators do regional level locations, not individual hotels or tour have some direct customers via their website, this operators. Thus, the issue of Tanzania's image as a only accounts for 5% of their business on average. country, and industry level coordination of brand- Their ability to generate greater volume is con- ing and publicity is essential. A tourism web portal strained by both the demands of corresponding can play a key role in building this brand. It also with potential customers and driving business allows tourists who enjoy planning their own trip to through the website, as well as the difficulty of build an itinerary and more easily secure bookings. securing a booking with a client after correspon- The Tanzania Tourist Board has spearheaded this dence. Often, correspondences that last for over type of initiative, a sign that the industry may be 4 months with potential customers do not lead to ready to embrace competitiveness and commit the a booking, as the competition among operators necessary resources to sector development. online is fierce and often caters to price-sensitive consumers. Tour operators simply cannot generate The Tanzania Tourist Board recently launched a the type of volume through their website which comprehensive tourism portal, with photos, infor- they enjoy when linked to agents. mation on activities, logistics, and links to accred- ited hotels and lodges. The Tourist Board used the Tehsin Takim confronted this reality directly. In site to begin re-branding Tanzania as "Tanzania: his 52 years of operation within Tanzania, Takims Authentic Africa," an attempt to project Tanzania Holidays' director, Tehsin Takim, has learned much as a one-stop tourist destination, not simply the about the tourism business. Like many tour opera- land of Kilimanjaro or the land of Zanzibar. When tors in Tanzania, Takims Holiday relies on out- done well, this type of web presence can have a sourced, agent-based customer procurement for its significant impact on potential visitors' impres- local tours and safaris. On the surface, this seems sions and desire to plan a trip to this destination. like an ideal area for ICT intervention and 48 . Improving competitiveness in Tanzania improvement. In 1996, Takims Holidays created business issue, not a technology issue."66 The its first website, consisting of eight pages at a cost Tanzanian market is not yet attractive for banks in of $600 for the basic design. It has since re-vamped terms of credit cards. However, debit card issuance the website, which can be found at www.takimshol- is a very new phenomenon, one that usually leads idays.com. While website development could in to a credit card market, and can create the neces- some ways substitute for an agent by finding new sary infrastructure for e-commerce. customers, advertising services and securing book- ings, Takims Holidays found that it is in no way In addition to this evolutionary path of financial a substitute for the business it receives through its services, there are several other factors that have relationships with foreign agents. ICT played a made the issuance of credit cards and the emer- supporting role, but could not provide a substitute gence of e-commerce, difficult. for the value of credibility and legitimacy conferred by home-country presence. 1. National Identification System. A more reli- able national system for personal identification One possible route for the development of the is required to identify and register people tourism sector is forward integration into the area within a country. Such a system is necessary to of foreign agents. Some of the larger tour operators ensure that credit can be adequately assessed would likely increase their profit margins and have and so that credit-granting institutions are much greater control over targeting customers if have recourse in situations of default. Tanzania they established a foreign presence as agents. has been in the process of developing a Depending on constraints of cost and scale, indus- National Identification System for the past ten try level collaboration should also be investigated. years, but it is not yet completed. According to Deputy Minister of Home Affairs, John 5.4.2 ICT in Banking and E-commerce Chiligati, the GOT is currently undertaking a The development of a sophisticated financial services feasibility study to assess the implementation industry that can facilitate secure transactions and of this system, which will cost an estimated efficient capital allocation is an essential component US$90 million.67 of an enabling environment conducive to private sec- 2. National Switch. A national switch system tor growth. Tanzania is at a nascent stage in terms of enables connectivity and network among financial services and products. The national bank- banks. It is one of the prerequisites to an ing industry consists of both foreign and local players, e-commerce gateway. Without a national including Citibank Tanzania, Barclay's, Standard switch, a country's ATM machines are not Chartered, National Bank of Commerce, National linked to the same network; similarly, it is Microfinance Bank, Azania Bank and CRDB Bank. impossible to transfer funds to someone who These institutions respond to local market conditions uses a different bank. Standard Chartered and provide products and services appropriate for Bank is currently investing in national switch Tanzania's level of ICT adoption. For example, components. Other banks can join the switch CRDB bank just introduced two new products: SMS and participate in network transactions. banking, which allows clients to use mobile phones to transfer funds and receive balance information, As detailed in Table 6, the cost of website design and and the "Tembo Card", a type of debit card. hosting for e-commerce is by no means exorbitant. For e-commerce, however, a website also needs to One significant use of ICT is to enhance competi- acquire a credit card merchant account and an online tiveness through e-commerce. Yet, e-commerce transaction provider. Those merchants without an only occurs in very limited settings, and generally account in a foreign bank would need a "cardholder Tanzanians are unable to buy and sell items or serv- not present" merchant account. This account obli- ices online. According to managers at Standard gates the merchant to requirements specified by the Chartered Bank, the absence of credit card issuance acquiring bank and is generally seen as a higher risk and emergence of e-commerce is fundamentally "a account and thus carries merchant fees. 66 OTF Group Interview. 67 Lyimo, Karl. "Tanzania Shouldn't Waste Money on ID Cards." August 2, 2005. Available at http://allafrica.com/stories/200508021061.html. Executive Summary . 49 TABLE 6. Costs of website design and hosting Basic, 1 Page Complex Online Website Shop Design Costs US$200 -1000 US$2000 ­ 3000 Hosting Costs/Month $50 $120 International Bandwidth 1.5 gigabytes 10 gigabytes Includes 25 email accounts 5 email accounts & Web-enabled & Web-enabled order forms order forms Total Annual Cost US$ 800 ­ 1,600 US$ 3,440 ­ 4,440 Source: Tanzania Internet Hosting, www.tih.co.tz/hosting.shtml 5.4.3 Future Development of Services Current key issues in BPO development in Sector: Business Process Outsourcing Tanzania are: Low Labor Productivity. Call centers from the The future development of an ICT Business U.S. will only outsource operations if they are Process Outsourcing (BPO) services industry in able to make at least 40% savings margins.69 The Tanzania could be attractive given that experts ability of Tanzania to provide these savings in project the market will exceed US$ 90 billion by BPO will depend on its ability to keep wages 2006.68 However, in order for BPO to be at all fea- low and productivity high. sible or profitable, careful consideration must be High Connectivity Costs. The cost of connec- given to the inputs and enabling environment nec- tivity and outbound calls remains an obstacle essary to foster a competitive BPO sector. for Tanzania in terms of BPO competitiveness. The absolute and relative costs, as shown in The most basic form of BPO is the call center, of Figure 32, are not competitive in supporting call which there are two major types: telemarketing, centers. These costs could be reduced with the which involves outbound calling to sell products and implementation of the Eastern African services, and customer service, mostly inbound calls Submarine Cable System (EASSY) and backbone answering concerns and solving customer problems. infrastructure projects discussed in Section 4. Telemarketing centers receive commissions on sales VOIP could also play a significant role in closed. Typically a call center will purchase call lists, reducing communication costs. and are paid, for instance, by a credit card company Accents. Although it may sound trivial, OTF based on how many credit cards are sold. In the tele- Group learned in its research with Caribbean marketing and low-level customer service industries, firms that a major obstacle faced by telemar- the ability to compete is based on low costs. keting call centers has been the lack of trust from US customers in giving credit card Other BPO opportunities exist that can either information to telemarketers with foreign build upon basic forms of BPO or be developed accents. We anticipate that Tanzania would independently. These include transactions services face this same challenge. such as invoicing and credit/debit card services. Due to growing global threats of terrorism and nat- When considering BPO projects, the ICT sector in ural disasters, an option for some remote locations Tanzania must remain wary of high-profile projects is to become a disaster recovery center for firms that work theoretically, but may be more realistic at operating in more sophisticated markets. a later stage of growth. One such project would be a 68 Zagada Markets Inc., 2002 69 Interview with CEO of Helen IT System in St Lucia, October 12th, 2004 50 . Improving competitiveness in Tanzania FIGURE 32. Outbound calls to the U.S. 1.05 Zambia 1.40 0.64 Kenya 0.90 0.80 Uganda 1.15 0.95 Tanzania 0.95 0.41 Mauritius 0.53 0.23 South Africa 0.26 $0.00 $0.20 $0.40 $0.60 $0.80 $1.00 $1.20 $1.40 $1.60 Off-Peak (US$; evenings/weekends) To US Peak (US$/minute; daytime) Cyber Park. Successful Cyber Parks are self-sufficient "towns" of technology that encompass residential BOX 6 Ebene cyber city - Mauritius neighborhoods, shopping centers, technical universi- The ICT sector in Mauritius took a giant step forward on April 8, ties, research centers, business incubators and tech- 2005 when the President of Mauritius and the Prime Minister of India inaugurated the first Ebene Cyber Tower. The Cyber Tower nologically modern business facilities. Cyber Parks is just one component of Mauritius's ambitious plan to become a rely heavily on the sustained creation of significant Cyber Island and act as an ICT hub for Africa and the Indian Ocean. Ebene Cyber City (Cyber City) is the flagship project of intellectual capital through universities, research cen- the program. The Cyber City is being developed by Business ters, and business incubators, all of which are costly Parks of Mauritius Ltd. (BPML), which also manages other ICT business parks. The presence of the Indian PM was due to the to develop and can take years to build. Given the commitment India made of a US$ 100 million line of credit size and sophistication of the Tanzanian ICT sector, towards the development of the Cyber Island. Tanzania would be challenged to provide not only The Cyber City, which covers 172 acres, has several components. the requisite connectivity, but also the ICT profes- The project's first landmark is the Cyber Tower, a 12 story, 42,274 sionals required to implement such a project. sq. meters "intelligent" building. The Tower targets three sectors: busi- ness processing outsourcing, call center operations and software development. The Tower is linked to the SAEF optical fiber that con- Due to the complexity of design and execution, suc- nects North America, Europe, South Africa, India and other countries. cessful Cyber parks require not only patience and The Cyber City will also include a business zone for the construction commitment from the country's leadership, but also of individual buildings by ICT firms, a commercial center, a business massive investments in infrastructure as can be seen class hotel, a knowledge center for offering educational courses, and a cyber village residential complex. In November, 2004, it from the experience of the Cyber Park in Mauritius. was announced that an administrative section of the Cyber City What is most important for Tanzania to learn from would be turned into a 10 story, 16,000 sq. meters Cyber Tower II the Mauritius example is that economic growth was due to the demand for office space in the Cyber City. first driven by sugar, then textiles and tourism. Once Source: www.e-cybercity.mu these sectors started generating significant revenues, the government was able to push investment in new sectors such as financial services and the Cyber City. local economy, they are also broadly important as a driver of individual prosperity. This section will evaluate current SME adoption of ICT in Tanzania, highlighting SMEs that are now using ICT to 5.5 BUILDING COMPETITIVE increase productivity and efficiency. SMES 5.5.1 Lack of SME Adoption In each of the key sectors previously analyzed, the SMEs in Tanzania have been slow to adopt ICT, role of SMEs is critical. Given their impact on the even as a means for communication. In one survey, Executive Summary . 51 FIGURE 33. Means of communication utilized by SMEs 120.0% 99.2% 100.0% 87.3% 80.0% 60.0% 50.0% Percent 44.9% 40.0% 30.5% 28.2% 20.0% 15.3% 0.0% Face Fax Mail to Based Telephone Delivery Courier Face Postal Personal Computer although 87.3% of SMEs used a telephone for surveyed SMEs in Tanzania have made no invest- communication, only 28.2% used a computer-based ment in ICT.71 Of the sectors surveyed in this medium such as email.70 Indeed, more than 99% of study, SMEs in the tourism sector invested the SME business takes place in person. See Figure 33. most. Since Tanzania's export sectors are competing on simple comparative advantages, SMEs who par- Though it does not indicate effective usage, anoth- ticipate in these sectors compete mostly on a low er meaningful measure of ICT adoption is the cost model. It is not unexpected that the tourism amount of investment in ICT made by SMEs. The sector has made the most investment in ICT since data in Figure 34 shows that more than 76% of it is creating a more complex, value-added offering. FIGURE 34. Annual total investment in ICT 100 90 80 60 firms of # 40 20 12 8 8 0 0 88-1,358 1,552-4,074 4,365-10,816 US$ 70 Kijo-Ringo, Natujwa Daniel, "Impact of Investment in and Utilization of Information and Communication Technologies on Market Extension: Overview of Small and Medium Enterprises in Tanzania", University of Dar es Salaam, November 2004 71 Ibid 52 . Improving competitiveness in Tanzania This data is not surprising. As discussed earlier, meeting in 2004 with Tanzanian SMEs. One issue SMEs can be divided into two broad categories, that many firms raised was the difficulty of market- those that directly support export sectors and those ing and selling their products and services without that are focused domestically. SMEs with a domes- a website. These firms, primarily retailers, use fairs, tic customer focus do not need to connect with exhibitions and conferences to show their goods, suppliers and customers overseas. Given the rela- and interested buyers increasingly wanted web tively low usage of the internet within Tanzania, addresses for additional information, follow-up firms that target domestic consumers may not be communication and confirmation of the firms' able to reach enough buyers to justify investing in legitimacy. The firms had very little information ICT hardware, training and maintenance. With the about a website's set-up and maintenance costs, or low connectivity of Tanzanian society, a compelling who could provide these types of services. need to use email or other computer-based means of communication does not exist. Many of these TDG selected seven businesses for a pilot project in firms are making very rational decisions to not which TDG worked with the firms to design a web- invest in ICT. site with logo, company profile, photos and categories of products, reference Box 7 Tanzania Development Even SMEs that have made investments in ICT Gateway (TDG) SME Website Pilot Project. do not seem to be capturing their full benefits. In TDG recently had a one­year anniversary workshop Figure 35 it is clear that computer investments are to measure the progress of the initiative. The results primarily being used for basic programs. This could have been mixed. There have been some challenges in be attributed to a number of variables: a need for introducing traditional artisan businesses to internet training, the cost of more complex software, limit- commerce. Several did not realize the importance of ed options for software in local content, limited checking their emails frequently and using the inter- access in terms of connectivity, low awareness of net to respond and correspond with customers. As the value of more sophisticated programs--or a discussed in Section 4, online payments are still a simple cost/benefit analysis that, again, does not challenge as wire transfers are the most common form justify the additional investment in the current of payment and can be cost prohibitive. environment. Though an internet presence can open a firm to As part of its mission to understand these issues, international possibilities, small or low price point the Tanzania Development Gateway (TDG) held a orders may not be worth the transaction costs. FIGURE 35. Computer usage Production Control 6.1% Inventory Management 9.5% WWW 14.2% Order Processing 20.3% Email 20.9% Bookkeeping 26.4% Word Processing 45.9% 0% 10% 20% 30% 40% 50% Executive Summary . 53 BOX 7. Tanzania Development Gateway SME website pilot project Marvelous Batiks The shining success of the TDG SME Website Pilot Project is Marvelous Batiks founded by Mrs. Flotea Massawe in June 1992. Mrs. Massawe was admittedly hesitant at first to put her designs on the internet as she feared that others would copy the Marvelous Batiks designs. Today she is doing nearly all of her business online and even has wireless network at home and a person dedicated to checking emails and following up on electronic orders. Sales increased by approximately 30% after implementation, and Marvelous now operates a second shift to satisfy the increase in orders. She ships the order after she has received 50% of payment by wire transfer. Afrika Sana Afrika Sana was established and officially registered in Tanzania in 1993 by two artists ­ Ailinda Sawe, a fashion designer and N.L. Merinyo, a textile designer and multitalented artist. The two artists started to train youths with talent in batik designing on cotton cloth from local factories. This had the dual benefit of getting young people engaged in meaningful employment and supplying local garment makers with beautiful and quality batik textiles. The company specializes in art, crafts and fashion commissions, including special corporate art pieces, official costumes for company executives and musical ensembles, theatrical costumes, campaigns and promotional outfits. Afrika Sana has not seen an increase in sales that it can attribute to its website. The company primarily uses print media in Tanzania to advertise its products. As most of its sales are in the domestic market and not bulk exports, this is the most appropriate marketing tool. Dina Flowers Dina Flowers was established in 1997 and currently has 16 employees with annual turnover of US$ 200,000. The firm provides fresh and artificial (dried and silk) floral arrangements, indoor plant placing and maintenance, landscape design and implementation, and gener- al garden and lawn maintenance. Dina Flowers has been contacted by a few overseas individuals and firms wishing to send flowers within Tanzania, but the logistics of such transactions are difficult. In such cases the wire transfer fee for payment can offset any profit made from the sale. Dina Flowers does use ICT frequently in the domestic market, as it receives 90% of its business over the mobile phone. Buyers frequently send orders via SMS. A digital camera has also proved useful and supplanted the photo album as the preferred method of showcasing products to customers. Source: OTF Group interview, www.tanzaniagateway.org, www.marveloustanzania.com, www.afrikasana.co.tz, www.tanzaniagate- way.org/dinaflowers For instance, a single local artisan selling wood organization and as a way to transact across the carvings or batiques on the internet from value chain. The former often requires a significant Tanzania will have to either pay for--or pass reorganization of work processes: incorporating effi- through--high international shipping costs that cient back up systems, knowledge-sharing databases, larger bulk exporters can spread across many and building seamless communications between units. She will also need to coordinate payment divisions. The latter involves connecting and inte- for the order, since the local Tanzania financial grating with others in the value chain to achieve services structure does not have the capability to better time-to-market, improved coordination of process credit card payments online. Expensive supply and demand, and excellent customer service. wire transfers are typically the payment solution unless the firm has a trusted partner abroad. Improving Logistics to Reduce Costs. Examples of current Tanzanian SMEs using ICT to reduce oper- ational costs may make the strongest case for the 5.5.2 Potential for SMEs to Build technologies' potential utility. Scandinavian Express Competitive Advantage Through ICT Services Ltd., is a passenger and cargo bus company A domestic or international focus is not always the based in Dar es Salaam. The firm operates a fleet of primary driver of ICT benefit, and there are many 70 buses serving 18 points of destination. domestic-focused SMEs that have successfully used Destinations include cities in all zones in Tanzania, ICT. One factor in determining ICT benefit is the as well as in Kenya, Uganda and Zambia. Passengers degree to which the business model rests on outper- can choose from a range of three service classes. forming competitors on levels of service and com- plex logistics. There are many ways in which ICT Since 2003, the firm has implemented a series of can improve logistics, and the greatest benefits ICT initiatives to improve customer service and accrue when they are incorporated both inside the logistics. The first was to computerize the ticketing 54 . Improving competitiveness in Tanzania system in Dar es Salaam. Using this system, manage- comprehensive database using Paradox, which it ment can run passenger counts on each route served eventually migrated to Access. The database tracked to determine occupancy and utilization so that distribution, delivery, sales, returns and banking. schedules can be adjusted accordingly. The firm Upon implementing this new system, The Business plans to connect all of its regional office to the sys- Times realized savings of US$ 200-300 thousand tem in the near future. This year, Scandinavia also per month. It now has a dedicated staff of five peo- introduced an online booking option for its passen- ple who monitor the system, performing audits and gers, computer logs for fleet maintenance, and GPS reconciling discrepancies. The system has been such systems to save time when locating a bus with a a success that a related company is being used to mechanical malfunction. GPS also helps to keep design similar systems for other local newspapers, drivers on track by making unscheduled stops visible. and the firm is handling the distribution services for other newspapers. Another firm that has used ICT to improve logistics is The Business Times. The Business Times is one of the Making Distance Irrelevant. The distance leading newspapers in Tanzania and has been a first- between buyers and sellers has been a strong mover in the use of ICT in the media industry. As comparative advantage since the beginning of early as the late 1980s, the company set up a network commerce. Historically, the time and effort to monitor the sales of its newspapers. The network required to move physical goods reduced­and covered ten regions in the country and used tele- often prevented - distant competition. As barriers phone lines to communicate. Two factors led the firm to movement and shipment reduced over time, to abandon this early initiative in 1992. First, the the difficulty of coordination and communication cost of using phone lines became prohibitive. Second, still gave the edge to local firms. With the advent the costs associated with maintaining a wide network of global ICT, distance has ceased to become a at that time were high and rising. Since there were major barrier to entry of competition in all but few qualified professionals who could correct mis- the most niche products or markets. ICT have takes, the company incurred large travel expenses to made it easier and less expensive to stay in touch send people to remote regions to make repairs. with customers and suppliers. As firm knowledge becomes an increasing percentage of the `value- In 1999 the company began to again consider how add' of products, ICT also create significant value to implement a cost-effective monitoring system, by allowing the free and instant transport of digi- since it had lost more than US$ 87,000 in the tized knowledge and insight. Geographically Tanga region alone over two years. It developed a based clusters are still fundamentally important BOX 8. Improving logistics & making distance irrelevant: Issam International Ltd. Mr. Suwedi Kassim was well aware of the impact that ICT were beginning to have in Tanzania in the 1990s. Trained as an engineer, he gained his MBA and began competing in the burgeoning ICT sector. He launched three internet cafes, but soon faced fierce competition and decided to shift to the retail sector, using his knowledge of ICT as an advantage. In June, 2000, he began operating Issam International Ltd., a used car and car parts business in Dar es Salaam. In its infancy the business reflected the traditional car sales business model in Tanzania, as Mr. Kassim bought used cars and hired many workers to assist in the logis- tics of the business, ranging from ordering, shipping, clearance and warehousing. Gradually over time, ICT was deployed as a tool to improve logistics and connect the outsourcing of supply worldwide. Mr. Kassim has never been to Japan, but it is where his suppliers are located. Mr. Kassim visits online car and car part wholesalers in Japan to order his supply online. He now researches and tracks his supply from shipping, clearance and warehousing online using the car's chassis number. While competitors of similar size can employ up to twenty people for these types of logistics, Issam International needs only four. On the sales side of the business, Issam International operates a website on which customers can view all the details of his supply. This has greatly increased the reach of the business, not only within Tanzania, but also to other countries. Issam International has sold cars to clients in Zimbabwe, Zambia and The Democratic Republic of Congo. Before developing the website the company averaged 2.93 car sales per month. Since the launch of the website in October, 2003, that number has grown by 96% to an average of 5.75 car sales per month. The business continues to grow steadily. The firm is currently having customized software developed that will further increase effi- ciencies from ordering, shipping, clearance, and warehousing. Source: OTF Group interview Executive Summary . 55 (in fact, some argue even more so). But this allows incubators to be financially viable until they importance is increasingly tied to the intercon- can transition to a more profitable business model. nections of human capital and other supply-side Many incubators eventually drive revenue by provid- factors, versus the proximity to final markets. ing training and consulting to more established firms, An illustration of this principle is found in Box 8 along with their original start-up clients. Improving Logistic & Making Distance Irrelevant: Issam International Ltd. Some African countries have looked at technology incubation centers to bolster their own ICT develop- The examples of these exceptional firms­and many ment efforts. In 2001 Mauritius set up The National others­provide an encouraging contrast to the dis- Computer Board's ICT Incubator Center to help appointing data for ICT adoption among transform entrepreneurship into viable commercial Tanzanian SMEs. The following section focuses on businesses and position the country as a "cyber how support can be given to the private sector to island".74 The center focuses on four areas: e-business, help drive this adoption more broadly. Internet content development, multimedia and bio- informatics, with start-ups generating businesses in e-learning, medical transcription, software and mul- 5.6 STRENGTHENING timedia development, e-tourism and e-commerce, THE PRIVATE SECTOR among others. It operates in partnership with the University of Mauritius, but 60% of its funds come from the government, and the remaining balance is 5.6.1 Incubation driven by rental income. A major part of this initia- The availability of support services and financing tive's success is the public-private cooperation in for start-ups and SMEs, particularly for technology- implementing a coordinated ICT policy. This part- based businesses, is almost non-existent in Tanzania. nership is guided by a clear, broadly shared vision of The failure to deliver funds and technical capabili- what Mauritius wants to become: it will not simply ties to small ventures is usually the largest impedi- be a technically proficient outsourcer for the global ment to their growth. This is a universal problem, economy, but it will also breed its own innovation. 75 and there are successful models throughout the developing world that Tanzania could replicate. 5.6.2 Business Development Services Another initiative that can strengthen SMEs and pro- Business incubation is one way to target and vide support for new business ventures is an invest- strengthen technology-based firms, or any type of ment in Business Development Services (BDS). This new business venture. Business incubators facilitate type of initiative begins with a deep understanding of the process of enterprise development by helping what SMEs in Tanzania need in terms of technical start-ups to survive and grow when they are most assistance. Access to financing is regularly cited as the vulnerable. They provide a range of services, from number one obstacle faced by new ventures in African hands-on technical assistance and access to finance, countries, but other areas of technical assistance also to support services and infrastructure such as office space and communication facilities.72 Although surface: (i) access to information (market information, new processes and technologies), (ii) training on basic their core competency is providing access to SME business skills and management skills, (iii) networking financing, they can be a potent mechanism for with the business community to form partnerships, intermediating venture capital and establishing networks of investors.73 and (iv) consulting on new business plan preparation, among others. Incubators usually partner with academic institutions Tanzania can address these needs by developing an or research centers, as well as government agencies. integrated National BDS Network that can impact By being housed in a university campus, incubators SMEs through clearly defined strategy, operations are better able to target young entrepreneurs and set and partners. The Tanzania Chamber of Commerce, up partnerships with research centers that foster Industries and Trade is a good example of an imple- innovation. Sponsorship by government agencies menting partner within this network. The network 72 NBIA www.nbia.org 73 Based on "Improving Competitiveness and Promoting Economic Diversification in the Caribbean: The Role of ICT". 74 Balancing Act New Update (2004), Issue 186. p. 2. http://ncb.intnet.mu/ncb/incubator/incubator_downloads/balancing-act_186.pdf 75 Ibid. p.3 56 . Improving competitiveness in Tanzania FIGURE 36. BDS services BDS VISION: First point of reference for private sector TRAINING INFORMATION FINANCE NETWORKING CONSULTANT MONITORING Objective: Build Objective: Provide Objective: Facilitate Objective: Create a capacity through and create relevant easier access to stronger regional Objective: Ensure training and information for financing and national quality of consultant mentoring private sector business community services Services: Services: Services: · Maintain database Services: Services: · Basic management · Library of basic of finance sources · Regular networking · Maintain techniques $ materials and terms events database of · Business idea · Custom research $ · Referral to · Mentoring program "certified" elaboration $ · Computer/ consultants for · Bulletin board consultants $ · Technical training $ internet access $ business plan · Showcase · Assist clients to · Regulatory issues $ · Company development speakers and access consultants · Accessing formation and · Connect clients regional BPC and structure financing $ regulatory issues with CEDP business winners engagement $ · Clustering and · Monthly newsletter plan competition · Annual business · Monitor quality of partnering · BRD representative award consultants and · Saving and in each office $ client satisfaction investment Coordination with Partners · Maintain consolidated calendar of events, maintain database of regional services, refer clients to partners, establish regional working group, and ensure continuation of NGO and government projects $ ­potential services to be billed to clients could serve as the first point of reference­or one- clear performance metrics in place to assess an stop-shop--for the private sector. In this manner, it agency's SME interventions, a continuous client could be used proactively to help certain priority feedback system to drive satisfaction and innova- sectors, but would be capable of supporting all tion, and a commitment to some type of "fee for SMEs. As this study has highlighted, agribusiness service" model. This methodology has been sup- will be one of the sectors that initially needs careful ported by the recent work in northern Tanzania of assistance as it transitions to more productive busi- the FAIDA SEP BDS support program, which cites ness models, especially in rural and provincial areas. market distortion of subsidized BDS providers as Given the existence of limited resources, service an issue.76 offerings could also be progressively introduced on a demand-driven basis as determined by the location Coordination with the government, private sector of each BDS (possibly starting with more informa- and the donor community is also crucial to ensure tion provision and moving towards advisory and an effective strategy and to ensure that no replica- training needs). This network can act as the first tion takes place in service offerings or project activ- point of reference for training, information, consult- ities. A partnering scheme should enable a BDS ing services, networking and access to finance, as post to, among other things: (i) maintain a consoli- illustrated in Figure 36. dated calendar of events for SME, (ii) maintain a regional database of services, (iii) refer clients to It is particularly important for BDS sites in rural partners and professional with key expertise, and areas to partner with local human resources, not (iv) ensure continuation of other NGO and gov- only to provide relevant services and solutions, but ernment projects. also to guarantee the sustainability of operations in these provinces. Entrepreneurial expertise will also 5.6.3 An ICT Alliance be a key part of upholding the legitimacy and suc- Business incubation and BDS services can address cess of these institutions. A system of checks and more general firm development, but the ICT sector balances needs to be in place to drive the opera- must also actively engage SMEs and other sectors tional excellence of any BDS post: there must be to foster rational investment in ICT solutions. 76 FAIDA SEP Tanzania BDS Support Project Case Study 2002 Executive Summary . 57 58 . Improving competitiveness TABLE 7. ICT alliance structure National Government Local Government Large Business Small and Medium Enterprise in Tanzania ICT Technology Develop after-sales support Identify economically viable Understand their client's busi- Develop and sell specific and Service programs for National solutions to rapid expansion ness processes better as part of solutions to SMEs that Providers Government agencies of rural telephony and elec- the solution selling process highlight the economic Learn ICT intensive aspect tricity services. Design appropriate ICT solu- value that these solutions of National Government Aggregate local community tions that demonstrate effi- can provide, despite the agency's operations demand for services (e.g., ciency improvements to the overwhelming variety of Develop/introduce cus- local gov't, schools, pay- business SMEs tomized software phones, businesses) to reach Sell solutions to both business Given the overwhelming Set-up continuous training critical mass. leaders and IT implementers variety of SMEs, providers mechanisms Mobilize local government within the firm must develop a "catego- authorities to partly invest in Develop an industry associa- rized" selling approach local telecom infrastructure tion of ICT providers that towards SME by business Develop and sell standard can validate the trustworthi- type (manufacturing, com- ICT based solutions to com- ness of providers bids to busi- mercial services, retail, etc.) mon local government tasks ness leaders To serve SMEs, providers Administration of local popu- must become ICT integra- lation (births, deaths, ID tors, not simple resellers of cards, notary services, etc) individual technology Finance (budgets and capital investments) Outreach (publications, bulletins) Education Coordinate contextual Coordinate contextual train- Develop partnerships with Educators must work with training with service ing with service providers to ICT providers before and providers to develop busi- Basic, Tertiary, and Specialized Training providers to train users train users during the solution selling ness specific public semi- Basic ICT training on PC Usage of basic ICT equip- process to large businesses nars and workshops to and application usage ment for new installations Build long-term, intensive, train similar SME types on Specialized training in the and expansion of existing user and customized training rela- ICT usage use of ICT to improve the base tionships with large businesses e.g., Workshop on MIS efficiency of day-to-day Specialized training in the use Regular group training ses- systems for retail manage- tasks of standard local government sions to new employees on ment Specialized training on ICT solutions systems and applications e.g., Workshop on finan- customize applications One-on-one training for sen- cial software and inventory ior leaders to help them rap- management idly integrate technology into their day-to-day activities Refresher courses as new ICT services are rolled-out within a business Government Review / amend current Set-up task force with Introduce tax credits / incen- Develop a clear strategy Policy, Regulation, and communications protocol providers and educators to tives for employee develop- and policy to facilitate Public to encourage ICT usages identify and roll out standard ment and training as is done investment in ICT by Infrastructure E-mail instead of paper- local government ICT solu- in many countries SMEs based memos tions Reduce import tariffs on tech- e.g., Tax breaks on equip- Electronic data transfer, Remove administrative, nology to facilitate further ment import not paper-based bureaucratic, and legal obsta- investment and upgrading of e.g., Tax credits for equip- Purchase customized ICT cles to rural utility service ICT. ment and training costs solutions to internal opera- expansion Partner with PS associa- tional problems (MIS sys- Identify and agree on stan- tions to provide advice to Executive tems / common PBX sys- dard procedures and ICT SMEs and ratings of ICT tems, etc.) solutions to improve efficien- providers (BBB) cy of local government Summary . 59 The local ICT industry is now relatively weak, together to target different types of users or "con- which affects its ability to create value for the pri- sumers" of ICT. The Alliance would not only target vate sector and SMEs. Profitable contracts and SMEs and larger private sector firms, but it would engagements are usually found in the public sector, also include local and national government. This either through government or donor initiatives. broadens the impact of ICT efforts in a structured There is little demand from the domestic private manner. See Table 7 for a more thorough explanation sector for ICT solutions beyond an internet con- of the roles of each supplier and the benefits they can nection or an off-the-shelf software program. One provide to each member of the Alliance. of the more successful local ICT firms in Tanzania is Soft-Tech. The firm's client list is primarily gov- The supplier coordination is critical to ensuring ernment and donor organizations. that efforts are maximized. For example, SMEs need appropriate ICT solutions that hardware and General access and ability improvements will software providers can provide, but they also need assist in creating a more conducive environment training on new systems and applications. for ICT adoption, as will increased awareness of Government has a role to play in ensuring that ICT benefits. But the key will be to engage SMEs ICT tools are available and that taxes and regula- and the private sector more broadly in a new way. tion are not overly burdensome to SME purchasers, The challenge becomes developing and selling for whom capital is typically scarce. specific solutions that highlight the economic value that these solutions can provide, despite This section has focused on the private sector, the overwhelming variety of firms. One way to including SMEs and distinct economic sectors. The address this challenge is to create an "ICT evolution of a developing economy also requires Alliance." supporting the government, particularly as it shifts roles as the private sector becomes more sophisti- The ICT Alliance would consist of hardware and cated. The next section will focus on the specific software suppliers, ICT trainers and government in role of government in employing ICT effectively-- its role as regulator. These entities can be viewed as not only for its own use, but also in building pri- the "supplier" side of the equation. They would work vate sector competitiveness. 60 . Improving competitiveness in Tanzania 6 THE ROLE OF improvement and investment. First, the GOT must prioritize initiatives within the National ICT Policy, GOVERNMENT: rather than trying to address a prohibitively large range and number of issues simultaneously. As the GOT determines these priorities, ICT initiatives that BUILDING AND benefit targeted sectors need to be at the top of the list. The GOT should also consider expanding the CAPITALIZING ON authority and scope of the office of the National ICT Coordinator to capture the full benefits of ICT poli- MOMENTUM cy coordination and implementation. The breadth of the National ICT Policy is impres- sive, including leadership, infrastructure, legal and regulatory framework, productive and service sec- This report has addressed the role of government in tors, and universal access. This reflects the relatively the new competitiveness model as a partner and advanced state of the professional network. supporter of the private sector. The same holds true Building on the ICT Policy document, the GOT is for government's specific involvement in supporting currently drafting a document entitled "Proposed the effective use of ICT. This section focuses on Implementation Strategies and Actions for the how government can facilitate an ICT platform that National ICT Policy". This document outlines increases productivity for the private sector, as well goals, strategies, actions, agencies and a timeframe as use ICT within the public sector to improve citi- for implementing the National ICT Policy. The zenship and governance. Mechanisms the GOT can political capital invested in this effort is commend- use to promote this agenda include creating an able, but the document and its successor, National actionable ICT Policy, adopting an informed role as ICT Policy Implementation Strategies, need to be regulator and actively using e-government solutions. more focused and actionable. For example, the latter document is detailed and comprehensive, but covers ten different policy areas, each of which includes up to twenty different 6.1 NATIONAL ICT POLICY components of large scale goals and actions. The areas cover all aspects of ICT, ranging from back- Tanzania has taken important first steps in docu- bone infrastructure to ICT company incubation to menting ICT policy, but there is room for policy the dangers of utilizing ICT to view pornography. improvement. On the positive side, Tanzania was These policy areas are not ranked or prioritized, the first country in the East African Community to and the responsible parties, financial resources, and put in place a blueprint to guide the adoption and timeline are not explicit, making action on these development of ICT in March, 2003, through the issues cumbersome. In order to capitalize on the adoption of National Information and momentum and resources invested in formulating Communications Technologies Policy (National ICT policy, the GOT must play a more proactive ICT Policy). The National ICT Policy's vision is for role in prioritizing its goals, actions and resources. "Tanzania to become a hub of ICT infrastructure Without this filtering process, policy will remain and ICT solutions that enhance sustainable socio-eco- unfocused and resources diluted among many proj- nomic development and accelerated poverty reduction ects of unequal value and contribution. both nationally and globally". Furthermore, a necessary addition to ICT policy is a The objective of the National ICT Policy is to pro- more detailed and actionable plan on how the ICT vide a national framework that enables the sector to platform will support the most productive sectors of contribute towards achieving national development the economy. The current strategies focus the discus- goals by exploiting ICT opportunities in a sustainable sion on issues of the ICT sector such as access, ability, way. Digging deeper into the actual impact of gov- training, and regulatory policy, but they do not make ernment interventions reveals several areas of potential the connection to how these changes will advance the Executive Summary . 61 TABLE 8. Productive sector ICT strategies Strategy Actions Prepare a sector specific prioritized action plan to Collect information on e-readiness address e-readiness recommendations Establish, maintain and monitor Take stock of usage of ICT in scope and span e-readiness of all productive sectors in industrial productive sectors Revisit and review legislations to support and Create a mechanism for ensuring sustain development of productive sectors sustainability of productive sectors Encourage expansion and modernization Attract investors in ICT infrastructure of ICT infrastructure to support development productive sectors Establishment of showcases centers Establish technology/ science parks of excellence that will reflect the impact of awareness of ICT. economic sectors that will be expected to employ Communications and Transport. The office lacks ICT to increase productivity and reduce poverty. financial and human resources, as well as authority over other ministries and public institutions that As Table 8 demonstrates, the National ICT Policy's would benefit from deeper ICT expertise. This sug- section on the Productive Sectors and how ICT will gests a lack of commitment. A more appropriate seat help them is vague and unfocused. The following for ICT decision-makers would be within the strategies and actions for the productive sectors are President's or Vice President's Office, where the ICT found in the document: Coordinator would have greater access to resources and greater ability to coordinate cross-ministerial Several broad GOT initiatives, such as facilitating efforts. Increasing resources within the office is a the development of a national payment system and starting place for increasing the efficacy of ICT poli- enacting legislation to provide a framework under cy. However, to truly coordinate ICT policy effective- which business can operate, rest within the outlined ly, a network of ICT professionals should be created strategies. However, to fully realize the potential of throughout Ministries and public institutions. These ICT in the sectors that matter most to Tanzania, the individuals would report to the ICT Coordination GOT and the private sector would need to come Office. Through these institution-level offices, ICT together choose priority sectors, understand these policy and ICT interventions in priority sectors could sectors' unique needs and design the appropriate be strengthened through a bottom-up approach. government support structures. The policy state- ments and economic sector targets in the National Strategy for Growth and Reduction of Poverty clear- ly demonstrate that the GOT understands the importance of adopting these principles of competi- 6.2 REGULATORY REGIME tiveness. Following this model, ICT policy and ICT REFORM solutions for the economic sectors would be informed by the sectors and firms themselves. The Tanzania's communications and broadcasting regu- ICT Alliance could also be leveraged in this fashion lator is The Tanzania Communications Regulatory to assist in the development of appropriate policy. Authority (TCRA). It was established on November 1, 2003, through a merger of the Two of the most important inputs into successful Tanzania Communications Commission and the policy are commitment and coordination. Currently, Tanzania Broadcasting Commission. TCRA now national ICT policy development and implementa- regulates telecommunications, broadcasting, ICT tion is owned by the undersized National ICT applications, provision of postal services, and man- Coordination Office in the Ministry of agement of the radio spectrum. Catching up to the 62 . Improving competitiveness in Tanzania FIGURE 37. TCRA converged licensing framework Network Application Content Type Network Facilities Services Services Service Activity: Construction and ownership of Operation of Reselling or Provision of public electronic communication electronic procurement of content. infrastructure. communication services from networks in order network service to deliver operators. services. Example: Earth stations, fixed links and Bandwidth Internet providers, Satellite cables, public payphone services, virtual mobile broadcasting, facilities, radio communications broadcasting provider, public broadcasting transmitters and links, satellite services, access cellular services, terrestrial, free to hubs, satellite control station, applications VolP, public air TV, terrestrial, space station, submarine cable services and payphones radio broadcasting landing center, switching space segment service, public and other electronic center, cover poles, ducts and service. switched data media. pits used in conjunction with services. other network facilities. Market Segment: International National Regional District Source: United Nationa & american society for Public Administration, "Benchmarking E-government: A Global Perspective". 2001 issues facing regulators worldwide, the TCRA also about improving the efficiency of market economies, needs to be vigilant in addressing new technologies and how changes in regulation can facilitate the and the issues they introduce into the competitive process".77 Regulation and legislation should address landscape. In general, TCRA actions can have a how to create not only a more efficient market for major impact on ensuring that ICT is used effec- ICT firms, but also an environment in which ICT- tively by the private sector. enabled firms can effectively access the benefits of ICT. In this context, regulation revisions being under- As many countries have revised telecommunications taken by TCRA should be measured by how effective regulations due to recent privatizations, the rapid they are in increasing the competitiveness of both the proliferation of ICT has forced policy makers to ICT sector and other economic sectors. The TCRA re-evaluate their current regulatory regimes. Current and its predecessors have been historically slow in pro- debates focus on the convergence of broadcast viding regulation that is appropriate for the commer- media and telecommunications. Convergence refers cial environment. This was the case with VOIP, which to the extent to which similarities exist among IT, was outlawed instead of licensed until this year, telecommunications and other forms of media impeding massive reductions in the cost of telephony. (print and broadcast) in such areas as content, infra- structure and networks. A common example is an One of the first effective steps TCRA is taking to internet site that is operated as a channel of a print address convergence is the drafting and adoption of or broadcast media company. Tanzania has chosen a converged licensing framework. One goal of the the path of convergence, and now needs clear regu- framework is coordination with infrastructure efforts lation as more brick and mortar companies delve so that the excess capacity of the institutionally into cyberspace as part of their business models. owned networks can be leased to customers. The converged licensing framework is also the framework The following perspective on convergence can be found under which VOIP was eventually allowed to be in another InfoDev sponsored report on the issue: offered by ISPs in Tanzania. Under the framework, "ICT and media convergence issues are primarily there are different types of licenses that are being 77 Henten, Samarajiva and Melody, "The Next Step for Telecom Regulation: ICT Convergence Regulation or Multisector Utilities Regulation?", World Dialogue on Regulation for Network Economies, 2003 Executive Summary . 63 issued, see Figure 38. Each type is further divided by government can use ICT to deliver services to busi- market segment and then classified as Individual, nesses and citizens more efficiently and cost-effec- Class or Exempt, depending on the size of social and tively. This is particularly significant in terms of economic impact. The new licensing framework was creating an enabling environment that is conducive designed to simplify procedures, ensure regulatory to private sector growth. The IFC's Doing Business flexibility and efficiency, and encourage the entry of in 2005 Report, cited that Tanzania was among the new operators, applications and services. five nations where government regulation and bureaucracy placed a huge tax on doing business. The next challenge that the TCRA may face is the For example, more than 50% of companies in national infrastructure backbone that will need to be Tanzania reported that more than 10% of manag- built by leveraging the ICT networks now owned by er's time is spent dealing with government regula- institutions such as the Tanzania Railway tion. Customs is one area that is particularly bur- Corporation, the Tanzania Electric Supply Company dened by bureaucratic delay. and Songo Songo Gas Supply. Moving forward, the TCRA will need to create incentives for these stake- As Tanzania begins its evolution towards a more holders and new entrants to engage in commercial developed economy, the GOT will be the primary last mile build-out using existing and new fiber optic actor. As such, any ICT improvement that allows it networks. In many countries, regulation often to cut costs, increase the speed of bureaucratic impedes the ability of networks to talk to each other, processes, and enhance transparency, will generate hand off communications and provide fair priced positive effects throughout the entire economy. The interconnectivity to increase competition. The GOT has been successful in adopting a few early e- licensing and introduction of telecom services in government initiatives that have provided these mainland Tanzania by Zantel, for example, demon- benefits, and its success can serve as a model for strates the positive impact that competition can have continued improvement. on price and service. Future opportunities will exist when Tanzania enhances its fiber optic international Governments can generally use ICT in three broad connectivity through the EASSY project or a compa- ways to improve their service delivery. First, they can rable venture. Similar network collaboration that fos- improve the transparency of the business environ- ters rather than hinders competition through fair ment by making data widely and freely available. competitive regulation should be encouraged. This is achieved through the use of technology such as database management and data entry systems, A final area of regulation that will become increas- which can manage data and increase the speed of ingly important as ICT develops in Tanzania is transactions. This increased transparency means that consumer rights regulation for the internet. This citizens have greater access to information, ranging area includes privacy, governance, advertising and from descriptions of legislation, tax codes, bills pro- related regulatory issues. While these issues seem posed in parliament to interactive transactions, such relatively tedious when compared to issues of infra- as applications for documents and benefits. See Box 9 structure investment and convergence regulation, it for an example of the GOT's investment in this type is important that they be addressed by the TCRA of ICT-enabled service. as, or preferably before, ICT becomes more perva- sive throughout the Tanzanian economy. Second, governments can help the business com- munity become more efficient, make more informed decisions, and spend less time and money on bureaucratic procedures when launching busi- nesses and transacting goods and services. ICT can 6.3 E-GOVERNMENT help government simplify transactions and regula- tory compliance, once the internal commitment for The movement towards digitizing government these changes exists. operations and using the internet as a medium for service provision and extension has been loosely One area that is currently under revision is the gov- described as "e-government." In the same way that ernment's procurement services. Government can private sector firms use ICT to increase productivity, use ICT to efficiently link its own need for services 64 . Improving competitiveness in Tanzania This new system is expected to broaden the base of BOX 9. The parliamentary on-line bidders and suppliers worldwide as firms will be information system able to subscribe to receive tender alerts based on a submitted business profile of the supplier. The The Parliament of Tanzania is established by the Constitution of GOT expects a savings of 60% once the system is the United Republic of Tanzania. The citizens elect the Members fully implemented. Suppliers will be able to view of Parliament (MP) every five years. The structure, composition, powers and functions of Parliament are also described by the summaries of all tender notices for free, but the full Constitution. text of tender notices (full bidding documents) will The Parliamentary overview process generates a large amount of only be available for a set fee, since the government information through different media. Over time, expeditious and generates significant revenues from selling bidding accurate retrieval of this information became a tedious assign- documents. However, bidding documents will be ment. In order to promote good governance, accountability and transparency, as well as to facilitate sharing of large amounts of available for local firms at no cost or at a lower Parliamentary outputs with the general public, the government price than for international firms. Due to the lack designed a robust information system. The Parliamentary On-Line Information Systems (POLIS) is the on-line home of Parliament. of an online payment system, purchasing the bid- The home page has a navigation panel whose links filter infor- ding documents will be done offline via wire bank mation that is stored in different modules of POLIS. The system constitutes the following four integrated modules: transfer, cash or check. Parliamentary Acts & Documents Management System: This mod- Third, investments in information management ule has an inbuilt retrieval mechanism to facilitate the overall management of statutes and other parliamentary documents, and data integration within and between agencies including standards. can significantly improve the quality of government services. This is true for both back office and con- Members of Parliament Profile Database System: The module is used for maintaining bio-data for Members of Parliament. It pro- stituent-facing processes, and can translate into vides a unique solution for searching MP's by name or constituen- powerful cost and time savings. The GOT's use of cy. There are also reports presenting MP's by political party, region, gender and other parameters. ICT to upgrade its payroll system generated an immediate return by catching more than 6,000 Bill Tracking System: This module keeps track of each bill from ghost workers. These technologies can also be used the time of submission to the Parliament to when it is assented. It is intended to show the efficiency of the bill discussion process to communicate across internal boundaries, such as before bills become acts. sharing relevant data about changes in citizen pro- Session Management System: The Session Management module files. When a person registers for an identity card is used to store the distilled Questions and Answers as presented in the Parliament and link them to the assigned unique identifier of the individual MP's. In addition, the system is used to store the decomposed sections of various documents. BOX 10. GOT increases efficiencies in Human Resource Management POLIS has other features such as the ability to create different information views for different users, and to restrict information from some groups of users. There is an offline system, available In 1997, the GOT began upgrading its payroll system, which on CD, which enables users to access archives of selected parts was viewed as outdated, not Y2K compliant, and insecure. The of the system and serves stakeholder segments that are not direct- initiative was a joint venture between The President's Office ­ ly connected to the internet. Public Service Management and The Ministry of Finance. The ini- tial upgrade was a collection of HR data, but continuous Source: OTF Group interview, www.parliament.go.tz. advances have led to the processing of a payroll of 290,000 public employees since April 2000 at a monthly average of US$ 13 million. During the upgrade, the system paid for itself by catching 6,000 ghost workers. The GOT used the correspon- with private sector suppliers, such as the potential ding savings to recruit 600 new teachers, the largest group on plan for a digital market being considered by the the public payroll. Central Tender Board. The Economic and Social Current improvements to the system include new printers that will Research Foundation, the Tanzania Development be able to print statutory payroll information and HR management Gateway and the Central Tender Board are collabo- information in 2-3 days instead of 2-3 weeks and the submission of payroll data electronically. The 1.5 ton printers will increase rating to create a unique, open source-based accountability as Heads of Departments and Line Mangers will dgMarket software to facilitate the government's receive HR and payroll reports for verification. Historically, line managers had no information on the salary levels of their employ- e-procurement system. Although not yet complete, ees and no automatic verification that vacant positions were the system will eventually allow suppliers to search removed from the payroll. The submission of the payroll electroni- for tender opportunities, provide a facility for con- cally will also reduce data input error. tract and profiles registration, and send out tender Source: OTF Group interview alerts based on GOT requirements. Executive Summary . 65 FIGURE 38. E-government stage vs. capacity, selected african countries 2001 3.0 2.5 2.0 Capacity South Africa 1.5 1.0 Kenya Tanzania Mauritius E-Government Zambia 0.5 Uganda 0.0 Emerging Enhanced Interactive Transactional Seamless Actual Stage of E-Government Tanzania Uganda Kenya Mauritius South Africa Zambia with a new address, for example, those updates instant access to any service in a "unified package." could automatically be shared with the revenue This stage, where ministerial and departmental authority (for tax purposes) and the interior min- lines are irrelevant, has yet to be reached by any istry (for census planning). country. Tanzania is ranked with Kenya, Uganda and Zambia in the enhanced stage. The United Nations has compiled e-government data that allows governments to benchmark their The United Nations survey concluded that Africa performance against other nations, and it is a useful has the lowest regional e-government capacity index measure of the GOT's progress. of all global regions at 0.84. The UN e-government capacity index is comprised of measures of a country's According to the United Nations, there are 5 stages web presence, telecommunications infrastructure, and of e-government: emerging, enhanced, interactive, human capital. The metrics, conceived by the UN, transactional and seamless.78 Countries with which compare a government's potential progress in e- emerging presence have a formal, but limited web government given its access to connectivity with its presence in the form of independent government achieved progress, are shown in Figure 39. Tanzania websites, with static organizational or political almost exactly reflects Africa's performance, with a information. An enhanced presence is defined by score of 0.83. Among the African comparison coun- an increased number of government sites with tries, Uganda scores lower at 0.46, with Zambia more specialized information and with links to nearby at 0.75. South Africa is clearly beyond the other government pages. Countries enter the inter- other African countries at 1.56. For context, coun- active stage when their websites allow for formal tries such as Belgium and Denmark are in the inter- interactions such as discussion areas and the search active stage, averaging 2.36, and France and Spain of specialized databases. Countries with transac- are at the transactional stage. tional presence have government websites that allow users to complete secure transactions such as As these and other metrics make clear, despite the obtaining visas and licenses, as well as pay utility anecdotal progress in Tanzania, there is still signifi- bills and taxes. Digital signatures can be recognized cant work to be done in capturing the true value of for procurement of government contracts. Lastly, digitizing government service. The individual seamless or fully integrated e-governments allow for charged with directing Management Information 78 United Nations & American Society for Public Administration " Benchmarking E-government: A Global Perspective" , 2001 66 . Improving competitiveness in Tanzania Systems and e-government sits within the President's continue to fall into three primary categories of out- Office, reflecting the seriousness of Tanzania's com- puts: transparency (relevant data, widely and freely mitment. The next step is for the government to available), simplicity (streamlined procedures for pri- concentrate on e-government reforms that will have vate sector transactions), and information efficiency the largest impact on the ability of firms to compete. (seamless integration of data, within and between The value of these reforms is broad, but they will government agencies). Executive Summary . 67 7. CONCLUSION conducting business. This will require investing in a broad ICT platform, as well as business development services that particularly target the SMEs that domi- nate the economy. As this study has tried to capture, the challenges to ICT have a critical role to play in achieving growth Tanzanian growth and competitiveness are tremen- and competitiveness in Tanzania, but they are not a dous. Tanzania is burdened by an economy that panacea. ICT are an enabler of competitive advan- generates very low GDP per capita and few com- tage and operational efficiency for firms, but they petitive exports. The goal for Tanzanian leaders are informed by policies, infrastructure and strate- must be to adopt a new economic growth model gies that government and the private sector help to that increases prosperity for the average citizen. create. The GOT has significant work to do before ICT can be a true competitive asset in Tanzania. At Achieving this goal will not be easy. Tanzania will this moment, there are no clear competitive strate- need to break from its traditional mode of compe- gies guiding the growth of key sectors. The current tition, exporting natural resources as commodities ICT platform is weak and must be upgraded. while capturing little value. The country must learn Access to telephony and the internet is very limit- to replace this cycle of poverty with a virtuous cycle ed, in comparison to other countries in Africa and based on competitive advantages. In this new cycle, beyond. Firms' ability to use relevant ICT is also increasing wealth for the average citizen is fueled by lacking, primarily due to low levels of secondary the sustained capacity to build and export complex education and training. SME usage is particularly products and services. The cycle drives individual low, and while it may be a rational firm-level deci- prosperity because it depends on increasing levels sion, it is also a function of the lack of coordina- of productivity and innovation, which can only be tion of ICT actors in supporting these firms. achieved by investing in human capital. While tak- ing a broad view of Tanzanian competitiveness, this One tension point of this study is that the real report has specifically explored the role of ICT in inputs to economic transformation range from the creating this virtuous circle. specific and measurable (increasing connectivity) to the psychological and often messy (adopting a new Should Tanzania's leaders choose this new model of competitiveness mindset). This journey depends on competition, they will need to make difficult choices. ICT, but it also depends on increased levels of They will need to focus scarce resources on where trust, creativity and leadership from disparate parts Tanzania can compete internationally. A construc- of Tanzanian society. It will require that people tive public/private sector dialogue will be a first step think differently, and then act differently. This shift in analyzing opportunities and evaluating the capac- is difficult for all countries, but history counters ity of Tanzanian firms to compete in global markets. that it is possible. The timeline can vary widely, This study suggests that Tanzania start this discus- however, often driven by these less tangible factors. sion in sectors that currently compete on compara- At the end of the day, it will require a delicate bal- tive advantages, but where building new competi- ance of both patience and impatience from all tive advantages is a distinct possibility, particularly stakeholders, including Tanzanian citizens. with targeted, relevant, cost-effective investments in ICT. Opportunities to use ICT to create value- The recommendations that follow focus on using added products for niche markets exist in the coffee, ICT to build more productive Tanzanian firms, to cashews, minerals and tourism sectors. help create an environment that enables competi- tiveness and growth. Given the nature of the study, Increasing firm productivity is not just an exercise for they are biased towards actions that can be imple- exporting firms, although they are vital to generating mented in the relative short-term by the govern- the foreign exchange that drives economic growth. ment and its development partners. It is not an Domestically focused firms must also adopt a new exhaustive list, but strives to be a useful input into competitiveness model. These firms must also seek an ongoing dialogue about ICT and Tanzanian out new opportunities and more efficient means of competitiveness. 68 . Improving competitiveness in Tanzania 8. RECOMMENDATIONS In the coffee sector, the first steps are being taken towards a more liberalized market structure with the adoption of a "second window" for direct export sales, rather than the state-run auction. Firms that are taking advantage of this opportunity are focus- 8.1 DEVELOP AND EXECUTE ing on higher quality clean coffee for the specialty COMPETITIVE STRATEGIES markets in the US and Europe. This is a sound IN KEY CLUSTERS strategy in the coffee market and should be encour- aged. At a primary level this requires investment in the construction of processing centers and the train- Tanzania's biggest challenge to achieving sustained ing of operators to ensure a consistent high quality economic growth and competitiveness is the product. ICT solutions will have relatively limited upgrade of its products and services. Given the chal- impact, but may include a GIS study to strategically lenges to becoming globally competitive in any locate processing centers, websites for more sophisti- industry, Tanzania's public and private sector leaders cated producers such as KILICAFE, and database must not only create a platform for successful firms, systems to capture data for appellation models. but also target support of key economic sectors that have the revenue and employment potential to help Experience with processing facilities in the cashew the country reach its national objectives. This sector highlights the need to adopt appropriate requires developing a coherent and coordinated technology. Value addition opportunities exist and industry-wide strategy for each of those target clus- are being successfully implemented by firms such as ters, and building public and private sector partner- Olam International. The processing method adopt- ships to attract support and investment. The initial ed by Olam is different from that used in the state- analysis in this report of the agribusiness, minerals built factories that have been privatized. This and tourism sectors can be used as a foundation to process ensures a higher quality product less kernel formalize selection of priority sectors and test breakage, which decreases value) and is labor inten- hypotheses about how to upgrade these industries. sive, providing employment opportunities. Again ICT solutions are not indispensable to affect this Once Tanzania develops competitive strategies in transition, but they can help to facilitate it. One promising sectors and sub-sectors with detailed technology that can be used in cashews and agricul- action and investment plans, specific firms and ture generally is the adoption of traceability systems institutions can then benefit from directed invest- that provide information to participants along the ment. The primary value of this directed invest- supply chain. Demand-driven traceability require- ment is that it allows the government, donors and ments include certifying organic products, tracking the private sector to prioritize high impact invest- insecticide/pesticide use, identifying genetically ments within constrained budgets. The inclusion of modified foods and backtracking sources in case of financial institutions and donor organizations in health hazards. the strategy and implementation process facilitates the access to capital required by private sector Within the minerals and mining sector there are entrepreneurs and cooperatives to execute the strat- forward integration opportunities in the lapidary egy. Local financial institutions feel more comfort- and jewelry industries. A victim of smuggling and able lending to investors who are following strate- adverse regulation in the past, this industry will gies that have been tested with the market, and have a new opportunity to thrive in January, 2006, have been explicitly endorsed by private and public when legislation prohibiting the export of uncut sector participants. Donor organizations are assured gemstones goes into effect. Donors and govern- that projects designed for the priority sectors will ment can prepare the domestic private sector to not be executed in a vacuum, but will be adding capture forward integration value. Activities include value to a comprehensive effort to strengthen the training domestic artisans, financing the purchase economy. of appropriate technology and conducting in-depth primary market research. One ICT solution that The outputs of this process will vary from sector to could be valuable is investing in electronic trace- sector, as shown by the analysis in this report of ability systems for tanzanite to efficiently conform several key economic sectors: Executive Summary . 69 to the Tucson Tanzanite Protocols. Another is pur- entrepreneurs launching businesses receive suffi- chasing and training the jewelry industry in CAD cient support. CAM methods of jewelry design. The tourism industry is a complex and important business in Tanzania, and one in which industry dynamics lead foreign tour agents to capture much 8.2 FACILITATE ACCESS of the total value. Most tour operators in Tanzania focus on developing relationships with tour agents Access to ICT is a critical part of a building a in foreign countries, who direct clients to the coun- strong competitive platform for all firms, whether try and maintain strong bargaining positions they compete domestically or internationally. against the domestic operators. As such, prices and Increased access can lead to better coordination and margins are not controlled by local players. communication across the value chain, increasing Domestic operators do not focus on targeting spe- productivity and efficiency. For all firms, but par- cific customer segments that can drive a more ticularly those competing in international markets, informed and competitive strategy in an industry increased access to ICT can also lower cost struc- that is approaching saturation in terms of the num- tures. Although this benefit may not translate into ber of players. Capturing valuable customer knowl- increased competitive advantage against firms from edge and increased communication with consumers other countries, it levels the playing field for will allow operators to customize experiences for Tanzanian companies. high value clients. Market research tools such as online surveys and analytical packages are an Access to ICT in Tanzania is currently weak. Fixed opportunity to leverage ICT to react to and satisfy and mobile teledensities are some of the lowest in customer preferences. Africa, even as the increasing use of mobiles phones compensates slightly for the slow growth in fixed- Finally, new sectors will become attractive in the line telephony. In terms of Internet hosts and com- near future, sectors in which Tanzania is not cur- puters, Tanzania has one of the lowest usage rates in rently competing. When these opportunities the region. Internet density is particularly low at develop, it will be important to ensure that 0.23 users per 100 inhabitants.79 The broadening of Action items Form a small committee to identify priority sectors and create public / private sector workgroups for each priority sector tasked with developing detailed action and investment plans, as well as concrete timelines for final sector strategies informed by international markets. Private- Public task forces should develop a system through which priority sector performance, investment and returns will be tracked against projections; this data will enhance transparency and be useful in refining resource allocation and strategy goals. In the coffee sector, government and institutions should encourage investment in a high quality strategy, including processing infrastruc- ture and operator training. In cashews, invest in appropriate technology for factories to process cashews for export. Research opportu- nities to invest in traceability systems that provide value to consumers and differentiation from competitors. Within the minerals sector, prepare the domestic private sector to capture forward integration value in lapidary and jewelry industries, including training domestic artisans, financing the purchase of appropriate technology (CAD CAM), and conducting in-depth primary market research. Invest in electronic traceability systems for tanzanite to efficiently conform to the Tucson Tanzanite Protocols. Build a more competitive tourism strategy through the framework discussed above. Invest in market research and analysis tools to increase customer knowledge and communications. 79 ITU 2003 70 . Improving competitiveness in Tanzania access to ICT can be accomplished by improving needs. The body will need to coordinate closely with international connectivity, building out the national TCRA and may need to be appointed by GOT. The backbone infrastructure and addressing the digital government should also work with key stakeholders divide between urban and rural areas. of incumbent networks, TANESCO, TTCL, TAZARA, TRC and SONGAS so that a framework The high costs of connectivity are mostly due to the can be adopted in which these owners provide access expense of connecting internationally through satel- to their networks to other potential providers, so lite. Tanzania must invest in alternative methods to that commercial connectivity can increase around access the international fiber optic backbone, which the existing networks. is a key determinant in making domestic connectiv- ity affordable and reliable. Expensive and unreliable In both efforts to improve international connectivi- connectivity adversely affects firms' ability to com- ty and the domestic infrastructure, TCRA will play pete in domestic, regional and international mar- a significant role. After liberalization, TCRA has kets. In order to address these issues, Tanzania must implemented progressive regulation around licens- not only focus on its infrastructure, but also on its ing, interconnectivity and switching costs for tele- regulatory and competitive environment. com. However, prices are still high for the region. In general, TCRA needs to implement and enforce Political support must be mobilized to facilitate fair competition among telecom and data services regional dialogue for the EASSY project via Eastern providers so that new players can enter the market African Community representatives and Zanzibar and effectively lower prices (seeking to duplicate Telecom, Tanzania's representative to the MOU. the effect Zantel has had in lowering prices and While the project is expected to be primarily improving service quality in Zanzibar). Regulations financed by MOU parties as equity holders and should ensure that all networks can talk to each strategic investors in the form of global carriers, other, hand off communications and provide fair- there will also be additional funding required in the priced interconnectivity. form of loans or bank guarantees to local private companies. This is an opportunity for donors to In preparation for more efficient international con- invest directly in improving Tanzanian and East nectivity, TCRA's regulation needs to ensure that African connectivity. In tandem with EASSY, cost savings are passed on to the private sector when support and regulation will also be required for EASSY and EADTS are operational. In some coun- the East African Digital Transmission System tries, access to these types of cables has been limited (EADTS), to link the EASSY fiber optic cable to by regulators, stifling the benefits of competition. In the mainland. these monopolistic models, providers have set access rates to cables so close to that of satellite rates that Access will further increase within the country once the private sector has not benefited from the reduc- Tanzania implements its national infrastructure tion in costs necessary to compete. backbone project. The government can play a key role in this endeavor and must leverage existing Access is particularly weak in rural areas, which networks while encouraging last-mile build out have significantly less ICT infrastructure, much services by the private sector. With large expensive lower quality of training in terms of education and infrastructure projects such as this one, Tanzania ability, and fewer opportunities for work. The cannot afford the duplication of effort or inefficient rural/urban divide in Tanzania will be exacerbated investments, which have plagued more developed under current competitive dynamics and condi- economies. The project would require the construc- tions. However, if resources are mobilized to tion of 6,997 km of fiber optic cables and approxi- improve connectivity, the gap between these popu- mately 3,475 km of links, at a total estimated cost lations can be decreased. Government and donors of more than US$ 169 million. have a role to play in bringing access to rural areas, although this investment must be rationalized. An institution needs to be identified that will oversee Social service providers and SMEs in rural areas the management of the new backbone operation and that have a demonstrated need for ICT solutions its leasing agreements. This institution will need to should be supported with access to capital. The operate effectively and be responsive to private sector same is true of ICT sector players that can fulfill a Executive Summary . 71 market demand in these areas. When direct subsi- 8.3 IMPROVE ABILITY dies or grants are used, it should be clear that doing so will not suppress private sector development in While Tanzania advances in making ICT more providing the same services. accessible to its population, it must also address the limited ability of many people to use ICT effective- For example, although the Sengerema Telecentre is ly. This is another important factor responsible for a current success story, and its rising revenues are a the lack of ICT adoption in Tanzania and is a testament to the market demand for such services, direct function of education levels and technology caution should be taken in replicating this model. training. Long-term success of access initiatives The private sector has demonstrated its ability to rests on firms' capacity to effectively take advantage serve similar market demands, as shown by the pro- of an enhanced ICT platform throughout the liferation of Internet cafes throughout the country country. To this end, Tanzania needs to focus on that also provide other basic business solutions. two critical areas: improving basic education and Donors must work to assist the private sector in enhancing training in applied ICT business skills. providing these services, while not distorting a competitive market environment. While recent reforms in primary education have lifted enrollment rates in Tanzania, secondary When bridging the rural/urban digital divide, school enrollments and tertiary education remain investment in rural areas must also be appropriate among the lowest in Sub-Saharan Africa at just to the environment. According to the analysis of 5% and 1% respectively.80 As the government rural ICT usage done by DFID in June, 2005, addresses the low levels of education generally, it rural areas are more dependent on radio broad- also needs to incorporate ICT into its education casts and mobile phones than personal computers strategy. It is beyond the scope of this report to or Internet. Social service providers will find these advise on general education policy, but there are technologies most useful in promoting healthcare tangible actions that can be taken to support the education such as infant immunization in rural inclusion of ICT in basic education curricula. areas. Given the low levels of basic education, Since basic education is a broad social good, this accessing more sophisticated forms of ICT should is an area in which donor institutions can imple- not be dependent on speaking English, and web- ment programs without worrying about distorting sites should be developed with Swahili content. markets. Action items Ensure that domestic firms wishing to invest in EASSY rollouts have access to financing, and research donor support of telecom financ- ing in other countries. Possibilities include direct loans or loan guarantee programs with domestic commercial banks. Invest in a detailed feasibility study of the backhaul logistics to landlocked partners of EASSY project such as Rwanda and Uganda that will traverse Tanzania. Create incentives through TCRA that will encourage commercial last mile build-out using the existing and new fiber optic networks. Invite ICT-focused NGOs and experts such as ICANN to diffuse best practices in pro-competition policy and regulatory frameworks to help guide TCRA. Support demand-driven ICT investments by SMEs in rural areas, subsidizing rural firms' investment in ICT hardware and connectivity. Accelerate the realization of the Rural Telecommunications Fund and the institutional framework by guiding implementation. Facilitate joint ventures between international ICT-focused companies and local ones. Subsidize training for ICT sector firms in areas such as English writing skills, training of trainers, customer care service, and technical abilities. Rationalize and revise taxes and tariffs on technology adoption. Analyze impact on ICT firms of 15% withholding tax on services such as bandwidth. 80 World Development Indicators, World Bank 2004 72 . Improving competitiveness in Tanzania Secondary schools represent a good opportunity for needs assessment of the private sector, especially introducing students to computers and their value. SMEs, to evaluate their specific issues. This study will This can be a more complicated process than it ini- inform the content of specialized ICT training pro- tially appears. Integrating ICT into the educational grams, which can be implemented by both academia system goes beyond setting up a few computers and and the private sector. exposing students to the Internet or basic pro- grams. It even goes further than conducting com- puter training sessions by experienced professionals. Action items To introduce students to the benefits of ICT in a rigorous way, teachers must incorporate ICT into Designate a task force that can develop methods by which their own teaching methods. This means that secondary school teachers can incorporate ICT into their cur- ricula and give teachers incentives to enroll in ICT training. teachers themselves may need ICT training. Invite experts such as SchoolNet Africa to provide counsel. The types of projects that support ICT training in Invite and sponsor private sector leaders to hold honorary professorships to teach courses that highlight the business secondary schools should be carefully evaluated. As applications of ICT at the University of Dar es Salaam's evidenced by the experience of the DILES project, Department of Engineering, as well as at the University Computing Centre, Ltd. which made secondary school examinations and syllabi available on the Internet, not all methods are appropriate. This project depended on passive methods of attraction and assumed greater access than was realistic. The Thin Client Terminal 8.4 STRENGTHEN SMES Project currently being implemented in a pilot phase may better facilitate access and improve ability at a THROUGH INCUBATION reduced cost. As mentioned previously, however, AND BDS providing hardware is only a first step in truly capturing the benefits if computers in schools Given the importance of SMEs in private sector development, both in exports and domestic contri- Improving ability in the private sector requires a dif- bution, these firms must be supported generally ferent approach, and should focus on enhancing and in regards to ICT adoption. Providing well training in applied ICT business skills. The annual thought out business incubators and business private rates of return to these types of technical and development services (BDS) to SMEs is one step on the job training are very high compared to basic that can strengthen these enterprises. education, as shown in a 1997 World Bank study. The study concludes that while return rates on Business incubators can help start-ups survive in University training average 9%, the rates for voca- this crucial phase by providing a range of services, tional training are 19.4%, and 35.2% for on-the-job from hands-on management/technical assistance training.81 Unfortunately there are not enough strong and access to finance, to support services and infra- links between academic trainings and the private sec- structure, such as office space and communication tor's needs. Though some initiatives have addressed facilities. Research on business incubation has led the disconnect between academic training and the to a set of general principles to secure the financial needs of private businesses, wider ICT literacy will stability and success rate of this business develop- only become a reality when both individuals and ment model. First, the less an incubator relies on organizations use them in their day to day lives, and subsidies, the more successful and viable it is.82 when the appropriate advanced training becomes This calls for a clear strategy of how the incubator available to those in or entering the workforce. will reach financial sustainability, by both providing services to in-house start-ups, and offering training There are several ways in which Tanzania can and consulting services to other private and public improve coordination between academia and the pri- sector firms. As incubators move towards financial vate sector, but the first step must be to conduct a viability, it is typical for their budget to move away 81 World Bank 1997 82 NBAI www.nbai.org Executive Summary . 73 from depending heavily on grants from govern- 8.5 CREATE AN ICT ALLIANCE ment or multilateral organizations, to relying exclusively on own revenue-generating activities An ICT Alliance for Tanzania would improve ICT (consulting services, training, etc). adoption by providing a structure through which Second, incubators must be run by entrepreneurs ICT actors can effectively engage target user seg- who have the management experience, private sector ments. This `task force' would include ICT actors connections, energy and charisma to lead other entre- or "suppliers," which would consist of hardware preneurs through the difficult terrain of launching and software suppliers, ICT trainers and govern- new ventures. Although connections with the govern- ment in its role as regulator, as well as discrete tar- ment are an advantage, this cannot be these man- get segments, including SMEs, large enterprises, agers' defining asset.83 Incubators should also draw a local government and national government (a more clear line between financial intermediation and pro- detailed structure of the ICT Alliance can be found viding direct financial assistance to firms. The former in Section 5.5.3, "An ICT Alliance"). The suppliers is their actual role, and they should focus on setting would work together to provide complete ICT up networks of potential investors that can help to solutions to firms and government, ensuring that raise start-up capital. Finally, the incubator should trainers are available and that government in its have clear and unfailing performance metrics, as well role of regulator is supporting private sector efforts. as clear success and "graduation" criteria.84 This structure promotes the relevance of ICT solutions for target segments. This is particularly Another initiative that can strengthen SMEs and important for the SME sector, which has shown shore up new business ventures are BDS. As this resistance to adopting ICT solutions that do not report highlighted, some of the major obstacles offer a clear return on investment cost. faced by Tanzanian SMEs are the lack of relevant service offerings and in-house technical capabilities, The ICT Alliance structure that this report endors- and the difficulty of rural and provincial SMEs in es is ambitious, requiring the coordinated effort of accessing them. For best results, BDS organizations many institutions. Fortunately, Tanzania has a rela- should have (i) a clear mandate and strategic vision, tively strong informal ICT network from which (ii) the in-house technical capabilities to execute on to draw upon for the ICT Alliance. For example, that mandate, (iii) private sector involvement, a two-day workshop sponsored by AITEC and including personnel with entrepreneurial expertise SWOPNET in Mbeya gathered individuals and that can engender institutional legitimacy, strategic firms from the private sector, government, NGOs guidance, and demand-driven services; and (iv) and local ICT professionals. During the seminar, clear performance metrics to assess an agency's suc- several simple but significant linkages and solutions cessful interventions with SMEs. Furthermore, the were developed. This type of coordination can be development of a BDS National Network with spe- facilitated by the ICT Alliance, though with more cific service offerings, can start to close the urban/ structure and depth. rural divide in terms of SME technical assistance. Action items Action items Conduct an SME needs assessment to evaluate ICT usage and training requirements to guide BDS service offerings, with Identify coordinating body to own the ICT Alliance effort and particular focus on the urban/rural divide. begin organizing specific actions. This role could be filled by the improved ICT Coordination Office in the national govern- Assess the strategic potential of Business Incubation as a tool ment, or by a separate body. to strengthen technology-based companies. Gather knowl- edge and best practices from other African Business Provide specialized training programs, forums and workshops Incubation centers (Mauritius National Computer Board's ICT that tie ICT solutions to the competitive strategy needs of target Incubation Center and South Africa's Voxel Innovation Support groups. Facilitators could include AITEC, Tanzania Center). Development Gateway and SWOPNET. 83 Balancing Act New Update (2004), Issue 186. Pag 2. http://ncb.intnet.mu/ncb/incubator/incubator_downloads/balancing-act_186.pdf 84 Based on "Improving Competitiveness and Promoting Economic Diversification in the Caribbean: The Role of ICT". 74 . Improving competitiveness in Tanzania 8.6 UPGRADE PUBLIC The National ICT Policy itself is extensive, but SECTOR AND ENABLING what is required now is a more detailed agenda on how the ICT platform will enhance the productive ENVIRONMENT sectors of the economy, the sectors charged with using ICT to increase productivity and reduce Upgrading the public sector and the enabling poverty. Creating ICT offices in relevant Ministries environment for ICT in Tanzania will take the and institutions is a starting point for tailoring pol- coordinated efforts of several institutions within icy for these key sectors. A next step is facilitating and outside of government. ICT coordination conferences for public and private sector players, within the government needs to be reinforced to ICT industry members and key players identified exploit synergies and make ICT policy more within the priority economic sectors. These forums focused and relevant. This can be accomplished by will foster public-private dialogue regarding ICT adopting a new coordination structure. The gov- policy and inform government leaders on private ernment also needs to continue to invest in appro- sector challenges and needs. A strong competitive priate e-government solutions. environment for the private sector rests on an informed and responsive public sector, and this To empower ICT coordination throughout govern- exercise would be one part of building cross-sector ment, the National ICT Coordination Office coordination. should be given additional resources and relocated within the President's or Vice President's Office, In terms of e-government, the GOT has already where the ICT Coordinator would have greater begun to realize the tangible benefits of ICT with ability to impact cross-ministerial effort. To truly the upgrade of its Human Resources systems and the coordinate ICT policy effectively, a network of ICT launch of the Parliamentary Online Information professionals should be created throughout System to better serve constituents. However, firms Ministries and public institutions. These individuals still complain of bureaucratic delays, and ICT can be would report to the National ICT Coordination a powerful tool in addressing this issue. By learning Office. The institution-level offices would recom- from regional and global best practices, Tanzania's mend ICT policy inputs and determine appropriate government could continue to transition from paper ICT interventions and investment in priority sector to electronic systems, centralize back office networks, strategies through a bottom-up approach. To ensure and simplify procedures related to trade, customs, execution, these offices would be expected to and new business registration, among others. The strengthen linkages with private sector firms and act ability to provide these services effectively to citizens as members of the ICT Alliance. In tourism, for will increasingly depend on the government's ability example, workshops could be sponsored by the ICT to digitize its own processes. To drive the transition office within the Ministry of Natural Resources and to a working e-government, investment will be Tourism, in which private sector operators could needed in hardware, software and training for discuss their challenges in using ICT effectively. government workers. Action items Restructure and relocate National ICT Coordination Office to a cross-cutting office and create ICT offices with Ministries and public sec- tor institutions that report to the National Office. Facilitate conferences for public and private sector players to inform government policy. Continuously train government and regulator's staff so that they can anticipate ICT environmental changes and be proactive rather than reactive. Invest in hardware and software to transition government back office from paper based to electronic systems. Accelerate National Identification System for purposes of credit assessment and financial accounting and facilitate development of National Switch through coordination support and/or investment. Executive Summary . 75 A final undertaking that will enhance the ICT some and expensive, since wire fees must be environment--and the business environment as a included in the transaction. The adoption of whole­will be enabling e-commerce throughout credit cards will require the completion of the Tanzania. Specifically, this means providing a National Identification System so that credit can financial framework for electronic transactions. be adequately assessed and credit-granting institu- This is a growing demand among firms that target tions can have recourse in situations of default. international buyers and use websites to support Another important component is a national their enterprises, but do not have access to credit switch system that enables connectivity and net- card payment capability. These websites increase working among banks. access to buyers globally, but payment is cumber- 76 . Improving competitiveness in Tanzania 9 APPENDICES 9.1 ECONOMIC AND TRADE STATISTICS TABLE 9. Tanzania exports US$ millions 1999 2000 2001 2002 2003 Minerals 73.3 177.4 302.2 383.7 540.2 Cashew Nuts 100.9 84.4 56.6 46.6 39.4 Coffee 76.6 83.7 57.1 35.3 49.9 Manufactured Goods 30.1 43.1 56.2 65.9 99.9 Tobacco 43.3 38.4 35.7 55.5 40.8 Cotton 28.5 38 33.7 28.6 46.5 Tea 24.6 32.8 29 29.6 24.7 Source: Bank of Tanzania, Economic Bulletin TABLE 10. Marketed production of major export commodities (zanzibar) Metric tons 1996 1997 1998 1999 2000 2001 2002 Cloves 10,339.00 2,506.00 204 8,027.30 460 2,061.90 5,959.80 Copra 2,223.00 1,207.00 93.6 296.4 972.4 254.6 n/a Chilles 3.5 0.2 0.3 0.2 0 0 0 Clove Stems Oil* 1,624.00 771 19.9 10 242.1 323.5 187.3 Seaweeds 4,861.00 3,667.00 3,394.00 4,834.00 4,990.70 8,117.00 9,090.70 * Includes other essential oils, cardamom, eucalyptus, etc Source: Ministry of Planning Zanzibar TABLE 11. Inbound tourism arrivals Thousands 1999 2000 2001 2002 2003 Tanzania 627 501 525 575 576 Uganda 189 193 205 254 305 Kenya 969 1,037 995 1,001 1,146 Mauritius 600 678 675 709 722 South Africa 6,026 6,001 5,908 6,550 6,640 Zambia 404 457 492 565 578 Source: World Tourism Organization Executive Summary . 77 TABLE 12. Average length of stay Days 1999 2000 2001 2002 2003 Tanzania 7.7 8 8 10 11 Uganda na na na na na Kenya 9.4 8.7 8.4 8.5 8.4 Mauritius 10.4 10.4 10.4 10.5 10.4 South Africa na na na na na Zambia 8 6 9 8 8 Source: World Tourism Organization TABLE 13. In-country tourism expenditure US$ millions 1999 2000 2001 2002 2003 Tanzania 467 381 424 441 na Uganda 151 165 165 171 189 Kenya 485 500 536 513 631 Mauritius 718 732 821 829 946 South Africa 3,407 3,339 3,257 3,695 5,232 Zambia 85 111 117 134 149 Source: World Tourism Organization TABLE 14. Spend per day US$ 1999 2000 2001 2002 2003 Tanzania * 96.73 95.06 100.95 76.7 71.65 Uganda na na na na na Kenya 53.25 55.42 64.13 60.29 65.55 Mauritius 115.06 103.81 116.95 111.36 125.99 South Africa na na na na na Zambia 26.3 40.48 26.42 29.65 32.22 * Assumption: Passenger transport 8.0% of travel expenditure in 2003 Source: World Tourism Organization 78 . Improving competitiveness in Tanzania FIGURE 39. Tanzania top 5 exports by country export value Concentration of Exports by Country Export Value 54.9% 53.3% 50% 46.5% Top 5 Exports by Country Export Value, 2003 · Nonmon Gld Unwrt, Semimfd · Fish Fillets, Fresh, Chlld · Precious Metal Ores, Conc Percent25% · Coffee Green, Husks, Skins · Fish Fillets, Frozen 0% 1998 2001 2003 FIGURE 40. Tanzania top 5 exports by country and world share, 2003 Concentration of Exports by Concentration of Exports by World Country Export Value, 2003 Market Share, 2003 91.8% 81.1% 81.7% 75% 70.2% 75% 69.4% 54.9% 50% 50% Percent Percent 25% 25% 16.8% 10.3% 0% 0% Top 5 Top 10 Top 20 Top 50 Top 5 Top 10 Top20 Top 50 Executive Summary . 79 80 . Improving FIGURE 41. Distribution of tanzanian imports competitiveness Forest Petroleum/ Semiconductors/ Materials/Metals Upstream Industries Products Chemicals Computers 1998 in 2001 29% Tanzania 23% 2003 20% 18% 13% 7% 8% 9% 8% 1% 1% 1% 1% 1% 2% Power Generation Industrial and Multiple Business Transportation Office Telecommunications Defense and Distribution Supporting Functions 39% 36% 35% 23% 24% 19% 6% 5% 4% 4% 5% 3% 1% 2% 3% 5% 3% 3% 0% 0% 0% Health Entertainment/ Final Consumption Food/Beverages Housing/Household Textiles/Apparel Personal Care Leisure Goods and Services 40% 37% 41% 24% 20% 22% 4% 4% 4% 6% 6% 7% 3% 4% 4% 1% 2% 2% 2% 1% 2% FIGURE 42. Tanzania's trade balance Forest Petroleum/ Semiconductors/ Materials/Metals Upstream Industries Products Chemicals Computers 1998 2001 -15.10-17.04 -23.07 -18.71 -24.24 2003 -77.99 -31.59 -137.81 -211.93 -310.37 -231.84 -328.08 Power Generation Industrial and Multiple Business Transportation Office Telecommunications Defense and Distribution Supporting Functions -5.69 -6.78 -4.77 -75.99-68.32 -107.68 -66.27 -44.70 -43.17 -85.55 -24.32-31.53 -48.77 -77.49-50.44 -354.43 -320.33-413.46 Health Entertainment/ Final Consumption Food/Beverages Housing/Household Textiles/Apparel Personal Care Leisure Goods and Services 39.97 61.66 53.84 61.63 Executive -43.84-17.62 -49.19 -46.97 -48.46 -36.31-36.63-32.85 -18.63 -56.47 -23.77 -60.27-68.27 -33.51 -81.49 -131.42 -142.78 Summary . 81 Figure 43 Tanzania's current stage and broad cluster state Primary Goods 38.39% 34.82% 1.18% Stage Machinery 0.05% 0.05% 0.09% Vertical 5.64% 0.46% 19.32% Specialty Inputs Upstream Industries Industrial and Final Consumption Supporting Functions Broad Cluster 82 . Improving competitiveness in Tanzania TABLE 15 Top 50 Tanzanian Industries (1­50) by Export Value, 2003 Name Broad Cluster Detailed Cluster 2003 Country 2003 Country 2003 World Export Value Export Share Export Share Nonmon Gld Unwrt , Semimfd Materials / Metals Gold 438,078,528 36.43% 1.63% Fish Fillets, Fresh, Chlld Food / Beverages Fish 67,171,272 5.59% 3.36% Precious Metal Ores, Conc Materials / Metals Base Metal Concentrts,Ores 60,992,308 5.07% 5.57% Coffee Green, Husks, Skins Food / Beverages Coffee,Tea,Cocoa 49,135,848 4.09% 0.97% Fish Fillets, Frozen Food / Beverages Fish 44,126,128 3.67% 0.85% Cashew Nuts, Fresh, Dried Food / Beverages Nuts 43,536,680 3.62% 5.80% Raw Cotton, Excl Linters Textiles / Apparel Cotton 40,793,480 3.39% 0.64% Tobacco Stripped Or Part Personal Tobacco 33,458,180 2.78% 0.92% Other Wheat Etc Unmilled Food / Beverages Unmilled Cereals 32,543,736 2.71% 0.23% Tea Food / Beverages Coffee,Tea,Cocoa 24,755,712 2.06% 0.88% Diamonds Nonindust , Unset --Cut Etc Not Set Personal Precious,Semi -P Stones 24,325,708 2.02% 0.07% Leguminous Vegetbles Dry Food / Beverages Vegetables 23,688,808 1.97% 0.93% Prec -, Semi -Pr Stones NES Personal Precious,Semi -P Stones 18,579,978 1.55% 1.04% Maize Unmilled Food / Beverages Unmilled Cereals 18,509,628 1.54% 0.17% Shell Fish Fresh, Frozen Food / Beverages Fish 17,532,588 1.46% 0.11% Cloves Food / Beverages Spices 10,140,651 0.84% 9.46% Sesame Seeds Food / Beverages Seeds, Beans 9,328,344 0.78% 2.23% Tobacco, Not Stripped Personal Tobacco 8,646,238 0.72% 0.56% Refined Sugar Etc Food / Beverages Sugar, Processed 8,429,140 0.70% 0.15% Cocoa Beans, Raw, Roasted Food / Beverages Coffee,Tea,Cocoa 7,972,236 0.66% 0.27% Glass Bottles Etc Nonvac Food / Beverages Food Packaging 7,953,141 0.66% 0.20% Cut Flowers Housing / Household Plants,Flowers 7,389,387 0.61% 0.15% Flour Of Wheat Or Meslin Food / Beverages Rice, Cereals 7,050,881 0.59% 0.40% Sisal, Agave Fibres , Waste Textiles / Apparel Other Fibres,Tow,Waste 6,677,782 0.56% 12.79% Soaps Housing / Household Cleaning Agents,Waxes 6,151,250 0.51% 0.17% Live Plants NES Housing / Household Plants,Flowers 6,038,779 0.50% 0.12% Executive Groundnuts, Green Food / Beverages Nuts 5,943,786 0.49% 0.71% Cigarettes Personal Tobacco, Combustibles 5,717,101 0.48% 0.05% Knitted Etc Articles NES Textiles / Apparel Misc Textiles 5,440,865 0.45% 0.42% Summary Fresh Vegetables NES Food / Beverages Vegetables 5,399,574 0.45% 0.05% Cotton, Carded Or Combed Textiles / Apparel Cotton 5,166,850 0.43% 2.29% Footwear Rubber, Plastic Textiles / Apparel Footwear 5,124,146 0.43% 0.05% Under Garments Knitted -- Of Cotton Non Elastic -- O Textiles / Apparel Othr Undr Garmnts 3,857,098 0.32% 0.02% . Bovine, Equine Hides, Raw Textiles / Apparel Hides,Skins,Raw 3,771,292 0.31% 0.13% 83 Continued 84 . Improving competitiveness TABLE 15 Top 50 Tanzanian Industries (1­50) by Export Value, 2003 (Continued) in Tanzania Tobacco Refuse Personal Tobacco, Combustibles 3,751,767 0.31% 2.41% Industrial Diamonds Materials / Metals Slphr,Irn Pyrite,Natural Abras 3,675,488 0.31% 0.70% Parts, Acces NES Of 785 Transportation Transportation Parts 3,074,223 0.26% 0.04% Fish Frozen, Excl Fillets Food / Beverages Fish 3,054,730 0.25% 0.03% Cordage, Cable, Rope, Twine Textiles / Apparel Rope,Cordage 2,824,753 0.23% 0.28% Oilcake And Oth Residues --Of Cotton Seeds Food / Beverages Oilcake 2,484,068 0.21% 5.68% Common Salt, Etc Materials / Metals Other Crude Minerals 2,395,406 0.20% 0.21% Other Knit Etc Fab Nonel Textiles / Apparel Knitted Fabrics 2,274,766 0.19% 0.02% Oth Irn , Stl Plates, Sheet -- Of Iron Or Simple Stl Materials / Metals Plate Sheet, Flat Rolled 2,231,517 0.19% 0.01% Irn , Stl Thin Plate, Rolld -- Of Iron Or Simple Stl Materials / Metals Plate Sheet, Flat Rolled 2,197,399 0.18% 0.02% Cotton Yarn Textiles / Apparel Yarn,Other 2,141,181 0.18% 0.02% Palm Oil Food / Beverages Other Oils 2,018,882 0.17% 0.02% Bulk Text Wste , Old Clthg Textiles / Apparel Other Fibres,Tow,Waste 2,018,859 0.17% 0.15% Fish Frsh , Chlld , Ex Fillt Food / Beverages Fish 2,000,731 0.17% 0.03% Oth Crd Veg Material NES Food / Beverages Crude Vgtble Matrls 1,959,194 0.16% 0.33% Synthetc Tanning Prodcts Textiles / Apparel Non Synthetic Dyes 1,945,132 0.16% 0.34% SUBTOTAL (1 -50) 1,103,475,219 91.77% Source : OTF Group; COMTRADE / UN Trade Statistics SITC (Rev. 3) @ 3 -digit accuracy 9.2 ICT ASSESSMENT- 2003 ITU INDICATORS TABLE 16. Main telephone lines Per 100 Total (000s) CAGR (%) Inhabitants CAGR (%) Country 2002 1997­2002 2002 1997­2002 Uganda 55 0.3 0.22 3.2 Kenya 328.1 3.8 1.03 0.5 Mauritius 327.2 8 27.03 6.7 South Africa 4844 0.8 10.66 1.1 Tanzania 161.6 9 0.47 6 Zambia 87.7 2.5 0.82 0 Lower Income 68329.5 14.7 2.83 12.5 Lower Middle Income 394271.5 15.3 16.48 14.4 Upper Middle Income 66305.9 5.6 20.05 4.4 High Income 562668.9 1.9 58.54 1.2 Americas 293448.8 3.8 34.73 2.3 WORLD 1091575.7 6.7 17.9 5.3 Source: ITU 2003 TABLE 17. Local telephone network 2002 Main Telephone Lines Faults Per 100 Capacity Residential Main Lines Per Country Used (%) Automatic (%) (%) Year Uganda -- -- 80 -- Kenya 66.7 99 43.6 220.9 Mauritius 85.9 100 80 56.8 South Africa -- -- 51 48.2 Tanzania 68.9 97 63 24 Zambia 60.9 100 67 -- Lower Income 78.7 99.7 77.9 105.1 Lower Middle Income 78.8 99.8 79.7 20.8 Upper Middle Income 80 85.2 76 15.1 High Income 83.6 100 70.7 10.5 Americas 81.4 100 69.2 11.7 WORLD 79.4 99 74.7 23.3 Source: ITU 2003 Executive Summary . 85 TABLE 18. Teleaccessibility 2002 Residential Mainlines Public Telephones Per 100 % Households with Per 1000 as % of Country Total (000s) Households a Telephone Total (000s) Inhabitants Mainlines Uganda 21.6 0.4 2.7 3.24 0.13 5.9 Kenya 142.3 2.1 -- 9.6 0.3 2.9 Mauritius 261.8 84.4 80 2.92 2.41 0.09 South Africa 2511.5 25.1 31 179 3.94 3.7 Tanzania 101.8 1.5 2 2 0.06 1.2 Zambia 44.8 2.2 3.8 0.88 0.08 1 Lower Income 15954.4 8.2 8.2 2622.63 1.13 3.9 Lower Middle Income 310811.5 49.8 49.4 12689.83 5.31 3.2 Upper Middle Income 50273.7 58.4 59 1586.89 4.9 2.5 High Income 356235.5 120.5 96.1 4119.85 4.36 0.07 Americas 204646.6 84.5 70.8 4353.44 5.25 1.5 WORLD 733275.1 61 49.8 21009.19 3.52 1.9 Source: ITU 2003 TABLE 19. Telephone tariffs 2002 Residential Business Subscription as Connection Subscription Connection Subscriptions Local Calls % GDP Per Country (US$) (US$) (US$) (US$) (US$) Capita* Uganda 61 5.6 61 5.6 0.21 27.4 Kenya 29 5.6 29 5.6 0.07 17.4 Mauritius 33 2.5 67 7 0.04 0.8 South Africa 23 6.4 23 8.5 0.09 3.4 Tanzania 41 3.6 41 3.6 0.12 16 Zambia 11 1.1 34 2.3 0.09 4.4 Lower Income 54 3.1 66 4.2 0.08 11.8 Lower Middle Income 84 4.4 121 7.4 0.05 3.7 Upper Middle Income 62 7.5 82 12.2 0.09 1.9 High Income 83 11.8 94 16.7 0.1 0.7 Americas 88 7.6 115 15.7 0.07 3.1 WORLD 71 6.2 91 9.4 0.08 5.2 Source: ITU 2003 86 . Improving competitiveness in Tanzania TABLE 20. Mobile cellular subscribers 2002 Cellular Mobile Subscribers As % of Total Per 100 CAGR (%) Prepaid Population Telephone Country Total (000s) Inhabitants 1997­2002 Subscribers (%) Coverage (%) Subscribers Uganda 393 1.59 139.4 -- 55.0 87.7 Kenya 1325 4.15 187.3 93.2 -- 80.2 Mauritius 350 28.91 52.4 79.1 99.8 51.7 South Africa 13814 30.39 49.7 75.4 95.1 74 Tanzania 670 1.95 101.4 -- -- 80.6 Zambia 139 1.3 98.2 -- 50.5 61.3 Lower Income 42298 1.75 76.5 79.9 63.6 38.3 Lower Middle Income 380000 15.88 67.6 40.7 82.3 49.1 Upper Middle Income 102297 30.94 57.4 74.6 93 59.7 High Income 638079 66.39 29.9 44.5 97.7 53.1 Americas 252642 29.9 28.7 33.5 91.8 46.3 WORLD 1162675 19.07 40.2 46.7 84 51.5 Source: ITU 2003 TABLE 21. Information Technology 2002 Internet PCs Hosts Per 100 Users Per 100 Users Per 100 Users Per 100 Country Total Hosts Inhabitants Inhabitants Inhabitants Total (000s) Inhabitants Uganda 2242 0.01 100 0.40 82 0.33 Kenya 2963 0.01 400 1.25 204 0.64 Mauritius 3462 0.29 120 9.91 141 11.65 South Africa 198853 0.44 3100 6.82 3300 7.26 Tanzania 1731 0.01 80 0.23 144 0.42 Zambia 1621 0.02 52 0.49 80 0.75 Lower Income 201028 0.01 32112 1.33 16594 0.72 Lower Middle Income 3683093 0.15 116234 4.86 89202 3.8 Upper Middle Income 3327987 1.01 46678 14.13 33305 10.08 High Income 150369694 15.64 427999 44.53 448416 46.68 Americas 122555360 14.5 217649 25.76 239717 28.95 WORLD 157581802 2.59 623023 10.22 587518 9.91 Source: ITU 2003 Executive Summary . 87 10 BIBLIOGRAPHY 13. The Economist Intelligence Unit "Country Report April 2005­Tanzania", © 2005 The Economist Intelligence Unit Limited 14. FAIDA SEP Tanzania BDS Support Project Case 1. Albaladejo, Manuel "Drivers of Industrial Study 2002 Competitiveness in Tanzania: A capability and sec- toral approach", July 2003 15. FAOSTAT 2. Aubert, Jean-Eric; Wanga, Goodwill, "Innovation 16. Hancock, Simon, "Mobile Phones Boom in in Tanzania: Insights, Issues and Policies" Tanzania," BBC News. Available at http://news.bbc.co.uk/2/hi/programmes/ 3. Baffes, John "Tanzania's Coffee Sector: Constraints click_online/4706437.stm. and Challenges in a Global Environment" The World Bank, Africa Region Working Paper no. 56, June 2003 17. Heeks, Richard; "eGovernment in Africa: Promise and Practice", iGovernment Working Paper Series, 4. Bank of Tanzania , Economic Bulletin 2002 5. Block, Robert and Daniel Pearl. "Underground 18. Henten, Samarajiva and Melody, "The Next Step Trade. Much Smuggled Gem Called Tanzanite for Telecom Regulation: ICT Convergence Helps Bin Laden Supporters." The Wall Street Regulation or Multisector Utilities Regulation?", Journal, November 16, 2001 World Dialogue on Regulation for Network Economies, 2003 6. Charles, Kathleen, "National AGOA Strategy", The East and Central Africa Global Competitiveness 19. Integrated Labour Force Survey 2000/2001, Trade Hub, August 2004 Ministry of Labour, Youth Development and Sports and the National Bureau of Statistics, The 7. Click Online, Tanzania Mobile, BBC World, July United Republic of Tanzania 26, 2005 20. Kijo-Ringo, Natujwa Daniel, "Impact of Investment in 8. Department for International Development and Utilization of Information and Communication (DFID) KaR Project 8347: The Economic Impact Technologies on Market Extension: Overview of Small of Telecommunications on Rural Livelihoods and and Medium Enterprises in Tanzania", University of Poverty Reduction: A Study of Rural Communities Dar es Salaam, November 2004 in India (Gujarat), Mozambique and Tanzania 21. Luhanga, Matthew L., "The Tanzanian Experience 9. Duncombe, Richard and Heeks, Richard "Information, in Initiating and Sustaining Tertiary Education ICTs and Small Enterprise Findings From Botswana", Reforms. A case study prepared for a Regional Institute for Development Policy and Management, Training Conference on Improving Tertiary University of Manchester, Manchester, UK, 1999 Education in Sub-Saharan Africa: Things That Work!" Accra, Ghana, September 23-25, 2003 10. EASSy Secretariat Office, Telkom Kenya. Available at http://www.eassy.org/index.htm. 22. Lyimo, Karl. "Tanzania Shouldn't Waste Money on ID Cards." August 2, 2005. Available at 11. Economic and Social Research Foundation and http://allafrica.com/stories/200508021061.html International Business Initiatives Corp., Tanzania's Precious Minerals Boom: Issues in Mining and 23. Miller Esselaar and Associates "A Country ICT Survey Marketing, January 2001 for Tanzania", Swedish International Development Cooperation Agency (SIDA), November 2001 12. The Economist Intelligence Unit "Country Profile 2004­Tanzania", © 2004 The Economist 24. Ministry of Communications & Transport. Intelligence Unit Limited "Technical Report on Feasibility Study for 88 . Improving competitiveness in Tanzania Implementation of the National ICT Backbone 38. USAID Tanzania Data Sheet 2004 Infrastructure," June 2005 39. USDA Advisory Committee "Global Traceability 25. Ministry of Natural Resources and Tourism Forestry and Labeling Requirements for Agricultural & Beekeeping Division, "Investment Opportunities Biotechnology-Derived Products: Impacts and in Forestry & Beekeeping", September 1999 Implications for the United States 26. Mwalyosi, Raphael B.B., "Impact Assessment and 40. Wakabi, Wairagala "African Cashews to Indian the Mining Industry: Perspectives from Tanzania, Factories: How the Continent Exports its April 2004 Profits and Jobs", The East African,December 20, 2004 27. Mwanza, Joel "Implementation of Integrated Financial Management System in Tanzania" 41. The World Bank Institute, "Fostering Innovation, Productivity and Technological Change: Tanzania 28. National Information and Communications in the Knowledge Economy", October 22, 2004 Technologies Policy, The Ministry of Communications and Transport, The United 42. World Bank, Tanzania at the Turn of the Century. Republic of Tanzania, March 2003 Background Papers and Statistics- A World Bank Country Study. The World Bank, Washington 29. Nielinger, Olaf, "Fact Sheet­ICT Utilisation of D.C. February 2002 Small and Medium Sized Enterprises in Tanzania" April 2003 43. World Bank, Improving Tertiary Education in Sub- Saharan: Things that Work. Report on a Regional 30. Nzuki, Margareth "Tanzania Business Portal­an training conference held in Accra, Ghana. Africa Initiative to enhance Tanzania SMEs Participation Working Paper Series No. 29713. The World Bank, in the Digital and Knowledge Based Economy Washington D.C. 31. Semboja, Haji Hatibu Haji; Mbelwa, Rhoda; 44. World Bank, Skills Development in Sub-Saharan Bonaventura, Charles, The Cut-Flower Industry in Africa. World Bank Regional and Sectoral Units. Tanzania, Working Paper 152, International Labor The World Bank, Washington D.C. 2004 Office Geneva, March 2000 45. World Bank, World Development Indicators 2004 32. Southern Africa Documentation and Cooperation Centre, Business Times article, "Tanzania loses bil- 46. World of Information Business Intelligence Reports, lions in raw cashew exports", 22 August 2003 " Tanzania", © 2001 Walden Publishing Ltd. 33. Stevenson, Louis & St-Onge, Annette. "Support 47. World Telecommunication Development Report; for Growth-oriented Women Entrepreneurs in Access Indicators for the Information Age, Tanzania". International Labor Organization. 2005. International Telecommunications Union 2003 34. Tanzania Internet Hosting, www.tih.co.tz/hosting. 48. World Tourism Organization Country Statistics shtml 49. Yager, Thomas R. "The Mineral Industry of 35. UNESCO, "Tanzania Country Profile", 2005 Tanzania," U.S. Geological Survey Minerals Yearbook, 2003 36. United Nations & American Society for Public Administration, " Benchmarking E-government: A 50. Zagada Markets Inc., 2002 Global Perspective" , 2001 51. http://uga.edu/fruit/cashew.htm 37. United Republic of Tanzania Vice-President's Office, "National Strategy for Growth and Reduction of 52. www.afrikasana.co.tz Poverty", April 2005 Executive Summary . 89 53 www.aitecafrica.com implications, and future prospects" Food, Agriculture & Environment Vol. 1 (1) 101-106.2003 54. www.bistanzania.com 63. www.parliament.go.tz. 55. www.cgisnur.org 64. www.rwandatourism.com 56. www.distancelearning-tz.org 65. www.scandinaviagroup.com 57. www.e-cybercity.mu 66. www.stcl.com 58. www.gemvision.com 67. www.swopnet.org 59. www.kilicafe.com 68. www.takimsholidays.com 60. www.marveloustanzania.com 69. www.tanzaniagateway.org 61. www.olamonline.com 70. www.tanzaniagateway.org/dinaflowers 62. Opara, Linus "Traceability in agriculture and food supply chain; a review of basic concepts, technological 71. www.tanzaniatouristboard.com 90 . Improving competitiveness in Tanzania Innovation in Tanzania: Insights, Issues and Policies1 Jean-Eric Aubert ­ World Bank Institute (with Godwill Wanga, Consultant) Introduction This note discusses the innovation climate in Tanzania and the policies being pursued there. This is part of the background work carried out for the Country Economic Memorandum exercise, with a focus on the growth determinants in that country. This background work also includes a report analyzing the country's readiness to the knowledge economy2, providing benchmarking elements in the broader context of Tanzania's technological development, business environment, governance, the educational basis and ICT infrastructure, as well as the innovation system. This note was essentially prepared on the basis of a number of field visits in Tanzania made between February 6 and 15, 2005. The field visits took place in Dar es Salaam, Arusha, and Moshi as well as Mwanza at the end of February. We also visited numerous organizations (enterprises, R&D institutes, universities, government agencies, NGOs). A list of these organizations and persons met is given in Annex 1. This note is in three parts: Part 1 discusses the main issues bearing on the innovation climate present in Tanzania; Part 2 illustrates them with some examples taken from the places and organizations visited, and Part 3 makes some policy recommendations. Annexes 2 and 3 include detailed information on the fishing industry and the cut flower industry.3 1This report was mostly completed in April 2005 and has served as the basis for the section of the World Bank Country Economic Memorandum dealing with innovation policy (Chapter 8). The author wishes to thank all the persons he met during the mission and who provided very valuable information, as well as the World Bank Country office in Dar Es Salaam, more specifically Mrs. Mary-Anne Mwakangale, for her very useful assistance. 2 Anuja Utz, Fostering Innovation, Productivity and Technological Change: Tanzania in the Knowledge Economy, Knowledge for Development Program, World Bank Institute, October 22, 2004 3 The annexes on the fishing industry and on the cut flower industry have been prepared by Godwill Wanga, Consultant, Economic and Social Research Foundation 1 1. The Innovation Climate: Main Issues Innovation and poverty reduction Innovation should be understood as the spread of a new or improved technology or practice, but in a given environment and not in absolute terms. In this perspective two levels can be identified: - at a micro level, there is the spread of available technologies for use by firms, individuals or households which improve their productivity, welfare, living conditions, etc. - at a meso level, there is the development of new industries, based on foreign or local technologies and a source of new jobs, income, exports, etc. Innovation contributes in the reduction of poverty in several ways: - it fuels economic growth through the generation of wealth and jobs, by providing the basis for new activities, industries or services. - it induces productivity gains, which is a source of wealth as well as a means for maintaining existing jobs against foreign or other competition. - And, last but not least, innovation is helping to maintain self sustainability of local communities, for instance through appropriate technologies, in preventing the destructive process of urbanization. This does not necessarily translate into monetary gains and does not formally contribute to growth, but it is essential in preventing these vicious circles that lead to poverty and dependency. A lack of innovative dynamism Tanzania has benefited from a relatively strong economic growth in recent years. Although new sectors such as the fishing industry have grown spectacularly, there has not been, however, much innovation in recent years in Tanzania. The main changes have resulted in the wider distribution of commodities and services that significantly impact on the day-to-day life of the population, such as cell phones and small commuting vehicles (dala-dalas) ­ but do not change, even marginally, the conditions of development of the country. A very low RD/GDP ratio of about 0,25 per cent4 is a reflection of this lack of innovative dynamism. It should be kept in mind that in developing countries innovation always precedes research, because research is developed principally to improve imported technologies; also, the capacity to invest in research depends on the capability to generate wealth which in its turn depends on achievements in innovation. Foreign Direct Investment can be an important source of technology-transfer. In Tanzania, compared to neighboring countries, this Foreign Direct Investment is relatively high but the impact on the technology and on innovation capabilities is modest. As we shall see later, no specific mechanisms seem to be in place for promoting the transfer of 4The national S&T expenditure is estimated at $ 22 million for a GDP of some $ 8.5 billion. 2 knowledge and technology (nor for channeling a large part of the income generated by lucrative industries such as gold mines). A problematic governance and business environment A number of factors explain this lack of innovative dynamism. Firstly there is a problematic overall environment. In the last decade or so significant improvements have been made after the socialist model was gradually abandoned. However: - There is a lack of effectiveness and implementation in government action despite a number of policy plans. This has affected many policy areas and slowed down the reform process. - The overall governance and business environment continues to be beset by corruption. - The infrastructure is inadequate: the country's development is compromised because of poor regional and feeder roads, frequent power cuts and an insufficient telecommunication network. - There is a weak banking system notably for providing credit to business, and capital markets are almost nonexistent. The chart below shows how costs affect the business environment in Tanzania as compared to four other countries. Costs as percentage of sales In increasing order of magnitude: Poland, China, Brazil, Indonesia and Tanzania. Costs due to unreliable infrastructure and bribery practices are particularly high. 3 0 C o n t r a c t e n fo r c e m e n t d iffi c u lt ie s R e g u la t io n 2 5 B r ib e s C r im e U n r e lia b le in fr a s t r u c t u r e 2 0 sel safot 1 5 cenreP1 0 5 0 P o la n d C h in a B r a z il In d o n e s ia T a n z a n ia 3 Inadequate Innovation Policies The S&T infrastructure in Tanzania is relatively developed with a number of technological institutes and universities that cover most basic sectors and disciplines. The institutes are also relatively well-organized with a Commission for S&T (COSTECH) responsible for defining and implementing national S&T policies along with organisms in charge of promoting industry and SMEs, notably the SME agency (SIDO). However, this set-up suffers from a serious lack of means and resources and, more importantly, is bogged down by inadequate policies and ideas. The innovation policies lack both vision and funds. The innovation system in place has been inherited from the centrally planned regime. Innovation is wrongly seen as beginning with R&D ­ to be principally developed in public institutes ­ and then being passed over to enterprises. This outdated linear vision is far removed from a real innovation process, particularly in view of the Tanzanian reality as we shall see later. On top of this, there has been a deliberate policy, inspired by the principles of liberalization and privatization, that wrongly seeks a complete hands-off stand from the government. This has meant pushing the public R&D institutes to market their competencies and technologies and in cutting resources for S&T, without providing any needed incentives to entrepreneurs to put in greater innovative efforts, and without making any investment in public goods or setting down any appropriate regulations for ensuring and encouraging a climate of stimulating innovation (competition, quality promotion, protection of property rights, etc). A misused R&D infrastructure As a result of the policy outlined above, the R&D and technology infrastructure is misused. This infrastructure, a legacy of the socialist regime, is relatively large, with an adequate set-up for metrology, measurement, prototype making, etc. The government has cut resources to the institutes by pushing them to market their services and technologies. But this has been done without creating any mechanism for the articulation of demands by enterprises, public organizations, or local communities. So the institutes have proposed a lot of inapplicable/unapplied inventions (often the same between one institute and another), and have unjustifiably ventured into production activities. The limitation imposed on investment in S&T has also affected the development of S&T potential in new technologies such as micro-electronics and biotechnology, although some competencies have been developed at the level of university structures. The advice and support given by donors in regard to innovation and innovation policy are also questionable. Generally, there has been an encouragement of the liberalization approach without understanding the key roles to be played by government in the creation or maintenance of a climate conducive to innovation; in fact, it has encouraged inertia among government bodies. There is insufficient and scattered support of innovative 4 activities offered by bilateral and multilateral programs, lacking a critical mass effect and being poorly coordinated, although a number of programs are well designed as shown later. Some programs tend also to be related to the donors' self-interest (for instance to favor cooperation with their own country's firms). Finally policy advice is very fashion-sensitive: today there is a hype for clusters, but the policies to boost clusters are unclear, even in advanced countries where the concept was first elaborated. Conclusion: A New Role for the Government In conclusion, we can say that there are a lot of ideas and goodwill, and a decent technology infrastructure. But there is a lack of innovation in various forms, and notably in agriculture, welfare, etc where we have seen very modest improvement. Adequate responses are practically non-existent. In order to improve this situation the government role needs to be reassessed in a truly new perspective. The government has a significant role to play in the maintenance and upgrading of the knowledge and technology infrastructure, in organizing an institutional and legal framework that supports innovation and in creating the adequate financial incentives to motivate individuals and groups in innovative and entrepreneurial efforts. This is not the type of direct support and supply-driven approach characteristic of a State controlled economy. What needs to be done is to transfer the policy principles which inspired innovation policies in developed countries to the specific context of Tanzania responding to the country's needs and opportunities. To better understand what those needs and opportunities better let's examine more closely a few illustrative examples. II. Innovation Processes: Selected Examples Let's present and discuss these examples in the two categories outlined above: - The dissemination of technologies which are largely known and operative, but not much used so far in the Tanzanian context - The development of new industries introducing new technologies, services and activities within Tanzania Technology Dissemination Interlocking bricks This example is particularly illustrative of the issues at stake. A technology has been designed and successfully tested to make bricks with local materials and compressed without cement. This is a very relevant technology for (poor) rural communities. It originated in South Africa and has been perfected by several inventors in Tanzania, 5 notably by the Housing RD center. The machine to manufacture such bricks costs about USD 450. This is not a considerable sum, but it is still relatively expensive for a Tanzanian individual or a household with an average annual income of some USD 300. However, no scheme exists for helping interested individuals or communities either to use or buy the machine. Moreover, there is no mechanism to inform and familiarize potential users with the technology. Therefore, the technology does not get disseminated throughout the country, even though this would have a considerable positive impact on the economy, the society and the environment. Meanwhile the Housing RD institute ­ which has a budget of USD 400,000 (funded at 60 % on outside contracts) ­ has received a USD 250,000 support for ICT training and management improvement from the World Bank, but nothing to help in the dissemination of the brick technology: is it the best way to fight poverty? The generic problem illustrated by the brick technology example affects more broadly the dissemination of innovations in the agricultural machinery sector. Agricultural machinery There is a plethora of (similar) "inventions" supplied by RD institutes, NGOs, etc... in response to definite, crucial needs, but very few mechanisms to help in their selection and dissemination (information, testing, credit, etc). An institute devoted to agricultural machinery RD (CAMARTECH) was established in the early 80s near Arusha. With a staff of about 50 workers, it is relatively well equipped, but because of very mediocre approach road, almost inaccessible and without resources. Like other public institutes, it has been pushed to manufacturing and selling prototypes. It does this on an ad hoc basis, crowding out potential entrepreneurs in the business sector. Nevertheless these efforts have facilitated the dissemination of several locally significant technologies such as solar stoves, biogas boilers, etc (in thousands). Note that when it was established, the institute initially had eight workshops spread out in Tanzania where local farmers could come, test their ideas, repair their machinery, check out new technologies, etc. But these workshops have gradually disappeared due to lack of funds, and insufficient interest on the part of the government, etc. Development of new industries Let's examine three examples in this category: the rapid development of the fishing industry around the Victoria Lake, generally considered to be a success story but which is not without serious social and ecological drawbacks, the modest development of the cut flower industry, and the emerging cultivation of a plant for a new anti malaria drug ­ which could prove to be a globally significant innovation. 6 The Fishing Industry The development of the fishing industry has principally taken place around the Victoria Lake, leading to the creation of several hundred thousand jobs (including those in fish processing, transport, etc). About 80,000 fishermen are permanently employed. Exports ­ principally toward the European markets ­ are flourishing, amounting annually to over 100 million USD. The industry took off about a decade ago after the successful introduction in the Victoria Lake of a new species--the Nile perch ­ which prospers very well. Tanzanian fishermen developed the industry successfully to the detriment of their Kenyan and Ugandan competitors by meeting the strict quality-control norms in force for accessing EU markets. They have also modernized their processing (for fillet making) and conservation equipment and installed efficient logistics and transport systems at the Mwanza airport to fly out their production. Ships have also been motorized, with appropriate credit facilities offered to fishermen. Therefore it is a well-coordinated action, of a systemic nature, involving various types of actors and segments of activities which have been the key in making the industry successful. For more details see annex 2. This, however, has not happened without some drawbacks and problems. Firstly, the ecological balance of the lake has been very seriously disturbed by the extensive development of the Nile perch which has, according to certain sources, destroyed over 200 other species. Secondly, the overall social setting has been seriously destabilized. Villages around have lost their population that has migrated to the cities which has led to over-urbanization with its consequent problems of violence, drug trafficking, prostitution and addiction-related behavior. Finally shipping activities abroad are in the hands of dubious (foreign) flight companies which encourage the traffic of weapons to the neighboring countries affected by conflicts.5 The Cut Flower Industry Tanzania would appear to be an ideal place for a dynamic cut flower industry. There is a perfect climate, a largely mastered technology, willing investors. However Tanzania today lags behind, although first initiatives for implanting these green-house based activities started 15 years ago. There are in Tanzania today only 85 hectares of green- houses, while Kenya has 1200 hectares, Ethiopia 1500 hectares (in two years), and Uganda 200 hectares. Why is this? A number of factors seem to explain this situation. There is an over regulatory burden and bureaucracy: licenses to establish new green-houses are hampered by inappropriate land property laws, the authorities (TPRI) in charge of checking and authorizing use of fertilizers are slow and troublesome. There is a lack of appropriate credit for would-be entrepreneurs. 5These issues are convincingly documented in a film entitled "Darwin's Nightmare" (made by Hubert Sauper, an Austrian director). 7 The airport infrastructure is good with a very modern airport built near the Kilimanjaro, the main site for flower cultivation. However, this private status airport, has an agreement of monopolistic use with the Dutch airline (KLM) which transports the flower-freight to Amsterdam ­ the European market hub that centralizes the whole auction process. As there is not enough regular freight, and KLM cannot ensure a fully reliable service, conditions of export are uncertain and producers prefer to send their stuff to Nairobi or Dar es Salaam: this means high transport costs, reducing profitability and the motivation for producing in Tanzania. However, Tanzanian producers have been trying to organize themselves for surviving, also by cooperating with Kenyan producers for promoting research and development activities (e.g. tests of new plants). For more details on the factors affecting the cut flower industry see annex 3. A New Crop of Global Significance Let's end this series of examples by mentioning a new plant ­ arthemisia -- useful in an anti-malaria drug which is being developed in the Arusha region and could become a veritable success story. An American NGO (Technoserve) working in Tanzania (and other African and developing countries) has identified opportunities for cultivating a plant that can provide a new drug for malaria; it works closely with the World Health Organization which furnishes market requirements and certifications. Tanzania has much higher yields than China (so far the only potential competitor). A market of USD 2 to 5 million over the next 3 years has been identified. If successful, the cultivation of this new plant offers an excellent opportunity of diversification for coffee producers, many of whom are in a crisis following the decline in world coffee prices and the deterioration of their production quality. If this success materializes, it will owe much to the role played by the American NGO, which is very well plugged into global organizations and strongly business-oriented. The role of foreign investors Before making any general inferences from this series of examples, it is worth commenting on the role of FDI in the upgrading of technology in the country. FDI is universally recognized as the most rapid way to upgrade the technological level of a country and to increase its level of activity. However, the long term impact on the country's development depends largely on the effective transfer of knowledge ­ both technical and managerial ­ which flows out from FDI-generated activities and the linkages established with the local industry (sub-suppliers), R&D institutes, and educational institutions. Although there is a significant FDI in Tanzania, the long term impact, from this viewpoint, on the innovative competency of the country is not obvious. FDI undoubtedly plays a great role in several Tanzanian industries, including the fish, hotel, tourist, as well as the cut flower industry mentioned above. It is also a determining factor in textiles and clothing as well as in the rapidly growing mining activities (notably gold). However, the 8 number of knowledgeable executives of Tanzanian origin is still very small. Most, if not all managerial positions, tend to be occupied by foreign expatriates with a few exceptions. Linkages with local suppliers of technology do not seem important and, in any event, they do not seem to be stimulated by any governmental schemes. Therefore the long term transfer of knowledge and technology are not ensured, and the long term sustainability of the concerned industries is in doubt, if foreign investors decide to withdraw from the country. Conclusions To conclude, a number of policy orientations emerge from the above observations. There is a need to: · Reverse current policy approaches based on technology push without resources or markets · Establish mechanisms and incentives facilitating the expression and credit- worthiness of demands by "clients", entrepreneurs, local communities, etc. · Put in place critical masses; the current scale of effort has to be multiplied several times. · Have a systemic approach to tackle the different components of innovation promotion. III. Policy Recommendations Instruments and mechanisms need to be put in place to concretize the policy principles outlined above. Against this background, there is a large pool of experience in advanced countries from which policy makers in Tanzania can draw in order to adapt it to the specific conditions of their country and, more generally, to the challenges of the developing world. The main lessons of this experience of advanced countries are summarized in Box 1. 9 Box 1. Innovation Policy Schemes ­ Relevant OECD Countries' Experience Innovation policies in OECD countries have been developed since the late sixties and a considerable experience has been accumulated since then. Of particular relevance to Tanzania and other developing countries are those schemes which concern financial and other supports to SMEs and the management of R&D and technology infrastructure. Financial and other support to SME SMEs have been the first target of innovation policies, either because they were considered as an important potential source of innovations or a major provider of jobs. As these enterprises are scattered throughout the territory and need to be approached through close contacts, a key instrument was the establishment of local offices, connected to central bodies. A key element for having these structures functioning well was to have them funded partly by local authorities or organizations such as the Chamber of commerce, as a means for ensuring local ownership. As these enterprises also generally need a combined support of technical, financial, managerial and marketing assistance, these offices were equipped to provide such packages as were needed, by drawing upon expertise, if necessary, in universities, public laboratories or consultant networks. And as SMEs generally do not have financial resources for developing their innovation projects, ad hoc mechanisms had to be established. They were generally of a twofold nature: a) provision of seed money in form of grants to develop ideas, generally up to the prototypes, b) development money for the next phase leading to technical realization and market tests, either in form of matching funds given on condition that equivalent resources be provided by the innovator or other interested parties, or in form of subsidies reimbursable in case of success. In addition, measures were taken to facilitate the provision of venture money either for organizing venture capital funds (organizations supported by public money) or for regional development companies. Financing of R&D Structures and Research-Industry Relations The second important aspect of innovation policy has been the management of R&D and technology infrastructure. Such infrastructure was conceived as a public good to be supported by the government. However it was also understood that it needs to develop activities relevant to the market and conversely that the market should be in demand for these activities and therefore ready to pay for part of the costs. Two types of action follow from such a perspective: - First, funding of RD laboratories combining both sure and precarious resources, the first type being provided on a regular basis by governments, and the second type being obtained by contracts with the business sector or government agents; but the balance between the two would have to be adapted to the nature of research being pursued. Ratios between sure and precarious resources therefore evolved between 50/50 per cent for those laboratories providing technical research and services (such as metrology, testing, quality control, etc) that can and have to be paid by the business sector; and 70/30 for those structures pursuing a research of a more basic nature. - Complementary to these funding principles, incentives were provided to facilitate the development of a competitive market for the use of these structures and also to stimulate research-industry interactions which are at the heart of the innovation process. These incentives have been principally of two sorts; firstly there were small subsidies offered to small and medium firms paying 50 per cent of their contracts with accredited public or university laboratories, and/or RD expenses spent in the employment of an in-house researcher seconded by these laboratories; secondly there were significant grants provided to projects developed through university-industry collaboration conditioned to the provision of matching funding by the private sector (sometimes these being focused on specific technologies that the government wanted to promote). 10 In this perspective, an image is commonly used among policy maker communities in advanced countries. They see their role as one of a gardener who has to care for flowers and ensure their growth. In order to grow, a flower needs to be watered ­ this is principally the role of financial incentives. It needs also an appropriate soil; this is, in the innovation perspective, the injection of all forms of knowledge through technical information, assistance, and research. It is then necessary to remove weeds that prevent or affect the growth of the plant: these are all the measures for eliminating bureaucratic practices, over-regulation, inappropriate conditions for redistribution of income and wealth generated by innovations, and so on. In a similar vein, a recent study by the World Bank on ten examples selected in low and medium income countries identifies essential government actions in the development of competitive industries in such countries6. The provision of appropriate skills and support for technology acquisition and development come first, then come actions of a regulatory nature (e.g. standards, quality control, etc) and of a lower importance various types of support to enterprises and industry organizations for export promotion, investment, etc. 6 The work involves undertaking a series of studies that cover the extremes of the industrial spectrum: wheat and maize in India; salmon and wine in Chile, Nile perch in Uganda, oil palm in Malaysia, cut flowers in Kenya, medium tech electronics in Malaysia, high tech electronics in Taiwan, and software exports from India. The industries are chosen based on exceptional comparative performance in the last decade, large contributions to overall growth and for the potential that technological change played an important role in its success. (Source: Vandana Chandra And Erik Tallroth, "Understanding the "How To" of Technological Change For Growth," Concept Note PREMEP-ESSD Partnership, The World Bank, Draft, April 19, 2004). Forthcoming publication. 11 ROLE OF THE PUBLIC SECTOR Electronic Electronic IT ­ Maize Grapes Oil palm Salmon Wine Nile Cut - - Taiwan India - - India ­ - Chile - Perch ­ flowers Malaysia India Malaysia Chile Uganda - Kenya Spin-offs X X X X X Export & inv X X X X prom. Tech X X X X X X X X X acq/development Regulation/ X X X X X X X compliance Support to X X X X X Industry organization Tech skills X X X X X X X X X X development This will guide our recommendations regarding the establishment of an effective government support to innovation. We will complement these by more general remarks about the broader innovation climate, and related to education and governance issues. Establishing Efficient Support for Innovation As observed above, there is a need for a systemic approach providing complementary support on three basic aspects: financial, technical and regulatory. Financial support There is, in the Tanzanian context, a crucial lack of resources on the demand side for accelerating the design, testing, use and dissemination of technology. It is suggested to establish two complementary schemes, based on simple matching fund principles, providing -- in grant form -- 50 per cent of the funding required for the development (R&D phase) of small and medium sized projects (e.g. up to USD 20,000): · User scheme: allowing in particular groups and communities to buy needed technologies, in providing, if appropriate, complementary funding in in-kind form (labor for community purpose) 12 · Developer scheme: would fund 50 per cent of technical services or R&D projects undertaken by SME with R&D institutes (public, academic, etc), and would, therefore, help indirectly but more effectively R&D institutes to put their competences in service of the communities (note that this type of scheme is in place in a number of advanced countries). In addition, it is important that the management of such support schemes be given primarily to sub-national levels. Regional and local commissions would screen and select projects with support of appropriate expertise (including foreign ones). This is a primary condition for efficient management, reducing the bureaucracy which would ineluctably affect schemes administered at the central level. The latter, however, would have to supervise and control the overall process. Technical support Based again on the experience accumulated in advanced countries, it is suggested to establish a network of locally-based and owned structures serving the needs of rural and urban communities for technical advice, information, assistance (design, marketing, etc) These structures should be adapted to different sectors; e.g. extension services for agriculture, design and manufacturing workshops for industry, etc.; and they should also be conceived and operated as antennas of central bodies to which they would be strongly connected (IT, data bases). They should be established on a clearly expressed demand from local communities, being funded on a 50/50 cost sharing basis, local organizations (municipalities, business or farmer association, etc) matching the resources put by the Central government. Regulatory support Regulatory-related actions have to be implemented for dealing with several issues: · The above mentioned actions that have been suggested aim at stimulating service-based contracts and formalizing new linkages between the business sector and the R&D infrastructure, and require the establishment of clear legal and administrative procedures. It is, therefore, recommended to review, adjust and standardize relations on appropriate models. This may concern many issues: use of public or university laboratory equipment and personnel by firms, temporary employment of university researchers by business enterprises, intellectual property rights, etc. · There is in many sectors a need for developing quality awareness and quality control, and related accreditation and certification procedures. A crash program should be implemented with this in view. It could yield very significant results within a short time span. · There is a major financial problem for those firms or individuals who are in the leading position of producers and can not get credit from the banking system. The government has recently established a credit guarantee 13 mechanism which is supposed to facilitate the mobilizing of banks and the financial sector. It does not seem to work well enough. An audit needs to be carried out to examine in detail what mechanisms can be put in place to complement this incentive: micro credit schemes, equity investment procedures (such as the Dutch PSOM which support 30% of investment of individual firms in the flower industry), and others. · Finally there should be a systematic review of obstacles of all nature ­ taxes, regulations, customs, bureaucratic requirements, corruption, etc. -- which affect the innovative process. The best approach is to establish sector audit commissions which would enquire in the concerned communities (entrepreneurs, farmers, etc) and would recommend to the concerned authorities the necessary changes. These commissions should be independent from the usual administrative channels and should not be disbanded after their reports for change and reform. They should be entrusted to examine subsequent follow up actions and report back. An Innovation Multipurpose Facility (IMF) In order to implement and fund efficiently the multi-component program proposed above, it is suggested to establish a new ad hoc facility, which would be administered at the Presidential office level. It should be endowed with an important corpus of funds and should operate with a maximum of flexibility and use for the different types of policy actions mentioned above. This instrument could be named the Innovation Multi-purpose Facility. Funding needs can be estimated at USD 10-15 million per year; if matched by an equivalent spending from the private sector (currently almost zero), this would lead to doubling the current national S&T expenditures. They would then reach 0.5 % of GDP. Such funds could be received from and administered by the WB with the concourse of bilateral and other multilateral organizations. It can reach its full scale within 3-4 years after an initial pilot phase. The impact on the economy would be perceptible within 5 years or so. We would suggest that the whole Government support mechanism be modeled on the TASAF scheme established for social purposes and very efficiently administered at the Presidential level (see Box 2 below). 14 Box 2. Tanzania Social Action Fund (TASAF) The Tanzania Social Action Fund (TASAF) is one of the main development programs that addresses the problem of poverty which is widespread in the country. The main objective of TASAF is to increase and enhance the capacity of poor communities and stakeholders to prioritize, implement or execute, and manage sustainable development initiatives/ projects and, in the process, improve socio-economic services and opportunities. The TASAF leads and contributes positively in reducing poverty by: · Providing extra resources for the creation of community assets at the village level (e.g., schools, bridges, clinics, water points, etc.; · Targeting incomes to very poor households working on public assets like roads, forest lots, and small dams; and · Addressing institutional development issues at the community level, districts, and central government for sustainable poverty reduction measures. TASAF Funding The main sources of funds to finance the activities implemented through TASAF are from IDA and the GOT. However, the Local Government and communities also contribute in cash and in kind. Cumulatively until December 31st 2004, IDA has disbursed US$ 61,216,884.18 and the GOT has paid TZS 3,902,000,000.00 (US$ 3,483,928.57) to match the World Bank funds. About 98% of the funds have been disbursed. Phase I of TASAF (November 2000-October 2004) was confined to 40 Tanzania Mainland districts and Unguja and Pemba (Tanzania Island). A total of 1704 sub-projects were executed in the four years through Community Development Initiatives (CDI- 1338 projects), Social Support Programs (SSP-61 projects) and Public Works Program (PWP-305 projects) arrangements. This is a significant contribution to the quality of life of the concerned communities. Capacity Building The TASAF has completed various training-programs to develop capabilities at all levels. In this respect, through several targeted capacity building interventions, TASAF has been able to train about 20,000 members of democratically elected Community Project committees (CPC); more than 1,500 District Facilitators; about 200 NGOs/CBOs as well as district/island management teams. The CPCs were trained on simple booking, procurement and project management. These CPCs enhanced the implementation of TASAF supported community sub-projects successfully. Key Factors for Success of TASAF · Recognition and Adaptation of Community Know-How: TASAF acknowledges that poor communities in Tanzania have the technical know-how among themselves which can help them implement their own projects. Thus, development of the people can be enhanced only if it is done by the people themselves. · Apolitical non-Partisanship: The TASAF has been apolitical, i.e. being aloof from politics. Thus, it promoted and facilitated all types of people to cooperate and work together for their own development. · Participatory Management Approach: The TASAF has various stakeholders including the Government of Tanzania (GOT), the revolutionary Government of Zanzibar (RGOZ), the World Bank, and the local councils and rural communities. The TASAF promoted and facilitated participation between stakeholders including poor communities in the implementation of the programs. The TASAF confirms that with appropriate guidance and facilitation, poor communities have the necessary capacities to identify, execute, and sustainably own basic services. · Direct Funding: The TASAF implemented the direct funding of the community managed (sub) projects. In addition, mandatory community contribution for community development initiatives has been enforced. · Decentralized Planning Process: The TASAF allowed and strengthened the decentralized planning process for all the activities. Thus, transparency and full public accountability have been promoted. 15 Improving the Broader Innovation Climate Education and Innovation In order to complement the actions proposed above and which provide a relatively direct support to innovation in a medium term perspective, action is recommended on the educational and training needs with a longer term perspective. Progress is noticeable in education enrolments in Tanzania both at the basic and higher levels7. However, a number of issues need to be considered with the innovation perspective in mind: · In the field interviews which we conducted in foreign firms and subsidiaries, a number of our representatives pointed out a lack of intellectual vitality among Tanzanian employees8. This might be a question of generations, as a large number of Tanzanians have been exposed to a socialist type of culture of which some aspects do not stimulate the intellectual inclination needed for innovation. In any event it is strongly suggested to reconsider and change the educational methods to make the youth more autonomous, and more risk- oriented. · In parallel, there is a need to develop a true technical culture (not a theoretical one). This means imparting a more practical knowledge to youth that they can relate to their day-to-day life. This effort should be put in a broader perspective looking to the nature of knowledge and know-how that is taught and acquired by the youth and the student population in Tanzania. The concept of functional literacy ­ going beyond the concept of simple literacy ­ has been elaborated in advanced countries and specific surveys are implemented for testing countries' achievements from this viewpoint. Some developing countries have begun to participate in such surveys. Tanzania could usefully join this group. · In supplement, there are some sectors which are already in need for specific skills. For instance in the cut flower industry, although it is affected by low growth rates, we have been informed of a skill deficit. Appropriate programs have to be designed with concerned colleges and universities in close collaboration with farmers and enterprises. · At a higher level, as in all African countries, there is a need to develop a cadre of well-trained scientists and engineers of international caliber well-plugged into the international networks. This has to take place in centers of excellence. In this connection, the plan to establish in Arusha one of the four 7See Anuja Utz, op.cit 8This was illustrated by comparisons of the number of managers needed to supervise employees in factories producing the same goods in Tanzania and in neighboring countries, and belonging to the same MNC. The ratio of supervisors/employees is significantly higher in Tanzania. 16 centers, currently envisaged, for the Africa ST Institute (Eastern Africa), supported by the Nelson Mandela Foundation, offers very interesting perspectives. · Finally one should think about the creative use of the expatriate diaspora and try and attract it back. Efficient schemes should be designed with this in view. In a similar vein one could think of putting in place mechanisms to attract on a permanent basis a high level of foreign expertise, following for instance the experience of Costa Rica attracting pensioners from advanced countries with powerful fiscal and other incentives. Governance and Innovation Inappropriate conditions of governance are recognized to be a major obstacle to the development of African countries, and Tanzania suffers from them as well, despite noticeable progress over the recent decades. They affect the business environment, as well as the management and development of basic infrastructure including the knowledge-related one such as education and research institutions, without mentioning the conditions of wealth redistribution, including incomes generated by new industries and activities. Experience shows that traditional structures of governance, notably those functioning at local levels, can be an essential element for improving the overall management of African countries, reducing corruption tendencies, facilitating appropriate redistribution of food and other basic goods, moderating actively the decision processes in local communities.... This should be kept in mind, notably when thinking about the institutional mechanisms which can facilitate a true re-appropriation of technology by the local communities and the population at large. Preventing the negative effects associated with the somewhat anarchic development of an industry such as the fishing industry around Lake Victoria requires efficient measures woven into the basic socio-political fabric of the country, in which local communities are fully associated with the key decisions and discuss modalities of development of new activities of local and national significance. To begin with, we would suggest establishing such mechanisms for the further development of the tourism industry which offers considerable opportunities to the country and needs adequate regulations for preventing an anarchic and destructive exploitation of its unique social, cultural and ecological assets. 17 Conclusions To sum up, the innovation climate in Tanzania is weak, and it is doubtful that the current trends of economic growth and poverty reduction will be sustained unless more innovation, in all forms, even the most modest ones, takes place. Government policies pursued in this area so far are insufficient and too passive. Nevertheless a technological infrastructure and a certain administrative capability exist which allow the putting in place of a proactive set of actions that take advantage of the dearth of innovations and goodwill and competencies of actors. This requires a significant investment in a new scheme for innovation promotion devoted principally to serving grass root needs and markets on a truly decentralized basis. For this it is recommended to begin with a pilot phase on a fast track procedure. 18 Annex 1. List of Organizations visited and Persons interviewed Dar Es Salaam 1. Dr. Asifa P. Nanyaro, Director General, TIRDO, Dar es Salaam 2. Mr. Mike Laizer, Director General, SIDO 3. Mr. Rogers Alfayo Msuya, Senior Scientific Officer, COSTECH 4. Eng. Benedict C. Mukama, Senior Scientific Officer, COSTECH 5. Dr. Raphael M.L., Director, Centre for Development and Transfer of Technology, COSTECH 6. Prof. Burton L. M. Mwamila, Dean, College of Engineering and Technology PCET, University of Dar es Salaam 7. Dr. A.K. Temu, Senior Lecturer, , PCET, University of Dar es Salaam 8. Col. M. B. Mashauri, Director General, Tanzania Automative Technology Centre 9. Mr. Rogers Sezinga, Managing Director, Tan Discovery Mineral Consultants Co. 10. Eng. Dr. G.M. Kawiche, Chief Executive, National Housing & Building Research Agency 11. Mr. M.P.J. Ulungi, Assistant Director of Immigration Services, Ministry of Home Affairs 12. Mr. Massawe, Head of Small and Micro Enterprises Unit, Ministry of Industry and Trade 13. Ms. Nkya, Assistant Director of Fishing Division, Ministry of Tourism and Natural Resources 14. Mr. Januarius Mrema, Director of Planning and Policy, Ministry of Industry and Trade 15. Mr. Juma Hamisini, Desk Officer, Science and Technology Division, Ministry of Higher Education, Science and Technology 16. Mr. Finias B. Magesa - Renewable Energy Coordinator, TATEDO (NGO) Arusha 17. Mr. Thomas Dixon, Technoserve (US-based NGO for technology and entrepreneurship) 18. Mr. Ajay Shah, General Manager, Sunflag (Textiles and Garments) Ltd.. 19. Mr. Fred Lyaruu, Cargo Manager, KLM 20. Mr. Eligi M. Mushi, Accountant, TCCIA 21. Mr. Joseph Giovinazzo, General Manager, Hortanzia Ltd.. 22. Mr. E. Ngwandu, Director General, CARMATEC 23. Mr. Wilson Baitam, Director of Agrotechnology, CARMATEC 24. Mr. Leon W. Malisa, Principal Marketing Officer, ESAMI (African Mangement Institute) 25. Mr. Andrew Mollel, Managing Director, Kijenge Animal Products Ltd. 19 26. Mr. Msola, Director General, TEMDO 27. Mr. Colman Ngallu, Chairman, Tanzania Horticulture Association (TAHA), and Managing Director, Tengeru Flowers Ltd. Moshi 28. Prof. Suleiman Chambo, Principal, Moshi Cooperative College 29. Prof. Mallya, KCMC 30. Mr. Kimaro, General Manager, KNCU Mwanza 31. Mr. Rwekaza Charles, Acting Regional Fishery Officer 32. Mr. Lawrence, Finance Manager, Meremeta Mining Company Limited, Musoma 33. Mr. Muthu Subbiah, General Manager, New Mwanza Hotel 34. Mr. Kennedy Chagu, Accountant and Administration Manager, Tancan Mining Services Ltd.. 35. Mr. M. Mbwelle, Administration and Finance Manager, Major Mining Company Ltd.. 20 Annex 2. The Fishery Industry in Tanzania Fishery Industry Potential: Tanzania has a potential both marine and inland fisheries. It is a coastal state endowed with fishery resources. The country has a coastline of about 800 kilometers declared as Exclusive Economic Zone (EEZ) but which has not yet been exploited. The marine water covers 64,000 square kilometers, which includes the Indian Ocean and the EEZ, which covers 223,000 square kilometers. The fresh water includes the riparian-shared waters of the great East African lakes namely Lake Victoria, Tanganyika and Nyasa. There are also other small natural lakes, manmade lakes, river systems and many wetlands with fish potential. All this covers 58,000 square kilometers. These lakes are rich in natural resources including fisheries, which are a major economic resource for Tanzanians in particular. It is estimated that the present annual fish catch is more than 350,000 metric tons. Fishery production from Lake Victoria (on the Tanzanian side) is about 60% of the total inland production. There are many fish processing industries in the country and seven of them are located in Mwanza. These include foreign (Omega Fishing Ltd., Mwanza Fishing Ltd., and Tan Perch Ltd.) and local majority shareholding industries (Vick Fish Ltd., Tanzania Fish Processors, Nile Perch Ltd. and Chain Food Ltd.). Importance of the Fishery Industry: GDP Contribution: The contribution of this sector to the GDP for the past five years has been staggering between 1.6 and 3.1%. However, the fishery sector has a multiplier effect on to other sectors of the economy. For instance, the increased number of engine powered boats and other vessels have increased fuel consumption and the number of operators, etc, and hence stimulated economic growth. Employment: About 500,000 persons are formally and informally employed in fishery activities. In Lake Victoria (on the Tanzanian side) alone the number of fishermen who are permanently employed is 80,000. Also, there are a few others who earn their livelihood from the sector by being employed in the fishing and fishery-related activities such as fish trade, processing, net manufacturing, boat building, etc. In addition, there are various secondary activities, which have emerged mostly in the islands. It is estimated that after every two years the fishery population has witnessed a doubling trend. Traditional fishermen: The traditional sector produces about 90% of the total fish catch in the country; only 10% is derived from industrial fishing. Most of the fish caught is consumed locally while Nile perch, sardines and prawns are for exports. 21 Food Contribution: The Fishery industry contributes about one third of the animal protein or 30% of the total intake to the Tanzanian population. Also, the fishery industry in Lake Victoria provides food for more than 4 Million people living in the Tanzanian catchment. Export Trade: In 2003, Tanzania earned US$ 112,056,067 (F.O.B value) on export of about 37,300 tons of Nile perch fillets and its products. Main Factors behind the Success of the Fishery Industry Mechanized Fishing: The country has witnessed a general shift from traditional to mechanized fishing. For instance, most of the fishermen have replaced the pedal and sail boats with powered vessels. This shift has been possible because of the income accrued to reliable markets and rising price of fish and fishery products in both local and international markets. In addition, the fishermen have received finance from foreign investors who are not allowed to own fishing vessels but only processing plants or industries. The fish processors have facilitated the procurement of fishing vessels, gears and other important tools in order to improve the fish catching and supply to their industries. There are also local manufacturers of fishery technologies including fishnets, boats and other gears. These have facilitated the realization of economies of scale in fishing and fish processing activities with competitive advantages. Compliance with EU Standards: The EU Commission sends experts every year to inspect the fish processing industries, fishermen and other fishery related facilities and resources and determine whether they comply with the EU approved standards. The EU standards are very sensitive on cleanliness, hygiene, environmental health and sanitary conditions, etc. Tanzania has harmonized most of its internal standards with those of the EU and now conducts all the inspections. The country's compliance with such international standards has improved the marketability of the fish and fishery products originating from Tanzania. Community Participation through Licensing Process: Although foreign investment is restricted to the EEZ, the joint venture arrangements between foreigners and local investors have rectified the conditionality. That is, the restricted fish processors use local partners as agents to do fishing on their behalf for their industries. Thus, foreign vessels within the joint venture umbrella can also fish in local lakes, particularly for large-scale exports. In addition, the vessel licensing of local fishermen has created increased opportunity for employment for local people whose earnings are being ploughed back into the economy. Further, the flat licensing fee for investment in EEZ has led to increased investment in fishing although it is still under-fished. Improved Transportation System: The industrial fish processors have been sub- contracting the transport service to carriers based in the EU and other international markets. For instance, during the field survey, we found one cargo flight at Mwanza Airport and the interview confirmed that it had been waiting for the fishery cargo for the past three days. However, this has been possible due to the horizontal networking among 22 the industrial fishery processors who team up their exportable goods and hire cargo flights according to the fishery volume. External and Government Support Schemes: There are various national projects supporting the development of fishery industry in Tanzania. For instance, the Lake Victoria riparian states of Kenya, Uganda and Tanzania have been sharing Lake Victoria under the Lake Victoria Fisheries Organisation (LVFO) supervision since 1994. The programs under the LVFO include resource, environment and socio-economic research and monitoring; Fisheries policy, legislation, institution and processes; Aquaculture research and development; Human resources and infrastructure capacity building; and Information and database. The three projects under implementation include the fisheries management plan (IFMP) financed by the EU (Euro 29.9 Million); the socio-economics of the Nile Perch fishery on Lake Victoria Phase II (NORAD/IUCN); and the production and marketing of value-added fishery products in East and Central Africa (CFC/COMESA/FAO). Expansion in the Internal Market: There is high demand for fish and fishery products such as fresh fillet, frozen fillet, etc. Such high demand is related to the high rate of population growth and the diminishing rate of other substitutes of fishery products like beef. The Acting Regional Fishery Officer in Mwanza stated that about TZS 400 million is running the fishery economy every day. Such high liquidity finances the economies of scale in fishery-related activities and hence its growth. Outstanding Problems in the Fishery Industry Overfishing: The development of export markets especially for the Nile perch has attracted many investors into the Lake Victoria fishery industry. However, this has resulted in high fishing pressure accompanied by unsustainable fishing practices and methods and degradation of the fish habitat, which endanger future sustainability of Lake Victoria fisheries. The recent research program (Lake Victoria Fisheries Research Project-LVFP) has found that the fish stocks in Lake Victoria are declining. The government has planned several projects to address identified shortfalls in the fishery industry with the aim of making it sustainable. Unsustainable fishing practices: Fish is renewable resources which are limited, and therefore, they have to be conserved, managed and developed on a sustainable basis. The interviews revealed that there is illegal fishing using inappropriate gears, which kills immature fish and hence threatens the fishery ecosystem. Also, the eutrophication processes, indigenous fish species and frequent fish kills put the lakes under intense pressure from both human and natural processes. Further, deforestation and poor control of industrial effluents in the potential resources (lakes, rivers and seas/ oceans) affects the fishery industry. Investment Discrimination: Foreign investors are discriminated against the local ones in terms of fees. The study has noted that there is a significant gap between the foreign and local fishing fees. In addition, the fish processors are not allowed to conduct fishing. 23 However, due to the high corruption of responsible authorities and conflicting views and decisions between politicians and experts, some of the laws and regulations are difficult to enforce. Bureaucratic Registration and Licensing Process: The government separates boat registration from licensing to provide for administrative steps, additional service to the private sector and supports government recordkeeping and sources. For instance, the procured boat could be registered and kept in the dock for about six months before its' licensing. Alternatively, the registration and licensing processes can be carried out simultaneously, requiring no additional time. High Post-Harvest Losses: The post harvest loss is one of the problems identified in our interviews. The interviews reported that most of traditional fishermen lack the capacity to procure modern fishing boats that would accommodate ice block, fish and fishing gears. This amalgamates food insecurity and unsustainable fishing practices thus resulting in poverty. Lack of capital among fishermen resulting in poor fishing equipment: There are traditional fishermen who lack resources to procure engine-powered boats and hence strain to catch immature fish in the offshore areas using pedal boats or other gears. Ineffective enforcement of fishery laws and its subsidiaries: As pointed out above, the fishery Act of 1970 has governed the fishing industry. This Act provides for parastatal and central executed economy at the expense of the private sector. With the recent reforms, the Act has almost been outdated. In a way, the fishery industry lacks effective legislation and regulatory mechanisms and hence its economic performance is limited. Lack of entrepreneurship and fish farming skills: The country has a low level of entrepreneurial and fish farming skills like the making of fishponds, etc. Most fishermen depend on natural systems like rivers, lakes and sea for fishery activities. For instance, fishermen have suffered due to inadequate training in fish farming skills, inadequate fish farming inputs, hatchery and pond construction and management techniques, lack of entrepreneurial skills in order to diversify fishing activities, etc. Poor infrastructure: The fishing industry suffers from poor landing sites and beaches and other social and physical infrastructures. For instance, the interviews reported that there is a high post-harvest loss due to lack of storage facilities, poor roads and landing sites. This could as well be attributed to poor fishery extension services provided by ill- trained fishery officers/ personnel. As a result, the quality of the netted fish has been poor. Furthermore, the lack of effective fisherman associations, cooperatives or groups is responsible for a gap in the participatory management of the fishery industry. These weaken the coordination and cooperation of fishery stakeholders to uphold their interests as compared to the other sectors of the economy. Low Market Price of Fish and Fish Products and High Tax Rates: Despite high demand and declining stocks, fishermen indicated that they were getting very low prices 24 compared to the international market (although no figure was mentioned). Also, the lack of value-adding processes adversely affect the market price of fishery products. 25 Annex 3: The Cut Flower Industry Background Tanzania has a young flower industry as compared to Kenya and other countries. It has started dealing in cut flowers since the early 1990s. Before the 90s, there were only two companies, which dealt with natural summer flowers planted and sold in Europe. The significant changes occurred in 1996/1997, mainly due to an external push, support and effective demand. Currently, Tanzania has the capacity to export 100 tons of commodities (mainly flowers) every week. Innovations in the Cut Flower Industry The hydroponic green-houses: After the 1990s, the new technology introduced normal green-houses. Currently, Tanzania has introduced hydroponic green-houses. These houses grow flowers on soil-less media, use plastic, drip lines, and plant flowers. It is believed that this new technology depends 100% on what you give it including fertilizer, water, and temperature. The technology also requires the investor to feed the green-house from the fertilization room (or a pump-house). This technological innovation ensures management of plant in a pre-planned way. That is, the flower-houses depend upon predetermined balanced levels of fertilizers, irrigation, and chemicals. For instance, the fertilization rooms under the field assistant have to determine the pH level of the ordinary water whether high or low by using the pH meter and predicing their effect on the plant. In case of deficiency or shortage of any required input, the computer sensor indicates the levels of required inputs in the field or the green-house. Cargo Flights: The KLM and other cargo flights facilitate trading between Tanzania and the external world. Of late, the KLM has been operating Boeing 747s and has recently introduced the MD11 for carrying passengers and a considerable volume of cargo to EU and other markets. Thus, the KLM is commended for providing a solution to international problems affecting fresh produce in terms of speedy transportation and adaptation of sanitary and phyto-sanitary (SPS) conditions with regard to flowers and vegetables. Joint Venture arrangements: most of the successful flower farms are joint ventures between local people and foreigners mainly from the Dutch and other EU countries. This arrangement, at least, ensures marketing, cost minimization and high returns for their fresh produce. Employment The flower industry is labor-intensive with many people employed at very low wages and salaries. For instance, in the Arusha and Kilimanjaro regions alone, the flower industry employs about 4000 people, the fresh vegetables (fruits have not been tried yet) about 26 3000 people, and 1000 people are employed in semi-processing (totaling 8000). The level of employment has been increasing to improve the performance in terms of export volumes. Since there has not been much assistance in terms of productivity, the flower growers have kept on adding volumes to take care of high costs and low margin earned with flowers. The flower industry survives through increase in export volume and wage- cuts for supporting continuous investment in order to expand farms and increase employment. Collaboration between Tanzania and Kenya The producers in both countries are collaborating in various areas. Implicitly, they share knowledge, skills and employment of workers who come from other farms. Of late, the leaders of Tanzania Horticultural Association Ltd. (THAL) and Kenya Flower Council (KFC) have met in Nairobi and prepared a joint proposal of identified issues for technical assistance by potential donors. The key issues identified include the market price and freight. They are also planning for a joint research in order to improve the general performance of the flower industry in both countries. Key Constraints on the Flower Industry Market Prices: The interviews suggest that Tanzanian trade in produces has not broken free from the monopoly of international markets against African growers. Tanzanian growers (for flowers included) have also no say at all in the pricing of the commodity, have no say in the pricing of agricultural inputs and prices of green-houses nor on the freight. Flower prices are determined in favor of the foreign investors. For instance, the auction market (e.g., Dutch auction) has worked out that the Dutch farmers get a net profit 10% for each kg of flowers. On the other hand, the same market has worked out the mechanism to ensure that Tanzanian (traditional) farmers get a gross margin of 16% per kg, despite the level of costs incurred. This mechanism entails that in order to survive in the flower industry local artisan farmers are forced to enter into a joint venture with the Dutch farmers. Their survival depends on this with regard to Dutch auctions. These issues have been adversely influencing the flower industry in Tanzania. High freight rate and landing, parking, handling and other fees: The major cost factor is freight. The freight is high due to high landing and parking fees in Tanzania. For instance, the KLM complains that Tanzania charges very high landing and parking fees and fuel prices, 15% more compared to Kenya, Uganda and Ethiopia. In addition, DAHACO levies US$0.04 for each kg handled by each carrier and exporter, creating a cost of US$0.08 per kg of cargo. Other charges/fees are attributed to prerequisites of international civil aviation organization (ICAO), Air service license (ASL) and air operator certificates (AOC), etc. Consequently, the profit margin is low and thus discourages many airlines from operating flights in the country. Unequal competition: In addition, artisan farmers incur a cost of about 45% of the market price while the Dutch farmers incur only 1%. The governments of the respective countries subsidize the balance. For instance, a consignment with market value of Euros 27 16,000 will generate a profit of Euro 15,000 to the Dutch Farmers, not comparable to Euros 8,000 earned by the traditional farmers of East African Countries. The main items contributing to such huge costs include terrorism, insurance, etc, in Africa. Unfair Price Competition: The pricing and costing aspects as carried out by the international flower auctions reveal that there is no competition. That is, a traditional grower who incurs 45% cost per kg and earns 16% as gross margin per kg can not compete with a Dutch farmer who earns a net margin of 10% while incurring only 1% of the total cost per kg of flowers. Poor Customs Procedures: The TRA has introduced a new post-shipment inspection, where all imports have to go through the T-SCAN system. Therefore, all imports have to go through the DIA for inspection, certification and approval for taxation. The KLM, which lands at KIA (Kilimanjaro International Airport) before DIA (Dar Es Salaam International Airport), has thus a limited space to carry even 20 tons of flowers per day. As a solution, the KLM has introduced the process of offloading imports at KIA and trucking them to DIA under hired motorcade or convoy with police and customs officials at huge costs. The KLM confirmed that this process would not be sustainable and would stop it soon because of the costs, delays, damages and sometimes loss of items belonging to importers. Unreliable and Insufficient Volume of Exports and the Kenyan Route: the export consignment of flower-cuttings and cut flowers has shown a declining trend in the past three years. There is no new farming project which has been implemented so far in the flower industry in this area. In addition, the Kenyan route, which carries about 40% of the export consignment, discourages the carriers from introducing a new cargo flight in the region. Besides, some advantages attributed to the improved export trade between countries have been grabbed by Kenya through this route at the expense of Tanzanian farmers. The KLM, which introduced the MD11 (bigger freight capacity than the Boeing 747 model), is going to stop it after 2 months of operation due to diseconomies of scale reflected in losses attributed to a limited volume of exportables. Growers can not afford cost pertaining to transportation of cargo to DIA for KLM carriers and instead they prefer transportation to Jomo Kenyata International Airport (JKIA) for other carriers despite the hassles experienced with regard to time-consuming custom procedures, road tolls and poor roads on the Kenyan route. 28 Executive Summary: Supply Chain Development in Tanzania Conceptual Framework: A review of recent development experience suggest that "liberalization" stimulates economic growth and that economies which are lightly regulated tend to grow faster than more intensely regulated economies. This is so apparently because markets allow private agents to direct scarce resources to their most productive applications when markets are unimpeded and unobstructed in their operation. This "superior targeting capability" is a direct consequence of competition and in particular of the kind of competition which takes place among independent private agents who bid for resources in open and contestable markets. Minimally restricting competition among producers assures under ideal circumstances that resources flow to their best and most efficient use. These ideal circumstances apply, however, only when information is equally available to buyers and sellers; when capital flows efficiently to both the "buy" and "sell" sides of transactions; when product storage and post harvest holding facilities are equally available to buyers and sellers; and, finally, when markets for management capabilities and for third party services also operate efficiently and equitably on both the "buy" and sell "side" of transactions within supply chains. When these conditions do not hold risks are unequally shared between buyers and sellers and opportunities to realize excessive profits, in the form of rents, accrue to those commercial partners who hold superior positions within chains---particularly within traditional chains whose linkages are based on one-off, arms length transactions. Under such circumstances incentives cease to be effective in motivating trading partners who enjoy superior positions within chains to invest in best technologies, to reorganize chains into more efficient industrial structures and to implement best management practices. Under circumstances where the advantages of " liberalization policies" tip over into disadvantages, trading partners who enjoy a superior position have the option of retreating into commercial safe harbors within the chains that they operate. In these safe harbors rents can be collected without contest and above average returns realized without putting fixed capital at risk. From these superior positions chain integrators are able to assign risks to other participants who are less well able to manage them effectively. Under ideal circumstances market prices established among equally well informed and equally well financed buyers and sellers reflect the marginal value of all goods and services traded. Unconstrained competition and the efficient resource allocation that follows from it should, over a significant period of time, accelerate economic growth. However, these "ideal" market conditions often fail to become operative in developing countries. Hence, efficient resource allocation is by no means assured when private sector agents are left to their own resources, unconstrained and unregulated. Indeed, as we find with the thee case studies reviewed here, the initial competitive conditions--the conditions that occur immediately upon loosening direct government involvement in specific sectors----matters a great deal in determining how efficiently resources are allocated to accelerate growth and to increase competitiveness within specific sectors. Under real world circumstances, government interventions themselves are often the cause of initial structural imbalances and subsequent inequities. Under circumstances where initial structural imbalances occur within supply chains, disengagement on the part of government may not suffice to move subsequent industrial organization toward pro-poor structures and more equable chains. Indeed, subsequent regulation and/or strong incentives may be required in order to reinitialize the system and hence facilitate the escape of disadvantaged commercial partners from what otherwise become a poverty trap. For economies who are seeking to accelerate their economic growth and at the same time assure that that growth is pro-poor, cooperation in the accumulation of social capital among trading partners can be almost as important as competition in assuring the resources are efficiently allocated. Development economists have much less to say about cooperation than they do about competition. Why is cooperation important? Cooperation is essential to assure that transaction costs, incurred between trading partners, are minimized. Wealth does not accumulate as quickly or, indeed, at all in a high transaction cost sector. Moreover, competition itself is learned within sectors through some implicit or explicit form of cooperation among competing parties. Thus, cooperation is important first and foremost in establishing norms and standards for legitimate cooperation---competition which is socially beneficial in its net impact. Cooperation is essential, as well, for assuring that redundant commercial processes are avoided, that institutions ( including both market and supply chain institutions) are organized in ways that allows products, information, payments ( and credits) and ownership rights to flow efficiently among trading partners and that strategically interrelated industrial processes are allowed to combine into efficient industrial organizational structures. Increasingly, private companies compete with other private companies not on a one-on-one basis but as parts of larger industrial systems. Indeed, they compete as parts of supply chain systems. Supply chain systems that have rent collecting activities imbedded within them cannot compete effectively in global markets where new competing supply chain surface every year from new and unexpected directions. Innovative supply chain structures, moreover, are increasingly the primary means for creating and sustaining competitive advantage. Witness the global supermarket revolution which is underway in Tanzania and which is quickly substituting sophisticated supply chain structures for traditional farm market institutions and traditional trading chains. Importantly as well, supply chains have become the principle agents for the transfer of technologies and of advanced management techniques In the 21st century it is supply chains that provide the basic mechanisms through which supply and demand are matched in retail end markets which are located some distance from where products are produced. For all these reasons cooperation in the form of refined industrial structures matters a great deal for development. The study which follows deals with these and related issues in the context of the development of the agro-industrial sector of Tanzania's economy. First, however, some background information on supply chains....what they are and what they do......may be useful to set the stage for the more specific-to-Tanzania findings that follow. What then are supply chains? Supply chains are forms of industrial organization, which link together economic agents all of whom add value to primary products as these products move from primary production points to retail markets. Supply chains operate primarily within specific products or commodity sectors. Indeed, they are the industrial organization that informs specific product sectors. Supply chain participants typically include the producers of primary products, value adding processors, manufacturers, intermediate merchandisers and retailers. Importantly, supply chains are supported by specialized third party service suppliers who furnish essential services of various kinds to chain participants. These third parties operate on the periphery of most chains and their relationship to chain participants is typically arms length and transactional. Both the level of competition among third parties and the state of customized service development greatly effect the level of transaction cost absorbed by chain participants. Thus the state of third party service providers materially effects supply chain productivity and agility. Arms length relationships between service providers and supply chain principals evolve over time as the need for more specialized and more dedicated service support intensifies. Thus, in response to external competition supply chain principals and third party service providers sometimes share risks with one another in ways which reduces the net investment risk for the entire chain. Thus, they form service partnerships of various kinds. Risk sharing based on the assignment of specific risks to specific supply chain partners who are best equipped to manage them is an important source of competitive advantage which supply chains provide. In high risk economies like Tanzania's supply chain organizational structures can provide a natural immune system against multiple categories of systematic risk. Like other forms of industrial organization supply chains establish and enforce rules which effect the behaviors and the performance of individual economic agents who become principals in the chains in which they participate.. Together with corporations (which create value by coordinating multiple value adding functions efficiently and by accessing different forms of investment capital efficiently), product markets and industrial clusters, supply chains populate the diverse ecosystems of most developing economies. Importantly, in the ecosystems which they populate they either enhance or degrade industrial competitiveness depending on the way that they are structured, depending on their collective capacity to absorb new productivity enhancing management methods and technologies and their adaptability to respond quickly to new types of competition. The least developed forms of supply chains and the form which predominates in most sectors of Tanzania's economy involve "arms length", transaction relationships among a set of economic agents whose chain relationships are essentially reinvented with each individual buy/sell transaction. These are high cost and high risk forms of industrial structure. Within these structures risks tend to reside with the weakest link, instead of being assigned to those links which are best able to manage them. The supply chain structures found within specific economies can and do have a significant impact on economic development. How do supply chains help to reduce transaction cost and transaction risks? Supply chains allow participants to impose contingent claims on each others and to enforce those claims effectively. Mutual and contingent claims are enforced in one of four primary ways: i) though contracts which contain incentives for risk sharing and for efficient process coordination; ii) through unified corporate control and joint or cross ownership of corporate assets; iii) through the joint development and management of specific supply chain assets, including distribution channels, brands and order fulfillment systems; increasingly, iv) through the mutual adoption of common IT and business management and control software, and finally v) through the development of social capital within trading communities and the leveraging of that capital to effect compliance on the part of individual agents. Compliance is induced typically under threat of exclusion from the trade or other forms of social sanctions. One of the reasons that supply chains seem to develop more often among ethnic minorities in Tanzania is the social capital that accumulates among principals who share rich community ties. Different forms of organizational adhesion apply in different competitive circumstances. Supply chain development is highly circumstantial. Since supply chains are typically anchored at both their production and the consumption ends in contestable markets, under normal competitive circumstances prices at both the primary production and the retail distribution end of chains are market determined and set independently of the policies of the chain. Under these circumstances it is the efficiency with which chains themselves operate more than any other single factor that determines both the competitiveness of chain participants and the well being of affiliated, primary producers who are, in the Tanzanian context, small scale farmers or small scale fishermen. Supply chains operate more or less efficiently depending on the velocity with which working capital absorbed within the supply chain turns over or, said another way, depending on the return which is realized on end-to-end investment in working capital. Supply chains which are effectively integrated move products quickly, they move them at low cost and with a minimum of physical loss. Supply chains create internal incentives which encourage the rapid movement of goods, the minimization of excess inventory and the efficient use of working capital. Their effectiveness in integrating the work activities of multiple economic agents who are broadly disbursed geographically is determined, in significant part, by the kinds of control systems used within chains, as well as by the effectiveness of the incentives which encourage collaboration and process coordination among separate agents. What is the relationship between supply chain development and poverty alleviation? Three parameters are crucially important in determining the "pro-poor" or the "anti-poor" effects of specific supply chain structures. These three defining parameters include: i) efficiency; ii) adaptability; and iii) equity. Supply chain "efficiency" relates to the speed and the integrity with which products, information, ownership rights and credit ( or cash) more through supply chains. From one perspective supply chains can be viewed as conduits through which these four flows move more or less rapidly and more or less securely. "Adaptability" has both a short term and a long term aspect. In the short term adaptability relates to the flexibility with which supply at one end of the chain is matched with demand at the other end. To make this point in more operational terms, adaptability relates to how production is scheduled and how closely it is matched with demand. Over the longer term adaptability relates to how efficiently new technology is incorporate into chains, to how quickly new products are designed and delivered to market, and to how efficiently basic business models are changed in response to changes in the business environment. For example, business model changes may relate to how responsively critical process steps within chains are either in-sourced or out-sourced depending on where essential work can most efficiently be performed or how quickly new niche markets can be accessed through strategic affiliation with new distribution channel partners. Finally, supply chain "equity" relates to the economic power relationships which prevail within chains. Operationally this parameter relates to whether information and working capital credits flow with equal efficiency in both directions within chains, to whether the efficiency with which risks incurred by individual participants within chains are assigned to chain participants who have the best capability to manage these risks, to how open are supply systems to new participants and, finally, to how contestably chains themselves are governed and to how they are organized in ways that benefit one set of participants over others. Relative returns on working capital among individual participants typically reflect the prevailing economic power relationships within chains. Equal returns among chain participants signify high levels of equity. Whereas, skewed returns signal low levels of equity. Thus, for example, chain integrators who originally organize the industrial structures of chains typically enjoy higher returns than do passive chain participants whose participation is more subject to possible competitive replacement. What holds supply chains together? Supply chains are typically integrated in one of four ways. Said another way, the organizational " glue" which binds value adding activities together within a single chain include, either: i) ownership or cross ownership interests among chain participants. Thus, chains expand their vertical domain within vertically integrated corporations through some combination of in- sourcing, merger, acquisition or joint venturing; ii) agreement to act jointly and to coordinate production/ distribution/ merchandising processes across the economic boundary lines of individual agents. Contracts in the form of marketing franchises, long term supply agreements and preferred vendor contracts are examples of this form of integration ; iii) shared assets including both fixed and network assets. Assets can be differentiated by the function they perform. Supply chain assets typically facilitate the work of linking production functions to sales functions. Productive investment in this category of assets depends on the strength of the linkages which exists among end to end participants in chains. Thus, investment preference for production assets signals weak supply chain linkages whereas investment in distribution, order fulfillment and inventory management assets signals strong supply chain linkages. Increasingly, it is the joint and shared use of IT networks, process control software, resource management systems and other "management ware" which distinguishes the "inside" from the "outside" of supply chain structures; iv) social capital developed between and among trading partners and end to end commercial agents. Social capital is important in reducing transaction costs and in setting the background conditions for productive investment in other forms of supply chain assets. Social capital takes, however, multiple forms. To take one example: One form of social capital are standards for value differentiated products or product quality grades which are shared among trading partners. These standards together with the quality control systems which preservation value grade distinctions are critically important to improving chain competitiveness. Similarly generally accepted terms and conditions for product credit sales and institutions which create and preserve security interests in product inventories can substantially reduce supply chain financial risk and improve chain bank-ability. Social capital is typically developed within a community of trading and/or processing partners with the objective of excluding value subtracting agents and value subtracting process steps from chains and hence improving the residual returns on working capital realized by the remaining value adding agents. Where social capital is depreciated supply chains operate as ` zero sum' or as ` negative sum' games among participants. Reversing this situation and creating incentives for social capital accumulation is one path that can lead to escape from a poverty trap. What is the relationship between individual companies, product markets and supply chains? The role of supply chains is critical both in enhancing the competitiveness of individual companies and in assuring the efficiency of specific product markets. Without effective integration into supply chains individual economic agents experience high transaction costs. Thus, supply chain integration almost always reduces transaction costs by improving coordination among participants and by excluding rent collecting ( value subtracting) agents from participating. Importantly, individual agents are limited in their ability to enhance their own productivity by the limited number of factors within the control of their managements. Supply chain integration extends effective management control forward to retail customers and backward to primary supplies of essential inputs. Efficiently designed chains always improve access of participants to information, to customers, to productivity enhancing technology and to additional degrees of freedom with respect to potential management responses to unanticipated adverse circumstances. Supply chains allow producers both to differentiate their products based on quality parameters and to adapt their product offerings to changing consumer demands Supply chains likewise improve the efficiency of the markets that they serve. Without being effectively supported by supply chains, markets fail to clear at production levels where marginal cost and benefit are equal. Supply chain systems schedule production and manage inventory so that demand and supply balance at the retail end of chains. Importantly as well, when supply chain support is weak markets are limited in their geographic scope and are products are further limited in their ability to adjust to dynamic price changes through in-effective spatial and temporal arbitrage. When chains fail to preserve and protect product value which results from quality differentiating product characteristics, refined market segments cannot emerge and the wealth creating opportunities that niche market development can open are simply foreclosed. The bottom line is simply this: Effective integration of value adding activities through the development of efficient and flexible supply chain organizational structures is essential to rapid economic development. In its need to improve both the efficient and flexibility of its supply chain structures, Tanzania, is representative of many other African countries. Study Objectives: The study which is summarized below was undertaken because the Government of Tanzania has become increasingly concerned about the growing differences that exist between farm-gate prices, at one end of traditional supply chains, and retail or export prices, on the other end. Indeed, evidence cited in the study suggests that at least for one important farm sector marketing margins were increasing....this in a sector which has been " liberalized." The policies of market liberalization and of regulatory disengagement implemented by the Government of Tanzania over the past ten years may, indeed, have lead to an increase in marketing margins for maize because of the structural effects these policies have had on farm-to-market chains. The implication of increasing farm-to-market price spreads are that transaction costs within traditional farm to market chains are growing and that value is being somehow removed from these chains. The initial objective of this study then was to provide an etiology for these "rent" collection activities. If rents were being removed from specific product/ commodity chains what was the cause or causes? Thus, one line of inquiry was to determine whether some form of monopoly control had taken effect over specific key activities in farm to market chains---activities which for some reason had become closed to contest and to new entry. Another potential set of explorations involved determining whether redundant processes had somehow become structured into farm to market chains---processes which removed more value than they created within chains. Thus, for example, one hypothesis was that redundant trading was taking place because information flows among chain participants was asymmetric or because credit was unequally accessible to different categories of traders. Still a third line of inquiry was that the provision of third party services inefficient and that for various reasons markets for chain support services had failed. Thus, for example, it might be the case that transport service providers for a number of reasons had failed to price on a marginal cost basis or that intermodal competition was insufficient to assure that the limited supplies of rail or barge services were being allocated efficiently. However, if initial set of objectives provided an initial entry point, the supply chain studies that followed lead to other objectives that revealed themselves only in the context and the process of carrying out the study. The comparative nature of the study, provided a opportunity to assess what forms of government intervention appeared to be most effective in guiding the structural development of specific sectors of Tanzania's economy. A second objective than was to assess what policies appeared to be most effective and which least effective in guiding pro-poverty growth. As it turns out the regulatory role which government plays at the level of specific commodity or product sectors appears to have a major effect not only on chain efficiency but also on chain adaptability and equity. Still a third objective was to envision a program of post-liberalization, post-privatization reform for specific sectors of Tanzania's economy. This agenda was to be based on the findings that emerged from this study. Study Approach: The study which is summarized below focuses on three supply chain structures which the Government of Tanzania selected jointly with the World Bank for critical analytic scrutiny. The basis for selecting these three chains was threefold: i) The three should provide a degree of diversity in the competitive market challenges which each chain faces in accessing either domestic, regional and/or global markets or some combination thereof; ii) The three should provide a degree of diversity in their level of organizational development and in the initial conditions under which each chain was launched or re-launched as a result of a shift in government policy, privatization action, market liberalization, etc.; iii) direct and significant effects derived from the structure and efficiency of each chain selected should be significant on the welfare, income and potential for asset accumulation of poor farmers, factory laborers and fishermen. On this basis three products/ commodities were initially selected. These were discussed with and ultimately agreed upon with the Government of Tanzania. The three include supply chains through which maize, fish and sugar move from producers in Tanzania to market. The table below summarizes the basis on which these chains were assessed and subsequently selected for further study. Number of Primary Ratio of Policy/ Degree of Annual Degrees of Poor Markets Farm Regulatory/ Competitiveness Rate of Freedom Families into Gate or Investment vis a vis Production Available to Involved which Fisherman's Conditions Alternative Growth Primary in Production/ the Dock Sources Producers, Distribution Product ( for fish) Which Allow is Sold. Price to Them to Align Retail Themselves Price with Alternative Chains Maize 500,000 local Still to be Liberalization of the sector and minimal erratic Limited government disengagement from direct and added interventions in maize markets except regional for the emergency food program. No system of governance or of consultation exists between either the Food Security Department or other branches of government involved with maize markets and the private sector. Fish 110,000 local Still to be The Fisheries Act sets out the legal significant 3-4% Expanding framework for the environmental and and added economically sustainable management of global fisheries in Tanzania. It provides specific authorities of government to determine who can fish, what fishing and processing methods can be used and what commercial methods can be used to process and trade fish. The Act does not create any specific counterpart institutions in the private sector with whom the Fisheries Department is required to consult or liaise. Sugar 90,000 local Still to be moderate 5% Open and added Expanding The methodology used to complete the ESW project is described in detail elsewhere in the study. Very briefly, however, the approach taken was to apply systematically a methodology for measuring and assessing the economic value addition which takes place in each link in each end-to-end supply chain. To this end, each chain was first mapped into all of its component value adding" economic activities. Next, cash flow requirements or " activity costs" were estimated for each value added process step in each chain. In addition, the timing of each of these working capital investment activities were estimated, so that a full cash in to cash out cycle was constructed for each chain. At the same time, redundant, non value adding and/or inefficient processes were identified and isolated. Comparisons were made with respect to the unit cost of providing essential activities across production platforms where these comparisons were available. On this basis, inefficient processes or non competitive third party provided services were identified. Returns on working capital were then calculated for the entire end-to-end chain and for the activities of each discrete economic agent within the chain. On this basis, the efficiency of the entire end to end industrial structure was determined and the relative economic power of each of the principle agents in the chain determined. Next a set of " what if questions" were posed concerning alternative chain structures, improved process efficiency, new market and other institutional developments, etc. Again returns on working capital were calculated in response to each " what if scenario." On the basis of this kind of interactive testing of initiatives, reforms and industrial restructuring initiatives an agenda of policy recommendations, initiatives and tactical programs emerged. In addition, an assessment was made of the adequacy of key institutions which support efficient chains in specific sectors. A great deal of supply chain development involves the accumulation of social capital in form of shared quality standards, shared commercial rules and trading protocols, trusting and transparent transactions and mutual support among cooperating members of a production/ processing/ merchandising chain. Institutions are the reservoirs that contain social capital. Thus, an important element of the study involved the assessment of existing institutions. Thus, for example, the study found that relatively few mezzo level organizations ( e.g. organizations which fall half way between public sector organizations and individual private firms) have emerged in Tanzania. The prevailing cultural inclination is to compete and to act on private access to opportunity and on privileged information. Thus, a set of negative sum games have been organized among trading partners who perceive that the best way to succeed in business is to diminish the value received from or deliver to other trading partners. Reversing this prevailing predilection is only possible through a series of social capital accumulation exercises and through institutional renewal. The end result will be the partial transformation of Tanzania's negative sum game commercial culture into a culture that supports a positive sum game among supply chain partners. Fish Supply Chain Two very different supply chain systems, which support lake and ocean to market commercial operations, co-exist in Tanzania. The two systems serve two distinct markets: one international and one domestic. One of these supply chain systems is sophisticated. It entails low transaction costs, is well invested and falls just a half step below best international practice. The other is rudimentary in technique and technology and is poorly organized. It entails high transaction costs and incurs substantial risk for fishermen and boat operators. The amount of investment in the domestic supply chain, moreover, is de minimus compared to the level of investment in the parallel export chain. The same primary products move through both channels. However, the way in which they are processed and marketed is quite different. Some of the markets into which Tanzanian fish move are domestic local markets and others are international markets. A substantial portion of Tanzania's fish move in a fresh product form, some move to market in a fast frozen form and still others are processed and distributed in a salted, smoked or cooked form. Distinct distribution channels access each of these two markets. Thus, fish distribution channels in Tanzania vary in sophistication from "near-World Class" to rudimentary, and from "best in class" food safety practice to "hit or miss" practices which are neither controlled nor regulated. Moreover, the service industries which have developed on the periphery of these two supply chains have also developed along separate tracks. Little cross over service provision and even less common use of assets ( e.g. cold chain, warehousing, information systems, etc.) takes place beyond the point at which primary fish products are landed, sorted and sold. In the chain which supports the domestic market, for example, transactions are executed directly between fishermen and a relatively small number of retailers, wholesalers, buyers agents and, even, consumers who buy at fish landing sites and pay for them on a cash and carry basis. Multiple buy/sell transactions separate producers from consumer. Market intermediaries in the domestic market are highly specialized by type of fish and by end market. They typically resell the fish they buy into specific markets in which they believe their superior knowledge and access to local commercial networks can realize above market returns. In many parts of Tanzania, fish marketing is limited to the local village in which fishermen are domiciled or, indeed, to neighboring communities within walking distance. Little trading takes place in support of long distance commerce. As a result no integrated market exists in Tanzania for fish. Wholesale markets develop where a minimum level of logistics and transportation competency develops and where price information services emerge. The two primary commercial clusters where these two pre-requisites exist in Tanzania are situated in Mwanza and Dar es Salaam. Thus, the Banda Beach Market in Dar es Salaam is probably the most sophisticated market center in the country for reef fish. Importantly it is also the primary wholesale center for reselling into Dar es Salaam which is the largest retail market in the country. The Kirumba Market in Mwanza has assumed a parallel function for Lake fish. It has become the logistics center, the focal point for price discovery and the primary base for wholesalers who are trying to develop a national market for artisan processed fresh water fish. In both of these market centers wholesalers, retailers, storehouse operators, transport operators and retail customers co-mingle and cooperate in forming the most efficient of the ad hoc domestic supply chains that serve Tanzania.. The Banda market handles more than 10, 000 tons annually of a variety of fish products. The Kirumba market handles 15,000 tons annually, including mostly artisan processed dagaa and Nile perch. In both markets larger lots are reserved for sale to wholesalers recognized by the market itself and are auctioned to these wholesalers using a set of rules that conform approximately to Dutch auction principles. Smaller lots of fish are sold through retail market stalls and informally through other wholesalers. For these transactions prices are negotiated on a sale by sale basis. Both markets acts as a focal points for price formation. Competition on both markets is quite intense. However, prices are not effectively transmitted beyond these two market centers and the governance principles which apply to trades within the two centers continue to advantage local traders vis a vis others. Other markets for fresh fish remain under developed and local. With that said, it must be pointed out that the wholesale link in the domestic supply chain for fish products is under developed. Thus, for example, only basic ice cooled surface transport exists to move fresh fish long distances. Significantly, no inter coastal or cross Lake shipping capacity, interior air or rail services operate to support the national fish trade. No security interest are created in fish inventories, no intermediated transfers of fish products take place. Rather all trades are on a cash and carry, quid pro quo basis and hence require an expensive amount of time to negotiate. No regulatory framework exists either to protect customers or buyers/ resellers with respect to trading protocols, contract enforcement and/or liabilities related to safe food handling practices. Moreover, only limited market information is available and no third parties have entered the market for market information services. In other words, many of the essential components required to develop an integrated national market are missing. Hence, trading in fish remains an inefficient activity with high product losses, high trading risks and extremely high transaction costs. The only national market which has emerged to date is one for processes fish, e.g. dried, salted, fried and otherwise preserved fish products. With that said a growing market is developing within Tanzania for fresh fish. Supermarkets and fast food chains are leading this market development. In addition, an array of fish dealers, peddlers, small scale fried fish processor, fish retailers and restaurant owners buy fresh fish every day either directly from fishermen or indirectly through agents or, indeed, buy from each other. The short economic life of fish products requires expeditious trading, internal quality controls and rapid disposal of the product once it is purchased. Whatever other limitations they may have participants in the domestic fish supply chain clearly understand the need for expeditious trading, transporting and sale. Because of a lack of recorded data, the precise number of small-scale traders involved in the fish business is unknown. However, a rough estimate provided by one knowledgeable industry participant has this number exceeding 60,000. An informal survey of customers at the Kirumba market in Mawanza revealed that, with the exception of special species, most of the fish catch was consumed locally. From this survey the study team inferred that the distribution of fish products is on a small ­scale and that 80% is limited to buyers who are based within one trip day of the market. The exception is a few species of fish for which strong demand exists throughout the country. These are fish species that Tanzanian's are prepared to pay a premium to consume or to use as animal feed. Thus, Nile Perch and Dagaa, are marketed over a much more extended geography than most other species from Mawanza. . Of the supply chains assessed in this study, the chain which supports export fish processing and international distribution is the most developed. The technologies which it employs are near the global technology frontier and the industrial organizational structures through which it operates have adapted flexibly to several challenges associated with maintaining their global competitiveness. The rivate sector companies who have invested in the sector and who have developed sophisticated supply chain capabilities are locally based for the most part. Most of them operate not only in Tanzania but in Uganda and, in some cases, in Kenya as well. Most Nile perch continues to be purchased by the major fish processors, as are most of the prawns. Indeed, due to sharp increases in demand for both fishes in overseas export markets, prices have risen slightly and demand continues to exceed supply. Export markets have a strong preference for oily white fish like Nile Perch and as its market acceptance continues to develop into new export markets its price continues to climb. The supply chains through which Nile Perch and Shrimp are purchased are more intensely managed than the chains through which fish for domestic consumption move. Export processors purchase their inputs through specialized agents. These agents typically provide fuel, equipment and credit to fishermen who are affiliated with them and who importantly are trained and whose fishing processes are certified with respect both to food quality and ecological compliance The export chains are integrated by the large export processing companies. Inventory moves transparently through these chains, prices are set for extended periods and are based on published guidelines which the industrial processors set collectively and which normally entail a significant premium above domestic market prices. Quality control standards are rigorously enforced and as a result a substantial volume of fish rejected by export processors finds its way back into the domestic market. Processors provide trade credits and preferential leasing terms to quality suppliers ( the agents referred to above) who have established themselves with the large processors. Purchases for the entire export supply chain are limited to the supply of high quality fish. High quality fish are required to filleting. Once processed fish fillets are boxed and fast frozen. A relatively large inventory ( more than 90 days) is maintained in Tanzania and export sales are completed from this stock. The competitiveness of the country's fisheries sector is based primarily on the large renewable supplies of white table fish from Lake Victoria which is extremely popular in OECD countries and which provides a near substitute for Cod. Near universal popularity of cod and its progressively diminishing global supply set the market background conditions against which the local fishing industry has thrived. In recent years, the local industry has been challenged by several factors which emerged its business environment. All of these required " supply chain" organizational responses which over the medium term have had the effect of further strengthening the industry's competitiveness. . Thus, for example, Tanzanian producers scrambled during the 1990's to comply with new and progressively more exacting food safety requirements. Today, the industry enjoys full compliance not only with EU fish safety standards but with those of Japan and the US, as well. Complying with all three of these standard regimes provides Tanzanian producers with de facto access to all international markets. Shortly, after passing the food safety compliance test, Tanzanian producers were confronted with the need comply with the new regional ecological regulations which affect fish replacement rates and resource renewal on the Great Lakes. Tanzanian producers have make more compliance progress than have producers based in either Tanzania or Kenya and are leading the way to assure that take out standards and equipment requirements are fully enforced within the regional Lake System. The industry's response to these challenges has strengthened mezzo level organizations and provided the sector with new sources of competitive advantage. The fish export industry is made up of less than half a dozen large processor/exporters and another dozen or so small scale exporters. Twenty fish processing plants are licensed to operate in Tanzania. The industry possesses sufficient resources and competencies to meet the "standards" challenge. All of the licensed producers have attained HCAAP certification. However, only the top three or four exporters have managed to continue to reposition themselves into profitable niche markets. These are the same firms who are developing plans for investment in aqua culture and to consider investment in down stream value added processing. Indeed, the sector has developed to a point where its competencies now far transcend simple, low cost production and include more difficult to match capabilities in complying with international food quality standards and in making "fail safe" deliveries of both fresh and frozen food products around the world. The industry's special logistics competencies together with the supply chain systems that support them have helped to position it for future growth and to guide its development onto the next, higher level of value added production. Moreover, now that concerns over the renewable nature of the resource base on which the industry rests appear to be resolved at least over the short and medium term a confident fish processing industry is considering several strategic investments--in value added manufacturing of ready to serve fish products and table ready fish meals. If made these investments would enhance the industry's competitiveness. Tanzania is uniquely well endowed with fishery resources. It has a coastline of 800km and a narrow continental shelf (mostly 7 to 20km). Its inland fishery resources include three of the African Great Lakes which total approximately 54,000 sq. km of fishing domain. Most (e.g. 51% ) of the world's second largest lake lies within Tanzania's national boundaries. Fish taken from Lake Victoria contribute fully 50% to the country's total fish production and the uniquely valuable species of soft white meat found in the Lake provided the industry with its original source of competitive advantage. In 2003 Tanzania adopted a new regulatory framework for the fishing sector. The regulatory framework adopted in 2003 for the fishing sector is production oriented. It focuses primarily on assuring the industry's renewable resource based remains renewable by constraining the growth in primary product take out to the limits of the sustainable the ecology of the Great Lakes. The regulatory framework is designed to prevent over-fishing through constraints imposed on licensees, as well as through limitations imposed on fish extraction methods. As envisioned in the legislation another important resource management activity is the seeding of lakes, streams and reservoirs with new fish. What is most notable, however, about the new regulatory framework are the fisheries development issues which it fails to address. It fails to address the vertical structure of the industry, internal linkages among and between specific participants in the fishery to market chain, the development of cornerstone institutions within the sector, the efficiency with which both domestic and international supply chains operate and the collective bargaining ( or other ) basis for determining equitable agreements between organized fishermen and organized processors. In other words, the existing regulatory framework is mute with respect to key issues of supply chain development. The new regulatory framework provides no assurance that the country's abundant fishery resources will be actively developed in ways which improve the livelihoods of fishermen, small scale traders, artisan processors and other impoverished stakeholders. To the goal of alleviating poverty,. the best and most likely to succeed policies are those which facilitate the building up of institutions and then once strengthened, the application of these institutions to the develop of efficient local markets. However, an appropriate legal platform does not appear to exist from which to launch such a reform agenda. Supply chains for fish products are labor intensive. However, to operate efficiently they also require the intensive management coordination of product flows, of storage, transport and processing. As the schematic below demonstrates, supply chains for export fish and for domestic fish are integrated by the processors who schedule, program and coordinate each of the key activities within the chain including, most importantly, procurement of raw fish, production scheduling and outbound logistics. The set of value adding processes outlined in dark green fall within the direct corporate control of the processors. The value adding processes defined in light green correspond to those which are managed indirectly by the processors though their affiliated agents. Within the existing system fresh fish supply chains for export fish operate on a demand pull basis. Processing of fresh fish is scheduled to satisfy purchase orders and only once orders are received is product shipped directly to third party customers. When orders are shipped to affiliated retailers in overseas countries ( at least one of the major processors has developed its own set of fish retailing shops on a joint venture basis) shipments from Tanzania are made in order to maintain minimum inventory levels in the foreign countries where retailing affiliates exist. For fresh fish the "reorder point"....the point where actual and expected demand are reconciled....exists at the forward end of the chain where orders are taken. The export supply chain for fast frozen fish is quite different. It is designed around a supply push principle. Demand and supply which peak during different seasons are disconnected in the frozen product chain. Buffer inventories of flash frozen fish, based in Tanzania, delink demand and supply. Most frozen fish is purchased when raw fish stocks are abundant; during peak fishing seasons fillets and parts are produced for inventory and are stored in large cold storage warehouses located in Mwanza and Dar es Salaam. Frozen fish is shipped via ocean freight, in full container quantities on a minimum outbound logistics cost basis. Standing orders exist with most overseas customers and shipment, storage and transparent replenishment are all conducted on a lowest economic lot sizes basis in order to maintain minimum inventory levels in the entire export/ import system. Domestic supply chains which support fish distribution for consumption within Tanzania lack effective integration. Each process step in the chain is completed before the next process is begun. No active management integration takes place within the chain primarily because no single agent or set of agents has end to end control. Buffer inventories build up between individual process steps so that individual steps can be effectively decoupled. Moreover, multiple buy sell transactions take place in order to provide working capital to chain participants and to broaden geographic access within the country. However, the resulting inventory accumulations and the consequent high level of post catch losses which occur within domestic channels cause the working capital required to push a unit of product through the entire chain to be very high. Risks associate with lost and spoiled product are equally high. Working capital is provided to domestic chains through the multiple set of buyer/ resellers who participate in any end to end movement. However, the resulting multiple transaction chains also drive up transaction costs, increases market margins and makes traditional domestic fish chains inherently inefficient. The result is a loss of value both for artisan fishermen and for consumers of fish products in Tanzania. The most fundamental constraints inhibiting the development of more efficient distribution channels for the domestic sale of fish products are three fold: i) missing institutions which can do the work of channel integration; ii) missing skills and competencies and iii) financial instruments or modes of investment through which capital provided by third parties could enter domestic chains. In Tanzania extremely efficient chains applying near best global practice operate parallel to extremely inefficient chains which are inefficiently structured and inefficiently operated. Chains for domestic fish production/ distribution constitute a poverty trap. The objective of policy and public/ private investment interventions in this sector should be to facilitate the transfer of knowledge and of capabilities for the export chain to the domestic one. Conceptual Map for the Fish Supply Chain in Tanzania Export Chain Plan Store Sell Provide Procure Process Schedule Fish and Wholesalers Customers Inputs to Catch Fish Fish Fish Control Ship Retailers Fishermen Price Domestic Chain Upgrade Transfer technology infrastructure and management Upgrade infrastructure methods Develop cold chain Input Buyer Buyer Buyer Wholesalers Providers Fishermen Dock Side Reseller Reseller Reseller Municipal Retailers Customers Market Market Develop new Extend Develop systems, Extend Business models. chain methods and supporting chain Develop collaterals forwards ITC to balance demand s backwards And financable securities and supply The primary recommendations which emerged from the study with regard to the fish supply chain include the following: As we noted above, fish supply chains in Tanzania have developed in two quite distinct tracks..... with one track serving local and regional markets and one international markets. Companies which are most active in the international market have developed efficient and competitive chains and are prepared to develop further sources of competitive advantage in aqua culture and in value added food processing. Small scale fishermen, processors, traders and ancillary service providers who cluster around the second set of chains, however, are not advancing at the same pace as the more sophisticated segment of the industry. The prospect that efficient chains will spontaneously develop along this second track appear to be remote. In this second tier of competitors, a role exits for donors in partnership with the government to build up institutional capacity. Thus, stronger linkages are required within domestic channels, new business models need to be tested and refined and new and more effective market institutions need to be created. Investment in transport, storage and cold chain handling is required, as well, but these investments will naturally evolve once appropriate commercial institutions have developed and certainly not before. A potential leverage point for the kind of development which is required are the two major wholesale fish markets-- Banda Beach Market in Dar es Salaam and Kirumba Market in Mwanza -- which have emerged within the country. Modifying the underlying business models and the enterprise governance structures on which these two institutions rest provides a particularly interesting entry point for institutional reform. Currently both market institutions are severely constrained by the local purview of the government entities who sponsor them as well as by the strictures and limitation inherent in a tenant/ landlord commercial relationship. Alternative business models need to be tested for both institutions which are not traditional. Thus, for example, both markets might be redeveloped as concessions--concession which would require private sector concessionaires to invest in stronger backward linkages to fishermen, provide affiliated fishermen with working capital, with leased equipment or with productivity enhancing fishing gear. The concessionaire might also invest in improved cold chains or in improved scheduling of fish products onto the market, etc. Better yet, the two wholesale markets might be re-developed as concessions which included elements of a service network franchise. Thus, the terms of a long term concession agreement would require the concessionaire to invest in forward linkages in satellite wholesale markets which were linked to mother market through information systems, frequently scheduled logistics management services, pay-at-either end transfers of product and end to end cold chain coverage. Another possible leverage point are the product sourcing requirements of the new and aggressively growing supermarket chains who have recently arrived in Tanzania or, indeed, the seafood procurement requirements of fast food outlets. Both sets of outlets require significant supplies of fresh fish product. Long term, commercial linkages in the form of " quality food providers" or " core fish vendors" could be developed, initially on a demonstration project basis, around the requirements of these two niche markets. Once initial proof of business concept had been achieved, a larger procurement contract could be specified and competed. An important aspect of this kind of retail-up development is the adoption of food safety standards. Initially these could take the form of procurement specifications but subsequently they might assume the more permanent form of supermarket food safety protocols ( e.g. EUREPGAP Light !) These protocols would be similar to but less compulsory and hence less costly to enforce than the ones which apply to fish products sold into the EU, Japan and the USA. The local protocol analogs of these other food safety standard regimes would include standards for process conformance with " best practices" , real time product quality testing within the chain, occasional audits, third party certification procedures, etc. In this way, a formal fish product chain could be developed which included internal incentives for superior quality and for healthful food delivery. An important aspect of this kind of retail-up development is the adoption of food safety standards. Initially these could take the form of procurement specifications but subsequently they might assume the more permanent form of supermarket food safety protocols ( e.g. EUREPGAP Light !) These protocols would be similar to but less compulsory and hence less costly to enforce than the ones which apply to fish products sold into the EU, Japan and the USA. The local protocol analogs of these other food safety standard regimes would include standards for process conformance with " best practices" , real time product quality testing within the chain, occasional audits, third party certification procedures, etc. In this way, a formal fish product chain could be developed which included internal incentives for superior quality and for healthful food delivery. Still another potential leverage point within the legacy fish supply chain are the landing/ market/ processing clusters and commercial focal points around fishing communities which we discussed above. . Again opportunities exist to fundamentally reinvent the institutional foundations on which these commercial focal points rest. New business models could be demonstrated and tested which entail linkages from these fishermen controlled businesses to market centers, specialized market niches and/or consumer focused markets institutions. Ways could be developed to create bankable collateral within the new experimental, fish producing institutions, to cross guarantee bank repayment or to create commercial structures which might be attractive to a joint venture investment partner. In these ways and others, capital could be mobilized for investment in more productive fish production, processing and distribution. In addition, external markets for specialized services--e.g. equipment leasing, logistics management, transport, market information, storage and banking---need to be developed around the periphery of the domestic fishing sector. This can be done proactively by defining service specifications and then outsourcing the desired services to qualified service providers, by joint venturing the development of new services with qualified service providers or simply by demonstrating the commercial viability of new service launches through feasibility studies, business plans or procurement documents. Importantly, as well, new market network services need to be envisioned, specified both commercially and technically and rolled out on a nation wide basis. These network services would provide the superstructure around which a national market could be developed for fish products. Among other services the national network might include the following: i) third party inventory management and temperature controlled storage, ii) cash management, iii) price discovery and data dissemination , iv) insurance and v) transport delivery which was compatible with commodity basis point pricing. Sugar Supply Chain The country's sugar industry is currently going through a significant transformation in its structure, product mix and development strategy. This fundamental transformation has been triggered by the privatization and subsequent reorganization of the sugar industry. Competitive sugar industries require strong and well integrated supply chains with well organized backward linkages from refineries as well as forward linkages to retail markets. When the right set of factors come together so that productive supply chains operate and, just as importantly, so that these chains compete effectively with one another for expanded markets, as they are beginning to do in Tanzania, the wealth creation effects can be quite significant for poor farmers. The plant material, the basic planting and harvesting methods, the rain fall and the prevailing climate are such that at least three growing zones in Tanzania enjoy a comparative cost advantage in the production of cane sugar. Transforming this comparative cost advantage into a competitive market advantage and expanding the markets which sugar producers can serve from a Tanzania base is the strategic agenda being pressed forward by the new privately owned and operated sugar industry in Tanzania. New private management within the industry is realizing significant success in the following areas: I) the further lowering of production costs ( through the use of high yield cultivars and better coordination of production, harvesting and inbound logistics), ii) increased plant production capacity and the development of downstream industrial sugar capabilities; iii) improved product packaging and the development of quality differentiated " brands" and iv) more effective distribution channel management, through volume buying incentives, joint retail promotions with third party wholesalers, negotiation of exclusive distribution rights, etc. Through different combinations of strategic initiatives which link them both backwards to their out grower suppliers and forward to their wholesalers and retailers Tanzania's recently privatized sugar producers are rapidly becoming effective supply chain managers and in the process they are substantially improving their competitiveness. Moreover, as they have developed different producers have begun to adopt slightly different marketing and product positioning strategies. Importantly, the industry competes aggressively on a wholesale and retail price basis into the major Dar es Salaam market but through its recently developed trade association cooperates in developing capabilities which are enhance the competitiveness of all of the industry participants. Although the industry has not yet begun to export any substantial volume of sugar beyond the country's borders, producers are quickly saturating local markets and beginning to position themselves for regional growth beyond Tanzania's borders into less productive regional sugar markets. Trade policy, together with the basis for collective bargaining agreements with affiliated out growers, are the two primary issues of mutual concern to industry participants. These are also the issues which the newly formed industry trade association is most concerned to pursue. Three private companies currently dominate sugar production in Tanzania. A fourth company is preparing to enter the market subsequent to the financial closure of its privatization transaction with the government. In addition, several small scale producers continue to operate within the industry as do a limited number of cooperatives. As the four larger private companies continue to enhance their production capacities, they will begin to outgrow local markets which are, in any case, quite small. What happens next, once domestic supply in Tanzania equals domestic demand will have a major impact on the future development of the industry. Kenya is the largest regional market in which Tanzanian producers may be able to ship product competitively. Hence, once domestic markets become saturated and value added processes are added to primary refining processes, regional trade policy will become a key growth determining variable for the domestic industry. One aspect of this baseline policy involves a compromise between sugar manufacturers and industrial sugar users under which users of " industrial sugar" are granted preferential tariffs while higher tariffs apply to consumer or "raw sugar." This compromise will soon be tested as the new private owners of the nation's large scale sugar refineries increase their production to levels where it begins to meet or exceed domestic demand for raw sugar. This point of converge appears to be only two or thee years ahead. Sugar refining is both labor and capital intensive. Importantly as well, it is highly management and transport intensive. sugar production is also transport intensive. Hence, the location of refining facilities both vis a vis both primary inputs and end markets is critical for sustaining competitiveness. In developing countries refiners are most efficient from a supply chain design perspective when they locate close to inputs, e.g. to farm level production of cane. .Because the cost of transporting the primary input is extremely high compared to the cost of transporting the output particularly when transport infrastructure is poor and water transport is unavailable most refineries are located close to their input supplies. Still, product distribution costs ( outbound logistics costs) are also a significant factor in determining industry competitiveness especially in a large country like Tanzania. However, as it turns out they are not as significant as the costs associated with inbound logistics. The industrial geography of national sugar economies is a critical factor in their sustainable growth. Clearly all of these issues apply to Tanzania where the primary refiners are separated somewhat from one another and from the major consumer market in Dar. Importantly, all major producers are well located near productive cane growing areas. The labor intensity of sugar production makes sugar refining a particularly effective tool for economic development. Once they are built sugar refining facilities need to be supported over their entire economic life by labor intensive sugar cane production. Hence, efficient sugar processing can realize a significant welfare dividend for small scale farmers. The development of the sugar industry is important from the point of view of poverty alleviation. Sugar processing has the unique capacity to improve the well being of large numbers of poor farmers. The industry is just beginning to have a positive development impact on the 10,000 plus out growers who are affiliated closely with the three large producers and on the more than additional 80,000 employees who currently work within the industry. Currently, most sugar cane is grown on estates which the large sugar processing factories own and which are part of the state owned legacy which they inherited. However, new risk sharing relationships are being tested and refined between processor/ distributors and contractually affiliated cane producers. Sugar production is highly seasonal and the working capital requirements associated with producing sufficient supply over a six month period to satisfy twelve months of demand recommend alternative risk sharing arrangements which divide fixed asset investment and working capital investment responsibilities between farmers and processors in new and innovative ways. Thus, as it expands the industry is fundamentally restructuring its business model with contract, small holder farmers increasingly being substituted for direct employees of private sugar processing companies and with small holding being substituted for processor owned or leased cane plantations. However, the industry is expanding its production capacity so rapidly at the present time that it is both adding jobs and, at the same time, affiliating with new out growers at the rates respectively of 3% and 5% per year. Importantly, new forms of supply chain affiliation which are beginning to emerge. The conceptual map of the sugar supply chain represented below identifies value adding elements that are controlled directly by the large sugar producers within the green portion of the chain schematic. As the diagram suggests opportunities exist to further strengthen both backward and forward linkages from those which are currently being in sourced. e.g. the activities located within the green area. The diagram further identifies all of the weak links and functions requiring further reform within the rapidly evolving chain. These are laid out in a set of boxes. Significant, as well, is what the diagram does not identify. The diagram does not identify, for example, the provision of third party services, e.g. transport, warehousing, financing, cash management, insurance, as a major issue. Markets for third party services are developing rapidly on the periphery of the Tanzania's supply chains. Thus, specialized third party providers of inbound logistics services and of road way services have rapidly emerged. In addition, the basic industrial organization for efficient production to retail sale chains appears to be already in place. Although experimentation and testing of new business models continues, most of this is by way of refinement and does not go to the way in which essential functions are organized within the chain. The hatched lines in the schematic below represent systems and controls for receiving feedback from the retail and wholesale end of the chain and making that information operationally relevant for production scheduling in the short term and for capacity planning over the longer term. These systems, incentives and mutual information exchanges appear to be developing quickly in Tanzania. As we noted above most of the sugar refiners have adopted business models which entail the engagement under contract of out growers. These agreements are regulated under the Sugar Act and are being codified in the form of collective bargaining agreements subject to the regulatory enforcement and judicatory interpretation of the Sugar Board which includes as members representatives of all of the key stake holder sectors involved in the sugar supply chain. Indeed, the regulatory framework which was developed for a post privatized sugar industry lays out the regulatory and policy mandates, contract interpretation and enforcement mechanisms required to facilitate rapid industrial restructuring and supply chain reengineering. Table 1 below represents the basic regulatory functions and enforcement mechanisms that exist under the Sugar Act and compares these with the regulatory authorities which operate in the two other sectors studied. The Regulatory Framework for sugar is still relatively new and methods and protocols are still being worked out as the law is being applied, tested and refined in actually practice. However, it is not soon to observe among the three sectors studied, the regulatory mechanisms developed to support the sugar industry are clearly the most forward looking, the most supply chain supportive and the most open to private sector participation. Conceptual Map of the Tanzania Sugar Supply Chain Build/rehab local farm-to-factory road and bridge Achieve break through infrastructure improvements in core production functions Facilitate intra chain Extend the reporting, information Develop processes, supply exchange and methods and chain Build up mutually supportive systems to improve backwards Suppliers economic incentives the balancing of supply and demand Sell to Indepepdent Wholesalers Input Primary Inbound Refining Plan Take Retail Customers Suppliers to Production Logistics: And Schedule Orders Sell to Customers Customer Outgrowers From Both Out Growers Harvest Storing Control And Ship Affiliated And And Refinery Price Wholesalers Own Growers Coordination Promote Sell to Industrial Customers Extend the supply Implement Build up chain Improve cultivors, Coherent Distribution forwards harmonize input quality Supply Chain Channels standards, establish fair Development Liberalize trade procurement and payment Strategies regime and hence s improve export and ,methods, extend harvest Develop Alternative season, coordinate harvest, industrial market collection and competitiveness Distribution refining processes Channels including Importantly export and industrial channels The primary recommendations which emerged from the study with regard to the sugar supply chain include the following: The primary public policy issue with which the Sugar Board is likely to become involved in the years immediately ahead involves the country's trade policy effecting sugar. Since privatization, production capacity continues to increase apace within the industry and competitiveness continues to improve. As the industry grows it absorbs more inputs and generates ever increasing benefits for employees, affiliated out growers and third party service providers. However, unless the government takes effective steps to open new market segments for the industry this rapid growth may be prematurely constrained. The challenge facing the government is how best to assist the industry in opening new markets . One set of potential new markets are domestic markets for industrial sugar. Another are regional markets for sugar imports, primarily into Uganda and Kenya. In the Uganda and Tanzanian markets Tanzanian producers enjoy a competitive advantage vis a vis lower production cost competitors to the South by virtue of their closer proximity and consequent lower distribution cost. However, liberalizing regional market access is a "double edged" sword since the sugar economies of Malawi, South Africa, Swaziland, etc. are still more competitive than Tanzania's at the moment. The "double edge" granting, as well as seeking more open market access. Interestingly, the two primary market opening options entail quite different trade policies. Opening internal industrial markets entails rolling back the liberal tariff preferences that the food processing industry enjoys. Opening regional markets, on the other hand, entails quite an opposite approach. Livelihood gains that out growers, employees and ancillary service providers have been able to realize in recent years might be threatened in a head to head confrontation with lower cost producers. The issues here are complex and the ultimate outcomes are path dependent. The need for an in depth analysis of both the welfare and development to guide the political debate is clear. As noted above one area where additional reform is clearly called for involves the development of the local infrastructure which is required to support sugar production. The rural road and drainage systems which are currently in place around the country's major sugar refineries are rapidly deteriorating. In addition, new infrastructure needs to be build into the ancillary cane growing lands which are being developed anew as refinery production increases. The private companies which have emerged from the privatization process are unwilling to assume any additional responsibility for infrastructure development. The terms and conditions of their privatization agreements are clear on the point what responsibilities they have assumed. The failure appears to be with the multi-level, federal system of fiscal control and accountability for public investment. Cesses are imposed to provide for infrastructure renewal. However, value for money has not been forthcoming to date. The entire area of agricultural research needs to be strengthened. High yield clones, such as those bred in Hawaii, need to be tested for adaptability to Tanzanian climate and soil as a way of improving sugarcane yield. A need for the sugarcane industry in the country to develop research units in some form of public/private partnerships to refine the repertoire of agronomic, plant science and crop management " best practices" available to contract farmers.. A complementary need exists for a strengthened extension service which can deliver technical support regarding best practice to the most vulnerable farmers. Again the issues this whole arena of technology transfer and its practical application are complex and the required policy reforms nuanced. The study team suggests that an in dept study be undertaken which completes a benchmark assessment of prevailing agricultural practices, determines any significant " gap" and recommends ways to close this gap which are cost effective and which entail a form of public/private partnership. The final area in which fundamental reform is called for involves the Balkanized system of taxation which prevails in Tanzania. In a previous section of this chapter we discusses how the existing system of cesses, fees, tariffs, levies and taxes creates incentives among supply chain participants which are perverse. The table below suggests some potential responses and remedies to this problem. Again, this is a complex set of issues which call for additional, in depth study. Maize Supply Chain Tanzania has several natural advantages in the maize trade. It is well positioned in the center of a large regional market for export maize. It has abundant rain fed land which should allow its maize producers to be more productive then they are currently. However, the commercial environment in which farm producers, merchandisers and millers operate does not provide adequate incentives for the productive development of farm to market linkages and thus frustrates the full potential development of the maize chain. In comparison with the other two supply chains which the team studied, i.e. fish and sugar, maize appears to be the least efficient and the least developed both from an institutional and poverty impact perspective. Moreover, maize markets in Tanzania are less developed than markets for sugar markets and no better developed than national markets for fish. Moreover, Tanzanian maize markets are significantly less developed than the parallel maize markets that operate in neighboring economies. Maize is the principal staple food in Tanzania. It dominates the diets of both the rural and urban poor.1 Unlike other consumer markets in East Africa ( e.g. Uganda where maize is readily substitutable for other cereals and tubers), Tanzanian consumers have a strong preference for white maize. They refuse to substitute other staples such as rice, cassava, sorghum, millet and beans except during periods of severe food shortage. Hence, the price elasticity of maize in Tanzania's local markets is relatively low Average, marketed and traded maize in Tanzania averaged 650 thousand MT during the period 1999 to 2003. The Tanzanian market is among the largest in East Africa. Most of Tanzania's maize is produced by small holder farmers who lack on-farm storage capacity and who have no access to credit and only limited access to price and other market information. Significantly, most of the farm level enterprises, which produce maize in Tanzania are very small in scale. An average farm has less than one acre of maize under cultivation. Each farm produces on average only 1.6 tons per year. Importantly, these small scale, family farms lack both capital and organizational structures which would allow them to diversify production and effectively manage their market risks and/or sell those risks to third parties. Moreover, unlike in Kenya and Uganda relatively little commercial re-organization has taken place among Tanzania small holder farmers which would begin to reverse their marketing disability. Consequently, most farm organizations are generally ill equipped to deal with issues of risk management, production planning, post harvest loss management, marketing and pricing, working capital financing, transport management and storage of maize. Most primary sales of maize take place on a cash and carry basis and most farmers sell either at the farm gate or, alternatively, at the nearest village market center. In both settings they are " price takers." Weakness on the supply side of maize markets is only exacerbated by concentration on the demand side. Tanzanian maize markets are thin and illiquid, except for the liquidity that is provided by a limited number of large scale traders. Commercial bank financing is not available to either farmers, input providers or small scale traders within Tanzania and only the capital of large scale trading principals is available in local markets. Maize production is concentrated within Tanzania in two primary growing areas: the southern and northern highlands. Significantly, both of these growing regions are a significant distance from the nation's major urban markets in Dar and Mwanza. Moreover, of the country's twenty regions, only six highland regions regularly produce surpluses. Hence a great deal of cross trading and transport is involved in matching local demand with supply. Together the two primary urban markets account for 55% of total consumption. Significantly these two urban markets are the principle focal points for price formation within Tanzania. However, unlike the fish market maize prices are not exclusively set in urban market clusters in either Dar or Mwanza. In this large country in which net producer and net consumer populations are broadly dispersed both trading capabilities ( e.g. access to capital and to market information) and transportation are essential inputs into the maize marketing function. Because, the largest portion of maize production takes place at a significant distance from town and village markets in deficit areas. Farmers face the decision of either selling at their farm gate ( or local village) to third party traders or transporting their grain a significant distance to local market centers where sale prices are net of transport costs. Small scale farmers also lack the management competencies with which to operate in any market making role other than that of "price taker." Hence, even in a liberal policy environment small scale producers are disadvantaged vis a vis other more sophisticated chain participants who better comprehend regional market dynamics, the likely effects of local demand and supply balances on price and the price consequences of government and donor interventions. In this large market small sources of competitive advantage translate into a significant trading edge. In the long and inefficient chains that operate in Tanzania, large scale traders enjoy not only the advantage of superior access to information but, importantly, as well the advantage of owning or controlling a significant portion of the country's cereal storage capacity. Large scale traders control significantly more storage capacity than do small scale traders. The portion of the nation's maize storage capacity that is not controlled by traders is controlled primarily by the Strategic Grain Reserve Agency. The GRA, however, does not offer a competitive alternative to trader controlled storage, as a third party storage provider, to small scale farmers who are pressed for cash. One of the reasons that the market for maize is thinly capitalized in Tanzania is that grains cannot be effectively collateralized against bank lending. Another reason, as noted above, is that relatively little third party storage capacity exists in the country. Moreover, the function of third party asset management which involves securitizing maize and managing inventory assets has not yet developed within the country. Still a third reason is the with the exception of the two primary grain traders based in Tanzania investments in grain inventories are simply not financable. Only, two large traders have developed balance sheets which are strong enough to satisfy the requirements of commercial lenders. Tanzanian maize markets have important regional, as well as national features. Since Tanzania is a member of the ECA its maize traders enjoy liberal access to markets both in Uganda and Kenya. Importantly, traders in those two countries enjoy equally liberal access into the Tanzanian market. Trading rights are reciprocated under the ECA. This circumstance has clearly helped in increasing competition and injecting additional trading capital into the local maize market. However, for reasons explained below traders based outside Tanzania are reluctant to buy Tanzanian maize and resell it into the Tanzanian market. Importantly, other major maize deficient markets in the region, in addition to Kenya where maize is in chronic short supply, are also open to Tanzania traders. These include most importantly Zambia, Malawi and the DRC. In recent years, tariffs barriers among ECA trading partners have been reduced to sufficiently low levels so that an efficient regional market is beginning to develop among and between Kenya, Uganda and Tanzania. The dynamics of this EAC three way trade significantly effects both supply and demand within Tanzania, itself. Similarly, prevailing tariffs for grain imports into Zambia, the DRC and Malawi are quite low for Tanzania maize traders. In recent years Tanzania's cross border trade with its Southern neighbors has continued to grow. Much of this trade, however, is informal. Prices received by Tanzanian traders are increasingly subject to overall demand/ supply balances within the parameters of regional markets. Additional reasons for the rapid emergence of a regional market include, the following: i) Tanzania consumes marginally more maize than it produces and what is produced in Tanzania cannot be stored for more than a month or so because of capacity limitations; ii) Tanzania producers ( unlike Ugandan producers) enjoy only a single harvest in June/ July and are net exporters only during this harvest season, and iii) demand and supply balances vary significantly through the rest of the year among the several countries within the region. Depending both on the timing of their harvests and their capacities to store maize, they are alternatively importers or exporters. These background conditions set the stage for increased regional market integration. Strong evidence exists that an integrated regional cereal market is taking form and that liquidity, market information and risk management constraints within that regional market are in the process of being removed. All of these appear to be positive developments for both farm producers and consumers in Tanzania. Still the agenda associated with regional market development is far from complete and, significantly, Tanzania remains well behind Uganda and Kenya in pursuing this agenda. One consequence of regional market integration is that prices are more closely linked, for example, among regional trading centers in Dar and Nairobi or Mwanza and Kampala than they are between Tanzania's own primary maize consumption centers or, indeed, among maize surplus or shortage areas within Tanzania itself. Even the availability of maize in Tanzania's peripheral rural economy--particularly late in the country's annual production/storage/ consumption cycle---is affected by production and inventory levels in neighboring countries and hence by the import requirements of these other countries. During the immediate post harvest season of 2004, for example, much of Tanzania's maize stocks were bid away to Kenya. When Kenya's subsequent production fell well below expectations and its own later season import reliance proved to be much greater than was expected earlier, prices jumped in Tanzania. The larger point is this: Maize markets in Tanzania are increasingly linked to parallel markets in Uganda and Kenya and even in Malawi and Zambia and grain brokers based in these countries operate alternatively as buyers and sellers within Tanzania's own markets. Ironically, larger problems persist in Tanzania's own internal markets. For one thing local market are thinly capitalized, as we noted above, and, in addition, as a number of econometric studies have demonstrated they not particularly efficient. The commercial implications of these market failures are that opportunities persist to realize above average profits through both spatial and temporal arbitrage in Tanzania for those few traders who are able to act on opportunities in the local market. Indeed, as we might expect the majority of maize traded in Tanzania is bought and re-sold by relatively few large buyer/resellers in Tanzania. Major grain traders who are best positioned to do so have little incentive to invest in more efficient supply chains, preferring instead to buy in spot markets, hold maize until seasonal prices rise and then resell into chronically deficit grain areas. . This circumstance persists for several reasons: i) because, in most years, aggregate demand is effectively articulated through relatively few buyers who formulate the prices which other, smaller scale market participants accept; ii) because, again in most years, local demand exceeds local supply and the price of imported maize sets the marginal price in the local market late in the season at least in Dar and Mwanza markets. Again it is the larger traders who participate in this import market and who consequently best understand what factors are likely to influence the marginal price of maize in the local market; iii) because few areas in Tanzania produce a net surplus of grain and these areas are separated from maize deficit areas by long distances and poor infrastructure; iv) because storage capacity is limited compared with aggregate demand, so buying and holding strategies have limited effect. Moreover, the two largest grain traders both other nation wide buying and storage networks and hence are best positioned to buy, hold and sell; v) because regional markets themselves remain thinly capitalized with no trade credits being made available from net exporting countries like Uganda and, finally; vi) because, little opportunity exists to sell risk associate with holding maize inventories outside of local markets. Indeed, no effective risk hedging instruments exist with which to facilitate such a transfer. The government has developed no effective risk management instruments or effective modes for assuring food security. The large market is highly subject to price perturbations and on the sell side farmers face rapid price moves between the marginal cost of production and the import parity. For all of these reasons shortages occur with some regularity within Tanzania. These shortages in turn trigger direct government and donor interventions in local markets.. These interventions, in turn, only inject additional uncertainty into local market operations. These additional uncertainties concern the volume, timing and price effects of additional food aid. As a results a perverse "virtual cycle" is perpetuated of under investment by the private sector in risk resilient supply chains, chronic cereal shortages, chronic public and donor intervention and further market destabilization. The maize short months in Tanzania are March, April and parts of May. This is also the period when prices are typically the highest for cereals. Intervention which are not announced publicly in advance of actual buy commitments or, indeed, of actual delivery or interventions which are announced but not implemented limit the ability of local market makers to efficiently discount new information into local market prices. Such interventions not only disrupt the local market but, even worse, they allow trading parties with superior access to information to use that information in their trading operations and hence create invisible barriers to entry for other traders. As a result of ad hoc interventions during crisis periods and tight demand and supply balances within Tanzania itself, prices within the local market bounce between import parity, on the market top side, and the marginal cost of local production/ storage, on the market bottom side. This circumstance, in turn increases risks for private investment in the maize chain and causes the cost of capital for potential investors in the chain to rise. It also sends precisely the wrong signal to farmers who experience a boom and bust cycle when they respond to increased harvest values by increasing their plantings in the next season, only to witness a decline in revenues to which they typically respond with lower planting levels. The end result, not surprisingly, is a general reluctance of invest in long term productive assets within the sector and the regression of the industry into a trading mode as contrasted with a pro-production/processing or supply chain development mode. Regulatory policy within the maize supply chain is effected primarily through three sets of interventions : i) food security and the supply of emergency food relief, ii) trade policy , and iii) food quality control. The policy backdrop against which these three sets of market intervention levers are exercised lacks strategic clarity and fails to support the development of efficient supply chains. Policy preferences revealed through actions and initiatives demonstrate clearly that the government is highly risk adverse, fixated on short term consequences, especially those associated with food shortages, lags other trading partners regionally with regard to trade liberalization. . The arms-length transaction based maize chain which has developed in Tanzania since 1991 had developed in a policy context that might best be described as "laissez faire, except during times of maize shortage" Except during declared periods of food security the government has allowed maize transactions to take place without much intervention. Unfortunately, Tanzania has been in a continuous state of food shortage since the implementation of the new liberal policies. According, export permits have been difficult to secure, as have import permits during harvest seasons. During harvest season the Emergency Grain Reserve has been a significant buyer and during grain shortage periods a significant seller. Significantly, the way in which both the Emergency Grain Reserve and the WFP have elected to develop their procurement programs has advantaged large volume traders who can qualify to participate in these programs by virtue of their large standing inventories, storage and transport capabilities and/or access to financing. We commented above on some of the adverse market development effects that result from interventions during periods of food shortage which tend to tilt an otherwise level playing field to the benefit of large traders.. Adverse fiscal effects have also resulted from the way that the food security program is managed. In spite of the strategic market timing strategy which the Emergency Grain Reserves apparently employs: e.g. buying when prices are seasonally low and selling when they are high, the Emergency Food Reserve requires a regular annual budget allocation. Strategic timing interventions in the maize market which should result in a net profit from annual maize sales. Instead it appears to result in regular annual trading losses to the government.. Significantly, the Emergency Food Reserve represents the largest single line item in the budget of the Ministry of Agriculture. Thus, financial support of the current maize strategy squeezes out a number of other programs which might reasonably be expected to have a much greater impact on alleviating poverty among farmers. The most critical policy variable, however, which remains unregulated is industy structure. The " initial conditions" which prevailed when the maize trade was initially liberalized were such that only two trading companies emerged to fill the gap left behind by government's retreat from direct intervention in maize markets. In the interim period the two trading companies have managed to develop a nation wide storage and trading network, to develop effective information management, hedging and trading systems, and to develop effective arbitrage systems for selling risk outside of their own supply chains. Both companies have developed effective strategic responses to maximize the profits they realize from trading and thus have less interest in investing in supply chain systems which entail reassigning risks from those least able to manage them to those best able to manage them and to dampening overall systemic risk by improving the way in which demand and supply are balanced over time. . The structural response to liberalization in Tanzania is quite different from the one which exists, for example, in Uganda where a large number of grain traders operate. Uganda traders have recently formed a consortium but this consortium is for the purpose of completing gross border sales. Competition within the domestic market remains quite keen. The structure of the grain trading and processing industry in Tanzania is the primary reason that trading profits, as contrasted with supply chain management profits and corresponding incentives for stronger backward and forward linkages persist. Moreover, because of the smaller scale of local demand for maize within Tanzania vis a vis the larger scale of maize demand in Kenya and further because of the small scale of product available for export in Tanzania vis a vis the much larger scale of maize available for export from Uganda, Tanzania's own maize trading policies have, in effect, been overwhelmed by the policies of its two primary maize trading partners. Said another way: Tanzania's own policy framework, which might be described as "laissez faire, except for food security," is diminished in its consequences for local producers and consumers by the policies of its two major maize trading partners, whose food security dependence ( e.g. Kenya) and whose exportable surpluses ( e.g. Uganda) substantially exceed Tanzania's own. Moreover, the Government's quasi-liberal policies based on budget constrained and infrequent interventions into local maize markets might have had a more benign effect on small holders if local markets within Tanzania itself were truly competitive. However, overwhelming evidence exists that effective competition is not present within Tanzania itself. Certainly competition is not sufficient to compel private sector participants in the nation's maize markets to invest in productivity enhancing supply chain assets or in stronger backward linkages to farmers. In contrast, considerable evidence exists that limited competition has allowed strategically positioned participants within Tanzania's maize chains to extract significant rents from the chain, to trade advantageously based on superior information, to control the buy side of the market though ownership both of strategically located storage capacity and superior collection networks and to insulate themselves effectively from market risks, e.g. large scale traders appear to be able to maintain attractive margins in both strong and weak markets. The existing institutional and the governing framework of the maize sector makes it the least effective of the three chains studied with regard to raising farmer's incomes and thus alleviating poverty. In effect, the maize supply chain operates as a poverty trap. It exposes small scale farmers to significant price and market risk and affords them little opportunity for wealth creation or for reinvestment in productive assets, improved processes or better market linked production systems. Moreover, only very weak backward and forward linkages exist within the Tanzania maize chain, itself. For example, input providers extend very little trade credit to farmers to support fertilizer purchases. Merchandisers purchase maize from farmers exclusively on a quid pro quo basis and even merchandisers extend little credit to millers. Moreover, neither input providers nor commercial banks are willing to assume any measurable supply chain risk, either in the form of trade credits, inventory backed loans or input credits mutually secured by groups of farmers. This situation is significantly different for the two other product supply chains which the task team studied. Indeed, a great deal can be learned from the fish and sugar sectors that has immediate application to the maize sector of Tanzania's economy. The market for maize in Tanzania is composed of a set of quite distinct market segments each of which is unique in its supply requirements. The table below identifies the largest of these segments. Supply chains which serve distinct segments in Tanzania have not adapted in specialized ways to serve each segment. Indeed, supply side capabilities remain highly undifferentiated. Two sets of supply structures have emerged: I) the integrated trader segment and ii) traditional buy/ resell chains made up of multiple trader intermediaries. Both of these chains compete in most of the distinct segments identified below with the exception of the emergency food relief segment. Access to that segment is conditioned by a number of procurement requirements manifested by donor and government buyers. The net effect of which is to provide superior market access to integrated traders. Little industrial organization currently exists within traditional supply chains. Most of the interactions which take place within legacy chains are arms length and transactional, as contrasted with longer term commercial relationships which can justify investment, contractual relationships which allow risk to be effectively transferred from parties who are less able to manage it to parties who are better able to manage it and process improvements which benefit both buyers and sellers by reducing transaction cost. Within legacy chains, commercial structures are essentially developed anew with each serial transaction. The result is extremely high transaction cost and little woven-in incentive for productivity improvements, conveyance of information or end to end efficient process alignment. Few incentives exist within legacy chains, for example, to remove process redundancies, to reduce the number of value confirming inspections and tests etc. When incentives are absent to squeeze costs out of the system, these costs naturally accumulate Marke t Niche Integrated Independent Munic ipa l Export E mergency Ani ma l Traders Grain Mil lers Markets Marke ts Food Relief Feed Numbe r o f intermediate 1-2 3-4 3-4 3-4. 1-6 3-4 transactions Information received Receive and disseminate Transmit information Transmission of Advertisements in Transmit and P rima ry me ans from farms via own information through local primarily through local information to and from Newspapers, posted receive marketing network. traders. traders and local farm trading partners via the tenders and outreach information fo r transmitting However, market and marketing cooperatives. internet and receipt of on the WWW. through local information production information Several municipal markets broadcast information traders within the retained internally. are experimenting with vis cell phone SCM and channel Web sites. the WWW. Market information form foreign markets is collected via own agent network. Depend on own network Depend extensively on Depend extensively on Depend extensively on Exclusive Depend extensively for most services, except third party service P rima ry third party service third party service dependence on on third party for transport and telecoms providers, middle men and providers, middle men providers, middle men third parties. All service providers, dependence on which are purchased from independent traders. and independent traders. and independent traders. services are middle men third party third parties However, none of these However, none of these However, none of these outsourced and independent se rvice relationships are long term relationships are long term relationships are long term traders. providers within or contractual. or contractual or contractual the channel The schematic below represents the supply chain which currently operates within Tanzania. The chain includes two major functions: maize production and maize marketing. Production is land and labor intensive, while marketing is information and transportation intensive. These two functions are organizationally distinct and separated in Tanzania. No linkages exist between production activities and marketing activities. Neither information, credit nor technology flows between the two key sets of economic activities. Only product moves between the two functions and then only at very high cost. Transport costs for moving grain in Tanzania from where it is produced in excess to where it is consumed are extremely high. In part this is because maize moves primarily in small lot sizes whose unit transport cost is high. Among the several participants in the Tanzania maize chain only the integrated traders (who are represented within the green box) have been effective in realizing break though process improvements. Their integration through vertical ownership of grain marketing functions under a single corporate control has allowed them to realize above average returns. However, since the two integrators who dominate the market have not been challenged effectively they have not been forced to surrender the productivity gains they have realized in the form of higher maize buying costs or lover maize selling costs. All of the gains realized through private investment in market information systems, storage capacity and transport capacity to date have been captured within the domain of individual integrated grain trading companies. As the schematic suggests two competition enhancing initiatives should be considered in an effort to allow maize producers to break out of the poverty trap in which they are currently stuck: i) increasing competition among integrated grain trading companies, by opening the Tanzanian emergency food relief market to trading companies licensed in other parts of the EAC, and ii) strengthening third party service providers, both through demonstration development projects and subsequently through liberal licensing of specialized maize rail based logistics management companies, storage and asset management companies and specialized market information service providers . Conceptual Map for the Maize Supply Chain in Tanzania Develop systems, management methods and ITC to balance demand and supply Increase competition Develop new among integrated supply business models. chains through procurement Market Develop collaterals of maize for emergency relief Segments and financable securities for these models Plan purchases, Buy Independent hedge Transport Maize Fumigate Mill and Grain Millers positions, to maize During and store package secure deficient Harvest maize transport, price areas to market Export Markets Provide Grow and Inputs to World Food Program Farmers Harvest and Other Donors Maize Emergency Food Local Intermediate Town and City Relief Department Trader Trader Trader Strengthen specialized Other Markets: service Supermarkets and delivery FastFood Outlets Specialized Storage Market And Extend Third Party Information Asset Railway Trucking . chain Services Management Services Services Municipal Services backwards Services Markets From a comparative regional perspective it appears that both market institutions and producer organizations within Tanzania's maize supply chain are less well developed than their counterparts in either Kenya or Uganda. The process of social capital development and of market institutional strengthening have barely commenced in Tanzania. Markets are further distorted by donor interventions in the form of local food aid purchases and local food aid deliveries, by limited access to market information and by poor rural infrastructure all of which inject further transaction cost and risk into maize trading and processing activities. The primary recommendations which emerged from the study with regard to the sugar supply chain include the following: A clear need exists for a zero based review of all of the government functions that affect the maize supply chain. These functions have developed at different times and in response to quite different needs. They lack a coherent architecture and a one of them a contemporary raison d'etre. They include processes and activities which are mutually off setting and which increase transaction costs for private participants in local markets who attempt to comply with them. Taken jointly they create incentives for informal market trading and smuggling and disadvantage formal sector market participants vis a vis informal market participants. Most importantly regulations in this key sector entails little private sector governance, little decision making transparency or accountability. Including the private sector on decision making and advice proving boards and vetting regulations for public commentary before enforcing them would much better integrate what is happening within government with what is happening within Tanzania's several niche markets for maize. Increasingly as regional markets develop within East Africa, food security has become more a regional issue and less a national one. Opportunities for linking up sources of maize surplus with needs for maize in deficit areas are most likely to be found when the circumference within which open market operations are allowed to prevail is expanded. In the case of Tanzania this truism is all the more applicable, because much of Tanzania's maize is produced near the country's borders--a significant distance from its primary national market in Dar and close to other areas of chronic maize deficiency, including much of Malawi, the Nairobi metropolitan market, etc. Given these circumstances open regional trade in maize may well be the most certain and efficient strategy for assuring both grain sufficiency within Tanzania and maximum maize production incentive within the country as well. An in depth assessment needs to be completed of the ways in which emergency food security supply chains operate within Tanzania and that this assessment focus specifically at reconciling different policies and procurement/ distribution methods among the several agencies which provide emergency food relief. Once an assessment of the welfare effects of various food relief policies and mechanisms is complete, monitoring and regulating these policies and methods should then become a primary regulatory function of government. Allowing agents involved in providing food aid to operate independently and without some regulatory constraints within Tanzania is to forfeit one of the most critical market development controls within the grasp of government. In any case, the fact is that several NGO's, multilateral agencies and Tanzania's own Food Security Department currently execute similar functions which are not well coordinated. These need to be reconciled and harmonized so that not only the short term function of emergency food relief is executed in a way that assures better value for money but that longer term maize market and supply chain development objectives are similarly advanced across multiple organizational fronts. This is not happening at the moment. The country's emergency food relief program is closely linked to its trade policy. A determination that an emergency food shortage is forthcoming is sufficient cause within government to shut down export licensing. However, the precise preconditions and/or empirical findings which trigger this draconian measure are neither clearly enumerated nor systematically applied. Similarly, excessively high or low internal prices for maize appear to one of if not the sole sufficient cause for government to diminish or to facilitate import licensing. Again clear criteria and transparent assessments are missing. The recently completed COMESA "Maize Market Study" recommends that regionally acceptable, objectively measurable parameters be used to invoke maize export bans or to trigger maize import restriction and that these be developed within overall framework of the ,Safeguards Clause' of both the EAC and COMESA Treaties. In addition it recommends that a regional food security information clearing house become the source for statistics which would be used to calculate the ban triggering parameters. The authors of this study enthusiastically concur. The COMESA study further recommends that regional policies be harmonized on export and import regulations, that a regional crop forecasting system be developed to provide reliable information on maize availability. Again the study team enthusiastically concurs. Mezzo level organizations also need to be created within the private sector in order to do the work of standards development and enforcement, professional certification and harmonization of commercial trading protocols within the several local market niches which exist within Tanzania. It is critical that these organizations define their membership over the entire set of value adding activities which define a value chain. The government can facilitate the creation of mezzo level organizations by explicitly devolving its responsibilities to one licensed and certified organization for each niche market linked chain. More specifically, efforts need to be made in concert with trade associations representing each of the market niches discussed above to develop standards for food safety and for grain quality which are harmonized to the extend possible with standards which apply to imports and exports and which are broadly accepted within the local markets. Ideally these standards should differentiate multiple levels of quality and thus create incentives for investment in quality production, protection and securitization within entire farm to market chains. The markets for maize in Tanzania are very thinly capitalized. The only working capital available in these markets is the equity and balance sheet backed debt which second parties supply to the market. Very little third or fourth party financing is available to create market depth and stability. The thinness of the Tanzanian market and its relative isolation from other markets are two of the factors that most adversely effect risks associated with investing in it. One potential solution is to subordinate the Tanzanian grain market to a much larger regional market and to allow the market institutions which operate at a pan regional level to subsume trading in Tanzania itself. Several promising new market institutions are emerging within the region which depend on the internet and on internet based clearing mechanisms to operate. Alternatively, a strategic affiliation with the Johannesburg Commodity Exchange is worth considering. It may be possible to create one or more new trading basis for pricing and ownership transfer on the Johannesburg Exchange. The new pricing and trading bases would be located within Tanzania but contracts would be sold in Johannesburg and local store housemen who created the Tanzanian grain securities would operate as licensees of the Exchange. The development of deeper and more stable commodity markets within Tanzania is multifaceted and requires first the development of a network of independent asset managers or public warehousemen who are empowered by financial institutions to create security interests in grain inventories. These asset managers would ideally operate at strategic junctions on the railway system and thus facilitate the inter-linkage of rail and commodity prices in ways which were at the same time transparent, contestable and The efficiency with which external markets operate on the periphery of supply chains is almost as important as how efficiency the value adding activities which are performed which specific supply chains internalize within themselves. Some activities on which supply chains are dependent are more efficiently out-sourced ( externalized) and some are more efficiently in-sourced ( internalized). Chains which become dependent on third party providers of services need these services to be supplied in truly competitive markets less the service suppliers manage to capture a disproportionate share of the value being created in specific chains. Thus, in the case of the maize supply chains key external services such as banking, rail and highway transport, warehousing and market information services are furnished by third parties. It is critically important that each of these providers either operate in a highly competitive and efficient market or that the option be maintained that first and second parties within the chain can enter these markets and provide their own essential services with own work forces, management and capital Lessons Learned A number of lessons were learned from studying the three supply chains, lessons that can be beneficially transferred from one sector another and lessons that can be learned as well from better practice and better policies which are being effected by neighboring countries who confront the same economic and market conditions as Tanzania itself. These lessons are reviewed extensively in Chapter 5 of the supply chain study. The reader is encouraged to scan that chapter. Very briefly, the key lessons that can be usefully taken away from the study include the following: Industrial structures within sectors have a significant effect on sector level productivity, adaptability and poverty impacts. Industrial structures once established are difficult to change, particularly when private interests are able to realize above normal returns on their investment by preventing competitive challenges to the status quo or thorough going restructuring of the sector. Typically, change resistant industrial structures reinforce " zero sum" supply chain systems. Government procurement, fiscal and capital inventive policy, as well as donor sponsored demonstration programs, can be used to accelerate structural change and to begin to move from " zero sum" supply chains to " positive sum" chains. Regulation can and does effect supply chain structure. The ways in which regulations are interpreted and applied as well as the substance of these regulations can and do have structural effects on the underlying sectors, as experience in all three sectors demonstrates. Thus, even is no explicit regulatory authorities are granted over issues of supply chain development, the tacit effects of policies and regulations do, in fact, effect supply chain development. Given this background, explicit regulatory review and discussion of industry structural issues may be more appropriate and beneficial over the long term than their avoidance. Because structural issues have a significant effect on the livelihoods and welfare of primary producers, these issues need to be drawn explicitly into the regulatory domain. In this respect, the Sugar Industry Act defines best regulatory practice. Market institutions complement the development of supply chains and dynamic market institutions challenge supply chains to develop in more efficient ways. This complementarily between market institutions and more productive supply chain structures can entail both a " pull" and a " push" interaction. Potent leverage points for pro-poor private sector development involve facilitating and accelerating the development of markets which are larger in their domain, lower cost, better supported by asset management, cash management and market information systems. The study suggests numerous concrete ways that market institutions can be further developed, deepened and interconnected. Every supply chains requires a unique set of specialized services to support it. These services may include, for example, inbound and outbound logistics services, information services, warehousing and asset management services, etc. Some of these services are supplied out side the chain itself by third parties. This is called "outsourcing." Outsourcing is the prevailing paradigm for the most supply chains in Tanzania. However, when third party service providers fail to supply specialized services to supply chain principals, " in sourcing" or " own sourcing" may be the most economic option. However, in-sourcing typically requires reforms in sector level regulatory regimes, issues of rights and entitlements to non traditional service providers, etc. Issues of service support to specialized supply chains are particularly important for governments to take up at the time that they are considering the privatization or devolution to private parties of management control over state owned assets. Chapter five discusses the implications of supply chain development for the post privatization agenda of the GoT. A substantial amount of working capital in the form both of inventory (including raw material, in process and finished goods inventory) and, to a lesser extent, accounts receivable is tied up in each of the chains. All three commodities/ products have large working capital requirements on account of their seasonal production/ consumption imbalances and also on account of their inefficiency. The net result is a need for external financing. Access to bank credit is a key issue which effects negotiating leverage and equities within all three of the supply chains analyzed. In all three cases upstream supply chain participants had less difficulty than downstream participants in securing access to capital. Chapter five discussed several ways in which this problem might be addressed, including the development of bankable collateral, cross guarantees within chains and within producer organizations ( e.g. grain banks), development of non traditional lending instruments and the development of non traditional companies who sell and finance key inputs to producers in partnership with financial institutions Chapter 1 Supply Chains Development in Tanzania An Assessment of Three Products/ Commodities 1.0 Background Since Tanzania's independence in 1961, its economy has grown at an annual rate of 3.8 percent while its population has increased at 3.1 percent. As a result, per capita income has increased at a mere.7 percent per year and remains among the lowest in SSA at $270. However, in 2000 and again in 2001 the country's growth accelerated to 5.5% while its population growth declined marginally with the result that per capita income increased more rapidly. Still, the distribution of this income remains highly skewed between urban and rural areas. No economic growth appears to have trickled down to the agricultural sector. The economy's response to the rigorous macro economic reform it has undertaken in recent years does not appear to be sufficient by itself to alleviate poverty which is concentrated in Tanzania's rural economy. Even during the country's socialist experiment, the private sector has been the primary contributor to Tanzania's economic growth. However, today, the need for private sector driven growth is more acute than at any time since the 1960's. In 2001 private sector production accounted for more than 85 percent of GDP and for more than 80 percent of fixed asset investment. However, at 13.6 percent of GDP, private sector investment in Tanzania has only attained the average level for SSA countries overall and falls significantly below the 25 percent level that is required for growth sufficient to alleviate poverty within one generation. Since 1995 foreign direct investment has increased markedly to the level of $ 150 million to 220 million per year and accounts for fully 12-14 percent of fixed capital formation per year. At the same time, Tanzania enterprises have become a more attractive target for acquisition and merger. Privatization of state owned enterprises has been the primary magnate for private investment. However, privatization itself has involved more a transformation of ownership than a net increase in capital stock, unless of course, specific privatization transactions mandated asset rehabilitation. In most instances, however, only marginal investments have been needed to transform value subtracting former state owned operations into competitive, value enhancing ones. Moreover, most direct private sector investment has found its way into the enclave industries of tourism and mining, where little collateral value enhancing effects have resulted from either backward or forward linkages. Most importantly, little of the country's private sector investment has found its way into agricultural production or agribusiness. And, indeed, some of the investment which has taken place provides highly questionable collateral benefits. Some direct investment in supermarket retailing, for example, may actually strengthen linkages to and from the agricultural sectors of neighboring economies more than to Tanzania's own. Although it is the largest and most significant sector of Tanzania's economy, the agricultural sector appears to be the least dynamic and the least invested by the private sector. Fully 41.7 percent of the economy's value added comes from agriculture. Significantly, agriculture employs 78 percent of the country's male population and 90 percent of its female population. However, among SSA countries Tanzania ranks near the bottom in terms of food production per capita. Since 1990 the area under permanent cultivation has increased by 1 per cent per year while annual fertilizer imports have declined by 5.8 per cent per year. Over the same period the country's population has increased by 2.8 percent per year. Rather than being enhanced through private investment Tanzania's agricultural production base appears to be declining . Reversing this situation is a high priority for government. One bright spot in Tanzania's recent economic development has been the accelerated participation of the private sector in agricultural trade. During the 1990's the value of agricultural exports from Tanzania has increased at an annual rate of 7.5 percent . Moreover the year to year variability of this growth has declined as Tanzanian farmers have diversified their dependence of cash cropping systems. At the time of Independence cash crops were limited primarily to sisal, coffee and cotton. However, since then tobacco, tea, pyrethrum, cashew nuts and more recently high value horticulture have been added to that list. The private sector has moved rapidly to develop merchandising and distribution channels since agricultural markets was liberalized in the 1980's. Control over output and input was surrendered to the private sector in 1985. Today a core of trading companies specializes in importing and distributing fertilizer and other farm inputs. A separate set of holding companies specializes in merchandising the country's key agricultural exports. . However, in building up trading networks and in facilities which enable cross border trade the private sector has failed to create extensive collateral benefits for small scale farmers. Indeed, the Government of Tanzania is concerned about the large differences that appear to exist between farm-gate prices, on the one end of traditional supply chains, and retail or export prices, on the other end.1 Some evidence exists that market liberalization has lead to a increase and not a decrease in market margins. The implication of large price spreads are that transaction costs within traditional farm to market chains are growing and that substantial value is being removed from chains though inefficient, value depleting intermediate processes, including but not limited to transport and intermediate trading. As a result of this value subtraction, farmers receive less payment for the products they produce, consumers pay more for the food products they consume and pricing incentives within traditional chains distort resource allocation decisions--with the result that supply chain management assets are under invested, production/shipment schedules are mismatched with demand schedules and risks are assigned within chains to the economically weakest participant rather than to the participant best able to manage specific risks.2 Moreover, a " glass ceiling" effect results for the entire set of productive agents within market channels when the supply chains in which they participate are poorly aligned and transaction costs remain relatively high. This effect limits the ability of farm producers and intermediary processors to move up the value ladder from low value commodities to high value specialty products, e.g. high end horticulture, organic, fair trade, branded products, etc. 1.1 Administrative Arrangements Three sectors---fish, sugar, and maize---were initially selected for in- depth study, subject to the approval of government. During a fact finding mission in April, 2004 the study team held a series of meetings with key government officials. The primary objective underlying these meetings was to explain and justify the basis for selecting the three sub-sectors and thus to gain concurrence regarding the selection of the three.. Government officials in the President's office agreed subsequent to these meetings that the three sub sectors did, indeed, provide a reasonable starting point for supply chain analysis. 1Between 1982 and 1999, producer prices for export products as a percent of FOB prices declined sharply from 1982 to 1999 for pyrethrum, tea, tobacco, cotton and coffee and bounced back at the end of the period after declining sharply during the middle of the period. See Table 3.8 in Agriculture in Tanzania Since 1986, World Bank Publication. Absolute spatial margins are also relatively high for staple crops. However, time series analysis of marketing margins over the period 1986 to 1998 indicate that market margins between primary crop production areas and Dar have been declining gradually for most staples. Among staple foods, price spreads remain highest for wheat and rice. Next comes cassava root. Significantly, market margins appear to be increasing over time for cassava. Market spreads appear to be lowest for maize. See p. 37 Ibid. 2One of the key findings of the Agriculture in Tanzania Since 1986 study was that " Market-mediated structural reforms will continue to be difficult to implement until spatial margin margins can be brought down further, through infrastructure improvements and rural transportation policies that reduce transport costs" It can be added temporal margins can also be reduced and not only by reducing transport costs but by reducing transaction costs more generally through the development of more efficient supply chain organizations. The government has a strong sense that reforms at the macro level are nearly complete but that these macro level reforms, in and of themselves, are not adequate to stimulate private sector driven growth. Thus, macro level reforms need to be complemented and extended by sector and sub sector level reforms. The government has essentially requested an conceptual road map to help guide its sector level reform work. While in Dar, the study team also had an opportunity to address interested donors. Donors and NGO's have come together to form a Poverty Alleviation Research Review Board. The mission was invited to present its study plan and its early findings to that board. The review session involved a lively exchange of ideals and views. At its conclusion donors made it quite clear that they would like the Bank to move as quickly as possible in completing the pilot supply chain study. Further, they made it equally clear that if the formal study results confirm and amplify some of the preliminary conclusions which were discussed with them, additional work on other commodities should be pursued . 1.2 Relevance of the Supply Chain Paradigm to Poverty Alleviation Developing integrated supply chain structures is not only pro-competitive but more importantly it is also pro poor. Appropriately designed ago- industrial structures can and do allow poor farmers to break out of the low level equilibrium in which they are frequently stuck and to begin to complete for resources from a higher and more stable equilibrium level. In other words, affecting supply chain structural changes can substantially raise the competitive game of chain participants. Not only does it allow them collectively to match the requirements of specific niche markets into which they may want to sell but are constrained from selling, but also it increases the value available to farmers at the production end of chains.. Supply chain development improves the livelihoods of poor farmers and food processors in four ways: i) It allows them to realize higher prices at the farm gate by reducing market margins; ii) It allows them to differentiate their products by quality category and thus price their products differentially; iii) It allows them to break through the " glass ceiling" of marginal participation in inefficient chains mentioned above. And n additional, vi) integrated chain organizational structures help to interconnect production processes which rely on renewable natural resources with the precise product quality, service and delivery time requirements of specialized niche markets. This latter point needs to be underscored: Integrated supply chains allow producers of primary farm products to participate in market channels which can provide access to high end retail market segments. Furthermore, integrated supply chains allow downstream processors to move into the marketing of more labor and technology intensive products. From an economic perspective, one can say that supply chains accomplish the work of correcting market distortions. Chain restructuring programs allow productive players within chains to "internalize" rent-deducting externalities and hence remove unproductive, value subtracting activities. Chain restructuring involves substituting more productive, third party service providers for less productive third party service providers or, lacking third party alternatives, taking up the provision of essential services with first or second party employees and first or second party supplied capital assets. From a financial management perspective, supply chains minimize the amount of working capital required to move a unit of product from one end of the chain to the other. They substitute processes which are leaner, faster and more agile for processes that are redundant, inefficient and that, because they are decoupled from one another, require buffer inventories to be built up between them. Empowered supply chains minimize waste: They remove redundant processing, redundant inventories and reduce order to delivery times and minimize working capital requirements. They also facilitate the distribution of efficient information on a "need-to-know" basis among trading partners up and down the chain and thus remove rent collecting opportunities that derive from superior access to superior information. More generally, they transform community "bads" which generate negative externalities and into community "goods" which produce positive externalities for chain participants. Importantly, they also substitute "win-win" opportunities and positive sum games among trading partners for negative sum games....for example, for poverty traps which entail " diminishing of product quality " as products move from the farm producers to primary traders and on to secondary traders, etc.. In other words, empowered chains act as scavengers which relentlessly clean up the eco system in which farm to market commerce is conducted. Supply chains are forms of industrial organization that operate at the super enterprise level. Building up these competitive organizational structures needs to be come a high priority for development in Tanzania. Increasingly, competition in global markets is based on the integrated efficiency of contending supply chains and not on the productivity of individual farms, processors or firms that are sandwiched into these chains. Thus, farm producers, increasingly, compete for supermarket shelf space or for "core vendor" status with supermarkets or as " core suppliers" to food processors based on their ability to integrate into agro-industrial chains. Farm based organizations themselves compete on a learning and adaptability basis to become parts of these larger integrated systems. Ultimately it is integrated chains that compete with one another on a price, retail format, market positioning and quality basis rather than individual production units. As supermarket chains become increasingly active in the procurement management of fresh fruits and vegetables and as processed food manufacturers become increasingly adept in their management of sourcing and inbound logistics, supply chain adaptability ( process alignment, standards enforcement, rapid order fulfillment, etc.) on the part of farm producers and food processors become correspondingly important criteria for being selected into integrated systems. Modern farm based chains require close synchronization of all of the constituent value adding activities, including, for example, i) staggered timing of new crop planting and harvesting, ii) selection of hybrid seed which produces products with extended shelf life, iii) selection of specific cropping systems which assure more uniform production, iv) quality systems that further assure uniform shipment, timing and use of particular fertilizers and insecticides, v) order fulfillment systems keyed to continuous replenishment, vi) packaging and branding systems, vii) frequent store direct delivery of small slots, etc. Three supermarket chains are currently competing for market share within major urban markets in Tanzania and this competition is increasingly taking the form of supply chain development. Vendors who become affiliated with one or more of these chains, of necessity, must become more chain management intensive and require increased proficiency with new technologies, including importantly information technology. Competitive and hence sustainable development entails some form of vertical integration or linkage among the formerly disconnected processes of "sourcing", "production" and "distribution." These processes are typically de-linked in traditional chains which require much less process integration and which entail many more intermediate, arms length buy-sell transactions and correspondingly more inventory accumulation and transaction cost than do integrated chains. In the process of integrating and rationalizing value adding process steps, supply chains reformers exclude value neutral ( e.g. simply trading "access to working capital" for discounted purchase prices) or, more frequently value subtracting ( e.g. rent collecting) participants. They squeeze unproductive participants out of competitive chains. In lieu of value diminishing, activities, reformers substitute solid value creating activities. That, very simply, is what supply chain reform is all about: Linking up the value adding activities carried out by and through individual enterprises and aligning these discrete processes with one another through internal systems, incentives, monitors and other modes of behavioral conditioning. As a consequence of better integration, processes mutually reinforce one another and process redundancy and shared information ambiguity are reduced to a minimum and all of the chain participants are able, motivated and incentivized to act more efficiently together. Participants who do not add value to the chain should be omitted from the chain. Third-party supplied service which are not competitive should be replaced by first or second party supplied services which enhance competitiveness. In this way, competitive weaknesses within the chain can be transformed into competitive strengths. The key challenge is to activate this reform process. The study presented here is designed not only to deliver a practical and analytically rigorous " findings" but more important to trigger a sustainable " reform process." If modern retail markets depend for supply on broadly dispersed production, order fulfillment and shipment activities each with long lead times, who should take on this work of chain integration?. Who should integrate and coordinate the activities of all of the participants in chains, including input providers, seed suppliers, farmers, processors, suppliers of packaging material, transporters, retailers, wholesalers and exporters? Typically supply chains are formed, controlled and managed either by a single supply chain integrator (like a Shop Rite or a British Tobacco) or alternatively they form organically from the bottom up, through the active development efforts of mezzo level organizations. In the later case, the provision of supportive services and the sourcing of intermediate inputs into the production process emerge either as joint ventures, cooperative undertakings or stand alone enterprises which are supportive of the strategic development of a particular product chain. The degree of complexity and the structure of competitive supply chains may vary significantly by product type or by niche market. Thus, very different chains operate to supply supermarkets, than to supply export merchandisers, millers, traditional farm markets, street hawkers or indeed fast food outlets. Factors within the larger business environment open up opportunities for integrating new value adding activities--in addition to the minimum ones carried out within traditional chain-- into chains which serve new niche markets . These same factors demand that new supply chain structures be designed to provide secure superstructure for these new value adding activities. Thus, for example, within the niche market for high end horticulture products, compliance with a diversity of new food safety standards enforced by sophisticated overseas buyers has raised transaction costs and chain entry requirements for small scale farmers. Similarly, new process steps like cleaning Irish potatoes and sorting them by size and extent of damage allow one low value line of product to broaden itself into 20 distinct product categories each of which can be differentially priced. Closely related examples include the application and management of appellation control systems and the separation of related but value differentiated products though parallel chains. Still other examples involve the application and monitoring of complex standards for both product (quality) and process (pesticide residue). While externally imposed standards may have driven up farm to market transaction costs other internally developed standards have allowed for effective product differentiation and quality differentiated new product development ( e.g. fair trade, eco-friendly, pro-poor, etc.) Tradition farm-to-market transaction costs are relatively high in Tanzania in any case due to high unit transport costs, high taxes, poor transport asset utilization, etc. Additional standard setting, inspection and qualify verification testing steps that do no enhance product value only exacerbate this high transaction cost circumstance. In general, high value markets require more precise and standardized cultivation, production, selection and packing processes and more IT and more sophisticated packaging. Serving these growing markets is often more capital and skills intensive. However, this not always the case. Hence, all supply chain process innovations must be subject to an economic test of value addition. For example, smallholder farmers find it difficult to compete in high end markets where the marginal cost of quality compliance imposed by local chains and overseas buyers exceeds the benefit associated with imposing additional quality compliance costs on smallholder farmers. Under such circumstance new modes of commercial organization may be required to distribute unit compliance costs over a large number of separate farms. One such example, involves the local supermarket chain Shop Rite Checkers which is reluctant to deal with multiple small scale producers of fresh fruit and vegetables and which prefers to deal with agglomerated suppliers who are fully capable of operating within the product quality control, packaging and store direct delivery requirements imposes by the chain on its " quality vendors." Yet another example involves the World Food Program which is buying maize in Tanzania primarily from large scale commercial producers and middle men who can again fully comply with its procurement and minimum lot size delivery requirements. As farm product markets become increasingly supply chain intensive in Tanzania, economies of scale in production, collection and physical distribution and in quality control compliance appear to be key criteria for market entry. The types of commercial organizational arrangements that can be effectively developed and implemented in Tanzania among small holder farmers will determine in large part the extent to which smallholders can compete in increasingly demanding local urban, regional, and international markets. A key question that the study addresses is this: Can agro-industrial structures be designed and implemented which reduce the competitive disadvantages that small-holders increasingly face in non-labor markets? Further can these new agro-industrial structures be organized to reduce transaction costs and to capture scale economies in transport, marketing, quality control, order fulfillment, etc.? Increasingly, specific distribution channels, e.g. selling fresh fruit and vegetables locally to Shop Rite, selling staples to the World Food Organization, or exporting to the supermarket chains Sans bury in England, require specific organizational structures, technologies, and third-party services to supplement high productivity farm production. A diversity in buyer requirements translates into the need for a corresponding diversity in supply chain responses. Hence, a related question is how many and what forms of supply chain organization are needed to penetrate key growth markets? 1.3 Project Objectives: The objective of this study is to support the Government's strategy articulated in the PRSP to ensure that the private sector plays its role in improving the livelihoods of the rural poor.. The study provides a better understanding of the bottlenecks which hamper the development of stronger linkages between rural and urban segments of Tanzania's economy. IT identifies ways in which new public-private institutions can take up the work of integrating farm to market chains and hence improving farm livelihoods. . The study identifies measures to improve farm productivity and strengthen forward and backward business linkages between and among farmers, SMEs and larger export-oriented companies.. The recommendations and findings which it contains are intended to be integrated into PSD strategies for Tanzania. They should provide the basis for a lending operation inFY05/06. The study which follows assesses the state of supply chains in Tanzania both functionally and strategically. It diagnoses deficiencies in structure and organization of existing chains; It critically reviews the government policy foundations on which chains rest-- including fiscal incentives for strengthening backward and forward linkages and incentives for improving the supporting third party services on which chains rely and the supporting infrastructure on which third parties depend to provide reliable and low cost serviced. Finally it assesses the social capital base on which pro-poor supply chains are most often built from the bottom up. 1.4 Summary of Findings The study's primary finding is that the organization of supply chain structures and both the efficiency and adaptability of supply chain operations significantly determine both chain competitiveness and pro-poor development outcomes in Tanzania. All three of the sub-sectors studied afford employment, business development and potential wealth creation opportunities for large numbers of Tanzanian farmers and fishermen, as well as for small and medium scale enterprises which are not being fully realized. To a greater or lesser extent, the efficiency and industrial structure of all three of these sub sectors studies also affect food affordability and consumer welfare in the country. Said another way, supply chain organization has consequences both on the demand and supply ends of the chain for impoverished Tanzanians. Thus, although fish is produced primarily for export in Tanzania, a significant and growing volume of fish is also produced for domestic consumption. The other two commodities--sugar and maize---are primarily produced for domestic consumption, although a significant volume of both are exported regionally as well. Another primary finding of the work is that industrial organizations ( e.g. vertical and/or horizontal organization within sub sectors, types and efficiencies of backward and forward linkages, levels of competition and/or cooperation within supply chains) for each sector differs substantially one from another. Importantly, no other single factor, other than industrial organization, appears to determine simultaneously levels of competitiveness, transaction cost, levels of rents absorbed within chains and residual value left to farmers or fishermen . Legacy chains operate in Tanzania in ways which effectively constitute "poverty traps." In these legacy chains the level of residual value that is left to poor farmers or fishermen is the residual value which is left after all other supply chain claimants have removed their share. Farmers and fishermen located at the end of long chains are effectively price takers. These legacy chains have evolved in ways which allow multiple intermediate participants to remove more value than they add to the chain and in ways that marginalizes poor farmers and fishermen and severely restricts their ability to accumulate assets and improve their livelihoods. Supply chain systems, which are organized in ways that assure that the residual value left to farmers and fishermen is minimal and, indeed, highly uncertain, impose the conditions of a " poverty trap" on primary producers. Conversely, supply chain systems, which are organized in ways that allow risks to be distributed among supply chain participants, based not on economic strength or weakness, but rather based on the relative abilities of participants to manage or to sell specific categories of risk outside the chain, these chains define positive sum economic games and generally allow poor farmers or fishermen to accumulate assets and wealth. The study found examples of both kinds of chains in Tanzania. The three sets of production, distribution and marketing activities associated with each of the three sectors studied differ fundamentally in the basis on which they allow mutual dependencies and progressively increased specialization to evolve within chains. The possible bases for linking separate production, distribution and marketing activities are: i) cross ownership, ii) contractual affiliation or iii) shared standards, shared information systems and more generally shared social capital. Strong vertical linkages within chains assure resilience, effective risk sharing, and ready access to capital for chain participants. Chains themselves operate as the instrument through which production schedules are matched with expected demand schedules, though which information concerning customers, competitors and new technologies are disseminated, and through which working capital is provided through trade credits and risks are distributed among chain participants through contingent outcome agreements. Importantly, the three sectors differ fundamentally in the levels of social capital which has accumulated within them, in the types and efficiency of information exchanges that take place. Thus, for example, in the case of the maize chain no effective integration exists within the chain or more precisely no effective integration exists which includes consumers and poor farmers; significant information asymmetries persist; end to end transaction costs are high and remain high; third party services have failed to evolve in pro-poor directions; uncovered price and production risks reside fully with farm producers and no financial or risk management instruments have evolved to ameliorate this situation; government policy is neutral on issues of supply chain structure and provides no regulatory framework for the development of pro-poor structures.. The situation in the case of the sugar supply chain, on the other hand, is quite different. The level of integration within sugar chains between out growers and sugar refiners is increasing measurably year over year; a strong regulatory framework exists in the Sugar Act of 2001 for assuring that farmers fully and equitably participate in the division of value within the chain; collective bargaining agreements are evolving as a basis for formalizing internal supply chain operating accords and incentives; consequently, the types and depth of process integration within the chain is gradually increasing. Thus, for example, process steps at the farm level--planting, harvesting and delivery to the refinery---are closely programmed and scheduled in the case of the sugar chain and completely independent and non programmed in the case of maize cultivation. Among the three, the fish supply chain is both the best integrated and the most globally competitive. Its competitive advantages vis a vis other sources of global supply are quite significant and the returns which it provides chain integrators exceed their cost of capital. Sources of competitive advantage in the fish production/ distribution chain are deeply embedded within the chain's structure. Its ability to absorb risk and to continue to compete internationally have been refined and tested against several recent near catastrophic developments in its business environment. The chain's investment in core competencies ( some of which are embedded in individual enterprises which participate in the chain and some are embedded in artisan input providers while others exist for the common use of chain participants and reside within the external market which has developed for specialized services, packaging, etc. that has developed on the periphery of the chain) has proved both flexible and adaptable. They have adapted, for example, both to internally imposed environmental standards and to externally imposed food safety standards, e.g. EUROGAP standards. Importantly, development investment downstream in the chain has proved profitable, as well. The chain integrators who are the fish packing houses have invested in a set of intermediary agents. These agents provide credit and capital assets to small scale artisan fishermen both in fresh and salt waters. The agents collect and delivery their catch to large, modern processing plants but they also manage the production scheduling, productivity enhancement and compliance with new ecological standards of the corps of artisan fishers who operate at the base of this chain.. Significantly, investments in enterprise owned primary processing and frozen food storage have proved equally profitable as have investments by third parties in specialized packaging facilities and in specialized digital telecommunications and transport services. These same investments in essential competencies, by the way, have productive applications not only in fish production/ processing/ distribution but also in related food processing activities. For example, food processing activities which require efficient and highly reliable producer to table cold chains--such as the production and distribution of tropical fruit pulp and concentrated juice. Indeed, at least one of the major fish producers is beginning to diversity into tropical fruit juices. Clearly, the dynamics and structure of the fish supply chain is different in fundamental ways from that of the maize chain. The next most developed supply chain after fish is the raw sugar chain. The raw sugar industry is just emerging from privatization. It is still in a nascent stage of industrial development. Competition is still based on production and distribution costs and market parameters are principally national. However, in 2005 when industrial restructuring and refinery reinvestment commitments are complete, the four private firms who have taken over sugar processing facilities within the country will for the first time have more production capacity for raw sugar than exists demand within the country. Producers within the market still compete primarily based on the full long term cost of producing raw sugar. However, once industry restructuring is complete and once production capacity for the entire industry begins to exceed national demand one or more private enterprises will be forced to price down to the short term variable production cost level and an industry shake out may commence during which none of the enterprises may be able to recoup its full cost of capital. The industry is currently at a strategic cross road where government policy regarding the imposition of trade barriers against refined sugar products (e.g. industrial sugar which is used primarily in brewing and soft drink manufacturing) and/or the lowering of trade barriers vis a vis Kenya may provide additional development trajectories for the industry either toward downstream investment in refined sugar production and value added manufacture of products that require refined sugar as their starting point or into direct competition with Kenyan sugar producers who are less efficient that their Tanzanian counterparts. Backward linkages are beginning to strengthen as the industry expands its production base. All of the major processors are becoming increasingly dependent on out growers for their supplies of cane sugar. Out growers currently supply significantly more than 60% of refinery inputs and this level of reliance is expected to increase as the industry continues to increase its production. The first formal collective bargaining agreement has been negotiated between Kilambaro Sugar and its affiliated out growers. Other similar formal agreements between processors and their supportive out growers are likely to follow. The newly implemented regulatory framework envisioned in the Tanzania Sugar Act empowers government to act as an arbitrator in the resolution of disputes may arise either in the interpretation and implementation or in the negotiation of collective bargaining agreements. The Act provides a best practice example of the intermediating and envisioning role which government can and should play during a period of post privatization, supply chain development. A number of issues involving local infrastructure development, fiscal and tax incentives for production expansion and trade policy still need to be resolved within the industry. However, by and large the sugar sub sector appears to be developing quickly and equitably in Tanzania. Of the three supply chains the one for maize is the least developed and the most problematic. Competition within this chain is based on delivered price; backward and forward linkages from merchandisers to farmers and back to millers are extremely weak; government policy vacillates but generally tends to increase risks to private parties and does nothing to ameliorate the substantial risks that farmers face as the primary producers at the end of long trading chains. Internal markets for maize both within Tanzania and more generally within the ECA are very thin and very weak. Market integration and time/ spatial arbitrage are incomplete and only partially effective over Tanzania's large geography. Thus, for example, linkages with the financial sector of Tanzania's economy are almost non existent. Maize inventory serve neither as bankable collateral nor as a basis for value retention. Post harvest losses are very large ( approaching 30%) , on farm storage capacity is extremely limited and little able to extend the economic life of stored inventory. And, by virtue of their lack of liquidity farmers are forced to sell when prices are low only to have to buy back grain stapes later in the season at highly inflated prices. Large rents are removed from the maize chain by agents who have superior access to information, capital and who qualify for public and donor procurement processes. Maize market institutions are less well developed in Tanzania than either in Kenya or in Uganda. Indeed, the process of social capital development and of market institutional strengthening have barely begun in Tanzania in this sector,. Food aid operations which are extensive within Tanzania only exacerbate the risks that the private sector faces. In summary: Lots of problems persist with the maize supply chain. 1.5 Sources of Competitive Advantage Embedded within Each Chain As we noted above the three supply chains exist in very different states of development and depend on very different sources of competitive advantage. Of the three, the fish chain has developed the most sophisticated and difficult to emulate sources of competitive advantage. Thus, for example, most of the fish processors/ exporters and their affiliated sets of artisan providers in Tanzania have gained full compliance with strict food safety standards---including those promulgated both in the EU and in Japan---- require very tight process integration and control. Fish exporters based in Tanzania have managed to develop deep sources of competitive advantage based on effective process integration, supply chain flexibility and adaptability. These deep sources of advantage are reflected in the sub sector's high compliance rate with Euro gap.....not the other way around . In addition, they have managed to carve out niche markets ( e.g. for Nile Perch fillets in the EU and for fresh large prawns and lobster in East Asia). They have effectively disconnected demand and supply ( both of which are highly seasonal but out of phase in their peaks and valleys) by building up large stocks of frozen fish filets within the country. In the process, they have developed best in class production/ distribution cold chains and in other ways they have managed to create significant barriers to entry into the niche markets they have pioneered for their still non compliant competition based on Cambodia and Viet Nam, Madagascar and Egypt. The industry consists of five or six medium sized enterprises each of which operates its own vertically integrated fish processing/ storage and shipment chains. These leading edge companies are complemented by a small number of fledgling enterprises which operate within specific links in the chain and which collaborate with and depend on larger companies for essential supply chain functions. Significantly, both the larger and the middle scale fish processors depend heavily on artisan fishermen for their supply of both lake fish and ocean fish. The symbiotic commercial relationships which have emerged between the processors and the artisan fishermen are ad hoc, unregulated and subject neither to mandatory arbitration nor to formal dispute resolution. However, these relationships have evolved in ways which allow significant numbers of artisan fishermen to acquire assets, to accumulate savings and to make an above average income, and to enjoy continuously improving livelihoods. The recently enacted environmental legislation which protects Lake Victoria catches acts to strengthen the hand of small scale artisan fishermen who are advantaged under the regulations. The fact that the regulations appear to be working in restoring the renewable natural resource base of the Lake is a big plus. Each of the larger companies in the sub sector manages its own lake to port, lake to airport and ocean to port cold chains. Each, for example, operates its own truck fleet and its own cold storage ware houses. The industries core competency is in precise and "fail safe" cold chain management. The first tier participants in this chain are preparing to step up to a higher level of value added competition, both by producing value added fish products, e.g. fish fingers, fish cakes, single serve micro-wave able potions, pre-cooked fish medleys, etc. and by entering new market niches, e.g. Chinese markets for live fin fish and crustaceans. Currently, the industry produces only fresh fish and fast frozen fish fillets. Little value addition takes place within Tanzania. The critical issue which appears to be retarding move into a higher level domain of value addition are the multiple and continuously increasing taxes which several independent authorities, local, regional and national level tax collecting authorities impose on the fish industry. Through its industry level association the industry wishes to clarify and lock in place ad valorum tax bases before it invests in fixed assets which will increase value created within the industry in Tanzania. Sugar producers have not developed the same deep sources of competitive advantage which the fish sub-sector has developed. Sugar producers continue to compete on a production cost basis, in commodity markets defined for the most part by Tanzania's own borders. Developing low cost, efficient out growers is a cornerstone element in the strategies of each of the major sugar producers. The experimentation that takes place in refining business models and the testing of alternative modes of backward linkage that is currently underway will be highly beneficial for the industry and good for out growers. Existing levels of comparative cost advantage do not allow them to overcome the tariff protection boundaries which fence off markets in Kenya, for example, from formal development by Tanzanian producers. However, informal trade in sugar is certainly active within the ECA and more generally within the region. Still, specific processors within the industry are developing their own strategies for competing in the near future on a value added basis, either within Tanzanian local markets or beyond the countries own domestic markets. Dialogue between government and the Sugar Refiners Association on issues of trade policy and market expansion strategy are particularly important at this juncture in the sub sectors development. Indications are good that this dialogue is both forthright and effective. Maize producers/ millers/ traders have developed no sources of comparative advantage what so ever. Linkages are reinvented with each transaction within this chain. Transaction costs are extremely high; farmers are price takers and market margins are large and growing larger. Little opportunity and less incentive exists to invest in productive assets within this chain. Comparative advantage to the extent that it exists among merchandisers ( buyer/ resellers) is based on superior buying networks, monopoly ownership of storage capacity and superior access to trading information. In summary, the competitive advantages of the fish producers are based on differentiated products, economies of scale and external economies of specialization as well and on continuing investment in cutting edge competences ( including new competencies in fish farming and in value added product processing. Comparative advantage (advantage based on production and distribution cost alone) still applies within the sugar supply chain. However, that chain is quickly evolving into higher order sources of competitive advantage, such as brand differentiated products, value adding processes, etc. The maize supply chain is the weakest of the set. . The comparative advantage of maize marketers is developed anew with each separate transaction and has little permanence or continuity. Given its structure the chain does not seem to justify investment in fixed, long lived assets. Indeed, little evidence exists that effective competition is driving private sector participants to develop sources of competitive advantage which add value. On the contrary a significant body of evidence seems to exist that limited competition has caused significant rents to be extracted from this value chain.. Lessons can be learned from the way that the fish sector has developed which have relevance to the other two sectors. These lessons are elaborated in the sections which follow. 1.6 Sector Specific Regulation: Support for Supply Chain Development Government policy effects supply chain development differently in each sector which the study team assessed. Thus, for example, important differences exist with respect to sector level regulatory regimes, trade policies, tax regimes, sector level chain governance models, tax regimes, the licensing and permitting of specific economic activities, implementation of the terms and conditions of privatization transactions and requirements for the implementation of collective bargaining agreements within supply chains. No clearly defined policy body exists within the maize supply chain. Rather a laissez faire policy applies except during periods of "food shortage" when government the government reserves the right to intervene actively in national maize markets, embargo exports and mobilize the resources of the strategic food reserve. The problem is that a state of food shortage appears to have become a chronic condition in Tanzania and the activation of emergency measures has become routine policy. What the government is lacking are policies and program designed to address the root causes of the chronic grain shortage which the nation suffers in the context of its other commitments to become part of an integrated regional market for maize. At the opposite extreme stands the sugar supply chain for which a clearly framed legislative charter exists. Under the Sugar Industry Act of 2001, a Tanzania Sugar Board is created and empowered to address all of the primary development issues facing the industry, including its vertical integration. Thus, the Sugar Board is authorized to recommend trade policies which balance the interests of the several stakeholder groups represented on the Board, to assist with the development of collective bargaining agreement between out growers and industrial processors, to adjudicate disputes which may arise between parties to these agreements, to allocation preferential access export quotas provided by the EU, and to deal with issues of licensing, industrial structure and internal competition . One of the great advantages that each of the new private entrants into the sugar industry enjoys is that the terms and conditions under which ownership was transferred and privatization transactions completed provides constraints on taxation and user fees and hence provides more investment certainty than is enjoyed by other industries which did not have a similar birth event. In the fish sub sector, in particular, the lack of certainty with respect to future tax liability is holding up additional investment. Fish industry executives would like to invest in additional value added processing to their primary export products as well as in expanding their production chain to include fish farming or aqua culture. However, their recent experience with "tax creep" undermines their confidence and that of their bankers that they can predict with certainty the shares of cash flow which are likely to accrue to investors and those which are likely to accrue to government. It is subject to multiple levels and multiple sources of taxation, user charges, cesses, etc. The industry is concerned that it may die the death of a thousand stings unless the government clarifies and stabilizes the fiscal ground rules under which the industry operates. The fish sub sector is highly regulated with respect to the effects that its operations have on the environment. Thus, existing regulations effect take out rates, size of catch, types of fishing technology employed, etc. In order to fully conform to these requirements that industry has had to build stronger backward linkages to the artisan fishermen who catch most of the fish which it processes and exports. Not only that but the major processor/exporters who are the chain integrators in the fish sector have developed programs designed to increase the productivity of artisan fishermen within the regulated constraints which apply. Thus, two quite different but effective regulatory paradigms exist in the sugar and fish sectors which offer valuable object lessons in how the government might be able to apply regulatory controls, fiscal incentives and other leverage to strengthen linkages within the maize supply chain. 1.7 State of Mezzo Level Development: Institutional Development Assessment Much of the work of supply chain development involves investment in social capital. . The nature of trading relationships themselves vary significantly along a broad spectrum that defines degrees of social capital investments---from supply chains, which are social capital intensive, at one end, and which engender `positive sum' games among trading partners and an industry ethos of mutual dependency to social capital deficient chains, and an industry ethos of sharp dealing and a legacy of product quality deterioration.. Significant problems exist in this arena in the maize supply chain. Smaller problems exist for fish and positive gains in social capital appear to be developing in the sugar sector. Mezzo level organizations are the institutional manifestation of effective social capital formation. Mezzo level organization can and should take up self regulation, investment in community goods ( which generate large externalities for supply chain participants but are always underinvested unless strong supply chain structures have emerged) , development of specialized third party services and development of new complementary goods and services on the periphery of the chain.. Mezzo organizations can take over some of the regulatory work associated with chain development from government or, indeed, take up this work de novo. In the maize supply chain little cooperation or affiliation currently exists between farm level producers, on the one hand, and the primary and secondary merchandisers and processors, on the other. Most farm to market linkages in Tanzania are based on arms length transactional interactions. Little information, credit or product development knowledge passes through existing traditional farm to market chains. The result is uncertainty in supply, minimum product quality control, modest technology transfer and difficulty in securing chain length conformance with various food safety standards which are increasingly essential as pre-conditions for export into key overseas markets. No institutional remedies are anywhere to be found for these problems. Incentives that are currently available to align farm level activities and the activities of primary processors and merchandisers are very weak or simply do not existent. Market price signals are the primary incentive for process alignment. However, Price signals are extremely blunt instruments with which to attempt to control chain co-ordination and alignment. Indeed, the kinds of arms length transactions that prevail in the maize chain are the antithesis of chain integration and provide inadequate incentive for progressive improvement in farm productivity or for market margin reduction as recent experience has revealed. Other mechanisms may be required to facilitate technology transfer, product quality improvement, food security standard conformance, use of intellectual property rights ( e.g. to hybrid seed, genetically reengineered plant stock, hybrid pigs, etc.) within farm to market chains. Mezzo level organizations are most developed in the sugar supply chain both for out grower and for processors. The Sugar Industry Act not only mandates that mezzo organizations be created to select representatives on it board but also lays out at least the initial agenda of the authorized mezzo organizations--the successful development of collective bargaining agreements. However, even without a statutory mandate mezzo level organizations have developing in the fish supply chain. They are, in fact, well developed and active behind the scenes in behalf of large scale processors/ exporters and artisan fishermen. Both sets of mezzo level organization are deeply involved in charting and planning the development of their respective sub sectors The institutional development of supply chains in the maize sub sector may, for example, require the substitution of new farm level business models for traditional small holder farms or the creation of new agro-industrial organizations or the substitution of first and second party provided services for third party provided services at key process interfaces along chains. The development of working chains may call for the substitution of long term, strategic exclusive supplier relationships for short term buy-sell relationships among multiple trading partners. In order to test these and other potential institutional remedies additional work is required in the form of workshops, interaction with sounding board groups, focus group interviews and supply chain development working sessions. This institutional design work constitutes part two of this study effort. . 1.8 Assessment of Third Party Service Delivery Capabilities The ability of third party service providers to design and offer competitively priced services to supply chain dependent producers, merchandisers and processors significantly determines the competitiveness of the environment in which supply chains operate. Among services that are provided by third parties banking, transport and IT services are critically important. However, third party warehousing, asset management and cash and credit management services are critical, as well. The ability of legacy service providers to customize and adapt their offerings to the specialized needs and requirements of real sector producers, merchandisers and processors is particularly important. These are all cross cutting issues that effect all three of the supply chains under study. In general, the post privatization development agenda for services in Tanzania appear to be a significantly limited in vision, in network development leverage and in interconnection enforcement powers and in competition enhancement, more generally. The full report will elaborate on these limitations. Off the shelf service offerings are typically not adequate to the needs of highly competitive supply chains. Customization, time performance incentives, risk sharing and consignment risk transfers, alternative terms of sale, etc. are critical in maintaining competitiveness in the Global Economy. In highly competitive product environments, underlying and supporting service markets need to be just as competitive and just as flexible. Where intense competitiveness among third party service providers is missing it is all the more important that first and second party service provider options be explored. The ability of real sector participants to generate their own designed service solutions, however, may be limited by existing regulatory, licensing and other modes of government control. The flexibility, adequacy and competitiveness of key third party service providers needs to be assessed in the context of each specific supply chain. 1.9 Study Outline The supply chain study which follows is organized into four additional chapters. Chapter 2 deals with the description and diagnosis of the maize supply chain. Chapter 3 deals with the supply chain for sugar and Chapter 4 with the supply chain for fish. Chapter 5 deals with cross cutting issues and with lessons which can be taken from the assessment of the three chains. Chapter 2 Maize Supply Chain Development in Tanzania 1 2.0 Background: Maize Production/ Distribution in Tanzania This section addresses issues that relate to the efficiency of supply chains through which maize produced in Tanzania moves to market. Actually, Tanzanian maize moves to several markets--not just one--and each distinct market has associated with it one or more equally distinct distribution channels or supply chains. The efficiency with which these channels or supply chains operate is the primary topic dealt with in this section. Supply chain efficiency in the maize supply chain is particularly important for development since it directly effects the incomes and livelihoods of approximately 1.5 million farmers in Tanzania and another 300,000 thousand workers who are involved in ancillary activities within the chain.1 More specifically, this section focuses of several interrelated issues that affect the efficiency and adaptability of maize supply chains and of the markets that they serve, including: i) the industrial organization of specific distribution channels and the level of competitive entry, exit and operation into, from and within the niche markets which they serve; ii) the organization of specialized service providers who support each chain and the efficiency of the markets for third party support services on which supply chain participants depend to minimize their transaction costs; and iii) the effects of public sector (and donor induced) policy, regulation and intervention in each niche market/distribution channels that operate in Tanzania. The subsections which follow this background discussion are organized around the following topics: i) the defining characteristics of the several niche markets into which Tanzanian maize is sold, as well as the critical organizational features of each distribution channel(s) which serve(s) each niche market; ii) the key factors which effect transaction costs within maize supply chains--including key factors which are uniquely associated with marketing and transporting maize via each of the several distribution channels defined in (i) This topic includes a discussion of the efficiency of market institutions through which maize is sold and of other supporting institutions that affect both the efficiency and integrity of maize supply chains ( e.g. grades and standards, protection of the security interests of lenders and of absent buyers, etc.); iii) factors which limit the competitiveness and productivity of maize producers in Tanzania; and importantly; vi) supportive ( or not-so-supportive) roles played by several key participants in the maize supply chains, including: a) government agencies that operate at both local and national levels; b) specialized third party providers of essential services to real sector participants in supply chains; c) farmers through their production and marketing organizations; d) private merchandisers and processors and e) donors, including most importantly the World Food Program. Also included in this section is a brief review, under topic (v), of the legal framework that effects the efficiency of maize buy/sell transactions within the nation's maize production/ distribution/ merchandising system. Based on a critical review of these several topics, the section goes on to set out its recommendations in subsection (vi). These recommendations for both policy reform and investment are intended to raise farmer's incomes and to squeeze value out of existing supply chains at both the farm and consumer ends--this as a combination of higher farm gate prices and lower consumer costs. 1Estimates developed from conversations with Tanzanian traders and millers. 2 In comparison with the other two supply chains which the team studied, i.e. fish and sugar, maize appears to be the least efficient and the least developed from both an institutional and poverty impact perspective. Moreover, markets for maize produced in Tanzania are less developed than either fish or sugar markets or, indeed, than the parallel maize markets that operate in neighboring economies. These are the most general conclusions that emerge from this chapter. 2.0.a Overview of Maize Supply Chains Most of Tanzania's maize is produced by small holder farmers who lack on-farm storage capacity and who have only limited access to price and other market information.2 Very little commercial organization has taken place among Tanzania small holder farmers. Consequently, farm organizations are generally ill equipped to deal with issues of risk management, production planning, post harvest loss management, marketing and pricing, working capital financing, transport management and storage of maize. Weakness on the supply side of maize markets is only exacerbated by under development and by concentration on the demand side. Tanzanian maize markets are thin and illiquid except for the liquidity provided directly by a limited number of large scale traders. Commercial bank financing is not available within Tanzania and only the capital of trader principals is available in local markets were most sales are based on cash for product or quid pro quo terms. 3 The farmers who participate in these cash markets have limited savings and typically have an immediate requirement for liquidity. Maize itself plays the role of a near currency in the Tanzanian rural economy. It is a store of value, but one with a limited economic life since post harvest loss rates for maize stored on farms exceed 30%. Farmers typically sell their maize when they have expenses to cover. Maize production occurs within Tanzania in two primary growing areas, the southern and northern highlands. Both of these growing regions are a significant distance from the nation's major urban markets in Dar and Mwanza. Moreover, of the country's twenty regions, only six highland regions regularly produce surpluses.4 Together the two urban markets account for 55% of total consumption.5 Even more significantly these two urban markets are the primary focal points for price formation within Tanzania. 6 In this large country in which net producer and net consumer populations are broadly dispersed both trading and transportation are essential inputs into the maize 2These circumstances on the " sell side" of Tanzania maize markets are confirmed in the household surveys which are being conducted as part of the Poverty Assessment process. See " Is there Life after Liberalization? Transaction Costs Analysis of Maize and Cotton Marketing in Zambia and Tanzania" by Satu Kahkonen and Howard Leathers, 1997 3During the study, team members had several discussions with commercial bankers whose primary interest was in advocating legislation which would have created collateral in farm property including farmer owned inventory and a legal basis for claiming and liquidating that collateral. Subsequently this legislation passed the Tanzanian Parliament. However, the next step--that of converting this new bankable opportunity which the legislation opens up into actual investment in maize supply chains--- is a large one and will require some new thinking and new forms of real sector/ financial sector collaboration. 4See David A. Nyange, " Estimation of Inter ­regional Maize Market Integration in Tanzania and Its Deteminants, " Journal of Rural Problems, ( Japan), Vol.35, No.2 pp59-71 5Dar es Salaam and Mwanza are the primary market price reference points in Tanzania, because of their population ( Dar has 3 million residence and Mwanza has 1 million). Populations in both cities enjoy significantly above average purchasing power and large scale milling, agro-processing and livestock processing facilities are located in both urban centers. 6See Nyange above 3 production function. The graphic representation below highlights the major flows of maize from production centers to consumption centers within the country. The geography of maize marketing in Tanzania Key: -Major producing area -Supply to major market areas -Major market area -Distribution from major Markets to regions and export Markets -Importation of maize & food aid 4 Econometric evidence with respect to the efficiency of underlying markets in transmitting price signals and the effectiveness of temporal and spatial arbitrage in integrating local markets with the major consumer markets in Dar and Mwanza is mixed.7 However, based on anecdotal evidence collected during the study, it appears that price formation in rural areas of Western and Southern Tanzania, appears to be somewhat independent from prices developed in Dar or Mwanza. Even more significantly opportunities for realizing excessive short term profits through temporal and spatial arbitrage appear to be abundant within Tanzania for large scale traders. The analysis contained in Appendix 2 confirms this finding. In the long and inefficient chains that operate in Tanzania, traders enjoy not only the advantage of superior access to information but, importantly, the additional advantage of owning or controlling a significant portion of the country's cereal storage capacity8. Moreover, large scale traders appear to control significantly more storage capacity than do small scale traders. The portion of the nation's maize storage capacity that is not controlled by traders is controlled primarily by the Strategic Grain Reserve Agency. The GRA, however, does not offer a competitive alternative, as a third party storage agent, to small scale farmers who are pressed for cash. One of the reasons that the market for maize is thinly capitalized in Tanzania is that grains cannot be effectively collateralized against bank lending. Another reason is that relatively little third party storage capacity ( as contrasted with first party storage which controlled by farmers or second party storage which is controlled by traders) exists in the country. Moreover, the function of third party asset management which involves securitizing and guaranteeing maize inventory has not yet developed within the country. Indeed, the total storage capacity to sales volume ratio within the country is extremely low and storage capacity offered by third parties as part of an asset management 9 service is nil.10 As a result, prices for grain are more volatile than they would otherwise be and opportunities for temporal arbitrage are highly profitable for traders who have both sufficient liquidity and sufficient storage capacity to participate. Prices tend to raise from low levels during harvest periods in July, August and September to highs during grain shortage periods during March and April. There in lies the opportunity for temporal arbitrage. Again see Appendix 2. 7Several economic tests have been conducted to determine the degree of spatial and temporal integration of Tanzanian maize markets. The conclusions of these tests are mixed.. Their results seem to depend on the conceptual frameworks and the statistical tests applied by the analysts. Thus, .econometric work conducted in developing the World Bank Agricultural Development Strategy for Tanzania suggested that national markets were better integrated than subsequent econometric analysis ( e.g. " Interregional Maize Market Integration in Tanzania" by David Nyange, See http://www.foodnet.cgiar.org/Post%Harvfest/P.... mkt%20 integration. ) The later study found that regional markets in the West of Tanzania not integrated with markets in Dar and that markets in the East of the country are not integared with markets in Mwanza and that markets in the East and West of the country operate somewhat separately. Of the 17 market pairs that were tested for spatial and temporal integration using a Timmer-Ravallion Model, 6 pairs were found to be segregated. Moreover, the model found that market integration declined generally with distance between market pairs and with the quality of the transport infrastructure available. 8The total storage capacity in Tanzania is estimated to be 300 thousand metric tons. Of this total maize buyer/ resellers own, lease or otherwise control approximately 60%. 9The ratio of annual production to storage capacity is 1: 10 this corresponds to sufficient capacity to store just 30 days of consumption. 10Correspondence with Mr. Farid Mohammed, an investment advisor and with Mr. Daniel Kanyi an Apdf- IFC official who are attempting to develop a network of third party maize storage facilities in Kenya. According to Mr. Mohammed Tanzania lags well behind Kenya in the development of third party storage and asset management services. 5 Tanzanian maize markets have important regional, as well as national features. Since Tanzania is a member of the ECA its maize traders enjoy liberal access to markets both in Uganda and Kenya. Importantly, traders in those two countries enjoy equally liberal access into the Tanzanian market. Trading rights are reciprocated. This circumstance has clearly helped in increasing competition and injecting additional trading capital into the local market. Importantly, other major maize deficient markets..... that is in addition to Kenya where maize is in chronic short supply...... are also open to Tanzania traders. These include most importantly Zambia, Malawi and the DRC. In recent years, tariffs barriers among ECA trading partners have been reduced to sufficiently low levels so that an efficient regional market is beginning to develop among and between Kenya, Uganda and Tanzania.11 As discussed below the dynamics of this EAC three way trade significantly effects both supply and demand within Tanzania, itself. Similarly, prevailing tariffs for grain imports into Zambia, the DRC and Malawi are quite low for Tanzania maize traders.12 In recent years Tanzania's cross border trade with its Southern neighbors has continued to grow. Much of this trade, however, is informal.13 For several reasons, prices received by Tanzanian traders are increasingly subject to overall demand/ supply balances within the parameters of regional markets. These reasons include, the following 14 i) Tanzania consumes marginally more maize than it produces and what is produced cannot be stored for more than a month or so because of capacity limitations; ii) Tanzania producers ( unlike Ugandan producers) enjoy only a single harvest in June/ July and are net exporters only during this harvest season, and iii) demand and supply balances vary significantly through the rest of the year among the several countries within the region. Depending both on the timing of their harvests and their capacities to store maize, they are alternatively importers or exporters. These background conditions set the stage for increased regional market integration. One consequence of regional market integration is that prices are more closely linked, for example, among regional trading centers in Dar and Nairobi or Mwanza and Kampala than they are between Tanzania's own primary maize consumption centers or, indeed, among maize surplus or shortage areas within Tanzania itself.15 Even the availability of maize in Tanzania's peripheral rural economy--particularly late in the country's annual production/storage/ consumption cycle---is affected by production and inventory levels in neighboring countries and hence by their own import requirements. During the immediate post harvest season of 2004, for example, much of Tanzania's 11Duties of 3% apply to Tanzanian exports into Kenya and 4% for exports into Uganda. In addition, import declaration fees of 2.75 and 2% respectively apply to exports into the two countries. In the opposite direction, duties on Tanzanian imports from both Kenya and Uganda of 5% prevail. 12Malawi duties on Tanzanian maize imports are zero rated. Zambian duties are 5%. However, much of the Tanzanian/Zambian trade remains informal. 13 See "Regional Maize Trade Policy Paper, Regional Agricultural Trade Expansion Support Program", August 2003, funded by USAID 14This claim is based on correspondence with principals of Foodnet concerning the operation of regional cereals markets in East Africa 15Based on Nyange's work the correlation coefficient between Dar and Mwanza maize prices is only .66 . The two major market centers apparently serve as independent focal points for price formation within Tanzania. Nyange goes on to suggest ( though he does not test the hypothesis) that prices in Mwanza are more closely linked to prices in Kampala which is much more accessible via Lake Victoria than to prices in Dar. The corresponding correlation coefficient for Dar to Narobi maize prices is ___ and for Dar to Kampala maize prices ____. 6 maize stocks were bid away to Kenya. When Kenya's subsequent production fell well below expectations and its own later season import reliance proved to be much greater than was expected earlier, prices jumped in Tanzania. 16 The larger point is this: Maize markets in Tanzania are increasingly linked to parallel markets in Uganda and Kenya and even in Malawi and Zambia17 and grain brokers based in these countries operate alternatively as buyers and sellers within Tanzania's own markets. Strong evidence exists that an integrated regional cereal market is taking form and that liquidity, market information and risk management constraints within that regional market are in the process of being removed. These appear to be positive developments for both farm producers and consumers in Tanzania. Still the agenda associated with regional market development is far from complete and, significantly, Tanzania remains well behind Uganda and Kenya in pursuing this agenda. However, opportunities exist to play the role of "free rider" with respect to regional market institution building. This options appears to be open for Tanzanian participation. Of course, Tanzania's own internal markets have significant problems of their own. For one thing they are thinly capitalized, as we noted above, and not particularly efficient as several of econometric studies have proved. Indeed, within Tanzania local markets are made by relatively few large buyer/resellers. For several reasons: i) because, in most years, aggregate demand is effectively articulated through relatively few buyers who formulate the prices which other, smaller scale market participants accept; ii) because, again in most years, local demand exceeds local supply and the price of imported maize sets the marginal price in the local market late in the season at least in Dar and Mwanza markets. Again it is the larger traders who participate in this import market and who consequently best understand what factors are likely to influence marginal price; iii) because few areas in Tanzania produce a net surplus of grain and these areas are separated from maize deficit areas by long distances and poor infrastructure; iv) because storage capacity is limited compared with aggregate demand, so buying and holding strategies have limited effect; v) because regional markets themselves remain quite thinly capitalized with no trade credits being made available from net exporting countries like Uganda and, finally; vi) because, little opportunity exists to sell risk outside local markets. Indeed, no effective risk hedging instruments exist with which to facilitate such a transfer. For all of these reasons shortages occur with some regularity within Tanzania. These shortages in turn trigger direct government and donor interventions. These interventions, in turn, inject additional uncertainty into local market operations-- uncertainties concerning the volume, timing and price effects of additional food aid. As a results a perverse virtual cycle is perpetuated of under investment by the private sector in risk resilient supply chains, chronic cereal shortages, chronic public and donor intervention and further market destabilization. The maize short months in Tanzania are March, April and parts of May. This is also the period when prices are typically the highest for cereals. Intervention which are not announced publicly in advance of actual buy commitments or, indeed, of actual delivery or interventions which are announced but not implemented limit the ability of local market makers to efficiently discount new information into local market prices. Such interventions not only disrupt the local market but, even worse, they allow trading 17Ibid 7 parties with superior access to information to use that information in their trading operations and hence create invisible barriers to entry for other traders. As a result of ad hoc interventions during crisis periods and tight demand and supply balances within Tanzania itself, prices within the local market bounce between import parity, on the market top, and the marginal cost of local production/ storage, on the market bottom. This circumstance, in turn increases risks for private investment in the maize chain and causes the cost of capital for potential investors in the chain to rise. The end result, not surprisingly, is a general reluctance of invest in long term productive assets within the sector and the regression of the industry into a trading mode as contrasted with a pro-production/processing or supply chain development mode. The maize chain has developed in Tanzania since 1991 in a policy context that might best be described as "laissez faire, except for food security."18 The government allows buyers and sellers to trade with a minimum amount of intervention expect during harvest times when the Emergency Grain Reserve becomes a buyer and during grain shortage periods when it becomes a seller. We commented above on some of the adverse market development effects cause by this intervention. Adverse fiscal effects have also resulted. In spite of the strategic market timing which apparently the Emergency Grain Reserves employs in buying when seasonal prices should be low and selling when they should be high, the Emergency Food Reserve requires a regular annual budget allocation. Strategic market interventions which should result in net profits from annual maize sales instead appear to result in annual trading losses. Interestingly, the Emergency Food Reserve is the largest single line item in the budget of the Ministry of Agriculture. The table in the footnote below reviews the past 10 years of the EFR's budget and estimates the volume of maize which the agency bought and sold during each of those years.19 Significantly, most of the farm level enterprises, which produce maize in Tanzania are very small in scale. In fact, an average farm has less than one acre of maize under cultivation. Each farm produces on average only 1.6 tons per year. Importantly, these small scale, family farms lack both capital and organizational structures which would allow them to diversify production and effectively manage their market risks and/or sell those risks to third parties. Importantly, the largest portion of maize production takes place at a significant distance from town and village markets20 Farmers face the decision of selling at their farm gate 18See " Marketing Liberalization: The rise and fall of Agricultural Liberalization in Tanzania," Brian Cooksey, Development Policy Review, 2003, 21 ( 1): 67-91. Cooksey argues that markets for maize have experienced a " significant and sustained liberalization" and that internal markets for maize are as a result " relatively efficient and competitive." At the same time he argues that markets for other farm products have not been as effectively liberalized with the result that they are less efficient. The Cooksey position, however, runs contrary to the conclusions of this study. 19 1994 1995 1996 1997 1998 1999 2000 2001 2002 Budget 28,000 56,000 105,000 90,000 57,000 25,000 24,000 22,00 25,000 Volume 27798 24275 73109 59154 43145 3500 3000 3700 244,229 of Maize Purchased Source: Crop Monitoring and Early Warning Unit. 20According to Kahkonen and Leathers 88% of maize is sold to traders and transported on average from 20 to 40 km to a town or village market center. Nyange finds that transport costs are more than 50 percent of 8 to traders or transporting their grain a significant distance to local market centers where sale prices are net of transport costs. Small scale farmers also lack the management competencies with which to operate in any market making role other than that of "price taker." Hence, even in a liberal policy environment small scale producers are disadvantaged vis a vis other more sophisticated chain participants who better comprehend regional market dynamics, the likely effects of local demand and supply balances on price and the price consequences of government and donor interventions.21 Moreover, because of the relatively small scale of local demand for maize within Tanzania itself vis a vis the scale of demand for maize in Kenya22 and further because of the small scale of product available for export in Tanzania vis a vis the much larger scale of maize available for export from Uganda, Tanzania's own maize market policies are in effect overwhelmed by the policies adopted by its two primary maize trading partners. Said another way: Tanzania's own policy framework, which might be described as "laissez faire, except for food security," is diminished in its effects on local producers and consumers by the policies of its two major maize trading partners, whose food security dependence ( e.g. Kenya) and whose exportable surpluses ( e.g. Uganda) substantially exceed Tanzania's own. Moreover, the Government's quasi-liberal policies based on budget constrained and infrequent interventions into local maize markets might have had a more benign effect on small holders if local markets within Tanzania itself were truly competitive. However, overwhelming evidence exists that effective competition is not present within Tanzania itself. Certainly competition is not sufficient to compel private sector participants in the nation's maize markets to invest in productivity enhancing supply chain assets or in stronger backward linkages to farmers. In contrast, considerable evidence exists that limited competition has allowed strategically placed participants within Tanzania's maize chains to extract significant rents from the chain, to trade advantageously based on superior information, to control the buy side of the market though ownership both of strategically located storage capacity and superior collection networks and to insulate themselves effectively from market risks, e.g. large scale traders appear to be able to maintain attractive margins in both strong and weak markets. Moreover, only very weak backward and forward linkages exist within the Tanzania maize chain. For example, input providers extend very little trade credit to farmers to support fertilizer purchases. Merchandisers purchase maize from farmers exclusively the farm to market margin for maize and as much as 70 percent for farmers who suffer from poor rural roads.. 21See Cooksey on the effects of donor interventions and Kahkonen and Leathers on the effects of poor transport infrastructure and high transport costs. 22The market for maize delivered though farm to market chains in Kenya is approximately three times larger than the market in Tanzania. Moreover, over the past ten years import demand in Kenya has been growing much faster than import demand in Tanzania. See the table below. Mutual Dependence of Kenya and Tanzania Maize Trade 1998 1999 2000 2001 2002 2003 Kenyan 4 1,201 1,324 2,533 2.608 3,010 Imports Tanzania 0.4 240.2 251.5 264.8 573.7 632.1 Exports to Kenya Tanzanian 61477 18,197 9,124 7,605 8,505 8,567 Imports Kenyan 9221.5 2729.5 1186.1 760.5 936 867.8 Exports to Tanzania 9 on a quid pro quo basis and even merchandisers extend little credit to millers. Moreover, neither input providers nor commercial banks are willing to assume any measurable supply chain risk, either in the form of trade credits, inventory backed loans or input credits mutually secured by groups of farmers. This situation is significantly different for the two other product supply chains which the task team studied. Indeed, a great deal can be learned from the fish and sugar sectors that has immediate application to the maize sector of Tanzania's economy. The existing institutional and the governing framework of the maize sector make it the least effective of the three chains studied with regard to raising farmer's incomes and thus alleviating poverty. In effect, the maize supply chain operates as a poverty trap. It exposes small scale farmers to significant price and market risk and affords them little opportunity for wealth creation or for reinvestment in productive assets, improved processes or better market linked production systems. Much of the discussion which follows involves a critical review of the institutional, regulatory and infrastructure foundations on which the maize supply chain rests. From this discussion the authors have endeavored to extract lessons which are relevant to improving the lot of small scale formers--lessons which may enable them to break out of the poverty trap in which they currently appear to be caught. From a comparative regional perspective it appears that both market institutions and producer organizations within Tanzania's maize supply chain are less well developed than their counterparts in either Kenya or Uganda. The process of social capital development and of market institutional strengthening have barely commenced in Tanzania. Markets are further distorted by donor interventions in the form of local food aid purchases and local food aid deliveries, by limited access to market information and by poor rural infrastructure all of which inject further transaction cost and risk into maize trading and processing activities. 2.1 Characteristics of Niche Markets into which Tanzanian Maize is Sold and the Characteristics of the Distribution Channels which Link Farmers to these Niches This section deals with the structure of markets and the organizational design of supply chains for maize in Tanzania. It is organized into five subsections, which deal respectively with aggregate demand, market structures and market niches, distribution channels which link farms to each distinct market niche, supply chain organization and internal controls, and a conceptual " map" of the value adding processes included in farm to market chains for maize. 2.1.a Aggregate Demand for Maize. Maize is the principal staple food in Tanzania. It dominates the diets of both the rural and urban poor.23 Unlike other consumer markets in East Africa ( e.g. Uganda where maize is readily substitutable for other cereals and tubers), Tanzanian consumers have a strong preference for white maize. They refuse to substitute other staples such as rice, cassava, sorghum, millet and beans except during periods of severe food shortage. Hence, the price elasticity of maize in Tanzania's local markets is relatively low Maize 23Nyange estimates that is contributes 42% of the population's total dietary calories. 10 also accounts for a relatively small portion of animal feed. Approximately 44,000 tons per year of maize are used for animal feed and seed retention.24 Per capita maize consumption is estimated to range from 30 to 35 kg/year.25 Thus, total annual demand for maize is estimated to be approximately 2,700 thousand tons. In recent years, demand for maize has been growing at the rate of overall population increase--approximately 2.8 % annually .26 However, much of this demand is satisfied through on-farm production e.g. subsistence farming. Of total demand approximately 700 thousand tons is satisfied through market channels.27 This is the volume of maize, together with relatively small export volumes, which is traded through the country's supply chains. Average, marketed and traded maize in Tanzania averaged 650 thousand MT during the period 1999 to 2003. The Tanzanian market is among the largest in East Africa. By way of reference, consumption of traded maize in Tanzania is only slightly less than consumption is Kenya or Ethiopia. Of Tanzania's total marketed maize 93 percent is for food consumption. The remainder of marketed demand is accounted for as follows: i) livestock feed--5%; ii) seed retention--1.5 % ; and iii) food processing-- .4%28. In a typical production year the total supply of maize falls slightly short of local demand. However, this deficit balance is an annualized figure and does not account for month to month imbalances of supply and demand within the region and indeed within Tanzania itself. The timing of maize harvests varies with local elevations and local rainfalls among regional producers and this circumstance is conducive of cross border trading. Moreover, Tanzania's long borders with its maize producing and consuming neighbors are quite porous to the informal trade. Significantly, limited storage capacity exists not only within Tanzania but more generally within the region. Limited storage capacity is a cause for much of the seasonal trading which takes place back and forth across the same borders in different directions over a 12 month period.29 The table in the footnote below represents national aggregate demand and supply balances over the past ten years.30 24Estimate based on data from FAO. 25Estimated per capita consumption is based on data developed in the National Bureau of Statistics on Household Census and Budgetary Survey. _____________________. 26As of 1988 according to the census statistics, Tanzania's population was 23.2 million people and had an estimated growth rate of 2.8%. According to the Household Budgetary Survey (HBS) report of 1991/2 urban populations in 1988 was 18% of the total population or 4.176 million. With an average urban growth rate of 10%, the urban population as at 1988 can be estimated at 9.8 million. Assuming maize per capita consumption of 102.2 and 85.95 kgs for rural and urban areas respectively, as highlighted by Household Budgetary Survey (HBS) report 1991/2 and the Strategic Grain Reserve (SGR) report of 1989 and a population growth rate of over 2%, the country's maize consumption level outstrips her production levels by approximately 561 thousand tons and this deficit is growing at the approximate rate of 28 thousand tons per year. 27Estimate of marketed maize provided by National Sample Census of Agriculture, Ministry of Agriculture and livestock development 28Based on FAO Data base. 29Insight provided by Sophie Walker of Foodnet, Uganda 30 Demand/ Local Production Balances Within Tanzania 1994 1995 1996 1997 1998 1999 2000 2001 2002 Demand 2902. 2672. 1904. 2757.7 2510.9 2052.4 2,698 2,777. 2,501. 2 5 4 1 3 3 11 Maize cultivation is more extensive than intensive in Tanzania and most of the arable land available within the country has already been put into production. Unless, the use of fertilizer and hybrid seed inputs is increased significantly the future production is likely to decline marginally. FAO/WFP estimates that maize production will need to grow by approximately 10 percent per year to keep up with demand.31 This growth is possible only if ways are found to recapitalize Tanzania's nutrient depleted soils. A net marginal deficit situation has rendered Tanzania a marginal net annual importer of white maize and of maize related products. Imports occur during portions of the year when local production falls off to zero and when limited storage capacity does not allow for surpluses generated during harvest periods to be carried forward into deficit periods. See the schematic below that describes annual production cycles within the region. According to FAO data, average annual imports from 1998 to 2002 were 89,800 MT into Tanzania. During this period, imports reached a high of 269,600 MT in 1998 because of drought and poor harvests, as well as strong early season sales to Kenya. The same scenario played itself out in 2004. The data are not yet available for 2004, however. Over the same period, imports reached a low of 31,000 MT in 2001 because of plentiful rain, good harvests and low import requirements from Kenya.32 The country's total imports over the last five year period for which data are available were almost half a million .33 This corresponds to a value of approximately $ 150 million. Local 2,874. 2,648. 1,831. 2,684.6 2,451. 2,009. 2733. 2818. 2,540. Productio n 4 2 2 7 3 2 1 0 Defecit 27.8 24.3 73.1 59.15 43.14 43.1 ( Surplus) 35.2 40.8 35.7 Importantly, this deficit closing growth rate substantially exceeds the 2.0% growth in production which has been recorded over the past two decades. In order to achieve a growth rate of 3.5% according to the Ministry of Agriculture and Livestock development, maize production must become both more extensive ( maize growing areas must be expanded) and, at the same time, more intensive (yields must be increased). This will require significant incremental investment and significant technology transfer. To succeed in this, an amendment to the current land policy may be considered--specifically one which encourages foreign investment in irrigation and commercial farming, third party and public warehousing , supply chain financing and market information systems, among others. 31FAO/WFP Special Report on Crop and Food Supply. Substantial improvements in maize yields are possible because the use of fertilizer in Tanzania is not only extremely low ( 6.13 Kg/ha in Tanzania vs 54 Kg/ha in Latin America and 80 Kg/ha in South Asia) but declining. Between 1990-95 and 1996-2000 fertilizer application fell off from 9.02 to 6.13 Kg/ha when subsidies were discontinued. See:" Fertilizer Consumption Patterns in Sub-Saharan Africa," T.S Jayne, Valerie Kelly, and Eric Crawford, July 2003, USAID, African Bureau, Office of Sustainable Development 32Import and export levels appear to be inversely related in Tanzania. Thus, in 1998 when imports reached an extremely high level official exports were nil. In large part this manifests the tight export licensing controls imposed by the government. Conversely, in 2001 when imports were low, export volume exceeded 150,000 MT. The government loosened its export licensing restrictions that year. This inverse import/ export effect reflected in the official statistics is probably overstated , however, because a significant informal cross border trade exists with Kenya, Malawi and the DRC during periods either of food shortage or food deficit. 33The table below represents volumes of imported and exported maize over the past eight years. 1996 1997 1998 1999 2000 2001 2002 2003 Imports 61,477 18,197 9,124 7,605 8,505 8,678 61,477 18,197 Exports 2 1,201 1,324 2,533 2,608 3,010 2 1,201 12 Of this total, the largest share was imported from within the region. Regional imports have accounted for 44.03 % of total Tanzania maize imports over the past five years.34 The remainder is primarily emergency food relief imported by the World Food Program. Regional Maize Production Schedules Jan Feb Mar Apr May June July Aug Sept. Oct. Nov Dec. Kenya Malawi Uganda Tanzania Zambia Source: RATES Country study reports as represented in the RATES Regional Maize Trade Policy Study 2.1.b Market Niches Tanzanian produced maize moves into several distinct markets--not just one. Each distinct market niche has associated with it one or more distribution channels or supply chains. The market institutions which operate in each of these market niches can be defined as `prevailing rules of the commercial game'35. Market institutions create and enforce incentives/disincentives for enhanced transaction efficiency, for assuring that prices are discovered more or less efficiently, that payments are made as agreed and that goods of expected quality are surrendered to buyers. Importantly, they also sanction the non compliant behavior of market participants. Some market institutions are strong in enforcing discipline among participants. These are well invested with social capital. Other institutions are much weaker in enforcing discipline. Thus, for example, legitimacy and some limited degree of disciplinary effectiveness comes with the authority vested by participants in open markets which are frequently organized as collectively owned for profit enterprises in Tanzania. These public market institutions have many of the features of public-private partnerships. Other market institutions are controlled by specific participants in the trade and operate privately in support of their own enterprise interests. These market institutions operate through procurement rules, terms of purchase and product quality guidelines adopted unilaterally by monopsonists. This is the market development approach adopted, for example, by large traders in Tanzania. Typically the scope and scale of the buying and selling activities which these large traders undertake and the sheer volume of their purchasing power allows them to enforce rules within the markets in which they participate. Still other markets operate in response to formal procurement and tendering guidelines which are established and promulgated broadly. The rules which apply in markets Net 61,475 16,996 7800 5072 5897 5668 61,475 16,996 Imports 34See: RATES Regional Maize Trade Policy Study 35See: See: North,D.C (1990): "Institutions, Institutional Change and Economic Performance," Cambridge. and , Hall and Soskice, 2001, "Varieties of Capitalism, The Institutional Foundations of Comparative Advantage," Oxford University Press 13 developed by emergency food relief agencies are the rules of public procurement. These markets should be among the most open and contestable within Tanzania since the interests of the agents who set and enforce these rules is to maximize the use of the scarce public resources with which they are entrusted and to afford all potential sellers equal opportunity to make a sale. Transparency, contestability and openness should be the hallmark characteristics of these markets. The first market niche then is the vertically integrated market niche. This niche is dominated by only two large scale traders who are well organized and well capitalized. These two traders purchase maize through nationwide networks of purchasing agents and store it in nationwide networks of local storehouses.36 They both buy and sell maize to and from second parties and supply maize to their own owned mills which process it into value added products for local sale. The two large scale operators who dominate this niche are Mohammed Enterprises and Export Trading. The table below outlines their maize related activities. Company Estimated Total Grain Number of Other Maize Processing / Volume of Storage Grain Mill Use Assets Maize Capacity Facilities Purchases (1000 tons ) ( 1000 tons) Mohammed 14 20-25 5; all in Dar · 27 Regional storage Enterprises facilities Export Trading 30 50-60 2; all in Dar · 24 Regional storage facilities · Operates cross boarder storage facilities in Zambia and Malawi These two operators buy maize during the harvest seasons (from June to September) directly from large scale producers and indirectly through secondary agents from smaller scale producers. They purchase grain only in the spot market and they pay cash for their purchases. Both have established their own product standards and discount the sale price of maize which does not measure up to these standards. Their marketing and processing activities are backward integrated into a number of additional value chain functions in addition to merchandising. When they buy in the Tanzania market their risk is lower than that of other buyers because they are able to cover a portion of their position by virtue of their own internal demands for mill-able input. These two big companies not only trade domestically in maize but also in neighboring countries. When they are able to secure export permits, which is not often, they are also able to export Tanzanian maize. Thus, for example, both companies buy and sell directly and through joint venture partners in all of the major Southern and Northern maize producing / consuming areas surrounding Tanzania. Due to the volume of trade they conduct and the way they factor multiple factors into their decision making they are considered by other market participants to be "price setters." Other traders buy in reference to the price points which these two integrated 36 Mohammed Enterprises maintains 27 main agencies for purchasing maize in Tanzania while Export Trading maintains 24. 14 traders set weekly. Mohammed Enterprises operates 27 buying posts based in key maize surplus locations within the country. Each of these has a dry storage and fumigation capacity of 1000 MT or more. Each is managed by the company's own staff. Field staff also buy through affiliated agents. Export Trading has developed a look alike network to that of Mohammed Enterprises. Export Trading is larger than Mohammed Enterprises in terms of storage infrastructure and trading volume. The company operates 24 buying posts-each strategically located in key maize production areas. The minimum storage capacity of these posts is approximately 2000 MT. . The second niche market is the large scale, independent grain millers market. This market consists of a relatively small number of milling companies, most of them based in Dar. Participants in the industry are profiled in the table below. Millers typically purchase maize from either large scale producers directly or from the large traders. Millers are large purchasers of imported maize and wheat, as well as buyers of local maize. It is estimated that millers have imported on average 50,000 tons of maize per year over the past three years. Maize imports tend to peak in March, April, and May when local stocks are drawn down and when regional supplies are low. The primary business of the millers is supplying bakeries, food processors, supermarkets and feedlot operators with milled products. The market which large scale millers make is an informal market which typically includes two elements: i) large lot purchases for which the terms and conditions are negotiated between millers and established traders and ii) small lot deliveries which comply with the terms and conditions which millers typically post at their mills. Respondents to these open tenders are primarily smaller traders. In these instances, prices for purchased maize are set by the miller and accepted by the supplier. Miller's margins are determined by the difference between their cost of inputs and the final sale price of their products. Pricing is as dynamic as the underlying market for inputs and is typically done on a "cost plus basis" with price leaders and price followers in the specific milled product market passing input cost adjustments though on a weekly basis. Thus, prices of milled product tend to move up and down with the price of maize delivered to Dodoma and Dar es Salaam. "Cost plus" pricing is possible since, independent miller have no more than 5 days of production capacity on hand. Total annual volume for maize in this market segment is approximately 150,000 MT. Company Number of Installed Storage Location Mills Processing Capacity Capacity (Days Tons/ Day Production) E R 2 120 2-3 Dar es Salaam Investments, Ltd Kizota 1 60 5-6 Dodoma Prime Products Zainabu Grain 1 60 3-4 Dar es Salaam Millers Coast Miller, Ltd 1 120 5-6 Dar es Salaam 15 The third niche market includes open, municipal markets. Dar es Salaam has three municipal markets, one each in Buguruni, Tandale and Mbagala. Municipal markets operate in other cities, as well: For example, Kirumba and Nyakato Markets in Mwanza and Soko-Kuu Market in Dodoma. Municipal markets are typically organized as follows: Agencies of municipal government operate at landlords and lease out stalls and storage houses within market centers to local traders on a fixed fee basis. Payments are typically due monthly. Local traders develop their own cooperative market organizations around market centers and these coops in turn invest collectively in community goods and community services. Collectively as well, they sometimes create collateral which allows market organizations to access credit and to on-lend to individual traders. Local traders buy grain from other traders as well as directly from larger farmers who are located near to specific markets. Thus, traders based in the Tandale market buy primarily from farmers, agro-transporters and middlemen. In the Tandale market, which the project team visited during the study, more than 100 tons of maize are delivered into the market on a weekly basis. Most of this is sold FOM by farmers who absorb the transport costs. However, some of it is purchased at the farm gate by local market traders who buy regularly from a circuit of farm coops with whom they deal regularly. All of the bid and ask prices in this market are negotiated on a transaction by transaction basis. However, according to traders in the Tandale Market few traders are willing to move very significantly beyond the bid or ask prices set by the integrated traders. Both consumers and retail traders purchase maize in 90 Kg sacks and make arrangements with micro millers many of whom who operate on the periphery of municipal markets to mill their maize on a fee or barter basis. Relatively little storage capacity exists within municipal markets as the table below suggests. Approximately 350 thousand tons of maize are traded through municipal markets annually. Municipal Estimated Estimated Storage Location Market Number of Annual Sales Capacity Traders Buguruni 350 300 million TS 2,000 tons Dar Tandale 560 910 million TS 7,000 tons Dar Mbagala. 300 300 million TS 2,000 tons Dar The fourth niche market is the export market. An informal export trade in maize has existed for a very long time and is well entrenched in local traditions, as the table below suggests. In 2001, the Tanzanian government first lifted a preexisting ban on maize exports. This policy coincided with high food demand in neighboring countries. However, the ban has now been reinstated. Still, owing to Tanzania's extremely porous borders, substantial volumes continued to be traded out of the country through the informal trade. Officially, a license is required to export maize. Approximately 15 traders are licensed by the Ministry of Agriculture. Of these the most active maize exporters were Export Trading Company, Mohammed Enterprises and Azam Bakhresa. These enterprises account for approximately 8 % of Tanzania's total exports. The table below represents the destinations for most of Tanzania's export grain over the past three years. Kenya is the most regular market and exporters in Tanzania especially ones based in the North sell regularly to millers based in Kenya. In recent years, Zambia, Malawi, the DRC and Zimbabwe have become major export markets as well. 16 The table below describes the flow of maize within the regional market. Note the central place which Tanzania plays in this trade. As was pointed out in the previous section all of the major regional buyers and sellers of maize effect local markets within Tanzania either directly or indirectly. Maize Trade Flows Among East African Countries Source: COMESA Regional Trade Study The fifth niche is the one defined around the procurement methods of the World Food Program (WFP) disaster fund, and to a lesser extent the Strategic Grain Reserve (SGR). The rules promulgated by these two institutions constitute substantial markets in and of themselves. Prior to liberalization, SGR constituted the single market channel though which Tanzania maize moved to market.. After liberalization its role has diminished progressively due to competition and internal constraints, e.g. lack of funds, 17 bureaucratic limitations, etc. However, because of the priority access to limited transport and storage assets which both programs enjoy and because of the nature of their marginal demand and supply balancing interventions this particular segment causes significant carry over effects for all of the other market segments discussed above . SGR procures maize in response to competitive tenders. The terms and conditions of a typical SGR tender are outlines in the table below. Typically, SGR buys in the months of March through May and sells, again through competitive tender, in the months of September through December. On average SGR has purchased 125 thousand tons per year over the past three tons per year over the past three years. SGR Tender Terms and Conditions Tender Terms and Bonds and Third Minimum Other Conditions Party Tender Size Qualifications Certifications Buying Type of maize: Trade license Minimum na from SGR delivery Quality of maize Bank statement for quantity: 3 credit access metric tons Delivery specifications Insurance policy Selling to Item: White maize Trade license Minimum Delivery Location SGR delivery Delivery Bank statement for quantity: 7 Deliveries may be Specifications: credit access metric tons made to specific SGR Depot Delivery in 90 kg Insurance policy locations. Tender bags; Moisture bids should specie content not exceeding locations to which 12% maize would be delivered WFP is quite a different story. WFP buys maize in Tanzania for food relief both for food relief in other neighboring countries and for food relief within Tanzania itself. For many large farmers, WFP is a preferred buyers, since the UN Agency pays a premium price for good quality maize. Significantly, the terms and conditions under which the WFP purchases grain are different from the terms and conditions required in any of the other market segments. Sellers to the WFP must actually hold and be able to demonstrate their ownership of maize which they are offering to the WFP before submitting a tender. Moreover, under its procurement guidelines, the WFP essentially purchases a " call option" and not a spot purchase. The striking price of the option is set at the offering price tendered by the seller. The call is good typically for 30 days. However, the WFP often requests that sellers extend the term of their price commitment. Frequently, the WFP finds itself working against prices set in other market segments by requiring its offerors to extend their set terms in the face of rising prices. In addition, all tendering parties are required at the time that they tender their offer to post an earnest payment of 10% of the offered price. This payment and collateral bonds can be confiscated by the WFP is the seller fails to make good their product delivery when the WFP exercises its option. Finally, WFP sometimes requires that tendering parties deliver to WFP leased warehouses, sometimes to port warehouses 18 and sometimes that tendering parties deliver forward to aid recipients. Failure to comply fully with WFP guidelines can and does result in severe sanctions including exclusion from the list of pre-approved bidders. WFP has tried but failed to exclude the two large integrated trades from participating in its procurements.... but without success. They are apparently too large and too significant for the WFP to do without. WFP operates two procurement programs: One for large-scale producers, traders, and one for small-scale producers and traders. The terms and conditions of both types of tenders are outlined in the table below. Typically, WFP buys in the months of December through February and distributes grain during the months of April through June. The distribution process is intended to have minimum market impact. It is based on the mechanisms of Food Security of Households within a given region of the country. However, some WFP maize invariably finds its way back into commercial channels. On average WFP has purchased over 230 thousand tons per year in Tanzania over the past three years. WFP Tender Terms and Conditions Type of Offeror's Bonds and Typical Maize Specs. Other Qualifications Offeror Pre-qualifications Third Party Minimum Certifications Tender Size Large Offerors should be able to Trade license Minimum 20 tons White maize Must post an earnest Scale provide a general payment of 10% of the Sellers description of the nature of Bank guarantee Moisture content not offered price the services to be provided for credit access exceeding 12% and bidders are invited to submit proposals for Insurance policy supplying the produce Collateral bonds Small Offerors should be able to Collateral bond Maximum 5 tons White maize Must post an earnest Scale provide a general payment of 10% of the Offerors description of the nature of Moisture content not offered price the services to be provided exceeding 12% and bidders are invited to submit proposals for supplying the produce The sixth niche market involves sales through local town and village markets through which many small producers sell their surplus maize primarily for local consumption. However, local village collectors frequently accumulate stocks in local markets and brokers sell these stocks to larger traders. Part of this channel also involves small scale wholesalers who mainly buy from village collectors. They again provide town posho shops and sometimes even sell to small exporters. Small traders normally trade in 30 tons trucks lots(for the export market) The exact volume traded by the small traders is not known since most of their transactions go un- registered. The final market niche is the market for animal feed. This market operates in modes similar to municipal markets. Consumers and retail traders purchase animal feeds and make arrangements from micro millers many of whom who operate on the periphery of municipal markets to mill their maize on a fee or barter basis. The maize husks are what are usually sold as animal feed. The feed is sold by both retail traders and wholesalers. The table below summarizes some to the key parameters that characterize each of the major niche markets described above. 19 Market Niche Integrated Independent Grain Municipal Export Markets Emergency Food Animal Feed Traders Millers Relief Operations Estimated level An actual figure can Transactions between On average over On the buy side On average during Estimates of transaction not be provided but Tandale millers and 1500 transactions 20 and on the catastrophic range from 20- activity during the level of farmers and ultimate are conducted on sell side one or periods such as 100 the harvest transactions between customers average 50 daily basis two... drought and floods transactions season for a integrated traders and transactions on a between farmers intermediate on daily basis single their customers is daily basis. An and ultimate transactions in Tandale representative quite low compared additional 10 customers average 300 on trader to the level in transactions are daily basis municipal markets conducted with large- scale processors or traders Estimated 2 Six large miller in 40 markets 8-12 active Five: WFP; SGR; Not Available. number of addition to in the country participants UN Children's However, market making approximately 50-250 including traders Fund and UNCEF; approximately participants small scale millers in based outside Disaster Fund 65% of regional markets of Tanzania independent Tanzania grain millers also participate in animal feed Degree of Relatively strong; Relatively weak; since Relatively strong; Relatively weak; On the whole Linkages are backward Purchase from famers they purchase from Local market Trading linkages are weak linkage to through a network of large-scale producers traders buying relationships negligible Farmers local agents or large-scale traders directly from based on the farmers annual production Degree of price High: Prices are Low: Market prices High: This is Medium: High: They work Low transparency posted weekly and tend to differ slightly basing on the fact Discrepancies on tender basis and differ based only on from one region to that prices are exist. Prices pay a premium the quality of the another, based on the generally set by offered by price for good produce quality, scarcity and the existing different quality maize socio-economic market conditions exporters differ conditions in a given in their different region sourcing markets Degree of price High High High Fair Low High elasticity Approximate 15 50 60 25 125 6 Annual Volume( 1000 metric tones) 2.1.c Distribution Channels An extensive network of affiliated traders and warehousemen, as well as of non- affiliated traders, service providers, warehousemen and asset managers operate the distribution channels which feed each of the niche markets identified above. The ratio 20 of workers involved in maize distribution and marketing to the number of farmers involved in maize production is estimated to be approximately 20 : 100 37 With the exception of company employed traders and warehousemen, who support the efforts of integrated trading companies, most of the rest of work force involved in maize distribution and marketing do not associate exclusively with a specific niche market or specific distribution channel. Agents, brokers and traders.....and even mini millers..... associate with buyers or seller for a single transaction and then re-associate with other buyers and seller for subsequent transactions. Thus, they may buy from large/medium farmers, as well as from small farmers, either directly or from village collectors and/or small wholesalers. With respect to the outlets which they supply they typically serve various alternatives including, millers, exporters, WFP and even the integrated traders. This " atomic aspect" of supply chains for maize in Tanzania is their defining feature. Supply chain associations are configured and reconfigured on a transaction by transaction basis. This is not the most efficient way to reduce transaction costs, to build network capital or to encourage long term investment in supply chain efficiency improving assets. Indeed, much of the value addition within maize supply chains involves realizing economies of scale in transportation through consolidation of small production lots into larger more economic lots. The process is referred to as " bulking up". The earlier in the chain that large lot sizes form for subsequent movement through the rest of the chain the larger the scope which exists for securing economies of scale within the channel. The significance of this opportunity becomes immediately apparent when we reflect that fully 85% of maize production comes from small holders. Small holders only sell their surplus maize and consume the rest within their households, often after having it processed through a village posho mill. For the most part, these farmers trade only in extremely small lot sizes. Because of the extremely high unit transaction cost associated with dealing at this level in the chain, their business is of little interest to the larger traders who require that the consignments they purchase first be "bulked up". In this way, the organizational structures through which maize moves to market determines the high transaction costs and large market margins that are incurred within specific channels. Another source of value addition within value chains is the preservation of value created by cultivating, harvesting and processing maize of superior value. If high quality maize is not segmented and separately marketed its value will inevitably be blended down to the level of the lowest common denominator. A few, limited opportunities for differentiating maize products exist in Tanzania. These include the packaging, branding and differential pricing of maize products for different grades. Unfortunately little effort is being expended by millers or anyone else to develop quality standards or quality differentiated channels. The fact is that distribution channels have not yet been designed in Tanzania which can protect the value content of products derived from superior quality maize. Little incentive exists anywhere in the chain to invest in superior products if the quality of those products cannot not be preserved within the chain. The distribution channel for maize which comes closest to a quality management capability is the one which supports the large traders/processors--- Mohammed Enterprises and Export Trading. These two enterprises buy directly only from a few large producers. These producers operate farms which range between 1 to 5 hectares. 37Based on conversations with millers and traders 21 For the most part they employ traditional labor intensive technology in all farm processes from cultivation to harvesting. However the large trader also deal with a few farmers who apply mechanized technology from cultivation to harvest and who farm larger plots of 5 to 10 hectares. Explicit backward linkages have not formed with these farmers in the formal sense of a contract or joint venture. However, a commercial rapport has developed based on trust and long term and mutually beneficial trading relationships. The two integrated trading companies internalize a number of value enhancing functions within their own enterprises and outsource relatively few value adding functions. Thus, for example, they apply their own quality standards at the time of purchase. 38These big companies not only trade in maize but also process and export maize.. As noted above they both operate their own buying networks, which they staff with their own salaried employees. Importantly, they own large, near farm, go-downs, which enable them to buy large quantities of maize when the price is low and to leverage their buying power further when it comes time to transport their grain to market. Significantly, the two large trader/processors are also the two largest grain hauling customers of the Tanzania/ Zambian Railway. All other channels are supported by independent agents, brokers and traders, many of whom are able to handle reasonably large quantities. They buy from large /medium farmers as well as from small farmers, either directly or from village collectors and small wholesalers. With respect to their outlets they have various alternative s including millers, exporters, WFP and also large traders. Even though the participants in specific channels may, to a significant extent, be interchangeable, the structure and industrial organization of specific channels does vary as the table below high lights. 38Mohammed Enterprises differentiates two categories of maize, which are distinguished by the packaging provided. These are mainly 50 and 25 kg packs.. Export Trading likewise has developed its own quality standards and brand. The company specializes in white maize and provides it in packages of 100 kg. 22 Market Integrated Independent Municipal Export Emergency Animal Niche Traders Grain Millers Markets Markets Food Relief Feed Number of 1-2 3-4 3-4 3-4. 1-6 3-4 intermediate transactions Information Receive and Transmit information Transmission of Advertisements Transmit and primarily through information to and receive Primary from farms via own information through Newspapers, information means for marketing network. traders and local trading partners via through local transmitting However, market traders. tenders and traders marketing internet and receipt information production on the WWW. within the Several municipal broadcast retained internally. channel are experimenting vis cell phone SCM Web sites. the WWW. Market information form markets is collected own agent network. Depend on own Depend extensively on Depend extensively Depend extensively Exclusive Depend third party service dependence on Primary For most services, providers, middle men third party service third party service third parties. All on third party dependence except for transport providers, middle providers, middle services are service and telecoms which independent traders. outsourced on third are purchased from However, none of and independent and independent middle men party service third parties traders. and independent providers relationships are long However, none of However, none of traders. term within the or contractual. relationships are long relationships are channel term or contractual long term or contractual 2.1. d Supply Chain Organization and Channel Controls All of the supply chains for maize which operate in Tanzania are of necessity" market facing." Critical market interfaces in these chains operate both at the production and the retail distribution ends, without any process integration and little systematic information sharing among activities imbedde in the chains. For example, no scheduling of product delivery onto market exists, with the exception of the large traders. As a result of their unique capacity to actively manage the storage, shipment and selling functions, the large traders can act on the superior information to which they have access and schedule their purchases and their sales to coincide with optimal market conditions at both farm and retail ends of their integrated chains. Yet even the integrated chains operated by the large traders are market facing. Their integrated operations exclude both food retailers and farm producers. The net result of an anomic supply chain structure is that processes imbedded within chains are only gradually and indirectly responsive to market, technology and process changes. Thus, prices at both the farm gate and the retail ends of the chain induce buyers and sellers to adjust, but to adjust only gradually to shortages, blockages, excess inventory positions, etc which take place in other links of the chain . Moreover, coordination among processes is low with the result that fixed asset utilization is low, inventory accumulation between process steps is high, working capital requirements to 23 finance the chain are extremely high and losses from wastage, spoilage and pilferage are also high. Process responses at the level of discrete logistics activities within the chain are slow because response require an interpretation of price signals before specific actions can be taken. Moreover, because price signals are subject to multiple interpretations price changes can trigger divergent responses. Moreover, in most cases these responses are significantly time lagged or simply un-enforced by commercial agents in the chain who are separated spatially and organizationally from one another. Again the notable exception is the chain which supports the activities of integrated traders. Significantly, uncertainties associated with the need to reconfigure end to end commercial relationships with each transaction increase transactions costs and transaction risks. Moreover, investment in supply chain productivity enhancing assets have failed to attain the level where marginal costs from incremental investment equal marginal benefits, because these externalities are not being captured by supply chain integrators who have emerged in the private sector. Moreover, the GofT has committed itself to market non intervention except for food security, so public investment in the community goods needed to improve supply chain performance is not forth coming either. Thus, assets which could improve supply chain efficiency for all participants remain underinvested. Since the benefit stream from these investments cannot be easily captured by single enterprises or internalized by investors in long term commercial relationships, they are not financially feasible . The result, over time, is little or no improvement in farm to market transaction costs and, indeed, a tendency to slip backwards with respect to transaction cost to levels which prevailed during the pre-liberalization era when the government managed the single channel through which all maize moved. These phenomena have been recognized and commented on in two papers: In " Is There Life after Liberalization?"39 the authors note that market margins for Tanzanian maize have increased over time. They ascribed this increase in market margins to the government's failure to invest adequately in transportation infrastructure and to the collateral failure of the private sector to develop competitiveness enhancing service support systems. This is essentially a version of the same argument which the authors are presenting in this chapter but with a bigger picture perspective. In " Marketing Reform: The Rise and Fall of Agricultural Liberalization in Tanzania"40 the author argues that the liberalization of agricultural markets in Tanzania has been followed by "recrudescence statist legislation, policies and practices." He suggests essentially that incremental subversive operations within government have had the net effect of reversing the gains realized through early market liberalization. His argument is that the accumulation of various implementation failures are in part of a concerted effort on the part of government and government insiders to reverse the reforms of the 1990's. Among the specific product markets which he assesses, however, the author holds out the maize market as a singular example of prudential restraint. In the maize market in particular the author argues the general backward slide into a statist agricultural regime has been halted. However, the author fails to offer an explanation for the fact that transaction costs between farms and markets are high and appear to be increasing. 39" Is There Life After Liberalization? Transaction Cost Analysis of Maize and Cotton Marketing in Zambia and Tanzania;" by Satu Kahkonen and Howard Leathers, Prepared for USAID 40" Marketing Reform? The Rise and Fall of Agricultural Liberalization in Tanzania," by Brian Cooksey, Development Policy Review, 2003, 21 (1):67-91 24 The interpretation offered in this paper is that a failure has taken place not only in markets or, indeed, in government policy but rather in the agro industrial structure which has emerged since liberalization. Inadequate tax and fiscal incentives and faulty policy have not provided any incentive for strong forward and backward linkages within the maize trade. The result not surprisingly is the set of anomic supply chains that we described above. In these chains, key integrating activities are missing, critical assets are under funded, processes continue to lack integration, information is horded rather than shared and little social capital has accumulated. Specific examples of what is missing include the following: i) precisely forecast production requirements which are shared with stakeholders, ii) maize quality improvement programs which align incentives for producing and marketing high quality maize, iii) financing instruments for inputs and trade credits for fertilizer and hybrid seed, vi) programmed off take and shipment scheduling, v) forward contracting for production under out grower and similar schemes, vi) active management of marketing channels, vii) third party production management for increased yield and quality, vii) planned and programmed third party provided storage, vii) third party channel coordination, forward order fulfillment, etc. ix), third party insurance against crop failure and price spikes, etc. Moreover, little risk management and/or credit provision currently takes place within legacy maize chains. The notable exceptions to this generalization are the integrated traders who have effectively become supply chain integrators. However, since the level of competition within the segment of the market which they occupy is limited and since significant barriers exist for new entrants who may wish to emulate their success, the large traders are not forced by market pressures to compete away the gains which they have realized from integrating intermediate processes in their own managed chains. Moreover, they have declined to integrate their chains any further into either retailing or farming. Hence, farmers and retailers continue to operate independently of other chain participants, with the net effect that gains realized from improved chain continue to escape them. Sole reliance on market mechanisms for the coordinating and integrating intermediate processes in long chains is typically not as effective as including these processed within industrial structures in which they can be effectively motivated and controlled by fiat, by performance incentive and by self calibrating process control. However, prototype agro-industrial structures ( business modelsa) which include and encase investment in supply chain assets and which span multiple value adding processes in maize chains have failed to emerge in Tanzania. Total reliance on arms length transaction mechanisms to reduce costs, working captital requirements and risks has not proved to be noteably successful since the maize sector was liberalized in 1981. During the interim period instead of strengthening, relationship between buyers and sellers have remained transient and information exchanges remain asymmetric. Little long term contracting exists within legacy chains. Time horizons for forward buying ( at the retail/ processing end) or postponed selling ( at the production end) are limited by three factors in legacy chains: i) limited liquidity and zero finance-ability of stocks through third parties ( e.g. commercial banks) ; ii) limited storage capacity to buy and hold dried maize; and iii) the failure of a market to emerge for risk management instrument that could be used to sell specific categories of risk outside traditional chains. 25 The short-lived, "arms length" transaction relationships which persist within Tanzania's legacy maize supply chains have an additional adverse effect: Arms length trading relationships amplify rather than diminish trading and pricing risks. Importantly, they also reduce incentives for: i) conveying accurate and timely information among trading partners, ii) investing in community assets which advantage multiple trading partners and iii) transferring technology advances through the chain, especially if these technologies threaten traditional balances. Since liberalization, trading within the maize chain has increasingly taken the form of a "zero sum game" among trading partners. Cooperative activities and positive sum outcomes have little place in these chains. 2.1.e Process Coordination and Risk Management Within Legacy Chains. Some additional discussion is in order concerning the way in which the maize supply chain operates at the retail end.. No effective national retail markets exist within Tanzania. Supermarket chains are beginning to develop an integrated national market but one does not yet exist. Rather all retail markets for maize and for most other food products as well are local markets in Tanzania. Dar es Salaam is the largest of these local markets. The retail market centered in Dar is highly competitive . It supports multiple retailing formats and multiple outlets, too numerous to enumerate. In spite of the intensity of competition within the metropolitan area, however, retail markets are highly differentiated and different market formats appeal to different segments of the total population. Thus, in the spring of 2004, maize retailers based in Dar es Salaam were paying prices which ranged between 400 and 800 TS/ Kilo depending on market segment, product type and price elasticity of the principle consumers in each segment. Consumers in Dar, for example, have the option of purchasing a diversity of finished milled, unfinished milled and un-milled maize products, as well as multiple options for purchasing these products in a diversity of retail and/or wholesale formats. As underscored in the table below, consumer preferences for maize products and their corresponding willingness to pay vary widely depending on several factors, including the perceived quality of the product, its packaging, its advertised brand, the sale of complementary products in the same retail venue, impulse buying or promotion induced sales, and importantly the disposable income of the target consumer. With that said, in Dar the retail venue itself is a significant determinate of price for milled maize products. Thus, prices for similar maize products tend to vary among retail outlets with, for example, prices varying among chain stores, wholesale and traditional municipal markets. Retail prices that apply among several retail venues during February 2004 for coarse maize meal included the following: 26 Other supply chain issues process management issues also effect retailers. For example, retailers who offer maize to consumers do not schedule or program its delivery into their warehouses. Rather they make arrangements for the purchase of maize products and its pick up at the warehouse dock of wholesalers. Neither do retailers share any risks associated with production and maintenance of safety stocks with the wholesalers or millers who supply them. Downstream pricing is done based on "bid and ask" negotiations. The concept of " every day low pricing" does not apply in these channels. Higher than market prices offered to competing offerors who operate in the same segment is an frequent justification for changing suppliers. Margins realized from milling maize differ widely among processors.......especially between large scale processors, on the one hand, and small-scale processors, on the other. Small scale processors control the largest share of the milled product market both in urban and rural settings. They are the primary market makers for milled product at the wholesale level. Not only is their cost structure lower than that of large scale millers but their services are generally more "customer oriented" and thus preferred by consumers. Indeed, their services appear to be preferred even by Aid agencies and by the SGR. Donor agencies, such as WFP, for example prefer to rely on hammer mill operators to process the maize they buy than on large scale mill operators. However, hammer millers are not able to fully comply with the WFP's large volume processing and precise delivery tender pre- requirements. Another source of competitive advantage enjoyed by micro millers are the terms of sale and the terms of payment which they offer. Micro millers sell either ready to cook milled products or milling services with the raw dried maize being supplied by customers. They hold themselves out to be compensated based on a percent of milled product take back or to extend credit to long standing customers. Larger millers are not prepared to offer their services on a comparable basis. Traders within legacy supply chains based on Tanzania perform several important functions. They help to set prices by buying and selling for their own accounts. Importantly as well they provide the liquidity that exists within the maize channel. The working capital requirements associated with financing the maize supply chain can be estimated and that estimate as explained in Attachment 2.1 Almost all of this capital comes from the traders who buy, hold and sell maize within the country. Traders equity provides the working capital that floats the entire supply chain in Tanzania. Very little, third party ( e.g. commercial bank) financing is available to participants in the chain. Traders perform another valuable service for all chain participants. They extend the economic life of maize inventories. They do this by properly drying, fumigating and storing maize products. It is estimated that the economic life of maize stored on farms in Tanzania is approximately 3 months. The economic life of maize suitable for human consumption when it is property treated and stored can be extended to 18 months. Efforts of farmers as well to dry their maize after harvesting are crucial to extending its economic life sufficiently so that it can be handed off to traders. Ideally the post harvest moisture content of maize should be about 22%. Typically, this can be reached through field drying and open storage in top covered, side open cribs. If this level of dryness is not reached, a good chance exists that the product will rot even before it is sold at the farm gate. Since high levels of humidity prevail in Tanzania especially 27 during the rainy season, additional processing before storage is essential for preserving value, extending economic life and assuring quality control. Drying, cleaning and fumigating maize before storage prevents losses associated with insect infestation, rodent and pest infestation and spoilage caused by rot and mildew. Together these losses can exceed 30%. Pesticides including acetic super, Thiodin dust and Super 50 EC applied in field are extremely useful in reducing subsequent insect and pest infestation. However, inadequate incentives again exist within legacy chains to invest in these applications since the ultimate beneficiary is frequently not the party who must absorb the cost 2.1. f Map of Value Adding Activities which Take Place within Legacy Maize Supply Chains The schematic diagram below depicts the structure and organization of supply chains which support farm to market linkages in Tanzania. 28 The Tanzania Maize Value Chain Map Large/Medium Producers Small Producers I II III IV Agents Brokers Village Collector Village Posho Mills Large/Medium Size Traders SGR/Grain Reserves Wholesalers Disaster Fund/WFP Town Posho Mill Medium Medium Size Scale Small Millers Exporters Scale Exporter Supermarkets Border Retail Agents Kiosks Consumers Consumer Large Scale National Medium Scale Small Scale Importers Food Importers Importers Security **Rounded rectangle - Producers Oval - Traders; Rectangle - Market outlets; Hexagon - Final Consumers. 29 The remainder of this section presents a step by step assessment of activity costs for each of the value adding process steps in the schematic above. The value adding activities which are presented in this section correspond to actually costs incurred by participants in maize supply chains during the spring of 2004. Production: Production encompasses a number of discrete activities which are broken out and assessed separately in Section 2.2 which follows. To recap that sections bottom line: Farmer who use traditional cropping techniques can, if the rains are favorable, harvest 3 to 7 tons per hectare. On the other hand, a mechanized farmer may produce 2 to 3 times as much compared to a farmer using manual hoe- tilling techniques. Loading : Loading of maize is usually performed by casual laborers who are paid by the buyer of the product (Middleman). It should be acknowledged that loading requires that 90 km sacks of maize be moved from one farm gate to another as part of the buy and collect process step. However laborers employed in this activity typically also carry out the sack filling and offloading actities. These together costs are about 3 Tsh per kilogram . Transportation from the village to trading centers: The primary transport activity includes direct variable transport costs as well as overhead costs for drivers, laborers, and fuel. The overheads are included in our calculation since the product is rarely obtained from a single farm and hence some management overhead costs are absorbed in planning and scheduling. Distances from villages to trading centers tend to range from 10 to 20 km. A middleman/trader may buy grain at the gate of several farmers and move from one farm gate to another in collecting lots which are sufficiently large to justify the cost of transportation of trading centers. This circumstance applies more often than not when growing conditions are un-favorable and when the outputs of individual farms are very low. In these cases the middle man/ trader will incur a range of costs including: the village development levy, over heads and fuel, labor for loading and packing material-- if not provided by the farmer, himself. From a village like Igoma which is about 20 km away from the nearest trading centre, the following set of costs would be absorbed by a middleman/ trader. Primary Phase Transport and "Buy" Transaction Costs Activity Cost per kg from Igoma Development levy 10/= Labour loading offloading 3/= Sacks 3/= Overheads for people & fuel 10/= Total cost 26/= 30 Storage: Storage is the next stage in the typical maize supply chain. Storage is typically accompanied by other closely related activities such as taxation, off loading, unpacking, sorting and drying, packing as well as fumigation. Storage within trading centers is usually the responsibility of the primary buyer--the buyer who dealt with the farmer in the first transaction. Storage at trading centers is usually accomplished by renting space. Most trading centers in Tanzania are mixed with residential areas. Thus, most traders rent houses in which, they store their product before transporting it to major market areas. Rent for product storage ranges between 30,000/= to 50,000/ Tsh for a full Swahili house per month. This translates into 1000/= to 1600/= Tsh per day. Additional costs for fumigation add another 0.45/=Tsh per kilogram. Trading Center Transaction Costs Activity Cost/ kg in T shs Offloading/loading 3/= Spreading/drying 1.5/= Re-Packing/ 1.5/= Fumigation 0.45/= Storage 5/= Levy 5/= Sorting 1/= Total 17.45/= Transportation from trading centers to regional markets: Local third party transporters typically provide transportation from trading centers to the major markets within the country.. The most commonly engaged third parties are long haulers, who can carry between 10 and 20 tonnes, and shorter haul truckers whose three and four axel trucks can haul 7 to 10 tons. Significantly, some of the companies engaged in agro processing operate their own truck fleets which they use to transport their products as well as inputs into their processing. Examples include FIDA Hussein importers and Exporters Limited, AP & AP Limited and SAID SALIM BAKHRESA&CO. LTD. Storage in major market areas: As has been noted above Dar es Salaam is the principle agro market in Tanzania.. Storage not only requires sheltering maize from damp conditions, moisture, rodents, and pests but fumigating it and thus extending its economic life. The quality of maize and ,even more, the quality of the products derived from it are dramatically affected by pre-processing storage conditions. The average cost of warehousing maize is about 1/=Tsh per kilogram . Additional costs associated with fumigation add another 0.45/=Tsh . 31 Costs of storage in major market areas. Activity Cost/kg Offloading 5/= Municipal levy 1/= Storage 1/= Fumigation 0.45/= Total 7.45 Milling: Milling is the common terminology used for maize processing....the processes, for example, that transform maize into maize flour and maize meal. Both small scale and large-scale processors offer milling services. The flour derivative is referred to as "sembe" in Swahili. Sembe has 3 main grades, which include: low quality sembe that involves milling the whole maize seed (bean, endosperm and germ), medium sembe that involves milling the endosperm and super sembe which involves milling the germ. The best quality is super sembe though all are sold for food purposes. Large- scale producers mainly do super sembe milling. It should be noted that the small-scale millers, however, control approximately 95% of the milling market. Milling costs vary between small scale and large-scale processors. Costs cover milling, shortages due to impurities, packaging and overheads Average costs of processing Activity Cost per kilogram Milling 12/= Shortage due to impurities 5% 0.6/= Packing material 6/= Overhead/ administration 6/= Labour 5/= Total Cost 30/= Processed maize is then transported to retail markets. The main retail market for maize produce in Dar es Salaam is the Kariakoo market. Here retailers who buy milled product from either of the already mentioned wholesale produce markets incur the following costs: transport, packaging and market levy. 32 Costs incurred by retailers Item Cost per kilogram in Tsh Buying produce 200/= Loading& unloading 1/= Transport 3/= Levy 2.5/= Packing material 2.5/= Total costs 209/= . Retail prices: As already noted above prices tend to vary among different store types and formats. Ultimately prices are set by demand and supply balances for specific maize derived product categories within comparable retail formats. Thus for example, prices in chain stores, wholesale and retail outlets differ significantly as the table below demonstrates. Retail prices for milled products Retailer and wholesalers Selling per kilogram Kariakoo retail market 400/= for maize produce Rutham & Brothers wholesalers 600-800/= for super sembe Shop Rite 500/= as per estimates Also, because no integrated planning and/or scheduling takes place within legacy chains, the specialized resources required to support chain operations are systematically under utilized. What specialized resources exist are allocated based primarily on bid and ask offers to provide a specific services. In offering services bidders separate their service delivery systems from those of all other service providers and separate each service delivery event effectively from the prior and subsequent service delivery events. The most significant of these under utilized resources are transport resources, but others as well, including storage, fumigation, information and bagging resources, fall under the same category. These resources are typically in short supply either in one direction, or to or from a specific region or during the harvest season. Conversely, they are in surplus supply in the opposite direction or to and from adjacent regions or during off harvest periods. Thus, for example, the Tanzania-Zambian Railway continuously operates in a perpetual mode of equipment supply shortage and inadequate cargo hauling capacity. Sharing risks across commercial organization lines, for example, 33 between the railway and its largest maize shippers, is not an option for capacity expansion that has been seriously considered by government policy makers 2.2 Factors Effecting Supply Chain Efficiency This section deals with factors which most directly effect maize supply chain efficiency in Tanzania. The section is divided into seven subsections. These included: i) attributes of the maize produced in Tanzania; ii) trade policy; iii) trade facilitation; iv) transport cost and delivery time; v) market information; vi) forward and backward linkages to the from farms; vii) food aid; and viii) transport infrastructure. 2.2.a Key Attributes of Maize Produced in Tanzania Several attributes of maize affect directly the efficiency with which it moves through supply chains. Still other attributes affect the value of maize when it arrives into retail and wholesale markets at the end of the supply chains through which it passes. The ability or rather the inability of the supply chain to maintain the integrity of value chains through internal incentives and quality controls greatly effects the value which chains are capable of delivering. The attributes that most affect distribution efficiency have to do with density, losses associated with handling and transfer facility. The attributes which should most effect market value are moisture content, incidence of foreign matter, adulterants and toxic substances, size of kernels, color, and incidence of broken kernels. ``` The most important attribute in determining transport and distribution cost is density. The density of dried maize produced in Tanzania is approximately 720 kg per cu. m. by the time that it is transported long distance. Density is diminished when the percent of foreign material is high, as it is at the point of initial sale into village market centers where foreign matter can be as much as 10%. The value to weight ratio of dried maize as it moves via long haul transport is 100- 120Tsh/ Kilo. This is relatively low compared with other commodities which are transported long distance within Tanzania. This circumstance makes the maize supply chain extremely transport price elastic. The value to weight ratio directly effects the choice of transport mode though its effect on the trade off between opportunity costs related to transit time and transport cost. Low value commodities like maize move in the largest economic lots and via the least expensive mode of transport..... which in Tanzania is rail. Throughout Tanzania, maize is primarily bought and sold in 90 Kg sacks. However, as the study team discovered, strict standards for universal 90 Kg sacks are only loosely maintained with the consequence that a " seller's" 90 Kg sack can be quite a bit smaller than a " buyer's" 90 Kg sack. Custodial responsibilities for maintaining product integrity in every link in the chain is a significant issue in Tanzania. Neither third party carriers nor warehousemen offer fail safe asset protection. Consequently, neither a transport waybill nor a warehouse receipt is negotiable within the county. Not surprisingly, the incidence of loss and damage experienced by maize traders is relatively high. Approximately 3 % of maize inventory is lost in each transfer or handling. On average 5 or 6 handlings take place between farms and retail markets or mills. Thus, only 85 percent of the grain initially harvested and sold finds its way into the retail market 34 From these basic product attributes the following economic parameters can be determined: The inventory holding cost associated with storing a ton of maize for one month is 3000-4000Tsh. A 60 ton rail car can transport 800 sacks of maize before it cubes out. This lading weight corresponds to 792,000 Tsh/ car load as measured by the product's wholesale value. A 18 ton road trailer can transport 200 sacks of maize before it cubes out. This total lading weight corresponds to 264,000 Tsh/ trailers load. Given the differences in transport cost and holding costs associated with delayed delivery, a movement of maize would require more than 100 days longer in transit time via rail than via truck in order to justify the higher transport cost associated with truck transport. Hence for movements for which modal choices are available, rail is almost always the more economic option. Unfortunately, much of Tanzania's grain is grown a significant distance from railway lines. Thus, the choice of highway transport is more a mater of network necessity than of rational economic choice. Of course, the preferred mode for transporting maize is via water since it is always the lowest cost option. However, little coastal or, indeed interior water transport of maize takes place except for imported maize. The marine stowage factor for maize varies based on whether it is shipped in bulk or in bags. Maize is a relatively heavy grain. Hence its stowage factor is approximately 1.416, while the stowage factor for lighter grains, which include barley, oats and linseed, runs between 1.558 and 2.408. If the cargo is to be shipped in bags, these figures need to be increased by 10%. Within Tanzania grain millers do not, for the most part, recognize different quality categories for dried maize. Rather, particular consignments of the product are either accepted or rejected based upon physical inspection by buyers who embrace very different standards. Mini millers enforce no standards whatsoever and large scale millers enforce their own individual standards. Thus, for example, the buying standards enforced at Azam Bakhresa millers deal with moisture content, percent of broken kernels and percent of foreign matter and no other parameters. Generally accepted standards and third party inspection regimes apply primarily to the formal, cross border trade and not at all to the domestic trade. This is quite significant for poverty alleviation. Without embracing objective standards and without being able to transmit differential price incentives associated with a hierarchy of quality standards backwards through supply chains to farmers, farmers are not able to move up the value ladder from marginal production to more profitable production. In other words, they are stuck in a "low product quality " trap. For the purpose of human consumption, maize has a limited shelf life which is significantly determined by the moisture content of the grain itself. Maize dries from the inside out, so early efforts to protect and dry the product materially effect its subsequent shelf life. Maize with a high moisture content has a much shorter shelf life than maize with a low moisture content. Shelf life can be extended up to 18 months if maize is dried to 11-12 % and if it is fumigated every 6 months. The cost of holding a ton of maize in a well equipped warehouse in Dar is approximately 1500 Tsh/' month41 . At this rate the price of maize would have to increase by 60 plus % between the harvest season in July and the peak price period in March in order to justify third party storage costs. 2.2 b Trade Policy 41This cost includes the cost of fumigation 35 The volume of regional maize trade is constrained by current trade policy, as is forcefully argued in the most recent COMESA study of regional maize trade policies and practices.42 However, as the study points out regional policy is beginning to liberalize. Still it should be pointed out, Tanzania appears to be moving more slowly in the direction of liberalization than for its neighbors. The country's regulations effect trade both within the COMESA and EAC in two principle ways, through: i) the issuance of import and export permits, and ii) tariff and non tariff charges All traders must obtain a permit from the government before they can export maize from Tanzania. During both the 2002/2003 and the 2003/2004 crop seasons, however, the government banned maize exports completely because of internal shortages. Normally when no ban is in effect, a permit can be secured from the Food Security Department based in Dar es Salaam when exporting from the northern regions and when exporting from the south from the appropriate Regional Agriculture Department. This export permit takes the form of a letter to the customs department. The letter- permit represents the quantity of export which the trader is allowed and the timing during which the permit remains in effect which is typically one month. Stringent permitting and export bans discourage private sector investment in maize trading. Traders perceive a refusal to provide a permit and/or an absolute ban as a significant transaction risk in Tanzania. It is most costly when it happens after traders have entered into contractual commitments to deliver maize to foreign buyers. Imports are also regulated in Tanzania. A permit is required for all imports, as is prior registration of the importer. Permits must be secured and prior registration must be completed with the National Food Control Commission in Dar. Permits are allegedly issued in order to more closely monitor maize movements in and out of the country. Every permit request to the NFCC must be accompanied by a sample of the maize which is being imported. When the NZCC grants the import permit it is good only via a specific gateway and customs office. Consequently, imported maize must be routed vis this gateway even if it is not the most direct and lowest cost route. Both the administrative formality that surrounds the issuance and use of import permits and the routing conditions they impose are costly and time consuming and hence favors informal trading over formal trading. Tanzania's tariffs on maize are also the highest within the region. Thus, among the EAC countries, maize imports are subject to tariffs of 3%, 4% and 5% respectively in Kenya, Uganda and Tanzania . The MFN rate which Tanzania applies to maize exports from elsewhere in the region is 25%. Tanzania's exports of maize into the DRC, Malawi and Zimbabwi is duty free into and subjected to the MFN rate of 5% into Ethiopia and Zambia 2.2 c Trade Facilitation Other regulated aspects of maize trade impose significant additional costs on cross border transfers, as least within the formal trade sector. The primary trade facilitation obstacles are primarily three: i)quality and food safety standards; ii) phytosantiary requirements; and iii) customs clearance procedures. The EAC countries recently harmonized their maize quality standards and, indeed, adopted standards for two distinct qualities of maize. This represents a positive step forward because previously all three EAC countries used different and incommensurate 42Regional Maize Trade Policy Paper, COMESA, August 2003 36 standards as all of Tanzania's other maize trading partners continue to do. The new standards address the following criteria: I) moisture content; ii) foreign matter; iii) broken grains; iv) insect damage; v) incidence of rotten, diseased and discolored grains, vi) incidence of immature or shriveled kernels, and vii) packaging. Phytosanitary standards are intended to protect Tanzania's farm ecology from infection by pest and diseases which originate outside. Phytosanitary standards are enforced though certification by a an authorized Phytosanitary Agency in the exporting country prior to shipment . Tanzania's phytosanitary standards are unique within the region and require the exporter to secure a certificate that his maize is free from Erwinia Stewartii a bacterial wilt. In order to clear customs exporters of maize into Tanzania must produce ten documents upon arriving at a designated customs station. These include: i) the original invoice; ii) an import declaration form; iii) a pre-shipment inspection; iv) certificate of origin; v) the phytosanitary certificate; vi) the quality standards certificate; vii) a safety standards certificate; viii) an export permit; ix) an import permit; x) certificate from a licensed customs broker that documents have been lodged with customs. 2.2.d Transport and Delivery Cycles As we noted above maize production/ distribution is transport intensive in Tanzania in large part because the primary consumer market is a substantial distance from the primary production centers and also because net surplus and net deficit areas are scattered broadly throughout the country so that a large portion of the maize consumed is requires transportation. The centralized nature of the permitting process required for both import and export described above only add to this transport intensity. Not only is transport cost significant in the overall cost structure of marketed maize but it is equally significant in determining how risks are managed within supply chains, themselves. The terms of sale and the party designated as the beneficial owner of the maize cargo determines the incidence of risk within specific channels. This incidence in turn is determined by which chain participants are able to pre-pay freight charges and thus absorb the transaction risk. All of the maize which moves long distances in Tanzania move on the basis of pre-paid freight charges. This section adds temporal and spatial utility dimensions to the marketing and handling cost analysis which was discussed in the section above. It tracks these two dimensions through the same value chain which was laid out in the previous section. The production stage for maize is followed by the first transportation stage as produce moves from the farm gate to the nearest trading centre. In anticipation of transport the farmer packs his produce in sacks without necessarily sorting the maize thought in most cases he has dried it. He then sells his produce to a middleman at an agreed price. With the exception of the drying all of these other activities can be completed in less than one day. 37 Time from village to trading centre: Source: FIDA Hussein importers and Exporters Ltd. Upon taking ownership of maize at the trading centre middlemen see to it that the produce is offloaded, unpacked, spread for drying and sorting, packing and stitching, loading and storage. These activities require both time and other resources as shown in the table below. Time required to store maize at the Trading Centre Transportation arrangements from trading centers to major market areas differ depending on the mode of transport used. Truck transport is faster than rail transport and requires less reservation time for ordering equipment, making a shipment commitment, etc. A middleman or trader in maize produce will typically absorb the transport cost. He will incur different costs for rail or road transport and different modal choices imply different transit times for transporting produce to major market areas, as shown below. 38 Transport times via Road Transport from selected Regions to Dar es Salaam Source: Estimates provided by local transporters. Data obtained from small transporters based in Jangwani and from processing firms, 2003/04. Estimates assume the use of 10 ton trucks which are loaded to full capacity. If rail is used ­in lieu of highway transport---the costs incurred and time consumed will differ significantly from those associated with highway transport. Importantly, costs also vary with the size of the container which is used to transport maize. The unit cost for transport in a 40ft container is the same as the unit cost for a 20 ft container, in spite of the fact that the 20 container is almost as costly as the 40 ft container to move. Significantly as well, rail freight rates are quoted and collected in US dollars in Tanzania. Dollarizing freight charges is the easiest way to manage join line and interline rates and divisions of revenue. Transportation costs of maize by Rail: 40 ft container with capacity of carrying 30 tons fully utilized 39 Transportation costs for maize by Rail: 20 ft container with capacity of carrying 15 tons fully utilized. Source: TAZARA Dar es Salaam The overall farm to DAR movement cycle requires 1 to 2 days of transit time to complete and costs Tsh 105,000-120,000 Tsh per ton. The table below compares the price of maize at the farm gate and at the retail end of the chain ( in DAR) with the level of freight charges absorbed in moving the product. Note that cycle time and costs are both greater for rail than for truck. This is because the rail movement of maize involves a longer distance. The key takeaway point which this table clearly reveals is that transport costs account for a very large share of the delivered price in Dar. Transport costs are three or four times the farm gate price for maize. Farm Gate Price Dar Retail Price Rail Transport Road Transport Time to Market (days for Charges Charges rail and highway) 7.5-10,000 Tsh 30-40,000 Tsh 28-30,000 Tsh 13-15,000 Tsh 1-2 days by road 2-4 days by rail 2.2.e Modal Options Available to Maize Shippers Tanzania's two railways---the Tanzania Railway Corporation ( TRC) and the Tanzania- Zambia Railway ( TAZARA)---provide extensive access to the several major maize growing areas of the country as the map below indicates. Indeed, maize is one of the primary commodities that both railways originate. Pricing mechanisms used by the two railways differ significantly, with the TRC applying rates which vary from point to point depending on levels of truck competition and TAZARA applying rates which vary uniformly with distance--so called scale rates. TAZARA faces much less effective competition from trucks than does TRC b because of the degraded condition of the Central Highway Corridor which parallels it primary route. The map below characterizes the primary maize surplus areas in Tanzania ( in yellow) and the major maize defecit areas ( in red). In addition it identifies the primary WFP maize storage facilities ( red.triangles). WFP warehouses are located primarily in grain deficit areas but a few are also located in grain deficit areas. The relationship between Tanzania's extensive rail network and key elements of the underlying national market for maize are very close. It is the efficiency of railway operations and not the railway network itself that inhibits better market integration Maps below illustrate rail maize routing in TZ, regions of surplus & deficits & strategic reserves as well as WFP aid storage points 40 Maps Illustrating the Railway Routing in the Movement of Maize in, through and within Tanzania as well as Maize surplus and deficit regions. The two Tanzanian Railways are part of a larger regional network of railways which interconnect the entire regional market for maize. More than 90% of the formal, regional cross border trade in maize takes place via rail transport. As the map below demonstrates a North South Network of railways exists which links chronically grain surplus markets in Uganda with chronically grain deficit markets in Kenya and to more 41 recently grain deficit markets in Malawi and Zambia. Tanzania plays a keystone role in this regional network with reasonably good access both to sources of supply in Uganda and to markets in Kenya, Malawi, Zambia and more recently Zimbabwi. This regional network is made up of five state owned carriers: i) the Uganda Railways Corporation (URC), ii) the Kenyan Railway Corporation ( KRC) iii) the Tanzania Railway Corporation (TRC), iv) the Tanzania Zambia Railways (TAZARA), and the v) the Zambia railway corporation (ZRC). This network is supplemented by a car float operation across Lake.Victoria from Port Bell in Uganda to Mwanza in Tanzania. Another critical feature is the transshipment facility which exists at Kidatu in Tanzania. The TRC and the TAZARA railways operate with different track gauges. Consequently it is necessary to transfer cargoes from railway to the other and this manual transfer takes place at Kidatu. The routing for maize moving north to south over the entire length of this mult-line network is the following: i) Truck from all corners of Uganda to Kampala, ii)Kampala ­ Port bell ­ Uganda Railways Corporation-Mwanza ­ Tabora ­ Dodoma ­Dar Es Salaam ­ Kilosa ­ Kidatu ­ TRC, iii) Kidatu ­ Kapiripomshi ­ TAZARA, iv) Beyond Kapiripomshi ­ ZRC. At the present time four of the carriers which make up this route are being prepared for privatization, including the two Tanzanian railways. Significantly, a privatization transaction is being prepared between two of the carriers which will combine them under common private ownership or concession control--the URC and the KRC. However, the time required to prepare these transactions has been excessive, investment in the railways has diminished to a tiny trickle during the transaction preparation process and the market response to the first of the transactions offered for private control ( TAZARA) has been disappointing. During the interim period the capacity of all five railways to operate efficiently and to haul increased volumes of maize has progressively declined, as the number of " slow orders" has increased, the number of serviceable freight cars has declined, the number of operational locomotives has continued to fall off and the service provided by the carriers become more and more risky for shippers. At the same time, the share of maize which the five railways handle has become increasingly " local" as contrasted with " interline" in its orgin/ destination characteristics. This reflects a general inward focus among the carriers. Individual railways have attempted to rconserve their limited capital resources to first protect local line movements before allowing their equipment to move " off line." At the same time contractual and dispute resolution mechanisms which had applied to the buying and selling of car hire, locomotive services and maintenance services and to the sharing of revenues and interline pricing of through services and through prices have broken down. The net result is a transport network which is operating in a suboptimal Balkanized mode. Rather than facilitating regional maize trade it is frustrating the efforts of maize traders who must assume the coordination responsibilities and risks for interline movements on their own in lieu of the carriers. The interregional rail network could become the primary instrument for the integration of regional grain markets if several fundamental problems were corrected in the operation, marketing and pricing of railway services. These problems and issues include the following: i) the current lack of clear and enforceable carrier custodial liability for secure delivery of cargoes. A key obligation which carriers must assume and faithfully discharge as agents of the beneficial owner to freight is the custodial responsibility of loss coverage if the value of commodities delivered is less than the value of commodities accepted for transport. Common carriers need to operate as asset 42 managers for the purpose of completing buy-sell transactions. Rail carriers based in Tanzania do not assume this responsibility either on an interline or local basis today. ii) Interline pricing needs to be developed which relates more to opening new market niches than to competing with marginally reduced rates with truckers. Part of this new " market based" pricing philosophy needs to create partners out of shippers and allow shippers to assume some of the risks ( e.g revenue equipment supply) currently assumed by railways in return for incentive rates. iii) Joint line through services developed on a point to point basis need to become the standard and not the exception for regional rail service. All of the issues of equity and equal treatment that are a prerequisite to coordinated through train service need to be regulated at the regional level perhaps under EAC initially and under COMESA subsequently. Historically the primary emergency backup supply of maize for Tanzania has been South Africa which, indeed, serves as the supplier of last resort for the entire region. South Africa merchandisers hold large stocks of maize in high quality storage and are able to fill orders within a reasonably short period from Tanzania. Rapid discharge of bulk grain, rapid fill of bulk grain vessels in South African Ports and delivery to Dar requires a little as 8 days. Facilities and land side discharge capabilities exist with the Port of Dar to make emergency food relief from South Africa Tanzania's best guarantee against food shortages. The weakest links in this supply chain again is inland railway service from the Port of Dar to interior points which are food deficient. The limited car and locomotive supplies available to both TARZARA and to TRC have only exacerbated an already desperate situation. Clearly, the first priority for fixing the maize supply chain is first fixing the railway problem in Tanzania. Long haul truck capacity is not adequate to take the slack which declining car supply has created for the two railways. Low utilization of truck equipment is caused by poor highway infrastructure, lack of third and fourth party logistics management services to broker, direct and match supply and demanhy, too many imbalanced loads and too many one way hauls and long quequeing delays bewteen loads all tend to drive up unit transport costs. Low utilization of trucks and tractors like low utilization of railway equipment makes transport service provision extremely captial intensive in Tanzania. Moreover, very little cooperation has been institutionalized commercially between modes, so that the best operating aspects of rail, water and truck are not being combined on a case by case basis. Intermodal transportatation in Tanzania is dominated by ocean shipping oriented carriers and forwarders and not by transportation intermediaries who focus on the maize trade. 43 O CEON IN DIAN 2.2.f Storage Capacity for Maize Warehousemen perform several important functions within maize supply chains. Depending on which value adding agents they serve within the chain these services can have quite different economic consequences. Thus, for example, privately owned warehouses allow owners of stocks --which are properly dried and fumigated--- to postpone the sale of those stocks until a time during the annual production, harvesting merchandising cycle when demand exceeds supply and favorable prices can be realized for stock sales. In this way, they help to smooth out annual price peaks and troughs. Warehouses are the critical fixed asset which market arbitrageurs require in order to 44 balance supply and demand over time. Market arbitrageurs buy for their own accounts and lease or own warehousing capacity so that they will have a safe store for the inventories which they control. Third party warehousemen provide the same opportunity to farmers and small scale traders which are available to market arbitrageurs. Third party warehousemen provide storage, cleaning and fumigation services on a cost for service basis. They do buy maize for their own account but rather rent out their storage capacity to other chain participants. Privately owned warehouses also allow processors to buy critical grain inputs forward when prices are rising and hence they provide one way in which farm process dependent industries like milling can manage one of its primary risk in a volatile market like the one that exists in Tanzania. A corollary benefit is that storage capacity allows serial processes within the maize supply chain to be "de-coupled" from one another-- for examples, different milling and blending processes or the importation, inland transportation and local distribution of emergency food relief maize by the World Food Program. Storage capacities allow buffer inventory stocks to build up between process steps and thus allow their independent scheduling and programming.. This is important when serial processes cannot otherwise be precisely managed. Farmers provide for some storage capacity on their own farms. This capacity, however, is limited and not well designed for long term storage. After each harvest, farmers store t produce in either their houses or homestead. The storage facilities differ in size especially with production scale. However, small scale farmers usually have relatively small storage facilities of approximately 20-30m2. Storage is also vital in the trading centres. Most traders prefer to store produce upcountry both because it removes it from " competitive market view" and also because rental costs are much lower than in Dar and other larger cities. In the trading centers houses are rented by the traders, and these differ in size though are usually low-density plot houses and may provide 150-200m2 of storage. Storage at Kibaigwa trading centre 45 Large processors and traders also operate in the trading centers and upcountry towns. They have invested in go-downs and storage facilities specifically designed for maize storage, cleaning and fumigation in all principle producing regions. These storage facilities typically offer more than 1000m2 of space. The major wholesale markets include clusters of shore houses as well. A market the size of Tandale grain market which covers an approximate area of about 4 to 6 ha and has a throughput of 7000 tons per week and has a 2 to 3 week storage capacity. The traders store their produce in rented houses adjacent to the markets. However, most traders maintain more storage capacity and larger stocks up country were the cost of land, construction and developed real estate is much less. The large-scale traders/processors maintain large silos within their premises. Their storage and fumigation capacity is quite large as the table below suggests. Storage at Tandale: rented facilities The Ministry of Agriculture owns 15 silos which are controlled through the Food Security Department. These are strategically throughout the country both in chronically food surplus and food deficit parts of the country.. The silos operating under the Strategic Grain Reserve were established in 1977 and their use for various purposed directed by the Food Security Act No 10 of 1991. They were originally built with the goal of maintaining reserve stocks of up to 150,000 tons, which at the time was considered adequate to meet emergency food needs for three months--adequate time to negotiate and have delivered imported maize. However, the reserve was also been used as an instrument for capping prices by selling stored maize into tight markets. Since the 1990s the capacity of SGR to meet emergency needs has gradually eroded. Silos and warehouses under the control of the SGR have the rated capacity to store up to 241,000 tons. However, of this 35,500 tons of capacity have been leased out and SGR is currently using less than 25% of the remaining capacity. The fact is they lack of capacity and budget resources to have much of a significant impact in the local maize market. During 2001,2002 season, for example, they managed to procure only 50,448 tons which is less than eight days of consumption in Tanzania. 46 Total storage capacity Total storage capacity Primary Locations compared with annual (tons) sales (days) Integrated traders 50-60,000 tons 100 days Major agro trade market such as DAR, Mwanza and Dodoma Medium and small 5-10,000 tons 2-5 days Both in upcountry markets and Municipal scale traders markets Processors 20-25,000 tons 15 days Major producer regions such as Mbeya, Dodoma, Kgoma and in the major gro- markets such as DAR SGR 150,000 tons na Major producer areas and major agro- markets Third Party 10-15,000 tons 5-10 days Major agro-markets Warehousemen On farm 1-20,000 tons na Rural areas in the producer regions such as Igoma 2.2. g Availability and Timeliness of Market Information Market information services have gradually evolved in Tanzania from being a government dominated activity to one which is open to private sector participation. At the same time, the information requirements of private sector traders and processors have set the stage for innovative responses from third party service providers. The result has been a gradual improvement in the market relevance, timeliness and value of the information available to supply chain participants. This development in complicated by the ambiguity of the public vs. private nature of market information and the relative strength of a limited number of integrated traders in the Tanzania market, who use private market information systems effectively as source of competitive advantage. .For this reason alone rapid improvements needs to be made in opening the market for third party information services more broadly and for creating and disseminating new market information services to the public through some forms of public-private investment partnership and specifically to farmers and small scale traders. In 1970, a Marketing Development Bureau ( MDB) was set up under the Ministry of Agriculture. The FAO assisted with the initial organization of MDB. One of the objectives of the new bureau was to develop a system for monitoring both production costs and market prices throughout the country. Initially, under the Government's central planning policies, the role of the MDB was to disseminate and enforce official prices set by other agencies within the Ministry of Agriculture. Gradually, however, the role of the agency evolved into one of monitoring and reporting on market determined prices. In 1984, with the dawn of market liberalization the functions performed by MDB together with its organizational structure, the distribution channel that it used to disseminate information and the commodity market metrics which it tracked all began to evolve toward what they are today.. Eventually MDB's functions were transferred to the Ministry of Cooperatives and Markets from the Ministry of Agriculture and its name changed to Market Information Service Bureau ( MIS) . 47 The functions for which the MIS is currently responsible include conducting market surveys with wholesalers, retailers and farmers, analyzing market data and disseminating that data . Some of this data involve: i) commodity prices, ii) trade volumes, and iii) distribution and marketing costs at various stages of the supply chain. MIS disseminates the information which it collects through a number of different media channels, including principally newspapers and trade periodicals. In maize and other cereal markets, the Crop Monitoring and Early Warning System within the Food Security Unit of the Ministry of Agriculture complements the data collection activities of MIS. The CMEWS monitors and projects production, stocks and demand for maize. The information that it develops, however, is not broadly disseminated. The government supported MIS and CMEWS utilities, however, lag behind changes which are taking place rapidly in Tanzania's evolving maize markets. The timeliness and breath of dissemination of information services offered by these two government agencies do not serve the needs of supply chain participants as well as they need to be served. Indeed, an information divide has developed within the maize market in Tanzania between better informed buyers and less informed sellers. A clear need exists for the development and offering of market information services as a public supply chain management utility. Several promising developments are emerging to address this need. Foodnet--an NGO initially funded by USAID and DIFID--is striving to become financially self sufficient. Foodnet's primary business mission is the creation and dissemination of agricultural market information which creates value for both buyers and sellers. Foodnet is primarily an agricultural marketing service company. Its sources of information are diverse and its information quality control is very tight. Foodnet is deploying several strategies to expand the base of its operations from Uganda and Kenya into Tanzania. It applies the internet extensively to provide up to date and accurate market information. At the same time Foodnet continues to roll out a diversity of new market information products and in rolling out these new products it continues to test new business models and to team with a diversity of other network oriented agricultural information service providers. For example, Foodnet is working with an NGO---Radio works ­to create and distribute Radio Spots to local radio stations in Uganda, Kenya and soon Tanzania as well in multiple languages which deal with agricultural marketing and pricing. Foodnet has also teamed with local cellular phone service providers to offer a Text Messaging Service that provides up to date information on maize bid and ask prices. This cellular service originated in Uganda and is beginning to be offered in Kenya. Tanzania appears to be the next market opening for the Agribusiness TMS. In addition, Foodnet has networked with another donor sponsored enterprise--Trade Africa Biz.com--to develop a internet based electronic exchange for maize among other products. Footnet also works closely with the Kenya Agricultural Commodity Exchage, Ltd. ( KACE) which operates as an electronic brokerage for the purpose of bringing buyers and sellers to maize and other commodities together. What is beginning to emerge within East Africa is a cluster of IT based market information service providers who are alternatively competing and collaborating with one another, spinning off new information technology based services and working collectively at improving the integration of both markets and supply chains within the region. In this process, Uganda appears to be in the lead followed by Kenya with Tanzania rapidly catching up with the other two. 48 Radioworks Targeted Coverage Areas for 2004 2.2. f Linkages Between Value Adding Process Steps Redundant processing and simple buying, holding and reselling subtract rather than add value to distribution channels or supply chains when more efficient commercial activities are for various reasons unable to serve chains. Thus, for example, intermediate processors operate in lieu of efficient information service providers to discover prices and to bring sellers and buyers together. Similarly buyer/ resellers operate in lieu of financial institutions to supply working capital and to provide liquidity in long supply chains in lieu of credit provided by financial institutions which could do the same liquidity providing work more efficiently. Frequently, the opportunity to engage more efficient suppliers of essential services with distribution channels or supply chains depends of the ability of processors to link up with other participants and in the process to engage their special competencies directly, efficiently and under favorable terms. This section deals with demonstrated capacity of maize supply chains in Tanzania to create and support linkages with non traditional chain partners. Backward Linkages: Backward linkages--in the form of cross ownership, contractual affiliations, joint ownership of operating assets ( joint ventures) or common use of information systems- --between market centers, wholesale markets and large scale traders, on the one hand, and farmers, on the other hand, have not yet begun to form in Tanzania. As a result 49 risks associated with purchasing productivity improving inputs continue to be absorbed by the weakest link in the supply chain. With that said apparently several marketing groups are considering the possibility of backward linkages to farmers who are reliable and regular suppliers. Thus, for example, while preparing this report the study team members met with the board of directors of Tendale Market in Dar es Salaam. Tendale Market is the largest wholesale grain market in Tanzania. It is organized into a corporate structure with jointly owned equity which the board has invested in shares of utility stocks traded on the Dar es Salaam stock exchange. Their intention is to develop a sufficient equity base to allow the corporation to become bankable by pledging its shares as collateral against its borrowings. With respect to the boards investment intentions, they involve several projects all of which entail lowering transaction costs and strengthening backward linkages to farmers with whom traders in the Tendale market have developed long term relationships. One of the possibilities being considered is pre-paying for maize and other grains in well advance of their harvest. This would entail the provision of credit for input purchases from farmers who would subsequently payback their credit out of the proceeds of subsequent post harvest sales. Another possibility being considered is investment in near farm grain storage facilities which could be used to extend the shelf life of grains purchased in large quantities during the harvest. A variation on this theme is investment in third party storage and fumigation facilities which would operate on a fee for storage basis rather than as private store housemen who stored only already owned grains, in the mode of the integrated traders. Similar conversations which the study team had with the President of the Commercial Bankers Association of Tanzania and with one of the largest and most aggressive commercial banks in Dar pointed toward similar exploratory plans. Commercial bankers are considering various ways in which they might be able to make equity investments in joint ventures or stand along subsidiaries of bank holding companies whose value premise was the capitalization of farm to market supply chains or the provision of credit to existing chains which was secured against tradeable inventories. Their concerns with regard to the later possibility is the thin and volatile nature of secondary markets for maize and other grains in Tanzania. Both sets of potential investors appeared to welcome some form of technical assistance in articulating in business plans and testing pro formal financial statements with which to justify the financial feasibility of non- traditional business models that entailed strong backward linkages to grain farmers. Clearly developing business models and transaction structures which foster stronger backward linkages from bankers, processors and merchandisers to farmers is a practical and desirable set of initiatives which hold out the prospect of reducing poverty, transferring risk and improving the integration of traditional maize supply chains. Forward Linkages: Forward linkages from farmers to marketers, processors and merchandisers are even less conspicuous on the Tanzania development landscape than are backward linkages. This is most unfortunate because forward linkages are inherently pro-poor. Typically they advantage farmers who initiate them more than backward linkages which are initiated elsewhere in the chain and hence are biased in the relative advantages they provide to other chain stakeholders. 50 Unlike Uganda and Keya where significant levels of organizational experimentation with farm business models which entail strong forward market linkages are underway, little commercial testing appears to be underway in Tanzania. Forward linkages initiated at the farm level of the supply chain require as their foundation agro-industrial organizational structures which are larger and more substantial than small scale, family owned farms. They require a corporate or other legal, organizational base which creates sufficient economies and sufficient production scale to allow for the inclusion of specialized competencies in market management, finance, production management and risk management with the business model. Few of these foundation structures exist in Tanzania. Thus, by way of contrast, several notable farm level business models with forward integration being their most defining feature have been started up in Kenya.43 In 2000 for example, SACRED-Africa, an NGO, developed a " Cereal Bank" model to assist farmers in Western Kenya's Bungoma District . After two years of operation, this approach is being introduced to neighboring areas through the efforts of several NGO's--through Resource Projects Kenya , the Rural Outreach Program and the Sustainable Community-Oriented Development Project. Together, these individual projects constitute the "Smallholders Marketing Movement." Depending on its agro- ecological and socioeconomic conditions, each group has adopted slightly different business models. Cereal banking is a collective system for storage, bulking and marketing maize. Each local Cereal Bank is a storage and market center. Under the concept each farmers organization registers its members, elects officials and establish an account with a commercial bank. One of the NGO's mentioned above provides training to its sponsored Cereal Bank in post-harvest storage and processing and in quality control. In addition, the NGO assists its Cereal Bank is gaining access to credit so that each can acquire essential processing equipment and sufficient working capital to allow them to bulk and trade. The Smallholders' Network in Kenya has developed a reputation for high quality maize. The njectetwork sells most of its product year round to Unga Millers, Kenya's largest processor of maize meal. It also sells direct to local schools and other public institutions. The benefits of forward linkages to both farmers and their supply chain partners are many. They can result, for example, in better quality products and superior production and distribution planning. They offer win-win outcomes to parties at both end of the chain. Farmers margins increase their gross margins and processors/ consumers pay lower prices.44 They allow for risks to be transferred within the chain from the weakest link to the link best positioned to either managed the risk or to sell the risk to a third party insurer or risk arbitrageur. They reduce transaction costs by assuring that only value adding activities are retained within the chain and that redundant processes, redundant inventories and uncovered risks are removed from the chain. Importantly, efficient forward linkages also facilitate the efficient distribution of information concerning demand levels, product preferences, production technology and competition within the chain. Finally, they greatly improve chain adaptability and hence chain competitiveness. 43 See Woomer,P.L.,Omare, M.N. and Mukhwana, E.J.2003. The operations of rural self-help groups. pp.131-145. In: C.E.N.Savala, M.N.Omare and P.L. Woomer (eds.) Organic Resource Management in Kenya: Perspectives and Guidelines. The Forum for Organic Resource Management and Agricultural Technologies, Nairobi. 44 See "{Cereal Banks in Kenya" by 51 Moreover, unlike the circumstances which prevail in the fish and sugar sectors no legitimizing regulatory basis exists for building out forward linkages in the maize supply chain. In the case of sugar and fish a regulatory presumption exists that the industry will develop vertically and that some minimum set of regulatory safeguards need to be put in place to protect equities among affiliating parties. Thus, the Sugar Act facilitates collective bargaining between small holder farmers who supply sugar care and large scale processors. The presumption implicit in the legislation is that fair but productive inter-sector equity determinations will be assures through regulatory intervention. Thus, the regulation sets the stage for investment both on the part of small holder regional cooperatives and the sugar processors with whom they deal contractually. No such regulatory framework exists for the maize supply chain and, not suprisingly little innovation forward integration is conspicuous in that supply chain. 2.2. g Effects of Food Aid Both government and non-government agencies intervene in Tanzanian grain markets and to a lesser extend in farm input markets in pursuit of eleemosynary objectives and in the process of pursuing these objectives indirectly effect the efficiency with which markets operate. Two sets of agencies are particularly active in the Tanzania maize market--the Strategic Grain Reserve ( SGR) and the World Food Program. Together these two agencies have purchased approximately ____ tons of grain per year in Tanzania over the past three years and over the same period provided ____ tons of grain into the national market. The disruptive effects which these two agencies have on underlying markets can be parsed into four distinct categories: i) the timing, location and scale of their interventions undercut market clearing processes and distort the price at which markets would clear if the two agencies where not active. This disruption is only amplified to the extent that information about planned interventions are more and earlier available to some market participants than to others; ii) the agencies have first call on essential resources which third parties require to supply services to maize supply chains, including, for example, rail cars and maize storage facilities. Accordingly, they impose additional costs over and above the marginal cost of providing these services on a market clearing bid ask basis to private sector buyers and sellers. They effectively bid up the price of operating supply chains in Tanzania and undercut incentives for planning or allocating the use of limited resources based on both vertical and horizontal process integration. iii) efforts to isolate beneficiaries and to prevent beneficiaries from selling maize delivered as food maize back into open markets is only partially successful. It is interesting to note that a portion of the maize distributed as food aid finds its way back into channels which resell these stocks back to WFP. The net result of this round tripping is a diminution of demand for maize which private suppliers might otherwise supply and a substantial reduction is the beneficial results realized from the use of aid resources. This result would be diminished if all of the maize furnished to beneficiaries were produced in Tanzania or even within the region. However, over the past three years fully ____% of the food aid distributed within Tanzania has originated overseas and outside the region. iv) Participants in the specialized local market niche which the two primary aid agencies have developed are selected and pre- qualified based on ad hoc " procurement" criteria which these two agencies have developed. Frequently, the WFP as wells requires that the maize which it procures be delivered upcountry, be stored at bidders expense and/or be processed by bidders. The net effect of these requirements is to advantage large scale and vertically integrated traders over others, including new market entrants. Pre-qualification of bidders and 52 criteria for bid participation such as the prerequisite that offers hold sufficiently large stocks before they bid to cover their entire bid amount, tend to fence out new entrants from the food aid market niche and create a competitive advantage that spills over into other market niches as well based on established incumbency, market power and fully integrated operations which include merchandising, storage, transport and processing. The net result of these four factors is to make entry and operation within the Tanzanian maize market much less competitive than it would otherwise be. The food aid supply situation in the country is heavily influenced by transport capacity, storage capacity and market information. In areas which are not affected by transport capacity limitations, which have adequate storage capacity and where competitive sources of maize from outside the country are available, markets function relatively efficiently and appear to stabilize prices effectively. Thus, for example, in Dar es Salaam, Morogoro, and Arusha food supply/demand balance remains reasonably stable from year to year. Even during periods of virtual shortage in other parts of the country, prices remain stable in these three markets. Indeed, in these three efficient market centers wholesale prices for remained below international (CIF) prices. In contrast, in areas that suffered a poor harvest last year and which were isolated from the national market because of transport limitations, not only did food shortages occur but also prices spiked sharply as well in March and April above the inernationa CIF price????. The efficient management of food aid is highly constrained by Tanzania's inadequate internal transport system. Access to rural areas remains particularly difficult, with only an estimated 10 percent of roads paved into chronically food short areas. The problem of poor road infrastructure is compounded by the limited availability of rail wagons, the physically depreciated condition of core rail infrastructure such as rail passing sidings and repair shops and the almost complete lack of good feeder roads with which to link together road and rail transport. The ability of transport providers to offer integrated multimodal services that could support maize movements is limited, moreover, both by policy and practice which has a single modal orientation. 2.2. h Effects of Limited Transport Infrastructure Overall food aid appears to be most disruptive in areas of chronic short supply where policies for self help and for market sustainability are most needed. Markets in isolated parts of Tanzania, such as____ and _____, have increasingly become disconnected from other markets, with local prices reflecting local supply conditions only. Due to limited accessibility to Dar or other distribution centers effective temporal and spatial arbitrage cannot take place and prices in these areas do not reflect overall national or regional supply scarcity. In other parts of Tanzania maize may be available but simply cannot be transported to areas of immediate need. These are not so much issues which require emergency food relief as a solution but which rather require better market integration, more competitive supply chains and lower transaction costs. A distinct effect of recently implemented government and donor policies has been that on prices in areas of surplus have fallen and have correspondingly increased in deficit areas for contrary reasons. 45 At the beginning of the second quarter of 2004 the emergency food situation appeared to be particularly dire for Tanzania. As noted above, most of the stock in Tanzania is controlled by large traders and to a lesser extent by the Strategic Grain Reserve (SGR). Against available stocks at the end of the first quarter of 2004, demand for the 45Based on_____ 53 remainder of the marketing year, including food, seed, feed, waste and closing strategic stocks requirements are estimated at 909 000 tons. In order to meet all these needs, the country has an estimated import requirement of approximately 561 000 tons. Of this total commercial imports already contracted on 3/31/04 are estimated at 75 000 tons and food aid in the pipeline accounts for an additional 3 000 tons. This leaves the country with an uncovered import requirement of 483 000 tons, of which 20 000 will be requested as further food assistance through WFP. In the meantime the Government has released 10 000 tons through the SGR in relief assistance. The supply of maize in Tanzania is highly influenced by weather conditions. Thus, on the pattern of rainfall for last year maize production was estimated at only 228 000 tons, approximately 60 percent less than in 2002 which was a good production year and fully 40 percent below the long term average. Taking into account estimated stocks of 120 000 tons, available at the beginning of February, the domestic availability of maize for the remaining four months of the marketing year 04 is projected at 348 000 tons.46This corresponds to ___months consumption. The tables below compare the level of infrastructure development within Tanzania with the level of its development among other African countries. Note that the use of rail infrastructure has been increasing in Tanzania in a trend which is quite opposite that of all other countries in the region. This increase in goods movements via rail has a great deal to do with the entrepreneurship demonstrated by the managements of TAZARA and to a lesser extent the TRC. Neither carrier has enjoyed any significant infusion of new capital during the 1990's . However both managements have succeeded in producing more rail traffic with fewer resources. That trend, however, has reversed itself over the past two years as neither carrier has been able to continue to increase ton km's hauled without any capital infusions. The story for highway transport is not so different. It should be noted that Tanzania's highway infrastructure, particularly in the west; north and far south is in a dire state of repair.. It condition is so severely deteriorated that farmers in the south find it easier and less costly to engage in cross border trade rather than to deliver their maize into areas of scarcity within the country. Poor roadway infrastructure partly facilitates the informal trade. As the below tables confirm, Tanzania lags behind its neighbors in road way construction. The percent of paved roads in Tanzania is extremely low, well below the level of the other African countries with which it is compared in the Table. The absolute size of the road network, however, compares more favorably. It is roughly equal to that of that of Ghana and Kenya and significantly greater than that of Uganda when road network to population ratios are compared. Significantly, the number of road vehicles per 1000 population for Tanzania is significantly less ( 5 units/ 1000 population) than either Kenya ( 13) or Ghana ( 8). On the basis of how fully utilized its roadnetwork operates Tanzania is slightly higher than Uganda ( 4 units/ 1000 population) 46Source WFP and FAO. See______ 54 Rail Goods to GDP 500000 450000 400000 350000 Ghana 300000 Cote d' Ivoire 250000 Kenya 200000 South Africa 150000 Uganda 100000 Tanzania 50000 0 75-84 85-89 90-MR Road Network 1000km/million persons 8 7 6 Ghana 5 Cote d' Ivoire 4 Kenya 3 South Africa Uganda 2 Tanzania 1 0 75-84 85-89 90-MR 55 Paved Primary Roads Percent 30 25 20 Ghana Cote d' Ivoire 15 Kenya South Africa 10 Uganda Tanzania 5 0 1995 1997 1999 WB77328 C:\Hussein2.2.doc May 15, 2004 12:17 PM 2. 3 Factors which Effect the Competitiveness of Tanzania Maize Producers Farm production is a the most important process step in the maize supply chain. This section deals with farm production. It is organized into two subsections, one of which describes Maize production methods in Tanzania and one of which compares production productivity in Tanzania with that of other countries in the EAC. 2.3. a Maize Production in Tanzania Of Tanzania's total area of 94.3 million ha, 5.1% is used for crop cultivation and of this arable land, 0.5 % is used to cultivate maize. As noted, small-scale farmers, usually peasants with farms varying in size from 0.5 acres to 8 acres, produce almost 95 % of Tanzania's total production. About 70 percent of Tanzania's total maize crop area is cultivated by hand hoe, 20 percent by ox plough and the remaining 10 percent by tractor. Maize is grown almost exclusively under rain fed conditions in Tanzania. It grows from sea level to 2,400 meters above sea level. Production technology varies greatly with the agro-ecology, cultural background, resource availability, and environmental stresses but is generally traditional. Traditional cultivation refers to the following manual farming methods: Ground breaking and furrow forming done with a hand held hoe; Planting and weeding are manually done or done with the help of an ox plough; Fertilizer and insecticide applications are done with the help of a manual sprayer . Harvesting is also done manually. Traditional cultivation results in low productivity in most agro-ecological zones except in the transitional highlands where conditions for growing maize are excellent. 56 Output to a large extent is dependent primarily on two factors: i) the amount of rainfall in a any given growing season and ii) the natural fertility of soils. Very little fertilizer is applied to maize growing lands in Tanzania, less than 50,000mt annually.47 Where maize is intercropped ( and maize is often intercropped in Tanzania) it is planted at a spacing of 120 cm x 60 cm, which translates into a plant population density of 27.777 plants per hectare. If maize is not properly spaced the number of cobs per plant and the size of the plant itself will be diminished. The principal maize producing regions are located in the southern highlands of Tanzania near the Zambian, Malawi and DRC borders and in the northern and central highlands near the Kenyan border. As noted below the significant distances that exist between Tanzania's major production and consumption centers make the maize supply chain in the country relatively transport intensive. See the map below.48 Mbeya, Rukwa, Ruvuma and Iringa located in the southern highlands are the nation's primary grain production areas. The so-called "big four" accounted for fully 40.3%(810.2MT) and 34.6% (933.6 MT) of the total maize produced in the country respectively in 1999/00 and 2000/01. The next largest production areas are the central and northern highland regions which together accounted for 53% of total maize production in 2001. 47In Kenya total fertilizer consumption is five times that of Tanzania on maize growing areas which are comparable is size in both countries: 1,500,000 ha in Kenya and 1,564,000 ha in Tanzania. On average only 10% of the maize land under cultivation in Tanzania is fertilized. On these fertilized lands 80 kg/.ha of Nitrogen and 40 kg/ha of P2O5 based fertilizer is applied. In Kenya 30% of maize land under cultivation is fertilized and 40 kg/.ha of Nitrogen and 30kg/ha of P2O5 are applied to the fertilized lands. The corresponding percentages for Zimbabwe are 61% and 153 kg/.ha and 22 kg/.ha respectively. 48Map was based on FEW Net market analysis and was represented in the "Regional Maize Trade Study" undertaken by RATES for COMESA, August 2003, p. 10 57 Occasionally from Ethiop To Southern Sudan N From DRC MaizeProduction ategories To DRC High Medium L To Zambia and Malaw As noted above, maize is rain feed in Tanzania. Rain fall can vary significantly from year to year among various traditional maize growing areas within the country.. For robust growth, Maize requires 800 to 1000 mm per annum. Rainfall of less than 800 mm stunts growth and substantially reduces yields. Less than normal rainfall in the south of Tanzania explains why the harvest year of 2003/4 has been so spare.. Increasingly, advanced forecasting of weather patterns allows maize growers to anticipate future production. Converting this knowledge into useful and actionable form and disseminating it broadly are critical needs. Importantly, productive maize cropping also requires timely weeding and spraying of pesticides and fungicides. For example, a need exists for specific pesticides and fungicides to combat specific insect pests and plant diseases which affect maize. Some of the agricultural chemicals that have proved successful in combating specific diseases and specific pest infestations include: Pirimiphos Methyl, Permethrine and Dimethoate. Insecticide application is crucial not only for maximizing post harvest yield but also for assuring a marketable appearance. Sprouting maize crop on a cultivated farm 58 As demonstrated in Appendix 2.1 prices fluctuate sharply from one year to another due to natural factors such as rainfall availability, outbreaks of plant diseases and infestation with specific pests----all of which greatly affect domestic output levels. Typically, the lower the out put in a given season the higher the selling price at the end of the year and vice versa. Contrary to the prevailing trend in this crop year ( 2003/04) national production is off sharply due to the fact that the southern highlands were badly hit by drought. A rural farmer with a sample of maize ready for harvest The remainder of this section describes the set of on farm activities that collectively make up the production step in the supply chain. The stylized description which follows is based on a operations which could be found on a 5-hectare farm located in Mbeya Distict in the villages of Igoma or Inyalla. This is a maize growing area. The stylized description further assumes that the farmer has no access to modern farm equipment and that he tills his soil, plants, weeds and harvests using traditional labor intensive methods. The farmer will typically begin the production cycle by cultivating between October and November. Preparing the soil requires approximately a month to complete and is facilitated when the early rains arrive on their normal schedule. The farmer proceeds with planting his seeds during the months of December to January. Manually seedling a 5- 59 hectare plot requires about 1 month. The underlying assumption is that the farmer has stored his seeds and thus does not have to incur any out of pocket costs to buy new seed. Again this is typical of prevailing practice in rural Tanzania. Weeding as soon as planting is complete and continues between February and April. Again our assumption is that the farmer uses his sweat, his hoe and possibly a rake to weed . The application of fertilizers follows. The farmer may apply either or both traditional fertilizer such as animal waste and compost and industrial fertilizers, such as NPK, DUP, UREA, SA, CAN and TSP. The application of fertilizers requires approximately one month and the application is usually done in March. Application of pesticides is another important productivity enhancing activity. Pesticides, if they are applied are applied at all, are typically applied during February and March. An application of pesticides also requires a month to complete. Harvesting usually begins in June or July. Once harvesting is complete, the soil is rested or fallowed to regain its production capability before the next farming cycle begins again.. If our hypothetical farmer had access to mechanized equipment, which would be marginally justified on a 5 hectare farm, the costs he incurs differ from those of a farmer who applies traditional methods. However the time frame for each of the key process steps from cultivation through harvesting remain the same. The tables which follow review the activity costs and cycle times associated with each of the processes that make up the production stage of the supply chain. Separate tables are presented for traditional and mechanized farming.. Average cost and cycle times of producing maize per acre (in southern high land Mbeya, Rukwa, Iringa and Songea) using non mechanized technology. Activity Implements involved Cost per acre in Tsh Time in days/months Cultivating -Hand hoe -2000/= October-November. Planting Seeds December- January. Weeding Hand hoe and Rake 3000/= February-April. Applying fertilizers Fertilizers 5,000/= March. Applying pesticides Pesticides February and march Cost of harvesting Labour paid in kind if 5,000/= June- July. monetized Cost 13,000/= Approximately 10 month Note that the out of pocket cost differences between traditional and mechanized farming are quite significant--six fold---the gains in production yield are less significant--- two- three fold Average cost and cycle times of producing maize per acre (in southern high land Mbeya, Rukwa, Iringa and Songea) using mechanized technology 60 Characteristics of smallholders conditioned by poverty those and empowered to produce and market farm surpluses Farm attribute Conditioned by poverty Effectively empowered Farming system Subsistence food crops Mixed-enterprise agriculture Crop varieties Reliance on traditional species Seed of improved varieties routinely and cultivars purchased Cropping system Continuous cropping of staple Complex rotations to exploit market food intercrops opportunities Nutrient No external inputs or Integrated use of organic and management occasional natural fallow, mineral resources leading to leading to nutrient depletion improved soils Livestock Lacks livestock or owns few Confines improved livestock and management unimproved livestock markets products and offspring Composting Does not produce compost Produces and regularly applies fortified compost Group Does not belong to, or is Active participation in cooperative or membership passive within a farmers group marketing group Crop and market Operate with little or no outside Base farming decision upon current information information and reliable information 2.3.b Regional Production Cost Comparisons As the East African regional market for maize becomes increasingly integrated and as cross border supply chain costs begin to fall, production cost advantages among primary producers within the region may increasingly affect local prices, product flows and the differential levels of compensation realized by farmers. As noted above some parts of Tanzania are better integrated into national and regional markets than others. With that said, regional trade arrangements such as the PTA and the rejuvenated East African Community agreement are likely to accelerate the integration of regional markets for maize for the very reasons that very large disparities exist within the region with regard to production cost. One case in point is the growing informal market in cross border trade for both maize and beans between Tanzania and Kenya, on the one hand, and Tanzania and Malawi and the DRC, on the other. 61 A benchmarking analysis within the region reveals that Ugandan farmers produce maize at the lowest cost per hectare but at a medium cost per kilogram because of the low and declining yields which they realize. The table below compares the cost of production for Kenyan large scale, Kenyan small scale, Ugandan and Tanzanian producers. In terms of the industrial organization of farm enterprises and scale of farm production, Uganda and Tanzania quite similar. Small holders predominate in both economies. The table reveals that producers in the Kenyan large-scale sector have a significant comparative disadvantage in the regional market. The marketability of their maize derives completely from advantages which they enjoy in lower transport costs into the local Nairobi market. Indeed, in terms of economic return and capital productivity the sustainability of large scale maize cultivation in Kenya is highly suspect. As the table reveals, Kenyan growers--especially large scale growers-- realize financial DRC ratios of 0.52 even on sales into the Nairobi market.49 The corresponding DRC ratio for Uganda producer sales into the Kampala market is 1.72 and for Tanzanian producer sales into the Dar es Salaam market 1.27 and for sales into the international market 12.45. Smallholder produced maize in Kenya is much more competitive than large scale farm produced maize. However, smallholder production in Kenya is largely subsistence based. Little surplus is realized is realized by small holders in that country and what is produced is generally sold locally. Only insignificant quantities are left over for export. It is large-scale producers in Kenya who sometimes produce surpluses for export in good years although they seem to have little comparative advantage in production cost. With that said, in the few years in which Kenyan commercial farmers have produced bumper crops, small scale farmers in Uganda and to a lesser extent Tanzania have faced sharp reductions in local prices. Both sets of farmers have become increasingly dependent on Kenyan exports. Tanzanian maize is clearly the most competitive source in the local market of Dar Es Salam. Only production capacity, storage and production timing--all discussed above---prevent local farmers from completely dominating domestic markets. In cross border market, however, Tanzanian producers realize a DRC ratio of 0.95. In this instance, cross border border markets refers to Zambia, Malawi and Burundi. Tanzania producers enjoy a significant competitive advantage in accessing these markets vis a vis producers in either Uganda or Kenya because of the additional transport costs that these producer would need to absorb. Still more detailed comparisons of production costs are revealing. With regards to the cost per hectare, Kenyan large scale producers have the highest cost in the region. They spend more than $US 750 per hectare and are followed by smallholder Kenya producers who spend about half of that amount per hectare. Next in order follow Ugandan smallholders with their low input technology. Their cost per hectare is $ US 242 . Tanzania producers are the cheapest in the region with production costs of $ US 239 per hectare. However, when costs are calculated on a per kg basis the picture changes. Per kg costs of production differs from cost per hectare because of differences in product yield. These are quite different accoss the region. When analyzed in terms of production cost per kilogram large scale farmers in Kenya vis with smallholder farmers 49Based on a regional study of production efficiencies by Shetty, et al 62 in Uganda for the dubious distinction of cost producers, at USD 0.12 per kg. When calculated on a per kilo basis, costs of production are quite close for small scale Kenyan and Tanzania producers. With Kenyan producers being slightly higher at $ US 0.08 per kg for pure stand cultivation and $ US 0.07 for intercropped maize. Comparable costs in Tanzania are estimated to be $ US 0.05 kg. Moreover, maize production in Uganda is more labor intensive than maize production in any of the other regional production venues at 4,392 man-hours per hectare. The Uganda production system in less labor intensive at 3,048 man-hours per hectare for low input technology and 2,568 man-hours per hectare of high input technology. Kenyan maize producers are also much more productive with 2,028 hours per hectare for smallholders and only 56 man-days ( 560 hours) for the large-scale farmers. Much more so than their neighbors Kenyan commercial farmers operate on a large scale and apply a lot of machinery services in their farm operations. In some cases, Kenyan smallholders also hire in the services of farm machinery suppliers, especially for land preparation. This practice is not common in either Tanzania or Uganda among smallholders. Rather methods such as the use of Oxen for land preparation are more common among smallholders in Tanzania. A regional benchmark comparison of key input productivity (All prices are per hectare in (US $)) Maize Cost of Production: A Regional Comparison 63 NB:With regards to the ranking,(1) indicate lowest cost producer and (5) the highest cost producers. LS-large scale, SS-smallscale U1-aasumption regarding input use(land clearance necessary, pesticides are used and farmers must pay some interest on borrowed capital. U2 are the reverse assumptions.(Source BOU 1993) 2.4 Roles Played by Government, Donors and Third Party Service Providers The efficiency and adaptability of maize supply chains anchored in Tanzania is significantly influenced by the support of several critical agencies including those of government, donors, private sector trade self help organizations and third party providers of essential services. This section deals with external support systems that buttress maize supply chains more or less well in Tanzania. 2.4.a Government Role Government interventions has several direct and indirect effects on maize supply chain participants. The most apparent and overt is in the taxes which government at several levels remove from supply chains and the incentives for various kinds of supply adaptive behavior which they create in the process of imposing and collecting these taxes. The second major impact which governments have on supply chain 64 development is in the setting and circumstances which fosters backward and forward linkages within farm centered supply chains. The third effect has to do with the quality of the basic infrastructure which government providers for both supply chain participants and for the third party service providers who support them. The fourth major effect that derives from government is in the active development and dissemination of best farm and farm marketing practice through government sponsored agricultural extension services. The greater the extent to which extension services develop competencies and facilitate the dissemination of knowledge which leads to vertical linkages and efficient process integration across commercial boundaries within maize supply chains the greater the efficiency of local chains. The fifth aspect of government role is in removing market barriers and facilitating trade. The remainder of this section deals with these four issues. Taxes The total tax bill paid by supply chain participants in Tanzania accounts for fully 30% of the farm gate price of maize as demonstrated in the supply chain analysis contained in Appendix 2.2. . Most of the taxes imposed on maize supply chain participants in Tanzania are transaction or ad valorum based. Tax liability begins with the first transaction is the chain. And, it continues with each subsequent transaction. Village executive officers or district officials have the right to impose taxes and cesses on maize producers and/or maize traders. One form of tax which is ubiquitous is the District/ Village development levy. The tax is withheld from the net amount which traders pay to farmers for their maize. Proceeds from this levy are used to fund infrastructure in the rural areas as well as provision of extension services. Farmers generally feel they are receiving little value for their tax contributions. Based on interviews conducted in Igoma and Kibaigwa the levy for maize per kilogram appears to range between 7.50/= and 10/= Tsh. Because this is a cash cost of production and one which is payable upon the sale of maize it imposes a substantial penalty on farmers. Other taxes to which maize traders and processors are exposed I[tream in the chain include the following: · District levies · Municipal levies · VAT Commercial Linkages A Ministry of Agricultural Cooperatives and Marketing operates within government and the strategic mission of this Ministry is fully revealed in its title. Its role is the incubation, development and support of farm based cooperates. The Government of Tanzania is primarily concerned with the development of one type of organizational structure at the farm level and much less interested in other non traditional farm based business models, contractual affiliations, marketing joint ventures, vertical enterprises based on cross ownership or even virtual enterprises whose basis for coherence is shared information systems or Internet interconnections. This monochromatic vision which is fixed on one form of farm based industrial organization may over the long run have a significant adverse effect on farm production, on farm to market transaction costs and on the market adaptability of marginal farm production units. Diversity in 65 business models just like diversification in niche markets and in maize products is a positive development which should be encouraged. Supply chain development is all about competition based on superior business model design. Unlike the sugar and fish sectors no regulatory authority exists within the maize sector to foster backward and forward linkages within farm centered supply chains and/or to resolve disputes when these arise among chain partners. Thus, no piece of legislations comparable to the Sugar Act exists with respect to maize to assist with the development of collective bargaining agreements which facilitated vertical integration. Moreover, little pro-active effort has been expended on the part of either donors or government to test and develop innovative farm based business models with are vertically aligned with larger supply chains. Provision of Infrastructure As noted above little progress appears to have been made in improving the roadway system in Tanzania on in expanding that system into major maize production areas. Likewise the railway sector has been allowed to slide backwards through underinvestment. Railways have approximately a fifteen year economic life and both the TANZARA and the TRC are approaching the end of their economic lives.50 Extension Services The extension services provided by the Ministry of Agriculture are primarily production oriented as contrasted with market oriented. They assist farmers assist only indirectly with developing skills and competencies associated with supply chain management. They have little input either in terms of substantive content or in terms of needs to fill specific weak skills categories with the private sector. Trade Policy and Trade Facilitation The COMESA and EAC trade regimes have charted a policy course under which maize markets can be progressively liberalized and integrated at the regional level. According to this regimes, member countries must committed themselves to charging preferential tariffs for maize which originates within the region. However, Tanzania MFN tariffs and non tariff barriers remain relatively high by regional standards. Its MFN for maize is 25%. Moreover, the embargoes which Tanzania has imposed on its own maize exports over the past three years have effectively removed Tanzanian producers and small scale traders from the regional market, at least formally. Within COMESA, nine countries, which have already ratified the Free Trade Area (FTA) protocol. They are levying zero duty on goods that originate and terminate within the region. These countries include: Kenya, Malawi, Zambia, Zimbabwe, Egypt, Djibouti, Madagascar, Mauritius and Sudan. However, Tanzania is not a participant in COMESA. Under the EAC trade regime, in which Tanzania does participate, Kenya grants market access to commodities imported from Uganda and Tanzania with a 90% tax reduction below its MFN rate. Tanzania and Uganda on the other hand grant an 80% tariff reduction on goods originating from Kenya. EAC member countries sign the EAC Customs Union Protocol in November 2003. Envisaged under this protocol is a 50Best Methods of Railway Restructuring and Privatization, Ron Kopicki and Louis S. Thompson, World Bank, 1997. 66 progressive tariff reduction to zero on goods sourced from within the EAC. A common external tariff on goods from non EAC countries is also envisaged. Imposition of arbitrary measures, such as import restrictions to prevent market access, is prohibited under the COMESA trade regime, unless such an action is approved by the council of ministers under the safeguard clause of the Treaty. In EAC a similar provision is contained in the EAC Customs Union Protocol. However, this provision remains to be fire tested in actual practice. Currently trader who sell into Tanzania are required to secure an import permit from the Food Security Department prior to shipping. During the study it emerged it is only the largest traders and the integrated trader/ processors who can secure these permits and fully comply with the law. An exporter of maize from Tanzania requires an export permit from the Food Security Department. However, as we noted above the granting of export permits has been mute for the past three years during which the government has embargoed maize exports because of internal grain shortages. Non-Tariff Charges: An additional array of non-tariff charges must be absorbed by importers and/or exporters of maize. · Port Wharfage Fees: The THA (Tanzania Harbor Authority) requires a fee of 1.5% of CIF value for services provided to the ship while docked or leaving the port. In addition, importers/ exporters are liable for tally fees of US$1.00 per ton payable to the shipping agencies. · Pre-inspection charges: At the port of entry, pre-inspection is mandatory for each consignment valued at more than US$ 5,000 by COTECNA. This requirements entails a fee of 1.2% of the FOB value of the consignment. It is payable at the time of applying for the IDF. (Import Declaration Form). · Tanzania Central Freight Bureau (TFCB): The TFCB is a body responsible for ensuring that freight charges both from and into Tanzania are reasonable and competitive. For this service, the body charges 2.5% as a booking fees for ships from. and to Tanzania. · Phyto-sanitary charges: Phyto-sanitary certification costs US$ 15 per export consignment. This is a representative fee. Actual fees vary depending on the type of inspection required--which as we pointed about above varies from country to country in East and Southern Africa---and on infection/ infestation precautions which must be taken. Charges may increase, for example, if fumigation is required or if cargo is placed under quarantine.:Since maize is considered high-risk plant material that can accidentally bring along with it quarantine pests, importers of maize are required to: a) Apply for a Plant Permit (PIP). The permit specifies conditions for import prior to shipment. Once provided the permit is sent to the 67 exporting country in order to confirm that the conditions stipulated in the permit have been satisfied. b) The consignment must be accompanied by a phytosanitary certificate, which will be issued by the quarantine authority from the exporting country, in order to certify cleanness of the material, as well as full compliance with the PIP . In addition, declaration must be made that the maize is free from Erwinia Sterwarti, a bacterial wilt of maize. The consignment is also subject to phytosanitary screening by the Post Entry Plant Quarantine Station prior to release. · Ancillary Charges: In the event that an customs agent is engaged to assist with the clearance of cargo, the following additional costs will be incurred: --Documentation fees TShs 80,000= -- A negotiable agency fee which could be a %ge on CIF. --A surcharge where the clearing agent pays duty on behalf of the importer. Taken all together it is difficult to describe a progressive trade policy to the Government of Tanzania at least as that policy effects maize. 2.4.b Donor Role Several donor institutions have been involved in efforts to reform and support the maize supply chain in Tanzania. These include the Food and Agricultural Organization (FAO), UNDP, the World Food Programme (WFP) as well as the World Bank (WB) and a long list of bilateral donors. Many of these donors have contributed to food security for Tanzania, through the provision of food aid in times of shortage. They have contributed greatly to capacity development through training--particularly training with respect to food distribution and food supply chain management. Importantly, as well they have sponsored research into better production and marketing techniques. Donors have become increasingly sensitive to the adverse effects that subsidies of farm inputs, input credits and grain output sales can have on private sector development. They understand and increasingly reflect in their support of Tanzania that support which is delivered below marginal cost undercuttes market mechanisms, undercuttes the value of privately financed assets for which donor aid acts as a substitute and over the long term is not sustatinable. Still additional progress needs to be made on this front. 2.4.c The Role of Financial Institutions A new collateral law has just been passed in Tanzania which affords opportunities for financial institutions to increase their lending to maize supply chains with much less risk than they incurred previously when their lending was of necessity unsecured. However, the full effect and impact of this new legislation and the subsequent development of new lending instruments and new modes of secure lending to maize chain participants remain to be tested, tried and institutionalized. 2.4.d Third Party Service Providers Role 68 Transport inputs are required in every stage of the maize supply chain. Farmers require transportation to move their products from farm gate to storage facilities and from storage facilities to village markets. In this first stage of the chain most farmers use animal drawn and human portage, which is most often provided either on a first or third party basis. The cost of local portage or animal carriage is the opportunity cost of productive alternative activities. However, even in the first stage of the chain opportunities exist to better match transport service demand and supply and to more fully utilize carrying capacity through consolidation and the bulking up of consignments. Transportation service is also essential in the movement of the maize from the villages to trading centers and finally from trading centers the wholesale markets. Out of the 10 transporters interviewed at the Jangwani Parking yard, during the study team's field research 6 carried maize cargo that was being transported to Mtwara, Lindi and Tanga. These were all one off negotiated movements that entailed little prior or subsequent trip planning and backhaul arranging. Forward trip planning, more effective brokerage, consolidation of loads and full truck or rail car capacity utilization all afford opportunities for cost savings which are not being fully realized currently. They are not being realized because no efficient market exists for third or fourth party transport management services. Thus the two railways in Tanzania insist on selling direct to shippers, as do most of the motor carriers. Significantly, transport supply for maize is highly seasonal. Peak transport demand occurs during the maize harvest season. During this season opportunity costs for carriers whose equipment is poorly utilized increases dramatically. As a result, carriers typically price up sustanially more than average during peak harvest periods in areas which have poor rural infrastructure to reflect the lower asset utilization that they expect to be able to achieve. Indeed,. in much of rural Tanzania during the June-July harvest season risks of breakdowns on poor maintained rural roads are quite high. Thus, for example, transporters have a strong preference which is reflected both in the equipment they make available and the prices that they offer to transport maize from Inyalla village located near the Tanzam Highway rather to transport it from Igoma which is less accessible but which usually produces much more maize than Inyalla. 2.4. e Private Sector Institutions. Private institutions that operate in the maize sector are all informal, organized around local markets and are based primarily in the major market centers of Tanzania. Relatively few institutions encompass multiple participants in supply chains. Rather they principally represent a single function within the chain, such as traders, millers or, less frequently, farmers. A noteable exception is TAMAGRASAI, which is an association of farmers, traders, transporters and small-scale processors of maize. 2.5 The Legal and Regulatory Framework The legal and regulatory framework which support maize supply chains in the 21st century in Tanzania has developed in a piecemeal fashion and in response to a number of different problems and issues. It is time now to reexamine this mosaic of legal constructs and to assess its adequacy in the context of recent developments within the region and within the private sector. 69 In 1986, Tanzania made a firm commitment to its international donors to pursue policies and programs supportive of a market economy and to undertake a strict Structural Adjustment Program (SAP). The new policy placed a clear restriction on actions which the government could take to achieve its political and development objectives. Except in a limited cases such as restocking the emergency grain reserve, the government is not supposed to intervene in grain markets. Rather its role was to be limited to one of facilitating and promoting the participation of the private sector in efficient markets. Specific policy reforms which the Government embraced for the agricultural sector included the following: Cease from fixing producer and consumer prices Reduce export duties Remove agricultural subsidies on inputs including fertilizer and seed Remove quantitative restrictions on the internal movement of agricultural commodities, and Liberalize farm commodity markets and end the direct marketing of farm products At least with respect to maize the Government of Tanzania has complied with all of these reform commitments. However, in other significant ways the end result of efficient supply chains and effective well functioning markets has not been forthcoming at least not as forthcoming as was hoped. To some extent the failure to fulfill completely the expectations of the late 1980's has been the failure of the Government to surrender control over maize supply during periods of national food emergency. The Food Security Act of 1991 established a Food Security Department and empowered that Department of the Ministry of Agriculture with overseeing a strategic grain reserve. The Department was also vested with responsibility for dealing with all matters incidental to the establishment and management of the reserve. The declared objectives of the Act were to provide an effective food security system and to prevent food shortages like those which occurred in 1999 and again in 2004. Today, the Government's Strategic Food program in part competes with and in part complements the parallel programs of the WFO, CARE, etc. The ways in which these programs intersect and interact requires a new and "zero based" functional and legal review. The Tanzania Bureau of Standards (TBS) and National Food Control Commission (NFCC) are responsible for overseeing food safety and quality standards. The two bodies possess authorities which overlap to some extent. They impose, for example, separate but functionally similar permit compliance requirements on importers and thus increases transaction costs through double processing. It is a health and food safety requirement, for example, that, before maize can be imported into Tanzania, a food import permit must first be obtained from the NFCC. The procedure for the issuance of the permit involves submission of an application along with a sample of the maize which will be imported. (Stating the point of entry into Tanzania). The processing of applications is centralized at DAR. This is costly in terms of time and travel expenses-- particularly for upcountry traders as shown in Appendix 2.2. Tsh 1,000 is payable upon lodging an application and additional fees are payable when all tests are completed on the sample. An import permit is issued subject to positive sample test results. The control of other significant aspects of trade facilitation are likewise scattered throughout government as was discussed in the section above. A clear need exists to assess the way in which these programs intersect and interact with respect to facilitating grain 70 trade. The several functions which effect non-tariff trade restraints requires a new and "zero based" functional and legal review. Market Information Systems development in Tanzania dates back to 1970. That year that the Marketing Development Bureau (MDB) was first established under the Ministry of Agriculture. From its inception MDB pursued the following objectives: Provide expert advise to government on marketing policy Provide training in marketing to Agricultural Ministry staff and subsequent to the staff of the Ministry of Marketing and Cooperatives Provide a market news service When the MDB became a part of the Ministry of Marketing and Cooperatives additional functional responsibilities were added to its mandate, including: Monitor consumer prices Research the cost of crop production Since 1986 MDB and the Agricultural Market Services Department, whose name it assumed in 1998, has gradually transformed itself, its functions, organization structure and commodity coverage. Today, however, the department finds itself competing with private sector providers of market information services. Its competitors have a larger ( regional) market purview are effectively interconnected with a number of media companies who can effectively distribute MIS and have plans to developing new information based market service which promise to increase the efficiency with which Tanzanian grain markets operate. 2.6 Findings and Recommendations Maize is the primary food staple produced in Tanzania. Peasant producers who depend on maize are most adversely affected by the slow pace of development for those institutions which should strengthen farm to market linkages. It is their incomes, their living standards and their morale which are most adversely effected by this institutional failure. The consequences of stultified maize chain development are depressed production, low incomes for farmers and volatile price fluctuations at the farm end of the chain. Maize production, processing and marketing has fallen short of its full commercial potential in large part because of the way in which supply chains within the country have developed. As we have argued above economic value subtracted from maize chains between the farm gate and the retail customer create a poverty trap from which it is extremely difficult for the 1.5 million farmers who depend on maize production to escape. Tanzania has several natural advantages in the maize trade. It is well positioned in the center of a large regional market for export maize. It has abundant rain fed land which should allow its maize producers to be more productive then they are currently. However, the controls which the government has imposed---both on maize exports and on maize emergency relief stocks---undercut incentives for productive development of farm to market linkages and frustrate the full potential development of the maize chain. 71 Significantly, they also subject the country to secularly declining terms of trade. The fall off of maize export earnings only reduces the country's capacity to import. When the government cannot predict its export earnings with any certainty, planning and reallocation of discretionary resources to programs designed to combat poverty becomes all the more difficult. The remainder of this section suggests some potential solutions to the sector level structural problems which persist in Tanzania's maize sector. 2.6.a Supply Chain Structure Little industrial organization currently exists within Tanzanian maize chains. Most of the commercial interactions which take place within legacy chains are "arms length" and transactional, as contrasted with long term, contractual and incentivized to progressively improve process integration. Within these legacy chains, commercial structures need essentially to be developed anew with each serial transaction. The result is high transaction cost and little internalized incentive for productivity improvements, conveyance of information or end to end process alignment. Few incentives exist within legacy chains, for example, to remove process redundancies, to reduce the number of value confirming inspections, to upgrade product quality or to differentiate maize based products. When incentives are absent to squeeze costs out of the system or to innovate in mutually beneficial ways, costs naturally accumulate and innovations are spurned. Among the several participants in the Tanzania maize chain only the integrated traders whom we described in Section 2.1 have been effective in process improvement and in harvesting the productivity gains which have resulted. Their integration is the integration that results from vertical ownership under single corporate control and this integration does, indeed, create internal incentives for productivity enhancement, capital investment in supply chain assets and transaction cost reduction. These gains, however, are captured completely within the domain of proprietary activities that individual integrated traders control for their own benefit. However, since these structures exclude the farm and retail outlet ends of the chain, these links remain disconnected from intermediate links and they do not participate in the productivity gains which integrated traders have been able to realize. There is very little "trickle down" effect from the gains realized by the integrated traders and no investment in common goods which would benefit multiple participants in the maize chain on a non proprietary basis. This situation can be corrected by decisive government action. Government has three means for creating incentives for strengthening backward and forward linkages within legacy chains. These include: i) tax policy, ii) trade policy and specifically export licensing and iii) procurement for emergency food relief . To date none of these incentives have been applied actively to realize structural changes within legacy supply chains for maize. Indeed, as we have argued above these policy levers-- and emergency food relief procurement in particular--- have been used instead to affect economies of exclusion and of preferential dealing which have strengthened integrated dealers at the expense of other market participants and other less well organized supply chain structures. Opportunities exist, for example, to assist major local wholesale market institutions, such as the Tandale and Buguruni markets to build backward linkages to the farmers who supply them. Both of these municipal markets currently suffer from poor hygiene conditions with uncertain food safety consequences. However, they do provide 72 substantial storage capacity. Arrangements might be considered to recapitalize and reorganize these municipal markets as concessions and to link them with farm producers via investments made by concessionaires in new chain aligning systems and new upcountry storage. At the present time, the key missing link is the link backwards from traders and processors to the farmer. In the future, the three policy leverage points could be used to induce some measure of industrial structural realignment which would entail the inclusion of farmers in commercial chains, risk sharing between farmers and other chain participants and facilitating the transfer of technology among chains participants. Still other programs could be developed which began at the farm end of the chain. These programs would test and refine farm based commercial business models which entailed strong farm to market linkages and effective forward integration from the farm end of the chain. Bottom up development of this sort is being actively pursued in both Western Kenya and in Uganda. Similar experimentation, business model refinement and scaling up should become a key aspect of the development agenda of the Ministry of Cooperatives and Marketing in partnership with donors and technically qualified NGO's. 2.6.b Regulation Regulatory policy within the maize supply chain is effected primarily through three sets of interrelated interventions, including : i) food security and the supply of emergency food relief, ii) trade policy , and iii) food security and food quality control. The policy backdrop against which these three sets of market intervention levers are exercised currently lacks clarity and definition. However, policy preferences which are revealed through actions and initiatives clearly reveal that government policy is highly risk adverse, fixated on short term political fall out associated with food shortages, lagging with regard to trade liberalization as contrasted with leading, and tactical as contrasted with visionary and forward looking. As we suggested in the preceding section, a clear need exists for a "zero based" review of all of the government functions that affect the maize supply chain. These functions have developed at different times and in response to quite different needs. They lack a coherent architecture and a contemporary raison d'etre. They include processes and activities which are mutually off setting and which increase transaction costs for private participants in local markets who attempt to comply with them. Taken jointly they create incentives for informal market trading and smuggling and disadvantage formal sector market participants vis a vis informal market participants. Most importantly regulations in this key sector entails little engagement of the private sector in a governance role and little transparency or accountability in decision making ( e.g. under what conditions does a food shortage exist and under what conditions should maize exports be embargoed). Including the private sector in the decision making process and on advice providing boards and vetting regulations for public commentary before enforcing them would much better integrate what is happening within government with what is happening within Tanzania's several private sector markets for maize. Food Security 73 Increasingly as regional markets develop within East Africa, food security has become more a regional issue and less a national one. Opportunities for linking up sources of maize surplus with needy consumers lacking grains in deficit areas are most likely to be found when the circumference within which open market operations are allowed to prevail is expanded and not contracted. In the case of Tanzania this truism is all the more applicable, because much of Tanzania's maize is produced in only 6 of the nation's 20 regions, hence interregional as well as international transfers from surplus to deficit regions is essentially important for assuring food security.. Given these circumstances open regional trade in maize may well be the most certain and efficient strategy for assuring both grain sufficiency within Tanzania and maximum production incentives within the country as well. In any case, the fact is that several NGO's, multilateral agencies and Tanzania's own Food Security Department execute similar and overlapping functions, in intervening in food market operations. These need to be reconciled and harmonized so that not only can functions of emergency food relief be executed in a way that assures better value for money in the short term but also that longer supply chain development objectives can be similarly advanced. This is not happening at the moment. The study team recommends that an in depth assessment be made of the ways in which emergency food security supply chains operate within Tanzania and that this assessment focus specifically on reconciling different policies and procurement/ distribution methods among the several agencies which provide emergency food relief. Once an assessment of the longer term development effects of various food relief policies and mechanisms is complete, monitoring and regulating these policies and methods should then become a primary regulatory function of government. Allowing agents of food aid to operate independently and without some regulatory constraints within Tanzania is to forfeit one of the most critical market development controls within the grasp of government. Trade Policy The country's emergency food relief program is closely linked to its trade policy, since a determination that an emergency food shortage is forthcoming is sufficient cause for govenment to shut down export licensing. However, the precise preconditions and/or empirical findings which trigger this draconian measure are neither clearly enumerated nor systematically applied. Similarly, excessively high or low internal prices for maize appear to one of if not the sole cause for government either to constrain or alternately to facilitate import licensing. Again, clear criteria and transparent assessments are missing with respect to how and when import licenses are granted. The recently completed COMESA "Maize Market Study" recommends that regionally accepted and objectively measurable parameters be used to invoke maize export bans or to trigger maize import restrictions. It further recommends that these be developed within the overall framework of the ,Safeguards Clause' of both the EAC and COMESA Treaties. In addition the study recommends that a regional food security information clearing house become the source for statistics and forecasts which would be used to calculate ban triggering parameters. The authors of this study enthusiastically concur with these recommendations. The COMESA study further recommends that regional policies be 74 harmonized on export and import regulations, that a regional crop forecasting system be developed to provide reliable information on maize availability. Again the authors enthusiastically concur. Additional recommendations which emerge from the COMESA study that have direct relevance and application to Tanzania's trade policy include the following: i) A need exists to harmonize Tanzanian tariffs on maize in line with those of neighboring countries in COMESA and EAC region and to reduce maize tariffs with EAC trading partners progressively to zero, ii) Tanzania needs to accede to harmonized EAC maize quality standards and testing methods; iii) similarly, Tanzania needs to harmonize its phytosanitary regulations and requirements with those of its EAC trading partners, including the adoption of common diagnostic and inspection procedures and a common regional formal for phytosanitary certification; iv) Tanzania needs to conform its customs clearance practices to regionally accepted norms and to adopt a harmonized Single Entry Document for cross border trade which does not discourage small scale traders. Quality Control Efforts need to be made in concert with trade associations representing each of the market niches discussed above to develop standards for food safety and for grain quality which are harmonized to the extend possible with standards which apply to imports and exports and which are broadly accepted within the local markets. Ideally these standards should differentiate multiple levels of quality and thus create incentives for investment in quality production, protection and securitization within entire farm to market chains. 2.6.c More Productive Use of Key Assets The Food Security Department of the Ministry of Agriculture remains the single largest owner of maize storage capacity within Tanzania. This agency does not make productive use of the storage assets under its control. Similarly, the TAZARA and TRC railways are the largest owners of grain carrying freight car equipment within the country. Initiatives need to be undertaken with respect to these two key sets of supply chain assets and others as well to create third party leasing or service companies to supply these assets to first and second parties involved in maize supply chains at prices which reflect the economic cost of these assets. Not only will such a policy increase the productive supply of critical assets but also create incentives for their future investment and provision though business models whose mission definition is the supply of short assets to the maize chain. 2.6.d Market Institutions The markets for maize in Tanzania are very thinly capitalized. The only working capital available in these markets is the equity and balance sheet backed debt which second parties supply to the market. Very little third or fourth party financing is available to create market depth and stability. The thinness of the Tanzanian market and its relative isolation from other markets are two of the factors that most adversely effect risks associated with investing in it. One potential solution is to subordinate the Tanzanian grain market to a much larger regional market and to allow the market institutions which operate at a pan regional level to subsume trading in Tanzania itself. Several promising new market institutions 75 are emerging within the region which depend on the internet and on internet based clearing mechanisms to operate. Alternatively, a strategic affiliation with the Johannesburg Commodity Exchange is worth considering. It may be possible to create one or more new trading basis for pricing and ownership transfer on the Johannesburg Exchange. The new pricing and trading bases would be located within Tanzania but contracts would be sold in Johannesburg and local store housemen who created the Tanzanian grain securities would operate as licensees of the Exchange. The development of deeper and more stable commodity markets within Tanzania is multifaceted and requires first the development of a network of independent asset managers or public warehousemen who are empowered by financial institutions to create security interests in grain inventories. These asset managers would ideally operate at strategic junctions on the railway system and thus facilitate the inter-linkage of rail and commodity prices in ways which were at the same time transparent, contestable and competitive. Mezzo Level Organizations Mezzo level organizations also need to be created within the private sector in order to do the work of standards development and enforcement, professional certification and harmonization of commercial trading protocols within the several local market niches which exist within Tanzania. It is critical that these organizations define their membership over the entire set of value adding activities which define a value chain. The government can facilitate the creation of mezzo level organizations by explicitly devolving its responsibilities to one licensed and certified organization for each niche market linked chain. 2.6. e External Markets for Specialized Services The efficiency with which external markets operate on the periphery of supply chains is almost as important as how efficiency the value adding activities which are performed which specific supply chains internalize within themselves. Some activities on which supply chains are dependent are more efficiently out-sourced ( externalized) and some are more efficiently in-sourced ( internalized). Chains which become dependent on third party providers of services need these services to be supplied in truly competitive markets less the service suppliers manage to capture a disproportionate share of the value being created in specific chains. Thus, in the case of the maize supply chains key external services such as banking, rail and highway transport, warehousing and market information services are furnished by third parties. It is critically important that each of these providers either operate in a highly competitive and efficient market or that the option be maintained that first and second parties within the chain can enter these markets and provide their own essential services with own work forces, management and capital Banking As we noted above liquidity is in very short supply for the maize supply chains which are anchored in Tanzania. The level of investment on the part of commercial bankers in agricultural infrastructure, farm inputs and farm equipment is extremely low. Part of the problem has to do with the lack of bankable collateral within the agricultural and agro-business sectors. However, recently passed legislation addresses this issue and should no longer provide a fig leaf for commercial banks to hide behind. 76 Commercial banks do not manage agricultural or supply chains risks well. Moreover, they are highly risk adverse. Tight prudential controls and constraints enforced through the Central Bank afford little incentive for commercial banks to invest in assets other than those whose performance is low risk or better yet guaranteed by government. Few models exist in Tanzania's financial sector to point the way toward successful financial entrepreneurship and few notably successful private enterprises in the maize supply chain have development which can serve as starting points for further commercial experimentation, business process refinement and institutional learning. The two integrated trading companies referred to above are extremely well financed. However, they are almost alone. Thus, government aligned institutions, government owned enterprises and enterprises affiliated with food aid grain supply have much less difficulty securing financing than do independent private enterprises. The reputation of the borrower rather than the merits of the project for which he is lending too often determine the willingness of banks to assume commercial risk. For may of Tanzania's financial institutions "relationship lending" supersedes rigorous risk management as a basis for proceeding with a loan. For all of these reasons, country's savings are not being effectively rationed and channeled based on highest expected returns. Hence, economic growth is slow At the same time, the country's embryonic micro financial institutions are heavily risk exposed by virtue of their un- diversified local loan portfolios, their limited equity and their high transaction costs. Another fundamental deficiency of both commercial banks and of micro lending institutions is their limited ability to assess and manage farm and agricultural market risk. Neither set of institutions is equipped to manage the significant production, weather and price risks that are directly associated with the financing farm inputs. So one key challenge for the Tanzanian financial sector is to link together risk takers in the country's emerging private sector whose core business is trading in farm inputs and outputs with financial institutions who are either risk adverse or risk management capacity constrained. The need to form these alliances is based on the fact that the two partners have both complementary competences and complementary needs. Commercial banks are able to mobilize savings but are not able to effectively manage risks associated with selling farm outputs and recouping the cost of farm inputs. Micro lending institutions are limited in their ability to mobilize funds in poor farm areas and to effectively manage investment risk in the same areas. Real sector actors, on the other hand, may be able to assess and manage farm input and output risks but are not able to mobilize savings. Indeed, it should be pointed out that most of the credit that is expended to the maize supply chain currently is trade credit of one form or another. By default a lot of supply chain financing is already taking place in Tanzania. A worthwhile objective is to facilitate, legitimize and main stream this kind of activity through explicitly developed and closely monitored strategic alliances between commercial banks and real sector players. Effective commercial management of credit risk will entail effective control both over input merchandising and of farm product off taking. In other words petty and medium scale traders in maize are the most likely candidates to become effective managers as well of input credit risk. Unfortunately input merchandising has been separated from output merchandising for the most part in Tanzania. Managing the risks associated with input credit sales relates primarily to understanding and managing risks associated with output merchandising. 77 The kinds of strategic alliances that might be tested would involve one of three possible models: i) strategic alliances in the form of joint ventures or special purpose limited liability companies formed at the regional level between merchandisers of grain, wholesalers and retailers of farm inputs and commercial banks. These strategic alliances would work " back to back" with partner merchandisers and commercial banks to provide input credit. Commercial bank investment would most likely come though a bank holding company structure rather than through any direct commercial bank investment. The business case for these joint venture would be defined initially through a detailed business plan and constraints on business activities would be agreed and enforced through corporate charters. Liabilities for investors would be limited to paid in equity and ownership would be divided among strategic partners based on their financial capacities, but, in any case, would be adequate to provide incentives for all joint venture partners to exercise effective prudential control over credit extension. Management would be provided under contract and each management contract would include incentives for profitable credit management. Credit managers would offer input credits to qualified farmers or to farm cooperatives based on their capacity to use these inputs productively and to repay them promptly once outputs are harvested and sold. Repayment of credits may take the form of offsets against contracted, previously priced output sales. ii) The second option would involve commercial banks buying into going concerns whose business was the merchandising of farm products and the sale of inputs to farmers and to farm cooperatives. This second option would expose commercial banks more directly to farm commodity production, selling, storage and pricing risks than would the first option. It would afford the benefits and exposures of this more intimate linkage to the farm economy. Of course, once again the investment vehicle would be a limited liability company and the corporate route for commercial bank investment would be through a holding company. Thus, the exposure of commercial banks would be limited to their paid in equity and their relationship with the invested company would be arms length. This option might be well suited to rolling out and extending the scope and scale of the farm equipment leasing company that we discussed briefly above. iii) The third option is different in kind. Instead of developing an alliance in the form of a jointly owned enterprise, it would involve the development of a special purpose investment security. This security would be sold to banks and to other financial institutions. In general, each security would specify the co-investment requirements, guarantees and collaterals pledged and committed by the issuers who would be qualified farm product input/ output merchants. Co-investment requirements, guarantees and/or collaterals would be enforceable against the entire set of issuers who effectively would provide cross guarantees. " Fertilizer securities" would be sold through an auction to individual commercial banks---like government securities--- or to sold to a consortium of bank's through a negotiated sale. 78 Collateral Required for Commercial Banking in Tanzania During the study period , on behalf of the banking industry, the president of the Commercial Bankers Association highlighted some of his fellow banker's concerns regarding the provisions of the LAND ACT NO.4 of 1999 which relates to liens against farm land. The following among other sections were highlighted in this regard: Section 20(3) provides that "Subject to section 37(8),all lands acquired prior to the enactment of this act shall be deemed to have no value ,save for unexhausted improvements for which compensation may be paid under this Act or any other law". Section 37(8) provides that "Unless otherwise provided by this Act, no sale of land without unexhausted improvements shall be approved ,and as such land shall be deemed to have no value". Section 124(1) provides that " any rule of law written or unwritten, entitling a mortgagee (lender) to foreclose the equity of redemption in mortgaged land is abolished". Section 48(1) of the civil Procedure Act, 1996 provides that: "the following shall not be liable for attachment or sale ,in execution of a decree namely: a)Any land used for agricultural purpose by a village , an ujamaa village, a co-operative society, or an individual whose livelihood is wholly dependent upon use of such land; b)Any residential house or building ,or part of a house or building occupied by the judgment debtor ,his wife and dependant children for residential purposes. The above provision did not exist in the bill; and its inclusion in the Act is in contradiction with the National Land Policy, which is incorporated in the Act by reference The practical problems embedded in this arrangement is the following: -Land itself cannot be valued in order to form the basis for mortgage or sale -Land cannot form collateral for the initial draw down. -The National Land Policy and current practice recognize that empty land has value while the Act does not. -Holder cannot use land as equity with a foreign investor .(The Act should clearly state that land has value and it can be determined for purpose of disposition). Shouldn't the law be changed, there are real risks that befall the banking industry and ultimately the macro-economic stability of the country .These among other include: - A continued rising cost of funds due to the risk of premium inherent in the cost of managing credit especially the recovery process and protracted litigation. - A severe squeeze thus depriving the household and businesses of any financial resources needed for economic development. - Hampering of mortgage finance aimed at creating home ownership. For the above reasons a review of the Land Act No.4 of 1999 is paramount in order to create an enabling environment one which is required to kick-start the concept of supply chain financing in the country. Transport The railway network that exists within Tanzania has the potential to serve as the backbone for a regional grain market. Unfortunately the way that railway assets are being managed and made available to grain shippers substantially diminished the full potential which exists in these assets. The first priority for maize market development should be the realignment of strategy, business focus, specialized assets and profit and loss orientation of the two railways along line of business organizational structures. 79 The business of supporting maize shippers is a large, growing and already substantial one in Tanzania. An opportunity exists to develop a grain logistics service company which is initially rail based and whose underlying business premise is the supply of cost effective transport and logistics service to maize shippers. Such as line of business organization would buy train services from the two railways, operate specialized grain loading and unloading facilities and operate train sets of specialize grain hauling rail cars. The company would be rail based but would operate off the rail network as well as on the rail network. It would provide water and highway transport to its shipper customers. Moreover, the new line of business company would attempt to maximize the use of its core assets by bulking, consolidating and gathering large lots of maize into train load quantities for train load shipment beyond it terminals. Over time the company might begin to develop partnership relationships with its key customers and encourage them to invest in specialized grain storage and transshipment facilities, rapid discharge cars specifically designed to haul maize, etc. In this way customers with a short term view of mutual commercial linkages could be transformed in to long term partners who had invested in specialized assets which were most useful both to the carrier and the shipper when used in a close partnership mode of operation. The development and launch of a grain oriented line of business organization could be undertaken as part of the restructuring and privatization preparation process which is currently underway on both the TRC and the TAZARA. Still other opportunities in the transport exist to license third and fourth party transport market managers and supply chain integrators whose business mission would be to fill under utilized backhaul capacity and to improve the asset utilization of slow moving trucks and containers. The application of internet technology to the development of regional freight markets holds out great potential benefit. However, to date no regulatory or legal framework has been put into place in East Africa to determine what authorities a virtual carrier could exercise, what liabilities it must assume and on what basis it can pass though or not pass through the common carrier obligations of real carriers and real transporters whose equipment and drivers it manages. A positive step forward would be to assess the feasibility of a virtual market for highway services which was grain oriented. Warehousing Another set of critically underdeveloped assets are storehouse assets. Tanzanian supply chains badly need a set of third party warehouse and fumigation service providers who operate in the mode of a public utility and whose business mission is the provision of storage, cleaning, fumigating and grain inventory securitizing services for a fee. Importantly this regulated service industry would not buy and resell for its own account but rather would operate exclusively as a third party service provider, possibly under the terms of a franchise or joint venture with a commercial bank or commodity exchange. At the present time no third party storage services are available in Tanzania. First and second parties own and operate their own storehouses but fee for service warehousing is missing. Developing a third party warehouse industry is also the key to creating a bankable collateral in maize inventory. Commercial banks will never be willing to lend against grain which is stored in the warehouse of a borrower. However, they 80 might be willing to lend against the same inventory if it were insured against secure release and securitized by a licensed public warehousemen. At the present time several different and incompatible warehouse receipt legal frameworks are under consideration or have recently been made law in Uganda and Kenya. Communalizing these several warehouse receipt regimes should be a high priority. It will not only increase bank ability and tradability but will serve as a logical stepping stone for the development of a regional grain trading exchange As a set of prudent collateral development initiatives, a warehouse receipt system might work best be launched if it were developed jointly with as established producer group, in the context of collateral quality control systems implementation efforts and jointly with the use of high quality storage technology and a good and reliable information system which could be used to monitor and then disseminate price and inventory information to relevant stakeholders. Underwriting such a system might prove quite useful in helping increase creditor confidence in the maize chain. Market Information Services As noted above several private sector initiatives are underway within the region to develop and disseminate timely and accurate market data. Increasingly, these initiatives are having spillover effects on Tanzania. Pushing these initiatives to the forefront and facilitating their development through a series of public-private sector partnerships should be a very high priority for government. 2.6.f Infrastructure Redevelopment Priorities for roadway development need to be better aligned with priorities for the development of underlying agricultural markets. Both investment priority setting and accountability for service improvement results should become the explicit responsibility of mezzo level organizations which are deeply involved in critical trades including the maize trade. 2.6.g Production Enhancement at the Farm Level Community development is imperative in order to improve productivity, supply chain agility and equities within long chains. The soil nutrient deprecation is a major issue in most of the maize growing areas of Tanzania. Methods for re-capitalizing depleted soils require strengthening backward and forward linkages from the farm. Being able to identify success stories and concrete examples of rural commercial restructuring efforts which have realized substantial gains in production, livelihoods and incomes is critical to this end. Equally important is developing the institutional capacity within government to scale up successful farm business experiments to a substantial commercial scale without diminishing the benefits realized in behalf of poor farmers. When the commercial vision of a new kind of farm organization is consistent with basic human needs, growth-with-equity will invariably be realized. Demonstration projects should be initiated which test the merits of new farm business models and of 81 new forms of supply chain partnerships with the purpose of elevating the living standards of the rural poor.. At the individual community level, its crucial to unleash the resourcefulness and innovative spirit of the people. This can be realized through demonstration projects which focus on capacity building, strengthening backward and forward linkages and private sector co-investment. The table below outlines some of the kinds of initiatives that need to be considered. Some of the characteristics that distinguish small holders who are conditioned for poverty form those who are empowered to produce and market surpluses, include the following: · Farm strategy: Traditional farmers concentrate on subsistence crops and sell surpluses when they are available. Non traditional farmers undertake mixed enterprise agriculture. They diversify their production and price risk over multiple products and services. · Crop varieties: Traditional farmers rely on traditional cultivars. Non traditional farmers routinely purchase seed for new hybrids. · Cropping systems: Traditional farmers apply continuous cropping of subsistence crops. Non traditional farmers rotate their crops and frequently intercrop in order to feed the soil and also to exploit market opportunities. · Nutrient management: Traditional farmers allow soil to lay fallow. They apply little fertilizer or chemicals. Non traditional farmers use both organic and mineral materials to restore soil productivity. · Composting: Traditional farmers do not compost . Non traditional farmers compost and regularly apply compost to their fields. · Organization: Traditional farmers are not actively involved in organizations outside their family farms. Non traditional farmers actively participate in farm affiliated input buying and output marketing organizations. · Market information: Traditional farmers use little or no outside information. Non traditional farmers make all of their decisions based on best and most timely information available. 2.6.h Skills Development Most important in developing new and more efficient supply chains for maize are the competencies and skills required to diagnose, remediate and perpetually refine the business models which include, complement and support competitive chains. These skills are developed both through formal training at the university level and though continuing professional education. The study team recommends that a two pronged effort be made to leverage both sources of skills enhancement. Specifically, the team recommends that a degree program be started up at one of the National Universities and that tenured professorship be dedicated to supply chain organizational development. In addition, the team recommends that a professional association of agricultural supply chain managers be formed and that this organization be assisted by donors in developing a skills upgrading and professional certification process. Both institutions should be encouraged to network with counterpart institutions in other parts of the world and, of course with one another. 82 . 83 Chapter 3 Sugar Supply Chain Development in Tanzania 3.0 Background: Sugar Production/ Distribution in Tanzania This chapter deals will issues that relate to the efficiency with which supply chains for sugar production/ distribution operate in Tanzania. The country's sugar industry is currently going through a significant transformation in structure, production capacity and strategy. This fundamental transformation has been caused by the recent privatization and subsequent reorganization of the sugar industry. Tanzania's sugar industry is rapidly becoming regionally competitive and is developing both production costs and marketing capabilities which are beginning to rival those of other regional producers. Still, the industry has not yet begun to export any substantial volume of sugar beyond the country's borders. Three private companies currently dominate sugar production in the country and a forth company is preparing to enter the market subsequent to the financial closure of its privatization transaction. In addition, several small scale producers, who produce 1000- 2000 tons per year, continue to operate within the industry as do a limited number of cooperatives. As the three larger private companies enhance their production capacities, they will begin to outgrowth the parameters of the local market which is quite small. What happens next, once domestic supply in Tanzania equals domestic demand, will have a major impact on the future of the industry. The development of the sugar industry is important from the point of view of poverty alleviation. Sugar processing has the unique capacity to improve the well being of large numbers of poor farmers. The industry is just beginning to have a positive development impact on the 10,000 plus out growers who are affiliated closely with the three large producers and on the more than 80,000 employees who currently work within the industry. The industry is adding jobs and affiliating with new out growers at the rates respectively of 3% and 5% per year. Until 1999 while still under public sector ownership, the sugar industry was in sharp decline. Then as now the industry served almost exclusively domestic markets for raw sugar , with the exception of the small quota of sugar imports allocated to Tanzanian producers by the EU. A two tiered tariff protection arrangement allowed state owned sugar refineries to satisfy most of the country's internal demand for "raw sugar" but none of its demand for "industrial sugar." For the most part these same protections remain in place and, in fact, were part of the negotiated basis on which the privatization of the sugar industry has taken place. Today, under private ownership the sugar industry is growing its production capacity and, at the same time, is improving its productivity and cost competitiveness. Future trade policy will have a major effect on the industry's continued growth or possible future decline. The trade policy which existed at the time of privatization maintains in effect. One aspect of this baseline policy involves a compromise between sugar manufacturers and industrial sugar users under which users of " industrial sugar" are granted preferential tariffs while higher tariffs apply to consumer or "raw sugar." This compromise will soon be tested as the new private owners of the nation's five large scale sugar refineries increase their production to levels where it begins to meet or exceed domestic demand for raw sugar. At that point leaders in the newly privatized industry would prefer to open local industrial sugar markets and / or possibly open regional markets, particularly in Kenya, rather than precipitate an internecine price war and even possibly provoke an industry shake out. This policy "tipping point" is not far away. A complicating factor is Tanzania's own commitments to regional trade integration. Tanzania is a signatory to two regional trade agreements--the EAC and the SADC. Members of these two trade blocks have quite different interests with respect to sugar trade liberalization since their underlying sugar industries differ dramatically in their competitiveness and hence their exposure to further trade liberalization. Tanzania stands between the two blocks, both geographically and philosophically. The speed and direction which regional trade liberalization takes effect is probably the most determinative factor with respect to the future of Tanzania's sugar sector. This chapter focuses of several additional interrelated issues that affect the efficiency and adaptability of sugar supply chains in Tanzania, including: i) the rapidly evolving industrial organization of sugar supply chains; ii) the kinds of competition/cooperation actions/ reactions which are beginning to emerge among industry participants in the post privatization era; iii) effects of specialized service providers on sugar production/ distribution chains; and finally iv) the effects of public sector policy, regulation and infrastructure provision on the development of the industry. The subsections which follow this background discussion are organized around the following topics: i) the industrial and market structure within the sugar industry and, importantly, the evolving role of small scale farm producers of sugar cane within that structure; ii) the operation and economics of the supply chain for sugar; iii) the key factors which affect sugar supply chain competitiveness in Tanzania; iv) factors which limit the productivity of Tanzanian sugar producers; and importantly; v) the supportive ( or not-so-supportive) roles played by new key institutional participants in the sugar supply chains, including mezzo level organizations which have formed to represent the collective interests both of out grower farmers and of large scale refiner/ distributors. Also included in this section is a brief review, under topic (vi), of the legal framework that affects the efficiency and market protection , as well as the equities the prevail among stakeholders in the nation's sugar production/ distribution/ merchandising system. This section also reviews in some detail the basis for taxation within the chain. Based on a critical review of these several topics, the section goes on to set out its recommendations in subsection (vii). These recommendations for both policy reform and investment are intended to raise the incomes of small scale farmers. 3.0.a Overview of the Sugar Supply Chain World wide the sugar industry is Balkanized into a number of discrete national markets with a global market overhanging this ensemble. National markets for sugar are disconnected from one another by tariffs, quotas and other trade barriers. As a result of this separation, sugar prices vary significantly from one market to another. Prices in global markets reflect residual supply and demand outside of protected national market domains. Consequently global prices are lower and more volatile than they might otherwise be.1 Behind these trade barriers a large international sugar processing industry has grown up in a patchwork with greater or lesser processing efficiency prevailing within individual countries. 2 This market patchwork is clearly the legacy in East Africa and it 1See" Sugar Policy and Reform" World Bank Working Paper, 2001, Donald F. Larson and Brent Borrel 2See " Sugar Policies: Opportunity for Change," World Bank Policy Paper, 2003, Donald Mitchell is the backdrop against which the Tanzanian sugar industry is developing under new private ownership.3 Sugar refining is both labor and capital intensive. Importantly as well, it is highly management intensive4. Competitive sugar industries require strong and well integrated supply chains with well organized backward linkages from refineries as well as forward linkages to retail markets. When the right set of input factors come together so that productive supply chains can operate, as they are beginning to in Tanzania, the wealth creation effects of productive sugar sectors can be quite significant for poor farmers. The labor intensity of sugar production makes sugar refining a particularly effective tool for economic development. Once they are built sugar refining facilities need to be supported over their entire economic life by labor intensive sugar cane production. Hence, efficient sugar processing can realize a significant welfare dividend for small scale farmers. Most sugar industries in Africa were originally built on economic policies which emphasized import substitution. This philosophy had the unfortunate effect of breaking large regional markets up into much smaller ones in which high cost producers could survive behind trade barriers. However, significant economies of scale were forfeit by regional sugar supply chains when processing facilities were down sized to match smaller national markets and forward and backward linkages were developed on this smaller scale. If it were move effectively integrated, the East African market could provide significant incentives to realize economies of scale and to move to lower cost production and distribution within the region. However, any transition to such a development state is fraught with adjustment challenges and social costs. The place of Tanzania's own industry in a liberalized regional market depends completely on the pace and the path through which free market liberalization is achieved. Significantly, in addition to being labor and capital intensive, sugar production is also transport intensive. Hence, the location of refining facilities both vis a vis both primary inputs and end markets is critical for sustaining competitiveness. In developing countries refiners are most efficient from a supply chain design perspective when they locate close to inputs, e.g. to farm level production of cane.5 .Because the cost of transporting the primary input is extremely high compared to the cost of transporting the output particularly when transport infrastructure is poor and water transport is unavailable most refineries are located close to their input supplies. Still, product distribution costs ( outbound logistics costs) are also a significant factor in determining industry competitiveness especially in a large country like Tanzania. However, as it turns out they are not as significant as the costs associated with inbound logistics. The industrial geography of national sugar economies is a critical factor in their sustainable growth. Clearly all of these issues apply to Tanzania where the primary refiners are separated somewhat from one another and from the major consumer market in Dar. Most importantly, all major producers are well located near productive cane growing areas. As with most developing countries, so too for Tanzania, demand and supply for sugar clear within the boundaries of national markets and most product flows take place only within these national borders. Export parity establishes a cap on local prices and the 3See" The Sugar Industries of East and Southern Africa: The Challenge Ahead" International Sugar Journal, 2004, vol, 106, Jonathan Innes 4Ibid 5Larson marginal cost of the least efficient producer establishes a floor under national prices. Prices within the local market move within this range and are for the most case stable at the level or the least efficient producer because of the level of trade protection which producers enjoy. Since in Tanzania sugar production takes place over 6-7 months when cane sugar is available for processing producers are faced with the risks and challenges associated with producing for expected demand levels well in advance of their realization. Hence, hording or mis calculation of demand on the part of major producers can result in price spikes. Sill, national demand is predicable and has been increasing steadily year over year, in step with population growth. Until the recent privatization the supply response to this growth was lumpy and related to discrete additions of production capacity. However, the new private operators of Tanzania's sugar refineries have found ways to produce substantially more sugar without having to make large capital expenditures. These sugar entrepreneurs have found ways to produce more with less. How far the expansion of production capacity can go remains to be seem. Clearly, the supply of 100% of domestic demand is a critical threshold. Price wars can and, indeed, may break out within national markets once this threshold is achieved or surpassed.. This " tipping point" has not, however, yet been reached. However, it is quickly approaching. Indeed, individual industry participants are beginning to position themselves for this development. Importantly, the strategic options that are available to individual enterprises will be determined in large part by the policies and actions of the government. In the mean time, as the industry continues to increase its sugar production capacity incrementally, demand for cane inputs continue to increase apace. The national market for sugar is becoming more efficiently integrated and with this national market integration per capita consumption of sugar appears to be increasing. Sugar is no high priced or in short supply in Tanzania's rural districts. With this increase in demand for primary agricultural inputs, commercial linkages between processors/ distributors/ marketers and small scale farmers continue to evolve and to strengthen. Small scale farmers are the source of most of the marginal sugar cane inputs that the recently privatized refineries process. Indeed, the newly privatized sugar industry has become a major source of income and, indeed, of wealth creation for affiliated out growers. Moreover, the regulatory framework which supports the industry generally and ,specifically, in the area of collective bargaining between farmer organizations and the new private refiners appears to be working quite well in balancing conflicting economic interests. Lessons can be learned from this regulatory model which have application more generally to other agri-business sub sectors in Tanzania. 3.1 Industry and Market Structures This section describes the market context in which Tanzania's sugar industry operates. It describes the industry's structure and the organization of its distribution channels. It goes on to describe the state of development of markets for specialized, third party supplied services on which the industry depends more and more. 3.1 a Global Markets The world sugar market is characterized by volatile prices and widespread government intervention. In spite of the north/south juxtaposition of international sugar import demand and export supply during the 1990's and early 2000's trade restrictions remained in effect in many countries---including tariffs, quotas and other controls over imports, domestic prices, domestic demand and supply. These restrictions tend to make the pattern of world production, consumption, and trade less efficient than it might otherwise be. Still, with the dissolution of COMECON and the adoption of more liberal trade policies restrains have diminish somewhat, in the Former Soviet Union, the Middle East, North Africa and East Asia. Import demand for " global sugar" has increased accordingly.6 Thus, for example, Indonesia, Taiwan, Malaysia and the Philippines have transitioned from being self sufficient or occasional exporters, to becoming import dependent.7 Over the past two decades, production has slowly shifted from countries with higher- cost production 8 generally located in the Northern Hemisphere to the countries with lower costs of production primarily located in the Southern Hemisphere.9 At the same time, production has increased sharply in China, Thailand, India and Pakistan. Over the past two decades these four countries have transformed themselves from being large net importers to being large net exporters.10 Sugar is a basic food product. It is consumed in every country and produced in most countries. Indeed, almost 140 countries produce sugar of their own. Aggregate world demand for sugar increased by 63 percent between 1984/5 and 2002/3 while aggregate world supply increased by 57 percent.11 Still, stocks over this period increased by only 46 percent. The run up in stocks, moreover, had only a marginal effect in pushing prices down because most of the stock accumulation took place in India where relatively low product quality makes it difficult to sell off the surplus internationally.12 In spite of extensive residual trade protection sugar is widely traded and, indeed, it is traded in increasing volumes. Approximately, 32 percent of total world production was exported in 2003. 13 Moreover, the global sugar market is quite sophisticated with multiple hedging and risk management instruments available to participants. Still global prices remain volatile, in part, because of the legacy of national market protection and, in part, because of the significant differences in production cost which persist among countries. When incremental demand evokes supply responses from different parts of the global supply curve, some parts of which remain heavily subsidized, big demand differences can translate into big price jumps. Over time, incremental differences in growth rates between global production capacity and global residual demand can also have major effects on price levels, as well. However, it still remains true today just as it was two decades ago that it is the trade and agricultural subsidy policies of major producer and consumer countries that have the largest influence on volatile global prices. These are relatively few in number and 6" Changes in the World Sugar Situation," www.fas.usda.gov, ">, Robert Knapp, USDA 7Ibid 8Especially from the Caribbean ( e.g. Cuba and Jamaica) and from North America 9Especially from Brazil, Australia, South Africa, Zimbabwe and Swaziland 10Knapp 11Larson 12Knapp 13Larson today include China, India, Brazil, the EU and the US.14 Changes in the internal trade and production policies of these countries are quickly reflected in market price expectations. Indeed, among the largest producers of sugar only one, Australia, has adopted a liberal trade policy. In spite of recent moves within WTO to liberalize global trade in farm products, sugar remains among the most protected agricultural commodities. With all of this said, new information concerning price influencing factors are continuously and efficiently being discounted into the price of sugar, through sophisticated risk management instruments which operate in the global market. . The primary effect of global market prices on the sugar supply chains in Tanzania is through the special access arrangements for which Tanzanian producers qualify. The EU provides quota based access to several African countries, including Tanzania. The Tanzanian quota under the EU protocol is 11 thousand tons annually and under the EU SPS Basic allocation an additional 2 thousand tons.15 These quotas are quite valuable to producers. They allow Tanzanian producers to sell their product at the same protected price levels enjoyed by EU producers. This level substantially exceeds both the level of global prices and prevailing prices within Tanzania itself. Tanzania enjoys no US quota allocation, unlike most of its neighbors to the South.16 3.1 b Regional and Local Markets The individual national markets within East and Southern Africa define the primary context for the development of sugar supply chains in Tanzania. The nine countries which make up East and Southern Africa produce substantially more sugar than they consume. As we discussed above, some of these countries are among the fastest growing sugar exporters in the world. However, two of the East African countries--Kenya and Uganda--- remain the primary net importers within the region. Kenyan imports in 2002 were 149 thousand tons and Ugandan imports were 50 thousand tons. The supply deficiency situation in both Uganda and Tanzania is likely to remain as long as public sector control over the sugar industry and partial public sector ownership remains in effect.17 Moreover, the productivity of the various national cane sectors in East and Southern Africa differs significantly both with altitude and climate and with industry structure. Thus, in Uganda and Kenya cane growth maturity averages 15 to 17 months in the highlands where it is principally grown. Moreover, yields in the rain fed conditions which prevail in both countries generate 75-80 tons per hectare and plants produce only three ratoons there. In contrast, cane is harvested after only 12 months in Southern Africa, it yields 100 tons per hectare and a plant customarily produces 10 ratoons.18 All of this translates into higher unit production costs for East African sugar than for Southern African sugar. The East African industry includes seven factories in Kenya, three in Uganda and five in Tanzania. See the table below which describes the steady state production levels in 14Knapp 15Innes 16Innes 17" Kilombero Sugar Company and the Tanzanian Sugar Industry," Institute of Financial Management, Salim Haiji, Chuo Cha Usimamizi Wa Fedha, 18Innes Uganda and Tanzania for the past six years. From this chart it is clear that Tanzanian producers have been growing steadily since industry privatization in 1999. 600000 500000 400000 Kenya 300000 Tanzania, United Rep of Uganda 200000 100000 0 1998 1999 2000 2001 2002 2003 Tanzania itself has historically also been a net importing country, with most of its imports being purchased by food processors and taking the form of reduced tariff "industrial sugar." However, as we noted above, with the privatization and modernization of Tanzania's production base beginning in 1999, the " gap" between demand and supply had begun to close within Tanzania. Estimates vary with respect to when supply will match demand.19 However, clearly this is the direction in which the local industry is moving. Tanzania Raw Sugar Production 180000 160000 140000 120000 100000 Production 80000 60000 40000 20000 0 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Year 19 Interviews with industry management suggest that the point of self sufficiency may be as close as 2006 or 2007. Clearly more sugar production is coming on line in Tanzania. Of the five refineries which exist within the country only four have actually been in production over the past four years. The fifth is out of service pending its financial closure subsequent to its sale to a private company and the rehabilitation which has been committed as the part of the new investor. The table above represents production of cane sugar within Tanzania. On average this production has increased at a rate of 4.5% per year over the past six years. Approximately 15% of this total production is expected to continue to be exported to the EU under the quota program described above. The rest is available for domestic consumption. National consumption of sugar in Tanzania is estimated to be around 289 thousand tons annually ( based on 20 per capital kg/yr.) However, "real consumption is hard to estimate precisely" as a great deal of sugar enters the country illegally across Tanzania's open borders.20" Over the next few years demand is expected to increase in line with population. Of total domestic demand, 20 thousand tons is refined sugar for soft drink bottlers and processed food manufacturers. This represents a market segment which is not currently served by domestic producers. A small supermarket chain segment has developed in recent years21 and this segment has evolved investment in packaging, new product development and brand equity. Most of the market, however, is the traditional bulk market for brown sugar which controlled though a multi-tiered buyer/ reseller network. Indeed, the critical factor which will influence the industry's future development is what happens once a balance of demand and supply is achieved within the local market. At the moment production capacity remains below local market demand. However, as discussed in the footnote below this gap is narrowing and may be closed as a result of the continued restructuring, productivity enhancement and plant expansions, all of which are following from the privatization of the industry.22 One case in point is Kilombero Sugar Company whose production has gone up from merely 29,517 tons in 1997 to 98, 400 tons in 2002. This represents a production increase of 233 per cent in the five years since that company's privatization. Today, Kilombero Sugar is the biggest sugar producer in the country. It is projected to produce 105,000 tons in the 2003/04 season. 3.1.c Industry Structure 20See "The East African", Monday, May 15, 2000. 21Currently the supermarket market and restaurant segments combined are less than 10,000 tons. 22Production estimates supplied by management of sugar processors with whom the study team interacted. Also, projects of future sugar production capacity developed in " Infrastructure Development and Improvement Project in Outgrowers Cane Fields in Kilombero and Mtibwa," Doto D. Mahinga, AIMS Development Consultants Trust, Jan 2003. According to both sources domestic market break even is likely to be achieved in 2005/6 Year 2004/5 2005/6 2006/7 2007/8 Output with Infrastructure 247,500 321,745 434,313 521,236 Improvement Sugar cane production is concentrated mainly in three regions within Tanzania: Morogoro, Kagera, and Kilimanjaro. These regions are well suited to cane cropping due to their climatic conditions and soil as well as the abundant sources of water which is required for sugar processing. Currently, most sugar cane is grown on estates which the large sugar processing factories own and which are part of the state owned legacy which they inherited.. However a growing share of their factor inputs comes from contract growers who produce cane on their own lands which are located within close proximity of the refineries with whom they have developed stable supply chain relationships. A monopsony relationship exists between local small scale farmers within a region and the refinery they supply because of the very large inbound logistics costs which are absorbed As described below the three largest factories in Tanzania are located in Morogoro and Moshi. All of these refineries produce raw brown sugar. No white sugar is currently produced within the country. With that said, one of the refiners, TPC, has announced plans to begin to produce white sugar in the near future. Kilombero Sugar is also exploring this possibility. Owners/ Location Annual Current Approximate Company's Investors Production Capacity Percent of Name Capacity Utilization Cane from Smallholders KILOMBERO ILLOVO MOROGORO 150,000 130,000 44% tons TPC SUKARI MOSHI 100,000 87,000 30% INVESTMENT tons COMPANY MTIBWA METL CO. MOROGORO 80,000 66,000 26% LTD tons Kilombero Sugar actually operates two refineries which are located near each other, in the South Central part of Tanzania. It also owns and operates large storage facilities in both in Morogoro and in Dar. Kilombero Sugar emerged from privatization in 1999. The company sells its sugar through out the country, except in the Kilimanjaro and Arusha regions, where TPC is based. Transport costs into this region make it difficult for the two Morogoro refineries to compete with TPC in its" back yard" region. Kilombero markets its sugar under its own brand, " Bwana Sukari" which has become a mark of distinct quality within the national market, even if the company's brand does not find its way onto all of the retail offerings which emerge from its retail supply chain23. Kilombero has been the leader within the industry is taking control of distribution channels and product positioning within those channels from traditional wholesalers. The company has also pioneered in selling one kg packages as well as other specialty size packages through supermarket and restaurant channels . The company's management comes from one of the most efficient international sugar producers--Illovo--which is headquartered in South Africa. The management team is knowledgeable and effective. They understand what needs to be done to successfully reposition the company and they are making significant progress. 23Haji TPC emerged from a privatization transaction which was completed in 2000. The company is wholly owned by Sukari Investment Co. Ltd which is a Mauritian based holding company with extensive holdings in the sugar industry. TPC produces good quality sugar which is comparable to that of Kilombero.24 TPC typically matches Kilombero in price particularly in the large Dar es Salaam market. TPC sells its product only in 50 kg bags and has not been as active in marketing differentiated products or in developing new channels. Mitbwa emerged from privatization in 1999/2000. It is located between Dodoma and Morogoro near the two Kilombero factories. For this reason, Mitbwa and Kilombero Sugar serve the same set of geographic markets and have comparable outbound logistics costs. Mitbwa's sugar quality, however, falls below that of Kilombero. Its sugar is generally darker and more moist. Mtibwa has been forced to price its product at a slight discount below that of Kilombero. However, the company is learning quickly. It is beginning to diversify its packaged offerings, adopt a distinctive brand and improve the quality and appearance of its product. Cane inputs have two sources in Tanzania: i) farm estates which are owned farmed directly by the processing companies and ii) contract growers. All three companies are moving from primary reliance on the first source of cane supply to the second. Approximately 100,000 hectares are dedicated to cane cultivation within Tanzania and more than one million tons of cane inputs are produced and harvested annually. Currently, contract growers produce about 35 % of this total. Approximately 10,000 small farmers supply cane to and are located near one of four plants identified above.25 As noted the remaining supply of cane comes from factory owned estates. The " big three"--- Kilombero, Mtibwa and TPC-- currently account for 95% of total sugar production in Tanzania. Kilombero has the largest market share with 44% . TPC has 30% and Mtibwa has a 23% share.26 All three companies have developed similar business models and similar supply chains. Except, for differential interest and overhead costs their cost structures are also quite similar27. As noted below, the government has been able to realize a substantial benefit in the form of income taxes from the newly privatized companies. Year of 2000/2001 2001/2002 Name Divestiture She. Mill. She. May. Kilombero Sugar Company 1998/99 1,590.0 570.7 Mtibwa Sugar Estates 1999 889.3 892.5 TPC 2000 na na Income Taxes Paid to the Government of Tanzania 3.1.d Development of Distribution Channels 24Ibid 25Source Interview with Director of Planning Tanzania Sugar Board (TSB) 26Institute of Finance and Management and" Sugar Industry Development Plan 2000-2010", TSB 27Kilombero Sugar is more highly leveraged both financially and operationally than its two prime competitors. Since taking over production and storage facilities, all the new companies based in Tanzania have been busy building up nationwide dealer networks. Through these networks, sugar consignments are distributed primarily through independent agents. Sugar is bagged at the factory mostly into 50 kg or 25 kg lots and stored either near the factory or in the major consumer and food processing areas e.g. in Dar es Salaam or Morogoro). Most sugar is sold to Asian traders who typically trade in several different food commodities in addition to sugar. Larger wholesalers sell the product through their own networks of smaller regional wholesalers and retailers. Independent wholesalers provide an important function to the industry. They are its primary source of working capital. Most sales are on a cash and carry basis at warehouses owned and/or controlled by refiners.28 At the retail end of this traditional distribution channel sugar is sold in non standard quantities on a product weight basis in response to the specific demands of individual customers. Retailers who have purchased 50 kg sacks of sugar sell it in plain brown bags and in small quantities to their customers. The price of sugar on the local market ranges typically between 600 and 700/= Tsh per kilogram. The retailers supply the final consumer and their prices differ due to different factors such as: quality of the product, costs incurred in transporting sugar to the shops and outlets as well Retailer and wholesalers Price per kilogram Local vendors/Retailers 600/= for local sugar 400-500/= for Malawi and Zambia sugar Wholesalers 430-450/= for super sembe Shop Rite 700/= for local sugar 700/=-1500/= for imported sugar brands from South Africa However as we noted above, the refiners are attempting to strengthen their own managed distribution channels so that they can more effectively leverage the brand images they are building, the market positioning they are attempting to effect and the relative price positioning that they are attempting to impose on the retail market. In addition to selling into the local market, refiners also export a small volume of sugar. Most export sales are managed through refiners own sales forces or through foreign sales agents. Sugar is produced during a six month production window when cane sugar is available. Factory production is scheduled to meet a sales forecast and most production is scheduled "for inventory." Sales are made from storehouses whose stocks built up during the production season and fall during the off production season. Production planning is executed against an annual sales forecast and production scheduling is determined both by current inventory levels and continuously updated demand forecasts. 28Haji Because of the seasonality of production, the working capital requirements of sugar producers are great. Large expenditures of working capital must be made during the production season and working capital gradually recouped through sales during the ensuing months. Production scheduling under such circumstances can be risky. Production scheduling controls based on inventory levels at the upstream end of the supply chain and demand forecasts can be quite uncertain and unstable. The control system works this way: When inventories accumulate faster than expected given demand forecasts, production is slowed and conversely when inventories fail to accumulate as expected production is accelerated. However, because several constraining factors combine to retard production increases in the short term--e.g. plant capacity, yield of local cane, storage capacity---it is easier to reduce the system's production than it is to accelerate it. During the off-season, sugar prices within the country typically rise. Both Tanzania food processors and merchants import " industrial sugar" primarily from both Southern Africa and from Europe. Most of the storage and transport of imported sugar takes place in and around Dar. The country still suffers from hoarding and from illegal importation. Sugar smuggled from Malawi and Zambia, in particular, is quite common. A separate black market distribution channel has formed to support illegal smuggling and hording activities. . 3.1.e Specialized Service Providers Highly specialized services are required to support the sugar industry and an "external market" is emerging for these services within Tanzania. Since privatization in 1999/2000 several highly specialized third parties have emerged to support the sugar processing and distribution industry. Among these are specialized high way transport companies, such as DT Dobie and F.M. ABRI Transporters. Both carriers operate road equipment, which is specially designed around the loading and handling characteristics of sugar. Representative prices for these services are summarized below. Third Party Highway Transportation Costs: Sugar to Dar es Salaam From To Costs/ton in Tsh Morogoro Daresalaam 30,000/= Arusha Daresalaam 70,000/= Zanzibar Daresalaam 50,000/= Source: Local transporters at Jangwani in Dar es Salaam Other specialize service providers include operators of food grade public warehouses such as FIDA Hussein Importers & Exporters. Public warehousemen provide both storage and inventory securitizing services to sugar processors. As we noted above the sugar distribution channel requires a great deal of external financing. Hence securitization and third party storage of sugar inventories is useful in creating bankable collateral. Still other specialized third party service providers include inbound logistics and harvest collection and delivery service providers. Kilombero Sugar, for example, outsources to a South African company the collection, bulking, and inbound transportation of sugar cane into its two sugar processing plants during the harvest season. This specialized third party uses large off vehicles, specialized bulking and handling equipment to transport cane quickly and efficiently from the gathering points in the surrounding cane fields to Kilombero's refineries. This specialized service provider effectively programs and manages the entire operating interface between small-scale farmers and the factory. It brings in the harvest and in the process realizes higher yields . The most notable third party service failure is on the part of Tanzania's two railways which both provide essential transport services between the refineries and wholesale markets. All of the refineries are rail served. Neither the car supply nor the transport service reliability of either rail carrier is adequate to support the needs of the sugar processors. The failure of the government to recapitalize and to periodically renew the asset base of these two rail carrier has continued for some time pending the execution of plans for privatization which have not been realized. In the mean time, from the perspective of sugar shippers the point of complete service collapse does not appear to be far off. Still other specialized service arrangements which support sugar supply chains are worth discussing in more detail. .Delivery of industrial sugar to the food and beverage industries such as SBC Company Ltd (bottlers of Pepsi), Azam, and other confectionery industries is usually from the sugar warehouses in Dar es Salaam. The cost of this local delivery service within the Dar commercial zone is about 10,000/= per ton. It should be noted that industrial sugar is distributed to the food and beverage industries "in bond." Most of this sugar is imported as we noted above and is duty discounted. Hence it remains in bond until it is used. Specialized bonded warehousemen support procedures for release of industrial sugar imports prescribed by Tanzania Customs. Specific processor consignees are authorized by government to use imported industrial sugar which is duty discounted. This circumstance--effectively a distribution monopoly controlled by the processors themselves--allows processors though their third party agents--- the bonded warehousemen--- to maintain relatively tight control over the tariff discounted sugar. In other African countries which have recently privatized their sugar sectors ( e.g. Mozambique) effective control over industrial sugar has proved extremely difficult to implement. Outside Dar transport costs largely depend on the distances traveled and also on the modes of transportation which may either be road or rail. Thus, for example, moving sugar from Dar es Salaam to Mwanza by rail costs 1,888,800/= Tsh per 40 ft ton container. The container has a net weight of 30MT. Rail is largely used in this corridor due to the poor road conditions between major market areas and smaller towns.. The export of sugar by local producers is restricted to approximately 15,000 tons annually through the government managed EU quota program. In terms of physical distribution this program works the same way that any other export regime would work with the straight forward addition of documentation requirements which certify that the shipper enjoys the EU tariff exemption assigned to him by the Government of Tanzania. 3.1.f Product Flows The map below represents the primary sugar product flows which take place within Tanzania and which are facilitated by the physical distribution services described above. Principle Sugar Flows Within Tanzania Key: Major producing area Supply to major market areas Major market area Distribution from major Markets to regions and export Markets Importation of Sugar Key points for illegally imported sugar 3.2 Economics and Operation of the Supply Chain for Sugar This section deals with the operations and underlying economics of the sugar supply chain. 3.2. a Supply Chain Map The schematic diagram below represents the sequence of process steps and value adding agents that collectively define the supply chain for sugar. The remainder of this section discusses the underlying economics of this chain. Contract Growers Company Estates (SPF) Refinery Storage Distribution to Distribution to Export Import wholesalers industries Retail and chain stores Retail Customers Contract and estate growers incur different production costs. Both, however, undertake the same set of value adding activities, which include the following: · Land preparation . Land preparation costs depend on the type of farm and plant materials being used. However, farm preparation costs approximate 20,000 to 30,000/= Tsh per acre. · Cultivation. Cultivation involves either mechanised or non-mechanised techniques. Contract growers typically use non-mechanized techniques which cost about 10,000 to 15,000/=Tsh per acre while estate growers typically use mechanised techniques which cost approximately 20,000/= Tsh. · Harrowing. Harrowing involves pulverising and then smoothing the soil. This process is typically mechanised and costs approximately 20,000/= Tsh per acre. · Fallowing. In being fallowed the soil is left to rest without being planted or cultivated. On average contract growers spend 13,000/= Tsh per acre compared to an approximate 20,000/=Tsh per acre for estates growers. · Preparation of seeds: The preparation of seeds is an important enabling activity that must be carried out before planting. To prepare the seed, cane must be cut into seed stock and then loaded as well as offloaded before it is planted.. This costs approximately 31,680/= to 61,680/= Tsh per trip for 4 tons per acre. · Application of fertilizers: Before planting fertilizers such as DAP, TSP, UREA need to be applied. Fertilizers cost about 310/=/kg and about 60kgs are required per acre. · Planting . Planting is done in farrows and these cover about 53m2 per acre. Planting costs about 31,800/= Tsh per acre if planting is mechanised and approximately 21200/= Tsh per acre if it is not. Sprouting sugarcane · Application of pesticides: Pesticides such as Gramaxon and Durron are required to fight common pets that adversely effect cane growth. They cost approximately 15,000/= per gallon and a minimum of 3 gallons are required per acre. This is followed by weeding which costs about 10,000/= to 20,000/= Tsc per acre depending on whether mechanisation is used. · Harvesting: Harvesting takes place approximately 12 months after planting. Harvesting involves cutting the cane, loading and transportation it to the factory. In the table below each of the farm based activities that go into the value chain are summarized and compared on a cost per acre basis for mechanized and non mechanized processes. Note that the non mechanized cost is substantially lower than the mechanized costs. This explains the strong preference of the refiners for increased reliance on outgrower schemes. Activity Mechanised: Cost/acre in Non-mechanised: Cost/ acre in Tsh Tsh 1: Farm preparation 30,000/= 6000/= 2: Cultivation 20,000/= 10,000/= 3: Harrowing 20,000/= 4: Farrowing 20,000/= 13,000/= 5: Transportation of seeds 61,680/= 31,680/= 6: Application of fertilizers 18,600/= 18,600/= 7: Planting 31,800 21,200/= 8: Application of pesticides 45,000/= 45,000/= 9: Harvesting 30,000/= 10,000/= 10: Hiring land 130,000/= 130,000/= Total cost 407,080/= 285,480/= Source: Contract growers, and Agricultural manager Kilombero Sugar Co. Ltd Transportation costs of the cane from the fields to the factories largely depend on distances to the factories, as the table below respesents. Distance to industry in kilometres Cost per ton 0-10 km 2720/= 11-20 km 2980/= 21-Above km 3470/= Source: Agricultural manager Kilombero Sugar Co. Ltd Specific refiners costs vary from factory to factory and, indeed, are a primary source of competitive advantage for producers. . Hence precise information about the operating costs of specific plants is confidential . Moreover, specific companies based in Tanzania have quite different capital cost structures, overhead cost burdens and hence different profit/loss break even points. Moreover, as we noted above factory and farm level costs are jointly determined and hence vary together across a spectrum of levels determined by parameters whose discussion goes well beyond this immediate need to understand. With all of these caveats, a set of activity costs are presented below which represent approximately the variable costs incurred by an efficient and competitive factory currently operating in Tanzania. During the course of the study the management of one of the major refineries assisted the study team in developing these estimates which are summarized in the table below. Variable costs incurred in the processing of sugar Item Cost/kg Processing (Variable Supervision , Maintenance, Plant 20/= Replacement Costs, etc. Labor 3/= Power 80/= Packing material 3/= Ingredients/chemicals 100/= Storage 10/= Total 216/= The reference plant produced more than 90,000 tons of sugar in 2002/2003. It management has plans of increasing its production to 105,000 tons by the end of the 2003/04 season . For this facility the variable cost is approximately 216/= Tsh to process a kilogram of sugar. This figure includes storage costs of about 10/= Tsh per kilogram incurred at the factory premises. Kilombero Sugar Factory 3.2. b Physical Distribution As discussed above, sugar consignments are shipped directly to wholesalers, sugar processing industries as well as exported to foreign customers. Most sugar produced is transported to Dare salaam for storage and subsequent re-distribution to the various customers within the national market, including food processors. During the off- production season, sugar prices within the country rise in order to meet administrative and storage costs in Dar es Salaam. Individual refiners /distributors price their product seasonally based on estimates of remaining supplies and expected demand during the remainder of the off production season. Frequent sugar shortages occur during the year when factories close for major repairs or are delayed in beginning new season processing because they are awaiting the arrival of canes. During these periods prices tend to spike. According to the primary sugar transporter in Tanzania, D.T.Dobie, transportation to Dar from Morogoro costs about 30,000/= per ton by Road. Additional warehousing or storage costs are incurred in Dar es Salaam costs of about 450/= per ton. Additional handling costs (loading and unloading costs) are incurred with each modal transfer of about 3000/= per ton of sugar. For different sources of supply outbound logistics costs are summarized in the table below From To Costs/ton in Tsh per ton Morogoro Daresalaam 30,000/= Arusha Daresalaam 70,000/= Zanzibar Daresalaam 50,000/= Outbound transportation and logistics costs Source: Local transporters at Jangwani in Dar es Salaam Final delivery customers in the processed food and beverage industries such as SBC Company Ltd (Bottlers of Pepsi), Azam, and other confectionery industries is usually completed from warehouses located in Dar es Salaam. This local delivery costs are about 10,000/= per ton. Outside Dar delivery costs largely depend on distance and mode of transportation which might either be road or rail. Thus, for example, sugar shipped from Dare to Mwanza by rail will cost 1,888,800/= Tsh per 40 ft ton container with a net tare weight of 30MT. As the two exhibits presented below clearly demonstrate significant economies of scale exist not only in production but in physical distribution as well. Selling sugar in 30 ton 40ft container lots as constrasted with 15 ton 20ft container lots halves the unit cost of transport. The cost of transportation for sugar by Rail: 40 ft container with capacity From To Freight Distance in Cost per Time rate/ km km ton Dar es Kapiri 1.4 $ 1860 2604 $ 5 days Salaam Mposhi to carry 30 tons Source: TAZARA Daresalaam The cost of transportation for sugar by Rail: 20 ft container with capacity From To Freight Distance in Cost Time rate/ km km Dar es Kapiri 0.7 $ 1860 1302 $ 5 days Salaam Mposhi to carry 15 tons Source: TAZARA Daresalaam, Note that in both cases the areas shown include inland and cross boarder towns, e.g. Kapiri Mposhi (ICD) in Zambia 3.2.c Supply Chain Cycle Time The discussion which follows explains the timing and sequencing of acativities within the supply chain. From land preparation to harvesting a number of activities are completed and each different times as the table below explains. Sugar production is highly seasonal. Activity Time 1: Land preparation October 2: Cultivation November-December 3:Harrowing ** 4: Farrowing ** 5:Transportation of seeds December- January 6: Application of fertilizers March 7: Planting December-January 8: Application of pesticides February 9: Harvesting After 9-12 months from planting The farm production cycle in Tanzania requires a total of 12 months between initial land preparation and when sugarcane can be transported to the plant for processing. Only one harvest occurs per year in Tanzania. As we noted below this long farm production cycle is a source of competitive disadvantage vis a vis production bases in Southern Africa. Factory processing time, on the other hand, is very brief. Sugar is processed in batches and the final product bagged and stored. The production and bagging of a single batch requires less than one working day. It is the capacity of the factory and its match to the production of neighboring cane production which determines how long may be required to process an entire harvest . Transportation of finished product to distribution centers and to wholesale markets in Dar requires on day if transport is by road and five days plus or minus 3 days if it is by rail. 3.2. d Storage points along the Chain. Sugar has some special storage requirements: It should be kept in cool and dry conditions to avoid spoilage. Sugar as noted above is susceptible to spoilage from water due to its deliquescence. Its is this property that makes proper storage important. Storage of sugar occurs at specific stages along the supply chain. Thus, sugar is stored at the sugar processing factory. Warehouses are located at the factory premises. These factory are relatively small in size compared to the warehouses which are located in the major wholesale markets and are used to disconnect production and shipment processes within the chain by allowing buffer stocks to build up. Sugar storage warehouses at the Kilombero sugar factory, for example, cover approximately 2000m2. The facilities at the factory are inset to avoid spoilage. .The sugar producing companies operate large distribution centers in both Dar es Salaam and Arusha. Kilombero, for example, fills orders and ships all over the country from its distribution center in Dar. TPC does the same from its distribution center in Arusha town. Wholesale agents also have storage facilities and these are located in regions of country were they sell. Regional storage facilities for large buyer/reseller merchandisers typically are 400-600m2 in size. These warehouses serve as the sales point for local retailers who buy and take delivery at local warehouses. Abundant storage also exists for illegally imported sugar. This is, for the most part, in residential housing facilities, in Tunduma and Kyela, where illegally imported sugar from Malawi and Zambia is stored. Also other storage points of illegally imported sugar include the island of Pemba, as well as villages in Tanga and Bagamoyo. 3.3 Factors Effecting Supply Chain Competitiveness Several factors determine the competitiveness of sugar distribution within the Tanzania market and beyond. Most importantly, these include the commodity characteristics of the sugar cane which is produced within the country, the country's domestic trade policy, the incidence of smuggling, the deterating state of local infrastructure , backward linkages from refineries and the adaptability and quality of service provided by specialized third party service providers. 3.3.a Product Characteristics As already noted sugar cane is the primary input into the sugar refining process in Tanzania. Raw cane contains between 10 and 15% sucrose. The actual level depends on the hybrid plant material which the farmer plants and the cropping methods which he applies. The remainder of the cane is primarily water with fiber and soluble impurities making up approximately 14% and 3 % by weight respectively of the input. The refining process involves separating sucrose from the rest of the extraneous material and, in the case of industrial sugar, coloring the output white. "Refining" actually entails several process steps. Specifically it involves crushing, spinning, boiling and drying. Some refining processes are more efficient than others in extracting sucrose. Refining requires large volumes of water to complete the extraction efficiently. Energy costs are also significant in processing sugar. However, all of the refineries in Tanzania generate their own electricity and produce their own process steam by incinerating the fiber from the cane . The three most commercially significant attributes of the cane which is harvested for processing are: i) its sucrose concentration, ii) its shelf life and iii) its density. Cane is burned first in the field in order to reduce the amount of extraneous material which is brought into the refinery. The more concentrated the sugar content of the cane the less the unit cost associated with inbound logistics and processing. Through field burning the sucrose content of cane sugar can be raised to 5%. Importantly, sugar cane has a limited self life once it is harvested. The sucrose concentration in the fresh cane begins to deteriorate if the raw material rests too long in the field. Delays between cutting and refining of much more than 24 hours can result in reductions in the sucrose yield of the final output. Thus, forward scheduling and close coordination between harvesting, bulking, inbound transportation and refining activities is critical for productive end to end processing. Once cane inputs begin to arrive at the refinery early in the production season the plant operates continuously around the clock until the entire harvest is complete and the entire crop refined, bagged and stored. During this season, production planning must be closely synchronized with planning for field cutting, bulking and shipping of cane from the field to the plant. Two additional attributes of refined sugar directly affect the efficiency with which it moves upstream through its supply chain. These are the density of the finished product and losses--in the form of spoilage, pilferage and waste--- which are associated with handling, storage and transportation. All refined products which are produced at the refinery are bagged for outbound transport from the plant. The density of raw sugar produced in Tanzania is approximately 50 kg/ cu. meter. Sugar is a relatively dense product and in all modes of transportation in which it moves it weights out before it cubes out. Moreover, most raw sugar moves via its primary transport leg in 50 kg sacks and most transfers are manual. All of the finished product is packaged and some of it is branded under one of two distinct brands---" Bwana Sukari" and "KK Sukari. " The quality of the packaging is generally good in Tanzania and losses are correspondingly low. The value to weight ratio of sugar as it moves via long haul transport to urban markets is 70Tsh/ Kilo. This is relatively high compared with other agricultural commodities which are transported long distance within Tanzania. This high value makes the sugar supply chain less transport price elastic than it would otherwise be and is one of the reasons that a mostly integrated market exists for sugar that spans the entire country with the exception of the Northern border with Kenya as wed noted above. The value to weight ratio directly effects the choice of transport mode through its effect on the trade off between opportunity costs related to transit time and transport cost. High value commodities like sugar typically move in the smaller economic lot sizes and via the most expedient mode of transport even if that mode is more expensive. Opportunity costs associated ith lost sales and additional working capital needs outweigh marginally higher unit transport costs. In Tanzania the optimal transport mode is truck in some corridors and sometimes rail depending in others. Kilombero and Mtibwa sugar tend to be more rail dependent because the quality and reliability of truck transport is less in the central corridor where highway infrastructure is less developed than in the north 3.3.b Domestic Trade Policy Industrial sugar products enjoy preferential tariff treatment in Tanzania. " Industrial sugar" refers to the intermediate input of refined white sugar which is used extensively to produce food, beverages and other consumer products. Local food processors have traditionally been able to purchase this key input outside Tanzania in global sugar markets for significantly lower prices than were available within Tanzania . The functional tariff exemption they enjoy is intended to prevent food processors from being disadvantaged competitively. However, the trade policy compromise which the government maintains between sugar manufacturers and industrial sugar users will soon be tested as the new private owners of the nation's large scale sugar refineries increase production to levels where it begins to meet or exceed domestic demand for raw sugar. At that point leaders in the newly privatized industry would prefer to open domestic industrial sugar markets or advance into liberalized regional markets, particularly into Kenya, instead of precipitating an internecine price war and then even possibly an industry shake out. Indeed, this policy "tipping point" is not far away. Importantly, the competitive position of Tanzania's sugar producers is also effected materially by the regional trade agreements into which Tanzania has entered. The government of Tanzania, of course, entered into these agreements with the expectation that its economy would realize an overall net benefit as a result of open trade. However, the effects of open trade are highly uneven on a product by product basis. Indeed, regional trade commitments can effect the viability of entire industry sectors. Regional commitments on sugar are a particularly good example of the tough decisions and dire adjustment consequences which may result from freer regional trade. The table below reviews the current tariff levels which apply on sugar imports and exports, to and from Tanzania's primary regional trading partners. TZ Imports from TZ Exports to Country Prevailing Tariff Country Prevailing Tariff MFN Effectively MFN Effectively rates applied rates applied rates rates Kenya 25 5 Malawi 25 Uganda 25 Mozambique 7.5 DRC 25 Swaziland Malawi 25 5 Zambia 25 Zimbabwe 25 Source: UNCTAD TRAINS Database. This ex ante circumstance appears to provide strong protection for manufactures within most countries in the region and to reinforce strong firewalls between national markets. However, this set of conditions may be subject of substantial change in the future as the work of regional trade integration proceeds. As the graph below further demonstrates, individual trading partners within the region may be signatories to more than one of the regional trading agreements and this circumstance creates some additional complication with regard to the venue and instruments for future regional market integration, to the say nothing of the pace of that integration. Thus, Tanzania may have some difficulties meeting its dual commitments both to SADC and to EAC. Tanzania is caught on the horns of a dilemma when it comes to regional trade policy. Its sugar industry is arguably more productive and faster growing than that of other EAC countries. Significantly, its EAC neighbors imported more than 200k tons per year in 2002 and thus provide an attractive growth target for Tanzanian producers.. On the other hand, Tanzania's fledgling sugar industry is arguably significantly weaker and higher cost that that of its SADC neighbors which include some of the faster growing sugar producing economies in the world. They include, for example, Malawi, Swaziland, Mozambique, Zimbabwe and Zambia. These five countries exported more than 615 K tons of raw sugar in 2002, in addition to their quota allotments to the EU and the US. The way in which tariff policies and trade regulations evolve and the direction in which policy makers in Tanzania tilt---either toward the EAC which is likely to champion a "go slow" liberalization policy or toward the SADC which is likely to prefer a "go fast" policy will greatly influence the development of the Tanzanian sugar sector. Clearly that sector is at a cross roads. A real threat exists that regional trade agreements will force open the Tanzanian market prematurely and in the process undercut domestic prices. Such developments would not only diminish the value of the assets which have recently been privatized but would also diminish the attractiveness of additional investment in the sector. 3.3.c Incidence of Smuggling The smuggling of sugar into Tanzania poses a second significant treat to continued industrial development. Like premature trade liberalization, smuggling tends to diminish the value of the assets which have been transferred through privatization and undercuts incentives for additional investment in plant, equipment and out grower programs. All of Tanzania's borders are porous but those which separate it from neighboring low cost sugar producing economies in Malawi, Zambia and Mozambique are particularly porous. To help combat smuggling, the Tanzania Revenue Authority has directed that the size of vessels carrying imported sugar all be larger than 250 metric tons. Such big vessels can only anchor in major ports. This policy has been partially successful in steming smuggling activities. However, the practice is deeply imbedded in Tanzania's commercial culture and it continues to the detriment of private sector investors in the sugar chain. Sugar from Malawi being smuggled into Tanzania Smuggled sugar at Kyela ready for distribution Significantly, the government has written and the parliament is currently reviewing an Anti Dumping Act. This Act which is expected to be enacted in the fall of 2004 is designed to further check the flow of cheap sugar imports into the local market. 3.3.d Infrastructure and Community Development Producing sugar competitively requires a combination of intensive farming, community development, industrial coordination, and minimum local road and drainage infrastructure. Because sugar production is labor intensive it provides an ideal vehicle for poverty alleviation. Indeed, when all of the ingredients noted above are present in sugar projects, they can provide an ideal platform for pro-poor growth. However, when one or more indedients is missing opportunities for pro-poor growth cannot be fully realized. A key missing ingredient in Tanzania is local infrastructure. When each of the privatization transactions noted above were finalized with the Government of Tanzania each agreement included a delineation of responsibility for the provision of public goods within the local sugar producing community. These public goods included most importantly local roads, bridges and drainage ditches. In addition, each agreement specified the basis on which the newly privatized facilities would be taxed. The new private owners insisted that their tax liability be explicitly defined and constrained for a period of years after the privatization. Key issues which remained unresolved at the time of financial closure, however, involved the mechanisms for applying tax proceeds to the maintenance and new development of local infrastructure. New rural road development has become increasingly important as the demand for cane continues to increase and as out grower programs are expanded to accommodate this new demand. As discussed in a subsequent section local district governments have not succeeded in either maintaining or improving local infrastructure. The rapid deterioration of this infrastructure has become a significant constraint on future growth. . 3.3.e Backward Linkages from Production Facilities The first of a set of collective bargaining agreements was successfully negotiated in the spring of 2004 between the local farmers cooperative of Ruembe Out growers Association (ROA) and the Kilombero Sugar Company. This agreement complies with the statutory requirements mandated in the Sugar Act of 2001. Importantly it also sets a precedent and establishes a pattern for subsequent collective bargaining agreements. The agreement is binding on both sets of parties for 4 years. It defines a basis for compensation for the delivery of raw cane sugar with matches or exceeds a minimum set of product specifications. The agreements defines a basis of compensation for small holder farmers who are members of the cooperate based on a minimum price for delivered sugar which is linked to its assayed sucrose content and a profit share based on a formula included in the agreement. In addition it commits Kilombero Sugar to give preference to smallholder suppliers when receiving cane input. It further obliges Kilombero Sugar to organize and manage the inbound bulking and transport process and conversely obliges small scale farmers to respond precisely to the cane pick up schedule developed by the refiner. Further, under the agreement the company provides support services to its out growers in the form of agricultural inputs including fertilizers and insecticides which are provided at cost. The collective bargaining agreement effectively makes operational and provides explicit economic incentives to small scale farm producers who work in close cooperation with the processor. Provision of the agreement are subject to interpretation and enforcement by the Sugar Board. Other processors are following the lead of Kilombero in negotiating their own collective bargaining agreements. The precedents established for linking farm level production and processing of agricultural products is a positive one and should serve as a good example of best practice. It may, indeed, have application to other farm sectors as well. 3.3.f Specialized Third Party Service Providers Highly specialized services from third parties are required to support the sugar industry. Since privatization an external market has begun to develop on the periphery of the sugar industry made up of third party service providers who have begun to train personnel, purchase equipment and/or set up operations in Tanzania. Specialized third party services have begun to emerge for example in the management of inbound sugar cane logistics, in outbound sugar transportation, in storage and handing, and in commercial financing of the sugar supply chain. All of these developments are quite positive. The most notable third party service failure is on the part of Tanzania's two railways both of which provide refinery to wholesale transport services. Neither the car supply nor the transport service reliability of either rail carrier is adequate to support sugar processors. The failure of the government to recapitalize and to periodically renew the asset base of these two carrier has continued for some time. The result are chronic car supply and locomotive shortages, slow orders and embargoes over critical sections of the rail network and a progressive diminution of service reliability. 3. 4 Factors which Effect the Competitiveness of Tanzania Sugar Producers Sugar production entails two distinct cost elements: a farm and a factory element. As we noted above the two are intimately linked together within a joint production function. Thus, the scheduling of the crop and the resulting harvest and its coordination with the production schedule of the refinery determine not only the production capacities of the combined system but its efficiency as well.. Cane planting must be planned in order to assure an even and reliable supply of cane during the entire production season. Extending that season as much as possible directly determines the required peak processing capacity and hence the investment required in fixed plant and equipment. Moreover, the time which elapses between harvesting, bulking and processing directly effects operating yields, since sucrose yields begin to deteriorate within 24 hours of harvesting. Hence, cutting, bulking, transporting and refining ideally must be done in a just-in-time mode of coordination. Importantly as well, the plant stock from which the cane is grown and the cropping methods used to grow the cane directly effect sucrose content and hence the cost of production processing. Sugar cane is a bulky product of substantial density. Unit inbound logistics costs are determined in substantial part by the loading capacity of the primary conveyance used to move cane from field to refinery. For all of the reasons cited, inbound logistics operations must be tightly scheduled in order of assure that factory production proceeds uninterrupted on a 24 hour basis once it begins and that the assets used for this tightly scheduled operation are minimal. The section which follows discusses the key factors which determine production productivity in the specific context of Tanzania's restructured sugar industry. The discussion deals with the indirect effects on production efficiency of the new institutions and industrial structures which are emerging in Tanzania subsequent to the privatization of the sector. These new institutions and the economic incentives that they reinforce have had the effect of substantially improving the productivity of Tanzania's sugar production. 3.4.a Water Supply and Soil Adequate moisture and tropical temperature are the two most important ecological pre- requisites for productive sugarcane cropping. As we noted above Tanzanian sugar cane is grown in three areas in Tanzania. All three are suitable for growing cane. The soils in Kilombero Valley are a combination of loam and clay and are good for sugar cane production. Rain fall during the rainy season typically averages 1500-1600 mm in the Kilombero Valley. The rainy season lasts for six or seven months from November/ December to May/ June . The dry season includes the remaining five to six months. The Valley has a minimum temperature which ranges between 18 to 23 degrees C and a maximum temperature which ranges between 29 and 35 degrees C. Both are ideal for sugar cropping. The soils in Mtibwa--the second major production area-- are heterogeneous and good for growing cane. Rainfall in Mtibwa is bi-modal with a six month rainy season and a six month dry season . The rainy season lasts from October to May. Average rainfall during the rainy season varies significantly from year to year between 850mm and 1800mm. However, it averages 1200mm. The minimum temperature in Mtibwa ranges from 18 to 24 degrees and reaches a maximum of 33 degrees in lowland areas. Soil and rainfall in the Ruhembe region are also conducive to cane cultivation. In all three areas, sugar cane production is primarily rain fed. The duration and timing, as well as the absolute level of the rainfall are important for sugarcane growth. Sugar cane requires ideally between 1200 and 1500 mm per annum. Too much, as well as too little rainfall, can adversely effect yield. In freely draining soils precipitation greater than 1500 mm can be tolerated but not in water retaining clay soils. For example, on the Kilombero Sugarcane Estates annual rainfall can sometimes be as high as 1500 mm per annum and most of this falls between March and May. Too much rainfall and standing water have a particularly adverse effect on cane sprouts. Over an extended period standing water and poor drainage result in increased salinity in soils and significant resulting loss in production. Hence, good drainage is vital, particularly, in the sprouting stage of plant growth. If excess water is not drained during the sprouting stage, stem cuttings called sets will rot and will not be available to generate the next season's plantings. The table below represents the yields which have historically been realized in each of the three primary cane growing areas within Tanzania and compares these yields with ones which prevail in other cane growing areas within the region. Growing Area Crop Yield Production Yield (Tons of cane/ hectare) (Tons of Sugar/ Ton of Cane) Kilombero Valley Mtibwa Ruhembe Malawi South Africa 3.4.b Local Infrastructure Local infrastructure is critically important to assure the value adding processes which take place at the farm and factory are synchronized. Local infrastructure includes rural roads which ideally are laid out in square grids around the perimeter of cane fields in order to facilitate efficient harvesting, bulking and transporting from one end of the cultivation grid to the other. Poor original design of the estate plantations in the major sugarcane growing regions poses another challenge for optimizing the way in which the harvest in planned and programmed to be brought into refineries in ways which minimize inbound logistics cost. Moreover traditional drainage systems are more conducive for mono cropping which has traditionally been practiced and less well suited for intercropping which is of interest to small holder farmers. . Unfortunately local infrastructure in all of the primary growing regions of Tanzania has deteriorated badly. While sugar farming and processing assets were owned by the state they were maintained at the expense of state supported enterprises and the losses these enterprises absorbed were in turn passed on to the national treasury. However, during the last years of state ownership maintenance was progressively deferred. With the private sector assuming control of the refineries, both maintenance and new construction responsibilities devolved to district governments. The privatization agreements that formalized the transfer of control were explicit about the incidence of responsibility both for infrastructure maintenance and new construction. Infrastructure is to be maintained out of a special local account which is funded through a cess on farm and factory production. However, this system is not working very well and rural roads, bridges and drains have deteriorated to the point where significant economic losses are occurring. More specifically29: · Cane is not being harvested because of impassable roads · Fertile and potentially cultivatable land lies idle because it is prone to flood during the rainy season · Harvested cane cannot reach the factory before its quality deteriorates when culverts or bridges fail as they do with increasing frequency · Circuitous routing is needed in order to deliver the harvest with the result that additional costs are absorbed by the farm/factory chain · Delays of more than 24 hours from harvesting to processing occur with increasing frequency with the result that yields and diminished. 3.4.c Internal Supply Chain Incentives As we noted above sugar growing and refining is a joint production function and within this joint function a tension arises necessarily between growers and processors. Still in order for the farm/factory chain to operate optimally, a critical set of production parameters must be balanced, economic tradeoffs calibrated for maximum system efficiency and performance incentives negotiated among chain participants. The fact that economic tradeoffs are inherent in facilities planning, end to end process design and joint harvest and production planning means that returns to one or the other of the participants in the joint production activity may be diminished to the benefit of one party and disadvantage of the other party unless the incentive system is well designed. For this reason, both the process of collective bargaining and the creation of incentives which lead to globally optimal operations are critically important. A closely related is land ownership on the part of small holder farmers. Ownership tends to extend the investment time horizon of small scale farmers and provides then with incentives of investing in productivity enhancing improvements. 3.4.d Plant Stock The genetic plant stock which farmers rely on materially effects both sucrose yield and susceptibility to pest and disease infection. Low-yielding clones still dominate most of the sugarcane estates in Tanzania. Moreover, these stocks are susceptible to pests and diseases, such as the chaffer grub and leaf rusts. Public sector sponsored research initiatives have failed to address the need of sugar farmers in Tanzania for high-yield, pest resistant clones The refiners have begun to invest in new, higher yield stock and to facilitate its adoption by their affiliated out growers. 3.4.e Agronomy and Cropping Methods 29 Tradition agronomy techniques limit production. The privatization of the sugar industry has been instrumental in opening the door widely for technology transfer. The new private companies bring expertise, experience and relevant technologies with them. Spacing, for example, is important in cane cultivation because it increases yields when the growing season is short. Geometric planting in rows assures optimum exposure to solar radiation and hence increases growth and yield . Proper application of pesticides is also instrumental in increasing yield because seeds might otherwise be susceptible to pests and diseases. In particular, the chaffer grub and leaf rusts reduce sugarcane yield and quality of the processed sugar. 3.4.f Community Development The intimate and co-dependent nature of the monopoly/monopsony relationship between growers and refiners and the recent transformation of this relationship from a purely public sector venue to a mixed private/public sector venue, raises significant issues about defining new boundary lines, roles and rules of conduct. One of the needs which has clearly emerged as a result of the recent sugar industry restructuring is the need to build up social capital at the community level and to develop a shared mission which transcends boundary lines between out growers/ factories / trade associations/ village and district government. Change is more likely to be successfully implemented if all stakeholders believe in the good will and good intentions of change agents. Moreover, in the Tanzania context the right to impose taxes and cesses and the collateral obligation to maintain local infrastructure devolves to the district government. Assuring the open governance and democratic accountability of district government is essential to the competitiveness of local refineries and to the high level of comfort that new refinery owners will require to invest further in their factories. All of these considerations come down to community development. 3.5 Roles Played by Mezzo Level Organization in Developing the Sugar Supply Chain The clarification of new roles and the mutual accommodation of stakeholder groups who represent out growers, refiners, consumers and sugar dependent food processors is critically important in effecting the basic restructuring which is underway within the sugar supply chain. This section deals with issues related to institutional development within the sugar industry. 3. 5.a Mezzo Organizations Representing Outgrowers Several mezzo level organizations have successfully emerged within the sugar sector . These institutions serve as evocative agents of consensus views within sub sectors of the industry, as representatives of their sub sectors in policy dialogue with the government and as negotiating agents in collective bargaining with other subsectors. " Mezzo" refers to the location in the policy dialogue of these institutions half way between the public and private sectors. Most essentially their strategic role within the sector is to assure that community goods are adequately invested and that policies with community wide impacts are democratically considered. The most visible of these mezzo level organizations is the Tanzania Sugar Cane Growers Association (TASGA). TASGA is the apex organization which represents all of the sugarcane growers associations in Tanzania. It has the formal legal standing of a non-governmental organization and it was first registered in April 2000 under Societies Ordinance, Cap 337. The founding members of the TASGA were Kilombero Cane Growers Association in Kilombero District, Ruembe Outgrowers Association in Kilosa District and Mtibwa Outgrowers Association in Mvomero District. The table below summarizes the membership of the three district level mezzo organizations that make up TASGA. ASSOCIATION MEMBERS as of Jan, 2004 KCGA 1584 ROA 1840 MOA 4712 TOTAL 8136 TASGA was formed with the vision of creating a vibrant national organization which could effectively support the efforts of individual farmers to improve their incomes by adopting modern methods of cane production and by influencing equities in contracts with refiners and policy consequences in regulations implemented by the Sugar Board . Among the missions which are designated in TASGA's charter legislation are the following: · Foster economic organizations which support farming communities e.g. Community Banks, Transport and Machinery hire centers etc · Ensure that millable cane is harvested · Organise and support the formation of farming groups in villages · Ensure that out growers are paid timely and fair prices · Plan, organize and coordinate agricultural services for members · Provide and approve loans for farmer groups and individual members · Source funds, loans, grants and donations which sugarcane growers can use for cane development etc In Tanzania, sugarcane farmers are much better organized than are maize farmers. Because of their ability to organize they possess much more political strength, even though they are more limited in number. Their efforts have a long history. Small scale cane farmers have been organizing into groups and associations for more than 40 years in Tanzania. A substantial part of the reform agenda being pressed forward by TASGA and its district affiliates is the strengthening of the sugar supply chain. These out growers associations are aware of a number of critical weaknesses within exiting chains and they are expending their political capital on correcting these weaknesses, frequently in partnership with their counterpart producer organization. Some of these weakness include: i) low levels of farm productivity, ii) low cane price, iii) poor economic infrastructure, iv) lack of tangible collateral for cane growers, v) cane fields which because they have not been surveyed, lack clear title deeds and hence cannot be pledged or easily transferred. In addition to these functional weaknesses within the chain there are other critical issues that are also of concern to TASGA: i) the weak enforcement and administration of duties on sugar imports, ii) high bank interest rates and iii) adverse revenue sharing regimes traditionally imposed on out growers. These chain link weaknesses and threats to the sugar supply chain are the agenda which the TASGA representative is pursuing on the Sugar Board. under the umbrella of the associations has created and would continue to create a conducive borrowing environment. This would be the a step towards realizing the supply chain finance concept in Tanzania. In addition to its role is pressing the political and regulatory agenda of the sugar industry, TASGA has also been active in specialized service development for out growers. Thus, for example, it has a successful track record of helping its members comply fully with the loan repayment requirements of lenders and in particular with those of NBC Branches in Kilombero and Mtibwa and with various SACCOS.30 In addition, TASGA through its constituent district organizations has succeeded in forming SACCO' s units and more generally in creating a more conducive lending and borrowing environment. By helping to develop new financial instruments and various forms of debt repayment cross guarantees the district associations have succeeded in developing an array of supply chain financing alternatives. 3. 5.b Mezzo Organizations Representing Refiners The sugar producers have formed a mezzo level organization of their own. A member of this organization represents producers on the Sugar Board. The Sugar Act provides both the opportunity and the requirement that the private sector focus on issues of mutual concern and of mutual benefit. The Act also establishes the basis for formal collective bargaining between associations representing out growers and associations representing refiners. Significantly, the private sugar millers who have come to Tanzania from South Africa and Mauritius bring their own outside experience in developing mutually beneficial out grower schemes and in negotiating the equities that effect these schemes with mezzo level organizations like TASGA. To date however, relatively little collective action has been taken by the producers mezzo organization. The organization is still in its infancy. However, clearly the need will soon arise to form agreed positions among producers on issues involving trade policy and taxation and hence to take collective action on these issues. Clearly within the sugar industry, basic institutional structures are in place to facilitate a dialogue with government. 3. 6 The Industry's Legal and Regulatory Framework 3.6.a The Sugar Industry Act The Sugar Industry Act was passed in 2001. It lays out the regulatory framework within which the industry will operate over the foreseeable future. Significantly the Act vests significant powers in the Sugar Board. In addition, it specifies the general composition of the board which is desined to be represent all of the primary stakeholders within the sugar industry. 30Source Interview with Managing Director National Bank Of Commerce In particular, the Act devolved the following powers and authorities to the board: i) -License all sugar imports and exports; ii)-License sugar cane growers, manufactures and industrial users; iii)-Impose levies within the participants in the sugar chain; iv)-License the importation and use of new genetic material; v)-Appoint inspectors of processing facilities, fields, warehouses, etc. vi)-Monitor methods of pricing, selling and purchasing within the chain. vii)-Impose and collect excise taxes both on importers and manufacturers. Effectively the Act empowers the Board to make recommend decisions on all of the critical issues which are part of the industry's development agenda. Thus, it provides an excellent paradigm for regulatory action in other sectors of Tanzania's economy. If there is a defect in the Sugar Industry Act, it is the fact the law stops short of making the Sugar Board independent. As it is the board serves at the please of the Ministry of Agriculture and effectively reports to him, recommends actions and policies to him, etc. In other words the Board cannot act independently based on its own decisions and its own best judgment. Ultimately, responsibility for policies that effect the sugar industry reside with the government in power and in particular with the Minister of Agriculture. Importantly, the Sugar Industry Act establishes a basis for collective bargaining with the sugar industry between out growers associations and refiners. Increasingly the Board is playing the role of arbitrator and mediator in resolving differences in collective bargaining between these parties. The Sugar Act also designates the powers and authorities of a small secretariate which supports the board's decision making with fact finding, policy oriented analysis, etc. It also set up a National Sugar Institute whose primary responsibilities involve human resource development and training within all links of the sugar supply chain. The National Sugar Institute receives and disburses funds from training levy and payable by employees under the Training Authority Act, 1994 3.6.b Taxation Participants in Tanzania's sugar supply chain are subjected to numerous taxes and levies which are imposed at almost every transaction interface with the value chain. In aggregate these have a significant impact on the cost of production. Taxes are imposed starting at the level of farm preparation and continue through cane production, sugar manufacturing and the purchase of various supportive services associated with the sugar chain. During interviews conducted as part of this study several stakeholders pointed out that these taxes and levies which currently prevail create adverse incentives within the chain whose net effect is to diminish competitiveness. In general two types of levies and taxes are imposed on sugar value creating activities: i) taxes which are imposed on cane sales and on sugar production and ii) taxes and duties imposed on sugar imports. The discussion which follows is organized under these two rubrics. In that discussion we attempt first to identify each of the categories of tax which apply and next to assess the economic consequences of each. The first category of tax includes land rent on sugar producing lands. The land rent is Tsh 200 per acre per annum . It is absorbed by both out growers and estates. Because cane production requires a substantive area of land, it is a significant cost burden. A second kind of land tax is the survey fee. Survey fees of between Tsh 300,000 and Tsh 400,000 per acre are imposed by district governments on buyers of sugar producing land.. Land surveys need to be conduced in order to obtain a Certificate of Title. Survey fees at current levels are exorbitant. They tend to prohibit ownership or formal ownership transfer. Survey fees should be reduced and made affordable, so that farmers can get certificates of title that can be offered as collateral for loans and advance consideration by banks and other lending institutions. . The second set of levies include withholding taxes. These are taxes paid on the sale of sugar cane. Sugar manufacturers deduct withholding taxes from the moneys they pay out to growers for the purchase of their cane. The current withholding tax rate is 2% of the sale gross proceeds. Withholding tax on the income of small scale cane farmers takes no account of the cost of production. Withholding tax is sometimes levied on revenues, which do not even cover even the cost of production. A third tax is the Value Added Tax, which is calculated based on 20% of revenues paid to growers for their cane. The Value Added Tax accrues at a standard rate for those generating annual income of 20,000,000 Tshs or less. It applies both to individuals and organisations. Those who are VAT registered have an opportunity to reconcile their gross income and their expenses. By calculating value addition net of expenses VAT registered tax payers are often able to claim refunds. Individuals, however, who are not registered under VAT cannot claim refunds. They are forced to pay the entire 20% levy on their gross income. Cane out growers fall into this category. VAT is also imposed on services rendered to farmers, for cane cutting, loading and cane transport. Since these services are outsourced only when they are more efficient than insoucing alternatives and since they are an integral part of cane production, farmers are not registered under VAT are doubly disadvantaged when they try to find more efficient ways of producing cane. District authorities charge both farmers and producers a special set of Cane Cess Levies which are assessed both on the volume of cane and on the sugar produced from that cane. District government's have significant discretion over the cesses that they charge. No set uniform national standards apply. District authorities have been empowered under sections 7 and 13 of the Local Government Finances Act No.9 of 1982 as amended by Act No. 17 of 1985 and Act No. 18 of 1991 respectively, to fix though locally enacted by laws the rate of cane cess levies payable by sugar cane out growers and sugar estates. One case in point is the by- law enacted by the Kilombero District Council which has been effective from 1st January 1994. This by-law sets the cane cess levy at Tsh 110 per ton. This circumstance of broad local discretion creates competitive imbalances among competing sugar refineries. Other sugar factories which are located within districts, for example, are subject to different cess levy rates that those absorbed by Kilombero. These rates are set by their respective district authorities. Currently cane cess levies are as follows: · Kilombero Sugar Company is paying Tsh 120 per ton to its local district council. · Mtibwa Sugar Estate is paying 1.6% of its revenue to its district council. · TPC is paying Tsh 100 per ton to its local district council . During the season 2002/2003 the company paid Tsh 55 million on 55,233 tons of sugar produced. (Estimate ex.rate 1 US$ = 1100 Tsh) Participants in Tanzania's sugar supply chain argued convincingly during the fact finding phase of the study that rates charged as cane cess are both high and discriminatory. Moreover, as pointed out above, the value for money is low given the deteriorating state of local infrastructure. An additional set of taxes indirectly effect cane and sugar production costs. These include: i) "Pay as you earn" taxes, ii) a skill development levy, ii) a corporate income tax of 30%, iii) duties on imported services of 20% and iv) a duty on imported equipment to 20%. Import duties on imported farm implements and machinery imported are high and discourage more capital intensive farming. It has been suggested that import duties on farm implements and machinery for sugar production should be reduced. Sugar originating outside Tanzania incurs import duty calculated on the basis of a Minimum Dutiable Value (MDV) of 25%. The current MDV is US$ 390 per ton. In addition a suspended duty may apply which varies between 20% and 10%. Heavy tax burdens associated with sugar imports are the primary factor that provide incentives for smuggling and for black market trading in contraband sugar. 3.6 Findings and Recommendations The regulatory framework which operates within the sugar industry is still evolving. Precedents are still being tested and procedural methods still being refined. However, the overall design of the framework and its initial implementation appear to be exemplary. The regulatory framework outlined under the Sugar Industry Act was first set in place in 2001. It defines a post privatization role for government as rule maker and mediator between the interests of different interests within the private sector with a stake in sugar industry development. Importantly, it also frames at least partially issues concerning the responsibilities which public and private sectors agents need to take up for investment in public goods. Much more work, however, remains to be done in clarifying post privatization roles. Only time will tell with certainty, however, at this point it appears that the regulatory framework is working well, through a first round of collective bargaining agreements. The sugar regulatory framework facilitates the negotiation of collective bargaining agreements between cane out growers and refiners. It establishes the context, defines the parties and lays out a process for dispute resolution which leads to negotiated agreements and enforceable outcomes. As we noted above joint nature of the production function for sugar refining requires the development of flexible, but contractually certain mutual dependencies and the collective bargaining process which is emerging is Tanzania appears to be well suited to achieving equitable and fair outcomes in these negotiations. Importantly, mezzo level institutions are forming within the industry in ways which are mandated in the Sugar Act. These institutions are quickly developing effective capacities with which to represent the interest of both key stakeholders on the Sugar Board which is the pri ciple body within government vested with responsibility for dealing with all of the major issues facing the sugar industry's development. All of these developments appear to be quite positive. The primary public policy issue with which the Sugar Board is likely to become involved in the years immediately ahead involves the country's trade policy effecting sugar. Since privatization, production capacity continues to increase apace within the industry and competitiveness continues to improve. As the industry grows it absorbs more inputs and generates ever increasing benefits for employees, affiliated out growers and third party service providers. However, unless the government takes effective steps to open new market segments for the industry this rapid growth may be prematurely constrained. The challenge facing the government is how best to assist the industry in opening new markets . One set of potential new markets are domestic markets for industrial sugar. Another are regional markets for sugar imports, primarily into Uganda and Kenya. In the Uganda and Tanzanian markets Tanzanian producers enjoy a competitive advantage vis a vis lower production cost competitors to the South by viture of their closer proximity and consequent lower distribution cost. However, liberalizing regional market access is a "double edged" sword since the sugar economies of Malawi, South Africa, Swaziland, etc. are still more competitive than Tanzania's at the moment. The "double edge" granting, as well as seeking more open market access. Interestingly, the two primary market opening options entail quite different trade policies. Opening internal industrial markets entails rolling back the liberal tariff preferences that the food processing industry enjoys. Opening regional markets, on the other hand, entails quite an opposite approach. Moreover, some of the livelihood gains that out growers, employees and ancillary service providers have been able to realize in recent years might be threatened in a head to head confrontation with lower cost producers. The issues here are complex and the ultimate outcomes are path dependent. The need for an in depth analysis of both the welfare and development to guide the political debate is clear. As noted above one area where additional reform is clearly called for involves the development of the local infrastructure which is required to support sugar production. The rural road and drainage systems which are currently in place around the country's major sugar refineries are rapidly deteriorating. In addition, new infrastructure needs to be build into the ancillary cane growing lands which are being developed anew as refinery production increases. The private companies which have emerged from the privatization process are unwilling to assume any additional responsibility for infrastructure development. The terms and conditions of their privatization agreements are clear on the point what responsibilities they have assumed. The failure appears to be with the multi-level, federal system of fiscal control and accountability for public investment. Cesses are imposed to provide for infrastructure renewal. However, value for money has not been forthcoming todate. TASGA has undertaken an economic feasibility study of local infrastructure redevelopment need and has further floated a proposal for financing investment in this arena to donors. The study team concurs that investment is indeed required. It has responded to the TASGA proposal with a counter proposal which would attempt to fix the financial management problem which exists at the district level rather than trying to build a separate financing system around the core problem. Appendix A contains the fiscal reform proposal which the study team would request be considered for potential donor funding. Additional investment in public goods is need in the area of improved sugar cane production technologies, particularly in the management of irrigation, excess water drainage and soils. Soil fertility enhancement measures, such as fallowing and intercropping with leguminous species could improve productivity in the long run. The entire area of agricultural research needs to be strengthened. High yield clones, such as those bred in Hawaii, need to be tested for adaptability to Tanzanian climate and soil as a way of improving sugarcane yield. A need for the sugarcane industry in the country to develop research units in some form of public/private partnerships to refine the repertoire of agronomic, plant science and crop management " best practices" available to contract farmers.. A complementary need exists for a strengthened extension service which can deliver technical support regarding best practice to the most vulnerable farmers. Again the issues this whole arena of technology transfer and its practical application are complex and the required policy reforms nuanced. The study team suggests that an indept study be undertaken which completes a benchmark assessment of prevailing agricultural practices, determines any significant " gap" and recommends ways to close this gap which are cost effective and which entail a form of public/private partnership. The final area in which fundamental reform is called for involves the Balkanized system of taxation which prevails in Tanzania. In a previous section of this chapter we discusses how the existing system of cesses, fies, tariffs, levies and taxes creates incentives among supply chain participants which are pervse. The table below suggests some potential responses and remedies to this problem. Again, this is a complex set of issues which call for additional, in depth study. TAX REFORM PROPOSALS ISSUE PROPOSAL JUSTIFICATION PRODUCE CESS ON Rate be harmonised a) Current cess is a heavy tax burden to farmers CANE from 1.6% charged on especially the out growers gross to 0.33% to bring it at par with b) Produce cess differs between farmers producing industrial cess same crop hence need for harmonisation c) Produce cess is discriminatory leaving farmers more harshly taxed than their counterparts who pay industrial cess d) Cess does not take into account cost of production and is charged on gross revenue and not profit SURVEY FEES Survey fees need to a) Survey costs are exorbitant and prohibitive be reduced and made affordable by b) Survey methods have improved significantly majority farmers. A easing operations and consequently reducing costs fee of between Tsh 300,000 and Tsh c) Farmers have nothing else to offer as collateral for 400,000 per acre is loan or advance consideration by banks too high for a farmer, leaving most farms unsurveyed VAT Remove VAT on the a) The farmer is the primary creator of value, thus services rendered to should not be charged VAT farmers e.g. cane transport, cane b) The services are part and parcel of the cane cutting, cane loading production process and costs c) Farmers are unduly taxed as well as overtaxed d) Because farmers are not registered as VAT collectors they cannot claim for refund of VAT payment. WITHOLDING TAX a) Determination of threshold of income tax for farmer's income is ill based. Tax is collected without taking into account costs of production Remove withholding b) The cost of preparing statements of accounts for tax on cane farmers small holder farmers would exceed revenue c) Farmers of other crops such as maize are not subjected to withholding tax. Thus the practice is discriminatory against cane farmers. IMPORT DUTY Remove import duty a) Farmers are unable to buy new tractors due to their and VAT on tractor high costs. If the price of tractor parts and farm spare parts and farm machinery are reduced it will render them affordable to machinery owners of tractors TAX HOLIDAY Farming should be a) While investors in the sugar industry i.e. sugar tax exempted companies are given tax holiday of five years, small holder farmers investing in agricultural production have no incentive like their counterparts. Chapter 4 Fish Supply Chain Development in Tanzania 4.0 Background: Fish Production/ Distribution in Tanzania This chapter deals with issues that relate to the efficiency with which supply chains for fish production and distribution operate in Tanzania. The competitiveness of the country's fisheries sector is currently being tested by requirements to comply with new international market standards, new regional ecological regulations and, at the same time, the need to contend with intensified competition in global fish markets. However, the industry appears to possess sufficiently resilient sources of competitive advantage to meet these challenges and to continue to grow. It has developed to a point where its competencies now transcend low cost production and include distinct capabilities in complying with exacting international food quality standards and in making "fail safe" deliveries of both fresh and frozen food products around the world. The industry's special logistics competencies together with the supply chain systems that support them have helped to position it for future growth and to guide its development onto the next, higher level of value added production. Moreover, now that concerns over the renewable nature of the resource base on which the industry rests appear to be resolved at least over the short and medium term a confident fish processing industry is considering several strategic investments--in value added manufacturing ( e.g. ready to serve fish products) and in aquaculture. If made these investments would enhance the industry's competitiveness. This chapter focuses of several interrelated issues that affect the efficiency and adaptability of fish supply chains in Tanzania, including: i) the evolving industrial organization of chains within the country along two distinct tracts--one which links to domestic markets and one to international markets; ii) the kinds of interactions which take place between participants in these dual domestic/ international supply chain structures and opportunities to leverage these interactions further for greater mutual benefit; and iii) the effects of public sector policy, regulation and infrastructure provision on the future development of the industry. The subsections which follow this background discussion are organized around the following topics: i) demand and supply factors which effect the industry's development; ii) the operation of the fish supply chain and its organization; iii) the economics of fish production, processing and distributions; iv) the key factors which effect fish supply chain competitiveness; and importantly, v) the supportive ( or not-so-supportive) roles played by new institutional participants in the fish supply chain, including importantly the roles of mezzo level organizations, market institutions and organizations that include fishermen and artisan fish processors; (vi) the legal and regulatory framework within which the industry operates; and finally (vii) findings and recommendations. Recommendations presented in section ( vii) are both for policy reform and investment . They are intended to raise the incomes of fishermen, small scale merchandisers and processors who are involved in the fish supply chain. 4.1 Competitive Context Over the past two decades several trends can be discerned in the development of the global fishing industry.1 These include: i) at the low value end of the market, the development of large scale fish meal manufacturing plants which convert small, surface species of fish into animal feed. Typically, these facilities rely on fish caught in shallow, territorial waters; ii) at the high end of the market, the 1See: " Of Saviors and Punks: The Political Economy of the Nile Perch Marketing Chain in Tanzania," Peter Gibbon, CDR Working Paper 97.3, June 7,1997, Danish Academy of International Studies. development of technologically sophisticated fishing systems which track and catch high value species (e.g. tuna, salmon and shrimp)both territorial and in international waters; iii) the emergence of aquaculture as a basis for providing reliable sources of medium to high value products to fishing processing plants. Fish farming has proved to be sustainable for shrimp, crustaceans, salmon, sea bass, tilapia and a number of other high value species; and iv) the development of new sources of white table fish, including new species and new fishing ground locations, which initially served to replace cod. As cod and a series of substitutes have been fished out in specific habitats the global industry has continued to find new species and supplies to satisfy global market demand . The global market for table fish is quite large--2.4 million tons of exports annually2. The quest for new sources to supply this market has involved the serial development of a number of fishery resources and their eventual depletion. In the late 1990's one of the casualties of over fishing appeared to be the production of Nile Perch particularly from Lake Victoria. However, regulatory measures taken over the past ten years appear to have stemmed the tide of irreversible depletion and to have extended the economic life of the primary resource on which the fishing industry in Tanzania is based.3 As for the other major global trends, these too appear to be having some effect in Tanzania. Thus, several proposals are being floated by private fish processors to develop aquaculture production platforms on the islands which dot the country's Indian Ocean coast and the production of chicken feed meal from small scale species of Lake fish is already a growing industry in Tanzania. 4.1.a Demand for Tanzanian Fish As a nation, Tanzania is uniquely well endowed with fishery resources. It has a coastline of 800km and a narrow continental shelf (mostly 7 to 20km). Its inland fishery resources include three of the African Great Lakes which total approximately 54,000 sq. km of fishing domain. Most (e.g. 51% ) of the world's second largest lake lies within Tanzania's national boundaries. Fish taken from Lake Victoria contribute fully 50% to the country's total fish production and the uniquely valuable species of soft white meat found in the Lake provided the industry with its original source of competitive advantage. Fish caught in Tanzania find their way into several distinct and separate markets. In fact, the same products-- depending on the distribution channels though which they move---are processed and marketed quite differently in the country. Some of the markets into which Tanzanian fish move are local and others are international. A substantial portion of Tanzania's fish move to market in a fresh product form, some move to market in a fast frozen form and still others are processed and distributed in a salted, smoked or cooked form. Distribution channels in Tanzania vary in sophistication from "near-World Class" to rudimentary, and from "best in class" food safety practice to "hit or miss" practices which are neither controlled nor regulated. In Tanzania fish have traditionally been part of the basic diet and are widely consumed. Fresh fish are sold through a number of distinct wholesale/ retail distribution channels which operate in parallel to one another within the country and beyond its borders into other countries in the region. Domestic demand for fish is influenced by several factors, including local supply, preferences for specific species and preparation methods, population growth, and prevailing price levels. 2See: pp-3-4 " International Market for Fishery Products," Fatima Ferdouse, INFOFISH, Malaysia 3See: National Fisheries Sector Policy and Strategy Statement, Ministry of Natural Resources and Tourism, The United Republic of Tanzania, December, 1997. However, an integrated national market for fish has not yet developed. In general consumption of 4 fish products is high near fishery resources and low in parts of the country which are distant from these resources. Statistics show that on average, Tanzanian households consume about 25 kilograms of fish annually.5 Fish protein accounts for fully a third of the total consumed by Tanzanian's A national goal, has been set, however, to reach an even higher household consumption level of 30 kilograms of fish annually. That goal should be attainable given the scale of renewable fishery resources available in the country. Disturbingly, however, evidence exists to the effect that the local availability of fish for consumption is declining and that protein malnutrition is prevalent even among people of the Lake Victoria basin. Clearly, some aspects of the supply chain for domestic fish distribution are not functioning, as well as they could. Still, over 500 species of fish are consumed within Tanzania, with reef fishes being the most popular category. These, include such tasty species as emperors and snappers. Another very popular species which trades extensively throughout the country is Dagaa.6 Dagaa is the collective name for various types of sardine-like fish eaten in a dried form by poor and middle-income groups throughout eastern and southern Africa. Fishing and own consumption of fish, moreover, are an important subsistence activity in Tanzania. Other species of fish, however, are reserved primarily for export markets. Thus, for example, fully 90% of the prawns harvested in Tanzania are exported. The remaining 10% are sold into local open markets, as well as through chain stores. Local demand for prawns, however, is relatively low because prevailing prices for this item are much higher than the prices of alternative sources of protein. High prices discourage domestic consumption. However, with the expansion of the tourist industry this scenario may change, particularly as supplies of salt water fish continue to increase. For other reasons, most of the Nile perch and tilapia is exported. Nile perch and tilapia are caught with gear that local fishermen cannot afford and that the industrial fish processors make available to them in exchange for their catch. Even the local fishermen communities that catch these native fish do not benefit as consumers from their abundance. Of the total factory output of processed fish, fully 50 to 60% is sold chilled and is airlifted to Europe, principally to Holland, Ukraine and Belgium, as well as to Israel and Japan.7 Tanzanian 4 "The Master Plan Study on Fisheries Development of Tanzania," Ministry of Natural Resources and Tourism Fisheries Division ,2002. 5Ibid 6See: " The Poor Relation . A political economy of the marketing chain for dagaa in Tanzania" CDR Working Paper 97.2, June 1997, Danish Academy of International Studies, Peter Gibbon. The study found that: " Over the last decade the trade has increasingly fallen under the control of a group of about one hundred wholesaler-cum-brokers based at Kirumba market in Mwanza town. This group has centralized control over the national and regional trade through establishing a high degree of regulation on Lake Victoria itself, in a context of falling catches and high demand. Unlike other fisheries, where regulation is often organized through sharecropping-type relationships, Kirumba traders' basically regulate the trade through various forms of cartelization. The study examines how they work and what conditions they rest on." 7"The Master Plan Study on Fisheries Development of Tanzania," Ministry of Natural Resources and Tourism Fisheries Division ,2002 fresh fish moves via channels which are quite distinct from frozen fish which makes up the remainder of factory processed (40 to 50%) exports. Frozen fish generally moves by sea-freight and ends up primarily in the USA, Australia, South Africa, Malaysia, Hong Kong. as well, of course, as the EU. The export of Nile perch fillet and related products earns Tanzania about $90 million annually. 8 This corresponds to 15 per cent of the country's total foreign-exchange earnings. Industry insiders believe that if fish production/ distribution chains were better managed and if incremental investments were made in value added processing, export earnings could easily reach $120 million per year. With respect to fish product imports, these amount to only 0.3% of the export amount value. Most of the imported products include tinned foods and their contribution for consumption remains small. 4.1.b Renewable Resources: Tanzania has enormous water resources. The country shares three of the largest inland lakes in Africa. It is served, as well, by diverse river systems and has numerous wetlands. Its coastline stretches over 800 km. The country is rich in both marine and inland fishery resources and around this renewable resource base a significant fishery sector has developed and matured. From the perspective of prudent resource management, additional growth potential appears to exist both in captured and aquaculture methods of fish production.9 The present potential yield is estimated at 730,000 metric tones of fish, which could be harvested at sustainable rates from the country's natural waters as follows10: · Lake Victoria-200, 000 metric tons, · Lake Tanganyika-300, 000 metric tons, · Lake Nyasa-100, 000 metric tons, · Other lakes, dams, and reservoirs 30,000 metric tons, and · Marine waters 100,000 metric tons. 8 Ibid 9The issue of whether the " maximum sustainable yield rate" for Lake Victoria has been exceeded in terms of harvest levels is being studied in depth by another World Bank study team. That research project entitled: " Fish Exports from Lake Victoria--Export Diversification and Common Resource Management" is studying this among other issues related to the export dependency of fishermen in the Lake Victoria basin.. However, based on conversations and correspondence with two of the largest fish processors who operate in Tanzania we found that the conservation compromise which was reached in 2000, e.g. to regulate " fish slot sizes," has been effectively implemented both in Tanzania and in Uganda with positive effects but not in Kenya where allegedly 80 tons of fish per day are still being smuggled through the Mara Region. The slot size issue is complicated by the intermediation of " agents" who mix mature with immature fish and who shop consignments among processors some on the Kenyan and some on the Tanzanian side of the Lake. The ESW work which is underway will deal with the enforcement and regulatory management issues surrounding the " common pool" asset which is the Lake Victorian fishery. 10National Fisheries Sector Policy and Strategy Statement, Ministry of Natural Resources and Tourism, The United Republic of Tanzania, December, 1997 The actual annual catch falls significantly below this potential and had been oscillating between 350,000 and 400,000 MT .11 The major, fresh water commercial fish species are Nile perch, Tilapia and related species. The principle salt water species are prawns, lobster and squid. Hence, opportunities exist for substantially increased production apparently without diminishing the renewable base of the fisheries industry. Moreover, the nation has just begun to exploit its off shore Exclusive Economic Zone. Significant potential exists for increased landings of saltwater fish such as tuna and similar fish species caught on the country's continental shelf and for investing in fish farms in the shallow waters surrounding the country's many off shore islands.. However, by far the largest proportion of fish comes from the fresh waters of the lakes: Victoria, Tanganyika and Nyasa. The waters of the Indian Ocean are the main source of marine fish, lobsters, crabs, and squids. Nile perch is the number one fresh water export. It was introduced into the Great Lakes in the 1950 and has since then become the dominant predator fish in the Lakes. Prawns and lobsters top the salt water export product categories. Small-scale artisan fishers account for the largest share of the nation's total fish catch. More than 70, 000 artisan fishermen make their living in Tanzania. They operate principally in shallow waters within the continental shelf and in the Great Lakes Basin; They use both modern diesel powered trawlers, but more often traditional fishing vessels including dhows, canoes, outrigger canoes and dinghies. They apply various fishing techniques but primarily use uncomplicated passive fishing gear such as basket traps, fence traps and ring nets, as well as various hook and line techniques. Many artisan fishermen operate from island bases in the Indian Ocean which are 10 to 30 km off shore and from islands in Lake Victoria. They deliver their catch to shore landing sites such as the Dar es Salaam landing site and fish market and the Mwanza landing site on Lake Victoria. They have developed complex commercial relationships with other participants in the fish chain who process and market their products and provide them with fuel, boats and gear. They live in communities around which clusters of specialized support services and ancillary value adding industries develop and generally they are able to realize a better income than farmers. However, despite the high earnings levels which hard work can realize for fishermen, they frequently live in poor conditions, conditions which stem, in part, from over spending and the carefree life style for which fishermen are notorious but, in part as well, from the failure of local community organizations to fully support the development needs to the fishing community. Importantly, the level of fish supplied from different water sources varies seasonally and some fishing in Tanzania requires fishermen ( and their families) to follow the catch. Most fish species manifest an annual pattern of abundance alternating with relative scarcity. This seasonality in the catch can be explained based on the annual migratory habits and seasonal feeding patterns of specific species. In order to succeed at their elected profession fishermen require abundant knowledge of fish habits, patterns of migration and propensities to feed in specific local waters and a willingness to adapt as best they can to the uncertainties associated with the profession. . 4.1.c Market Institutions The most critical commercial exchanges in the Tanzanian fish trade takes place at the local fish landings around which local markets frequently cluster. Almost 600 of these exist in Tanzania. 11Ibid Here fish is transferred, stored briefly and redirected by traders into the various specialized retail and local food processing markets which they serve. It is from fish landing sites, as well, that fish is transported to the export processing plants. One case in point is the Dar es Salaam fish market. This market enjoys an ultra modern landing site/market center through which mostly marine fish is landed and traded.12 Most of the other landing sites on the various water bodies which provide fish are in much worse condition than is the Dar market. These facilities, for example, have poor fish handling and no cold storage capacities. With the exception of markets located in the principle urban centers which have been planned, designed and constructed specifically to perform fish trading functions, most fish ladings/markets/ artisan processing centers have developed spontaneously near landing sites and sandy harbors. Most of these markets have no running water, no electricity or no public toilets. In addition, the transportation infrastructure which links these landing site/markets to the rest of the roadway network are frequently non-existent or, if existent, poorly maintained. As these market have expanded and larger motor vehicles have begun to access them, the result has been chronic traffic jams. Further more, significant damage has resulted to surrounding roads from overweight vehicles and frequent off road excursions to avoid traffic. No city or regional planning is evident, for example, in the ancillary support systems, infrastructure integration or traffic flow management in the harbor areas of Mtwara or Tanga. Importantly, these fish landing/ market centers are organized and constituted for the most part agencies of local government. Government operates as a landlord in furnishing space for various kinds of retailing, wholesaling and auction trading activities. As we noted improvements are minimal. Typically, the local government agency responsible for managing the landing/ market also establishes the rules for the operation of the market and the nature of the exchanges which take place within it. These rules are established either by fiat or with the consultation and consent of market tenants. The agency responsible for the fish market, covers its overhead and maintenance budgets from fees with it collects on individual transactions and/or from rents which it collects from tenant/ traders. Significantly, no backward of forward linkages are typically developed between these landing/ market organizations and other activities in the fish chain. 4.1 d Supply of Fish Products As the table below demonstrates export sales of fish and fish products have been increasing at more than 20% over the past four years. The export industry is experiencing a boom. 12The Dar es Salaam market was redesigned and reconstructed with the help of a credit provided by the Government of Japan. Tanzanian Export Sales ($ US thousand ) 90,000 80,000 70,000 60,000 50,000 Fish--Fresh and Frozen 40,000 Fish-Salted, Dried and Smoked Crustaceans 30,000 20,000 10,000 0 1998 1999 2000 2001 The primary fish taken on Lake Victoria is the Nile Perch. Nile Perch is a large carnivorous species, which was introduced into the Lake in the early 1950s. By the early 1980's the Nile Perch had taken control of the Lake and accounted for most of its biomass. After some initial consumer resistance Nile Perch became popular for eating locally in smoked, fried and dried forms. The primary edible species which preceded the Nile Perch on the Lake was the tilapia. It is still produced in the Lake. However, the tilapia is badly over fished and as we noted has been overmatched in its struggle for survival with the Nile Perch which feed on the tilapia. Other commercial species, notably Bagrus and Labeo, rose briefly in prominence on the Lake before declining equally rapidly. Moreover, no real market could be found for the most common Lake species, the small but less tasty Furu. In the second half of the decade a northern hemisphere market developed for Nile Perch as a 'table fish' substitute for cod. Today, the Nile Perch is highly valued as a premium priced table fish in several markets, including Europe, the FSU, North America and Japan. Since its initial market acceptance its price has continued to increase. During the 1980's a number of export oriented industrial plants for filleting and freezing the fish opened in Kenya. In the early 1990s some of these plants moved into Tanzania , as an immediate consequence of the government's prohibition of unprocessed and semi-processed fish for re-export and as a consequence of treaties which clarified national fishing rights and take out rates which are significantly greater in Tanzania than in Kenya. In addition' the government provided several forms of investment incentive to the industry to build plants and facilities within the country-- including tax holidays, import rebates, forgiveness of sales tax on capital investments, facilitation of leaseholds for plant sites, etc. 13 13: " Of Saviors and Punks: The Political Economy of the Nile Perch Marketing Chain in Tanzania," Peter Gibbon, CDR Working Paper 97.3, June 7,1997, Danish Academy of International Studies. Eventually most of the Nile Perch catch and as well as its processing originated from the Tanzania side of Lake Victoria. In 1995, factory operators were prohibited by Tanzanian law from owning and operating their own trawlers and, thus, made to depend on artisan fishermen. This remains the situation today. The same traditions and regulations apply to ocean fishing within the territorial waters of Tanzania as well. Both because of improved prices and factory sponsorship, a huge expansion took place in fishing activities during the 1990's with a consequent upgrading of the 'artisan' fishing fleet. As the table below demonstrates the number of landing sites in Tanzania rose to its current level of 596 and the number of fishermen increased to 56,000.14 Essential support facilities required by the fishing industry increased apace. In the meantime, the trade in 'artisan ally' processed Nile Perch for local market sale also remained a significant one, although it was now confined to whole fish and fish parts which the factories had either rejected or were unable to process further ( e.g. heads, tails, skeletons, etc. . Great Lakes Fishermen and Fishing Facilities In 2003 I T E M C O U N T R Y KENY TANZANI UGAND LANDINGS Number of Landing Sites 29 59 59 FACILITIES Bandas 7 3 5 Cold Rooms 1 2 Pontoon/Jetty 1 3 3 Fish Stores 1 1 7 All Weather Roads 6 13 13 Boat Repair Facilities 5 22 22 Net Repair Facilities 5 24 18 Electricity Supply 2 2 1 FISHERIES STAFF Fisheries Staff (To be verified) 5 6 7 FISHERMEN Number of fishermen 33,03 56,06 34,88 As well as its importance in its own right, the Nile Perch fishery indirectly created the basis for the rise of the dagaa fishery. During the period of its proliferation, Nile Perch seems to have fed mainly on furu. Furu, in turn, fed on plankton. Hence, their reduction in numbers created an opening in the Lake's ecosystem for other species competing for this same food source. The chief beneficiaries were the dagaa, whose population increased dramatically. Unlike furu, dagaa shoal at different water levels than those in which the Nile Perch feed. Hence they avoid the Lake's number one predator.15 P. 16 14This figure does not include salt water fishermen of whom there are estimated to be an additional 20-30,000 15 " Of Saviors and Punks: The Political Economy of the Nile Perch Marketing Chain in Tanzania," Peter Gibbon, CDR Working Paper 97.3, June 7,1997, Danish Academy of International Studies. P 12 The most common fishing technique used on the Great Lakes is gill netting. Using this method, multiple nets are thread together by a cable and thus inter-connected into a chain. The chain is set out early in the evening and retrieved with the catch early in the morning. Gill netting allows significant economies of scale to be realized. By working together multiple and larger boats and multiple crews can raise more fish than can small boards with single crews. 16 Regulations currently in effect specific the minimum size of the netting mesh and increase the likelihood that small size fish will escape the gill nets. The same regulations prohibit the use of seine netting, however, this practice continues in some areas of the Lake. With that said several of the fish processors whom the study team interviewed told us the conservation practices are beginning to pay off in the increased yields of Nile Perch. The fishing method which is increasing most rapidly is the use of long baited lines. These long lines contain hundreds and sometimes thousands of hooks. They can be put out and taken in by fewer boats and are less costly to purchase new and to maintain than gill nets. Also, the size of the hook and the type of the bait used allows the fisherman to target a specific species and size of catch more precisely. Thus, this method offers conservation benefits, as well.. The table below describes the kinds of gear which are currently in use on the Great Lakes. Great Lakes Fishing Boats and Gear In 2003 C O U N T R Y KENY TANZANI UGAND FISHING CRAFTS Number of fishing vessels 10,01 15,48 15,54 PROPULSION Number of outboard engines 49 1,53 2,03 Number of inboard engines 1 1 No. of boats with paddles 6,57 11,62 12,84 No. of boats with sails 2,92 2,32 66 GEARS Gear type & size: Gill nets: < 2.5" 4,38 7,09 67 2.5" 5,17 3,12 32 3" 8,29 2,94 3,01 3.5" 6,71 2,30 9,64 4" 5,71 4,08 20,36 4.5" 2,82 5,67 20,43 5" 8,05 88,37 51,47 5.5" 9,96 27,09 16,29 6" 29,32 59,33 95,30 6.5" 8,85 8,80 8,06 7" 22,28 15,15 54,45 7.5" 1,99 1,39 8" 2,40 1,13 8,10 9" 2,50 19 1,77 16Ibid, p. 26 10" 3,52 48 5,70 > 10" 3,20 62 Total gill nets 125,22 225,80 297,66 Beach seines 5,24 1,01 81 Scoop nets 80 Dagaa seines 2 Cast nets 4,41 4 1,27 Lift nets 31 Number of hand line hooks 27,78 13,23 4,58 Number of long line hooks 972,08 2,212,57 254,45 Number of traps 3,19 2,55 11,34 Mosquito nets 11,26 3,26 2,45 Others (Unspecified) 1,70 1 7 TRANSPORT BOATS Transport boats 40 63 91 DERELICT BOATS Derelict boats 1,88 2,81 2,77 4.2 Overview of the Fish Supply Chain in Tanzania Relative to the maize supply chain, the fish chain is better developed. However, in comparison with other world-class fish production/distribution platforms a significant gap still remains between "best international practices" and the integration of fish catching, processing and shipping in Tanzania. It is also the case that two very different supply chain systems co-exist in Tanzania. One of these is sophisticated. It entails low transaction costs, is well invested and falls just a half step below best international practice. The other is rudimentary in technique and technology and is poorly organized. It entails high transaction costs and incurs substantial risk for fishermen and boat operators. The amount of investment in the domestic supply chain, moreover, is de minimus compared to the level of investment in the parallel export chain. Moreover, the service industries which have developed on the periphery of these two supply chains have also developed in separate dual tracks. Little cross over service provision and less use of common assets ( e.g. cold chain, warehousing, information systems, etc.) takes place beyond the point at which primary fish products are landed, sorted and sold. Thus, for example, in small, rural fishing communities, fish transactions are still executed directly between fishermen and a relatively small number of retailers, wholesalers, buyers agents and, even, consumers who come to buy at the fish landing sites. These market intermediaries are highly specialized. They resell the fish they buy into specific niche markets in which they believe their superior knowledge, commercial networks and access can realize above market returns. In many parts of Tanzania, fish marketing is simply limited to the local village in which fishermen are domiciled or, indeed, to neighboring communities within walking distance. Little trading in support of long distance commerce takes place. Wholesale market channels which entail a higher level of logistics and transportation sophistication and more complex price discovery are situated in Mwanza and Dar es Salaam. Among these wholesale markets, the Banda Beach Market in Dar es Salaam is probably the most sophisticated market center in the country for reef fish. Importantly it is also the primary wholesale center for sales into the local Dar es Salaam which is the largest retail market in the country. The Kirumba Market in Mwanza has assumed a parallel function for Lake fish. It has become the logistics center, the focal point for price discovery and the primary base for wholesalers who are trying to develop a national market for artisan processed fresh water fish. In both of these market centers wholesalers, retailers, storehouse operators, transport operators and retail customers co-mingle and cooperate in forming the best of the ad hoc supply chains that reach out to national markets.. The Banda market handles more than 10, 000 tons annually of a variety of fish products. The Kirumba market handles 15,000 tons annually, including mostly artisan processed dagaa and Nile perch. In both markets larger lots are reserved for sale to wholesalers recognized by the market itself and are auctioned to these wholesalers using a set of rules that conform to Dutch auction principles. Smaller lots of fish are sold through retail market stalls and informally through other wholesalers. For these transactions prices are negotiated on a sale by sale basis. Both markets acts as a focal points for price formation. Competition on both markets is quite intense. However, prices are not effectively transmitted beyond these two market centers and the governance principles which apply to trades within the two centers continue to advantage local traders vis a vis others. With that said, it needs to be pointed out that the wholesale link in the domestic supply chain for fish products is generally under developed. Thus, for example, only basic ice cooled surface transport exists over long distances. Significantly, no inter coastal or cross Lake shipping capacity, interior air or rail services operate to support the national fish trade. No security interest are created in fish inventories, no intermediated transfers of fish products take place. Rather all trades are on a quid pro quo basis and hence make extremely claims on traders time. Moreover, no regulatory framework exists either to protect customers or buyers/ resellers with respect to trading protocols, contract enforcement and/or liabilities related to safe food handling practices. Only limited market information is available and no third parties have entered the market for market information services. In other words, many of the essential components required to develop an integrated national market are missing. Hence, trading in fish remains primarily a inefficient activity with high product losses, high trading risks and high transaction costs. Moreover, the only national market which has emerged to date is one for processes fish, e.g. dried, salted, fried and otherwise preserved fish products. With that said a growing market is developing within Tanzania for fresh fish. Supermarkets and fast food chains are leading this market development. In addition, an array of fish dealers, peddlers, small scale fried fish processor, fish retailers and restaurant owners buy fresh fish every day either directly from fishermen or indirectly through agents or, indeed, buy from each other. The short economic life of fish products requires expeditious trading, internal quality controls and rapid disposal of the product once it is purchased. Whatever other limitations they may have participants in the domestic fish supply chain clearly understand the need for expeditious trading, transporting and sale. Still markets for fresh fish remain under developed and local. Because of a lack of recorded data, the number of small-scale traders involved in the fish business is unknown.17 However, an rough estimate provided by one knowledgeable industry participant has this number exceeding 60,000. Moreover, the majority of traders and processors remain 17One industry official estimated the total as just over 50,000. unlicensed and prefer, generally, to remain in the informal sector. The average daily transaction volume of petty fish peddlers is about 20kg. Their daily sales are estimated at Tsh.2, 000 ­4,000. The trader's margin is about 10% of the fish price.18 An informal survey of customers at the Kirumba market in Mawanza revealed that, with the exception of special species, most of the fish catch was consumed locally. From this survey the study team inferred that the distribution of fish products is on a small ­scale and that 80% is limited to buyers who are based within one trip day of the market. The exception is a few species of fish for which strong demand exists throughout the country. These are fish species that Tanzanian's are prepared to pay a premium to consume or to use as animal feed. Thus, Nile Perch and Dagaa, are marketed over a much more extended geography than most other species from Mawanza. As highlighter in the flow map below both Nile Perch surpluses sold be large processing factories and Dagaa are shipped to Dar es Salaam and other inland cities in quite substantial volume.19 Kirumba based traders control much of this traffic. . Moreover, most of the Nile perch continues to be purchased by the major fish processors as are most of the prawns. Indeed, due to sharp increases in demand for both fishes in overseas export markets, prices have remained firm and demand continues to exceed supply. Export markets have a strong preference for oily white fish like Nile Perch and as its market acceptance continues to develop into new export markets its price continues to climb. The supply chains through which Nile Perch and Shrimp are purchased are more intensely managed than the chains through which fish for domestic consumption move. The export chains are integrated by the large export processing companies, there are transparent to inventory flows, prices are set based primarily on guidelines set collectively by the industrial processors which normally entail a significant premium above local market prices, quality control standards are rigorously enforced and trade credits and preferential leasing terms are extended to quality vendors who have established themselves with the large processors. 4.2.a Artisan Fish Processing: No statistical data have been collected by the Fisheries Division on local fish processing activities or on processed fish product volumes. It is primarily processed fish that are distributed internally within the national market However, it becomes immediately apparent to even the most casual observer that small-scale processing activities are conducted principally in the or near the transient fishing centers or " makambi" where fishermen live, where fish are processed and traded. Every where in Tanzania where fish are caught " makambi" have developed as formally recognized political jurisdictions with their own elected officials. Significantly as well they have also evolved into complex and multi-functional industrial complexes where all of the specialized functions required to carry out fish commerce are co-located. However, obtaining creditable data on the harvest/production volumes and on the level and value added content of processing activities which take place in makambi is almost impossible. 18Data provided by several fish wholesalers who were interviewed during the study. 19Most of the Dagaa purchased by animal feed millers comes from Mawanza. Some of it this is purchased through a specialized trader who deals exclusively with below market fish products, rejected Nile Perch and distressed market dagaa. However, much of this volume is traded through the Kirumba market and delivered by traders based in the market. In areas where production sites are located close to consumption centers, fresh fish is sold in preference to processed fish. Tanzanian consumers, like most others, have a strong preference for fresh fish. In remote production areas/sites where fish consumption is limited to the immediate watershed market, however, surplus catch ( beyond the needs of the local market) is processed and preserved for sale into more distant markets. Because no cold chain operates in most parts of Tanzania, the processing/preparation methods used are intended to extent the economic life of the product. These methods include, among others, drying, salting and smoking. Specific product preparation processes differ from area to area. Along the coast, for example, drying with salt and grilling on the beach are the most common practices. On the other hand, smoking is the processing practice of preference applied to Tilapia, Clarias and Migebuka. Frying, however, is used for short period food preservation, for instance for 2 to 3 days. Frying is widely applied in both coastal and lake areas to a diversity of fish products. Nile perch, which does not have a domestic market for the reasons cited above is processed and exported almost exclusively by the industrial scale fish processing industries. A small volume of fresh refrigerated and fast frozen Nile perch, however, is sold to restraints, specialty food store and hotels by the export processors. Dagaa is one of the rare fish products that are marketed nation-wide. It is sold in small quantities in even the most remote inland parts of Tanzania. Owing to its small size, the freshness levels of Dagaa drops quickly once it is caught. Thus, 100% of the Dagaa which is catch is sun-dried immediately once the fish is landed. Almost 25,000 tons a year of dried dagaa find their way into animal feed, primarily for chickens and primary in the Dar es Salaam area. The informal distribution network that apparently works for Dagaa could be imitated to distribute other relatively high value processed fish products nation-wide, as well. 4.2.b Industrial Fish Processing Currently, fish which are processed in factories include primarily Nile Perch, pawns, octopus and squid. The largest portion of the Nile Perch catch is processed into fillets. Other product categories similarly are processed as commodities without a great deal of value addition ( e.g. no frozen fish dinners, fish patties, table ready preparations, etc. are manufactured in Tanzania) However, the industry is beginning to move in this direction and several of the industrial processors have developed their own brands in the countries to which they export. They have also commissioned market feasibility studies of ways to leverage their brand equity with a larger ensemble of products, including table ready products As the table below demonstrates the volume of frozen fillets processed in Tanzania has increased significantly in volume over the past ten years. The volume of shrimp on the other had declined since peaking in 1999. The volume of milled dried fish used in animal feed has continued to increase marginally year over year. Most of the first two products mentioned are for export. Dried and milled fish is for domestic use. Production of Selected Fish Products 60000 50000 40000 Frozen Fillets MT 30000 Frozen Shrimp Milled Fish for Animal Feed 20000 10000 0 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 The volume of processed fillets exported in 2000 was 38,868 MT. Approximately 8,000 tons of other ( non fillet) fish parts were also processed and exported. Fifteen fish processing factories and 23 trawlers are registered in Tanzania. Most of these are engaged either in Nile Perch or Prawn trawling, fishing, processing and exporting. The Tanzania Fish Processors Association represents the interests of the industrial processing sector. The TFPA lobbies the Government in behalf of its members, provides advice to the government's trade negotiators on issues effecting the fish trade and endeavors to assist the registered industrial companies that make up its membership to conform fully with the international standards which apply to imported fish products. All of the processing companies in Tanzania have individually obtained HACCP qualifications and individually implement sanitary management programs. However, the TFPA continues to be effective in identifying and disseminating best practice and most economic practice with regard to food safety.. Seven factories which process and export Nile Perch are located on Lake Victoria. The remainder are located on Tanzania's Indian Ocean coast. It should be noted that most Tanzania export factories can process either or both marine and fresh water products. It should further be noted that regulation prohibits factories from undertaking fishing activities with their own crews, employees and/or assets. Thus they depend on artisan fishermen to supply 100% of their basic inputs. All of the major processors engage " agents" under contract and these agents serve as intermediaries with groups of fishermen whom they supply with ice, fuel, fishing equipment and sometimes boats in return for their catch. The distribution of profits and the basis for procurement of primary fish inputs varies from agent to agent. These arrangements are more informal than formal and are enforced through the selective extension or removal of supplies, support and preference on the part of individual agents. Thus, a strong symbiotic relationship exists between artisan fishermen and processing factories such as ALPHA LTD. Risk and reward pass through these preferred supplier relationships. Thus, swings in export volume and in price levels for specific categories of product directly effect the livelihoods of artisan fishers who have only limited capacity to diversity out of one fish type into another . 4.2.c Utilization of By-Products The remained of the fish once the fillet has been removed including its skin, head, tail, and skeletal bones comprise approximately 50% of the total weight of the raw fish, About 46,000 tons of these by-products are generated each year by the fish processing factories in Tanzania. Approximately 5,000 tons of this volume is processed into fishmeal and exported. The remaining 40,000 tons are salted/dried and smoked by small-scale processors and exported to the Democratic Republic of Congo or sold into domestic markets. Although the export of salted and dried Nile Perch is officially recorded to be only 56 tons in published export statistics, the actual production volume of processed fish products which are derived from fish scrap generated in the processing plants is estimated to be about 20,000 tons. This tonnage is exported unofficially and shipped primarily from Dodoma and Morogoro. Processing plants for Nile Perch wastes are located in three areas: Kirumba, Kanyama and Mwanza. In those plants, several hundred small-scale processors, mainly women, work with various Nile perch by-products to convert these into processed food. Thus, fins and bones are dried under the sun, heads and chips are smoked, fried or dried with salts. Since there is a lack of proper drainage in these facilities and since the hygienic conditions are generally poor, the processed products are also generally of poor quality and of questionable food safety. According to an FAO Survey completed in 1998, the volume of by-products from prawn trawlers dumped on the sea was estimated to be from 5,000 to 7,500 tons. If these disposed fish by- products were effectively utilized, they could contribute a great deal to improving both the quantity and quality of domestic fish supply. The survey reported that prawn trawlers off load only 2,365 tons of by-products annually, this out of a total of 8,000 to 10,000 tons. The 2,365 tons are sold to small-scale processors. In Tanzania, explicit permission is required to transfer products from Prawn trawlers to other boats on sea. Yet, many fishers do not bother to secure this permissions preferring to dispose of their by product catch at sea. The culture of by product disposal without accountability and of administrative permit avoidance would have to change if an effective collection system for by-products were to be successfully implemented. 4.2.d Post Catch Losses The close proximity of fishing grounds to landing sites helps to minimize post catch losses at least in the first activity step in the chain. Most of the fishermen who are affiliate with industrial processors are furnished with ice and with insulated, "cold boxes" in which to store their catch. However, product quality does diminish rapidly in the hot Tanzanian sun, especially since no effective refrigerated " sub zero" chains exist. Still even when fish has deteriorated to the point where it is no longer fit for human consumption, it can still be used as chicken feed. Hence, post harvest losses take the form primarily of loss in product value. Still, this loss in value or price can be quite substantially. Its effect can be observed in small fishing villages, where when fresh fish supply exceeds local demand the surplus is processed. Lack of refrigerated transport makes the sale of fresh fish into distant markets where fish supplies are scarce and price high simply infeasible. Still, the price of processed fish is fully 40% less than that of fresh fish. The issue of value addition.....and conversely of value subtraction....is thus critical in the fish supply chain. Yet another case study of post harvest loss and value chain adaptation is that observed in Dagaa processing....particularly during the rainy season. The quality of processed Dagaa depends on prevailing weather conditions since most of the surplus Dagaa is dried under the sun. Neither artificial drying equipment or reefer equipment are available to support value chain alternatives. The post harvest loss of Dagaa in Kigoma is a particularly big issue because the volume of the catch during rainy season is larger than during any other time of year. Official data on post harvest losses are not available. However, based on interviews with local fishers and previous field researchers, 50 to 80% of this valuable and tasty fish is sold as chicken feed because of the low quality of sun drying which is possible during the rainy season.. The fall off in price between fresh Dagaa suitable for consumption (Tsh 700 to 1,600). and partially processed Dagaa sold for chicken feed (Tsh 200 toTsh 300/kg) is quite significant. and Dagaa for human consumption. The price of Dagaa is at its peak in July, which is the dry season, and this is almost twice compared to the price in the rainy season. Dagaa Kigoma, one of the important marketed species, make up 15% of total national fisheries production and therefore, its post harvest loss is an issue for securing both incomes for the Dagaa processors and continuity in food supply in the country. 4.2.e Quality Control Quality control measures exist within Tanzania's fish supply chain. However, they are enforced only for two categories of product: Nile Perch and Prawns both for export and not for other products which are consumed domestically. Food inspection was introduced in Tanzania in order to assure sanitary control of fishery products under the Quality Control and Sanitation Standards Department, Fisheries Division in the 1980's and has been progressively tightened. However, enforcement is uneven in domestic channels. Most retail fish, for example, is sold through street stalls and rarely is it handled through these shops with adequate preservation equipment. Since street stalls are mobile, enforcement by way of inspection systems or sanitary licenses or even by way of more general business licenses remains difficult in Tanzania. Most small and informal enterprises " fly below the quality control radar." With that said, as rigorous regime of bacterial testing applies to Nile Perch products for export. The food inspection laboratory of Fisheries Division in Mwanza conduct inspection surveys in the factories, and psytosanitary inspectors complete temperature examinations on products moving through the airports and the Port of Dar es Salaam. Sanitation controls are required on Nile Perch products to meet EU standards and a survey team from the EU regularly reviews and confirms the adequacy of local inspection, audit and monitoring practices. Since the ban of Nile Perch exports caused by cyanide contamination, periodic surveys are conducted of residual heavy metals and pesticides in fish/lake. These tests are the legal obligation of all exporters. Since equipment is still in short supply in Tanzania, samples are sent to a laboratory of South Africa. In the coastal areas where they are captured, Prawn products are inspected by the TSB (Tanzania Bureau of Standards) since the Fisheries Division lacks an inspection laboratory on the coast. All factories in Tanzania apply the HACCP system, part of which requires the implementation of secure sanitation standards. Yet factories in Europe and USA are now requiring ISO9000 certification and soon Tanzania will be required to follow suit. In Tanzania, 20 companies are engaged in the processing and exportation of fish to European Union markets. In addition, approximately 18 vessels are approved in their anti-septic food design for handing export fish for the EU. These are run by the major fish companies: Alpha Ltd, for example, operates with 6 vessels, BAHARI FOODS LTD with 4 and FRUITS DE LAMER with 4 vessels. Other major companies who are engaged in the airlift of fish to the export markets include TAN PERCH LTD, NILE PERCH LTD, and VICTORIA FISHERIES LTD. These companies specialize in Nile Perch and are located in Mwanza . Other exporters are based in Musoma and Mara regions. All of these companies are engaged in processing Nile perch. However, despite the fact most of the exports are fresh and frozen fish, one can not ignore the supply of smoked and dried fish produced in particular in the Western regions that boarder Lake Tanganyika. Local traders supply fish from the Lake to the main local consumer markets where lake sardines are a delicacy and in high demand. This fish product is principally distributed from Kigoma and Sumbawanga. 4.2.e Transactional Interfaces in the Export Supply Chain Settlements between fishermen and boat operators when fish are sold are either on a spot cash basis or, more commonly, in the form of debits provided against credit advances which boat operators had earlier provided fishermen. Thus, boat operator provide ice blocks, freezer cases, fishing nets, the boat itself, as well as fuel, oil, food and maintenance fees. When fishermen sell their fish to industrial plants or onto the auction market---also with the assistance of the boat operator--they receive debits against the credits they had earlier received for these services and supplies. In this way, the chain linked system provides its own internal credit based on the superior credit standing and access of chain participants who own the most substantial fixed assets. Total sale proceeds are shared among the parties after first deducting expenses owed to the boat operator. The remaining net income from total sales is split normally at a rate of 50-50% . The fishing crew will share among themselves the earned 50% and the boat operator will retain the remaining 50% as profit. The terms of sale for fish sold to the factory supply agent may be either a spot cash basis or again on the basis of debits provided against previous advances. Depending on the mutual understandings and level trust between the parties involved some factory supply agents are prepared to advance credit against subsequent delivery of product. . However during our research, it appeared that most agents pay cash once confirmation is made that the quality checks have been passed at the factory and that the consignment has been found to be acceptable for processing. Once a consignment has been accepted at the processing plant, the supply agent will get an commercial invoice, which directs intermediate banks to make payment. The invoice will state the amount and weight of the fish that were delivered to the processor. 4.2.f Key Storage Areas in the Chain. Temperature controlled storage is vital at almost all stage in the fish supply chain, due to the products perish ability and short shelf life. Ice blocks cool insulated storage containers both at the fishing grounds and at the landing site. On fishing vessels freezer cases are used to store product. These can carry more than 1 ton of fish each. They are packed with ice blocks weighing 100 kgs. This helps in preserving the fish from the fishing grounds to the landing site. Transporters typically use insulated vans on their trucks which are also block ice laden to help reduce the rate of spoilage. Most of the trucks used to transport fish can carry up to 7 tons of fish when filled to capacity. They are either owned by individual transporters or by the fish processing companies. At the fish processing plants, much of the processing as well as the cold storage takes place in a temperature controlled environment. Packaging is specifically designed to facilitate cold air flow in the case on fresh product or to prevent freezer burn in the case of fast frozen product. During storage a number of precautions are taken to assure quality control. For example, stocks are rotated, product temperatures are frequently measured, production batches are monitoring and tracked and a number of organoleptic checks are regularly executed. Cold storage during shipment involves specialized 20 ft or 40 ft containers with refrigeration units and self contained generator packs. At the dispatch point checks are conducted and verifications made regarding the quantity, time and date of shipment. 4.2.f Fish Flow Map Key Fishing Grounds Landing sites and processing plants In line with chain studied Major local markets Export by Air transport Export by shipment 4.3 Economics of the Fish Supply Chain The schematic below describes the flow of fish products through the supply chain based in Tanzania. The discussion which follows elaborates on structural and economic features of the chain. Production and supply chain for Fish: Fishermen Transport boat operators Auction market Transporters to processing plants Receiving of Transportation to fish local markets by fish Rejection Receiving at local markets Supplier Pre- processing Local consumption Accepted Processing Storage Local market Dispatch to export markets Overseas retailers Usually fishermen operate in conjunction with a boat owner. Agents facilitate these agreements. Crews of fishermen work on behalf of boat operators based on term agreements which specify the obligations of the parties, their liabilities and duties and their share of revenue. The picture below shows fishermen operating off shore on the Islands of Pemba, Mafia and Unguja. Fishermen heading for the fishing grounds Boat operators also own smaller craft such as fishing canoes which are used in Lake Fishing as well as the larger salt water craft which are used in ocean fishing. They organize systems for transferring product periodically from larger vessels which continuously catch fish to smaller vessels which relay the fresh catch to and from landing sites. In this way fishing, storing and marketing and transport facilities are often integrated. Boat owners or operators usually provide both the fishing equipment and the overhead services which fishermen require who ply their trade at sea, including fuel and oil, fish nets, ice blocks, freezer cases and food. The period of productive fishing will largely depend on the level of inputs which the boat owners makes available to his crews and on the capacity of the boat to catch and load fish. In the case of the island fishermen represented above an average fishing trip, in deep waters beyond the edge of the Continental shelf may take between 4 and 7 days, in relatively deep waters between 2 and 3 days and in shallow waters, 12 to 24 hours. Costs incurred at Fishing grounds Item Av. Costs per kilogram Labor 30/= Ice blocks 150/= Fuel and oil 20/= Repair and Maintenance 2/= Levy 0.5/= Overheads 2/= Total 204.5/= Source: These estimations have been made basing on the data gathered from the fishermen and boat operators Fish at landing site The fishermen hand off their catch to the boat operators at the landing sites and the owner may either sell it directly to the factories or sell it at the auction market. If taken to the auction market from landing site, different lots of fish of a single species are sold though an open outcry, bid and ask mechanism. Specific times are set in advance for the sale of particular lots. The prices are usually determined based on the availability of a given fish species in any given season. Buyers have an opportunity to inspect lots in advance of their sale. Seasonality is determining what price any given species fish will fetch in a given season. Auction market for fish in Dare salaam. Total Fish weights and sales prices at the Dar es Salaam fish market, 2003 Fish type Wt. of fish (kgs) Price/kg Chaa/ 927.0 688.24 Changu 22,414.0 998.10 Chewa/ codfish 1,245.0 1,028.11 Chuchunge/sword fish 368.0 576.09 Dagaa/sardines 220,709.5 194.9 Frozen fin fish 6,654.0 631.20 Fuatundu/ 3,968.0 1,156.75 Gongola/ 327.0 819.57 Jodari/ shark 5,315.0 1,103.29 Kaa/ crab 125.0 800.00 Kalamamba/ 1,214.0 667.22 Kambakochi/ Lobsters 57.0 5,614.04 Kambamti/ prawns 841.0 1,474.44 Koana/oyster 2,320.0 655.17 Kolekole/ 8,166.0 933.63 Mbono/ 13,056.0 843.44 Mishe/ 1,248.0 801.28 Mizia/ 1,351.0 843.82 Msusa/ 2,316.0 682.64 Mwatiko 112.0 1,250.00 Nguru/king fish 1,408.0 1,406.25 Numba/ redfish 764.0 848.17 Others/mixed 9,160.0 476.42 Pandu/ 1,551.0 618.96 Panji/ brown king fish 1,080.0 666.67 Papa/Sharks 714.0 420.17 Pono/blue fish 3,306.0 518.75 Pweza/octopus 2,529.0 719.65 Samsuli/ nduaro 1,120.0 982.14 Sangara/nile perch 12,250.0 343.67 Sehewa/ 816.0 612.75 Songolo/ 282.0 1,418.44 Tasi/ 2,975.0 777.31 Tembo Uzi/ squid 666.0 1,201.20 Vibua/ 16,479.0 664.72 GRANDTOTAL 347,833.5 Source: Statistics department, Dar es Salaam complex fish market. Export quality fish is transported from the landing site direct to the factories. Some of the suppliers may be independent and may furnish their own transport equipment. Others may be under contract with specialized fish transporters. Some of the factories provide their own trucks which collect fish from various fishing towns on daily " pedal runs" and transport it back to their factory in Dar es Salaam or Muwanza. One case in point is Alpha Fishing company, which transports a substantial portion of the fish which is buys from Tanga, Mtwara and Pemba with its own fleet of temperature controlled trucks.. Lobsters on sale at local market though are largely for export Transport operators that aren't tied to factories have to cover their own expenses of fuel, labour (driver), and ice. However, in both cases, transport agents are expected to provide their own working capital. Costs of transportation from Daresalaam landing site to processing plant Item Av. Cost/kg Labor for loading 30/= Ice blocks 40/= Overheads 10/= Transport 20-25/= Total 100-125/= Transportation of fish onto the local markets is equally important. In the case of domestic market distribution the small scale fish mongers, who buy on the market and then resell to retailers, principally utilize local public transport. A substantially volume of fish product moves onto the local market via public transport. Thus, for example, in Dar es Salaam public buses called "daladalas" shuttle fish mongers and their cargoes from Kivukoni where the fish market is located to Mwenge, Kariakoo, Buguruni and other areas where they on sell their products. Cost of transportation from landing site to local markets in Dar es Salaam From To Cost/ kg Dare salaam Mwenge 60/= Fish market ** Kariakoo 30/= ** Buguruni 80/= ** Tandale 80/= The data below is an estimate on the, hiring and average weights of the produce of approximately 50 kilograms for all the produce that may be transported by transporter. Various fish species on sale at a local market. Twenty fish processing factories are licensed to operate in the processing of fish and its related products. All of these are also licensed to export their fish to the EU and to other export market areas. Inn addition, 8 of the companies running these factories have more than two vessels that are authorised to export fish and fishery products to EU markets as of 2002. In total there are 21 licensed vessels in operation. These vessels are managed under similar " putting out" arrangements as those described above for smaller vessels. Upon arrival of fish at the factory, it is received, inspected and other processes carried out, such as: registration of the truck carrying the fish, recording of the temperature of the fish and carrying out organoleptic checks and checks of the product ID and label. Weighing of the product according to grades and product processing input requirements is done, as well as hygienic unloading, typically under shelter and in temperature controlled conditions. Pre-processing involves washing, sorting and application of ingredients or chemicals. As a means to combat bacterial contamination of the flesh, relatively high levels of chlorine are commonly to the water that is used to clean the catch. However the operators are aware of the maximum level of 0.2 ppm. Also, some plants use hydrogen peroxide--purchased under the trade name Glyroxyle--which slows the deterioration of the fish flesh during processing.. The International Organization has approved Glyroxyle under the ISO standards. The primary processing step involves filleting and dressing the fish. Because of variations in size, species, etc. all of the cutting is done by hand. Importantly, however, processing also entails coding trays for individual batches, organoleptic checks, precise control of temperatures (<40C), visual inspection of grades and product ID, packaging according to grades, as well as weighing and labeling of ID, batch, number, time, and date. The processors claim to implement the Hazard Analysis Critical Control Point (HACCP). They claimed to have EU number since they export to European market. In efforts aimed to ensure that the finished products meet the quality standards required in the export markets, fish is checked and may be rejected by plants both before and during processing. Individual plants also their own laboratories which are fully equipped and operating with trained in-house. They also depend on external facilities, however, to assure objective confirmation of test procedures and conclusions (e.g. independent government laboratories, and the Tanzania Bureau of Standards). Average costs incurred during pre-processing and processing Item Cost/kg Labor 100/= Power 80/= Packing material 175/= Ingredients 75/= Storage 40/= Total 470/= Source: Alpha fisheries limited basing on prawns and octopus. It should be noted that the costs of pre-processing tend to have a slight variance from one product to another. With processing and packaging complete, the plants will sell a fraction of their produce (approx. 5 to 10%) on the local market. The rest is for export. Exporters must obtain an export license from Fisheries Department. The license is valid for one year. These are issued to foreign exporters only if they can show that they have an investment in fishing vessels or shore facilities approved by the Tanzania Investment Center (TIC). Tanzania exporters are not required to make such investments in order to be issued with an export license . However, they may also process their products in the third party plants. Requirements for the export of fisheries products from Tanzania are the presentation of a Customs declaration, together with a commercial invoice, a certificate of health issued by Fisheries Department and payment of export royalty of 6% of the FOB value of the consignment. Item Av. Cost/kg of Av. Cost per kg frozen fish of fresh fish Dispatch/Export 300/= 1000/= (Holland) Royalties I60/= 160/= Total 460/= 1160/= Source: Alpha Ltd (TANPESCA) Several shipping lines serve Tanzanian ports. Some of these are reputable multinational companies with large reefer container fleets and with enough knowledge and facilities to handle perishable products well. Information gathered from the Fish Quality Control Department reveals that there are 17 vessels approved to export fish and fishery products to the EU markets which regularly call on the Port of Dar es Salaam. An average of five heavy lift cargo planes per week from Ukraine, Holland, and Belgium land at the airport in Mwanza, Tanzania's second biggest city, to ship approximately 400 tons of fresh Nile perch fillet to foreign markets. The increased number of heavy lift cargo aircraft underscores the rapid growth of the industry around Tanzania's Lake Victoria where 12 plants process and export Nile perch fillet. 4.3.a Profit and Loss Assessment This section assesses the profit and loss risks and opportunities included in the end to end supply chain for fish beginning with the activities of fishermen. These are recapitulated in the table below for each of the key participants in the end to end fish supply chain. The table below contains an estimate of the costs and margins incurred by a typical fishing boat operator, with a 6-ton capacity boat, which will gather approximately 4 tons of fish per trip, which takes 3 to 4 days to fill. Fishing boat operator costs and margins Tsh per year Tsh per kg of fish US $ per kg of fish % of total costs Operating costs: variable Raw material 145,500,00/= 840 0.84- Sub-total variable costs 145,500,000/= 840- 0.84- 60% Operating costs: fixed Labor 39,600,000/= 30/= 0.02 Fuel and oil 29,040,000/= 20/= 0.03 Repair and maintenance 250,000/= 2/= 0.002 Ice blocks 198,000,000/= 150/= 0.15 License 30,000/= 3/= 0.003 Subtotal fixed costs 88,670,000/= 205/= 0.205 37% Capital costs 6,600,000/= 5/= 0.005 3% Total costs 240,770,000/= 1050/= 0.10 100% Revenue 288,000,000/= 1200/= 0.12 Profit (loss) 50,770,000/= 150/= 0.15 The boat operator makes an annual profit of about Tsh 50 million (US $ 50,000) on a capital base of only Tsh 6.6 million. The rate of return on investments in fishing boats is extremely high and the trade in third party boat support services is highly profitable. Still the possibility of incurring both commercial and environmental risks should not be ignored. These include poor fishing conditions, inexperience crews, bad weather conditions, and the possibilities of a boat capsizing or running aground. The boat operators with capital access are in a very strong bargaining power and the fact that the operators provide a credit input to the fishermen, which enables them to perform their trade improves their bargaining position even more. It is the ability of boat owners to provide capital goods in a capital contained environment which explains their ability to realize extremely high profits. These excess profits, of course, are realized at the immediate expense of fishermen. Ultimately, however, their continuance will diminish the competitiveness of the entire sector. . Processors costs and margins Tsh per year Tsh per kg of fish US $ per kg of % Of costs (Av.) fish Operating costs- variable Raw material 1,108,800,000/= 840/= 0.84 Subtotal variable 1,108,800,000/= 840/= 0.84 64% costs Operating costs-fixed Labor 132,000,000/= 100/= 0.10 Power 105,600,000/= 80/= 0.08 Packaging 231,000,000/= 175/= 0.20 Ingredients 99,000,000/= 75/= 0.075 Storage 52,800,000/= 40/= 0.04 Subtotal fixed costs 620,400,000/= 470/= 0.45 36% Total costs 1,729,200,000/= 1310/= 1.31 100% Revenue 2,400,000,000/= 2000/= 2.00 Profit (loss) 670,800,000/= 90/= The profits of fish processors are sensitive to the cost of raw materials and the export prices of the fish products. The prices of fish at the landing site fluctuate wildly while at the other end of the chain export prices remain relatively static. For example fish produces could drop to as low as Tsh 300/= or rise to as high as Tsh 1300/= per kilogram for prawns. On average a processor who generates export sale equivalent to 1000 tons annually will earn an annual profit of about US $ 670,000. Thus the fish trade in Tanzania appears to be profitable at almost every stage in the supply chain. Moreover, profitability would increase even further with a reduction in the costs of power and packaging. 4.3.b Cost of Transport vs. Cost of Production Transport costs are the least controllable elements in the fish supply chain. Moreover, at least the international component of these costs are denominated in US dollars. As the table below suggests, transport services represent a substantial portion of total delivered cost for fish moving into export markets. The figures below represent shipments of frozen fish moving in containerized lots to the Netherlands. This is a representative movement. In an effort to control ocean shipping costs, several of the major producers have acquired and operate their own dedicated vessels. This effort to internalize risk factors which effect chain competitiveness inside the chain can be only partially effective since none of the processors can serve all of the destination markets into which they are selling with their own operated transport equipment and, of course, none of them can hedge the cost of fuel which represents a large portion of variable transport costs. Transport in Tsh/ kg Cost of Production in Tsh/ kg 1: Fishing ground: Labor: 30/= Transport from Pemba to Dar es Salaam: Ice blocks: 150/= 300/= Fuel and oil: 20/= Overheads: 2/= Levies: 0.5/= 300/= 202.5/= 2: Landing site to processing plant Labor: 20/= Dar es Salaam landing site to Alpha Fish Ice blocks: 50-150/= company at Vingunguti: 25/= Over heads: 1/= 25/= 221/= 3: Processing plant to Harbour: Processing: 470/= From Vingunguti to Dar es Salaam harbo 25/= 25/= 470/= 4: From Harbour to Export market: Royalties: 160/= From Dar es Salaam to Holland: 300/= 300/= 160/= Total 650/= Total 1053.5/= 4.4 Factors Effecting Supply Chain Competitiveness This section deals with the key factors which effect supply chain competitiveness in Tanzania. More specifically the section deals with the logistics characteristics of fish themselves, cycle time from catch to market delivery and modes of supply chain integration and institutions based in Tanzania who are able to integrate chains effectively. 4.4.a Logistic Characteristics of Fish Fish has a number of logistic characteristics which effect operations in each stage of the supply chain. They include the following: Seasonality is one of most important logistic characteristics of fish. Primary fish production is linked to the timing of tides and the migratory movement of fish. Because different fish species have movement seasons, as well as different spawning patters, the available catch tends to vary significantly from month to month.. Knowledge to all these factors is critical for successful fishing. However, since demand and supply for fish are fundamentally mismatched buffer inventories need to be build up within the chain to disconnect the two and to assure a smooth and steady supply of product to meet orders. The ability to manage demand variability is one of the key success factors in the development of Tanzania's fish sector. Perishability is another key logistic characteristic of fish. Fish have a very short economic life. Hence all of the processes within the supply chain need to be carefully paced and scheduled to work with the economic life parameters of the product. Efforts also need to be taken to extend that economic life to the extent that available technology makes possible. Thus, the sufficient availability of ice blocks and of freezer cases at the fishing grounds is vitally important and as is temperature control at all other stages of the production supply chain. Because fish is highly susceptibility to spoilage most supply chain operations from the factory forward must be conducted at very low and constant temperatures. Fish also requires a high degree of hygienic care at all stages of the production chain. This is because with poor hygiene fish likely to be rejected at the vital stages of the chain. Thus the chain requires organoleptic checks and visual inspections of the fish at all the stages of the chain. Susceptibility to damage is another important logistic characteristic for fish. Damage and loss of marketable appearance takes place primarily when fish is handled roughly, during loading, transportation, unloading, packing and storage. Careful handling determines determine the amount of marketable fish that moves from one stage in the production chain to the next. As such proper handling reduces losses caused through both careless and rough handling. Time to market is the most critical characteristic of fish. Competitive fish supply chains are high velocity chains. Dispatch to markets, especially export markets is vital. Freshness is determined by the factory to shelf velocity of the production/ distribution chain as is the residual shelf life of the product in the overseas market. Delays in delivery will most certainly lead to the loss of fish products or change in its most economic use. Its this condition of time competitiveness, that makes efficient and rapid supply chain operations the sine quo non of successful fish marketing. 4.4.b Order to Delivery and Production to Delivery Cycle Times. With that point clearly stated, it is useful to examine the various stages of the production-supply chain from the perspective of its time competitiveness. As noted above fishermen in conjunction with boat operators deliver fish at the landing sites and thus begin the production/distribution cycle. Catch delivery capacity largely depends on the inputs and time invested at the fishing grounds considering that all other factors are also favourable. In addition the location of the waters in which the fish is caught helps determine production time.. All of these factors contribute to elapsed time of catch to delivery at the landing site. Table showing average time spent at Fishing grounds Fishing grounds Distance to landing site Time for fishing km hours/days Shallow waters 15- 20 km 12 ­24 hours Average waters 20-50 km 1-3 days Deep waters Over 50 km 1-2 weeks Source: These estimations have been made basing on data gathered from the fishermen The fishermen and their affiliated boat operators may sell their fish either directly to fish processors' agents or onto the auction market. Usually in situations in which the boat operator has an understanding with factory agents the sale of large fish lots ( e.g. truck loads of 6-7 tons) might not take long once quality checks and weighing have been completed. During periods of low supply and high demand landing site transfer might take about 1 hour to sell off a crew's entire catch. However when fish is in abundant supply on the market, it might take 4 hours or so to sell off a catch in smaller individual lots. Processing plant agents hire or own their own insulated trucks and they normally provide transportation to the processing plants. It should be noted that the fish trucks of 7-ton capacity are filled with block ice so as to avoid spoilage before delivering to the processing plant. The cost of transportation of fish depends on the area of origin and the distance to the final destination. Typically transport requires half a day or less in travel time, depending on the location of the factory. Travel times much longer than this adversely effect product freshness and quality. A truck from Tanga, off loading at the Dare salaam landing station. Transportation costs from different sites to Dare salaam processing plants Area Transport/kg Ice/kg Total cost Tanga 225/= 60/= 285/= Mtwara 300/= 90/= 390/= Kilwa 180/= 60/= 240/= Dare salaam 25/= 40/= 64/= Source: Alpha Ltd (TANPESCA) Loading Fish for transportation to processing plant Fish which are transported to the local markets from landing sites also require hired transport services, but not necessarily ice-laden trucks. Usually public transport is used and small lot sizes are conveyed. The table below contains some representative figures. Public Transport of Domestic Market Fish From To Cost/ kg Distance in kms Dar es Salaam Mwenge 60/= 5 Fish market ** Kariakoo 30/= 3 ** Buguruni 80/= 13 ** Tandale 80/= 10 At the fish processing plant, the fish that are rejected are transported back to the local market, at the expense of the supply agent. However once accepted at the plant, pre-processing and processing stages of production commenced immediately. The entire production stream is paced by the arrival of fresh fish inputs. No buffer inventories exist in these chains because of the quality issues discussed above. Processing including packaging and storage requires approximately 12 to 24 hours to complete. Processing is closely monitored and product periodically tested as it moved though the production chain in order to meet the exacting requirements of export markets. Packing of fish at a processing plant Logistics coordination and facilitation of fish exports requires the presentation of a customs declaration from a commercial invoice, a certificate of health issued by Fisheries Department and payment of export royalty of 6% of the FOB value of the consignment. Customs processes have been streamlined. They operate in a stand by mode and no delays are incurred. Item Av. Cost/kg of Av. Cost per kg of frozen fish fresh fish Dispatch/Export (Holland) 300/= 1000-1200/= Royalties I60/= 160/= Total 460/= 1160-1360/= Exporting fresh fish is more costly than exporting frozen fish, because of the transport modes required. Fresh fish such as Nile perch is usually airlifted out of the country to export markets. Frozen fish that is shipped via ocean container. The costs of airlifting are considerably higher than for ocean shipping . Next and second morning air lift service is available into most markets. Container services, however, require 20 days or more for delivery to most European destinations. However the relative competition in the airlifting of fish in Tanzania, especially in Mwanza, makes the costs of airlifting US 1.2 $ per kg lower than in neighboring Uganda where airlifting costs about US 1.4 $ per kg. 4.4.c Transport Supply vs. Demand The level of transport supply which is available reliably for fish differs for various modes. Demand for air transport is highest in the Lake regions of Mwanza and Musoma from where most of the fresh fish product is shipped. The capacity of air transport is generally good from these origins--more than 20 flights are available per week. From Mwanza, for example, 5 flights per week are available to the European Union. However the processors would like to develop daily flights into a number of other key markets as well. It is highly desirable to have every day air lift service into key markets in order to assure reliable every day fresh product. All of the processors operate their own trucks or arrange for dedicated trucking services. So that availability of surface transport is not a major issues with them. No industrial processes fish moves by rail in Tanzania. This mode is transport is simply not sufficiently reliable or sufficiently flexible to match the needs of processors. The supply of ocean shipping services is also generally good as is the supply of well maintained refrigerated ocean containers. During most weeks 10 or more sailing windows are available with major liners. Most recently a for hire ro-ro direct service to Europe has started up and this service substantially reduces port to port transit time. The larger processors, however, provide their own temperature controlled transport with their own ocean vessels. Thus, Alpha Ltd operates 6 vessels, BAHARI FOODS LTD operates 4 and FRUITS DE LAMER operates 4 vessels. 4.4.d Modes of Supply Chain Integration. Supply chains for export fish and for domestic fish are integrated by the processor who schedule, program and coordinate each of the key activities in the chain including, most importantly, procurement of raw fish, production scheduling and outbound logistics. Most fresh fish production demand driven. Production of fresh fish is scheduled to satisfy purchase orders when product is shipped directly to customers or to maintain minimum inventory levels in foreign countries when local orders are satisfied through company affiliated agents. Fast frozen fish, on the other hand, are is supply driven. Most frozen fish is purchased when raw fish stocks are abundant; fillets and parts are produced for inventory, they are stored in large cold storage warehouses located in Mwanza and Dar es Salaam and they shipped on a minimum outbound logistics cost basis in most economic lot sizes to maintain minimum inventory levels countries to which the company is exporting. Supply chains which support processed fish distribution within Tanzania are not effectively integrated. Each process step in the chain is completed before the next process is begun. No active integration takes place within the chain primarily because no single agent or set of agents has end to end control. Buffer inventories build up between individual process steps so that they can be effectively decoupled. However, these inventories and relatively high level of post catch loss within these channels causes the working capital requirements associated to pushing a unit of product through the entire chain to be very high. The working capital is provided to the chain by the series of buyer/ resellers who participate in any end to end movement. However, the resulting multiple transaction chains also drive up transaction costs, increase market margins and make the chains inherently inefficient. The result is a loss of value both for artisan fishermen and for consumers of fish products in Tanzania. The most fundamental constraints inhibiting the development of more efficient distribution channels for the domestic sale of fish products are two fold: i) missing institutions which can do the work of channel integration and ii) missing skills and competencies. 4.4. e Taxation The Tanzanian Processors Association has taken a strong position that its competitive advantage vis a vis processors in Kenya who compete for the same raw fish resources as Tanzanian based processors is being eroded by progressively escalating taxes. Tanzanian processors pay the following taxes and liens which add an additional 120/- to 200/- to the cost of raw fish inputs: · Royalties--USD.15 per kg or the equivalent of 150 Tshs per kg of finished product · Levies vary in different districts from 7/- to 10/-per kg. In addition the Mwanza City Council imposes an additional, fish levy of 7/-. The total fish levy exceeds 14/- for most processors. · Mwanza Service Levy--Based on .3% of the value of the finished product or equivalent to Tshs 7/- per kg of finished product. This "service levy"is in addition to the "fish levy" noted above · Withholding Tax--Based on 2% of purchased price from the agent or fisherman · Stamp Tax--Based on 1.2% of the purchased price from the agent or fisherman · Multiple Licenses and Registration Fees--These include annual boat license fees of Tsh 40,000 ; A annual " boat fitness " certificate of Tshs. 109,000 per boat; Water rights for boats of Tshs 150,000 per filing plus Tshs 10,000 per boat; Boat parking fees of Tshs 150,000 per boat per month is some districts; and fish container placement fees of Tshs 50,000 per month in various districts. · Multiple Processing Fees and Establishment Licenses--These include a fish processing fee of Tshs 750,000; Import License Fee of Tshs 500,000; Export License Fee of Tshs 125,000 all paid to the Ministry of Industry and Trade. In addition, a Food License Fee of Tshs 50,000 paid to the Ministry of Health and an export license of Tshs. 200,000 paid to the Ministry of Natural Resources. Additional levies and fees include a waste disposal license, a dumping levy, a TBS annual subscription, a radio call license, and a water usage license from the Ministry of Water and Livestock. · Export related Fees and Charges--This include documentation charges; Certificate of Origin Charges; Movement Certificate Charges; Bank Charges for the Payment of Royalties ( the Government does not accept company cheques) · Business Taxes--These include corporate tax, payroll levy, NSSF contribution, land rent. The net effect the Tanzanian Fish Processors Association argues in the near term is to advantage Kenyan processors over themselves. Since the level of taxes, levies and fees in lower in Kenya the cost of raw fish is correspondingly lower. Over the long term, however, the effect may be more significant. Two of the largest processors with whom the study team met made it quite clear that unless a clear agreement could be reached with government at several levels about the basis and level of future taxes, fees and levies, that it was unlikely that they were prepared to invest in additional value adding processes in Tanzania. The cutting edge from the perspective of the potential investors is not only current levels of taxation but even more significantly: i) the basis of taxation which penalizes labor value adding inputs ( e.g. when the basis for royalties and fees is export value a tax is effectively imposed on value added inputs); ii) the progressively ratcheting up of multiple taxes and fees has the effect of increasing these over time at rates which exceed any offsetting increases in productivity or net profit increases. iii) taxes and fees can be independently imposed by multiple agencies, levels and jurisdictions of government without effective recourse or appeal. The fishing industry points to developments in the sugar industry as the way forward. Apparently, agreements were reached between the government and individual sugar processors before financial closure on their privation transactions which explicitly circumscribed the levels of future tax liability. The Tanzania Fish Processors Association would like to see a similar treaty negotiated with government before its members begin to invest in the next level of value added capacity--in fish farming and/or in value added manufacturing of fish products. 4.5 Institutions and Institutional Roles Several institutions have emerged in the fisheries sector to have a significant impact on the strategic development of the sector. These include most notably the Tanzania Fish Processors Association, the major fish markets which operate within the country and the local political jurisdictions which serve as the home bases for fishermen and the platform for a diversity of value adding and fish chain supporting commercial activities. These entities include but are not limited to the " makambi" which were described above. These institutions operate not simply as political entities whose role is to maintain civil order and to invest in public goods but, importantly as commercial entities as well who coordinate, govern and allocate credit to specialized commercial undertakings which support the artisan chain for processed fish at its base. 4.5.a Tanzania Fish Processors Association The Tanzania Fish Processors Association is a private sector institution which brings together all of the fish processing companies with an export orientation in the country to advance their common political agenda and to collaborate commercially in developing the industry. . The Tanzanian Fish Processors Association is closely linked to two other associations with similar strategic agendas which are base in Uganda and Kenya. Together these three represent the interests of private sector association members on issues involving the development of Great Lakes fisheries resources, regulation of fishing on the lakes and other issues effecting equities and competitive balances among the three Great Lakes national fishing platforms. Significantly, some of the same companies belong to two or three of the national associations by virtue of their operating processing plants in more than one national venue. The Association has been the primary focus and coordination mechanism for several of the distinctive features of the fish industrial cluster which has emerged in Tanzania. In particular the Association has been instrumental in developing and implementing the several quality control regimes which have taken effect within the national industry. Through the Association mutual support and enterprise learning has been extended and received by individual enterprises. The Association has likewise been instrumental in developing a generalized approach to the development of the agency/ fishermen relationship and in particular to sorting through disputes which arise among members who advance prepayments or inputs to agents and do not subsequently receive what they consider to be fair compensation in fish products delivered. Also, the Association has been instrumental in harmonizing basis of compensation terms and conditions for agents and fishermen among processors. More recently the association has begun to focus its attention on increasing equity participation in the sector and on further developing processing enterprises already owned by black Tanzanian business men and women. 4.5.b Major Fish Market Centers Most of Tanzania's major cities have spaces designated for the buying and selling of fish products within market centers. This institutions are managed by Municipal governments who act as landlords and lease spaces within the markets to individual retailers, wholesalers, ice makers, warehousemen, etc. Two of these markets --Banda Beach Market in Dar es Salaam and Kirumba Market in Mwanza--are distinct not only in their scale of operation and in the fact that they are 100% dedicated to trading fish and fish products, but also in the critical price discovery functions which they perform for the national market. With that said not efforts have been made to integrate these two market centers either operationally or in terms of exchanging market information between them. Both market centers moreover have outgrown their initial bases and require significant investment in urban infrastructure, specialized fish storage and receipt equipment and facilities. Moreover, both institutions have developed governance mechanisms for setting market rules and for establishing priorities for internal investment and growth. However, their ability to develop further as commercial entities is significantly constrained by their current status as a loose federation of tenants vis a vis a municipal government landlord. 4.5.c Fishermen Communities Fishermen lead a semi-migratory existence in Tanzania. Some of their community economic- political structures are designed for transfer from one seasonal fishing venue to another. Other community structures are more permanent. One version of these structures is the " makambi" which has a formal administrative status in the Tanzanian system of federal, provincial and village hierarchy which is near the bottom. Fishing communities--including the " makambi"--- are as much industrial/ commercial platforms as they are political jurisdictions. However, their capacity to enter into contracts, to borrow money and to manage liabilities in behalf of their constituents is significantly less than that of the district level out growers associations that we discussed in the previous chapter. Opportunities appear to exist to recharge and to reinvent these organizations so that they can enter into collective bargaining agreements directly with processors, commit their members to contractual terms and conditions involving supply of quality product and invest in technology upgrading and recapitalization of both fish catching and fish processing activities. 4.5.d NGOs A number of private sector development oriented institutions participate in the fisheries sector. They help fund programs related to marine resource management as well as to ensuring quality fish production (to date exclusively for exports markets). They fund research missions on the possible enhancement of the sector and on enhancing the welfare of participants in the sector. Most notable among these organizations are: i) United Nations International Development Organization (UNIDO) whose main program trust is to fund emerging small-scale enterprises. The fisheries industry is one of UNIDO's priority development sectors. ii)European Union whose main objective is to ensure that its policies and programs support socially and environmentally responsible development. 4.6 Legal Framework The Fisheries Act of 2003 is the basic legal framework that applies to the management of fisheries in Tanzania. The Act empowers government officials to determine "who can fish," "how they can fish" and" how they can trade in fishery products.". The 2003 legislation replaces the Fisheries Act of 1970 which was repealed although all licenses and permits granted under the provisions of the Fisheries Act of 1970 are deemed to have been made under the new Fisheries Act. In addition, all regulations made under the Fisheries Act of 2003 are deemed to have been made under the Act and shall remain in force and have effect until repealed, replaced or amended under and in accordance with the provisions of the Act. 4.6.a Regulation The Fisheries Act authorizes the Division of Natural Resources and Fisheries to regulate all commercial activities involving Tanzania's fisheries industry. The Division was established within the Ministry of Agriculture, Food and Cooperatives (MAFC) in 1964. At the present time, the Fisheries Division is divided into four sections each with distinct functional responsibilities and each with regional offices in each of the three Lake Victoria districts. In addition, the Division operates three training centers and a quality control laboratory. It employs approximately 200 full time staff members. The mandate of the Fisheries Division is broad and technical. However, most of the day to day activities of personnel within the division involve enforcement of quality controls and standards; surveying fish resources, fishermen, processors and traders; training; compiling statistics;, planning and public communications. The Division has only 59 staff members in its Headquarters offices to carry out this mandate. Of this total only 19 are officers with bachelor degrees or higher. The roles and responsibilities of the division are quite broad and include, the following · Provide for and regulate the conditions under which industrial fishing shall be undertaken; Regulate the structure, functions and powers of authorized associations; · Issue, suspend and cancel licenses or other authorities granted to private companies to catch fish, to process fish and to export fish and fish products from Tanzania. · Require all fishing vessels to be registered; Regulate and control the description, specifications and form of nets to be used in fishing and the size of their meshes; Limit or control the number and size of fishing vessels; ) Prohibit, regulate or control the activities of foreign fishing vessels within territorial waters; · Prohibit or regulate the use of specific types of fishing gear; Control and regulate importation, manufacturing and construction of fishing gears; · Prohibit, restrict or regulate the importation into main land Tanzania any live fish, other than fish indigenous to Mainland Tanzania · Prohibit or restrict the capturing, collection, removal or destruction of any variety of fish, aquatic flora, product or product of aquatic flora; Provide for the protection of critical habitats; Prevent the obstruction and pollution of territorial waters; · Regulate the marketing of fish, aquatic flora, fishery products or products of aquatic flora; Regulate the processing of fish, fish products or aquatic flora or products of aquatic flora; Prescribe conditions under which every processor of fish, fish products or products of aquatic flora shall comply; Prescribe specifications to which any factory building or other premises used for the purpose of fish processing, storage or sale of any fish, fish product, aquatic flora or product of aquatic flora shall conform. : The division appears to have a barely sufficient number of technically competent officers to discharge all of its responsibilities. As a result of recent, district level administrative streamlining, the number of fisheries providing new stock and the quality of services offered to fishermen have decreased. District fisheries officers are under the management of DED. Instead of patrolling fisheries, collecting data, and offering extension activities, which were the original responsibilities of the fisheries field staff, they are increasingly involved in collecting fees to continue to support district activities. Their activities have shrunk to ensuring that basic statistics are collected on fisheries production and that large processors are complying with regulations and restrictions in their licenses. Active management of fisheries and leadership with respect to improving the incomes of poor fishermen are not among the priorities which the Fisheries Division is pursuing. 1983 became 4.6.b Fisheries Research and Skills Development Another important institution to the development of supply chains is the Tanzania Fisheries Research Institution. (TAFIRI). The Research Institution was founded in 1980 to implement a fisheries research program. Initially it was a part of the Fisheries Division but soon became a separate entity with its own budget. Its objectives were to promote, supervise and coordinate fisheries related surveys in Tanzania, to develop and protect the fisheries industry by promoting and developing fish processing, aquaculture, and fishing techniques, to cooperate with both domestic and international governments in coordinating research programs and to provide management training. TAFIRI is accountable to an operations committee and is managed by a director. Like the director of the Fisheries Division the director of the TAFIRI is appointed by the president. The fisheries related educational institutions are scattered throughout Tanzania. These include primarily occupational training and fisheries training schools which were established in the latter half of the 1960s. The thee most important training centers include the Mbegani Fisheries Development Center, The Nyegezi Freshwater Fisheries Training Institute and the Kunduchai Fisheries Training Center. The student corps of these three institutions fall far short of this capacities, by 10 to 40% depending on specific institution. Still all three institutions have remained financially independent. However, because of the limited number of students who attend, their per capita budgets are extremely high at USD 795, USD 3,496 and USD 9,545 respectively. The management of the institutions have not managed to work out, on the one hand, the appropriate balance between public subsidy and tuition and, on the other hand, the appropriate balance between aggregate demand for and aggregate supply of institutional learning. The inefficient management of these institutions is reflected in the high student matriculation fees (USD 4,261 to USD 13,113 per annum). In view of the depreciated equipment and facilities, the centers lack the means to attract students .The current programs offerings of 1 to 3 years together with the high matriculation fees made attendance prohibitive for the artisan, who comprise 70% of the fisheries population, from using the centers. A reorganization of the educational and training centers may be called for. 4.6.c Local Government Authority Some of the authorities of Division of Natural Resources and Fisheries have devolved to the local level of government. Other authorities have continued on a de facto basis to reside at the local level of government. Significantly, different authorities are differently exercised in various specific local government contexts. In general, these include the following: · Provide for and manage fish landings, fish markets and community centers; · Issue by-laws and participate in the formulation of regulations; · Assist with the management of aquatic and coastal protected areas. e.g. marine parks, marine reserves, etc. · Promote aquaculture and quality seed production; · Develop fishermen oriented communities and empower them to act in the interests of their members. 4.7 Findings and Recommendations In 2003 Tanzania adopted a new regulatory framework for the fishing sector. That framework is currently being tested, elaborated and adapted to its practical applications. Significantly, the new framework is primarily production oriented. It focuses primarily on constraining the growth of Tanzania's fisheries to the limits of sustainable extraction established for the country given its renewable resource base. The regulatory framework is designed to prevent over-fishing through constraints imposed on licensees, as well as through limitations imposed on fish extraction methods. As envisioned in the legislation another important resource management activity is the seeding of lakes, streams and reservoirs with new fish. What is most marked, however, about the country's new regulatory framework are the fisheries development issues which it fails to address. These include, most importantly, the vertical structure of the industry, internal linkages among and between specific participants in the fishery to market chain, the development of cornerstone institutions within the sector, the efficiency with which both domestic and international supply chains operate and the collective bargaining ( or other ) basis for determining equitable agreements between organized fishermen and organized processors. In other words, the existing regulatory framework is mute with respect to key issues of supply chain development. Neither is the new regulatory framework any more explicit about assuring that the country's abundant fishery resources are actively developed in ways which improve the livelihoods of fishermen, small scale traders, artisan processors and other impoverished stakeholders. To the goal of alleviating poverty,. the best and most likely to succeed policies are those which facilitate the building up of institutions and then once strengthened, the application of these institutions to the develop of efficient local markets. However, an appropriate legal platform does not appear to exist from which to launch such a reform agenda. As we noted in the discussion above, fish supply chains in Tanzania have developed in two quite distinct tracks..... with one track focused on local and regional markets and one on international markets. Companies which are most active in the international market have developed efficient and competitive chains of their own and are well prepared to develop further sources of competitive advantage in aqua culture and in value added food processing. Small scale fishermen, processors, traders and ancillary service providers who cluster around the second set of chains, however, are not advancing at the same pace as the more sophisticated industrial sub sector. The prospect that efficient chains will spontaneously develop along this second track appear to be remote. In this second tier of competitors, a role exits for donors in partnership with the government to build up institutional capacity. Thus, stronger linkages are required within domestic channels, new business models need to be tested and refined and new and more effective market institutions need to be created. Investment in transport, storage and cold chain handling is required, as well, but these investments will naturally evolve once appropriate commercial institutions have developed and certainly not before. A potential leverage point for the kind of development which is required are the two major wholesale fish markets-- Banda Beach Market in Dar es Salaam and Kirumba Market in Mwanza -- which have emerged within the country. Modifying the underlying business models and the enterprise governance structures on which these two institutions rest provides a particularly interesting entry point for institutional reform. Currently both market institutions are severely constrained by the local purview of the government entities who sponsor them as well as by the strictures and limitation inherent in a tenant/ landlord commercial relationship. Alternative business models need to be tested for both institutions which are not traditional. Thus, for example, both markets might be redeveloped as concessions--concession which would require private sector concessionaires to invest in stronger backward linkages to fishermen, provide affiliated fishermen with working capital, with leased equipment or with productivity enhancing fishing gear. The concessionaire might also invest in improved cold chains or in improved scheduling of fish products onto the market, etc. Better yet, the two wholesale markets might be re-developed as concessions which included elements of a service network franchise. Thus, the terms of a long term concession agreement would require the concessionaire to invest in forward linkages in satellite wholesale markets which were linked to mother market through information systems, frequently scheduled logistics management services, pay-at-either end transfers of product and end to end cold chain coverage. Another possible leverage point are the product sourcing requirements of the new and aggressively growing supermarket chains who have recently arrived in Tanzania or, indeed, the seafood procurement requirements of fast food outlets. Both sets of outlets require significant supplies of fresh fish product. Long term, commercial linkages in the form of " quality food providers" or " core fish vendors" could be developed, initially on a demonstration project basis, around the requirements of these two niche markets. One proof of business concept had been achieved, a larger procurement contract could be specified and competed. An important aspect of this kind of retail-up development is the adoption of food safety standards. Initially these could take the form of procurement specifications but subsequently they might assume the more permanent form of supermarket food safety protocols ( e.g. EUREPGAP Light !) These protocols would be similar to but less compulsory and hence less costly to enforce than the ones which apply to fish products sold into the EU, Japan and the USA. The local protocol analogs of these other food safety standard regimes would include standards for process conformance with " best practices" , real time product quality testing within the chain, occasional audits, third party certification procedures, etc. In this way, a formal fish product chain could be developed which included internal incentives for superior quality and for healthful food delivery. Still another potential leverage point within the legacy fish supply chain are the landing/ market/ processing clusters and commercial focal points around fishing communities which we discussed above. . Again opportunities exist to fundamentally reinvent the institutional foundations on which these commercial focal points rest. New business models could be demonstrated and tested which entail linkages from these fishermen controlled businesses to market centers, specialized market niches and/or consumer focused markets institutions. Ways could be developed to create bankable collateral within the new experimental, fish producing institutions, to cross guarantee bank repayment or to create commercial structures which might be attractive to a joint venture investment partner. In these ways and others, capital could be mobilized for investment in more productive fish production, processing and distribution. In addition, external markets for specialized services--e.g. equipment leasing, logistics management, transport, market information, storage and banking---need to be developed around the periphery of the domestic fishing sector. This can be done proactively by defining service specifications and then outsourcing the desired services to qualified service providers, by joint venturing the development of new services with qualified service providers or simply by demonstrating the commercial viability of new service launches through feasibility studies, business plans or procurement documents. Importantly, as well, new market network services need to be envisioned, specified both commercially and technically and rolled out on a nation wide basis. These network services would provide the superstructure around which a national market could be developed for fish products. Among other services the national network might include the following: i) third party inventory management and temperature controlled storage, ii) cash management, iii) price discovery and data dissemination , iv) insurance and v) transport delivery which was compatible with commodity basis point pricing. Two additional needs clearly exist within the domestic segment of the sector--one of these needs is for technical skills enhancement and for the development of specialized competencies in marketing, financial management and supply chain management. Both the international and the domestic market segments of this Tanzania market require these new skills and competencies. One of the fish processors with whom the study team interacted suggested that the current curriculum for fish sciences and fish culture was much more theoretical than practical, more life sciences oriented than business oriented and more scientifically technical than commercial. He suggested that his company and others would be interested in developing an apprenticeship program with TAFIRI. The larger point that this industry executive was making was that too little cross fertilization and too little transfer of knowledge exists between institutions in the public sector and private sector participants with a need for new skills and new technical knowledge. Based on interviews with the private sector and government officials, the simple conclusion is that the existing system of skills enhancement is ineffective, inefficient and in a state of financial deterioration. The failure to build strong support in the private sector and to develop a practical curriculum together with the lack of a coherent strategic focus have resulted in a system is becoming increasingly irrelevant to the needs of industry and of the job market. The main issues and challenges are: a) Establishing a clear policy and strategy b) Reviewing the governance and private sector outreach framework c) Enhancing the role of the private sector in providing technical and vocational training d) Restructuring the financing of technical training e) Addressing the needs of the informal sector f) Enhance the skills and competencies of technical graduates Mezzo level organizations---like the Tanzanian Fish Processors--can productively be engaged to build bridges and to direct training resources so that they match priorities of the private sector for missing skills and competencies. Mezzo level organizations could also be productively engaged in certifying the professional credentials of trained technicians. In this context, an important role could be carved out for mezzo level institutions like the Tanzania Fish Processors Association in integrating labor markets with skills development programs. . In addition to guiding and directing skills development efforts, Mezzo level organizations can and should play an important role in the fish sector in articulating community needs, building social capital within chains, disseminating best organizational design practices, assimilating responsibility progressively from the government for industry regulation particularly on technical and marketing issues and begin to identify investment opportunities and to assess the risks associated with investment for private investment in supply chain assets which benefit entire communities of users. A organization similar to the Tanzanian Fish Processors Organization needs to be formed to represent the interests and to service as a focal point for development of participants in fish supply chains which serve local markets. Once formed mezzo level organizations need to become the designed representatives of their respective sub-sectors. The new mezzo organization can serve as the institutional conduit from and the counterpart to the Tanzania Fish Processors Organization for the purpose of transferring technology and best practice technique from the export fish processors to domestic fish processors, traders and transporters. The new mezzo organization needs to represent the "silent majority" in the fishing industry in formal regulatory rule makings and standard settings processes, as well as setting priorities for research and for technical knowledge dissemination, they could begin the migration of regulatory controls from the public to the private sector and they could empowered to work in partnership with government in forming new kinds of public-private partnerships. Public-private partnerships are a particularly effective way for sharing capital investment risks between the public and private sectors as well as for testing innovative new business models. Some of these new public-private partnerships need to focus on modernizing and recapitalizing fish landings and on better integrating them into local electric power and highway networks. Others need to focus on modernizing fish markets and improving the assets, skills and incentives which exist within -communities of fishermen and artisan fish processors to add value to the chains in which they participate.. Poor road conditions are a serious problem which inhibit the development of integrated national markets for fish. Public-private partnerships could be organized to support the development of micro infrastructure around markets, landings and fish processing plants. "Access" is a major impediment to the development of a national market for fish, which simply cannot be delivered economically in regions that have bad roads. A case in point is the transportation of fish from Mtwara to Dar es Salaam. The roads which connect these two are impassable during some seasons of the year. Poor road conditions tend to increase the costs of transportation and overheads. The final area in which fundamental reform is called for involves the Balkanized system of taxation which prevails within the fish sector.. In a previous section we discussed how the existing system of royalties, fees, levies and taxes creates perverse incentives among supply chain participants. This is a complex set of issues which call for additional, in depth study. However, in general some level of harmonization and rationalization of the basis for taxing and charging the industry appear to be in order and negotiations over the next generation of investment in the industry appears to be an appropriate window of opportunity for resolving multi-level federal taxation and value-for-money issues with government. . Chapter 5 Lessons Learned from Assessing Three Supply Chains in Tanzania 5.0 Introduction This chapter identifies and discusses the take away lessons that can be learned from analyzing the supply chains for maize, sugar and fish in Tanzania. It is significant that this comparative analysis of supply chains has been anchored exclusively in a Tanzanian context. This anchorage makes the relevance of lessons learned in the context of one commodity/product chain much greater to other commodity/ product chains which are based in Tanzania and assures that the probability of successful transferability of lessons from Tanzanian context to another will be high . Clearly, " best practices" can be identified and abstracted from the supply chains which have been studied. Just as clearly, important lessons can be taken from one commodity context which have relevance to others. The chapter then deals with identifying key take away lessons and with developing tactics for their successful transplant. The rest of this chapter is organized into six sections. Each of these sections deals with a distinct set of take away messages. Thus, section one deals with the need for institutional reengineering. It finds that a significant need and opportunity exist for institutional realignment, redesign and restructuring in Tanzania. Section two deals with the legal and regulatory frameworks within which the three commodity chains operate and finds that some of these frameworks are more effective for building up integrated supply chains than are others. Section three addresses that state of market institutional development in Tanzania. It characterizes some of the foundation institutions which offer prospects for further development and suggests ways to further develop these market foundations. Section four deals with the state of service delivery systems which support supply chains. It also discusses the effects of privatization on supply chain support of service companies who have emerged from privatization with exclusive rights over some specific niche market domain. It recommends ways in which post privatization service delivery can become better adapted to supply chain development needs. Section five assesses needs for and sources of external chain financing. It suggests ways in which chain linked industrial and neo industrial structures can make themselves more bankable. Finally, section six suggests tactics for effecting some of the new directions recommended in the study. 5.1 Need for Institutional Reengineering Commercial relationships between primary producers and processors/ merchandisers are differently structured in the three supply chains which the study team analyzed. In the maize supply chain, for example, almost all commercial relationships are " arms length" and "transactional." These relationships are developed and consummated opportunistically, on a "one-off basis." No risks and little information are transferred through the transactions which take place. Traders or millers buy from farmers at prices which are determined by what they are prepared to offer in the field or, less frequently, through reference to market clearing prices which both parties, buyer and seller, can independently confirm. The later form most frequently takes place in municipal trading centers where many buyers and sellers come together. Traders and millers have no incentive under this system to invest in assets which either enhance production or which reduce unit production costs. Neither do they have an incentive to share risks related to the provision of fertilizers, hybrid seed, modernized cropping methods or any other tangible or intangible pre-harvest investment in farm production. Transaction costs are high for participants in the maize trade, for several reasons. These include the following: i) the cost of discovering accurate price information is high in a market where few third parties have emerged to provide complete, accurate and broad coverage, market information, ii) the costs associated with procuring transportation on a single-truckload ( or single car load) one off basis are extremely high and significantly, iii) strong pressures exists on the "sell side" of these transactions to complete them quickly, upstream in the chain. Sellers have no alternative sources of cash or liquidity except transaction proceeds, and significantly sellers have no way of assuring against the physical loss of the inventory which they hold. As we noted above post harvest losses reach 30% overall and the longer farmers hold their inventory the greater the probability of loss. The result is an asymmetric relationship between buyers and sellers. Sellers prefer to transfer ownership of their product s at their farm gate rather than deliver their products into metropolitan markets where price discovery is more efficient and where multiple buyers compete with one another for new supplies. Transfers take place on a quid pro quo basis whether they be farm gate or nearest market transfers. The transient and " one-off" nature of these commercial relationships precludes any opportunity for risk transfer or for investment in assets which can improve overall supply chain efficiency. Moreover, no incentives operate within these channels to reward producers or traders for superior maize quality. The channels themselves lack both the integrity and the social capital ( in the form of quality control systems) needed to maintain price differences between high quality and low quality maize from one end of the chain to the other. The aggregate effect of these several channel design features is a "poverty trap." In this trap, risk adverse and illiquid farmers are price takers. These price takers have few options except to sell their surplus product when other farmers are selling theirs as well into a " buyers market." In these chains the producers of maize are by nature risk adverse. Risk adverse sellers are likely to move on the first opportunity to zero out any cash exposure they may have assumed earlier in input application/ production/harvest/ selling cycle, this in the form of production inputs. Typically, they have little appetite for additional risks in the form of further commodity price fluctuations, transport market price fluctuations, and the unpredictable willingness of traders to purchase their products for cash. Moreover, by virtue of their lack of industrial organization they have no financial capacity to absorb additional cash costs ( for transport or storage) at the end of their production/ harvest/ sale cash flow cycle. So they sell. This chain itself operates, in the terminology of game theorists, as a "zero sum game." The rules are established by the structure of the chain itself. Moreover, since farmers cannot easily move from low value production to high value production.......since no quality premiums exist in the markets they serve for their maize...... their capacity to escape from this zero sum game or poverty trap is limited. Commercial relationships in the sugar supply chain are significantly different. Here clearly a positive sum game is being played. Mutually dependent relationships are developing between farmers and processors. These are based on programmed and scheduled transfers of raw cane from farmers either directly to processors or indirectly to processors through their agents. In the case Kilombero Sugar, for example, contract agents do the harvesting and inbound logistics management work under contact to the processor. A great deal of coordination is required, in any case, to execute these serial processes efficiently. For example, when cane farmers plant their crops they know precisely when they are scheduled for delivery to the factory operator. Their production schedules are purposely staggered and scheduled to allow for the longest possible period of raw material delivery to the refinery. This is because harvested cane has a limited shelf life and hence the pace of production is determined by the availability of fresh cut cane. Importantly, cane farmers also know the basis on which they will be compensated in advance of their harvest or, indeed, even their planting. Importantly, the transfer price basis between the parties is related to the quality of the cane that is delivered: its sucrose content and the amount of fiber which accompanies the sucrose. The basis for transfer prices is set for a single season or increasingly for multiple seasons, under the terms and conditions of collective bargaining agreements. Thus, a substantial share of risk is transferred backward from farmers to producers. Moreover, a strong incentive is created and shared equitably between the two parties to move from low quality production to high quality cane production. The private sugar industry is still quite young in Tanzania and these agreements are still forming, through trial and error testing. However, as they mature these agreements are also becoming more formal and more contractual. Thus, the commercial relationships that operate in the sugar supply chain are developing in directions which continuously refine the parameters of a joint and collaborative production function. The agreements include the mutual acceptance of contingent risks. The contingencies encompassed include a board range of possible operating and environmental parameters. In this way, even though farmers and producers do now operate inside a single enterprise their joint production function is moved significantly toward optimum global outcomes over an extended set of contingencies defined by contract and controlled through incentive. This may sound a little theoretical but this is exactly what is going on. The sugar sector has three advantages over the other two sectors which set the stage for the development of economically efficient commercial linkages. These include: i) the regulatory requirement, which mandates that sugar refiners and out grower associations operate under collective bargaining agreements. The Sugar Act which we will discuss in more depth in the section which follows mandates that a collective bargaining agreement will govern the commercial relationships of farmers and producers within a specific time after the privatization of each sugar refinery; ii) the nature of farm level and factor level production requires a great deal of coordination and cooperation; and iii) the prior experience of foreign factory owners with out grower schemes which they developed and refined in other venues before coming to Tanzania. With this said, it clearly appears that commercial relationships are progressively strengthening between farm producer and factor operators in the sugar industry, that social capital is being accumulated, that investments both by farmers and by factory owners are being encouraged by the system of stable mutual interdependence which is evolving and that institutions are forming to further strengthen these commercial relationships and to correct problems that may jeopardize their further smooth development. The commercial relationships which have evolved in the export fish supply chain represent a half way point between the other two sets of chain relationships. Thus, export oriented fish processors relate to groups of fishermen exclusively through agents who are more or less independent. Depending on the terms of the incentives programs under which they operate these agents are aligned either with fishermen or processors. In either case, however, agents who are effective in their intermediary roles have build up social capital, on the one hand, with communities of fishermen and, on the other, with fish processors. The relationships between factories and agents are of long standing and are based on a great deal of accumulated trust. Factories will not deal with agents who do not have a " successful track record " for delivering quality products. Some of these relationship are codified in formal contracts. These specify the basis for pay and the minimum acceptable quality condition for the catch. They provide incentives for the delivery of good quality fish and disincentives for bad quality fish. Importantly, they also mandate methods for fishing which support applicable fish resource conservancy regulations. The bottom line is that they seem to work well in creating wealth both for fishermen and for large scale processors. In return for compliance with quality standards for both fish ( size and freshness) and fish catching techniques, these relationships reward fishermen with premium prices. Clearly a positive sum game is being played among the participants in this chain, as well. However, a second set of commercial relations are developing within the fish sector. A second commercial game is being played between fishermen, traders and processors. This game is defined by the rules which apply uniquely in the domestic market chain. Both in organization and in operation this chain closely resembles the one for maize. It appears to constitute a "zero sum game" for participants, in which benefits realized by one trading partner are benefits lost to the next. This chain relies on arms length impermanent, transaction based commercial relationships. It fails to share risk effectively among the participants. It creates no incentives for bringing down high transaction costs and it creates no incentives for high quality, price differentiated fish products. It constitutes a poverty trap for the fishermen who are unable to graduate from this chain to the one which is integrated by the large scale, export factory processors. In general, it appears that the way in which commercial relations evolve in each of these chains is path specific. The initial conditions under which the chain structure developed, the subsequent conditions which facilitate or frustrate further structural evolution, the regulatory context and the incentives created within the markets themselves ( e.g. price differentiated product quality) matter a great deal in determining precisely how chains are organized and what effect their organization has on the poor. The question remains then: Is it possible to transform a chain in which the underlying dynamic is a " zero sum" game into one which involves a " positive sum" game among the participants? Our preliminary response based on a study of these three supply chains is a tentative" yes." The development strategy that appears to have the highest probability of success for effecting this transformation is one that creates supply chain options, alternatives and commercial space for accessing specific market niches in behalf of small scale farmers. Creating this commercial space entails some" institutional reengineering" and this can be difficult, since legacy institutions--including legacy supply chains-- almost always resist change. Institutional reengineering is all about the 3R's: Realignment, Redesign and Restructured. The institutions which are being reengineered are the legacy distribution channels which link farmers and fishermen to specific niche markets. In some cases no linkages exist between farm producers and specific market niches. Strategy realignment involves identifying and then testing ways to access and serve the needs of customers in niche markets for processed fish and maize. Strategy realignment entails finding ways to respond to the specific needs of these niches better than they are being responded to currently. This goal almost always entails more direct connections between primary producers and ultimate customers, fewer layers of intermediation, improved communication in both directions, greater transparency in product flow, lower transport cost and, importantly, more rapid working capital turnover. The next step is to redesign discrete value adding processes so that they are better integrated and easier to manage. Through the process redesign step, value addition for each process is increased, linkages between serial processes are strengthened and community assets which provide this strengthening are invested in up to the point where chain length marginal benefits equal marginal costs.. Redundant or value subtracting processes are removed from the chain or replaced with value adding processes. Next, an organizational restructuring needs to be carried out. A new organizational housing needs to be set up which includes all of the newly aligned processes: New modes of governance and more effective process spanning modes of management need to be developed and implemented. Legacy organizations need to be restructured and new business model designs substituted. In so doing, key value adding processes steps, need to be " internalized" or in-sourced within the chain. Other process steps may need to be outsourced. The alternative forms which the new organizational housing may take include: internal growth in new process directions, corporate acquisition, contractual alignment, cross ownership, joint venture or shared information and control systems. Restructuring is precisely what the export fishing chain has done with transportation and cold storage warehousing functions, by pulling both of these activities into their supply chain. Industrial exporters have moved these functions which were value subtracting inside their domain of corporate control and thus made them over as value adding functions. In the process of restructuring chains that have become poverty traps it is important to assure that value adding activities which take place on the farm or in the fishing boat are combined with other value adding activities both upstream and downstream in the chain. Specific "institutional reengineering" activities which might be considered in the context of the domestic market fish and maize chains include the following: · Create incentives for strong backward linkages from established merchandisers/ processors to maize farmers and fishermen where no incentives currently exist. The " incentive" leverage points in the maize sector that are probably most immediately available fall under the food security programs. Thus, the pre-qualification criteria or final award criteria for procurement of maize under that program could be used to effect the restructuring of the underlying supply chain for maize. Over the longer term, however, tax and trade incentives might be used to leverage the strengthening of backward linkages from processors and large scale merchandisers. Through the incentives which it creates government can influence industry structure significantly. This influence is, in any case, tacitly at work already in both the fish and maize sectors. Unfortunately as we have pointed out above its current effect is the creation of an industry structure which does not support the interests of poor farmers or fishermen. · Invest in demonstration projects which involve building up marketing channels from the producer end. Demonstration projects are worth exploring if the Government is prepared to translate experiments, which prove successful, into scaled up programs. This entails nothing less than the redefinition of the mission of key agencies of government who are responsible for the development of specific sectors. The " grain bank" demonstration project example which we discussed in Chapter 2 is a good example of a " bottom up" marketing channel development initiative. " Grain banks" have been successfully organized both in Western Kenya and Uganda where they have proved effective in shortening supply chains, serving niche markets extremely well and working transaction cost out of the system. Other "bottom up" experiments are being tested in Uganda. These involve new forms of farm based business models. Similar organizational experiments might usefully be conducted in Tanzania. · Develop backward linkages from wet markets and from fish markets to farmer/ fishermen producers. Several of the municipal market institutions upon which we commented in chapters 2 and 4 have reconstituted themselves as profit making entities in recent years. Thus, the merchants organizations which unpin both the largest fish market and largest farmer's market in Dar have accumulating substantial community assets in the form of negotiable securities, fixed plant and equipment. The study team discussed with leaders of these organizations plans for future development and the possibility of their investing in backward linkages to the farmers and fishermen who supply them with inventory or, for that matter merging with other similarly constituted market organizations located in interior markets to form a merchandising network. Both groups are quite interested in moving in these directions. The backward integration of traditional market institutions is a particularly interesting way to begin to transfer risk from the weakest link in the chain to the participant in the chain who is best able to manage that risk. Thus, backward integration can provide working capital for investment in yield enhancing farm inputs and fishing gear in return for quality differentiated products and quality differentiated retail offerings. · Developing third party supply chain support services which make farmers less dependent on second party buyers. In the case of maize, the development of a new service industry would be most beneficial to farmers whose legacy supply chains are missing a critical value adding step. This new service industry would be made up of third party warehousemen who would furnish grain storage capacity, fumigation and other grain preservation services to farmers. These third party service providers would refrain from buying and selling maize for their own account and thus they would become supporting partners with farmers in positive sum games. Rather they would allow farmers to store their grain without treat of its value deterioration and also possibly allow them to create bankable security interest in grain inventories under their control. A similar set of third party services which would be most helpful in proving another exit from the poverty trap would entail the provision of market information services which are immediately relevant and accessible to farmers and fishermen. Still a third set of services would involve the development of third party maize or processed fish logistics management services compensation for which came from products sold into forward markets. All three of these sets of services could be developed through a competitive tender for business services. A market for these missing services could be created by first specifying them and then offering them to be fulfilled for ensembles of farmers or fishermen who were restructured into procurement cooperatives.. The tender would specify both the services required and the supply chain consequences which were anticipated. Compensation would be based on the successful realization of anticipated outcomes. 5.2 Legal Frameworks: Some Better, Some Worse The legal frameworks, which apply to each specific product/ commodity chains which the team studied, differ fundamentally one from another. Each of these frameworks is summarized in the tables below. The frameworks were developed at different times and this difference in timing may explain some of differences in underlying regulatory philosophy. Thus, the regulatory frameworks for sugar and fish were enacted relatively recently, both in 2001. The framework for maize, however, dates back to 1991. It is the oldest of the three. Two of the frameworks replaced predecessor marketing systems which were dominated by the state, either in the form of a marketing board ( e.g. the maize marketing board was abolished in 1987) or a state owned industrial sector ( e.g. the sugar industry was exclusively state owned until 1999.) Only the fishing industry which has grown from a near zero base in 1990 developed without the historical legacy of either direct or intrusive government intervention. In all three cases the current regulatory frameworks define a new boundary line between public and private sectors. They devolve specific roles for enforcing regulation in each of the three sub sectors and assign these roles to specific agencies of government Significantly, some of these agencies are more open to direct private sector participation in the regulatory process than others. Thus, the Sugar Industry Act, creates an independent "Sugar Board" which it entrusts with extensive regulatory powers. Importantly as well, it designates the specific private sector stakeholders who are to be represented on the new Sugar Board. Significantly, neither the Fish Act nor the Food Security Act provide for direct private sector involvement in regulatory decision making. Nothing in either Act pulls the private sector into either sector policy making or regulatory deliberation. Though as we discussed in Chapter 4 the Fish Processor Association of Tanzania is not reluctant to make its position on critical issues known to the Fisheries Division of the Ministry of Agriculture and Livestock Development which is the designated regulatory authority. The Food Security Act creates a board of directors and defines a liaison relationship for that board with the Food Security Department. In this way it allows for some measure of private sector involvement. However, the food security board has no decision making authority, all of which resides with the Minister of Agriculture and Live Stock Development. The board simply enjoys advisory powers. The Fish Act fails to reserve even an advisory role to the private sector in decision making matters. All powers under the Act reside with the Minister of Agriculture and Live Stock Development. The three frameworks also differ in the objectives which they espouse for regulatory intervention in private sector operations.. Thus, in the case of the maize market, intervention is limited to periods of food shortage and during these formally designated periods the objective of regulation and direct market intervention is to assure food security. In the case of fish, the regulatory objectives espoused have to do primarily with assuring fisheries resource sustainability. All other objects are subordinate to these. The objectives underlying sugar regulation, however, go to protecting and balancing the economic interests of the four sets of private stakeholders who are represented on the Sugar Board: consumers, out growers, sugar processors and manufacturers of processed food who depend on industrial sugar inputs. Thus, only one legal framework--the Sugar Industry Act-- deals explicitly with issues of supply chain efficiency and supply chain equity. The three legal frameworks also differ importantly in the means and modes which they provide for enforcing regulations. Thus, all three Acts empower a designated agency within government to issues licenses for operators, manufacturers and other participants in the chain. They also devolve to regulators the right to revoke these licenses and permits or to impose fines or penalties for a failure on the part of licensees to conform to regulatory decisions. The maize and sugar frameworks empower regulatory agencies, as well, to effect trade policy. In the case of maize the Food Security Department is empowered during a food shortage emergency to impose bans on exports, or conversely to limit the number of import licenses during periods of food surplus. As we noted above these powers appear to be a odds with Tanzania's commitments under the EAC trade agreement. The Sugar Board is empowered to raise or lower import tariffs, though practically these activities are undertaken only in close coordination with the Ministry of Foreign Trade. Most importantly, the frameworks differ in the way in which they support specific forms of industrial organization and of supply chain configuration. Thus, the regulatory frameworks for fish and for maize are mute on this subject. The regulatory framework for sugar, on the other hand, explicitly endorses and facilitates the development of a processor/ affiliated out grower structure for the industry. Moreover, the Sugar Industry Act specifies that processors shall enter into collective bargaining agreements with local organizations representing out growers and that any disputes arising either in the negotiation of these agreements or subsequently in their interpretation and enforcement will be arbitrated before the Sugar Board itself. Importantly, the Sugar Industry Act also designates that both a Sugar Processor Organization and an apex organization representing regional outgrows will be represented on the Sugar Board. Hence, it calls into existence two sets of mezzo level organization and fashions an agenda for them to work on. On the surface, the framework for maize regulation appears to be the most liberal and the most non interventionist and hence to have the least effect on industry structure. In 1987 the government formally deregulated grain markets in Tanzania and hence ended direct intervention and tactical control over grain marketing. The Food Security Act of 1991 set in place a Food Security Department which could " assume limited regulatory powers" during crisis periods of food shortage. The "Catch ­22", however, is that the Tanzanian economy has been in a state of almost perpetual food shortage since the Act was enacted.. Thus, the limited regulatory powers granted under the Act have been in effect for some time. Moreover, these powers have been exercised in ways which have caused significant market distortions. They have created incentives, for example, for a trader oriented industry structure and have seemingly advantaged one set of market participants vis a vis others. The Fish Act is mute on issues of issues of market intervention. Implicitly the Act embraces a laissez fair approach with respect to fish market access. Only when a series of issues concerning compliance with EU standards for food safety arouse five years ago did the Fisheries Division strenuously exert its regulatory prerogatives with some significant effects on the industry and its supporting supply chains. However, it asserted these with the full compliance and support of export processors. It is also worth recalling that the industry has grown rapidly since 1990 and during this period its structure has changed and, indeed, it continues to change in response to regulatory challenges primarily from outside. Second tier exporters, for example, have risen in their levels of global competitiveness as a direct result of industry wide efforts to gain conformance with EU and Japanese standards. In dealing with external challenges the industry has tended to respond as coherent group with the Fisheries Division acting as the spearhead for the group and its agent for external regulatory compliance. Moreover, the same processors who operate in Tanzania also operate in Uganda and Kenya. The largest of the export processors have, in fact, migrated from those two venues. Since the fish processing industry depends in large part on Lake Victoria resources and these resources are fungible among processing plants located in all three countries, the industry collectively exercises some counter leverage vis a vis Tanzanian regulators. This is a counter leverage which participants in the other two other sectors do not enjoy. Regulatory Framework for the Sugar Sector Definitive Principle Counterpart Powers and Key Recent Regulatory Legal Governance Private Sector Authorities Regulatory Decision Framework Mechanisms and Institutions Decisions Making and Regulatory Adjudication Institutions Processes Sugar Industry Sugar Board of Tanzania--Powers Board has statutory Extremely broad powers are Key regulatory decision The board possesses the Act, 2001 vested in a Board of Directors. authority to promote the devolved to the Board, made to date include: power to regulate its own Board, however, is subject to the formation of private sector including: i) Allocation among procedures with respect to direction and discipline of the associations within the i) -License all sugar imports producers of the EU export meetings and the proper 2001 Legislation Minister of Agriculture. sector and to monitor their and exports; quota; conduct of it business. All replaced completely Responsibilities are divided between activities. An association of ii)-License sugar cane ii) Development and decisions of the board are pervious legislation Board and Director of Crop Dev and sugar producers and three growers, manufactures and enforcement of a based on " majority rule" which vested Tech Services in M o A. A board of regional associations of industrial users; negotiation process and a and four members parastatal companies eight is appointed by the MoA. sugar producer out growers iii)-Impose levies within the complementary dispute constitutes a quorum for the with exclusive rights However, board members are have been formed since the chain; resolution process for purpose of voting. to farm and refine nominated and selected from/by implementation of the Act. iv)-License the importation effecting long term specific stake holder constituencies. The Act itself reserves and use of new genetic contracts between local The board has recently sugar. National Sugar Institute--Primary positions on both the Sugar material; associations and major published its rule making responsibilities involve human Board and the National v)-Appoint inspectors of sugar refiners; proceedings and has resource development and training Sugar Institute Board for processing facilities, fields, iii) Review and consideration established a process of within all links of the sugar supply representatives of both the warehouses, etc. of increasing import duties arbitration and dispute chain. Receives and disburses funds Tanzania Sugar Producers vi)-Monitor methods of on industrial sugar; settlement which operates from training levy and payable by Assoc and the Tanzania pricing, selling and iv) Development of more under the contracts that employees under the Training Sugar Growers Assoc. purchasing within the chain. effective enforcement have been negotiated Authority Act, 1994. A board of seven Increasingly the Board is vii)-Impose and collect methods for assuring that " between specific sugar is to be appointed by the MoA. playing a role of arbitrator excise taxes on importers industrial sugar" does not manufacturers and the out However, board members are and mediator in resolving and manufacturers. find its way back into the growers who provide them nominated and selected from/by differences between these general economy. with inputs. specific stake holder constituencies parties. (e.g. Tanzania Sugar Producers Assoc and the Tanzania Sugar Growers Assoc). Regulatory Framework for the Maize Sector Definitive Principle Counterpart Powers and Key Recent Regulatory Legal Governance Private Sector Authorities Regulatory Decision Framework Mechanisms and Institutions Decisions Making and Regulatory Adjudication Institutions Processes Food Security Act, Food Security Department: The No formal linkages exist Broad powers are granted to Key regulatory decisions The Board of Directors has 1991 FSD was established as an between the activities of the the Board, including: i) made to date include the no independent powers independent department within the Food Security Department coordinate food security following: i) Imposed ban from those of the Minister Ministry of Agriculture. and private sector policy; ii) implement food on cereal exports when of Agriculture, at whose Act No. 10 of 1991 is organizations. No system security programs; iii) major food shortages pleasure it serves and the primary The Board of Directors: The Board of governance or of monitor the country's food threaten. Subsequent bans through whose authority it legislation that deals of Directors of the FSD is composed consultation exists between situation; iv) cooperate and have remained in effect acts. No formal hearings, with the residual of Government officials and other either the Food Security exchange information at the since 2001; ii) Reject the fact finding reviews or other regulation of maize private persons approved by the Department or other international level; v) report importation of GM maize as checks and balances exist markets. Maize Minister of Agriculture and Livestock branches of government on the level of exports of well as the cultivation of GM on the advice provided by markets were Development. The Board oversees involved with maize markets major cereals; vi) manage maize within Tanzania; iii) the Board or on the decision and coordinates the activities of and the private sector. procurement and all other Identification of SGR assets making authority of the formally deregulated Government specifically with regard Policies which effect equities aspects of the strategic which can be privatized and Minister with respect to the in 1987. The Act to the procurement, storage and among private sector grain reserve; vii) impose the offer of these assets to establishment of a food set in place a Food release of grain for security suppliers are adopted and grain embargoes against private parties for sale or security crisis situation Security Department purposes. The Board gives advice to implemented without public exports; viii) monitor lease. subsequent to which finding to assume limited the Ministry regarding emergency comment. cultivation of food crops. the government has the regulatory powers food relief during periods of food right to ban export bans. during crisis periods shortage. Similarly not vetting or of food security and public review needs to take Minister of Agriculture and place with respect to the to oversee the Livestock Development: The procurement, storage and strategic grain Board of the FSD acts primarily release of maize, etc. reserve and to through and with the approval of the provide for other Minister whom the Board advises. matters incidental to the establishment and management of the grain reserve. Regulatory Framework for the Fish Sector Definitive Principle Counterpart Powers and Key Recent Regulatory Legal Governance Private Authorities Regulatory Decision Framework Mechanisms and Sector Decisions Making and Regulatory Institutions Adjudication Institutions Processes The Fisheries Act, The Act does not · Provide for and regulate the Key decisions recently made Section 3 of the Fisheries 2001 Fisheries Department: The designate, authorize or conditions under which by the Fisheries Act states that the Director Director of the Fisheries Department create any specific industrial fishing shall be Department include: may from time to time issue is an officer in the public service counterpart institutions undertaken in Tanzania. and publish circulars and The Fisheries Act sets who must be a professional with in the private sector · Issue, suspend and/or cancel · Lifting the ban on fish manifest directives that are out the legal expert credentials in fisheries with whom the licenses granted to private exports from Tanzania to in conformity with the framework for the sciences. He acts as advisor to the Fisheries Department is companies to catch fish, to the EU. provisions of this Act. environmental and Minister of Agriculture and Livestock required to consult or process fish and to export fish · Setting in place a in all matters related to management liaise. from Tanzania. regulatory framework to In Section (4) The Director economically of fisheries. The Fisheries · Regulate and control the be followed by fish is required to make a sustainable Department is the primary description, specifications and exporters to the EU. This determination affecting or management of implementation agency of form of nets to be used in was basically drawn on likely to affect the rights of fisheries in Tanzania. government. fishing and the size of their the lines of quality and any person or the mesh; Limit or control the hygiene. opportunity for any person It provides specific number and size of fishing · Imposing a ban on the to undertake a given activity authorities of Minister of Agriculture and vessels. used of illegal fishing and shall give that person government to Livestock Development: The · Prohibit, restrict or regulate the methods such as reasons regarding the public determine who can minister is responsible for policy importation into Tanzania of dynamite fishing and the interest for such decisions. formulation and ensuring the live fish, other than fish use of illegal fishnets. fish, what fishing and execution by officials in the ministry indigenous to the country. · Establishing a methods Section (5) provides that processing methods of all functions related to the · Prohibit or restrict the for assuring that fish any person aggrieved by the can be used and implementation of the Act. capturing, collection or resources will be decision of the Director what commercial destruction of any variety of conserved and sustained. made in accordance with Missing Mechanisms: No fish; Provide for the protection This methods involves subsection (4) may within methods can be used mechanisms exist under the Act-- of critical habitats; Prevent the designating the minimum thirty days appeal to the to process and trade either in the form of hearings, board pollution of territorial waters. size of fish which can Minister. fish. membership, formal consultation or legally be processed and comment on pending rulemakings-- ·Regulate the marketing and setting penalties for the for private sector inputs and for processing of fish; Prescribe catching and processing checks on government's authority. conditions under which every processor of fish shall comply. of immature fish. The lessons that can be learned from a review of these three regulatory frameworks can be summarized briefly in the following bullet points: · Regulation can and does effect supply chain structure. The ways in which regulations are interpreted and applied as well as the substance of these regulations can and do have structural effects on the underlying sectors, as experience in all three sectors demonstrates. Thus, even is no explicit regulatory authorities are granted over issues of supply chain development, the tacit effects of policies and regulations do, in fact, effect supply chain development. Given this background, explicit regulatory review and discussion of industry structural issues may be more appropriate and beneficial over the long term than their avoidance. Because structural issues have a significant effect on the livelihoods and welfare of primary producers, these issues need to be drawn explicitly into the regulatory domain. In this respect, the Sugar Industry Act defines best regulatory practice. · In spite of statutory limitations, the domain of issues over which regulation applies is variable from issue to issue and does, in fact, frequently encompass supply chain organization. No patents or prescriptions appears to exist with respect to what falls strictly "inside" and what falls "outside" the regulatory domain for the three commodities. As the discussion above demonstrates issues of regulatory compliance, for example, with the food safety regimes of importing counties, as well as the trade policy consequences of domestic grain sufficiency have multiple consequences for supply chain competitiveness, equity among participants and, even, consumer welfare. Issues of supply chain development are implicitly considered in most regulatory decision making and consequently should explicitly become matters of regulatory deliberation and of public debate. If a determination is made that supply chain issues fall within the domain of the nation's public interest, there are better and worse ways to deal with these issues. Again lessons can be learned from the example of the sugar sector where issues of equity and efficiency within the chain are dealt with in ways which are minimally intrusive. The sugar regulatory framework forces the private sector to deal with issues of backward and forward linkage without active intrusion on the part of government. Indeed, it is the threat of that intervention that induces supply chain partners to negotiate in good faith. · Best practice standards are set in the regulatory framework for sugar. The regulatory framework for sugar provides several best practice examples, including: a greater degree of regulatory independence from government, engagement of private sector stakeholders directly in regulatory and policy deliberations, the definition of a broad domain of issues for regulatory consideration, provision of alternative modes and means for the resolution of issues concerning supply chain structure and holding out the possibility of regulatory intervention only in the case of need for dispute resolution and arbitration among private parties. Lessons learned from the regulatory framework for sugar can and should be carried over to other sectors. 5.3 Market Institution Building Market structures and market institutions continue to evolve, albeit slowly, in Tanzania. Specific policies could be taken to accelerate this development. This section discusses the state of domestic market development for maize, fish and sugar. The basis on which these market structures currently operate is primary around the primary trading points of Dar es Salaam and Mwanza where large numbers of buyers and sellers come together. As we discussed in Chapter 2 several distinct niche markets have developed within the country, for all three commodities.. Prices lines for these several market niches follow one another generally with prices in specific niches leading and in other niches following. As we noted above the development of distinct niches is generally beneficial for primary producers. It allows them to differentiated their products and processes and hence creates additional commercial space and allows for wealth accumulation. A coherent geographical pricing structure has developed for maize which is centered around the two major trading points. In these two centers several important market making activities take place: i) "basis" prices are formed in day to day transactions between multiple trading parties, ii) third parties accumulate this price information in a timely and accurate form and transmit it to other, iii) traders in other parts of the country develop reference prices by deducting transport costs from the basis price which is clearing local markets in either Dar or Mwanza. In this way nation wide basis point pricing structures begin to emerge. These market pricing structures, however, are not fully integrated over the entire country, in part because transport costs represent an extremely large portion of the total cost of delivered maize in many corners of the country and because transport prices themselves are variable. Indeed, transport prices into some parts of Tanzania exceed the basis price of the maize itself, in which case trading from these production points into Dar and Mwanza breaks down. These production points are "out of the market." Still other costs, including those associated with trading with incomplete information and high levels of spoilage and loss, add to variable transport costs to make up total transaction costs, which are large for maize. Moreover, as we have commented above the ways in which both maize pricing and transport cost information is currently transmitted to and through traders is different from the way that it is transmitted to farmers. This is continuing source of market failure. It reinforces asymmetries between buyers and sellers which already exit in the procurement of transport and dry storage.. However, as we noted in Chapter 2, this failure is in the process of being corrected by third party providers of market information services like Food Net. The further development of market information service providers and the further refinement of their knowledge products should be a high priority for market development. Importantly, as well, the two railways which operate in Tanzania provide an extremely valuable market support service. Their published tariff rate structures represent a stable transport cost structure from which reference maize prices can be inferred from basis point prices. Still, as the two railways become progressively less able to deliver transport services to maize shippers the utility of their tariffs for maize marketing is correspondingly diminished. Prices for highway transport are more variable and more asymmetric, i.e. different rates are quoted to different " beneficial owners" of the cargo. A parallel set of developments are taking place in the domestic market for dried and otherwise processed fish. This market is beginning to develop to the stage of the maize market with broad national distribution of processed fish but only partial integration of pricing mechanism based on a basis point structure. The market for processed sugar is evolving somewhat differently. In the sugar market private processors are trying to alter the market rules They are endeavoring to develop proprietary channels through which they can move their branded products. In order be protected brand equity the two largest producers require strong control over their marketing channels. In order to justify investment in brands producers need to control the positioning of their products at the retail end of the chain. Thus, producers are trying to assume more direct control over retail price, shelf space and retail store promotions. To date, however, these efforts have only partially succeeded. A large whole sale market still exists in Dar which trades in commodity ( non branded) sugar. Significant organizational changes are underway, as well, with respect to the traditional market institutions--the institutions where prices are primarily negotiated within legacy chains. As we noted above, the organizational basis for traditional open or " wet" markets falls under the regulatory control of branches of municipal government. Governments have traditionally related to traders who deal in these markets as land lords relate to tenants. However, within these markets informal organizations have developed. A least in some of the larger of these markets, these informal organizations of resident traders have gradually accumulated assets. For the most part these assets are intended to serve as a collateral base for insuring the bank borrowings of individual traders. However, they are being considered for other strategic market development purposes as well. The largest " wet" market in Dar, for example, has evolved a organization which has a formal corporate charter, system of governance and which has accumulated substantial liquid assets, in the form of shares traded on the Dar es Salaam exchange. One of the priority development needs of market institutions within Tanzania is to improve their access to capital and thus improve their capacity to assume risk, their capacity to store and hold larger inventories of traded products and their ability to better hedge their positions. This later objective can partially be achieved by holding inventories in different trading locations. Market institutions appear to be evolving into geographic networks. However, accelerated development clearly should be a high priority. Just as their backward integration is a high priority. One of the limitations that continues to exist with commodity trading in Tanzania is that trades are executed only on a quid pro quo basis with both buyers and sellers having to be represented at the time and location of the transfer. This limitation imposes very high transaction costs on trades. As we noted in the case of maize these include, out of route transport costs, handling, loss and damage, re-bagging, etc. The future development for market institutions is, however, toward the creation of security interests in standing inventories. These security interests can be created with the assistance of third party public warehousemen or commodity market nominated asset managers. Securitization involves the creation of a new category of inventory backed asset. The value and commodity specifications of these assets are guaranteed by an asset manager and co-guaranteed either by the market itself or by a fourth party insurer. Once a security interest in created a more efficient market institution is required for the trading them. This institution is either a commodity exchange or a virtual internet market. Several initiatives have been launched within the EAC to develop new commodity exchanges and/or new virtual internet markets. None of these initiatives, however, have proved notably successful to date. The most successful, the best capitalized and most sophisticated commodity exchange in SSA remains the Johannesburg Exchange. Maize traders based in Uganda are exploring the opportunity to extend the trading reach of the Johannesburg Exchange into Kampala and to establish a franchised relationship with the Exchange based full compliance with its asset management guidelines and full conformance of specifications for traded products with the requirements of the Exchange. In effect a new Johannesburg trading point and pricing basis would be established for securities back by and traded through the Exchange in Kampala. This is a very interesting development to watch and its implications for the future development of markets in Tanzania is quite significant. Me-too opportunities may make themselves available to Tanzanian traders big and small if the Ugandan Grain Exporters succeed. 5.4 Refining Service Support Systems: The Post Privatization Agenda Each of the supply chains which the study team assessed requires a distinct set of specialized services to support it. These services may include, for example, inbound and outbound logistics services, information services, warehousing and asset management services, etc. Some of these services are supplied out side the chain itself by third parties. This is called "outsourcing." Outsourcing is the prevailing paradigm for the most supply chains in Tanzania. One example, is the supply of packaging materials and the management under contract of the inbound logistics for field cane supply to Kilombero Sugar. Almost all transport and IT services are outsourced. When service suppliers are ready and able to develop specialized services and to continue to support the chains in which they participate in ways that continually push the technology frontier, outsourcing is, indeed, the preferred strategy. It minimizes the application of limited capital resources within the chain. No competitive advantage is either lost or gained by using third parties who are able to operate at or near the competitive frontier of service delivery but resources are conserved for investment in other more strategically significant investments. The situation is quite different, however, when third party service providers are not able to operate at or near the competitive frontier. In these cases, the chain forfeits competitive advantage to other chains which are able to gain more value from their third party supplied services. Under circumstances where value is subtracted by third party providers of essential services, chains sometimes internalize or in-source the provision of these specialized services. On example of this kind of response is the in-sourcing of cold storage services within the export fish chain. Another is the in-sourcing of highway refrigerated transport and of ocean transport again by the same export fish chain. By in-sourcing chains are able to transform a value subtracting activity into a value adding activity or a value neutral activity. In general, the more competitive markets operate for specific support services the more likely that individual competitors will be operating at or near the competitive frontier. Moreover, the larger the number of service providers who are active in a specific service market the more likely it is that a supply chain will be able to find the kinds of customized services which it requires in that market and will not need to in-source. In the context of Tanzania, however, three sets of obstacles persist which prevent supply chains from operating as competitively as they might otherwise be able to operate without third party providers of essential services. One of these obstacles involves the state of physical infrastructure upon which private third party service providers operate. The condition of the roadway infrastructure in Tanzania, in particular, is very poor. The central corridor is a poses a real problem for third party truckers. It allows them to operate at only very low speeds, hence it reduces their asset utilization. In addition, it substantially increases their maintenance and repair costs. Both of these factors are reflected in inflated pricing schedules. Moreover, the market for trucking services in Tanzania has not developed to the point where truck capacity is sold forward in anticipation of where subsequent backhaul load capacity will be made available. In many trucking markets around the world, "fourth party" market makers have begun to use the internet and other forms of IT to produce trip plans, to sell freight hauling capacity forward and to actively assets for trucking companies on a real time basis. The use of IT has made the programming and scheduling of productive truck hauling capacity possible. The development of "fourth party" market management services has not yet begun in Tanzania, however. Similar missed opportunities more generally effect other specialized service markets, as well. A second set of obstacles involves the slow pace of privatization and the continued dominance of state owned service suppliers in specific sectors. This problem is most acute in the railway sector where two state owned rail carriers continue to operate and to progressively year over year subtract value from the supply chains they support. These carriers are managed by dedicated, heroic and capable people but they lack the capital, the management flexibility and the strategic commitment from their owners ( the Government of Tanzania) to develop and supply reliably customized transport services. Neither the TRC nor the TAZARA railways have received any capital funding from the government since plans were launch 7 years ago to prepare them for privatization. During that period, remarkably the traffic volume of the TRC actually increased. However, the condition of both rail car fleets, the locomotive fleets and of the track and structures of the two railways has continued to run down for lack of capital to the point where neither carrier is able to handle the traffic which is tender to it efficiently or reliably. In chapter 2, we discussed at some length the perverse effects which preferential allocation of rail equipment is having on the maize market. Other adverse effects on supply chain development are also taking place in other chains. For example, uncertainties associated with the future rehabilitation and renewal of the railway have an adverse effect on industries who may want to locate their facilities on the railway but who are uncertain about its ability to deliver reliable service in the future, let alone specialized and customized service. The third set of obstacles involve the constraints and limitations for service specialization and customized service design which are implanted in many of the privatization transactions which have been completed. This is an issue of basic issue of post privatization service network connectivity. Many of the privatization transactions that have been completed to date award exclusive service rights to the award winners. Thus, for example, the award of ground handling services at Dar es Salaam airport includes the provisions for service network under an exclusive concession.. The two railways are similarly being prepared for privatization as vertically integrated rail operating companies with exclusive market franchises over the entire domain of local stations served by the two carriers. However, in order to customize services for specific users or chains and to allow the possibility of competitive challenge motivate risk taking and innovation, it is frequently necessary to go well beyond standard one-size fits all, common carrier service delivery and beyond sole service providers exclusivity. Specialized or customization services entail incremental investments, the development of new processes, the effective integration of old processes into new service packages, new systems and new bases for dividing revenue with service partners. Most importantly they require motivation and incentive to experiment. If the incumbent service provider in a newly interconnected network either cannot or will not participate in a joint service delivery effort the customization effort is frustrated. Two examples from the air freight sector highlight this issue: KLM operates a Boeing 767 aircraft on a regular scheduled basis-- 7 flights per week--- from Amsterdam to Dar es Salaam via Kilimanjaro Airport and return direct to Amsterdam . In order to improve increase its available capacity for handling outbound cut flowers from Arusha, KLM attempted to discharge as much of its inbound dry freight as possible at Kilimanjaro and to move that freight via truck from Kilimanjaro to Dar. In Dar cut flower consignments, however, needed to be removed from the belly of the plane and stored in a cold warehouse because of the delay time between arriving and departing and the ambient temperate which is quite high in Dar. However, the concessionaire who has the exclusive right to provide ground handling services in Dar has insufficient cold storage capacity to make this relay operation work and apparently too little incentive to invest in additional cold chain capacity. . As a result both KLM and Tanzanian based flower shippers are penalized. KLM has offered to provide its own ground handling services in Dar and to invest in a cold storage warehouse but the concessionaire is unwilling to give up its exclusive market franchise right. The carrier has not regulatory or other recourse. A similar circumstance effects the handling of air freight from Dar es Salaam International Airport via BA to London. BA operates three flights a week between Heathrow and Dar and return. Its inbound and outbound schedule times via this route are tight and the same ground handling company is unable to unload and reload the 767' s belly with a full consist of freight in the time which the carrier has available. Hence many of the BA flights depart light. This is most unfortunate since most freight is inbound into Dar. Both airlines--BA and KLM--- would very much appreciate the incremental revenues which fall to their bottom line from incremental outbound cargoes as would Tanzanian shippers. The larger issues that emerges from these two examples are the need for a post privatization regulatory strategy which assures that privatize incumbents in new service franchises have both a regulated requirement to respond to opportunities for business growth which fall outside their comfort zone and the need to enforce equitable revenue shares in the case of new interconnected services which are fairly compensatory to all participants. This is an extremely important issue in the case of railways, container intermodal services and digital data services to mention just three. Resolving this issue is part of Tanzania's post privatization agenda. The box below provides one example of a more general strategy which might be considered when privatizing critically important network based service providers in Tanzania. An opportunity exists in the privatization process to " deconstruct" the vertically integrated service delivery systems which these providers operate and to allow users to reconstruct these systems from components subsystems and from component assets. In this way, traditional and hide bound network industries might be redeveloped in directions which significantly increase the value which they deliver to users. The opportunity to defrost frozen assets ( e.g rail car assets) and to use them more productivity in the development of customized rail based logistics services is the example cited in the box. Many other examples might have been similarly cited. ---------------------------------------------------------------------------------------------------------------------------------------------------------------- Specialized Rail Car Leasing Company The Tanzania Railway Company has more than 200 rail cars in "bad order" status, meaning that they are sidelined and in need of heavy repair in order to become operational again. The carrier has been unable to fund the repair and rebuilding of its rail car fleet over the past five years as its discretionary cash flows have continued to diminish. One of the most acute problems that shippers face with respect to TRC's service has been the declining availability of serviceable freight cars. This problem has only become exacerbated over the past 12 months with increased demand for rail service. T A closely related point is this: The existing, serviceable car fleet available on the TRC is a general purpose fleet whose cargo handling capabilities are mismatch with the freight hauling requirements, densities, product flow ability, etc. the specialized commodities, which shippers require. General service cars consist primarily of covered box cars, open gondolas and flat cars. None of this general purpose equipment is particularly well adapted to handling specialized commodities such as ISO containers, flow able maize, sugar, dried beans, cement, building materials, etc. The general purpose design of the existing fleet discourages shippers from investing in fast loading and unloading technology. This fact, in turn, further diminishes the asset productivity and the ton kms. per year that individual cars can generate. When a private sector railway operating company is finally selected to operate the TRC , one of the obligations that they can be expected to take up is the obligation to rehabilitate a large number of " bad order" cars and to return them to revenue generating service. The financial obligation associated with this portion of the proposed concession is $ one million. Indeed, this obligation represents a substantial risk factor from the point of view of potential investors and diminishes the value of the rest of the concession to them. An opportunity exists to provide specialized equipment to shippers and to make shippers into partners of the private rail operator. This opportunity would involve the development of a private sector rail car leasing company and a corresponding set of pricing incentives on the side of the new rail concessionaires which would induce shippers to lease and use specialized rail equipment. This opportunity could be developed and implemented under a project which would entail the following: · Bad order cars would be sold for scrap value by the Tanzania Privatization Commission to a rail car leasing company. The leasing company would make a commitment to rebuild and recapitalize the equipment and to make it available to shippers/ consignees on a long term lease basis. · The rail car leasing company would contract with a specialized rail car building company to rebuild the bad order equipment. · At the same time the rail car leasing company would enter into long term agreements with shippers/ consignees to provide specialized equipment under long term, wet leases. · Shippers and consignees who entered into these agreements would enter into back-to-back service contracts with the new railway concessionaires. Under the terms of these agreements, service standards would be set, price discounts below published tariff rates would be established and specific enforcement incentives, penalties for non- performance would be set up. The role of the World Bank under this proposed arrangement would be the following: · The Bank would facilitate the development and execution of a privatization transaction for the sale of the bad order equipment consistent with the commercial parameters needed to move the project forward. · The Tanzania Privatization Commission with support from the World Bank would develop a business plan, complete a due diligence review and tender the cars. · The Tanzania Privatization Commission would make necessary stipulations to the proposed railway concession agreement and would adapt the pending railway regulatory proposal to accommodate the use and operation of a private rail car fleet. · Though the Bank's affiliated venture capital arm, the IFC, a stand by equity investment commitment would be offered to qualified investors in the rail car leasing company subject to the IFC's final prudential review. · The Bank would provide resources for the engineering re-design and conversion of rail car hulks into rail cars suitable for handling specialized commodities. These designs would be developed to support the particular logistics and goods movement needs of local shippers/ consignees. They would also be designed to enable and facilitate the development of new and more efficient kinds of railway services. The project would move forward in parallel with the pending rail concession offer and would complement that offer in critical and valuable ways. ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ 5.5 Financing Supply Chain Development The need for working capital is significant in all three of the supply chains, which the team studied. Working capital in the form o both of inventory (including raw material, in process and finished goods inventory) and, to a lesser extent, accounts receivable is tied up in each of the chains. All three commodities/ products manifest high levels of working capital requirements on account of their seasonal production/ consumption imbalances. Moreover, all three have primary production/ consumption cycles which exceed 15 days from first process to final sale. The net result of these two factors (inventory accumulation and production/ distribution/ sale processing time) is a level of working capital per unit of final sale for the entire end-to-end supply chain which exceeds 100 days and which requires financing from outside the chain itself.1 For upstream participants in these chains no external financing is available. Liquidity constraints force farmers ( and to a lesser extent fishermen who sell into domestic markets) to affect a rapid transfer of product ownership. As we noted above, profit margins are sacrificed when primary producers are forced to sell. Liquidity constraints on the part of primary traders force them, as well, to sell to secondary traders and so on. In essence, trade operates in lieu of external financing to finance traditional chains. A series of buyer/ resellers provide working capital to the chains in which they participate and remove rents in the form of gross margins.2 Liquidity is more a problem for maize farmers than it is for sugar farmers or fishermen. Chain integrators enjoy superior capital market access and producers in integrated chains who operate just one transaction away from primary processors with whom they are affiliated may be able to qualify for third party financing themselves. Moreover, the integrators are prepared under specific conditions to extend trade credits themselves to primary producers who supply their inputs. In Tanzania's more traditional supply chains, however, in chains where social capital has not developed to the point where trade credits are available, a lack of third party financing manifests itself in multiple transfers of ownership between primary producers and ultimate consumers. In other words the limitation of external financing causes transaction costs to increase within traditional chains . Thus, financing from financial institutions is simply not available for small scale participants in Tanzania's traditional chains. Its absence severely limits the ability of farmers and fishermen to produce "for inventory." Instead they produce "for immediate sale." 1Working capital requirements for each of the three supply chains is much greater than 60 days. However, the study team could not develop precise data on averages for the three commodities. However, sugar and maize are since products whose production over a 3-4 month period must satisfy local demand over a 12 month period. Hence, in addition to a production cycle of 15-30 days, inventories of 8-9 months must be carried within the system. Hence the working capital requirements for these two commodities Hence, 9 months of working capital requires financing in these two chains. The fresh fish supply chain turns over more rapidly, but the frozen fish supply chain turns over more slowly. Production and shipment are decoupled in the frozen fish supply chain. The working capital requirements for this chain are estimated roughly to exceed 100 days. 2Value adding chain participants pay 100% or more in market margins rather than 30% or less in interest payments to financial institutions. Their supply chains are designed accordingly. Importantly, as well, a lack of third party financing limits the ability of primary producers in traditional chains to invest in productivity enhancing technologies and in superior production systems, e.g. most maize farming is done with minimum farm inputs and production systems which are suboptimal. Indeed, as we discussed above, the only sources available for funding capital requirements in the cases both of sugar farmers and fishermen are funds advanced by supply chain partners. Discussions which the study team held with the Banker's Association of Tanzania confirmed that very little financial sector intermediation takes place vis a vis any of the three chains. For small holder farmers and for fishermen who lack affiliation with export processors external sources of working capital are extremely limited and investment capital is non existent. The primary problem which inhibits third party financing is the lack of bankable collateral. Asset backed lending to small scale producers does not occur in Tanzania for two reasons: i) because systems for validating and subsequently perfecting contingent claims, which are activated by loan defaults against land, chattel and inventory, are missing; ii) because the cost of managing small credits to small scale farmers exceeds the returns which might reasonably be expected. The lending policies of different commercial lending institutions wary widely in Tanzania which since the successful privatization of it banking sector has evolved in the direction of a diversity of bank business models. 3 With respect to the "collateral" issue: Tanzania's system of property laws currently fails to provide sufficient security for bankers to justify the risks and high administrative costs associated with providing loans to farmers and fishermen. Thus, for example, no laws exist which allow for the creation of security interests in farm product inventories. In those few instances in which working capital loans or export loans are made against booked sales physical security is taken in lieu of legal security. Companies like SGS or Crown Agents are hired by commercial banks to physically control inventories of products being exporter, to identify themselves as the beneficial owners in all of the trade and export documentation and to receive the receipts of the export sale before factoring out interest and principal costs and passing the residual compensation back to exporters. In this way a fledgling trade financing segment is emerging to complete against off shore financing. However, this same set of arrangements does not work for domestic market sales nor for any other form of security interest in inventory. In fact, the development of a collateral interest even in chattel is problematic in Tanzania. Similar problems exist with respect to creating collateral interests in real property. A new Land Act was passed in 1999. The Act provides for the creation and transfer of 99 year leasehold rights in real property. It does not, however, recognize freehold land tenure. Moreover, the Act restricts the ability of mortgagees to foreclose on the properties for which they hold a mortgage claim. Significantly as well, the Land Act fails to recognize that land has value in its own right, separate from the improvements which have been made to it. Hence the Act makes it extremely difficult for commercial lenders to form a basis for mortgaging or for valuing real property which will satisfy their own regulators. 3As the recently elected president of the Tanzania Bankers Association put it: " The risk appetite of the ( banking) institution determines where it will invest and not any policy of government or any across the board regulatory guideline." The Banker's Association has discussed these issues with the Governor of the Bank of Tanzania , the Minister of Lands and the Minister of State with the intention of influencing the regulations which emerged pursuant to the Act itself. However, to date they have been only partially successful in organizing a system of property rights which would allow for the creation of bankable collaterals. The type of lending that financial institutions are prepared to undertake principally is cash flow lending and then only in circumstances where supply chain structures have begun to mature and where cash flows flow between chain partners with some predictable regularity. Thus, for example, bankers are prepared to provide loans for farm inputs against sales receipts to sisal growers some of whom, significantly, are also maize inter-croppers.. These loans are guaranteed against proceeds from the sale of sisal to the Sisal Authority. The banks in question administer the flow of funds between the principals and deduct their due payments from the resulting cash flow. A similar arrangement is has worked for some time with tea out growers who have a similar symbiotic chain relationship with a nuclear farm operator. All of these are ad hoc arrangements. Similar arrangements are under consideration with respect to financing outgrower investment in the sugar chain. When making a "loan/ no loan decision", the more risk tolerant of Tanzania's commercial banks seem to apply the following set of criteria to specific chains: · Commercial relationships between primary processors and processors/ buyers are stable and of sufficiently long standing to assure predictability in transaction cash flow between the first and second parties; · Control over physical inventory can be made secure through third party validation and control; · Prices and values for inventories can be readily determined in a secondary market; Again, the solution to developing` bankable' supply chains appears to be found in industrial organization. Bank's are prepared to finance supply chains in which collective action among primary producers is organized by or in response to the input needs of better established merchandisers/ processors. The implications for supply chain development are quite clear and include the following: · Collective action at the farm level ( e.g. farmers organizations( FO's) which are prepared to assume joint liability or to organize in industrial structures which pool risk and liability) particularly in directions which strengthen forward market linkages can improve the credit worthiness of the group if the regular cash flows which result can be made transparent, collectable and accessible to lenders. Legal frameworks can begin to developed around existing jerry rigged cash flow based financing arrangements which ultimately translate into a securities law and a formal framework for security based lending. · Additional development work needs to be done in the are of collateral law. Again, working from " concrete" successes to a more generalized framework may well be the most effective strategy here. · Leasing appears to hold out a great deal of potential as a financing instrument for injecting productivity enhancing assets into supply chains. In the arena of specialized equipment leasing, joint ventures between commercial banks and supply chain integrators might be worth pursuing. 5.6 The Way Forward In order to assure a sustainable commitment from Tanzania's private sector, opportunities to pursue some of the initiatives described in the sections above can best be realized by pursuing one or more of three potential development paths: i) preparing and executing transactions which link together the commercial interests of multiple parties along an integrated supply chain; ii) investment in the strengthening and further development of already existing institutions. Starting up new institutions is highly risky; iii) investment in demonstration projects which prove the commercial merit of innovative business models or the utility of new modes of supply chain integration. Importantly as well, these three development paths mark out the most effective application of donor assistance. The way forward in supply chain development in Tanzania entails only three approaches: developing and offering transactions for private investment, ii) affecting institutional reforms and iii) developing demonstration projects which are subsequently followed up with programs designed to roll out successful demonstrations. The objective underlying all of these strategies is induce private sector commitment by marginally reducing the risks of investment to levels where sustainable commitments from the private sector are forthcoming. The critical process of " co-investment" subjects each initiative which is underwritten in turn to a market test. If some minimum acceptable level of private sector co-investment in sustaining a particular initiative is not forthcoming, the merits of that particular initiative fall below the threshold of the market test. Moreover, competitive, open and contestable tendering of co-investment opportunities assures their allocation to private partners who are best positioned to realize their full value. As the decision tree diagram below suggests a number of alternatives can be productively considered for development of each of the three paths. Thus, for example, different transactions can productively be prepared for investment in supply chain assets which have the characteristics alternatively of "private goods", "public goods" or "community goods." Community goods are capital goods whose investment generates large positive externalities for an entire community of trading partners or supply chain partners but which remain underinvested because inadequate organization exists to capture and channel incremental benefits to investors.. The simple act of organizing a community of beneficiaries around this allocation issue, preventing free riders and committing the entire community to support an investment through an exclusive use agreement and though the community adoption of a user fee schedule should be sufficient for inducing private investment. In each of the transaction preparation options outlines in the schematic, donor resources could productively be used to define risks, suggest transaction modes and organizational means for managing both benefit flows and for assigning costs and, most importantly, for determining the financial merits of specific proposed private sector investments. Donors would invest in a transaction development process, in a due diligence reviews, in complementary business plans and legal documentation. In this way opportunities could be articulated, tested and implemented to auction parts of a " the supply chain development market" to qualified investors who might not otherwise be aware that a market opportunity existed. This proactive strategy involves essentially the outsourcing of missing supply chain functions under the terms of a franchise agreement, a concession or a joint marketing agreement. The second general development path entails investment in soft assets--in social capital, in organizations, in systems, and in industrial organization. The three alternatives identified in the second branch of the decision tree certainly do not exhaust the entire domain of potential investment objects but they do include the most prominent ones: markets, mezzo level organizations and specialized extensions of producer organizations though which specialized support services can be offered. As we discussed above several opportunities exist to reinvent and reengineer market institutions so that they operate in modes which are more pro-poor and pro-growth. No need exists to repeat what has already been discussed. The concept of " mezzo level organizations," however, may require some elaboration. Mezzo level organizations do the work of supply chain development. They fall half way between the public and private sectors. Hence their name. Mezzo level organizations concern themselves with investment in community , with the adoption of best practice techniques and technologies and with the reduction of transactions costs within chains. Mezzo level organizations define their membership over the entire set of chain participants and sometimes include providers of specialized support services as well. The final category of institutions which are worth developing are specialized service providers who emerge to fill missing service links in otherwise competitive chains, or who transform value subtracting links in chains into value adding links. In may that we examined failures within markets for specialized services on which chain development depends have caused chain development to be retarded or the realization of competitive advantage to be frustrated. In such circumstances, chains need to develop their own service solutions, to in- source the delivery of services, instead of outsourcing this delivery. Farmers organizations ( FO) provide an excellent base from which to develop specialized in sourcing delivery systems for specialized IT, logistics and cold chain management services. The final path involves developing and testing new farm and fisherman community based business models and new institutions. The operative word here is experimentation. Demonstration projects are the instrument of preference. As we noted above, both Uganda and Kenya have made substantially more progress than Tanzania in developing new commercial business models and in realigning the functions of government to support their further roll out. Several NGO's who have pioneered in the development of Farmer Organizations in all three countries have made more progress outside of Tanzania than inside. Primarily this is because legacy institutions which are left over from the pre-reform era still dominate key chain management functions and tend to challenge or to crowd out new private sector initiatives and NGO sponsored demonstrations. Secondarily, proof of commercial viability has been slow to test and prove because of a residual bias against private sector driven development and an ambivalence in the messages send on this account by government itself. Moreover, the legal framework on which non-cooperative farm and fishermen community based businesses can be based is more limited in Tanzania, than in the other two countries. Uganda for example allows pubic companies to be formed based on liability limited by the guarantee of the founders without shared capital. No comparable limited liability legal structure exists in Tanzania. With this said, very attractive opportunities to migrate successful demonstration projects which have started up across Tanzania's borders, to transplant these business models and the industrial-agricultural structures which appear to work best and to transplant them, refine and develop them further in a Tanzanian context. In this context resources might be made available by donors for assessments of best organizational design, for enterprise planning, for investment risk assessment and for proof of business concept. For its part the Government of Tanzania would provide a laboratory for business model experimentation, sufficient pre-approvals and licenses for non traditional business models to allow them to start up and further commitments to adapt and to adjust the activities of government agencies once demonstration project results become available to duplicating and scaling up successful experiments. In this way a thorough going, bottom up reform of farm and fishermen to market linkages could be implemented over the course of a decade. . Intervention Alternatives Public Goods Intervene 1Prepare and Execute Transactions for the To Develop Private Sector Community Goods Pro Poor Supply Strengthen and Re- Private Goods Chains 2 engineer Existing Markets Institutions Mezzo Level Organizations Specializes Services Demonstration Project 3 Investment: New Planning Business Models and New Institutional Risk Assessment Structures Proof of Business Concept The World Bank Study on Growth and Environment Links for Preparation of Country Economic Memorandum (CEM) Part 1: A Review of Relevant Literature and the CEM Concept Note Final report May 2005 Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 The World Bank Study on Growth and Environment Links for Preparation of Country Economic Memorandum (CEM) Part 1: A Review of Relevant Literature and the CEM Concept Note Final report May 2005 Report no. 1 Issue no. 1 Date of issue 18 May 2005 Prepared KEP Checked TNH Approved Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 Table of Contents 1. Introduction 1 2. Review of Relevant Literature 3 2.1. Natural Resources & Growth: The First Debate 4 2.2. Growth & Natural Resources: The Second Debate 8 2.3. Poverty/Environment Nexus: The Third Debate 10 2.4. Summary 12 3. Review of CEM Concept Note 14 3.1. Analysis of Theoretical Framework 14 3.2 General Comments 16 3.3 Proposed Revised Structure of the CEM 19 3.4. Specific Edits and Inserts 20 Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 1 1. Introduction A study was conducted to analyze the linkages between environment and growth in Tanzania with the objective to feed into the on-going preparation of the Country Economic Memorandum (CEM). The specific objective of the study was to bring together data to clarify the contribution environmental and natural resources (ENR) have to GDP and to identify the potential for increased growth from the sustainable use of ENR, and constraints to the achievement of growth. The Terms of References of this assignment (attached in Annex 1) comprised the following specific tasks: (a) To undertake a critical review of the CEM concept note to identify where issues of growth and environment linkage are required. (b) To review documentation on links between growth and poverty, and the environment provided by the World Bank and other Development Partners. This should include the study on poverty and environment links commis- sioned by the World Bank, the PER of environment, and the MKUKUTA. (c) To collate and review additional documentation and data on links between growth and poverty and the environment. (d) To prepare written comments on drafts of the CEM. (e) To prepare brief inserts and edits for the CEM relating to growth- environment linkages, this should include relevant data and tables where this data exists. The results of the study are presented in three parts: Part 1 summarizes a review of relevant literature on environment and growth as well as comments on the CEM concept note. The findings of the data collection exercise are presented in two separate documents as Part 2 and Part 3 of the study. Part 2 focuses on Forestry, Wildlife and Marine Fisheries, which seem to have been forgotten in the discussion on economic growth in the context of the CEM. The title of the report is hence accordingly " Forgotten Growth potential - Forestry and Wildlife". Part 3 presents background data on the Mining, Freshwater Fisheries and Tour- ism sector with a view to illustrate the externalities of the 'success stories' of growth in the CEM. Thus, the presentation of results has been aligned with two Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 main comments on the CEM: Firstly, the omission of the discussion of forestry and wildlife as potential growth sectors. Secondly, the omission of considera- tion of environmental and social externalities created by the high growth in Fisheries, Mining and Tourism. The reason why the Fisheries Sector is sub-divided into Freshwater and Marine Fisheries, is that their 'stories' in the context of economic growth are very dif- ferent. While high growth rates in the Freshwater Fisheries are due to over- exploitation of the resource, in particular Lake Victoria, most likely exceeding sustainable limits; Marine Fisheries still provide a large un-captured potential for economic growth. Furthermore, the governance regimes for the two Fisher- ies sub-sectors are quite different in the sense the Marine Resources are regu- lated by international conventions in addition to national legislation. The study relies entirely on the review of secondary data and existing literature. Consulted sources of information include official Government statistics and reports (respective sector ministries, Bank of Tanzania and National Bureau of Statistics, Tanzania Investment Centre), sector studies commissioned by re- search agencies and donors, economic analyses from independent agencies, case studies, as well as a few current news stories from the media. The exam- ined literature is listed in the Annex of each of the three reports. The desk study was conducted by COWI Tanzania in March/April 2005 through Ms. Kerstin Pfliegner with contributions from Dr. Kassim Kulindwa and Mr. Thomas Hansen. This report presents Part 1 of the study and focuses on a general literature re- view on "Growth and Environment" in Chapter 2 and comments on the draft concept note of the CEM (Version September 13, 2004) in Chapter 3. A revised outline of the CEM, encompassing ENR, as well as inserts and edits are also proposed. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 2. Review of Relevant Literature There is a wide body of literature on the complex links between growth, pov- erty, natural resources and the environment, which can be divided into three main debates: The first debate, examines the macro-level implications of natural resources for economic growth. The point of departure for the debate is the so-called re- source curse hypothesis, arguing that countries with abundance of natural re- sources are, ceteris paribus, determined to have lower growth rates as the more productive manufacturing sector is crowded out. The second debate focuses on the inverse causal relationship, i.e. the macro- level implications of growth for natural resources and the environment. The Kuznet's environmental curve hypothesis is at the centre of the debate. It argues that pollution and environmental degradation will increase with GDP growth until a certain point after which sophisticated technologies and demands for better environmental regulation will help reduce emissions and environmental damage. This curve-like relationship is however widely contested. Finally, the third debate, examines the implications of natural resources for poverty and vice-versa. One of the prevailing hypotheses is the pov- erty/environment nexus, which argues that a degrading environment will cause poverty, which in turn damages the environment further, thus creating a vicious circle. The three debates are closely related. First of all, a general association is as- sumed to exist between growth and poverty reduction, and so the main differ- ence between the first and second debate on the one hand, and the third debate on the other hand is the level of focus. Whereas the two first debates are mainly focused at macro-level, the third debate takes a closer look at the micro-level effects and implications for the poor. The model below presents the links to be examined at the macro level (the first and second debate). The model introduces the two main variables, natural re- sources (NR) and economic growth, and the expected logic of causality be- tween them. The model also introduces influencing factors (also referred to as `channels'), which may impact, positively or negatively, on the relationship between the two main variables. Whereas many scholars agree about the overall association between the key variables, there is substantial debate about the na- ture of these influencing factors. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 Influencing factors - the first debate Natural Re- Economic sources Growth Influencing factors - the second debate All of the debates are relevant to understanding the challenges faced by Tanza- nia. The first debate may, for example, help illustrate how the observed high growth in the mineral sector may impact on the overall GDP growth rate. The second debate illustrates the environmental challenges that Tanzania will face in the coming decades if current growth rates are maintained. Finally, building on the National Strategy for Growth and Reduction of Poverty (NSGRP), the third debate may help illustrate the potential of NR for those still trapped in poverty throughout rural Tanzania. This chapter provides a brief review of this literature with a view to identify the main theoretical trends relevant to the case of Tanzania. The review will mainly focus on recent (1995 and after) academic literature (journals, books and working papers) and a few publications from leading de- velopment institutions. The review will provide a general overview of the pre- vailing ideas, but is not meant to be exhaustive. This Chapter is divided into four main sections. Section 2.1. reviews literature analysing natural resources as independent variable, and economic growth as dependent variable (The First Debate). Section 2.2. looks at literature taking growth as point of departure with a view to examine implications for natural resources and the environment (The Second Debate). The Poverty/Environment nexus will be discussed in Section 2.3. Finally, Section 2.4. will summarise the review. The reviewed literature is listed in Annex 2. 2.1. Natural Resources & Growth: The First Debate This section will present and discuss the body of literature dealing with natural resources as independent variable. The point of departure for most of this litera- ture is the so-called resource curse hypothesis, which argues that natural re- source endowments and economic growth are negatively associated. Various empirical work supports this hypothesis, but there is a substantial on- going debate as to the reasons (`channels') behind this association. Some schol- Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 ars (albeit a minority) suggest that the association is merely spurious and thus representative of underlying factors and/or mal-specifications. The Resource Curse and Dutch Disease In their seminal article on "Natural Resource Abundance and Economic Growth", Sachs and Warner (1995) argue that natural resource endowments negatively impact on growth performance. In their cross-country analysis, they define NR endowments as the ratio of primary-products to GDP in 1970 and find that a high NR endowment is negatively correlated with growth in the fol- lowing 20 years. They find this association to be statistically significant even after controlling for spurious effects and a number of variables commonly asso- ciated with variations in cross-country economic growth. To substantiate this negative relationship, they summarise a number of prevailing hypotheses: · Social: Abundance of natural resources is perceived by local producers and authorities as `easy riches' and sloth will therefore easily spread in the economy. This social argument was initially advanced by Bodin (1576 in Sachs and Warner, 1995) but has also been proposed more recently by Holmes (1995 in Cervellati and Fortunato, 2004). · Primary exports: This hypothesis is based on theories originally formulated in the 1950s by Prebisch and Singer arguing that world market prices for primary exports in the long term tend to decline relative to manufacture prices. · Rent-seeking: A political economy explanation advanced by Lane and Tor- nell (1995). The argument is that abundance of NR will promote rent- seeking behaviour in the economy. Hence, (too many) competing agents are expected to fight for NR rents. This is, from a society point of view, viewed as an inefficient use of resources. · Domestic linkages: A theory originally formulated by Hirschmann (1958 in Sachs and Warner, 1995), Seers (1964 in Sachs and Warner, 1995), and Baldwin (1966 in Sachs and Warner, 1995). They argue that the primary re- sources sector has relatively few "forward and backward linkages" to the economy compared to the manufacturing sector where a much more ad- vanced division of labour is observed. The argument has been further re- fined by Sweder van Wijnbergen (1984 in Sachs and Warner, 1995) and Matsuyama (1992 in Sachs and Warner, 1995). Matsuyama, who distin- guishes between two sectors, agriculture and manufacturing, argue that learning is higher in the manufacturing sector and learning-induced growth will therefore be relatively lower in an NR-based economy. This is also known as endogenous growth theory. · The Dutch disease hypothesis is a variant of the linkages approach, which sees the economy as divided into three sectors: A tradable NR sector; a tradable non-NR manufacturing sector; and a non-trade sector. There are two parts to the hypothesis: a. Abundance of NR leads to overvaluation of the national currency, which in turn implies deterioration in the real exchange rate with nega- tive implications for the manufacturing sector's competitiveness. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 b. High NR endowments, it is argued, will attract a high proportion of capital and labour thus squeezing out the manufacturing sector. Hence, the growth observed in the NR sector per se will be more than off-set by losses in the manufacturing sector, which, as argued by the endoge- nous growth theory, is assumed to be associated with higher growth rates than the NR sector. To substantiate the negative association between NR and growth, Sachs and Warner perform regression analysis and find that NR not only works directly, but also through (lack of) openness of the economy: Hence they argue (as in the Dutch disease hypothesis) that a high NR endowment squeezes the manufactur- ing sector, which in turn leads policymakers to protect the domestic manufac- turing sector from more competitive foreign manufacturers. This in turn is be- lieved to translate into reduced efficiency and overall lower growth. Gylfason, Herbertson and Zoega also argue in favour of the Dutch disease hypothesis in their 1997 article: A Mixed Blessing: Natural Resources and Eco- nomic Growth. On the basis of a cross-country analysis of 125 countries for the period 1960-1992, they find, like Sachs and Warner (1995), a negative associa- tion between the size of the primary sector and economic growth. With regard to the influencing factors, they point to the importance of the exchange rate ap- preciation, which has negative implications for the development of what is pre- sumed to be a skill-intensive manufacturing sector. These arguments are further elaborated and applied to transition economies in Gylfason (2000). Gylfason and Zoega (2001) add further detail by suggesting that the proposed inverse relationship between NR abundance and growth is related to intermedi- ate savings and investment effects: They argue that abundance of NR a) crowds out physical capital and b) slows down the development of a financial system, thus inhibiting growth in the non-NR sectors. They find support for these hy- potheses in their cross-country analysis of 85 countries for the period 1965- 1998. Papyrakis & Gerlagh (2004a) like the above set out to explain the nature of the perceived negative association between NR abundance and economic growth. They argue in their cross-country study covering the 1975-96 period that the negative association is established through a so-called savings- investment channel. More precisely they argue on the basis of their empirical work that NR abundant countries tend see investments in the manufacturing sector decline, which in turn has long-term negative effects for overall growth and productivity ­ as endogenous growth theory would predict. Papyrakis & Gerlagh (2004b) add further sophistication to their arguments in their article: "Natural Resources, Innovation, and Growth". Here they suggest that innovation may be another important transmission channel, which helps explain the negative NR-growth association. The arguments are related to their 2004a work, but in 2004b they emphasise that an abundant NR sector tends to crowd out research and development activities - which has negative implica- tions for the wider economy, notably the manufacturing sector. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 In summary, it is clear that a significant body of literature exists to support the resource curse hypothesis. They all share the view that abundance of NR has a negative impact on economic growth. They also appear to largely agree that the manufacturing sector, ceteris paribus, holds more growth potential for the economy (endogenous growth theory), and the negative growth effect associ- ated with NR is typically explained as a crowding out effect, with the manufac- turing being squeezed out. The scholars differ however on the exact nature of this crowding out effect. Some point to the typical Dutch disease effects (cur- rency appreciation and factor reallocation), whereas others point more specifi- cally to investment and innovation channels. The following section will present a few scholars who disagree more funda- mentally with the resource curse thesis. The Resource Curse Questioned Stijns (2001) questions two fundamental aspects of the resource curse, notably as interpreted by Sachs & Warner. First of all he questions the validity of their working definition of NR, which, as mentioned, is the proportion of primary sector exports to overall GDP. Stijns argues in this context that the "claim that being a resource export dependent country slows down its expected rate of growth, is a different claim than arguing high mineral reserves or production of those reserves is associated with slower rates of growth" (Stijns, 2001: 9). Ac- cordingly, Stijns does not question the Sachs and Warner (1995) findings per se, but he argues that their operationalisation does not capture NR abundance. In consequence, he chooses to look at NR reserves and does so by sub sector, i.e. land, oil & gas, coal, and mineral reserves. The findings are different from Sachs and Warner (1995) in the sense that only land abundance seems to influ- ence GDP growth negatively, whereas the results are somewhat mixed for the other NR-categories. Stijn therefore also questions the broader validity and ro- bustness of the resource curse hypothesis. He limits his own work to conclud- ing that at least fuel and mineral reserves have not been a significant determi- nant for GDP growth between 1970-89 (2001: 35). In terms of interpretation of influencing factors, Stijns does not find any hard evidence, but he speculates that land abundance may impact growth negatively by locking countries into an agricultural trade export model, which in turn pre- vents them from developing their manufacturing sector. Again, this argument is similar to the earlier mentioned crowding-out theories. For the other NR categories, which as mentioned show mixed results, Stijns concludes that what matters is not endowment of NR, but rather what countries do with their resources. In this context he speculates about the importance of R&D and patent legislation. Boschini, Petterson and Roine (2003) work along the same lines as Stijns, suggesting that NR must be broken down into sub-areas to properly understand the dynamics behind natural resource abundance, notably their effects on eco- nomic growth. To do so they group NR along a continuum measuring `techni- Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 cal appropriability' which in turn captures the level of ease required to quickly realize large economic gains from exploiting the NR. Their hypothesis is that the kinds of NR, which can be translated into quick profits, will be susceptible to moral hazard, such as corrupt behaviour. This would typically be diamonds or precious metals. At the same time, they hypothesise that the probability of moral hazard behaviour would be significantly reduced in countries featuring a proper institutional framework. In their cross-country analysis (80 countries, not including Tanzania) they find support for these hypotheses and thus conclude that the resource curse only ap- plies under certain conditions: Countries, which feature NR of high technical appropriability are particularly susceptible to corrupt behaviour, especially so if the quality of the institutional framework is low. They point to Sierra Leone as a case in point. Hence, the analyses of Stijns and Boschini, Petterson & Roine both point to the importance of institutions as an intermediate influencing fac- tor. In case of a good institutional framework, the resource curse hypothesis, they argue, breaks down. Whereas the above scholars focus on institutional-political variables as inter- mediate influencing factors, Bravo-Ortega & Gregorio (2002) introduces hu- man capital as an important factor in the NR-growth relationship. The argument is that countries with high levels of human capital have been better positioned to counter the negative effects, which they otherwise associate with NR abun- dance. Hence, Bravo-Ortega & Gregorio are therefore more in line with the original resource curse thesis than the other scholars introduced in this section. The scope of their study is limited to a mainly Latin American and Scandina- vian context. 2.2. Growth & Natural Resources: The Second Debate The literature reviewed in this section turns the above-discussed causality up- side down in the sense that it focuses on growth and how it impacts upon natu- ral resources and the environment. Again the debate is mainly limited to the macro level and the key issue is the degree to which (if at all) economic growth impacts upon the NR. NR is in this context understood in a broader sense than in the previous case and includes common goods such as land, water and air. The relationship between growth and natural resources has been at the forefront of policy discussions since the early 1970s with the publication of the 1972 Meadows et al. report on limits to growth (Meadows et al. in Martinussen, 1999). This document expressed concern over the wider implications of eco- nomic growth with regard to pollution and environmental degradation. The pro- file of the debate was further raised with the 1987 Brundtland Report and the 1992 UNCED summit in Rio, which introduced the sustainable development discourse into the main policy debate. The policy debate has been informed by a rapidly growing body of scholarly literature trying to assess in detail the pos- sible environmental consequences of economic growth. The main trends of the debate are described below. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 The Environmental Kuznet's Curve Hypothesis The Environmental Kuznet's Curve hypothesis (EKC-hypothesis) builds on a contested theory about the long-term association between economic growth and environmental pollution. Pollution is in this context measured by the level of per capita emissions (typically sulphur and carbon dioxide). Kuznets Environmental Curve sn emissio itap ca Per Income The hypothesis basically argues that as economies start developing, based on increasing scales of economy and a changing composition from agricultural production towards industrial pollution, the level of pollution increases. This relationship is however assumed to break down at a certain point after which rising income will be negatively associated with per capita emissions. The as- sumption is that rising incomes eventually will give rise to demands for a better environment at the same time as a more advanced economy will make the de- velopment of more environmentally friendly technologies possible. See the graph above for a stylised and very basic presentation of the argument. The EKC-hypothesis has been advanced by several scholars, notably Grossman & Krueger (1991 & 1995). Lately it has also been championed by Lomborg (2001 in Smith, 2004). As noted by Verbeke and De Clerq (2002) the theory remains somewhat underspecified. Scholars struggle for example to identify the income point at which the positive association between growth and emissions is expected to break down. Verbeke and De Clerq (2002) submit the hypothesis to an empirical test and find general, albeit only indicative, support for the fact that emissions start decreasing at a certain income point. They hesitate to ex- trapolate these results and thus refrain from making general observations about the validity of the EKC-hypothesis. As pointed out by Smith, the notion of environment and natural resources em- bedded in the definition is very narrow and does not consider many other envi- ronmental aspects. The next section will present two scholars who take a more comprehensive look at the environmental aspects of economic growth. Towards sustainable development For Vosti & Reardon (1997) the focus is specifically on agricultural growth and the key question is how fast growth can occur without degrading the natu- ral resource base? They suggest in this context intensification, i.e. the use of more and better-quality inputs on the same amount of agricultural land. Oppor- tunities for doing so vary between agro-ecological zones, defined according to Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 their soils, rainfall, ground- and surface water, and sunshine. In environmen- tally fragile, less favoured agro-ecological zones it is difficult to intensify agri- culture. Food grain production will be more costly, suggesting that one should focus on livestock, agroforestry, and other staple production systems. In more favoured zones agricultural intensification is essential to meeting sustainability and poverty alleviation objectives. They suggest that national policies should not only promote zone-specific opportunities but also address the food security and inequity issues that can emerge as a result of these strategies. Like Vosti & Reardon, Pearce (2004) points to the need for balancing growth and environmental concerns. He argues that the goal for poor nations is for them to use resources more efficiently. Hence, the solution is to continue pur- suit of growth while having as much regard as possible for conserving what is left of the natural resources and environment. According to Pearce, this requires a proper policymaking and institutional framework, notably clear legislation and enforcement of legislation. He illustrates in this context how advanced countries have managed to decouple environmental impacts from economic growth on the basis of new environmental technologies - as predicted by the KEC-hypothesis. The Policy Discourse in Tanzania Tanzania's National Strategy for Growth and Reduction of Poverty (NSGRP) builds on the basic principles of the Brundtland Report noting that unsustainable use of NR is widespread in Tanzania (GoT, 2005: 6-7). Accord- ing to the text, examples include reckless tree-felling and soil erosion as a result of bad farming-methods. Korongo Ltd. (2003: 20) illustrates another example from Tanzania, the Lake Victoria Basin, where over-exploitation of fisheries resources through the ap- plication of allegedly environmentally damaging techniques has led to a serious environmental degradation of the Lake and its catchments. The question is how a country like Tanzania, which presumably does not have access to the same technologies as many advanced OECD countries, can in- crease growth while at the same time reduces negative impacts on the environ- ment. 2.3. Poverty/Environment Nexus: The Third Debate The two above sections have mainly considered economic growth from a macro-perspective. This section will demonstrate that there are several authors arguing that poverty and natural resources should be analysed separately (Das- gupta et al. 2003). They examine for example the direct implications of envi- ronmental changes for the poor or whether NR as a source of income holds par- ticular potential for the poorest groups. In line with these ideas, the NSGRP notes that Tanzania's NR-base holds a big potential for raising and sustaining rural incomes (GoT, 2005: 9), while also noting that poor people generally rely heavily on NR and are thus more suscep- Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 tible than other income groups to external shocks ­ such as weather extremes, deforestation etc (GoT, 2005: 46). The recent Korongo Ltd. (2003) survey of the links between environment and poverty in Tanzania likewise concludes that it is mainly through profitable and sustainable use of NR, that Tanzania can achieve significant poverty reduction. The below section will examine in more detail some of the prevailing hypothe- ses about the relationship between environment/natural resources and poverty. Sustainability, Growth and Poverty Reduction Vosti & Reardon, in their seminal book Sustainability, Growth and Poverty Alleviation ­ A policy and Agroecological Perspective (eds., 1997) set the overall framework for the debate: Policy makers are faced with the need to pur- sue three challenging goals simultaneously: Growth, poverty reduction and en- vironmental sustainability. Poverty alleviation, they argue, is essential as it un- dermines development, growth and the environment. At the same time, natural resources must be made sustainable as an input into sustained growth. The three goals are complementary: sustaining NR will help growth. Growth will help reduce poverty and improve environmental management. However, in the short-term there will be trade-offs among the three goals. They argue for exam- ple that poverty is unavoidable if the NR base is degraded. Still, as they point out, trade-offs are at play: Conservation of NR, in the sake of preservation, may hurt farmers living from the NR. The poverty/environment nexus, to be dis- cussed below, is part of this overall debate as it sheds light on the connection between poverty reduction and environmental sustainability. Poverty/Environment Nexus The point of departure for the poverty/environment nexus debate, as already outlined by Vosti and Reardon, is the assumption that the livelihoods of the poor depend crucially on access to NR. DFID (2002) outlines for example how agriculture is without comparison the major source of income for poor people in the developing world. Thus the study mentions that for many African countries the prospects for growth lies in agriculture, tourism and mining. The main concept behind the debate is the no- tion that "one problem is a significant determinant of the other": Hence, to the degree that the nature and scope of farming activities of the poor are environ- mentally unsustainable, the long-term resource base for their livelihoods will be eroded (2004: 8). The poor are therefore, more than other income groups, as- sumed to be utterly dependent on a sustainable NR-base, but they may them- selves be the prime (direct) reason why the base is depleted. The hypothesis has however proven difficult to test, notably, as pointed out by Dasgupta et al. (2003), because of the lack of valid and reliable data. Cavendish (1999) uses data from Zimbabwe to show how environmental re- sources have made a significant contribution to average rural incomes. He also argues that there is significant differentiation in the economic properties of dif- Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 ferent types of NR. According to Dasgupta et al., 2003 similar studies (with similar findings) have been carried out by Ambler (1999 in Dasgupta et al., 2003), Kepe (1999 in Dasgupta et al., 2003), and Reddy & Chakravarty (1999 in Dasgupta et al., 2003). Brocklesby and Hinshelwood, (2001 in Dasgupta et al., 2003) have conducted a broader study to show how the poor at least perceive the environment and NR in general to be an important determinant for their general welfare and liveli- hood. Dasgupta et al. (2003) submit the nexus to an empirical test in relation to the prevalence of five so-called principal environmental problems, viz. deforesta- tion, fragile soils, indoor air, pollution, unsafe water, and sanitation. The study, which is limited to Southeast Asia (Lao and Cambodia), finds mixed evidence: No clear association can be detected in the case of Cambodia whereas the ef- fects is more pronounced in the case of Lao. Thus one of their main conclusions is that the validity of the nexus varies significantly between countries. In the case of Tanzania, the NSGRP observes that under-employment has led to unsustainable use of NR (GoT, 2005: 10). Korongo Ltd. (2003: 35) likewise argues that environmental degradation in Tanzania is caused by local poverty and lack of alternative income opportunities. Still they also point to additional factors such as lack of awareness and inadequate tenure etc., and a more rigid analysis would thus be necessary to assess the relevance of the pov- erty/environment nexus for the case of Tanzania. 2.4. Summary This brief literature review has touched upon three major debates within the overall context of growth and environment. The three debates differ by their choice of independent variable and/or level of focus - macro or micro. The first debate, taking NR endowment as the independent variable, discusses the extent to which countries rich in NR are cursed. The evidence is mixed, ow- ing partly to different working definitions of NR and growth as well as differ- ences in choice of method and empirical focus. Most of the scholars appear to agree with the notion that NR abundance can, at least under certain conditions, have a negative impact on GDP growth. There are however may different inter- pretations as to why this negative association exists, with a majority pointing to crowding out effects (of the manufacturing sector). Further empirical studies would be required to test the validity of this hypothe- sis to the case of Tanzania, and the effects arguably differ by NR sector. In the case of the large-scale mining sector for example, the crowding out effect is difficult to identify, since a) the sector employs very few workers due to its capital-intensive nature, and b) the massive capital investments have mainly been financed by foreign investors in the course of a few years and therefore have not crowded out any existing domestic capital sources. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 The question whether Tanzania is faced with a resource curse may therefore be more relevant in relation to the agricultural sector, which employs the majority of the workforce under a relatively inefficient production mode. The second debate, by contrast looks into the long-term environmental effects of economic growth. A key question is the exact scope and validity of the EKC-hypothesis. Applied to Tanzania, which is essentially an agricultural country, the hypothesis would be that the country in the future would be crawl- ing (slowly) upwards the curve as industrial production expands. Of more immediate concern are the more general and broader concerns about sustainable development noted by for example Vosti & Reardon and Pearce. Examples of unsustainable development have for example been registered in the Lake Victoria region, and the NSGRP is therefore right to stress the need for a sustainable development path. There is little systematic evidence available to support the third debate, but anecdotal evidence seems to support the notion that the poor are particularly dependent on NR, and therefore also the most vulnerable group in case of envi- ronmental degradation. This also applies to Tanzania, albeit with significant regional differences: Re- gions in central Tanzania are for example more likely to be hit by external shocks such as lack of rainfall. These regions, at the same time, feature higher than average poverty incidence, which in turn makes them more vulnerable to shocks. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 3. Review of CEM Concept Note The provision of overall comments on the CEM is one of the outputs required under the TOR for this study. Since the CEM was in preparatory stages, at the time of this assignment, the September 13, 2004 draft concept note is the basis of the comments documented here. This chapter provides first in Section 3.1. comments on the theoretical frame- work of the CEM as far as the linkage of growth and ENR is concerned. Sec- tion 3.2. provides general observations on the concept note. Comments pro- vided by other development partners, in particular the Poverty and Environment Advisor of VPO, are included in this section. This is followed in Section 3.3. with a proposed adjusted structure of the CEM to fully incorporate ENR. Lastly, in Section 3.4., specific edits and inserts for the CEM based on the con- cept note are provided. These are backed-up with data, where available, in Parts 2 and 3 of this study, presented in separate documents. 3.1. Analysis of Theoretical Framework Commenting on the CEM concept note by providing edits and inserts will only be cosmetic if one does not look at the underlying theoretical model for eco- nomic growth and ENR employed by the CEM. It is obvious that the CEM concept note does not subscribe to the sustainable development school of thinking, although this would be in line with the NSGRP. The growth model underlying the CEM concept note is based on neo-classical economic theory, where output growth is understood (in a Cobb-Douglas pro- duction function) as 'function' of the product of share and growth of the two input factors labour and capital plus technical progress. Capital is understood in this context as physical capital (infrastructure, investments, savings) only. This growth model does not consider natural capital as an input factor to growth. An indication that the CEM is aligned with this school of thought is that physical and human capital are discussed as drivers of growth acceleration, whereas natural capital is not (p.12 ff). The thinking that economic growth is constrained by environmental limits, is an idea, which was introduced by classical economists as early as in the 18th century (Malthus 1766-1834, Ricardo 1772-1832, Mill 1806-1873), and later developed further through the schools of environmental and ecological eco- Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 nomics. However, ENR have not found their way as an endogenous variable into mainstream economic growth theory. An important difference introduced by environmental economics is the capital theory, which defines sustainable development in terms of the maintenance of the value of the capital stock over time while striving for high growth rates. Capital encompasses natural capital (functions, goods and services provided by the environment) and manufactured, human and institutional capital (ethical, moral capital and cultural capital). Neoclassical theory assumes perfect substitution between physical and natural capital. Hence any limits to growth through reduced availability of raw materi- als can be overcome through technical progress as a source of economic growth. In contrast, ecological economics defends well-defined limits of such substitution and advocates that separate stocks of aggregated natural and other capital must be maintained. There is 'critical natural' capital for which no tech- nological substitutes exist. There is no doubt that such 'critical capital' exists in Tanzania, considering for example its catchment forest and their biodiversity, whose destruction is clearly a binding constraint to be addressed (see also Part 2 of this study). The investigation of the `Dutch disease effect' by the CEM in the context of the gold boom (p. 14f) is an indication that the theoretical direction of the CEM is aligned towards the first of the three main broad debates on ENR and Growth, "the resource curse hypothesis", which was described in Section 2.1 above. This debate defends the view that there is a negative relationship between NR abundance and growth. The CEM concept note directly refers to one of the main hypotheses underlying this school of thought, namely that abundance of NR may lead to increase in the real exchange rate which implies a deterioration of a country's competitiveness" (p. 14) with negative implications for the manufacturing sector. As explained in Section 2.1 above, this theory divides the economy in a trad- able NR sector; a tradable non-NR manufacturing sector; and a non-trade sec- tor. Sharing this line of thinking, it is hence not surprising that non-tradable or non-monetary aspects of the NR sectors are not being considered in the CEM. However, as Chapter 2 above showed, the `resource curse hypothesis' has been questioned and alternative views exist, such as for example the discussion of sustainable development and growth. To that extent the CEM is hence not in line with the approach taken by the NSPRG, which reflects the importance of economic efficiency in natural re- source use, but also stresses that the benefits of development must be distrib- uted equitably, as this is relevant to poverty reduction. ENR are discussed in the CEM in the context of the analysis of Gold, Fisheries and Tourism as the main drivers of economic growth. However, within the above theoretical context, the discussion will have serious omissions. These are related in particular to the following aspects: Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 1. Consideration and efficient control of 'externalities'. These occur when the activities of economic agents have external consequences for other agents other than by affecting prices and these effects are not compensated for. An example in Tanzania for an externality would be the reduced dry season flow of the Great Ruaha River created through abundant irrigation of large commercial rice farms up-stream, which is negatively affecting water users downstream. Three reasons are provided by economic theory for external- ities: Incorrect prices, missing markets or imperfect property rights. 2. Measurement of the benefits and costs of environmental programmes, which requires valuation of natural assets and 'true costing' of economic growth. 3. Natural resource economics, which focuses on the sustainable use of re- newable resources and the optimal depletion of exhaustible resources to de- rive policy measures that provide incentives for their management on a sus- tained-yield basis. There are data constraints in Tanzania for all of those areas (in particular point 2) and fully applying these principles may seem ambitious. But demand for data also 'creates' data and the World Bank can be an important player in this regard. Part 2 and 3 of this study will investigate these points further. In conclusion, it is hence recommended that the CEM apply an environmental economics approach to the analysis of growth. Advocating for the inclusion of ENR by 'environmentalists' and also this study, will have no considerable impact, unless the CEM employs an approach of en- vironmental or even ecological economics to its analysis. This would entail a severe shift in perspective. However, it is the only way in which ENR can be- come an integral part of the analysis of growth in Tanzania. 3.2 General Comments The concept note acknowledges that the majority of GDP depends on the natu- ral resource base, which is 66% if tourism, energy, mining, agriculture are in- cluded. While acceleration and sustainability of growth are primarily discussed to achieve increased levels of growth, the extent to which Tanzania can achieve a path of sustainable development while increasing growth is not considered. This would imply including environmental factors and equity issues and help- ing government to determine policy choices on growth and not just to aim for high rates of growth. As the primary objective of the CEM is to assist the Government of Tanzania to operationalize the NSGRP (p. 2), not addressing these issues would be a serious omission, as the NSGRP seeks to place greater emphasis on the integration of crosscutting issues, among others, the environment (p.2). The sustainable ex- ploitation of these natural resources will be of paramount importance to achieve increased levels of growth in the near and long-term future and to the achieve- ment of sustainable development in Tanzania in general. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 Gold and Fisheries Gold is a non-renewable resource. Hence, in the context of growth, the question arises, how long the current reserves will last, what the overall stock is and if it is exploitable. Similarly, for fisheries, a renewable resource, the question of sustainable yields has not been addressed. Furthermore, describing the high growth rates of the gold and fisheries boom as 'success stories' (p.5) reveals a priority for short-term economic decision mak- ing. There are environmental (i.e. landscape destruction in gold mining, illegal fishing practices) and social (health and gender issues, workplace security is- sues, food security issues) trade-offs associated with gold mining and fisheries. The local employment effect of the gold boom is also questionable. One would need to take all these factors into consideration to assess whether gold and fish- eries are still success stories if sustainability and equity considerations are taken into account. Agriculture and Land Under the heading 'unlocking Tanzania's potential in agriculture' (p.24), it is assumed that the country has large underused land areas. This assumption does not take into consideration that not all of the land is fertile. Land degradation and soil fertility were highlighted as main limitations to growth in the agricul- tural sector in the Tanzanian Participatory Poverty Assessments (2003). Similarly, the note emphasizes the huge potentials in Tanzania due to untapped natural resource endowments (p.24). It is true that Tanzania is rich in natural resources and potential to exploit these is there, however a key factor to the long-term exploitation of renewable natural resources is their sustainable use. Signs of land degradation are increasingly visible in Tanzania (deforestation around Dar es Salaam, Morogoro, Iringa due to urban demands for charcoal). The CEM needs to give issues of sustainable land and natural resource use higher priority. In close relation to this, the concept note mentions issues of land in relation to assets (p.28). For the agriculture and natural resource sectors access and owner- ship of land are very important to the development of economic activities and livelihoods of people. In the development of NSGRP issues of property rights, using land as collateral were important and controversial issues. The discussion in the CEM will need to include tenure regimes and access rights and how these impact on growth. Furthermore, the note does not identify water as one of the constraints to agri- cultural productivity and advocates increased irrigation. Water is often the lim- iting issue and dependence upon rain fed agriculture is reflected in agricultural growth rates. Competition for water for productive (energy, agriculture) and environmental services is increasing. Irrigation will and already is resulting in conflict between water users, such as for example in the Usangu Plains. In rela- tion to growth it is important to keep in mind that water is not seen as a free good, and that costs of maintaining catchments is factored in, and the sustain- Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 able management of water resources is included. This would imply "true cost- ing" of economic growth. Forestry and Wildlife Forestry and Wildlife are not included in the CEM at all. The concept note identifies the need to investigate growth in agriculture and agro-processing, but gives no mention of growth in forestry and wildlife sectors. Although both sec- tors have not been high performers of growth in the last few years and their contribution to GDP is still low, there is potential for increased future growth in these sectors. At the same time there is a need to avoid the unsustainable ex- ploitation of these resources (see recent evidence from illegal logging in Rufiji). Energy and Water There is little discussion about the role of energy and water in supporting in- creased growth. The concept note mentions that electricity is the second most important constraint identified by business. But apart from a reference on page 30 in paragraph 101, there is no discussion on how the CEM will look at this sector. The provision of water and of reliable and affordable supplies of energy will be essential to growth, and also has major environmental implications. For example last year due to low water levels in the dams GoT took an emergency credit of $43 million to subsidise TANESCO to buy fuel for its turbines. Simi- larly, the major energy source for most people in rural and urban areas is bio- mass (charcoal or firewood). The main source of electricity in Tanzania is hydropower. Power rationing due to low water levels occur with certain regularity each year. This puts constraints on industrial processes and also on domestic water supply in urban areas. Thus, to promote growth in agricultural processing, manufacturing, tourism, and other sectors, an effective strategy on sustainable energy and water is needed. Environmental Externalities of Growth The analysis of trade offs is limited to trade offs between growth, types of em- ployment creation and poverty (page 17). Issues of trade off between these and the environment are not included. To maintain growth in tourism we need to conserve the natural environment on which the industry depends while at the same time minimizing the impact on other land users who may not be directly benefiting from tourism. Similarly, as mentioned earlier, there are externalities generated through growth in the commercial mining sector affecting growth and peoples' livelihoods in the artisanal mining and fisheries sector. The possibility of increased levels of pollution as an externality of increased growth, which often impacts poor people more needs to be included in the analysis, as well as mechanisms to control such externalities of growth. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 Environmental Risks Issues of vulnerability are indirectly mentioned in the section on risk as a cause of poverty (p. 27). However, there is not explicit mention of the relationship between growth and environmental risk. Agricultural growth rates are very de- pendent on extremes of weather, i.e. droughts and floods. The relationship be- tween growth and environmental risk needs to be investigated. A further point to consider is increasing risks and frequencies of drought and flooding (models suggest that this may be the case) and their implications on economic growth. 3.3 Proposed Revised Structure of the CEM Based on the arguments provided in the sections above, a revised structure for the CEM is proposed. The draft outline of the CEM provided in Annex 1 of the concept note is the basis for the revised structure proposed here. Part 1: Economic Growth ­ Recent Developments and Prospects Add Chapter G: Natural Resources - The basis of growth in Tanzania This additional chapter should discuss the current and future potential growth contributions of Forestry and Wildlife to economic growth; as well as trade-offs of the success stories of growth in the Mining and Fisheries sectors. Sustained yield considerations should be brought into the analysis of `sustained growth' as well as the importance of sound environmental management to maintain growth in Tourism and the Tanzanian economy at large. Part 2: From an Improved Environment for Private Sector Activity to In- creased Private Sector Activitiy According to page 11 of the concept note focus of this part of the CEM is on the elements of a transition to a sustained growth path. In order to achieve this, it is crucial to discuss policies and production methods that ensure management of resources on a sustained yield basis as well as minimizing environmental externalities. Add chapter: E. Environment and natural resource management as a key factor for a transi- tion to a sustained growth path. This additional chapter should discuss the importance of maintaining a critical stock of natural capital during the process of economic growth. It should em- phasize the importance of environmentally responsible investment activities (undertaking EIA, SEA and Auditing etc) Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 Part 3: Enhanced Participation of the Rural Population in Economic Growth According to page 11 of the note, part 3 will analyze the opportunities and risks for future growth. Although this is not really reflected in the title and chapters under part 3 listed in Annex 1, it is recommended to include the following in such a chapter: Risks: · Limited stocks of exhaustible resources and Maximum Sustainable Yield Considerations of renewable resources; · Environmental externalities created by economic growth ; · Data constraints and institutional capacity for environmental planning and monitoring. Opportunities: · Growth opportunities in natural resource based sectors, e.g. carbon trade, wildlife This chapter should also include the discussion of the role of the environmental services such as hydrological services and their impact on agriculture, hydropower generation and domestic use etc. Furthermore, natural resources as a source of rural livelihoods need to be discussed. All these are opportunities for growth and poverty reduction. Part 4: Poverty and Growth Interactions Add Chapter: I: Growth, Environment and Poverty reduction. The Chapter C on 'risk as cause of poverty' should include environmental risks and vulnerabilities. Part 5: Towards a Strategy for pro poor growth Needs to discuss the role of natural resources for pro poor growth and the im- portance of sound environmental policy and planning to minimize externalities in the growth process, which would affect the poor more severely than the rich due to lack of means to protect themselves. 3.4. Specific Edits and Inserts The inserts and edits to the CEM concept note are structured according to the headings of the note. For each insert or edit the precise page number and loca- tion is provided. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 Under point 5, page 2, insert after 'recently observed": "A more cautious interpretation also needs to focus on the fact that the main drivers of economic growth in Tanzania are natural resource based and will not sustain growth in the long run if utilized beyond maximum sustainable yield capacity." At the end of point 5 on page 2, add: ..., "as well as exploitation of natural resources based on sound natural re- source economics, in particular stock assessments, and sustainable yield calcu- lation and the monitoring of their implementation by industry in the respective sectors." In Box 1 on page 4 insert paragraphs as follows: " Forest Sectors. The revision of the Forest Policy (1998) and legislation (For- est Act of 2002 enacted in 2004) shows commitment to principles of sustainable development, privatization and transparency in the management of forest re- sources. The recent logging scandal in Rufiji was a serious drawback to this progress in recent policy reform, making clear that implementation of trans- parency is still not being realized. The introduction of an autonomous Execu- tive Forest Agency is envisaged to improve investments in private forest estate development." " Wildlife Sector. Introduction of guidelines on Wildlife Management Areas have paved the way for community based wildlife management, being piloted by government in 15 pilot sites in the country. True commitment of government leadership to this decentralized approach is yet to be demonstrated." At the end of box 1 after 'closely monitored' include: "As a prerequisite for Tanzania to achieve a path of sustainable economic growth, the finalization of the long outstanding Environmental Management Act of 2004 has clarified the administrative and institutional arrangements of environmental planning and management of Tanzania. It provides the tools for sound regulation and control of private and public investment and policy deci- sions at national, local and sector" After point 13. insert an additional point: "The three drivers of growth, mining, tourism and fisheries are fully based on natural resources. Sustained future growth in these sectors hence requires the management of these resources on a sustained-yield basis. Presently there is little information on stocks on both fisheries and gold and there is little moni- toring, control and surveillance of fisheries resources. The recent growth in fisheries is based on 80% fish exports from Lake Victoria to the EU. There are signs that these growth rates are not sustainable and that stocks of Nile Perch are reducing. While fisheries are a success story in pure economic terms, this is not necessarily the case in terms of poverty reduction. The positive impact on Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 income poverty through increased local employment in the fisheries industry around the lakes, may well have been set-off through a negative impact on live- lihoods of aritsanal fisher folk and on nutrition of communities whose liveli- hoods are based on the lake and where over-fishing is making an impact. "Another key area for growth in fisheries is the near and off shore coastal fish- eries. There are increasing numbers of foreign trawlers in Tanzania's EEZ. While there are improvements to the monitoring and surveillance of these wa- ters, there is probably a high degree of illegal fishing and loss in revenue (and growth) to Tanzania from this activity. In addition, illegal trawling in coastal waters is impacting on artisanal fisher folk and on fish stocks. The challenge is how the country can sustainably exploit its marine and lake resources, for in- creased sustainable levels of growth - much of this is the focus of the forthcom- ing Coastal Livelihoods Project supported by the WB." Under point 17 on page 8 insert a sentence: "Similarly in the tourism sector, the availability of trained personnel has been a constraint to realizing the full employment benefits for the local population." At the end of point 24 on page 10 include a sentence: "In addition assets, including land, are critical correlates of poverty. The 2002/03 Tanzania PPAs emphasises that their absence or decline in access can constitute an important impoverishing factor. The forthcoming results of the agricultural survey will allow more exact statistical calculations on the corre- lation of size of land holdings and poverty levels." After point 24 on page 10 include additional paragraph: "Although the correlation between regional differences in poverty levels and agro-ecological conditions has not been statistically tested, it is assumed that such correlation is high. Rural economies and livelihoods predominantly de- pend on the utilization of the natural resources base, this is for both on and off farm activities. In order to fully utilize the economic potential of natural re- sources for the rural poor now and in the years to come, the improvement of rural infrastructure and markets, reduction of taxes and levies, secure tenure policy and management of resources on a sustained-yield basis are of vital im- portance. Important trade-offs will need to be taken into account in economic decision making where positive economic development may impact negatively on the livelihoods of poorer groups and the larger natural environment. For example, the introduction of large scale paddy rice schemes in Usangu Plains, which was one of the main reasons for reduced dry-season flow of the great Ruaha River, impacted on smaller rice farmers downstream and the Ruaha Na- tional Park eco-system, which is also an important tourist site." Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 Point 25, Page 10, before the sentence starting with "Human capital..." insert: "Many of these off farm activities are natural resource based. Beekeeping for example, has become an increasingly important source of income in some ar- eas of Tanzania" Point 26, Page 11, insert edit: "... as well as policies that ensure that economic growth is based on sustained yield calculations and creates income earning..." Point 27, Page 11, include inserts: In part 1 of the CEM, which deals with analysis of recent growth performance, include a section on the role of natural resources as an input into economic growth. Part 2 of the CEM on elements of the transition to a sustained growth path, should include after "entrepreneurial activities" "...and the introduction of poli- cies and production methods that ensure management of resources on a sus- tained-yield basis as well as minimization of negative externalities of increased economic growth." In art 3 of the CEM dealing with opportunities and risks that may have bearing on future growth, insert: "...opportunities in the natural resources sector and risks the stocks of non-renewable resources and maximum sustainable yields of renewable resources'. Point 30, Page 11, insert: The spatial dimensions of economic growth and their correlation with agro- ecological conditions, will also be... Page 13, after point 34 insert paragraph on natural capital "Similarly quantity and quality of natural capital are important determinants of economic growth. In the medium and long term, growth in sectors, which are based on renewable natural resources, e.g. fisheries, can only be sustained if production methods and extraction rates follow maximum sustainable yield calculations. Freshwater fisheries in the Lakes of Tanzania, presently follows private profit maximising yields, which do not consider factors internal to the resources, in particular rate of renewal, as well as the external economy, e.g. price, discount rate and the institutional framework. Non-renewable resources, e.g. gold, have a fixed stock. Therefore the marginal product of the resource or 'growth' component is zero. Increased economic growth can hence only be based on improved productivity in the production process or discovery of new stocks. Based on Hotelling's rule, in the long-run, there can be no gain from shifting extraction period between time periods. Im- Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 portant factors that feed into consideration about optimal extraction rates are the extraction cost, change in discount rates and increase in demand on the resource. Page 13, after point 37 insert paragraph: "Despite recent progress of the institutional and legal reform in the Environ- ment sector through the finalization of the Act on the Environmental Frame- work Law (2004), environmental planning, management and enforcement of safeguarding policies is still at very initial stages. The CEM will therefore pre- sent an account of current environmental management and planning practice in Tanzania and compare with critical elements that would need to be in place to ensure that recent and future economic growth follows a path of sustainable development.". Page 14, after Point 41 insert: "This ambiguity is for a number of reasons. First... [existing text on Dutch dis- ease effect]. Second, The economic growth through the gold boom did not translate into income poverty reduction and is associated with environmental and social trade-offs. The local employment benefits generated through the gold industry are considered to be very limited in scope and timescale, and there may even be an overall negative impact through an increasing income gap between artisanal and large scale mining." Page 17, Point 51, first paragraph two text inserts: First sentence in brackets: ...(land, labour, natural and physical capital...) Fifth sentence ... to make explicit trade-offs between growth, employment crea- tion, environmental sustainability and poverty in the .... Page 17, Point 51, second paragraph. The section on overcoming con- straints to ...growth "will include analysis of the freshwater fisheries, where private profit maximis- ing strategies and illegal methods have put future growth at risk. A careful analysis will also be given to the coastal fisheries, in particular regarding gov- ernment institutional capacities as a key input into avoiding risk factors for fu- ture growth." In addition insert paragraph: "The analysis of trade-offs between growth, types of employment creation, and poverty will also include issues of trade off between these and environment." Page 17, add to point 53: "However, the 'success' stories of fisheries and gold exports rather tell a story where accelerated growth is a sign of government failure to control extraction Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 for a resource to sustain its use and to direct investment so that the local popu- lation gains maximum benefits and equity considerations are applied. Under such conditions accelerated growth will not be sustainable." Page 18, point 55 add 'natural capital' Page 21, add paragraph after point 69: " Part of the analysis will also investigate how public, parastatal and private sector agents have produced environmental externalities, in particular in urban areas, where increased pollution of air and water sources, lack of proper sew- erage facilities and waster management is becoming an increasing problem." Page 21, add to point 70: "It will also review the progress in the institutional and legal framework to en- sure sound environmental planning, management and enforcement to accom- pany future economic growth." Page 23, add point c) "Environmental and vulnerability assessment" · The relationship between growth, poverty and the environment, envi- ronmentally induced risks and vulnerability" Page 24, paragraphs 78-80, add after 79: "While Tanzania is rich in natural resources and potential to exploit these ex- ists, key to the exploitation of renewable natural resources is their sustainable use. Increase evidence of unsustainable use with increased land degradation is already visible in parts of the country (e.g. deforestation around Morogoro, Dar es Salaam and Iringa for charcoal production). Issues of sustainable use hence need to be given high priority when discussing the country's potential to unlock agriculture. " Page 27 section risk as a cause of poverty Will include environmental risks, in particular droughts and floods. Page 28, point 93, insert: "The discussion on asset ownership will include land rights and access and use of natural resources and how this impacts on growth." Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 Annexes Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 Annex 1: Terms of References Terms of Reference for Study on Growth and Environment Links as Con- tribution for Country Economic Memorandum, 8. March 2005 Background The World Bank and the Government of Tanzania are presently preparing a country economic memorandum (CEM) and poverty assessment. A concept note for the CEM has been prepared and a mission was held from 2 ­ 26 No- vember 2004 to discuss the preliminary work that has been carried out so far, and to seek additional documentation and discuss relevant issues with authori- ties, private sector organisations, firms, and other stakeholders. During this mission it was identified during consultations with the Research and Analysis Working Group (RAWG) that among other issues the CEM needed to include a better analysis of the linkages between sectoral growth and development initia- tives and environmental issues. From other initiatives, the GOT has clearly stated that growth-environment linkages are of concern. Tanzania's economy and its prospects for growth de- pend upon the use and exploitation of the country's natural resources. Over two thirds of the country's GDP is dependent upon natural resources. This includes Agriculture, Forestry, Wildlife and Fisheries, but also the sectors with the high- est growth rates tourism and mining. In addition another key sector for growth, energy, is dependent upon natural resources as the main source of energy - elec- tricity from hydropower schemes and biomass for the majority of the popula- tion. The new PRS, the National Strategy for Growth and Poverty Reduction, better known by its Kiswahili acronym MKUKUTA has further emphasised the na- tional importance of environment to growth and poverty reduction. Fourteen percent of the MKUKUTA's targets relate to environment and emphasis is placed on strategies: to promote sustainable growth in economic, environment and social terms; to reduce poor communities' vulnerability from environ- mental risk (e.g. drought); and, to implement the new Environmental Manage- ment Act to protect peoples' livelihoods, and the environment and to promote sustainable development. This study is being commissioned to ensure that growth-environment linkages are integrated into the overall analysis of economic growth in the CEM and to produce a background paper elaborating linkages of relevance, with a focus on sectoral issues dealing with forestry, fisheries and wildlife, provision of water and energy services, and reduction of vulnerability from environmental risk. Objectives The purpose of the study is to identify and elaborate on the linkages between sectoral growth and development initiatives as elaborated in the CEM, and en- vironmental issues of concern to the GoT and the World Bank. Specific objec- tives of the study are: Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 (a) To clarify the contribution environmental and natural resources (ENR) have to GDP and to identify the potential for increased growth from the sustainable use of ENR, and constraints to the achievement of this growth. Identify the how the sustainable use of ENR can contribute to growth and reduction in inequalities between rural and urban areas, and between different rural areas. (b) To outline the risks to the achievement of growth in key sectors from en- vironmental factors (e.g. drought and floods), and measures to be taken mitigate against these. (c) To identify the potential impact of growth in key sectors on the environ- ment, and any trade offs to be made between growth and the environ- ment. Specific Tasks and Outputs The study will consist of two phases with distinct tasks and outputs: Phase 1: (20 days) Specific tasks to be undertaken by the Consultant(s) are (f) To undertake a critical review of the CEM concept note to identify where issues of growth and environment linkage are required. (2) (g) To review documentation on links between growth and poverty, and the environment provided by the World Bank and other Development Partners. This should include the study on poverty and environment links commis- sioned by the World Bank, the PER of environment, and the MKUKUTA. (3) (h) To collate and review additional documentation and data on links between growth and poverty and the environment. (5) (i) To participate in a two-day workshop on growth and poverty reduction or- ganized in the context of preparation of the CEM and review background papers being discussed at the workshop. (2) (j) To prepare written comments on drafts of the CEM. (2) (k) To prepare brief inserts and edits for the CEM relating to growth- environment linkages, this should include relevant data and tables where this data exists. (4) (l) To discuss the findings of the study with relevant stakeholders such as the Development Partner Sub-group on Environment and Natural Resources and members of the Working Group on the Environment (2) The following outputs are expected: (a) A brief paper containing two sections: I) overall comments on the CEM and ii) insert and edits to the CEM (b) Both of the above delivered in softcopy Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 (c) Dissemination materials including power point presentation of the findings. Phase 2: (26 days) Specific tasks to be undertaken by the Consultant(s) are (a) To analyse existing economic and growth data (including economic survey) to determine the contribution of natural resource sectors to growth, and their potential for increased rates of sustainable growth. (6) (b) To interview key stakeholders from relevant sectors including agricul- ture, forestry, fisheries, wildlife, tourism, water and energy and miner- als to seek their views on growth and environment links and obtain relevant data on environment contribution to growth, and impact of growth on the environment. (10) (c) To prepare a technical paper elaborating linkages of relevance, with a focus on sectoral issues dealing with forestry, fisheries, wildlife, tour- ism and minerals, the provision of water and energy services, reduction of vulnerability from environmental risk, and actual and potential im- pacts from growth on the environment. The background paper will also include a brief review of existing initiatives, documents and persons consulted. (6) (d) To integrate comments received from the World Bank and the DPG. (2) (m)To prepare a power-point presentation and present the results to the De- velopment Partners Group (DPG). (2) The following outputs are expected: (a) A technical paper to contribute to the CEM and other macro processes such as the Joint Staff Assessment (JSA) and Joint Assessment Strategy (JAS) and for wider circulation and distribution. Consultant qualifications and profile The consultant(s) should have expertise in economic planning, environment and natural resources management, and environmental economics. The consult- ant(s) must have Tanzanian experience and be familiar with environmental re- forms and the MKUKUTA. Management Arrangements and Time Schedule The consultant will report to the World Bank task manager for the CEM and the Senior Environmental Specialist based in Dar es Salaam. Phase 1 shall commence on 7. March and a draft of the work should be made available to the World Bank by 23 March 2005 in time for incorporation into the final CEM. Phase 2 shall commence on 18 April and a draft technical paper shall be sub- Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 mitted to the World Bank on 16 May 2005. It is expected that will involve ap- proximately 46 days of work in total. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 Annex 2: List of References Reviewed literature marked with an asterisk(*) Atkinson, Giles and Hamilton, Kirk (2003) "Savings, Growth and the Resource Curse Hypothesis", World Development, vol.31, no.11, pp. 1793-1807. Auty, R. M. (1997) "Sustaining Development in Mineral Economies: The Re- source Curse Thesis", Journal of International Development, vol.9, no.4, p.651-663. *Boschini, Anne; Pettersson, Jan; and Roine, Jesper (2003) "Resource curse or not: A question of appropriability", Working Paper Series in Economics and Finance, no.534, full text available at http://swopec.hhs.se/hastef/papers/hastef0534.pdf, Stockholm: Stockholm School of Economics. *Bravo-Ortega, Claudio and De Gregorio, José (undated) "The Relative Rich- ness of the Poor? Natural Resources, Human Capital and Economic Growth" in its series Working Papers Central Bank of Chile, no.139, full text available at http://www.bcentral.cl/esp/estpub/estudios/dtbc/pdf/dtbc139.pdf, Santiago de Chile: Central Bank of Chile. Cavendish, W. (1999) "Empirical Regularities in the Poverty-Environment Re- lationship of African Rural Household", Working Papers Series, no.99-21, abstract available at http://ideas.repec.org/p/fth/oxesaf/99-21.html, Oxford: University of Oxford. *Cervellati, Matteo and Fortunato, Piergiuseppe (2004) "Natural resources and the wealth of nations in a globalized world economy" Cahiers de la MSE, no.v04068, full text available at ftp://mse.univ- paris1.fr/pub/mse/cahiers2004/V04068.pdf, Paris: Maison des Sciences Economiques, Université Paris Panthéon-Sorbonne. *Dasgupta, Susmita et al. (2003) "The Poverty/Environment Nexus in Cambo- dia and Lao People's Democratic Republic", Policy Research Working Pa- per Series, no.2960, with Uwe Deichmann, Craig Meisner, and David Wheeler, full text available at http://econ.worldbank.org/files/23318_wps2960.pdf, Washington D.C: The World Bank. De Long, J.B.and J.G Williamson (1994) "Natural Resources and Convergence in the 19th and 20th Centuries", unpublished paper, Harvard University. *DFID (2002) "Better livelihoods for poor people: The Role of Agriculture", Issues Paper ­ Consultation Document, prepared by DFID Rural Liveli- hoods Department, London: Department for International Development. *GoT (2005) National Strategy for Growth and Reduction of Poverty, full text available at http://www.povertymonitoring.go.tz/downloads/new/nsgrptext.pdf, January 2005 Grossman, G. M. & Krueger, A. B. (1991) The Environmental Impact of the North American Free Trade Agreement, Working Paper 3914, Massachu- setts: National Bureau of Economic Research. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 Grossman, G. M. & Krueger, A. B. (1995) "Economic Growth and the Envi- ronment", Quarterly Journal of Economics, vol.110, pp.353-377. Gylfason, Thorvaldur (2000) "Resources, Agriculture, and Economic Growth in Economies in Transition", CERGE-EI Working Papers, no.wp157, full text available at http://www.cerge-ei.cz/pdf/wp/Wp157.pdf, Prague: The Center for Economic Research and Graduate Education - Economic Insti- tute. Gylfason, Thorvaldur and Zoega, Gylfi (2001) "Natural Resources and Eco- nomic Growth: The Role of Investment", C.E.P.R. Discussion Papers, no.2743, full text available at http://www.cepr.org/pubs/dps/DP2743.asp, London: Centre for Economic Policy Research. Gylfason, Thorvaldur; Herbertsson Tryggvi, Thor and Zoega, Gylfi (1997) "A Mixed Blessing: Natural Resources and Economic Growth", C.E.P.R. Dis- cussion Papers, no.1668, full text available at http://www.cepr.org/pubs/dps/DP1668.asp, London: Centre for Economic Policy Research. *Heltberg, Rasmus and Nielsen, Uffe (2000) "Foreign aid, development and the environment" in Finn Tarp (ed.) Foreign Aid and Development, London: Routledge. *Hodler, Roland (2004) "The Curse of Natural Resources in Fractionalized Countries", Diskussionsschriften, no.dp0404, full text available at http://www.vwi.unibe.ch/publikationen/download/dp0404.pdf, Bern: Uni- versität Bern. *Korongo Ltd. (2003) Poverty and the Environment in Tanzania ­ A Prelimi- nary Study of Environment and Poverty Linkages, study commissioned by the World Bank, Dar es Salaam Country Office. Lane, P. and Tornell, A. (1995) "Power Concentration and Growth", Harvard Institute of Economic Research, Discussion Paper no.1720. *Martinussen, John Degnbol (1999) Samfund, Stat og Marked, Koebenhavn: Mllemfolkeligt Samvirke. Matsuyama, K. (1992) "Agricultural Productivity, Comparative Advantage, and Econmic Growth", Journal Of Economic Theory, no.58, pp.317-334. *Papyrakis, Elissaios and Gerlagh, Reyer (2004a) "Natural Resources, Invest- ment and Long-Term Income", Working Papers, no.2004.87, full text avail- able at http://www.feem.it/NR/rdonlyres/6A5BB1B4-80EF-44FE-8C15- 70510DC411CE/1178/8704.pdf, Milan: Fondazione Eni Enrico Mattei. *Papyrakis, Elissaios and Gerlagh, Reyer (2004b) "Natural Resources, Innova- tion, and Growth", Working Papers, no.2004.129, full text available at Pearce, David (2004) "Growth and Environment: Can We Have Both?", Envi- ronment Matters 2004 Annual Review, pp.14-15. *Sachs, Jeffrey D. and Warner, Andrew M. (1995) "Natural Resources Abun- dance and Economic Growth", NBER Working Papers, no.5398, full text available at http://www.nber.org/papers/w5398.pdf, Cambridge, MA: Na- tional Bureau of Economic Research, Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 1 Sachs, Jeffrey D. and Warner, Andrew M. (2001) "The curse of natural re- sources", European Economic Review, vol.46, pp.827-838. *Smith, Ian (200?) "Welfare, Growth and Environment: A Sceptical Review of the Skeptical Environmentalist, Discussion Paper Series, no.0204, full text at http://www.st-and.ac.uk/academic/economics/papers/dp0204.pdf, Univer- sity of St. Andrews. *Stijns, Jean-Philippe (2001) Natural Resource Abundance and Economic Growth Revisited, full text available at http://econwpa.wustl.edu:80/eps/dev/papers/0103/0103001.pdf., CA: Uni- versity of California at Berkeley. *Verbeke, Tom and De Clercq, M. (2003) "The income-environment relation- ship: Does a logit model offer an alternative empirical strategy?" Working Papers of Faculty of Economics and Business Administration, no.03/192, full text available at http://www.FEB.UGent.be/fac/research/WP/Papers/wp_03_192.pdf, Ghent: Ghent University Vosti, Stephen A. and Reardon, Thomas (2001) "Sustainability, Growth and Poverty Alleviation ­ A policy and Agroecological Perspective", Food Pol- icy Statements, no.25, full text available at http://ifpri.org/pubs/fps/fps25.htm, Washington, D.C.: International Food Policy Research Institute. The World Bank Study on Growth and Environment Links for Preparation of Country Economic Memorandum (CEM) Part 2: Uncaptured Growth Potential ­ Forestry, Wildlife and Marine Fisheries Final report May 2005 The World Bank Study on Growth and Environment Links for Preparation of Country Economic Memorandum (CEM) Part 2: Uncaptured Growth Potential ­ Forestry, Wildlife and Marine Fisheries Final report May 2005 Report no. 2 Issue no. 2 Date of issue 2 May 2005 Prepared KEP Checked TNH Approved Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 iii Table of Contents 1 Introduction 7 2 Forestry 8 2.1 GDP Contribution 9 2.2 Forest Resources 10 2.3 Sector Financing and Spending 15 2.4 Production and Consumption 22 2.5 Exports and Domestic Trade 24 2.6 Employment 27 2.7 Unaccounted Forest Services and Externalities 27 2.8 Future Growth and Recommendations 35 3 Wildlife 38 3.1 Policy Framework 39 3.2 GDP Contribution 40 3.3 Wildlife Resources 41 3.4 Contribution of Consumptive Wildlife Resources Use 42 3.5 Wildlife Division Income and Expenditure 44 3.6 Live Animals Export 50 3.7 Wildlife Resources and Poverty Reduction 52 3.8 Sustainability 55 3.9 Conclusions and Recommendations 58 4 Marine Fisheries 60 4.1 Contribution of Fisheries to GDP 61 4.2 Sector Financing and Spending 63 4.3 Marine Fisheries Resources 66 4.4 Revenue from Marine Fisheries 71 4.5 Marine Resources and Poverty Reduction 73 4.6 Sustainability of Marine Fisheries 75 4.7 Policy framework for Marine Fisheries 77 4.8 Recommendations 79 Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 iv List of boxes Box 1 Key estimates on the economic contribution of Forests 9 Box 2 Facts and figures on charcoal production 30 Box 3 Monetarizing the contribution of forests to water supply 31 Box 4 Contribution of Eastern Arc Forests to water and energy supply 32 Box 5 Ways to increase revenue from the Forestry sector 36 List of charts Chart 1 Overall GDP and contribution of Forestry 10 Chart 2 Regions in Tanzania, Road density and forest reserves, 2002 13 Chart 3 Regions in Tanzania, Income and forest reserves, 2002 14 Chart 4 Forestry sector revenue, 1987/88 - 2003/04 15 Chart 5 Distribution between sources of revenue, 2003/04 17 Chart 6 Revenue collected and retained at source, 2002/03 ­ 2003/04 18 Chart 7 Development budget, 2003/04 20 Chart 8 Recurrent versus development spending, 2001/02 ­ 2003/04 21 Chart 9 Degree of self-financing of FBD 22 Chart 10 Composition of Forestry exports by product, 2001/02 24 Chart 11 Beeswax exports, 1988/89 -2003/04 25 Chart 12 Honey exports, 1988/89 ­ 2003/04 25 Chart 13 Trend in gum and resins export and revenue, 1997 - 2003. 27 Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 v List of tables Table 1 Forestry reserves in Tanzania by type and land area 11 Table 2 Forest revenue by Forest type 16 Table 3 Revenue collection by source, 2002/03 ­ 2003/04 17 Table 4 Budget of MNRT, distribution by sub-sector, 2002/03 ­ 2003/04 19 Table 5 Industrial wood consumption, 1988, 1989 and early 1990s. 23 Table 6 Inputs of forests to Tobacco production 29 Table 7 Projection of charcoal production impact on forests, 2002-20 30 Table 8 Sources of income in villages, Tanzania, 2002 34 Table 9 Areas occupied by National Parks in Tanzania. 41 Table 10 Value of Tanzania's Wildlife Resources in regional comparison 42 Table 11 Tourist hunting and live animals export earnings, 1994 ­ 2003 43 Table 12 Revenue collection by source, 2002, 2003, 2004 46 Table 13 2003 Revenue by selected District Councils from Game Licenses 46 Table 14 Export of Government Trophies, Live Animals and Birds 51 Table 15 Export of Live Animals and Animal Tusks/Teeth 52 Table 16 Marine Fisheries resources and their value (base year 2000) 67 Table 17 Management advice from the IOTC 70 Table 18 Catch and Value of Tuna and Tuna-like big pelagis in EEZ, 2004 70 Table 19 HDI and HPI for coastal regions 74 Table 20 Contribution of fish to household subsistence in coastal areas 75 Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 vi List of abbreviations and acronyms CEM Country Economic Memorandum CFR Catchment Forest Reserve DWFN Distant Water Fishing Nations CITES Convention on International Trade in Endangered Species EEZ Exclusive Economic Zone ENR Environment and Natural Resources FA Fisheries Agreement FBD Forestry and Beekeeping Division (under MNRT) GDP Gross Domestic Product GoT Government of Tanzania GRT Gross Registered Tonnes IOTC Indian Ocean Tuna Commission IUU Illegal, Unreported and Unobserved LMDA Logging and Miscellaneous Development Account MNRT Ministry of Natural Resources & Tourism MSY Maximum sustainable Yield NFP National Forest Programme NTFP Non-Timer Forest Products PER Public Expenditure Review PFA Private Fisheries Agreement RFO Regional Fisheries Organisation SPM Southern Paper Mill TFCMP Tanzania Forest Conservation and Management Project TIC Tanzania Investment Centre Tsh. Tanzanian Shilling TWICO Tanzania Wood Industry Corporation URT United Republic of Tanzania Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 7 1 Introduction This paper is Part 2 of the Study on Growth and Environment Links for the preparation of the Country Economic Memorandum (CEM) of Tanzania by the World Bank. While Part 1 discussed the relevant literature and provided comments on the draft concept note of the CEM, Part 2 provides background data on the Forestry and Wildlife Sectors, both of which are excluded in the early drafts of the CEM. Part 2 also includes Fisheries with particular focus on Marine Resources. Part 3 of the study discusses the Mining, Freshwater Fishery, and Tourism sec- tors and potential externalities of the recent 'success stories' of growth in these sectors. Part 3 is presented in a separate paper. The Forestry and Wildlife Sectors combined contributed on average 3.3 percent to national GDP over the last ten years, which largely under represents their potential contribution to the national GDP, economic growth and rural wealth. While this under-valuation is partly due to under-accounting in the System of National Accounts, corruption and inefficient administration and management in both the Forestry and Wildlife sectors lead to losses in government revenue and livelihoods benefits for the local population. This report is a secondary data analysis, aimed at compiling quantitative infor- mation on the contribution of Forestry and Wildlife to economic growth as an input into the CEM. Data availability and quality were constraints to the analy- sis. The sources of information are statistics from the Ministry of Natural Re- sources and Tourism, Bureau of Statistics, existing literature and project re- ports. The examined literature is listed in Annex 2 of this report, while the Terms of Reference for this assignment are annexed to the separate Part 1 re- port. The report presents first in Chapter 2 the Forestry sector, in Chapter 3 the Wild- life sector and then in Chapter 4 the Fisheries Sector. While Chapter 4 presents some general data on the contribution of Fisheries to GDP and Sector Financ- ing and Spending, it focuses primarily on Marine Fisheries in the remaining Sections. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 8 2 Forestry The difficulty of examining Forestry in the context of economic growth arises through the fact no markets exist for many of the sectors contributions to the Tanzanian economy and population. Hence there is no objectively verifiable monetary value in which these services could be measured. Water services, biodiversity, climate regulation, and cultural values are examples of such forest services. Although these `ecosystem functions' of forests can be determinants to growth in other sectors, via the water and energy supply chain, these values are usually not reflected in the GDP. Many transactions related to forest prod- ucts and services, although markets exists, fall within the informal sector or are undertaken illegally and are hence not recorded either. These are for example, the sale of non-timber forest products or illegal logging for timber and charcoal production. Official GDP figures, on which the analysis of economic growth, and so the CEM, is usually based, do not reflect the 'true' economic importance of the forest sector to the national economy. The reason why this 'under valuation' matters is that the contribution to GDP and its growth determine decision making by the Government of Tanzania and also to some degree its development partners. Pearce describes an `asymmetry of values' (1991), in the sense that economic decision making is typically bi- ased in favour of development options which can be calculated. Thus, unless incentives are devised whereby the non-market benefits of forests are `internal- ized' into the land-use choice, eco-system benefits will be downgraded. In the absence of monitoring data, reliable statistics and forest inventories, the valuation of the sector's economic contribution remains speculation. National estimates exist, but it is questionable how reliable they are. There are a few comprehensive studies that have tried to quantify the contribu- tion of the forest sector. These include Aku et al. (2000) who is estimating the economic contribution for the Kilimanjaro Regional Forest Sector, Monela et al. (2004) who calculate the total economic value of catchment forest reserves, Mkanta and Chimtembo (2002) with a study on the contribution of natural for- ests to national income, and Norconsult (2004) with their attempt to calculate the externalities of charcoal production. Other major studies in the context of forest economics are Salmi and Monela (2000) who present an overview on Forest Sector Financing as an input into the preparation of the NFP. The collec- tion of data in "Forestry in Figures" by the Forestry and Beekeeping Division presents a useful overview on some hard data, is however not up-dated on a regular basis. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 9 Box 1 below summarises some key estimates on the contribution of forests to the Tanzanian economy, provided by these sources: Box 1 Key estimates on the economic contribution of Forests · Tanzania's Forests provide · Employment to about 1 million people officially and about 5-10 times more unofficially and part-time. · 10-15 percent share of Tanzania's registered export earnings. · 2-3 percent of GDP for officially recorded forest products with the major cash value be- ing derived from timber, customary products and fuel. · 95 percent of Tanzania's energy supply through woodfuels. · Potential for tourism, the pharmaceutical industry and carbon sequestion, which is not captured presently. The value of Tanzanian forests for recycling and fixing carbon diox- ide is estimated to be US$ 1,500 per ha by Salmi and Monela (2000) and US$ 664 per ha by Turpie (2000). · around 75 percent of building materials. · 100 percent of indigenous medicinal and supplementary food products. (Source: Norconsult 2002:14) The quantitative data collated for this report aimed to include those contribu- tions to the economy and peoples' livelihoods which are quantified and re- flected in official government statistics as well as those not yet officially re- corded as far as quantitative data could be found in the existing body of litera- ture. This Chapter is divided into 8 sections. Section 2.1. will provide information on the GDP contribution of the country's forest resources. Section 2.2. presents an overview on the country's forest resources. Section 2.3. presents officially re- corded data on forest sector financing and spending. Section 2.4 investigates production and consumption of forest products, followed by Section 2.5. on export and trade and Section 2.6. on employment. Section 2.7. deals with unac- counted forest services and externalities and lastly, Section 2.8. with the future growth potential of the sector. 2.1 GDP Contribution The estimates of the contribution of the forest sector to GDP differ. Most esti- mates are between 2 (Mushi 1999) and 3 percent (Salmi and Monela 2002, URT 2004b) of GDP. The Economic Survey (URT, 2004c) quotes an average percentage contribution for Forestry and Hunting combined to the GDP of 3.3 percent over the period 1995 to 2002. In comparison Agriculture contributed 35.3 percent, Tourism 13.0 percent and Fisheries 2.5 percent during the same time period. Chart 1 below compares overall GDP to the monetary contribution of the forest sector to GDP. It shows that while GDP has been increasing constantly between 1990 and 1999, the contribution of Forestry remained around 3.5 percent on Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 10 average. Its contribution to growth has hence been constant at 1 percent over the period of 1988 to 2003 as reflected in the assessment of recent growth per- formance of Forestry and Hunting in the context of the CEM (Utz 2005). Chart 1 Overall GDP and contribution of Forestry 6000000 5000000 4000000 Overall GDP '000,000 3000000 Forest Sector Tsh.2000000 1000000 0 89 90 91 92 93 94 95 96 97 98 99 1988/ 1989/ 1990/ 1991/ 1992/ 1993/ 1994/ 1995/ 1996/ 1997/ 1998/ (Source: based on MNRT 2000) Studies that try to take non-industrial Forestry into account, present higher val- ues. For example, a World Bank study conducted by Sharma in 1992 values the contribution of the forest-industry, non-industrial Forestry and logging to be 13.9 percent of GDP in 1989. The amount of uncounted wood-fuels alone is estimated to be more than 30 million m3 per year (FOSA Country report). In the system of National Accounts, Forestry is added up into an aggregate value with Hunting and then summarized under the 'Agriculture GDP' with Fisheries, Livestock and Crops. Furthermore, the GDP undervalues the contri- bution of Forestry to the national economy. Forests and woodlands that support commercial industry are included in the national accounts. However, values of forest goods and services are often underestimated, wrongly attributed to other sectors, or entirely omitted. These include non-market timber, non-timber forest products, tourism and recreational services, and ecosystem services such as positive influences of forests on agricultural production, water and energy, car- bon storage and biodiversity protection. Tanzania is in the initial stages of de- veloping a system of Forest Accounts with support from the CEEPA Natural Resource Accounting Program for Eastern and Southern Africa, 2003-2006. 1 While Sections 2.3. to 2.6. describe recorded contributions of Forestry, Section 2.7. deals with the unrecorded ones. The next Section provides an overview of the resources of the Forest Sector. 2.2 Forest Resources There are a number of estimates of Tanzania's total forest cover and its rate of change. In 2001, the National Forest Programme estimated the country's forest and woodland resources at 33.5 million hectares, which constitutes 38 percent 1 Centre for Environmental Economics and Policy in Africa at University of Pretoria. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 11 of the total land area of Tanzania (URT 2001: 19). One of the many published estimates of the deforestation rate is 91,276 ha/year (MNRT, 2002).2 Government forest reserves According to official statistics from 2002, the country's 815 forest reserves3 cover about 15 million ha of Tanzania's land area, the majority of which is pro- ductive forests, totalling approximately 13 million ha or 74 percent of the entire forest estate. The remaining 26 percent are protected forests (MNRT 2002). Table 1 below shows the Forestry estate in Tanzania by type and use. 600,000 ha of forest were owned and managed by local governments in 2001 (MNRT 2001a: 20). Table 1 Forestry reserves in Tanzania by type and land area Use Category Quantity Total Land area (ha) % Productive 394 11,134,558 74 Protective 421 3,956,210 26 Total 814 15,090,769 100 (Source: MNRT, 2002. Note: Protective includes two different categories of protective forests, which where not specified in the source and hence aggregated here. Furthermore, declared and proposed reserves have also been aggregated) Industrial plantations In 2001 about 83,000 hectares of forestland was in government owned indus- trial plantations, distributed in 16 units throughout the country (MNRT 2001a: 23). Major plantation species are pines, cypress, eucalyptus and teak. The an- nual cutting potential is estimated at about 1 mill m3 (MNRT 2001a: 23). The potential of industrial plantations to contribute to the national economy is presently not realized. The reasons for this are summarized (based on govern- ment sources) as: · Insufficient supply of quality wood to support modern forest industry. A plantation wood production and supply forecast by the FBD estimates the gap to approximately 930,000 m3 per year for the period 2000-2010 (MNRT 2002). · Lack of incentives to increase productivity and maximising net revenue. · Net planting area and stock are declining in area and quality. · Poor management, leading to outdated management plans and inventories causing under stocking and overstocking in certain areas. · Shortage of skilled staff and investment capital. 2 About 8 other estimates were found during the course of this study only. 3 Out of which 608 declared reserves represent close to 14 mill. ha and 207 proposed re- serves, 1.2 mill. ha.. MNRT (2002). Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 12 Private Forestry In 2001 the area under private and community Forestry was estimated to 70,000 ­ 150,000 ha including community woodlots of small sizes (MNRT 2001a: 20). This is only about 1 percent of the country's total forest reserves.4 Private plantation management is promoted by the new Forest Policy and Act, and is still a fairly recent phenomenon in Tanzania. Based on TIC (2001), there are currently three privately managed plantations, including · Kilombero Teak company (Kilombero District), which plans to expand its plantation area from 2,900 ha (2001) to 10,350 ha by 2011, · Its subsidiary Tanganyika Wattle Company (Njombe District) grows be- sides wattle also pine and eucalyptus. These joint ventures with the gov- ernment have been financed by the Commonwealth Development Corpora- tion, · Escarpment Co/Kilombero Forest Ltd.(Mufindi) with 1,800 ha of pine and eucalyptus plantations. The main objective is to sequester CO2 and generate revenue from trading carbon credits. It is reported that 22 Tshs per tree per year are being paid to about 4,000 farmers. 5 Private Forestry by individual farmers and communities is taking place throughout the country, although generally on a small scale. The HASHI pro- ject in Shinyanga reports that communities have recently reforested about 350,000 ha of woodland. Investment in tree growing is limited to household use. Restricting factors to commercial tree growing by individuals have been lack of capital, poor land tenure arrangements and long term time horizon before returns will be realized (Salmi and Monela 2000). In addition the supply of 'free' wood produce from forests on general lands, makes it difficult to price wood products from private farms competitively. Forests on general land Forests on general lands are classified under the Forest Policy, Land and Vil- lage Land Acts and were estimated in 2001 to cover 19 million ha (MNRT 2001a). Due to the pressure on these forests for competing land uses (agricul- ture, livestock grazing, settlements and industrial development) the loss of for- est cover is estimated to be high. Lack of clear ownership, tenure and user rights have provided little incentive for sustained management of these forests by the villagers. Due to imperfect property rights, forest resources on public lands are incorrectly prices leading to a financial incentive for investments in conservation. The establishment of village forest reserves, a category intro- duced by the new Forest Act, is considered an instrument to solve this degrada- tion problem. 4 150,000 ha of private and community Forestry is about 1% of the 15 million ha of forest reserves. 5 Comments received by A. Boehringer, FBD-FOPIS project Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 13 Regional distribution of forest resources The regional distribution of the forest resources in the country varies consid- erably. Table A in Annex 1 groups 19 Regions based on the forestland cover. The three Regions with the largest forest reserves are Tabora, Rukwi and Lindi, which all feature 450,000 ha or above. Chart 2 Regions in Tanzania, Road density and forest reserves, 2002 40 Dar es Salaam sqkm) 35 per 30 25 roads 20 (km 15 Tanga 10 density 5 Tabora Ruvuma Road 0 0 100 200 300 400 500 600 Forest reserves ('000 ha) (Source: Ministry of Communications and Transport, and Author based on Forestry on Figures, 2002, Pearson correlation coefficient -0.55) The forest coverage is plotted against road density in Chart 2 above. The Chart shows a negative correlation between these two variables, indicating that the regions with high forest coverage are at the same time regions with limited in- frastructure development (using road density as indicator for infrastructure de- velopment). Two extreme examples are Dar es Salaam (no forest combined with high road density) and Tabora (extensive forest coverage coupled with low road density). Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 14 Chart 3 Regions in Tanzania, Income and forest reserves, 2002 700,000 600,000 Dar es Salaam prices 500,000 currenth 400,000 Ts in 300,000 Arusha 200,000 /capita Kagera Tabora GDP100,000 0 0 100 200 300 400 500 600 Forest reserves ('000 ha) (Source: NBS and Author based on Forestry on Figures, 2002. Pearson correlation coefficient ­0.36) Similarly, Chart 3 above reveals a (less significant) negative association be- tween forest area and income per capita for the regions. This association sug- gests a negative relationship between forest cover and economic growth: Higher growth regions have less forest coverage and vice versa. Policy framework for the management of forest resources The reform of the Forestry sector over the last decade included the policy and legal framework, with a revision of the Forest Policy (1998) and the Forest Act (2002), which was enacted in 2004. Procedures and regulations for implement- ing the Act are currently under preparation. The Policy advocates private and community based Forestry and provides legal basis for Joint Management of Forest reserves with catchment or biodiversity values. In 2003 a forest area of 1,085,306 hectares was under co-management in public lands while 1,863,623 hectares were in the forest reserves (URT, 2004c).6 Institutional changes are underway with the transformation of the Forestry and Beekeeping Division into an Executive Agency. Privatization has also bee in- troduced in the sector with ten companies involved to date (TIC 2001). Centralised forest management has contributed to both market and policy fail- ures in the sector. The NFP describes the low capacity of government institu- tions to control and manage the forests as a central constraint. The country's forest reserves have been suffering from degradation due to encroachment, over-utilization, fires, unclear boundaries, lack of systematic management and inadequate resources for controlling illegal harvesting as well as inefficient revenue collection system (MNRT 2001a: 23). 6 These figures do not match the data in Table 2 on page 5 although both data sets originate from government sources. Differences could not be clarified. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 15 The NFP also argues that inadequate development of economic activity to ab- sorb increased labour surplus from agriculture and livestock increase pressure on the country's forest resources. The 2004 logging scandal in Rufiji demon- strated that ineffective control systems and government corruption are still by far the greatest challenge that the forest sector has to face. Lack of data and monitoring of forest resources has been a major constraint faced by the Forest Administration both centrally and in the Districts. Efforts are currently underway to streamline existing elements of monitoring into a comprehensive national monitoring system under the National Forest Pro- gramme. 2.3 Sector Financing and Spending National income from Forestry is generated primarily through the collection of revenue from fees, royalties and licenses charged or issued in respect of forest reserves by the Forestry and Beekeeping Division (FBD) under the Ministry of Natural Resources and Tourism. Local Governments also have mandate to col- lect revenue from forest produce in local authority forest reserves. The 2002 Forest Act, provides legal basis for the allocation of management responsibili- ties of national forest reserves, including collection of revenue, to the desig- nated forest manager (Forest Act 2002, part V, section 27). Revenue collection Chart 4 below shows that the revenue collection in the forest sector has been steadily increasing from Tsh.77 million. in 1987/88 to 5,818 million in 2003/04. Chart 4 Forestry sector revenue, 1987/88 - 2003/04 7000 6000 5000 Tsh 4000 Mill.3000 2000 1000 0 8 /99 1987/881988/891989/901990/911991/921992/931993/941994/951995/961996/971997/919981999 /20002000 /01 2001/022002/032003/04 (Source: MNRT 2002 up-dated with data for later years collected directly from FBD containing reve- nue from natural forests, royalty and LMDA (Logging and Miscellaneous Development Account) from plantations) Comparison of the data with Salmi and Monela (2000) reveals that for the years Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 16 1995 to 1999 the total figure presented in the Chart above reflects only the revenue from Central Government Forest Reserves. It could not be verified if this is also the case for previous years. If so, then the total revenue is even higher. Additional revenue collected for charcoal production amounted to Tsh.662 mil- lion or 19.4 percent of total collection in 2001/02 while revenue from collection of firewood was Tsh.89 million or 2.6 percent of the total collection (URT 2004b: 24). Sources of revenue Table 2 below shows the distribution of revenue between Local and Central Government Forest Reserves and Plantations for 1995 to 1999. Central Gov- ernment Reserves earn by far the largest proportion of government forest in- come, followed by plantations. The proportion of income of Local Government Forest Reserves is comparatively small, but has been steadily increasing be- tween 1995 and 1999, while plantation Forestry shows a downward trend dur- ing the same period. Revenue accrues also from forests on general land, for ex- ample from charcoal licenses, as quantified above. Table 2 Forest revenue by Forest type 1995/96 1996/97 1997/98 1998/99 ValueTsh. % ValueTsh % ValueTsh. % ValueTsh. % Gov.plantations 386,685,332 31.8 777,171,338 44 558,707,684 26 681,445,550 26 CentralGov.FR 822,878,000 67.7 961,938,000 54 1,528,270,000 71 1,733,540,000 66 LocalGov.FR 6,079,370 0.5 26,202,000 1 73,974,800 3 228,385,850 9 Total 1,215,642,702 100 1,765,311,338 100 2,160,952,484 100 2,643,371,400 100 (Source: author, based on Salmi and Monela 2000) Royalties from timber sales are the most important source of income for the sector. Chart 5 below shows the various sources of revenue collection and their relative importance. Royalties hold a share of 83 percent equal to Tsh.4.5 bil- lion in nominal terms without the revenue collected and retained at source. Similarly, as Table 3 below outlines, in the previous financial year, Royalties constituted 87 percent of the total revenue. Registration fees for dealers of for- est products were the second largest source of revenue. Revenue from registra- tion fees has been increasing steadily over the last three years. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 17 Table 3 Revenue collection by source, 2002/03 ­ 2003/04 Revenue Collection by Source 2002/03 2003/04 Source Amount in Amount in 1000 Tsh. % 1000 Tsh. % Export Certificate and Permit 35,000 0.74 155,239 3.41 Royalties 4,106,410 87.13 3,762,014 82.69 Registration Fees 261,440 5.55 570,001 12.53 Olmotonyi FTI 65,600 1.39 Compounding Fees 118,339 2.51 39,761 0.87 Receipts from Honey and Bees wax 1,668 0.04 6,687 0.15 Tree Seed Program 123,500 2.62 Recovery of stores and public money 2 0.00 Miscellaneous Receipts 1,000 0.02 16,073 0.35 Total 4,712,959 100.00 4,549,775 100.00 (Source: MNRT (2004). Note: this does not include revenue collected and retained at source) Royalties have in the past been fixed arbitrarily with no recognition of market values (Salmi and Monela 2000). The 2002 Forest Act, provides the basis to determine the level of royalties based on market value of the produce, accessi- bility, profitability and principles of sustainable harvesting (Section 78, the 2002 Forest Act). Furthermore, the Act authorises villages to sell timber from their own forest reserves, which will provide a new and additional source of forest revenue, directly accruing to the communities. Chart 5 Distribution between sources of revenue, 2003/04 Receipts from Miscellaneous Compounding Honey and Bees Receipts Fees wax 0% 1% 0% Export Certificate and Permit 3% Registration Fees 13% Royalties 83% (Source: MNRT, 2004. Note: this does not include revenue collected and retained at source) Chart 6 shows that while total revenue collection and actual collections have increased over the last 2 years, the amount retained at the source has decreased. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 18 Chart 6 Revenue collected and retained at source, 2002/03 ­ 2003/04 7,000,000 6,000,000 5,000,000 Tsh. 4,000,000 2003/2004 2002/2003 '0003,000,000 2,000,000 1,000,000 0 Actual Collected and Total collections retained at source (Source: MNRT 2004) Constraints to revenue collection Various sources report revenue under-collection in the Forest sector during the past decades (URT 2004b, Salmi and Monela 2000). The NFP estimates that only 5 to 10 percent of the revenue due from forest reserves and general lands is collected (p. 23). There have been several studies assessing the forest sector revenue collection in Tanzania, notably Kowero (1990 and 1991), Silviconsult (1991), Chatterjee and Mushi (1994), Buys et al. (1996), Koppers (1998), and Mushi (1999). The authors seem to agree that the reasons for the low performance are the poor management of FBD and recommend the development of transparent and ac- countable forest management without corruption (Salmi and Monela 2000). Corruption is still a problem in the forest sector, accounting for large revenue loss and destruction of the forest estate. While forest staff carrying out revenue collection have low salaries (around Tsh.30,000 per month, data based on the year 2000), corruption does not al- ways originate within the forest sector itself. Under the local government de- centralization reform in Tanzania, the forest sector has been severely affected in the sense that all Forestry staff based in the districts is not answerable to FBD anymore but to the President's Office Regional Administration and Local Gov- ernment. The FBD is understaffed to monitor the management of the country's national forest reserves, while decentralized management is far from being fully operational throughout the country. While forest sector revenue is one of the most important sources of income to many district councils, very little has been invested by local governments in developing the sector to secure future income. The lack of capacity at the dis- trict level is well recognised and documented. Local governments may not be the right entities to carry out profitable business in primary or secondary For- estry production. The 2002 Forest Act provides for contracting out of opera- Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 19 tions to private sector, and different types of competitive concessions and leases may be appropriate means for local governments to continue earning revenue from their forests, while keeping their own staff adequately small and affordable. A constraint to competitive bidding is the current lack of capacities for tendering, procurements and monitoring and evaluation, which contributes to the high levels of under-spending of the foreign development budget in For- estry (see below). Public financing National Budget Allocations to the Forest Sector have generally been inade- quate and subject to fluctuation. During the 90s FBD received 0.1-0.2 percent of the annual budget of the GOT. Allocations in real terms have been increas- ing more than four-fold during the same period (Salmi and Monela 2000). While FBD's allocation was 50 percent of the total MNRT budget in 1996/7, it decreased to 18 percent in 1998. This was due to the introduction of the For- estry Retention Scheme. Forestry and Wildlife each received 29 percent of the Budget in the last finan- cial year. This was followed closely by Fisheries (18 percent) and Tourism (11 percent) (see Table 4 below). Table 4 Budget of MNRT, distribution by sub-sector, 2002/03 ­ 2003/04 2002/03 2003/04 Sub-sector '000 Tsh. % 000 Tsh. % Forestry and Beekeeping 4,897,656 24 7,633,912 29 Wildlife 6,593,025 33 7,586,736 29 Fisheries 3,688,280 18 4,648,202 18 Tourism 2,208,073 11 2,880,761 11 other Total MNRT 20,243,165 26,257,352 Note: The Total includes other sub-sectors not listed here. The amounts are only recurrent expendi- tures. Based on MNRT 2004. The retention scheme allows the MNRT to retain 70 percent of the revenue col- lected from Forestry. The remaining 30 percent is submitted to the Treasury. After deductions of 14% of the retained revenue for the Ministry's central ad- ministration and other divisions, the remaining 56 percent of the originally col- lected revenue is allocated to FBD. With this income, the FBD finances all its recurrent (except for staff salaries which are paid by Treasury) and some devel- opment expenditures, though donor financing covers most. Local Governments are allowed to retain for their own purposes 5 percent of the sum above Tsh.1 million of collected revenue. There are indications that that the forest sector suffers from under-financing, not only at the central level but also at the local administrative level. The ap- proved local government budgets are much lower than the requested budgets and actual disbursements still lower. According to Kaduvage (2000), the actual disbursements were only 27 percent of the requested budgets in Mwanza Re- gion during the past four financial years. In 1999/2000 about Tsh.0.5 billion of Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 20 retained revenue was returned to the regions and districts by FBD (Salmi and Monela 2000). Foreign financing Chart 7 below illustrates the high degree of donor financing in the development budget of FBD with foreign allocations to FBD's development constituting 99 percent in 2003/04. This reflects a long-term trend. During the past two dec- ades, donor agencies have been providing the bulk of forest sector financing with contributions around 90 percent (FBD statistics; Kobb 1999). However, it is reported by the Development Partner's Group that this situation will change dramatically in 2007 when most partners will cease their support. Secondly Chart 7 below demonstrates the high degree of underspending of the development budget in the Forestry sector. Only 10 percent of the total budget allocation was spent. This reflects low capacities to absorb foreign funding and institutional inefficiencies, often induced through uncoordinated policies of the development partners. To tackle this problem, the Government and its devel- opment partners have developed a sector wide approach for the Forest Sector, being implemented this year for the first time. Yet, a SWAP capacity building and human resource development plan is missing to cater for restructuring in the medium term. Chart 7 Development budget, 2003/04 Development Budget Comparison local - foreign contributions and underspending 16000000 14000000 12000000 10000000 tsh Estimates 8000000 '000 Actual Expenditure 6000000 4000000 2000000 0 Local Foreign Total (Source: Author based on MNRT 2004) Sector spending The comparison of recurrent and development expenditures over the last 3 years in Chart 8 shows that recurrent expenditures are much larger than devel- opment related expenditures and have been steadily increasing over the 3 year period. In contrast, development expenditures were negligible during 2001/02, but increased to 29 percent in 2002/03 and decreased again to 16.5 percent in 2003/04. The nominal values are shown in Table B in Annex 1. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 21 Chart 8 Recurrent versus development spending, 2001/02 ­ 2003/04 Recurrent versus Development spending 9,000,000 8,000,000 7,000,000 6,000,000 tsh 5,000,000 recurrent '0004,000,000 development 3,000,000 2,000,000 1,000,000 0 2001/02 2002/03 2003/04 (Source: Author based on MNRT 2004) Chart A and Chart B in Annex 1 show the allocation of the FBD budget in broad terms: The development budget is dominated by service delivery expen- diture (92 percent), while the recurrent budget is dominated by capacity build- ing expenditure (61 percent) (see also Chart D in Annex 1 for allocation in the past two years) Over a two-year timeframe, the amount spent on service delivery in the recur- rent budget has even gone down further vis-à-vis the capacity building compo- nent (see Chart C in Annex 1). Similarly the component of service delivery has increased in a 2-year time frame within the development budget (see). Based on data provided by the Public Expenditure Review for Environment (URT, 2004b), about 31.8 percent of the recurrent budget is spend by FBD on aforestation, 18 percent on administration and 8 percent on research. The bal- ance is allocated to Districts, training, beekeeping and other environmental ac- tivities (p.25). The PER does not detail for which financial year these figures apply. In 2003/04 the allocation by FBD to Districts was Tsh.265,955 (MNRT 2004). Degree of self-financing Although Forestry is a productive sector, it is presently not a net contributor to the treasury. Chart 9 below shows how, during the last two years, the revenue collected by FBD has been insufficient to cover its budget, although the financ- ing gap has decreased recently.7 7 The actual collections do not include revenue collected and retained at source. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 22 Chart 9 Degree of self-financing of FBD, 2002/03 ­ 2003/04 "Degree of Self-financing" - Actual revenue collection versus budget amount - compared for the last 2 years 6,000,000 5,000,000 4,000,000 3,000,000 tsh 2003/2004 2,000,000 '000 2002/03 1,000,000 0 -1,000,000 Actual Budget Under collections Collection -2,000,000 (Source: Author based on MNRT 2004) The 2000 Forestry Sector Financing study assesses that the forest sector could become self-financing with limited potential to provide net contribution to the government income, mainly through secondary and multiplier effects (Salmi and Monela 2000: 23). 2.4 Production and Consumption Although the new Forest Policy and Act promotes private investment in the forest industry, through incentives and credit facilities for private investments and joint ventures, local and foreign private investment in the sector has so far been limited. A state owned company, Tanzania Wood Industry Corporation (TWICO), was responsible for harvesting and sealing wood products up to the early 1990s. To date ten logging companies previously owned by TWICO have been privatised (TIC 2001). Southern Paper Mill (SPM), the only large scale pulp and paper factory in Tanzania, East and Central Africa, was closed in 1997 and has been advertised by TIC for privatization (TIC 2001). Harvesting of forest products is carried out in both natural and plantation for- ests by private companies, pit sawyer and small-scale companies.8 The Forestry industry is dominated by wood processing through sawmilling, furniture marts and joinery. There are also small-scale paper and board produc- tion, matches manufacturing, poles production, chipboard, fibreboard, black- board manufacturing and tannin extraction (MNRT 2001a: 27). 8 Currently the main large operators include, Sao Hill Timber Ltd., TANSCAN Timber Company, Escarpment Forestry Ltd, Kilitimber Ltd, Mena Wood Ltd, and United Lamber Ltd. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 23 Table 5 Industrial wood consumption, 1988, 1989 and early 1990s. Type Annual consumption in m3 1988 % 1998 % Early 1990s % annual aver- age Industrial saw- 533,000 92.5 944,000 95.6 382,500 51 milling Pitsawing - - 150,000 20 Wood based 14,000 2.5 13,000 1.4 22,500 3 products (1990s), panels and matches (1988, 1989) Pulp and paper 29,000 5 30,000 3 157,500 21 mills Joinery and furni- - - 37,500 5 ture industry Total 576,000 100 987,000 100 750,000 100 (Source: based on NFP: 27 and MNRT 2002, adjusted by author.) Table 5 above provides information on the industrial wood consumption in 1988, 1998 and the annual average in the early 1990s. The data source did not specify if consumption data refers to domestic production or if imports are in- cluded. The industrial consumption of wood is predominantly for saw milling, although there has been a decline from over 90 percent in 1988 and 1998 to around 50 percent in the early 1990s. According to Ngaga (1998) private sector holds a 78 percent share in saw milling through small-scale units (1,000m3/annum). Simi- larly, wood working units for furniture and joinery are mainly owned by indi- viduals and families. Productivity in small-scale saw milling is low due to the use of traditional methods, e.g. mobile sawmills, saw benches and handsaws (pitsawing). Pitsaw- ing organized by individuals and accounts for 40 percent of the country's sawn wood (Skage and Naess 1994). In addition, there are problems with transportation due to poor infrastructure. The Forest industry also suffers from lack of information on material availabil- ity, market statistics and skilled labour (MNRT 2001a: 27). Since the closure of SPM there has been an oversupply of plantation grown softwood from the Sao Hill plantations. The National Forest Programme esti- mates the sustainable long-term supply of saw logs from industrial plantations at 540,000-600,000m3 a year (URT 2001:27). The long-term supply of saw logs will decline unless new planting and re-planting schemes are initiated. The quality of logs is considered to be poor (same source). Based on the 2003 Eco- nomic Survey, the total harvest of logs increased quite significantly from 4,556 Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 24 m3 in 2002 to 334,873 m3 in 2003 (315,853 m3 in plantations and 19,020 m3 in natural forests).9 Due to declining stocks and increasing royalties, the harvesting of hardwoods from natural forests is going down. There is a declining availability of accept- able log sizes of hardwood species, in particular Mninga. Machine cut hard- wood is becoming rare. A number of hardwood sawmills have been closing down, which has resulted in increased pitsawing. Due to ineffective control and management of natural forests, harvesting is mostly carried out illegally. 2.5 Exports and Domestic Trade It is difficult to establish the contribution of the Forestry sector to exports as each source consulted present different figures. According to data in Table C in Annex 1 the value of export of Forestry products was around 4,5 million US$ in 2002/03. The data presented is showing some fluctuations, which are not ex- plained. Also, the narrative part of the Economic Survey describes that the total value of exports of forest and bee products increased between 2002 and 2003 by 20 percent from Tsh.6 to 7.2 billion, which differs from the data in the Ta- ble. A similar table contained in a FBD source (MNRT 2002), establishes the total value of exports of Forest Products for the three years 1998/99, 1999/00 and 2000/01 at US$4,9m, US$5,9m and US$2,1m respectively. Other sources (see FOSA Country Report for example) argue that the Forestry sector contributes 10 percent of official foreign exchange earnings or 11 percent of the total mer- chandise exports.10 Chart 10 Composition of Forestry exports by product, 2001/02 Composition of forestry exports 2001/02 Beeswax Honey Logs - Teak 21% 0% Paurosa 21% Tree seeds 2% Other products 0% Wood Timber carvings/sculpt 19% ures 8% Ebony Floor boards 26% 3% (Source: author, based on URT 2004c) Chart 10 above shows the composition of Forestry exports by product on the basis of the financial year 2001/02. 9 The reliability of these data is questionable. 10 The year of these values is not specified in the source. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 25 Some of the fluctuations in the data might be due to a changing policy envi- ronment, such as the liberalisation of trade. Also, exports of some individual forest products such as some types of logs, flooring and blackwood have de- creased whereas the value of honey exports has increased substantially. Exports of logs and unprocessed products of some valuable species have been banned in 2000 (Salmi and Monela 2000). Official Beekeeping Statistics of the FBD are presented in Charts 11 and 12 below, showing the trend of beeswax and honey exports in Tanzania between 1988 and 2004. Statistics of the FBD summarize that the average annual export of honey between 1993-2003 was 348.8 tons of honey and 187 tons of bees- wax, which generated an average of US$ 1,158,220.3 for honey and 211,393.2 for beeswax in export earnings (FBD 2003). Chart 11 Beeswax exports, 1988/89 - 2003/04 Beewax exports from Tanzania, 1988/89 - 2003/04 900 3000 800 700 2500 tons 600 2000 500 ´000 400 1500 Metric300 1000 200 USD 100 500 0 0 /89 /90 /91 /92 /93 /94 /95 /96 /97 /98 /99 2000 2001 2002 2003 2004 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 19981999/2000/ 2001/2002/ 2003/ Metric tons Value in USD ´000 (Source: author based on FBD data) Presently, Tanzania produces about 4,860 tons of honey and more than 750 tons of beeswax per year. According to the National Beekeeping Programme 2001-2010 (p. vii) the current production is only about 3.5 percent of the coun- try's potential production from honeybees, which is estimated to reach Tsh.138 billion annually. The sudden increase of honey exports in 2002/03 is due to an EU export ban on honey imports from China, which led to an increase in world prices and opened an opportunity for Tanzania. This ban has been lifted and prices seem to have fallen back.11 11 Comments received by A. Cauldwell, DPG. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 26 Chart 12 Honey exports, 1988/89 ­ 2003/04 Honey exports from Tanzania, 1988/89 - 2003/04 900 1400 800 1200 700 1000 tons 600 500 800 '000 400 600 Metric300 USD 200 400 100 200 0 0 89 90 91 92 93 94 95 96 97 98 99 2000 2001 2002 2003 2004 1988/ 1989/ 1990/ 1991/ 1992/ 1993/ 1994/ 1995/ 1996/ 1997/ 1998/1999/2000/ 2001/2002/ 2003/ Metric tons Value in USD '000 (Source: author based on FBD data) Salmi and Monela (2000) estimate that the farmer income from honey and other forest produce from miombo woodlands is as much as Tsh.1,050 per ha. Based on a study conducted by Aku et al (2000:5), between 1990 and 1998 76m litres of honey and 780 tons of bees wax were extracted in Kilimanjaro Region. A total of Tsh.333,746,000 was accrued from sales of honey. Issango (2001) has done similar study on honey production and sales along the Dar es Salaam ­ Morogoro Highway. He found that between 1995 and 2000 a total of Tsh.584 million was accrued from selling 751 tons of honey and Tsh.2.5 mil- lion from selling 2 tons of bees wax. The new Forest Policy and the National Beekeeping Programme foresee the establishment of village beekeeping reserves. In 2003 villages in five districts have established such reserves, from which they will accrue direct cash income (unpublished FBD URT, 2004c). 12 The export of gums and resins, another non-timber forest product besides honey and beeswax, amounted to more than 7,000 metric tonnes in the period 1997-2003, corresponding to an average of 1,011 tonnes a year (Kagya 2004). See Chart 13 below for recent export figures. Industrial use of gum arabic is as an additive in a variety of foods, such as confectionary and beverages, pharma- ceutical, cosmetics, lithographic inks, paints and dye, adhesive in postage stamps. In addition, there are a number of local traditional uses, which include nutritional and medicinal purposes. 12 These are Kibondo, Tabora, Manyoni, Kondoa and Handeni districts. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 27 Chart 13 Trend in gum and resins export and revenue, 1997 - 2003. Exports figures for gums and resins 1,600,000 800,000 1,400,000 700,000 1,200,000 600,000 1,000,000 500,000 Weight 800,000 400,000 US$ Net 600,000 300,000 Kg 400,000 200,000 200,000 100,000 0 0 1997 1998 1999 2000 2001 2002 2003 Amount in Kg (Net Weight) Amount in US $ (Source: Tanzania Revenue Authority, Customs Statistics Section) According to the export statistics, the price offered by the world market for gum coming from Tanzania is low compared to other countries. The reason is, according to Kagya (2004), poor quality of gum resulting from mixing gums from different sources and exporting without grading. Furthermore, exporters may be paying a low price so that they can evade taxes. There is no local indus- try consuming gum, and therefore 99 percent of the collected gum is exported. With regard to the industrial use of NTFPs, common problems are lack of local value-added processing facilities. Hence products are exported in raw form at low prices. Also, marketing is usually weak. 2.6 Employment According to the Integrated Labour Force Survey 2000/01 Agriculture, Forestry and Fisheries combined provide 82.1 percent (13.8 million people) of employ- ment in Tanzania. Additional employment in the informal sector related to Forestry is most also likely to be significant. The Labour Force Survey accounts 0.4 percent of total informal employment to agricultural and forest services and 6.7 percent to fur- niture making and manufacturing of non-metallic mineral products. The National Forest Programme estimates that the sector employs about 3 per- cent of paid labour and an even bigger proportion of people in informal For- estry related sector activities (NFP). Salmi and Monela (2000) estimate that the sector provides 730,000 person years of employment. 2.7 Unaccounted Forest Services and Externalities In addition to the contributions of the Forest Sector that are accounted for in the previous chapters, forests provide goods and services that are not accounted for. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 28 These are the provision of energy sources; protection of watersheds; store of carbon dioxide; micro-climatic functions; provision of habitat to fauna and flora of high biodiversity values, recreational/tourism and cultural/religious values. These services are not readily calculable as they do not have directly associated cash flows. Over the last two decades a wide body of international literature has developed attempting to value various aspects of tropical forests and the goods and ser- vices they are supplying. Various methods have been used to place monetary value of forest services within environmental economics to capture spillover effects and intangibles not included in private economic valuations. These are for example, travel cost methods, contingent valuation, hedonic pricing etc. In Tanzania, most attempts to value these unrecorded forest goods and ecosystem services originate from site-based case studies. The few exceptions have been named in the introduction. The main products and services Tanzania's forests provide to the economy are as follows: Energy supply Forests contribute to the energy sector in two major ways. First, as a direct source for biomass energy, namely fuelwood and charcoal. Second, more indi- rectly, via water catchment and storage for the hydropower sector (discussed in the subsequent paragraph). All these forest inputs are basically provided as 'free goods' to the national economy and population. Biomass energy or woodfuel, which comprises fuelwood and charcoal from both natural forests and plantations, accounts for 93 percent of total energy consumption in the country. The estimated per capita consumption of woodfuel is 1m3 roundwood per an- num. This amounts to more than 30 million m3 per year. If valued with 3000 shilling per m3 (the November 2000 royalty rate) this amounts to 90 billion shillings per year (Salmi and Monela 2000). The NFP (MNRT 2001) estimates that 95 percent of Tanzania's total wood consumption was consumed by wood fuel in 1999 (40.4 million m3). Out of this 26 million m3 was consumed in the rural areas as fuelwood and 13.4 million m3 in urban areas, mainly as charcoal. A large number of rural industries rely on the use of woodfuel in their produc- tion processes. These are in order of priority, tobacco curing, fish smoking, salt production, brick making, bread baking, tea drying, pottery, lime production and processing of beeswax. Additional large sinks for fuelwood in rural areas are beer brewing and alcohol distillation. Mostly, these are non-farm activities and one needs to keep in mind their 'environmental cost' when promoting in- crease of non-farm activities in rural areas as a means of economic growth. Surprisingly, the NFP predicts that the annual woodfuel consumption is not ex- pected to increase, but to remain at 40 million m3, the level of 1999 (MNRT 2001a: 28). This estimate does not seem to take population growth into ac- count. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 29 Table 6 Inputs of forests to Tobacco production Tobacco Production and impact on forests, Tabora Region Year/Season Tobacco Estimated fuelwood Estimated area production in consumption m3 cleared 1000 ha tonnes 1985/86 9,673 173,529 4338.2 1986/87 6,614 118,641 2966.0 1987/88 6,176 110,803 2770.1 1988/89 5,641 101,186 2259.7 1989/90 4,633 83,112 2077.8 1990/91 5,961 106,947 2673.7 1991/92 12,123 217,485 5537.1 1992/93 14,443 259,114 6477.9 1993/94 11,499 206,301 5157.5 1994/95 17,133 307,371 7664.3 (Source: Mkanta and Chimtembo, 2002) A study on Tobacco Production in Tabora Region conducted in 2001 revealed that forests deliver significant input to tobacco production. The area cleared for tobacco growing has for example been increasing yearly in the period 1985/86 ­ 1994/95 (see Table 6 above). The National Forest Programme (MNRT, 2001a) estimates that around 27 per- cent of residents in Dar es Salaam use firewood, while over 86 percent use charcoal as a source of energy.13 Charcoal provides significant inputs both to rural and urban economies. It pro- vides employment and income to the rural producer and a source of energy for cooking to the rural and urban consumer, as well as a source of government revenue. It is difficult to quantify this contribution as much of the charcoal trade is undertaken illegally and only a fraction is captured through the official road blocks. Malimbwi et al. (2000) found that about 6,000 bags of charcoal are transported daily to Dar es Salaam. Other estimates are much higher, between 15,000 and 20,000 bags every 24 hours (Norconsult 2002). Dar es Salaam's charcoal intake converts 2 million tonnes of wood per year and presents about 50 percent of the urban demand (same source). In a study conducted by Lusambo in 2002 in Kilosa District, charcoal contrib- uted Tsh.175,765 to household annual income. Noah (2002) found that house- holds realize an income of Tsh.50,000-62,000 monthly from sale of charcoal. Malimbwi et al. (2000) likewise estimate that charcoal production provided 38 percent of the household income in six villages surveyed (converted into US$ 445 in cash). 13 The NFP does not explain why the total exceeds 100 percent, but it is assumed that this is due to the fact that some households use both energy sources at the same time. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 30 However, the environmental trade-offs of this economic contribution are high. A study "true cost of charcoal" (2002) provides data on charcoal consumption in the country and the associated environmental cost. Some of the key figures are presented in Box 2 below. Box 2 Facts and figures on charcoal production Facts and Figures on Charcoal Production · 15,000 ­ 20,000 bags of charcoal enter Dar es Salaam every 24 hours, 365 days a year. Equal amount enter other major Tanzanian cities combined. This adds up to about 1 mil- lion tonnes of charcoal per year. · To produce this charcoal tree had to be cut from 3320 km2 of forest, most of which will not regenerate as it is converted to other land uses. · Charcoal is subsidised by ineffective collection of dues, while liquefied petroleum gas is penalized by higher import duties. Due to incorrect pricing charcoal users are externalis- ing the negative impacts of their fuel choice on the environment. Government policies would need to correct this market failure through fiscal policies. · Forests are declining by 11.5 percent annually, 99, percent of which is for fuelwood and charcoal. · Burning charcoal provides high volumes of carbon dioxide, carbon monoxide and CH4. It adds to the load of carbon dioxide, while at the same time reducing carbon sinks. · The efficiency of charcoal consumption is very low. Charcoal stoves has an efficiency of 20-35 percent energy conversion compared to for example 45-65 percent for LPG. (Source: Based on Norconsult 2002) To the extent that charcoal production competes with other forest uses, it re- duces growth potential in the forest sector as it leads to a decline of the total stock. In many parts of Tanzania, e.g. Dar es Salaam, Morogoro, and Iringa, the sustainable yields seem to have been exceeded and thus the country's natural capital is sacrificed for charcoal production. Charcoal is hence being produced at a cost to society in terms of its present and future availability to meet wood biomass needs and wood needs for other purposes. A projection of the impact of charcoal production on the country's forests and woodlands, is provided by Norconsult (2002) and re-produced here in Table 7 below. It illustrates a possible scenario of present trends of charcoal production were to continue. Table 7 Projection of charcoal production impact on forests, 2002-20 urban charcoal woodland area needed woodland woodland surplus consumption in for sustainable remaining in km2 or deficit in km2 Year tonnes production, in km2 2002 926,000.00 98511 300000 201489 2005 1,071,961.00 114038 218700 104662 2010 1,368,124.00 145545 129140 -16405 2015 1,746,111.00 185757 76256 -109501 2020 22,285,529.00 237078 45028 -192050 (Source: Norconsult 2002:18) Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 31 Water catchment, storage and filtration Forest reserves known as catchment forests, occupy only 2 percent of the coun- try's land area. However, besides wetlands, they provide important watershed area for rural and urban water supply in Tanzania. The largest catchment forests are found on the volcanic and crystalline mountains of Tanzania and in lowland forests, such as Rau in Moshi, Minziro in Kagera Region and Kimboza in Morogoro (MNRT 2001a: 26). The interaction between forests and water quality and supply is complex but it is understood that forest loss reduces watershed moderation. Deep soil moisture storage and rate of aquifer recharge are usually enhanced through reduction in evapotranspiration. Woodland clearing may also increase overland runoff, ero- sion risk, flash floods in bottom areas and reservoir siltation. Many formerly perennial rivers in Tanzania now desiccate for several months per year, a phe- nomenon more likely to be caused by deforestation of the catchment than by climate change. Hydropower is besides petroleum and coal the major source of commercial en- ergy supply in the country. Hydroelectric energy has a potential of installed ca- pacity of 4.7 GW of which only about 10 percent is developed (TIC 2001). The inputs of forests into hydropower generation are indirect, through their ecosys- tem service of providing water catchment and storage as described above. An attempt to provide a monetary value for the role of Forestry to water supply provided by Norconsult (2002) is based on the cost of water and summarized in Box 3 below. Box 3 Monetarizing the contribution of forests to water supply Monetarizing the contribution of catchment forests to water supply · Based on existing studies a working figure of US$ 100 as average cost of water for domestic use per urban household per year is used to derive a conservative estimate of US$ 100 million current annual value of Tanzanian urban water supplies. · With reduction of water availability per capita on the basis of population growth, the cost of maintaining the same volume of urban water supply to each household as now is likely to double at least by 2020. · The predicted reduction of woodland coverage by 2020 would impair the water catchment (inter- ception, retention and slow release) function by an estimated 50 percent, thereby quadrupling the cost of water in 2020. Impairment of the water catchment function will therefore impose a cost rising to US$ 300 million per year by 2020. An estimated 10% of this cost (US$ 30 million per year) is attributed to charcoal production. · Engineering replacements of the lost natural storage of water in catchment basins could augment the cost to absurd figures, 1000 times higher at least, but such infrastructure would not be afford- able. · Tanzanian urban dwellers will pay more for less water and will forgo the development and health benefits of an adequate water supply (based on Norconsult 2002, p. 26-27) Monela et al. (2003) estimated the total economic value of Tanzania's Catch- ment Reserves, which is comprised of Use Values and Non-use values, at US$ Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 32 620.4 million.14 Box 4 below summarises some facts on the contribution of the Eastern Arc Forests to water and energy supply in the country. Box 4 Contribution of Eastern Arc Forests to water and energy supply · Major Rivers are flowing from the Eastern Arc Mountains, e.g. Pangani, Sigi, Wami, Ruvu, Parts of Rufiji and part of Greats Ruaha as well as smaller tributary streams. · At least 10 percent of the total population of Tanzania gets their water from these rivers. · Seven major hydropower plants using Eastern Arc Water are Kihansi, Kidatu, Part of Mtera and part of Pangani Falls. · Hydropower is 62 percent of the total energy supply of Tanzania (559 MW of 892 MW) in large part derived from water flowing from the Eastern Arc Mountains · In total at least 50 percent of Tanzanians' electricity is coming from water flowing from the Eastern Arc Mountain Forests. · Water from the Eastern Arc Mountains feeds large irrigation schemes, such as Liovu Sugar Company and various rice schemes. · The forested mountains maintain a suitable microclimate for growing vegetables, spices and fruits. Major export from mountains to Dar es Salaam and other cities of: banana, potato, peas, cardamom, cinnamon, pears, apples, plums, leeks, tomatoes. Provides boost to local peoples livelihoods. · Eco-tourism possibilities exist but are largely under explored. (Source: Eastern Arc Conservation Strategy Component on TFCMP) Soil conservation Tanzanian Agriculture depends largely on soil fertility, as the application of fertilisers is too expensive for most farmers. Tree loss on and around farmland premises deprives the soil of leaf-fall, thereby reducing fertility and accelerates soil erosion by wind and water. Rainfall runs off soil hardened by exposure much faster than before, removing humus and animal droppings, and carrying away the most fertile top layer of soil. The value of crops grown in what is ef- fectively sub-soil, after serious erosion, is greatly reduced, impacting nega- tively on agricultural productivity. Norconsult (2002:27) estimates a 50 percent reduction in output, equivalent to 25 percent of GDP. However, no timeframe or basis for this calculation is provided. Estimates for the East African Region are that typical erosion rates are 0 in ma- ture woodland, 5 tons per ha per year in wooded savannah, 30 tonnes/ha/y in maize fields, and 50-100+ tons/ha/y in degraded land. Soil formation occurs at about 1 ton/ha/y in the region, however rapidly eroded soil is not replaced within a human generation. (Norconsult 2002). Estimates from the West Usambara Mountains in Tanzania are that the land value under agroforestry was raised by 126.6 Euro per ha, using a time horizon of 20 years one would arrive at an increased land value of 6.3 Euro/ha/year (undiscounted and assuming equal value for all 79,000 ha of land converted).15 14 Use values include direct, indirect use values and option values. Non-use values include existence values and bequest values. 15 Study "Ten million Trees Later". Comments received by A. Boerhinger, DPG. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 33 Provision of habitats and conservation of biodiversity Tanzania is among the four countries in the world, classified as `megadiversity' nations,16 due to the variety of habitats found in the country. The biological diversity has been described in a number of studies and academic papers, in- cluding the Tanzania Country Study on Biological Diversity (1998). In the context of the discussion of economic growth, biological diversity of for- est resources is of relevance, as it is a so far uncaptured economic potential. Apart from global existence values, biodiversity is of national value for me- dicinal use. Other values associated with biodiversity are pharmaceutical and genetic resources. Ngaga et al (2003:56) shows that some local communities in Arusha Region ranked medicines as the third most important good or service derived from the forest reserve after `water for consumption' and `firewood' (between 71.4 and 57.1 percent in the surveyed villages responded like this). People in Nkoanekoli village send the medicines to Arusha town for sale and hence derive a direct income. These preferences and dependencies of communities on the forests will vary according to the local conditions. Attempts to monetarize biodiversity values contained in Tanzanian forests have been made by a number of studies. Norconsult (2002:27) estimate that forest products are worth at least US$ 100 per rural dweller per year in nutritional and medicinal value. According to conservationists, Tanzania's biodiversity is under severe threat. For example, the country's coastal forests are remnants of some of the world's oldest forests. Burgess et al (1992) describe the biodiversity potential hosted by these forests. Collectively they support many rare and poorly known plant spe- cies, including around 50 believed to be endemic to a single forest, seven bird species and subspecies of global conservation significance. They also contain several rare mammals, reptiles and amphibians, and an invertebrate fauna with many rare species. All Tanzanian coastal forests are currently being destroyed at a rate leading to complete removal of forest cover and biodiversity loss, fol- lowing a sequence of a) logging for timber; b) pole-cutting for building, c) wholesale burning for charcoal; and d) wholesale conversion into farm land. Non-timber forest products Direct income can also be achieved from forests through the sale of other forest products such as fruit, nuts, rubber, meat, honey, oils etc., summarized under the term non-timber forest products (NTFPs). NTFPs have a large potential to contribute to the local economy, only some of them ­ as far as they are traded through the formal sector ­ are recorded in the GDP. Some figures for gums and resins were for example provided in Section 2.4 above. To date, FBD does not address issues related to non-wood forest products apart from beekeeping. Hence, data to assess supply and demand as well as income 16 Other countries are Republic of Congo, Brazil, and Indonesia. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 34 generation potential of NTFS is not available on a national scale through offi- cial statistics, but only through individual case studies conducted by independ- ent or government commissioned research. Monela et al. (2000) estimate that honey, charcoal, fuelwood, and wild fruits contribute 58 percent of the cash incomes of farmers in six villages surveyed. Honey alone accounted for one third of all cash income in these villages. Ac- cording to the surveyed farmers, agriculture has become less profitable, thus inducing them to find other means to earn a living, e.g. collecting and selling forest produce. Infrastructure improvements have made it easier for them to bring their forest products to the markets of sale. Table 8 below illustrates that, among the income sources derived from the CFRss, NTFPs contribute the largest share followed by fuelwood. The contribution of forests to food security is indirect via soil quality services provided as an input to Agricultural production. A direct contribution to food security is through NTFPs that have nutritional value such as mushrooms or gum (used by pastoralists and hunter gatherers). Table 8 Sources of income in villages, Tanzania, 2002 Sources Lushoto Kilimanjaro Meru Sample District District District averages Crop production 157,704 138,010 360,000 218,571 Livestock 38,038 417,061 590,000 348,366 Off-farm 296,304 718,233 420,000 478,179 Env. incomes reserves 72,186 71,266 123,261 88,904 -NTFP 45,036 21,427 62,708 43,057 -Timber 1,200 2,100 504 1,268 -Poles 106 13,581 4,374 6,020 -Fuelwood 21,168 33,345 49,600 34,704 - Whithies 4,676 813 6,075 3,855 Sum 564,232 1,344,570 1,493,261 1,134,021 (Source: Field data , 2002 and Ngaga et al., 2003). According to Kagya (2002), 94 percent of households surveyed in a study con- ducted in Shinyanga utilize wild vegetables as relish. Theostina (2003) demon- strates that 51 percent of residents in Morogoro use money accrued from sales of NTFPs to purchase food in times of food shortage. Forests hence also have a 'safety net' function to poor people. In addition to NTFPs forests provide habi- tat to small mammals, hunted for protein supply, in particular in poorer vil- lages, such as for example in Linzi, Uluguru South, and Morogoro Rural Dis- trict (same source). Several studies evaluate the contribution of agroforestry to rural livelihoods (for example Idda 2003, Mwanahija 2003, Eustack 2003) and there seems to be consensus that agroforestry provides increased income compared to households not practicing agroforestry, and provides an important source of food in time of food shortage. For farmers, forests can also have positive value by reducing Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 35 risks of total crop failures in annual crop production (e.g. due to droughts or floods). Farmers often see trees as an insurance against unexpected cash needs, or as capital banks. 2.8 Future Growth and Recommendations So far Forestry has not been a very important part of the Tanzanian rural farm economy and the national economy at large. With long production cycles For- estry can hardly compete in profitability with agriculture or animal husbandry. At national scale GDP contribution is far below other sectors. However, there are number of competitive advantages associated with the For- estry sector: labour input needs which are low; timing of labour input which can take place outside the agricultural produc- tion season; land and soil requirements which are generally much more modest for trees than for other agricultural crops; and the combination of right tree species with agricultural crops (agroforestry systems) can increase the combined yields. The private sector, both national and foreign, provides the largest potential for future economic growth of the Forestry sector. This prognosis includes not only wood exploitation and timber and non-timber forest product processing and marketing, but also aforestation and sustainable forest management, and vari- ous other forest-based economic activities such as tourism (Salmi and Monela 2000). Also, local communities and individual farmers represent a valuable re- source which under right incentives can be mobilized to grow tress on a large scale and cost-efficiently. The domestic private sector may have a considerable potential for Forestry, if the structural barriers for profitable Forestry are removed.17 Pre-conditions for increasing private investment in the sector, include clarification and security of tenure rights to farmers, good road infrastructure and market development. However until corruption and policy and market failure are effectively tackled, it is unrealistic to expect major private capital flows into the sector. The bottlenecks for entry of foreign companies into the Forestry business are similar to other sectors, as identified in the CEM. In addition, the Forestry sec- tor has high transaction cost due to permit and license acquisition and limited knowledge of foreign investors of the local environment. The 2000 Study on Financing the Forestry Sector assessed that among others Malaysian and Chi- nese companies are among the likely investors due to the logging bans in China. As the 2004 logging scandal in Rufiji revealed, this prognosis came true as many of the illegally felled and exported logs were shipped off to China. The 17 Timber-traders, NTFP product producers, agro-industry, energy sector, tourism industry etc. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 36 value of the illegally harvested logs was valued at Tsh.382.65m (DPG Policy Brief: The Forestry Sector in Tanzania). Salmi and Monela (2000) suggest possible ways to increase revenue gained through the Forestry sector18. They are summarized in Box 5 below. Box 5 Ways to increase revenue from the Forestry sector 1. Reduce revenue loss · Resolving of competing revenue claims between central and local government. · Administrative strengthening of the revenue collection section in FBD. · Increase in transparency and accountability; eliminate corruption. · Improved reporting and monitoring of revenue. · Simplify licences and improve control check points, stock registers and transit passes 2. Increase revenue · Introduction of new revenue sources, i.e. watershed management fees from hy- dropower stations, eco-tourism fees, sale of carbon sequestion credits (CDM of Kyoto Protocol), sale of genetic resources etc. · Better enforcement of collection of royalites and fees from wood using indus- tries and inclusion of exempted industries such a tobacco, fishing, army etc. · Introduction of taxes for wood lot and plantation owners, in particular an in- come tax based on timber sales, as well as a property tax based on average pro- ductive capacity of different land categories. In addition, research could support the use of timber from lesser-used species. · Improvement of Forest Produce Pricing System through market-base pricing of forest produce; public auctions or tendering for timber lots; cheaper royalties to lesser-known species. 3. Increase domestic and foreign private sector investment · Reduction of bureaucracy in licensing system. · Clear investment guidelines. · Clearly defined ownership of all forestland. · Improving infrastructure. · Tax incentives; credit facilities, and technology transfer. (Source: Salmi & Monela, 2000) The Tanzania Investment Center makes the following observations about the economic opportunities in the Forest Sector (TIC 2001): · Domestic and foreign demand for wood products exceeds supply. · Potential for manufacturing of plywood and abundant supply of plantation softwood, such as Cyprus and Pines. · Low capacity utilization of the sector, despite great forest potential. · Potential for non-wood forest products, e.g. tourism, game, bee which are undeveloped. 18 The study also describes the optimising of foreign assistance to the forest sector, which is not considered a `genuine' revenue source in this report. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 37 · Establishment and management of bee reserves by private operators. The estimated potential of bee products is about 138,000 tons of honey and 9,200 tons of beeswax per annum from an estimated potential of 9.2m hon- eybee colonies. This prognosis does not match with the bee export data provided by FBD (see Section 2.5). · Opportunities for carbon sequestion projects. So far one Tanzanian com- pany, Kilombero Forestry Ltd. is among the three companies in the world certified in November 2000 to engage in carbon trading with greenhouse gas pollutant companies in developed countries. In order to capitalize the growth potential of Forestry now and in the future, this Chapter has emphasized the importance of policy measures that ensure management on sustained-yield basis, as well as correct pricing of traded eco- nomic goods and services that rely on Forestry resources and services. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 38 3 Wildlife This Chapter examines Wildlife's role and importance as a source of economic growth and rural development and whether its potential could be developed more fully. It focuses particularly on the wildlife hunting industry, drawing on available, yet limited sources. As such, it cannot be conclusive. The non-consumptive utilization of wildlife resources through trophy hunting industry and game viewing (`safari' tourism) are the two most important sources of wildlife related income in the Tanzanian economy. Game viewing tourism is discussed in the Tourism chapter in Part 3 of this report. Other wildlife-based activities are, with the exception of capture and trade of live animals, under-developed in Tanzania. These include trade in bushmeat, skins and other products such as biltong, and farming of specific wildlife spe- cies. These products could offer potential growth opportunities, but markets, production facilities and the necessary policy framework do presently not sup- port expansion of these activities. As this Chapter will illustrate, there are concerns about the governance regime regulating the hunting industry, which leads to losses in terms of growth and poverty reduction. Due to a lack of transparency and competition, wildlife re- sources are priced below the true market value, resulting in revenue losses and unsustainable exploitation. While there are a few powerful winners through this situation, the large majority of the population is loosing out. The hunting industry is one of the few non-farm industries with potential for economic development in remote rural areas of Tanzania. However, at present rural communities only see minimal benefit from an industry that operates on their land. Various sources quote, that there is resistance to reform of the indus- try both within the Wildlife Division (WD) and the private sector. Vested inter- ests between the Government and foreign private sector, lead to sub-optimal decision making that deprive the rural population of economic potential for growth and wealth. As in Forestry, studies quantifying the contribution of wildlife to the economy are scarce. Baldus and Cauldwell (2004) provide the first and only comprehen- sive empirical study with data for tourist hunting in Tanzania. This study is confidential and copies have been presented to the WD. Box 6 below summa- rizes some estimates on the economic contribution of wildlife. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 39 Box 6 Key estimates on the economic contribution of wildlife: Total income of the Wildlife Division generated from tourist hunting is about US$ 30 million p.a., based on 20,500 hunting days sold to 1,370 clients. The leasing companies generate about an additional US$ 9 million (Baldus and Cauldwell 2004); Gross value of wildlife (other than timber and fisheries) to the Tanzanian economy is estimated to US$ 128.50 million per year (IUCN 1989); Illegal wildlife hunting for wild meat is estimates to some US$ 50 million annually (IUCN 1989); Well over two-thirds of people eat wild meat, with up to 95 per cent of the rural popula- tion claiming it is their most important meat protein source (Barnett 200o); and During 1989 to 1999 at least 1.68 million birds, 521,000 reptiles, 148,000 amphibians and 12,000 mammals were exported from Tanzania (Roe 2002). (Source: Baldus and Cauldwell, 2004; IUCB, 1989; Barnett, 2000; and Roe, 2002) 3.1 Policy Framework The policy framework in Tanzania focuses mainly on wildlife conservation and not on utilization. The tradition of wildlife conservation in Tanzania dates back to the colonial era and the principle was manifested by former President Nyer- ere in the 1967 Arusha Declaration. The National Tourism Policy stresses wildlife conservation, due to the fact that Tanzania's tourism is largely wildlife based. It states that "... the government vows to ... improve and implement wildlife conservation regulations, and to protect other tourist attractions for the benefits of present and future genera- tions". Administration of wildlife resources in Tanzania falls within the Ministry of Natural Resources and Tourism. The two lead bodies under the Ministry are Tanzania National Parks (TANAPA) and the Wildlife Division (WD). Other institutions and parastatals include the Ngorongoro Conservation Authority, the Tanzania Wildlife Research Institute (TAWIRI), the Tanzania Wildlife Corpo- ration (TAWICO) and the College of Wildlife Management at Mweka. In addi- tion there is also administration through the district and regional administrative structures. The national legislation guiding the management of protected areas in Tanzania are the Wildlife Policy (1998) and the Wildlife Conservation Act (1974). The Wildlife Conservation Act is currently under revision. The WD is responsible for wildlife conservation in Tanzania in game reserves, game controlled areas, open areas and district game reserves. The two main functions of the division are regulation and co-ordination. Regulatory functions include safari quotas and other consumptive use of wildlife, licensing, prosecu- tion of offenders against the wildlife Act, gazettement of wildlife areas and su- pervision of photographic tourism in their areas. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 40 TANAPA is responsible for the management, conservation and use of all na- tional parks in Tanzania. TANAPA is a parastatal and corporate body under a Board of Trustees. They provide guide services, community conservation ac- tivities and anti-poaching in the national parks. 3.2 GDP Contribution The 1998 Wildlife Policy lays down a vision for the next 25 years, which is to raise the contribution of the Wildlife sector to GDP from about 2% to 5% and to contribute to poverty reduction and improvement of peoples' quality of life. Deriving the contribution of Wildlife resources to the GDP is not straightfor- ward. Within the System of National Accounts "hunting" combined with For- estry under the Agricultural GDP, is the closest proxy to measure the economic contribution of Wildlife resources. As outlined in Section 2.1 in Chapter 2 above, the average contribution of forestry and hunting combined to overall GDP has been 1 percent over the last 15 years and the average growth rate of the sub-sector has been 3.6 percent during the same period. Based on these figures, one might derive the misleading conclusion that Wild- life, is not of major importance to the national economy. However, similar to Forestry, Wildlife's true value is not reflected in the GDP. First of all, "hunting" is not the only way in which Wildlife contributes to growth. The Parks and Reserves attract high numbers of visitors each year for game viewing and photo tourism, all of which are non-consumptive activities that contribute significantly to the Tourism sector. In fact, the largest wildlife based income source is photographic tourism, which is not measured under the "hunting GDP", but under "Trade, Hotels and Restaurants", the closest proxy for tourism in the GDP. Secondly, even for hunting, the value recorded in the official statistics is largely under-representing the true income earned and the true value of the resource. The reasons for this are explained in this Chapter. The primary measure of growth of the wildlife sector is the total income gener- ated. This figure would need to be compiled of revenue from non-consumptive photo tourism, which provides revenue to TANAPA, primarily through park fees paid by international tourists and also, of revenue from tourist hunting and export of live animals, which accrues to the Wildlife Division. Tanzania is experiencing some of the fastest tourism growth in the world. The international and local wildlife tourism industry makes considerable contribu- tions to GDP, foreign exchange earnings, employment and to some extent the creation of related local business opportunities. The employment and local spin-off effects of tourism are discussed in part 3 of this study. Tanzania's considerable wildlife assets combined with the anticipated growth of tourism, will provide the country with a long-term comparative advantage in wildlife-based tourism, provided the sector is well managed and cost can be kept down. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 41 3.3 Wildlife Resources Tanzania is endowed with attractive national parks occupying approximately 47,000 square kilometres (see Table 9 ). Table 9 Areas occupied by National Parks in Tanzania. National Park Area (sq.km) Ngorongoro 8,320 Kilimanjaro 750 Serengeti 14,750 Manyara 325 Tarangire 2,600 Katavi 2,252 Udzungwa 1,000 Mahale Mt. national Park 400 Rubondo 460 Arusha 117 Mikumi 3,230 Ruaha 13,000 Gombe 52 Total 47,256 (Source: MNRT) Protected areas in Tanzania are habitat for a large variety of mammals, includ- ing the 'big five' (lion, elephant, leopard, rhino, cheetah), which attract tourists for game viewing as well as other species attracting national and international hunting tourism. Data on the abundance of these mammals differs. Most animals migrate and aerial surveys and ground surveys combined are necessary to attain reliable fig- ures for ecological monitoring. Estimates originate from wildlife conservation and research projects, however, regular and reliable ecological monitoring of wildlife resources on a national scale is not undertaken. A rough indication of the value placed by Tanzania on its wildlife resources is the value of trophy fees presented in Table 10 below. The hunter pays the tro- phy fee to the Government after killing a respective animal. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 42 Table 10 Value of Tanzania's Wildlife Resources in regional comparison Trophy Fees (US$) Animal Tanzania Zimbabwe Botswana Zambia Elephant 4,000 9,000 19,000 Lion 2,000 3,420 5,500 2,750 Leopard 2,000 2,000 4,100 1,750 Buffalo 720 1,680 2,500 1,000 Zebra 590 680 1,000 600 Sable 1,200 2,000 2,700 Hartebeest 370 1,000 650 Waterbuck 440 1,200 1,000 Wildebeest 320 600 950 650 Impala 240 120 400 100 Warthog 320 175 300 300 (Source: Baldus and Cauldwell 2004) Compared to Zimbabwe, Botswana and Zambia, Tanzania is offering its wild- life resources to the hunting industry at a low price. As a result buffaloes are heavily hunted in Tanzania. Income of the WD is heavily dependent on buffalo hunting and raising the fee would potentially have a large impact on the indus- try. Fee structures, restrictions and quota are a topic of on-going debate be- tween the WD and the hunting industry (Baldus and Cauldwell 2004). In the context of economic growth a crucial distinction needs to be made be- tween the consumptive utilization of wildlife resources, in Tanzania presently predominantly by the tourist hunting industry and the non-consumptive utiliza- tion through game viewing and photographic tourism. The monetary contribu- tion of the former is discussed in this Chapter, whereas the latter is discussed in the Tourism Chapter in Part 3 of this report. 3.4 Contribution of Consumptive Wildlife Resources Use Land areas that offer consumptive wildlife use, include game reserves occupy- ing approximately 95,000 km2 and Game Controlled Areas, comprising 58,565 km2 (Mabugu and Mugoya 2001). Consumptive uses of wildlife resources in Tanzania are primarily hunting and animal trade, which have been contributing significant revenue to the Treasury. The export of live animals increased in revenue terms from Tsh 29.7 million in 1995 to about Tsh 167.4 million in 2001 (see Table 11 ). Likewise, revenue col- lected from hunting increased from a level of US$ 6.4 million in 1995 to ap- proximately US$ 9 million in 2003 (see Table 11 ). Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 43 Table 11 Tourist hunting and live animals export earnings, 1994 ­ 2003 Year Tourist hunting earnings (US$ Exports of live animals (Tsh. mil- million) lion) 1995 6.44 29.72 1996 7.31 25.49 1997 8.21 32.47 1998 8.56 40.99 1999 9.02 93.66 2000 8.53 136.97 2001 9.12 167.42 2002 9.30 N/A 2003 8.80 N/A (Source: URT, 2004c) While animal exports have increased over fourfold over the eight year time pe- riod, income generated through the tourist hunting industry has not even dou- bled during the same period. As Chart 14 below shows, the number of hunters has been almost stagnant since 1997. The number of hunting companies de- clined from 43 in 1994 to 39 in the year 2000. Chart 14 Number of tourist and citizen hunters 1994-2000 Number of Tourist Hunters 1200 1000 800 600 400 200 0 1994/95 1995/96 1996/97 1997/98 1998 1999 2000 Number of hunter Foreign Number of hunter Tanzanians (Source: URT, 2004c) Some additional figures on tourist hunting from Baldus and Cauldwell (2004) are presented in Box 7 below. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 44 Box 7 Some key figures on tourist hunting Box 8Trophy fees represent 60% of the income generated by the WD from hunting; Box 9Average income to the WD per hunting client is approximately US$ 7,000; Income generation in the Selous Game Reserve has grown on average by 13.7% per year from 1988 to 2001. Income from trophy fees and conservation fees has grown by 14% and 17.5% respectively; Income generation per unit area from all hunting areas of Tanzania is approximately US$40/km2. Hunting income per unit area for the Selous Game Reserve is approx. US$ 70.km2; Photographic tourism in the Selous Game Reserve generates approx. US$ 130/ km2. Photographic tourism generates nearly double the income, but at more than 50 times the tourist density per unit area; The number of hunting clients visiting Selous Game Reserve has grown by nearly 400% from 1988 to 2001. Income generation has however grown by only 250% over the same period, despite an increased utilisation of concessions. Income generated by the Wildlife Division per hunter day has remained at approx. US$ 250 per hunter day; and 141 concessions are leased to 42 companies, which in turn have formed 32 groups. 51 concessions (36%) are leased to the 3 largest groups. (Source: Baldus and Cauldwell, 2004) 3.5 Wildlife Division Income and Expenditure Funding sources of the WD are allocations from the Treasury, donor support and the Tanzania Wildlife Protection Fund (TWPF). Chart 15 below shows that the total revenue collection by the WD has increased from Tsh. 8.3 to 9.5 billion over the last three financial years. There is inconsis- tency in the data presented in the MNRT Financial Statements and the data pre- sented in Table 11 above originating from the Economic Survey 2003. Likewise, the Guardian of 5 May 2005 disclosed that the Wildlife Division col- lected Tsh. 7.2 billion in 2000 and Tsh. 9.7 billion in 2004 through professional hunting. Baldus and Cauldwell (2004) in an independent study, estimate income accrued to the Wildlife Division to roughly US$ 10 million, equivalent to three times the level reported by the Authorities. The Author could not establish the exact reasons for this significant discrepancy. Annual income from hunting is subject to fluctuations as tourism is affected by external events. For example the attacks in 1998 (Dar es Salaam and Nairobi) and 2001 (New Your and Washington, D.C.) have caused an approximate 25% drop in income in the following years. Hunting quotas are set by the WD in ad- vance of the season. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 45 Chart 15 Wildlife Division Revenue 2002 to 2004 9,800,000,000 9,600,000,000 9,400,000,000 9,200,000,000 9,000,000,000 sh 8,800,000,000 Ts 8,600,000,000 8,400,000,000 8,200,000,000 8,000,000,000 7,800,000,000 7,600,000,000 2002 2003 2004 (Source: MNRT 2002,2003,2004) Over the last three financial years, hunting licenses made up between 80 and 96 percent of the total revenue generated by the WD. Hunting Licenses consist of concessions (or blocks), each charged at US$ 7,500. There are 130 hunting blocks (2001) covering areas in game reserves, Game Controlled Areas and open areas. (Mabugu and Mogoya 2001). The second source of revenue between 1996 and 2000 were capture permits, which generated between US$ 20,000 and 38,000 per year. Since the last two years, game licenses, which are trophy fees per animal, paid after the client shoots the animal and which are passed over to the government, have generated increasingly more revenue than capture permits. Income from Game Licenses has increased from 1.6 percent in 2002 to 2.6 per- cent in 2004, which is still a relatively modest contribution. Other revenue sources, such as for example, Trophy Dealer's Licenses, Certificate of Owner- ship, Trophy Export Certificate, Capture Permits and various receipts contrib- ute not more than about 1 percent of the total revenue (see Chart 16 below). 25 percent of the revenue generated through hunting licenses go to the Tanzania Wildlife Protection Fund, and 75 percent to Treasury. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 46 Table 12 Revenue collection by source, 2002, 2003, 2004 Revenue Collection by Source Source 2002 2004 Amount in 1000 Amount in 1000 Tsh % Tsh % Trophy Dealer's Licenses 4,555 0.1 0 Certificate of Ownership 549,872 16.8 1,627 0.0 Game Licenses 53,682 1.6 154,407 2.6 Hunting Licenses 2,573,523 78.5 5,622,347 96.0 Trophy Export Certificate 5,857 0.2 2,842 0.0 Capture Permits 62,675 1.9 47,298 0.8 Receipts from Compounding Fees 16,315 0.5 11,978 0.2 Receipts Ivory, trophies, Hippo Teeth 146 0.0 0 Misc. Receipts 13,791 0.4 13,221 0.2 Total 3,280,416 100.0 5,853,720 100.0 (Source: MNRT 2004. Note: The Total does not include revenue collected and retained at source) District Councils in turn receive 25 percent of hunting and game license fees collected by the central Authorities within the District's jurisdiction. See Table 13 below for amounts received by selected District in 2004. The balance ac- crues to TWPF, the WD, and finally the Treasury. Table 13 2003 Revenue by selected District Councils from Game Licenses District Amount received (Tsh.) Monduli 47,029,337 Simanjiro 32,162,534 Mbarali 16,041,957 Kiteto 15,042,119 (Source: Wildlife Division 2005, unpublished) District councils are supposed to use the allocated funds to finance wildlife management and social infrastructure. Many councils have not been able to di- rect funds this way and are dissatisfied with the amount they receive, saying that these are too little and they do not know how the revenue sharing arrange- ment is determined (Mabugu and Mugoya 2001). Conservation fees at US$ 100 per hunter per day accrue entirely to the TWFP as do all other remaining fees. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 47 Chart 16 Distribution between sources of revenue in 2004 1% 3% Others Game Licenses Hunting Licenses 96% (Source: MNRT 2004) Hunters are divided into three categories: International clients using local out- fitters, Tanzanian citizens, and resident non-citizen hunters. Fees for Hunting in open areas for citizens and non-citizen residents are pay- able to the district councils. For example, Monduli district council, a recipient of the biggest share of game fees from tourism for the period between 1992-93 and 1999/2000 collected from resident hunting activities Tsh. 4,8 million in 1998, close to 6 million in 1999 and 3,2 million in 2000 from these fees. Chart 17 shows that while total revenue collection and actual collections have increased, the amount retained at the source has decreased. The author could not clarify why in 2002 a larger amount was retained at source than actual col- lections. Chart 17 Revenue collected by WD and retained at source, 2002-2004 7,000,000,000 6,000,000,000 5,000,000,000 4,000,000,000 revenue collected Tshs 3,000,000,000 retained at source 2,000,000,000 1,000,000,000 0 2002 2003 2004 (Source: MNRT 2004) Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 48 Constraints to revenue collection Several sources quote that revenue generation by the WD is sub-optimal. There is loss of revenue due to widespread subleasing of concessions. It is estimated that up to 70% of companies leasing concession are subleasing in various ways. Sub-lessees account to about 40% of the industry's income, yet there is no spe- cific taxation controls on the sub-lessees who are gaining huge profits at the expense of the industry. Moreover, many foreign permit holders are bringing clients to sublet concessions, among these are permit holders entering as unde- clared observers, paying neither observer fees nor private hunting licenses (Baldus and Cauldwell 2004). As a consequence a considerable revenue loss may be incurred. Some observers suggest that the current system, with no transparent system for allocation of hunting concessions to the private sector, is being maintained by vested inter- ests, including government officials and business agents. Furthermore, income generation from the concessions is dependant solely on hunting and is not con- ducive to the development of other sources of income. Public Financing As shown in Chart 5 in Chapter 2, Government recurrent budget allocation by the MNRT to the WD has declined from 33% in 2003 to 29% in 2004, in rela- tive terms, but slightly increased in absolute terms. Chart 18 Comparison of Recurrent Budget versus Revenue 2002 to 2004 8,000,000,000 7,000,000,000 6,000,000,000 5,000,000,000 Recurrent Budget 4,000,000,000 Tshs Revenue 3,000,000,000 2,000,000,000 1,000,000,000 0 2002 2003 2004 (Source: MNRT 2004) While Forestry had a small government contribution to the development appro- priation account, the development budget of the Wildlife Division is entirely foreign financed but equally characterized by under-spending. Chart 19 below shows the underspending of the WD development budget in 2003 and 2004. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 49 Chart 19 WD Development Budget 2003 and 2004 and underspending 9,000,000,000 8,000,000,000 7,000,000,000 6,000,000,000 Development Budget 5,000,000,000 (100% foreign) Tshs4,000,000,000 actual expenditure 3,000,000,000 2,000,000,000 1,000,000,000 0 2003 2004 (Source: MNRT 2004) Chart 19 shows that in 2004 only 35 percent of the allocated development budget was spent. The reasons for this underspending might be similar to For- estry due to low capacities to absorb foreign funding and institutional ineffi- ciencies, often aggravated through lack of coordination among the development partners that jointly support the sector. In contrast to Forestry, no SWAP for Wildlife is being discussed. Similar to Forestry, the comparison of recurrent and development expenditures over the last 3 years in Chart 20 shows that recurrent expenditures exceed by far the development expenditures and have been increasing over the time pe- riod. In contrast development expenditures were small in 2003 but increased again in 2004. Chart 20 Recurrent versus development spending 2002-2004 8,000,000,000 7,000,000,000 6,000,000,000 5,000,000,000 recurrent budget 4,000,000,000 Tshs development budget 3,000,000,000 2,000,000,000 1,000,000,000 0 2002 2003 2004 (Source: MNRT 2004) Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 50 Degree of self-financing Although wildlife is a productive sector, it is presently not a net contributor to treasury. Chart 21 below shows how during the past two years, the revenue col- lected by the WD has been insufficient to cover its budget, although the financ- ing gap has decreased recently. Chart 21 Degree of self-financing of the WD, 2003 and 2004 10,000,000,000 8,000,000,000 6,000,000,000 4,000,000,000 Revenue actual 2,000,000,000 Budgeted amount Tshs Under-collection 0 2003 2004 -2,000,000,000 -4,000,000,000 -6,000,000,000 (Source: MNRT 2004) The main expenses of the WD are associated with management and protection cost, especially salaries, maintenance of wildlife areas, transport and equip- ment. 3.6 Live Animals Export Live specimen form the majority of animal exports from Tanzania by quantity, although there is also a substantial trade in trophies and skins. During the pe- riod 1989 and 1999 at least 1.68 million birds, 521,000 reptiles, 148,000 am- phibians and 12,000 mammals were exported from Tanzania. The major mechanism by which the live export trade is controlled is through a system of licenses, permits and quotas. The issuance of TDLs and Trapper Cards determines the number and quantity of trappers and traders, whilst per- mits and certificates are the legal instrument that must accompany all live ani- mals from the trapping ground to export. Trapper Cards and TDLs for trading in live animals are valid for one year and cost Tsh. 10,000. Four types of permit/certificate are used in Tanzania: Capture Permits, Owner- ship Certificates, CITES permits and Trophy Export Certificate. Their issuance is normally restricted to license holders and the precise natural of the consign- ments follows strict annual national quotas. National quotas are usually divided equally amongst TDL holders. Every species has a capture cost which must be paid before capture and is normally less than US$ 1, although it can reach sev- Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 51 eral hundred dollars for some mammals. After capture, the specimen details are transferred to Ownership Certificates, which cost Tsh. 500 each and are valid for six months, whilst non-CITES specimen are exported on a Trophy Export Certificate, which costs Tsh. 1,500. The trade in animals has been affected by CITES regulations. In addition, a va- riety of local, national and international measures, including trade restriction and harvest control, have affected animal trade dynamics. One example is the impact of the export ban of Fishers' Turaco Bird: This bird was generating a high price (US$ 20) in 1995 when it was an important income earner in some areas in the East Usambara Mountains, and then ceased to be traded due to an export ban. Profit increases almost exponentially going up the trade chain. Middlemen may receive almost double the amount received by the collectors, whilst export prices are a further six times greater. For some species, US retail prices average 12 times the minimum export prices and 146 times the original collector's price. A case study conducted by TRAFFIC in the East Usambara Mountains, re- vealed that regulatory and market changes have had the greatest impacts on animal trade and not species availability. Local prices in US dollar terms have continually declined since 1990 although the relative price of different species has changed little. Average income from animal traders dropped from US$ 830 in 1995 to US$ 125 in 2002. Table 14 Export of Government Trophies, Live Animals and Birds 1993 1994 1995 1996 1997 1998 Type of Trophy Value Value Value Value Value Value (Shs.'000) (Shs.'000) (Shs.'000) (Shs.'000) (Shs.'000) (Shs.'000) Live animals and birds 28,013 16,672 18,120 16,934 22,672 22,900 Hippo teeth 18,729 7,319 4,509 1,600 1,200 .. Tortoises .. 523 106 .. .. .. Crocodile skins 1,200 660 5,000 5,000 5,000 5,000 Other product 340 1,185 1,990 .. .. .. Total 48,282 26,359 29,725 23,534 28,872 27,900 (Source: URT, 2004c) Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 52 Table 15 Export of Live Animals and Animal Tusks/Teeth 1999 2000 Type of animal Qty/Weight Value Shs. Qty/Weight Value (000) Shs.(000) Primates 1,094.0 6,216.0 3,897.0 2,005.5 Other mammals 1,279.0 1,291.9 543.0 7,477.1 Birds 166,605.0 20,263.3 114,241.0 41,171.7 Amphibians 25,900.0 2,130.2 48,173.0 3,372.1 Reptiles 63,850.0 3,072.4 43,590.0 6,298.7 Insects 18,383.0 811.3 57,866.0 4,050.6 Animal Tusks and Teeth 1,739.0 5,217.8 23.0 1,929.6 Cropping problem animals .. 39,000.6 Miscelenous e.g Export Certificates .. 6,625.6 Total 39,002.9 111,931.5 (Source: URT, 2004c) 3.7 Wildlife Resources and Poverty Reduction Wildlife Management Areas (WMAs) are described in the 1998 Wildlife Policy as land areas managed by communities to provide substantial tangible benefits from wildlife conservation. The regulations detailing WMA procedures were released with a four-year delay at the end of 2002. These regulations list 16 pi- lot areas in Tanzania where the concept of WMAs is being tested over a three year period (2003-2005). A number of applications for establishment of WMAs within these pilot areas have been submitted to the Wildlife Division, but no WMA has so far been formally established. The primary sources of village income from WMAs are the sale of hunting quotas to resident hunters and a percentage of revenue obtained from tourist hunting. Case studies provide aggregate income data at community level. There is no data on household gross benefits, and the distributional effects of income accrued at community level. There are studies showing that elite capture is a constraint to achieving the desired effects of poverty reduction (Mabugu and Mugoya 2001, Gillingham 2001, Ashley et al. 2001, and Homewood et al. 2001) Walsh (2000) provides income data from the MBOMIPA project working with communities in Ruaha National Park in Iringa, one of the WMA test sites. The data is presented in Chart 22 and Chart 23 below. Chart 22 shows the income from sale of resident hunting quotas, which totalled Tsh. 15 million in 1999, out of which the Villages earned 80 and the District 20 percent.19 This divided over the nine participating villages would be an annual income of Tsh. 1,7 mil- lion per village, assuming equal distribution. 19 These are 9 villages of Idodi and Pawuga in Iringa District. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 53 Chart 22 Village and District Incomes from sale of Resident Hunting Quota 16,000,000 14,000,000 12,000,000 10,000,000 villages 8,000,000 Tshs district total 6,000,000 4,000,000 2,000,000 0 1996 1997 1998 1999 (Source: Walsh 2000) In addition, seven out of the nine villages earned Tsh. 4,1 million from the 25 percent share of license fees from tourist hunting, which is Tsh. 587,000 per village. For those seven villages who record income from both sources, the an- nual income from WMAs per village was hence, Tsh. 2,25 million (assuming equal distribution). The overall total gross income for all villages from WMAs was close to Tsh. 20 million. No data on cost encountered by the villages is given, hence the net income cannot be assessed. Local management cost can be substantial. Walsh 2000 describes how frictions with resident hunters and potential inves- tors, as well as problems of good governance and competition over resources at village level, as well as conflicts with the district level, were constraints in maximising the potential income. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 54 Chart 23 Village Total Incomes comprised of resident hunting quota and 25% share of license fees from tourist hunting 18,000,000 16,000,000 14,000,000 12,000,000 village income from tourist 10,000,000 hunting Tshs 8,000,000 village income from resident hunting 6,000,000 4,000,000 2,000,000 0 1996 1997 1998 1999 (Source: Walsh 2000) Ashley et al. ( 2001) likewise report considerable income figures in MBOMIPA villages. The income from hunting quotas was sufficient to treble village level communal income, enabling villages to pay district-level taxes though would otherwise be levied by households, as well as to carry out specific inter-village investments such as the building of a secondary school and road improvements (Ashley et al. 2001). One of the success factors identified was that the project has placed emphasis on institutional capacity building at village and inter- village levels. Negotiations have been initiated with private tourism developers and the project is trying to help communities retain some control over the nego- tiation process. By contrast, WMA data from the Selous Game Reserve shows that village in- come generated is small and several villages reported misuses of funds by vil- lage level institutions (Gillingham 1998). Gillingham quotes figures of total income of Tsh. 1.8 for nine villages for the 1994/95 hunting season. Total in- come is comprised of quota meat income plus Tsh. 75,000 per year to the vil- lage from the tourist hunting company operation in the buffer zone of the WMA. Depending on the total income per village and the cost, net income per village ranged between US$ 159 and 463. While in one village cost were 9.5% of total income the highest was 37.5%. (Gillingham 1998). Similarly, Ashely et al 2002 found that in case study villages along the northern boundary of Selous Game Reserve most benefits accrued to local institutions. In terms of cash earnings, the 19-village society set up to manage wildlife, plus the District council and village governments had earned approx. twice as much in total as individuals received in wages or allowances. The main focus has been on distributing revenue shares from trophy hunting. Less evident are at- tempts to create enterprise opportunities or build capacity to manage problems. The role of private sector development and tourism development is often for- gotten. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 55 These experiences are from isolated 'project islands' and as long as there is no national effort to make WMAs fully operational, lessons learned from these pilot sites will not be replicated. Despite the alleged commitment to community involvement and participation through the Wildlife Policy, in practice, the WD has not adhered to its own policy and legal framework. For example it is re- ported by resident hunters and the MIOMBO magazine of the Tanzania Wild- life Conservation Society, the a well known and functioning WMA in 'Utete', Rufiji District, was taken away from the village government and sold to a for- eign private entrepreneur. Communities who are the legitimate holders of the land upon which hunting takes place are excluded from the decision making process. As Baldus and Cauldwell 2004 describe "there is a distinct lack of will to implement the policy" (p. 30). Consequently, the development of WMAs is seriously delayed. There is no progress on development of an effective schedule for sharing of benefits from tourist hunting with the local communities on whose land hunting takes place. There is a general hesitation among outfitters to accept the WMA concept. It is possible that much of the delay in the development of WMAs is the result of a negative high-level influence by some hunting operators, who reportedly have influence with the Government to block the allocation process. Another criticism suggests that the above regulations fail to place any real con- trol of the WMAs in the hands of the communities as was originally envisaged. It is widely known that hunting will be the major source of income from WMAs, but scrutiny of the regulations reveals that the WD will regain full con- trol over the appointment of outfitters to operate the WMAs and what they are allowed to hunt by controlling the quotas. Further constraints to community involvement in wildlife related business is that commercial success in mainstream hunting and tourism business requires large up-front investment, commercial experience and substantial risk-taking behaviour. These often present barriers of entry for local entrepreneurs or community level organisations. Rural credit is scarce, familiarity with the in- ternational tourism sector limited and hospitality skills low. Community part- nerships could be a possible way of tackling some of these constraints. 3.8 Sustainability There seems for example to be agreement among conservationists and hunters that there is a general decline in wildlife populations in many areas of Tanzania. Typically, they point to increased settlement and illegal off take of bushmeat by local communities as the main reasons for the alleged decrease. It is argued that growing human population are turning game areas into islands of wildlife habi- tat surrounded by cultivated and semi-urban land, risking that these areas loose their economic value for the hunting industry and associated government alli- ances. In this context it is reported by the hunting industry that a number of hunting concessions have ceased to be viable (Baldus and Cauldwell 2004) The views of local communities, by contrast, has not been well represented I the debate and no reliable figures have been provided by the authorities. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 56 Hence, in the absence of reliable ecological data it is not possible to objec- tively conclude about current levels of wildlife resources and their future sus- tainability. However, the absence of community benefits from wildlife management pro- vides poor rural households with little incentive for conservation. High levels of illegal utilization of wildlife resources persist, leading to a loss of opportuni- ties for economic growth and poverty reduction and un-sustainable use of the resource. While poaching and illegal trafficking is often associated with adja- cent communities, a data analysis in Katavi National Park on the village of ori- gin of poachers, shows that the majority of poachers did not reside in the park adjacent villages but rather in the district capital or other further removed loca- tions (KRCD 1995 unpublished). Although fines from poaching present a source of government income, they will be outweighed through the cost of anti-poaching control. As reported by The Guardian on May 9, 2005, Illegal trade of ivory from Tan- zania into the Middle East, is a persisting problem. As reported by the MNRT, over a five year period, 3,704 kg of ivory in transit were intercepted and more than 700 guns impounded at various exit points. Valued at the price that a pro- fessional hunter pays for an elephant (US$ 5,000) the government revenue loss through a single illegal consignment of ivory is 200 million US$ for 40 ele- phants killed in sport hunting. As the sources have established there is a well- organized criminal network of illicit trade out of Tanga port that rapes the country of billions of shillings. While the illicit trade pays the porters only Tsh. 50,000 while exposing them to great risks, it brings prosperity to the middle- men. The last consignment weighted 800 kg stuffed in 10 boxes labelled as hor- ticultural produce. Since the consignments are usually under police escort, it is difficult to stop illicit traffickers. There is a declining elephant population although an increase would have been expected due to the export ban introduced in the 1980s. Elephants are on the CITES Annex 1 list, for which trade is prohibited. There is a dispute as SADC countries wanted to down lift elephants into Appendix 2, endangered with lim- ited trade allowed. Unsustainable use and over-exploitation of wildlife resources is also fostered through hunting quotas out of tune with true market values and with no scien- tific basis with respect to maintenance of critical stocks. Concessions are leased at rates far below true market value irrespective of size, quality or income potential. This represents a significant loss of income to the Wildlife Division (estimated at over UD$ 7 million). Many concessions are leased to outfitters without the capacity to market or manage their own hunting operations. The system thus promotes subleasing to foreigners with a result that much of the income generated by the industry never enters the country and the Tanzania Revenue Authority do not access much of the funds that should be due for taxation. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 57 There is no effective monitoring of wildlife populations in the hunting areas. There is no objective system for quota setting. Many hunting quotas are issued at levels that allow unsustainable take off. Hunting outside of quota and a gen- eral lack of respect for the law by the members of the private sector has prompted the Director of Wildlife to issue a Call for Compliance to all hunting outfitters in 2004. However, no serious effort to prosecute violations has been observed. There is unsubstantiated evidence from professional hunters that trophy quality and age of key species such as buffalo has declined in the last five years in some key areas. Many hunting outfitters also admit that it is becoming increas- ingly difficult to hunt good lion trophies. Many outfitters are not voluntarily restricting the numbers of lion hunter in their concessions to encourage an in- crease in the available lion trophies despite high quota allocations. While some operators are imposing their own tight standards on the age and quality of lion trophies taken, others are over-shooting their quotas and taking young animals. Baldus and Cauldwell 2004 report that, although a few species such as lion are being affected by trophy hunting, the vast majority of species is unaffected. The general decline in wildlife populations is attributed to population pressure and there is no reported evidence that the regulated tourist hunting industry has con- tributed to the general decline of wildlife populations. In the contrary, Baldus and Cauldwell report that there is plenty of evidence that the presence of a regulated hunting industry contributes significantly to reducing the illegal ac- tivities of poaching and provides an economic incentive to protect vast areas. Ecological monitoring efforts of wildlife populations tend to be concentrate in and around national parks and Ngorongoro Conservation Area. Most areas managed by the Wildlife Division, which include most of the country's hunting blocks, are not subject to any regular population monitoring, with the exception of the Selous Game Reserve. Developing an ecological basis for setting quotas is not easy. It is extremely costly to conduct regular aerial surveys countrywide. Furthermore, aerial sur- veys are unable to provide data for key species and do not provide consistently reliable trends of populations. In the absence of reliable ecological data, cumulative experience of setting quo- tas over many years relying on a number of indicators (e.g. trophy quality) pro- vide the basis for an adaptive management approach. Each year about 7,500 animals representing up to 60 species are hunted in over 130 hunting blocks generating large amounts of data. The WD has twice resisted attempts to com- puterise the system. It is doubtful that in the absence of proper records man- agement an adaptive management approach can be practices. The approach used to allocate quota is to a large extent relying on the knowl- edge of project managers and district game officers. Aerial survey data are taken into account together with past hunting records. There is a tendency for outfitters to be allocated their required hunting quotas, and that quotas are in- creasingly raised regardless of the population status. Hence, quotas are not Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 58 based on ecological indicators but rather economic interests of a few powerful players. In summary, the WD is not adhering to ethical principals that would support sustainable management of the resource. For example, under-sized trophies are legalized with export documentation although they do not meet the minimum standard. Also, it is reported that the WD issues additional quotas to outfitters upon request during the hunting season. In addition the environmental stan- dards for hunting camps are inadequate. In summary, an inflexible management system, rigid game fee schedules and strong emphasis of trophy fees leaves the Wildlife division with no option other than to increase quota off take to gener- ate increased revenue. An inventory of the true value of the hunting conces- sions to know the value of resources has never been conducted. 3.9 Conclusions and Recommendations Most of the issues raised in the context of the wildlife sector, refer to the tourist hunting industry and can be summarized as government failure to manage the sector effectively and transparently, preventing the realization of the true mar- ket value of resources, achievement of sustainable resource use and poverty reduction benefits. In more detail, some of the key issues include: Lack of monitoring of the wildlife resource and `off-take' by the hunting industry. This is associated with a lack of ethical standards for hunters and certification of hunters' competence. Lack of implementation of the WMA regulations, leaving government in control of resources on village land and preventing communities from con- trolling their own poverty. Lack of implementation of the policy and management plan for tourist hunting, developed in 1995 as a tool for reform in the industry. An out- dated and non-transparent system of issuing permits is maintained. In conclusion, the current management of the tourist hunting industry seems to serve the present hunting concession holders at the cost of potential competi- tion. Increased transparency and competition would bring many new players to the industry and thus boost current revenue levels for the potential benefit of the Government as well as local communities. Recommendations In order to exploit the potential for growth and rural development from wildlife resources, it is necessary to ensure greater coherence between different national policies and instruments, particularly community based wildlife management, tourism development, rural growth strategies, investment regulations and incen- tives and poverty reduction strategies. More specifically, the following recom- mendations are made: Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 59 Encourage a change of mindset at policy level for Wildlife to be seen as an asset for rural development and poverty reduction. Involvement of local communities through implementation of the WMA regime. Local communities should be allowed to negotiate directly with hunting outfitters through decentralised tendering, provision of personnel to supervise hunting on their land, participation in setting quotas, and receiv- ing and managing revenue from hunting; Improve monitoring, information and records systems at central Govern- ment level with a view to monitor wildlife stocks more precisely. This would also improve accountability and transparency in the sector overall and could also be supported through independent monitoring mechanisms. Improved information would also give the Authorities a basis for fixing fees and licenses at levels allowing for maintaining sustainable Wildlife stocks. Fees should also reflect market rates and shall include the conserva- tion and observer fee. Liberalise procedures for tendering hunting blocks to allow the best outfit- ters to bid competitively against each other for concessions. Provide ena- bling frameworks for partnerships between communities and the private sector. However, recognising the strong resistance to this particular issue, alternative allocation procedures may be necessary. The correct allocation and utilization of resources can lead to significantly higher levels of income generation, which should be used as a yardstick for measure of the effec- tiveness of reform. Implement control of subleasing, which may come naturally through effec- tive market-based competition. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 60 4 Marine Fisheries Marine and freshwater Fisheries are analysed separately in this study. While this Chapter provides some general data on the Fisheries sector as a whole, the focus is primarily on marine Fisheries, aligned with the overall heading of this paper of `un-captured growth potential`. Freshwater Fisheries, by contrast are dealt with in Part 3 of this report, where the significant growth observed in the Lake based fisheries is analysed and its wider externalities for employment, income and food security is discussed. Apart from Prawns, which are one of the major foreign exchange earners for the country, marine Fisheries has been slow to develop in Tanzania.. The slug- gish growth is partly due to the export ban on fin-fish, which has been lifted partly for certain species in early 2005. Still, as this Paper will argue, the sector may hold significant un-captured growth potentials. Investors are already starting to shift their operations from the Lake Zone to the Coast. It was reported for example by the Norwegian Embassy in Dar es Sa- laam that five letters from Tanzanian entrepreneurs were received during the last 24 months requesting assistance (credit and equipment) to shift their opera- tions to the coastline.20 Declining yields per boat in freshwater Fisheries place even more emphasis on the potential contribution of the marine Fisheries to economic growth. If the sector is well managed, the commercial Fisheries can potentially have a posi- tive impact on the country's economic development and the wealth of its popu- lation. It is hence important to learn from the lessons of the freshwater sub- sector to ensure that commercial Fisheries in the coastal zones can be devel- oped on a sustainable basis. However, even in marine Fisheries (as in the freshwater sub-sector), there are estimates that exploitation is approaching or already exceeding maximum sus- tainable yield (MSY) levels for certain species. It will hence be important for sustainable management of the Fisheries to include the retention and re- investment of revenue into the sector. This may allow control and surveillance monitoring of stocks to ensure sustained exploitation of the marine Fisheries resources. Also, it will be necessary to put in place certain `safeguards' for ar- 20 Personal interview with Eirik Jansen, Norwegian Embassy, 12. May 2005 Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 61 tisanal Fisheries to protect their rights and access to the resource, which sus- tains their livelihood. Another key area for discussion is the growth in near and off shore Fisheries. There are increasing numbers of foreign trawlers in Tanzania's EEZ. While there are improvements to the monitoring and surveillance of these waters, there is arguably a significant degree of illegal fishing and loss in revenue (and growth) to Tanzania from this activity. Box 10 Key estimates on Fisheries in Tanzania · Tanzania's Fisheries sector has grown at a rate of 6 to 7 percent annually since 2000; · Fisheries exports totalled US$ 130 million in 2003, corresponding to more than 10 percent of total exports · The export value of Nile Perch totalled US$ 100 million in 2003; · The number of artisanal fishermen has roughly doubled since 1995 and reached close to 120,000 in 2003; · Tsh.9.5 billion was collected in revenue from the Fisheries sector in 2003/04. This represents roughly a 50 percent increase from revenue collected in 2001/02; · Only 20 percent of revenue originates from marine fisheries, with 80 percent com- ing from freshwater fisheries (2003); · The sector registered a revenue over-collection of roughly Tsh.3 billion in 2003/04; and · The number of foreign vessels licensed to operate in the EEZ (Mainland and Zan- zibar) has increased from less than ten in 1998 to more than 170 in 2004 corre- sponding to a revenue of US$ 3.3 million (Source: IMF, URT, 2004c, MNRT, 2004) The structure of this Chapter is as follows: Section 4.1 provides key figures on the contribution of Fisheries to the GDP and Section 4.2 on Sector Financing and Spending. Section 4.3 describes Marine Fisheries Resources, followed by Section 4.4 on Revenue from Marine Fisheries. The linkages between Marine resources and poverty reduction will be discussed in Section 4.5 and Section 4.6 deals with Sustainability. The policy framework of marine Fisheries is de- scribed in Section 4.7 and finally some recommendations will be developed in Section 4.8. 4.1 Contribution of Fisheries to GDP The Fisheries sector in Tanzania has recorded high growth rates during the past four years. Whereas annual growth was averaging three to four percent during the 1990s, it jumped in 2001 to a level of six to seven percent annual growth. Hence, export of fish products is now a major source of foreign currency for Tanzania. In 2003, 11.9 percent of total export earnings came from fish prod- ucts, making it the second most important source of foreign currency to the minerals sector (IMF, 2004). Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 62 This growth has primarily been driven by freshwater Fisheries, especially har- vesting of Nile Perch in Lake Victoria for consumption in EU markets. Still, catches per boat in the Lake Zone has been declining in recent years and so. growth may not be sustainable in the long-term. Fisheries are a sub-sector under the Agriculture GDP in the System of National Accounts. There are no statistics that account separately for the contribution of marine or freshwater Fisheries to GDP. The rate of growth of the Fisheries sector under the agriculture GDP between 1990 and 2003 is shown in Chart 24 below. As mentioned, the increase in growth in 2001 is due to the lifting of an EU import ban on fish from Lake Vic- toria in November 2000. At the same time, the Chart shows (mildly) declining growth rates for 2002 and 2003: This is explained as follows by the Fisheries Department (FD): In 2002, due to a fall in prices of exported fish products in the world market and decrease in volume of fish, particularly prawns. In 2003, due to a shortage of financial capital to small scale fishermen, poor handling and processing of fish products, illegal methods and destruction of marine habitat (URT, 2004c). Despite the slightly lower growth rates in recent years, growth is still above 6 percent and remains an important source of income for the Tanzanian economy. Hence, it is still too early to say whether declining rates in recent years is in- dicative of a negative trend. Chart 24 Growth rate of Fisheries Sector 1990-2003 8 7 rate 6 5 growth 4 3 percentage2 1 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 (Source: URT, 2004c) Similar to forestry, the total contribution of the Fisheries sector to economic development and as a source of livelihoods is only partly reflected in the GDP. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 63 The GDP figures represent only commercial Fisheries, which are directly asso- ciated with a cash flow through either license revenue or foreign exchange earning from export. Artisanal Fisheries, which to a large extent take place on a subsistence basis and fall within the informal sector, are not considered in the National Accounts. In addition, the potential of coastal zones for increasing the share of foreign exchange earnings from sustainable development of the coastal tourism indus- try in both mainland Tanzania and Zanzibar is significant but not accounted for under the Fisheries GDP. Part of this contribution of marine resources might be captured under earnings of the tourism industry within the "Trade, Hotels and Restaurants" GDP, which is the closest proxy for tourism in the National Accounts. There is a trade-off between the use of marine resources as a source of revenue for Fisheries and as a source of foreign exchange earnings through tourism. Mainland Tanzania and Zanzibar combined include 1,380 Km2 of Marine Pro- tected Areas and Marine Management Areas (MMAs) within their terrestrial seas. Only if these areas are protected from commercial (illegal) over- exploitation for Fisheries can their tourism values be maintained. A delicate balance needs to be drawn, which requires sound environmental governance. Research will need to explore additional un-captured growth potentials from marine resources, such as for example bio-prospecting, values of sponges, soft corals, tunicates, and different sea-weeds. These are small but growing indus- tries just beginning exploratory activities along the eastern African coast. 4.2 Sector Financing and Spending While revenue collection is dealt with more specifically for the marine Fisher- ies in Section 4.3, some general information and data on sector financing is pre- sented here combined for fresh and marine water Fisheries. Sources of revenue The primary revenue raising instruments in Fisheries are vessel registration and licensing, export royalty and fish levy (charged on the sale of fish). Fish levy accrues to local governments, whilst export royalty and license fees for vessels greater than 11m/20GRT accrue to central government via the Fisheries Divi- sion. Charges for vessel registration, licensing and fishing vary with vessel size, flag and whether the owner has an approved onshore processing facility. Export royalty on principle Fisheries exports is charged on per kilo basis, approximat- ing six percent of free on board (FOB) value. Fish levy is charged at a maxi- mum of five percent of landed value. The legal basis for all charges in the sec- tor is the Fisheries Act of 1970. Chart 25 below shows the two main sources of revenue for 2003 and 2004: Ex- port royalties and export licenses (from vessels). Export royalties clearly domi- Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 64 nate (mainly derived export of from Nile Perch from Lake Victoria) while ex- port licenses, although small in comparison, has been increasing due to an in- creasing number of foreign vessels registering to operate in the Exclusive Eco- nomic Zones (EEZ). Chart 25 Sources of Fisheries revenue in 2004 9,000,000,000 8,000,000,000 7,000,000,000 6,000,000,000 5,000,000,000 Fishing export licenses Tsh.4,000,000,000 Export royalties 3,000,000,000 2,000,000,000 1,000,000,000 0 2002/2003 2003/2004 (Source: MNRT 2004) Revenue collected at decentralised level amounted to US$ 1,5 million in 2002. 99 percent of the decentralized revenue collection comes from fish levy, with 34 percent originating from marine Fisheries and 66 percent from freshwater Fisheries. The total revenue collected by the Fisheries sector was close to Tsh.7 billion. in 2002/0321 and close to Tsh.9.7 billion in 2003/04.22 As Table 4 in Section 2.3 above has shown, the Fisheries Department has re- ceived 18 percent of the budget of the MNRT in 2002/03 and 2003/04. In nominal terms, this was Tsh.3,688,280 and Tsh.4,648,202 respectively. Government spending on Fisheries depends on the revenue generated from ex- port levies and the issuing of fishing licenses. Chart 26 below compares the re- current budget of the FD to collected revenue. 21 exact amount: Tsh. 6,994,511,808.15 22 exact amount: Tsh. 9,698,498,793 including revenue retained at source. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 65 Chart 26 Comparison recurrent budget to revenue collected 12,000,000,000 10,000,000,000 8,000,000,000 Recurrent Budget Tsh 6,000,000,000 Revenue 4,000,000,000 2,000,000,000 0 2001/02 2002/03 2003/04 (Source: MNRT 2004) In contrast to Forestry and Wildlife, Fisheries has no recorded allocations or expenditures under the development appropriation account in the MTEF. There was hence no data readily available on the amount of foreign financing to the sector. In summary, Fisheries operate with a smaller recurrent budget than For- estry and Wildlife while they record higher revenue. Being a revenue-earning sector, the key fiscal objective for Fisheries is to be self-financing. In other words, the cost of regulating and developing the sector is to be covered by the earned revenue. Moreover, revenue generated from Fisheries on the mainland plays an important role in financing the sector and in raising revenue for the treasury and local administration. In comparison to the Forestry and Wildlife sectors, the Fisheries sector is the only one where revenue collection exceeds the GoT recurrent budget allocation. As Chart 27 below demonstrates, the Fisheries Division has recorded over- collection of revenue in the past three years. In 2003/04 the actual collections represented an over-collection of more than Tsh. 3 billion vis-à-vis budgeted amount. The Division explains this over- collection as a result of increased foreign vessel compliance, due to aerial sur- veillance done under the on-going SADC-MCS programme, and intensified patrols by revenue collectors. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 66 Chart 27 Over-collection in the Fisheries Division, 2001/02 ­ 2003/04 12,000,000,000 10,000,000,000 8,000,000,000 Revenue actual Tsh 6,000,000,000 Budgeted amount Overcollection 4,000,000,000 2,000,000,000 0 2001/02 2002/03 2003/04 (Source: MNRT 2004) The FD's budget ceiling is defined as 50 percent of estimated revenues. If ac- tual revenues are greater than estimated, the difference accrues to the treasury. This can lead to a substantial 'loss' to the potential FD budget. The sector con- tributes about one percent to total government revenue collection. The retention scheme in Fisheries allows the FD to retain about 48 percent of the earned revenue. 6 percent is taken as overhead by the MNRT and the remaining 46 percent retained by Treasury. The annual amount spent by the mainland FD is about US$ 5.2 million. The mainland more than covers its costs. The MNRT reports that FD achieved 96.1 percent of its targets. In terms of sector spending, only 2.2 percent of available resources was spent on protection of resources. 4.3 Marine Fisheries Resources Marine Fisheries in Tanzania are dealt with separately by the Fisheries Depart- ments of Mainland Tanzania and Zanzibar. Therefore, marine Fisheries can in principle be divided into two Territorial Seas and two Exclusive Zones: Mainland and Zanzibar. To simplify the analysis, there are five distinct marine Fisheries resources in Tanzania: Mainland prawns; Mainland artisanal; Mainland Exclusive Economic Zone (EEZ); Zanzibar artisanal; Zanzibar Exclusive Economic Zone (EEZ). Given that Zanzibar falls under a separate CEM process, this report will focus primarily on marine Fisheries administered by the Mainland. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 67 In the context of the economic growth potential of the marine Fisheries, there is a crucial distinction to be made between prawns and artisanal Fisheries on the one hand and Fisheries in the EEZ on the other. While the former are, based on estimates from the Government (UTR, 2004c) and the MCS-SADC project , currently operating at maximum sustainable yields levels, and are already showing declining yields; the latter present an important, emerging revenue source for Tanzania and the sector itself. However, as this Section will show, even for certain species in the EEZ, current catch rates are at MSY levels. An additional difference is the governance regime; in contrast to prawns and ar- tisanal Fisheries, EEZ Fisheries has to comply with international legislation beyond the one of the coastal state, in this case Tanzania (see Section 4.7 be- low). Table 16 below shows various marine Fisheries resources and their estimated value for Tanzania Mainland. Table 16 Marine Fisheries resources and their value (base year 2000) Fish- Type Total Value in US$ Export Value in Remarks eries US$ Prawn Industrial and Ar- Estimated at 7 mil- 5-6 million an- 13-21 trawlers all Tan- tisanal lion per year nually zanian flag. Closed Sea- son March ­ Nov. Ar- Artisanal reef and Estimated at 11 3.5 million Up to 13,000 traditional tisanal inshore pelagic million per year fishing vessels species, crusta- ceans EEZ/ Industrial ­ large Not known 1.9 million Boats from far East and Off pelagics the EU. Most do not land shore in Tanzania and records are incomplete. (Source: Fisheries Department) Prawn and artisanal Fisheries With regard to prawn fishery, there is evidence at TAFIRI, the Tanzania Fisher- ies Research Institute that the stocks in the Rufiji Delta are decreasing. How- ever, no quotas have been imposed. Albeit lack of government control, the Trawler Owners Association has voluntarily reduced the length of the season by one month (March to October) and reduced the number of nets used on its beam trawlers. The fact that this has happened is an indication that the prawn fishery has been operating at a level that cannot sustain increased exploitation. However, government has not yet imposed effective controls beyond these vol- untary measures. Prawn catch is also associated with fish by-catch, which provides an important source of fish supply to the domestic market. Chart 28 below shows the catch of prawn and fish by-catch between April and October 2004, an in-season pe- riod. There is a sharp decline from a total of 189,319 kg prawn catch in April down to 10,998 kg in October 2004. The decline is even more significant for the fish by-catch, which declined from 305,052 kg to 12,952 kg within the same period. The abovementioned voluntary restrictions may explain part of this decrease. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 68 Chart 28 2004 Prawn Catch and Fish by-catch in 2004 350,000 300,000 250,000 kg 200,000 in Prawns 150,000 Fish catch 100,000 50,000 0 April May June July August Sept. Oct. (Source: Fisheries Department) While the artisanal Fisheries present an important economic activity in coastal regions, there are no official statistics on the volume of business activity, em- ployment or catch. Still, Some indication can be derived from Chart 29 below, which shows number of seawater boats compared to the total number of boats in artisanal Fisheries and the total number of artisanal fishermen. While the number of registered seawater boats has remained around 4000 between 1995 and 1999, it jumped to roughly 5000 in 2000 where it now remains stable. The number of freshwater boats on the other hand has increased much more rapidly, and now counts more than 30,000 boats. This growth seems to be positively correlated with the number of fishermen, which has increased to 120,000. Chart 29 Artisanal fishermen, total no. of boats and sea water boats 140000 120000 100000 No. artisanal fishermen 80000 Seawater boats 60000 Total boats 40000 20000 0 1 2 3 4 5 6 7 8 9 (Source: Economic Survey 2003) Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 69 EEZ Fisheries There is a dearth of information on the Exclusive Economic Zone (EEZ) Fish- eries in the Tanzanian territorial seas. Foreign corporate bodies and/or indi- viduals are allowed operate within the EEZ water bodies provided that they en- ter into an agreement with the Government. Of the approximately 25 industrial boats fishing the EEZ in any given season, only four land in Tanzania (Dar es Salaam) and the fishing states supply only limited information on the fish removal from EEZ and their value. Although the FD closely monitors these registered vessels, they only represent about one sixth of the entire fishing effort. It could be surmised from the figures that are available from the FD that the potential value of the fishery is about US$ 12 million per annum but this is not substantiated by scientific research. The other major factor concerning EEZ Fisheries is that the stocks are generally migratory and their appearance in Tanzanian waters is seasonal. There is a dis- tinct season from September to February only. Regarding EEZ pelagic fishery, there is no baseline study of stocks and no figures for Maximum Sustainable Yield or Total Allowable Catch. However, some estimates exist for Tuna and Tuna like species in the Western Indian Ocean from the Indian Ocean Tuna Commission (IOTC). Yellowfin catches in the Indian Ocean were extraordinarily high during 2003 and 2004, while skipjack and bigeye remained at their average levels. 2003 was a record year for the catches from all fleets that reported to the IOTC Secre- tariat. These anomalous catches occurred all over the western Indian Ocean, in particular a small area in eastern Africa. The fish caught were of large sizes (100-150 cm). The management advice of the IOTC was that total catches of Yellowfin were close to or possibly above MSY. In these circumstances, any further increase in both effective fishing effort and catch above levels in 2000 should be avoided. While the MSY set by IOTC is 280,000-350,000 tons per year, the 2003 catch was 400,000-450,000 tons per year and the average catch over the last 5 years was 326,000 tons (Chopin 2005). For all main tuna species, 23 yearly catches have been increasing steadily since the early 1950s when the industrial tuna Fisheries in the Indian Ocean began. While the increase has been slow but steady until the late 1980s, there has been a sharp increase in fishing effort since the 1990s with exceptional high catches in 2004 recorded through voluntary reports by longline vessels and tuna sein- ers. The catch includes younger, lower weight tuna for canneries, caught by seine vessels, and larger fish for the sashimi market caught by longliners. Evidence suggests that 2004 was an anomalous year for tuna Fisheries with a significant shift in effort benefiting Tanzania as the fish moved into the EEZ. Accordingly, there is the possibility that this temporal shift may not occur in future years and interest for licenses may decline. 23 These are: yellowfin, skipjack, forskip jack, bigeye Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 70 Based on a review of the current status of stocks from IOTC data and on the Fisheries data submitted by vessels operating in the Tanzanian EEZ, there may be cause for concern with respect to sustainable development of three of the prime species harvested in the EEZ (see Table 17 below). In light of these es- timates, future controls may be organised by the IOTC. The fact that Tanzania is not an IOTC member and hence does not have a his- tory of catch reporting vis-à-vis the IOTC, the quota eventually allocated to Tanzania by the IOTC may underestimated since IOTC has no information on the level of current catchments. Table 17 Management advice from the IOTC Species Recommendations and Observations Bigeye Tuna Current catches above MSY Yellowfin Tuna Current catches close to or above MSY Increasing pressure on juvenile fish by tuna seiners is likely to be detrimental to the stock if it continues Skipjack No need for immediate concern Swordfish Current level of catch unlikely to be sustainable Controls recommended for the SW Indian Ocean. (Source: Robin 2005) The data coming from the Far East longliner fleets and the EU Tuna seiner fleet is scant and of dubious quality. This lack of data makes it extremely difficulty to assess the total removals of fish from the EEZ and their value. The catch of species listed as 'others' is very high (69 percent) and there is no indication what these species are, or what their value is. There is also a disparity between catch composition of longliners in the EEZ and data held by IOTC on all fleets. The data available on both catches and effort at the vessel level are extremely limited. There have been no scientific observers onboard the vessels to verify catch and effort information. Based on the data available, the following esti- mates are provided for the year 2004: Table 18 EEZ Catch and Value of Tuna and Tuna-like big pelagis, 2004 Estimated Catch Estimated Value of Catch*) Longliner (mainly Asian countries) Total Fleet 900 ­ 18,000 t out of which US$ 47.6 ­ 254.0 million 4,500-8,800 t tuna and billfish Per vessel per 70t of tuna and billfish and US$ 0.39 ­ 2.06 million year 150 t of other species Tuna seiner (mainly EU) Total Fleet 26,000 ­ 32,800 t of tuna US$ 27.5 ­ 85.6 million Per vessel per 900 ­ 1,000 t US$ 0.67-2.09 million year (Source: Chopin 2005) *) based on average annual prices for tuna for the sashimi market and canning. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 71 However, the data may underestimate actual catch, as some vessels reportedly are in breach of the license conditions by fishing in restricted areas, such as ma- rine protection areas. 4.4 Revenue from Marine Fisheries The greatest potential for increasing government revenue from fishery is in the licensing of vessels to fish in the EEZ. The FD started issuing licenses to for- eign vessels in 1998, mainly to European tuna seiners and Asian longliners. The license period ranges from a month to one year. The cost of a license de- pend on its duration, but the following general rules apply: The annual license fee is set at US$ 16,000; The registration fee is set at US$ 2,000; A license can be issued for a period of 1 month, 3 months, 6 months or one year; and Where the license period is less than one year, the license fee is pro-rated but the registration fee remains fixed at US$ 2,000. Since licenses were introduced in 1998 the numbers of foreign vessels fishing Tuna and Tunalike big pelagis has risen rapidly. In 2004 the total number of registered vessels was 171 of which 41 were tuna seiners and 123 longliners. The mainland issued 85 licenses and Zanzibar 86 (mainly to longliners from the Far East). Chart 30 EEZ licenses issued by mainland Tanzania in 2004 Equatorial China Guinea Spain 19 3 1 France 15 South Africa Indonesia 2 1 Seychelles Italy Panama 5 1 1 Mozambique Korea Japan 1 11 25 (Source: Fisheries Department) Chart 30 above shows to which countries the licenses are issued. The revenue accrued from these licenses (combined for Mainland and Zanzibar) is shown in Chart 31 below. The significant increase from 2002 to 2003 can be explained Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 72 by two factors: a) increased demand as a result of an increase in the presence of Tunas in the Tanzania controlled EEZ and b) improved enforcement mecha- nisms, which have been strengthened in the framework of the MCS-SADC pro- ject. Chart 31 Annual License Revenue and foreign vessels in Tanzanian EEZ 4,000,000 180 3,500,000 160 3,000,000 140 2,500,000 120 100 vessels US$2,000,000 80 of 1,500,000 60 No. 1,000,000 40 500,000 20 0 0 1998 1999 2000 2001 2002 2003 2004 US$ No of vessels (Source: Fisheries Department) Increased control and compliance in 2004 have increased the number of li- censes and reduced illegal fleets. In 2004 the total revenue from license fees that accrued to the Tanzanian FD was US$ 3.3 million (171 licenses @ US$18,000). This is not even reflecting the total amount that the government could earn. Experts estimate that the real catch is most likely much higher than what has been assumed as a basis to set the license fees (between 200 and over 400 tons a day per boat). Notably, there is no catch based license or fee, and the vessels are allowed unlimited catch once they are in possession of a valid li- cense. In addition, the Government earned US$ 300,000 in license fees from Tanza- nian Flag Prawn Trawlers (there are 25 with a license fee of 16,000 US$), which are not shown in Chart 31 above. Although the above revenue figure is a considerable amount, it is low compared to the estimated value of the catch by foreign vessels in Tanzanian waters sold on foreign markets (see Table 16 in Section 4.3 above). Also, in the search of efficiency, substantial cost may be incurred, especially in monitoring and sur- veillance. In addition, in order to maintain a healthy resource base, substantial scientific research will be required and membership of IOTC should be consid- ered. Taking these costs into consideration, the net revenue from EEZ Fisheries will be modest and certainly not sufficient to capture a resource rent to re-invest into the sector to ensure sustainable growth. Estimates on resource rent are pro- vided in Section 4.6 below. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 73 Chart 32 Revenue earned from Marine Water Catch in Tsh. and percent of total catch (incl. Freshwater) 40,000,000 50.0 35,000,000 45.0 40.0 30,000,000 35.0 value 25,000,000 30.0 h 20,000,000 25.0 catch Ts 15,000,000 20.0 15.0 total 10,000,000 of 10.0 % 5,000,000 5.0 0 0.0 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Value Tsh % (Source: URT, 2004c) Chart 32 shows that the revenue from marine water catch increased signifi- cantly between 1993 and 1995, and has also, except minor drops in 1996 and 2000, increased in absolute terms until present day, albeit at lower rates. Measured as a proportion to total revenue, revenue from freshwater also in- creased during the early 1990s and peaked to a level of roughly 40 percent in the 1990s, but has since dropped sharply to roughly 20 percent, following the previously mentioned boost in the freshwater Fisheries sector which led to a corresponding increase in revenue from freshwater catch by more than 150 per- cent24 in 2003. Revenue from Export royalties In 2003 there were 25 vessels for prawn fishing, which generated the catch vol- ume and export revenue listed in Table F, Annex 1. Prawns, fish maws and octopus are by far the largest income earners. The percentage of marine prod- ucts on total exports of fish products is small but has been increasing in the time period considered. 4.5 Marine Resources and Poverty Reduction Marine Fisheries have a significant importance for employment and income levels along the coastal regions in Tanzania. Presently, one third of the national GDP is produced in the coastal areas (Ruitenbeek et al. 2005). This however is mainly due to the fact that some 75 percent of the country's industries are located on the coast, mainly in the Dar es Salaam area. 24 From Tsh. 54,771,300 to 141,073,500 Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 74 Poverty along coastal regions in Tanzania is widespread, on average 85 per- cent of the sample population of a survey conducted for the "Blueprint 2050" study survives on less than a dollar a day. The average monthly per capita con- sumption expenditures in 2003 for the sample population of the same study was about US$ 21. This is roughly two-thirds of the established poverty line of US$ 30 (Ruitenbeek et al. 2005). Table 19 below shows the GDP per capita for the five coast regions, the Hu- man Development Index (HDI) and the Human Poverty Index (HPI). While Dar es Salaam is an exception and shown here for comparison, the other four coastal regions are among the middle and low ranked regions for both indica- tors, suggesting relatively high poverty levels. Table 19 HDI and HPI for coastal regions25 Indicator Tanga Pwani Lindi Mtwara Dar es Salaam GDP per capita 230,454 193,877 223,191 292,795 584,086 2002 in Tsh (cur- rent prices) GDP contribution 4.4 2.0 2.1 3.8 16.9 in percent Human Develop- 0.447 0.449 0.407 0.488 0.734 ment Index by Region Human Poverty 40.7 44.9 47.2 36.8 21.4 Index (Source: NBS website and URT 2002, p. 57 ff) Outside the urban centres, marine Fisheries are a vital source of employment and income as well as an important source of nutrition and alternative to expen- sive meat. Direct benefits of marine Fisheries resources to the local population are presently derived from artisanal Fisheries and fish by-catch from Prawn fishing, which is sold on the domestic market. The household survey in fishing communities conducted through the `Blueprint 2050 study' shows high levels of dependency in fishing communities on marine resources, in particular fish. As Table 20 below shows, among the resource based activities, fish is with 47.4 percent of all households the second largest source of household subsistence and employment after cassava farming which is undertaken by every second household (50.1 percent of the sample house- holds). 25 While a low value for HPI shows lower level of poverty, a high value of HDI show lower level of poverty. For computation of the two indices see URT 2002. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 75 Table 20 Contribution of fish to household subsistence in coastal areas Activity Percentage of households engaged in activity (total sample size 749) Marine fauna fish 47.4 crustaceans 15.6 sea cucumbers 4.3 mollusks 2.4 Seaweed farming 25.9 Farming - various Crops Cassava 50.1 bananas 32.7 rice 27.5 Farming - Agroforestry coconuts 14.7 Livestock keeping poultry 12 (Source: Ruitenbeek et al (2005), p. 53, adjusted) While commercial Fisheries, in particular in the EEZ, represent potential for economic growth, the impact on poverty reduction will depend on how the li- cense revenue earned by the Government is translated into benefits for local people. The effect of Fisheries Agreements on poverty reduction will depend on the creation of economic `spin-offs' and associated development activities. These are expected to be negligible as no fish is expected to be landed ashore and few supplies will be sourced from the country. It is for example unlikely that foreign vessels will employ Tanzanian nationals as crew in addition to what is stipulated in the Fisheries Agreements. If none such `spin-off' effects are created, the net impact of commercial Fisheries on poverty reduction may be negative, provided that it competes with artisanal Fisheries over the same resource. Although EEZ Fisheries and artisanal Fisheries cover different zones, foreign vessels sometimes transect from the EEZ into coastal zone. It will consequently be important to impose strict policy measures to protect the resource rights of the local population. Otherwise, poverty might be aggravated as a livelihood base is withdrawn while catch by foreign vessels, who can access the EEZ wa- ters and have more efficient technology, increases. 4.6 Sustainability of Marine Fisheries There are concerns about the sustainability of marine Fisheries. These originate from two kinds of illegal fishing: Firstly, infractions by industrial fishing fleets which often transgress into terri- torial waters disrupting the livelihood of the artisanal fishery. These vessels are at times in the zoned areas of the marine protected areas as well, causing the destruction of globally significant marine biodiversity. The second type of illegal fishing is related to dynamite fishery, which contin- ues unabated, causing economic and ecological loss with negative impacts for tourism potential and waste of the fishery resource. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 76 Recent reports from Muheza District in Tanga Region, suggests that dynamite fisheries remains a serious problem: After a significant fall in explosions to a level of three per month, allegedly as a result of Navy patrolling, the frequency has increased to 70 explosions in just three months following the cessation of patrolling by the Navy (Daily News 16 May 2005). In 2004, Private Fisheries Agreements (PFAs) were the basis for all longliner and tuna seiner fishing operations in the Tanzanian EEZ. The PFAs specify the financial compensation and license conditions associated with access to the re- source. There are significant deficiencies in PFAs regarding responsible man- agement of the Fisheries. Specifically the PFAs for these fleets have not · Set limits on the catches that can be made by individual vessels or fleets; · Generated the quantity and quality of Fisheries information required to re- liably assess the catch and effort of foreign fleets; · Provided reliable estimates of the value of fish removals from the EEZ; · Achieved a measurable level of compliance of the fishing fleets or deterred IUU fishing activity; · Enhanced the coastal state's understanding of the EEZ pelagic Fisheries nor fostered the development of a Tanzanian capacity to contribute to manage- ment of these Fisheries in a rational manner; · Contributed to establishing an accurate EEZ catch history that could be used to justify future resource sharing formulas in the RFOs; and · Provided the basis for a cooperative partnership between coastal and fishing states wherein the financial benefits are disbursed equitably between the relevant parties and the resource is exploited in a sustainable manner. With regard to the artisanal Fisheries, which are operating at the limits of the traditional technologies, there is evidence that catches are decreasing across the range of reef and inshore species. The destruction of habitats by dynamite fish- ing and poisoning causes concern for the future of these Fisheries. Artisanal Fisheries are socially and economically important activities for the coastal communities, but the systems put in place by the Authorities for monitoring the conservation of these areas appear inadequate. From the fragmentary data available, it seems that coastal and prawn Fisheries are surpassing MSY levels. Increasing effort would appear to be impracticable without causing possible terminal damage to the stocks. The EEZ fishery is presently an unknown from a scientific standpoint, and highly seasonable in nature. In such a scenario, sustainable management of the Fisheries resources is a sine qua non for sustaining or developing future economic growth in the sector. This would include research to establish stocks, regulation, control, monitoring and Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 77 surveillance. The resource rent generated by the sector would need to be high enough to allow re-investment into its protection against over-exploitation. Resource rents Estimates of resource rents from marine Fisheries computed from license fees as percentage of value of revenue are provided by Chopin (2005). Table D and Table E in Annex 1 show the detailed calculation of resource rent for longliner and tuna seiner operations. For longliners, estimates are that the gross resource rent is approximately 2.2 percent. This is slightly less than half of what might be expected in a western industrial fishery. Calculations for tuna seiners vary somewhat more depending on different catch scenarios. See Table D and Table E in Annex 1 for detailed estimates of resource rents. In summary, while the current license fee arrangements of PFAs in the EEZ generate a not insignificant amount of revenue, the level is too low to result in a reasonable return to capture resource rent (>5 to 7 percent of gross revenue) to the coastal state. The PFAs as currently offered to foreign fishing enterprises are hence considered untenable. 4.7 Policy framework for Marine Fisheries Marine Fisheries are governed by national legislation as well as international conventions. The international conventions include the UN Convention on the Law of the Sea (UNCLOS) and the related Fish Stock Agreements. These im- pose roles and responsibilities on the coastal state (Tanzania) as well as on the Distant Water Fishing Nations (DWFN). Therefore, illegal activities in the ma- rine context include illegal activities of vessels of DWFN which contravene national legislation and international obligations. The 1970 Fisheries Act provides the basis for the national policy framework. It was reviewed in 2002 with the aim to "ensure promotion of sustainable Fisher- ies, ensuring adhering to regulations and conservation of resources" (URT, 2004c: 152). The Act is currently re-written and expected to be back in force by 2005. On the mainland, 80 percent of the fishing effort is freshwater based and lake Fisheries (and export) dominate the policy agenda with marine Fisheries somewhat neglected. For example the Fisheries master plan (funded by JICA), is heavily weighted towards the lakes and rivers but there is a marine compo- nent (10 percent in real terms), which is designed to fit traditional craft with donated outboard motors and better net technology. The effect of these meas- ures on an already fragile stock situation along the coast has not been assessed. It should be acknowledged, however, that there are projects in the pipeline, which aim to deal more comprehensively with marine fisheries: An example is the World Bank/ Global Environmental Facility (GEF) supported ` Marine and Coastal Environmental Management Project' (MACEMP) designed to Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 78 strengthen governance of the EEZ and the near-shore marine environment as well as supporting livelihoods in the coastal communities. The fact that Fisheries is administered by two separate Fisheries Departments (Mainland and Zanzibar) increases transaction cost and the need for co- ordination and exchange of information. There is duplication of effort and nei- ther entity can afford to carry out surveillance unilaterally. The 1998 Deep Sea Fishing Authority Act by contrast has status as a United Republic Law and so applies equally to both Mainland and Zanzibar. The im- plementation of the Act is however stalled, due to reservations from the Zanzi- bar Government. An important vehicle for the management of the EEZ is there- fore missing. Dialogue is however ongoing between the two sides of the Union with a view to establish a common governance regime for EEZ. Efforts to regulate resource access in EEZ Fisheries are undertaken by each FD unilaterally by issuing licenses. However, enforcement is too lax and the fiscal regime inadequate to enable resource rent capture. There are no onboard ob- servers on these vessels although it is a condition of license. Moreover, there are no enforcement of catch reporting or vessel arrival or departure. The license agreements do not result in reasonable amount of catch and effort data flowing from the fishing state to the coastal state to allow it to perform its responsibili- ties for managing the Fisheries taking place in its EEZ. The existing licenses have failed to result in either fleet submitting comprehensive reports on their fishing activities within the EEZ. There are no catch limits attached to the licenses, allowing vessels to take as much fish as is available with scant information being returned to the coastal state. There is anecdotal information about irregularities in the Zanzibar licens- ing system. Supposedly some vessels are being registered through Muscat, Oman with the fees escaping the Zanzibar authorities. A new Fisheries agreement (FA) is currently being negotiated between Tanza- nia and the EU. The FA is a vehicle to regulate the sector and provides a reve- nue source to GoT. However, currently the average catch taken as a basis for the negotiations has been undervalued and negotiations are difficult in a sce- nario where for EEZ pelagic fishery, there is no baseline study of stocks, no figures for Maximum Sustainable Yield or Total Allowable Catch. Monitoring, Control and Surveillance Presently, Tanzania's Monitoring Control and Surveillance (MCS) capacity is dependent on the MCS SADC project. This is a regional programme that aims to build national and regional capacity in marine fishery monitoring, control and surveillance in SADC countries. Prior to 2002, fewer than 13 vessels were licensed to fishing in the Tanzanian EEZ and it was not considered necessary or cost effective. However, with the increasing effort over the last two years, the need for MCS has become appar- ent. Under the SADC programme, approximately 6 hours flying time per week Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 79 is required to search most of the EEZ. With the support of this project it was possible to apprehend nine suspects and three ships in 2003 through air surveil- lance. In 2003 a total of 26 government patrols were conducted along the coastal strip and in the EEZ zone (URT, 2004c). This is a reduction compared to 2002, when 135 patrols were conducted and 53 suspects were apprehended for fishing without license and undertaking dynamite fishing. In 2001, the FD reported that 90 rounds of patrol were undertaken and about 50 illegal fishermen were caught with 4,130 mesh fishnets, 250 purse seines and 600 kg of fish. Likewise, efforts to strengthen monitoring of commercial fishing activities were under- taken, whereby 27 observers were assigned to monitor fishing activities carried out by large fishing companies. Although some MSC activities are taking place, efforts are patchy. Effective MCS is difficult for the coastal state as long as the fishing state is not providing adequate quantity and quality of data. Vessels need to provide more compre- hensive and timely information on all fishing activities. 4.8 Recommendations Based on the analysis provided in the previous Sections, some of the key rec- ommendations include: Putting in place a regulatory framework and sound governance regime for marine Fisheries, comprising the EEZ and near-shore Fisheries; Strengthen capacity for MCS to address the illegal Fisheries; Punitive measures that are real deterrents to control unsustainable fishing practices, whether in EEZ or nearshore Fisheries; Safeguard rights and livelihoods for coastal communities, through for ex- ample demarcation of a Community Territorial Sea; Fisheries Sector review to assess the economic and social, ecological and fiscal perspectives, policy options etc. To inform policy makers and influ- ence the strengthening of the regulatory framework. Capacity building for research, MCS and assessing economic and fiscal as- pects. The Fisheries Agreement should support the scientific effort of establishing precisely what the EEZ Fishery will support. 'Spin-off' effects through re- lated on-shore development could also be promoted. Sponsoring some form of EEZ inspectorate in patrol terms could help to build up a much more accurate picture of what is available in the sector. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 80 Operational feedback from sea inspections is a vital element of any moni- toring effort. Investigate the potential for exports of (various) marine products and value adding of these products in order to promote growth in the coastal zone. Although some of the above recommendations are being addressed by the on- going SADC Regional MCS Programme, or will be addressed by the proposed MACEMP project, project activities are of limited duration and it will be im- portant that these recommendations are institutionalised. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 81 Annexes Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 82 Annex 1. Background data Table A. Forest distribution by regions Name of Region Forest estate in 1000 ha Group 1: ha > 400.000 1. Tabora around 540 2. Rukwa around 500 3. Lindi around 450 Group 2: 200 000< ha > 300.000 4. Arusha around 250 5. Coast around 250 6. Dodoma around 230 Group 3: 150.000< ha > 200.000 7. Tanga around 200 8. Morogoro around 190 9. Kigoma Group 4: 100.000 < ha > 150.000 10. Mbeya around 120 11. Singida around 120 12. Shinyanga around 110 Group 5: 50.000 < ha >100.000 13. Ruvuma around 90 14. Mtwara around 90 15. Iringa around 80 Group 6: ha < 50.000 16. Kagera around 30 17. Kilimanjaro around 20 18. Mwanza around 20 19. DSM 0 (Source: Author, based on Forestry on Figures 2002). Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 83 Table B. Actual annual expenditure of FBD, 2001/02 ­ 2003/04 Year Amount in 1000 Tsh and % by expenditure Type Recurrent % Development % Total % 2003/04 7,633,912 83.448 1,514,220 16.552 9,148,132 100 2002/03 4,897,656 70.880 2,012,171 29.120 6,909,827 100 2001/02 4,300,955 96.743 144,800 3.257 4,445,755 100 (Source: Author based on MNRT 2004) Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 84 Chart A Forestry: Development Spending, 2003/04 2003/04 Development Spending 8% 0% Capacity Building Fixed Cost Service Delivery 92% (Source: MNRT, 2004) Chart B Forestry: Recurrent spending, 2003/04 2003/04 Recurrent Spending 11% Capacity Building 28% Fixed Cost 61% Service Delivery (Source: MNRT, 2004) Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 85 Chart C Allocation of recurrent budget, Forestry, 2002/03 -2003/04 2 year trend in Recurrent spending 9,000,000 8,000,000 7,000,000 6,000,000 tsh 5,000,000 Recurrent Spending 2003/04 4,000,000 Recurrent Spending 2002/03 1000 3,000,000 2,000,000 1,000,000 0 Capacity Fixed Service Total Building Cost Delivery (Source: MNRT, 2004) Chart D Allocation of development budget, 2002/03 ­ 2003/04 2 year trend development spending 2,500,000 2,000,000 Development Spending tsh 1,500,000 2003/04 Development Spending 1000 1,000,000 2002/03 500,000 0 Capacity Capital Service Total Building Investment Delivery (Source: MNRT, 2004) Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 86 Table C. Volume and Value of Exports from the Forestry Sector, 1998 to 2003 1998/99 1999/2000 2000/2001 2001/02 2002/2003 Product Vol./Weight Value Vol./Weight Value Vol./Weight Value Vol./Weight Value Vol./Weight Value US$(000) US$(000) US$(000) US$(000) US$(000) Logs-TeakPaurosa(M3) 3897.0 718.0 1956.7 585.6 1724.4 453.4 934.6 299.0 - - Timber(M3) 8066.0 684.5 103.7 710.0 511.0 15.5 1230.0 271.0 4122.9 415.0 Ebony(M3) 122.0 1193.2 848.2 1355.5 62.8 294.3 38.5 368.4 57.8 528.7 Floorboards(M3) 68.0 128.4 6546.2 87.1 157.2 114.1 66.0 46.2 486.4 190.9 Woodcarvings(Pcs.) - - 258494.0 1210.7 83746.0 163.5 56254.0 114.5 165696.0 1201.0 Otherproducts - - 21.2 9.2 43.0 4.8 - - - - Treeseeds(Kg) - - - - 69.0 18.6 80.5 24.2 - - Beeswax(Tons) - - - 2405.6 431.0 1044.6 109.0 306.7 600.0 1800.0 Honey(Tons) - - - 167.7 12.0 5.5 1.8 2.4 12.0 12.0 Total 2724.1 6531.4 2114.3 1432.5 4147.6 (Source NBS, Economic Survey 2003) Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 87 Table D. Estimation of Gross Resource Rent for Tuna Seiners Gear type tuna seiner Fleet catch (t) 8,000 16,000 24,000 36,000 48,000 No. vessels buying licenses 39 39 39 39 39 No vessels fishing in the EEZ 39 39 39 39 39 Days spent in the EEZ 25 25 25 50 50 Nominal fishing effort 975 975 975 1950 1950 Vessel Catch per day (t) 8.2 16.4 24.6 18.5 24.6 License fee (US$ / t / yr) 0 0 0 0 0 License fee (US$ / t / yr) 17,550 17,550 17,550 17,550 17,550 Registration fees USD 1,950 1,950 1,950 1,950 1,950 Compensation 0 0 0 0 0 License fee paid (US$) 760,500 760,500 760,500 760,500 760,500 Gross Fee per tonne (US$/ t) 95.063 47.531 31.688 21.125 15.844 Revenue from seining @ 910 7,280,000 14,560,000 2,184,000 3,315,000 43,680,000 (US$/t) Revenue from seining @ 1040 8,320,000 16,640,000 24,960,000 37,440,000 49,920,000 (US$/t) Revenue from seining @ 1,170 9,360,000 18,720,000 28,080,000 42,120,000 56,160,000 (US$/t) License as % of Value @ 9.1% 4.6% 3.0% 2.0% 1.5% 1,040 (US$/t) (Source: Chopin, 2005) Interpretation: The gross resource rent ranges from 9.1 percent for 8,000t catch to 1.5 percent for 48,000 tons catch. The 5 percent gross resource rent is reached when the catch is limited to 14,300 tons. It can also be reached by increasing the license fee to US$ 42,000 and allowing a total catch of 28,000 tons. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 88 Table E. Estimate of Gross Resources Rent for Longliners Fleet catch 9225 18450 No. vessels buying licenses 123 123 Days spent in the EEZ 50 100 Nominal fishing effort 6150 12300 Vessel Catch per day t (tunas) 0.5 0.5 Vessel Catch per day t (others) 1.0 1.0 License fee (US$/tonne) 0 0 License fee (US$) 1775 1775 Registration fees (US$) 195 195 Compensation 0 0 License fee paid 2,878,200 2,878,200 Gross Fee per tonne (US$/tonne) 312 156 Revenue from fishing (@ 9,750 US$ /t) 89,943,750 179,887,500 Revnue from seining (@ 11,050 US$ /t) 101,936,250 203,872,500 Revenue from seining (@ 12,090 US$/t) 111,530,250 223,060,500 License as % of value (@ 9,750 US$/t) 2.8% 1.4% (Source: Chopin, 2005) Interpretation: To reach a 5 percent gross resource rent, the catch would either have to be limited to 4000 tons or the license fee in- creased to US$ 42,000 allowing a total catch of 9000 toms. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 89 Table F.Exports of Fish Products, Marine Resources, 1998-200 1998 1999 2000 2001 2002 2003 Quantity Value Quantity Value Quantity Value Quantity Value Quantity Value Quantity Value Marine Product Kg US $ Kg US $ Kg US $ Kg US $ Kg US $ Kg US $ Prawns 223,639 390,976 1,154,181 5,127,642 1,218,181 5,644,876 1,175,095 5,850,876 1,248,008 6,618,047 1,672,018 5,979,557 Lobsters 25,999 255,146 161,962 818,512 63,093 365,455 98,793 804,425 119,261 1,189,357 146,286 1,443,442 Live Lobsters .. .. .. .. 14,799 186,581 .. .... 63,589 852,366 165,580 2,050,997 Crabs 17,882 99,295 4,819 24,697 29,880 144,373 51,719 294,864 487,950 182,227 40,139 118,736 Live Crabs .. .. .. .. 67,261 134,623 .. .. 128,242 806,503 168,002 925,586 Octopus 595,783 1,055,637 574,522 1,357,381 492,763 1,177,630 275,286 667,788 355,310 1,023,762 1,603,456 5,044,536 Squids 12,567 30,494 35,005 107,572 26,750 66,194 55,164 282,993 59,359 182.523.61 298,398 1,239,705 Sea shells 154,893 29,859 250,006 85,539 345,164 105,501 433,643 120,345 253,189 113,763 895,748 380,010 Beche de mer 872,926 21,722 93,439 255,514 124,478 463,883 49,947 185,483 6,800 5,409 12,301 40,330 Fish maws .. .. .. .. .. .. .. .. 1,081,123 2,919,410 1,353,604 5,774,464 Fish offals 199,974 389,169 90,050 181,575 1,171,660 717,666 120,804 286,497 .. .. 123,900 172,666 Shark jaws 254 4,274 .. .. .. .. .. .. .. .. .. .. Marine Fish Fillet 10,321 24,248 .. .. .. .. .. .. .. .. .. .. Sea shells .. .. .. .. 5228 pcs 5,680 .. .. 253,189 113,736 .. .. Sub-Total 2,114,238 2,300,820 2,363,984 7,958,430 3,554,028 9,012,463 2,260,451 8,493,272 4,056,020 13,824,580 6,479,432 23,170,030 % of total fisheries export 4.8 3.2 8.2 12.9 8.9 14.9 5.4 8.9 13.1 15.2 16.1 18.3 (Source: Economic Surveys 2001-2003) Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 90 Annex 2. References Ashley, C., Mdoe, N., and Reynolds, L. (2002): Rethinking Wildlife for Livelihoods and Diver- sification in Rural Tanzania: A Case Study from Northern Selous. LADDER Working paper No. 15, Norwich, University of East Anglia. Aku O'kting'ati, Monela G.C. & Nyella H. (2000). Contribution of Kilimanjaro Regional For- est Sector to the Economy of Tanzania (1990-1998). Barnett, R. (ed.) 2000: Food for Thought: The Utilization of Wild Meat in Eastern and Southern Africa. Nairobi, TRAFFIC East/Southern Africa. Buys, Hans; Moshi, Izzet and Mariki. (1996) Long-term Financing of Forestry, Report prepared for FCMP. Chatterjee & Mushi (1994) Study on improvement and monitoring royalty collection systems, Report no.1., Evaluation, reformulation and design of recommendations of the Silviconsult study. Chopin, F. (2005): Review and assessment of the DRAFT United Republic of Tanzania ­ Euro- pean Community Fisheries Agreement. Draft. Restricted Circulation. DFID (2003): odi Natural Resource Perspectives, Just Wildlife? Or a Source of local Develop- ment? Number 85, April 2003. DPG (2005): Development Partners Group Tanzania, Brief on Issues Pertaining to Tourist Hunt- ing, April 2005. Emmanuel, Juma (2001). Tourism in Mikumi National Park (MNRP) Morogoro, Tanzania. Spe- cial Project. Eustack, Mtui Boniface (2004). Assessment of the contribution of agroforestry to poverty alle- viation in Lushoto District. M.Sc. thesis. Gillingham, S. (1998): Giving Wildlife Value: A Case Study of Community Wildlife Manage- ment Around Selous Game Reserve, Tanzania. Dissertation submitted to the University Col- lege of Cambridge. Homewood, K., Lambin, E.F>, Coast, E., Kariuki et al (2001): Long Term Changes in Seren- geti-Mara Wildbeest and Land Cover: Pastoralism, Population or Policies? PNAS 98 (22). Idda, Makawia Amani (2003). Contribution of Agroforestry to human nutrition. A case of Arusha Region. M.Sc. thesis. Issango, Juma (2001). Honey production and sales along Dar es Salaam ­ Morogoro High way. Special project. IUCN (2000): The effectiveness of Trade Measures Contained in the convention on international trade in endangered species of wild fauna and flora, UNEP Economics, Trade and Environment Unit. Kaduvage, B. L. (2000) Study on the present status and future prospects for mechanical and chemical industries in Tanzania, Project report. Kallonga, Emmanual; Rodgers, Alan; Nelson, Fred; Ndoinyo, Yannick and Rugemeleza, Nshala (2003) Reforming Environmental Governance in Tanzania: Natural Resource Management and the Rural Economy, non-commissioned paper presented at the Inaugural Tanzania Biennial Development Forum, 24-25th April 2003, Dar es Salaam. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 91 Kagya, M. A. (2004): Gums and Resins production and marketing in Tanzania. FAO project report, Strengthening the Production and Quality Control of Gums and Resins in Africa, Sep- tember 2004. Kagya, Monica Andrea Nkunja (2002). Contribution of non-wood forest products to household economy and welfare of women in Meatu District, Shinyanga Region, Tanzania. M.Sc. thesis. Khamis Fatma Ali (2003). Financial Analysis of Ecotourism industry. A case study of Jozan Chwaka Bay conservation area, Zanzibar. Special Project. Kobb, Daniel (1999) Forestry royalties in Tanga Region: Paper vs. Reality, Paper prepared for East Usambara Catchment Forestry Project, Natural Resources Management and Buffer Zone Development Program, Tanga Coastal Zone. Koppers, B. (1998) End of Assignment Report, Danagro. Kowero (1990) Some Aspects of Tanzanian Forest Royalties, TFAP, Working Paper no.26, FBD. Kowero (1991) "Management and utilisation of forest estate in Tanzania: some policy issues", Journal of World Forest Resource Management, vol.8. (10), pp.15-27. KRCD (2005): Establishment of a socio-economic monitoring system for WMAs. Mission re- port. Unpublished. Luoga,E.J.; Witkowski, E.T.F. and Balkwill, K. (2000). "Economics of charcoal production in Miombo woodland of eastern Tanzania: some hidden costs associated with commercialization of the resources". Ecol.Econ.35 (2000) 243-257. Lusambo, L.P. (2002). Socio-Economic Analysis of Land use Factors causing Degradation and Deforestation of Miombo woodlands. Case study of Kilosa District, Tanzania. M.Sc. Thesis. Lyimo Jacobo Ophoro (2002). Production economics of Transmission poles in Tanzania. Spe- cial project. Mabugu, R. and Mugoya, P. (2001): Financing, Revenue sharing and Taxation issues in Wildlife Management Areas. Report prepared for USAID, October 2001. Makundi, W.R. and Aku O'kting'ati (1991). Carbon flows and economic evaluation of mitiga- tion options in Tanzania's forest sector. Malimbwi, R.E.; S.Misana, G.C. Monela, G. Jambiya & E. Zahabu (2000). Socio-economics of charcoal extraction to the forest resources of Tanzania: The case study of Kitulangalo Area, Tanzania. Mkanta & Chimtembo (2002) Towards Natural Resource Accounting in Tanzania, A study on the Contribution of Natural Forests to National Income, CEEPA Discussion Paper Series. MNRT (2001a). National Forest Programme in Tanzania (2001-2010), Ministry of Natural Re- sources and Tourism . Forestry and Beekeeping Division. MNRT (2001b). National Beekeeping Programme (2001-2010), Ministry of Natural Re- sources and Tourism, Forestry and Beekeeping Division. MNRT (2002) Forestry in Figures, Ministry of Natural Resources and Tourism, Forestry and Beekeeping Division. MNRT (2004) MTEF Budget of the Ministry of Natural Resources and Tourism, Vote 69 Monela, G.C., S.A.O. Chamshama, R. Mwaipopo and D.M. Gamassa (2004). A study on the social, economic, and environmental impacts of forest landscape restoration in Shinyanga Re- gion. Report. MNRT & IUCN. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 92 Mushi, J.A. (1999) Evaluation of Revnue Collection and Monitoring Systems in Singida, Ta- bora, Shinyanga, and Mwanza Regions, Report prepared for Forest Resources Management Project (FRMP), IDA Credit no. 2335-TA. Mwanahija, Salehe Shalli (2003). Contribution of agroforestry Vs Rural households food inse- curity in Kibaha Disrict, Coast Region, Tanzania. M.Sc. thesis. Ngaga Y.M, Chamshama S.A.O. & Monela G.C. (2003). Resource Economic Analysis of Catchment Forest Reserves in Tanzania. Report. Ngaga, Y. M. (1990). Present consumption and future requirement of sawn wood in two urban centers of Tanzania. M.Sc. thesis. Ngaga, Y. M. (1998). Analysis of Production and trade in Forest products of Tanzania. Doctoral Scientiarium Thesis. Department of Forest Sciences, Agricultural University of Norway. Noah Edward (2002). Impact of production and consumption of wood fuel in Miombo wood- lands of Makanya Division, in Kilimanjaro Region. Special project. Noel, Maeda (2001). Statistics of game hunting and revenues; and its implication to the econ- omy of Tanzania 1997-2000. Special project. Norconsult (2002) The True Cost of Charcoal, Norconsult Ltd., May 2002. Paullo, Teddy (2002). Market survey and economic values of useful edible mushrooms con- sumed in Iringa Municipal. Special Project. Pearce (1991) Philbert, Michael (2002). Impact of tourism on employment. A case study of Kilimanjaro Na- tional Park (KINAPA). Special project. Robert, Modest (2003). Economic potential of trophy export in Tanzania. Special project. Roe, D., Mulliken, T.; Milledge, S., Mremi, J., Mosha, S. and Grieg-Gran, M. (2002): Making a killing or making a living? Wildlife trade, trade controls and rural livelihoods, Biodiversity and Livelihoods Issues No. 6, Traffic, IIED. Rutenge, Catherine (2004). Assessment of the influences of live small animals trade in Arusha Region. Special Project. Ruitenbeck, J.; Hewawasam I.; Ngoile, M. eds. (2005): Blueprint 2050, Sustaining the Marine Environment in Mainland Tanzania and Zanzibar, The World Bank. Salmi and Monela (2000) Study on Financing in Forestry, Formulation of National Forest Pro- gramme in Tanzania, Final Report by Salmi and Monela, Dar es Salaam, November 2000 for Ministry of Natural Resources and Tourism (MNRT), Forestry and Beekeeping Division. Sathaye, J. & W.R.Makundi (1995). Biomass and Bio-energy; special issue forestry and climate change. Special issue. Sekandende, A.N. (2001). Economic benefits of wildlife-based tourism. A case study of Tarangile National Park. Special project. Severin, Ansbert (2001). The economic implication of tourism towards the economy of Tanza- nia. Special project. Silviconsult (1991) Forest revenue collection in Tanzania. FBD, World Bank and ODA. Skage and Naess (1994) Pitsawing and sustainable forest management, a case study of ecologi- cal and social considerations from Morogoro District. MSc Thesis in Forestry, Dept. of For- estry, AUN. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 2 93 TANAPA (1999): Review of Efficiency and Effectiveness, Draft Report of 6 December 1999, For internal review. TANAPA (no date): Maximizing Revenues in Tanzania National Parks: Towards a better under- standing of park choice and nature tourism in Tanzania. TaTEDO (2003). Scaling-up Dissemination, Partnership and Networking for Increased uptake and use of Sustainable Energy Technologies. Annual Report 2002/2003. Theostina, Anthony Lema (2003). The role of NTFPs in household food security and women income in Morogoro rural. Special Project. TIC (2001) Turpie, J.K. (2000): The use and value of natural resources of the Rufiji floodplain and delta, Rufiji district, Tanzania. Rufiji Environmental Management Project, Dar es Salaam, Technical Report No. 17 URT (1997) Ruvuma Region Socio-economic Profile. Dar es Salaam, Tanzania. pp 1-57, United Republic of Tanzania. URT (2002) Poverty and Human Development Report 2002, United Republic of Tanzania. URT (2004a) Budgetary Speech (2004/2005) by the Minister for Natural Resources and Tour- ism, United Republic of Tanzania. URT (2004b) Public Expenditure Review of Environment, Financial Year 2004. United Republic of Tanzania. URT (2004c), Economic Survey for 2003, Planning and Privatisation Commission, Dar es Sa- laam, Tanzania. Utz, R. (2005). Tanzania, Recent Growth Performance and Prospects. Walsh, M.T (2000): The development of community wildlife management in Tanzania, Lessons Learned from the Ruaha Ecosystem. Paper presented to the conference on African Wildlife Management in the New Millenium College of African Wildlife Management, Mweka, Tanza- nia, 13-15 December 2000. The World Bank Study on Growth and Environment Links for Preparation of Country Economic Memorandum (CEM) Part 3: Success Stories of Growth? Mining, Freshwater Fisheries and Tourism Final report May 2005 The World Bank Study on Growth and Environment Links for Preparation of Country Economic Memorandum Part 3: Success Stories of Growth? Mining, Freshwater Fisheries and Tourism Final report May 2005 Report no. 3 Issue no. 2 Date of issue 18 May 2005 Prepared KEP Checked TNH Approved Table of Contents 1. Introduction 1 2. Mining 2 2.1. Questions Arising in the Context of the CEM 2 2.2. Data and Definitions 3 2.3. Is the Growth Sustainable? 4 2.4. Employment and Income 5 2.5. Social Impacts 12 2.6. Environmental Impacts 14 2.7. Conclusion 16 3. Freshwater Fisheries 18 3.1. Questions Arising in the Context of the CEM 18 3.2. Trends in Catches and Revenue 18 3.3. Contribution to Export Earnings 20 3.4. Employment and Income 21 3.5. Sustainability 23 3.6. Conclusion 25 4. Tourism 26 4.1. Tourist Arrivals and Earnings 27 4.2. Contribution to Exports and GDP 29 4.3. Tourism Attractions 31 4.4. Employment and Income 37 4.5. Environmental Impacts 38 4.6. Tourism and poverty reduction 40 4.7. Recommendations 41 Annexes 42 List of Boxes Box 1 Tanzanite production 4 Box 2 Key Figures on Earnings from Tourism 26 Box 3 Figures on tourist visiting Tanzania's National Parks 32 List of Charts Chart 1 Employed persons by employment status 6 Chart 2 Employment in large-scale mining sector and gold production 7 Chart 3 Average monthly income of paid employees by age and selected industries 8 Chart 4 Mean monthly income of self-employed (without employees) by selected industries 2000/01 10 Chart 5 Statutory taxes and other contributions paid to the Government - Commercial mines in Tanzania 12 Chart 6 Income of self-employed women by selected industry, absolute and as ratio to self-employed men, 2000/01 14 Chart 7 Freshwater and marine revenues as share of total fisheries revenues 20 Chart 8 Proportion of freshwater fish export earnings on total exports, 1996 ­ 2003 20 Chart 9 Fishermen and boats in freshwater Fisheries, 21 Chart 10 Fishnet production and import 1998 ­ 2003 22 Chart 11 Freshwater Fisheries, CPUE and total catch, 1995-2003 23 Chart 12 Number of tourists, 1990 - 2003 28 Chart 13 Revenue trend from tourism services for 1990 - 2003 28 Chart 14 Tourism arrivals, tourism revenue and revenue per tourist ratio growth rate for 1990 ­ 2003 29 Chart 15 Proportion of tourism earnings to total exports for the year 1990 - 2003 30 Chart 16 Proportion of tourist earnings to the GDP for the year 1990 - 2003 31 Chart 17 Proportion of revenue from National Parks to the total tourism revenue for the year 1990 ­ 2001 32 Chart 18 Number of foreign visitor to each park from 1991 to 1999 34 Chart 19 TANAPA 2001/2002 Primary Revenue 35 Chart 20 Number of employees in the tourism sector 37 Chart 21 Employment rate in the tourism sector 38 Chart 22 Income from payments by one of four tour operators, Olosokwan Village and Ngorongoro District Council, 1999/2000 - 2002/03 41 List of Tables Table 1. Freshwater catches and revenue collected for the year 1993 ­ 2003 19 Table 2. Number of fishermen, catches and revenues, 1995-2003 24 Table 3. Tourists arrivals and revenue collected for the year 1990 ­ 2003 29 Table 4. Projected Revenue and expenditure by park for 1999/2000 33 Table 5. Visits to selected historical sites for the year 2000 ­ 2001 37 List of Abbreviations and Acronyms BBC British Broadcasting Corporation CEM Country Economic Memorandum CPUE Catches Per Unit Effort DFID Department for International Development EIA Environmental Impact Assessment EIU Economist Intelligence Unit EMP Environmental Management Plan ENR Environment and Natural Resources EU European Union FOB Free On Board GDP Gross Domestic Product GoT Government of Tanzania HIV/AIDS Human Immunodeficiency Virus/Acquired Immunodeficiency Syndrome IMF International Monetary Fund ISO International Standards Organisation LFVRP Lake Victoria Fisheries Research Project MSY Maximum Sustainable Yield NBS National Bureau of Statistics NCAA Ngorongoro Conservation Area Authority NEMC National Environment Management Council NP Nile Perch NSGRP National Strategy for Growth and Poverty Reduction TANAPA Tanzania's National Parks TSH Tanzania Shilling UNIDO United Nations Industrial Development Organisation URT United Republic of Tanzania US$ United States Dollar Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 1 1. Introduction This paper presents Part 3 of a report on growth and environmental links com- missioned by the World Bank to serve as an input to the Country Economic Memorandum (CEM), which is currently being prepared. The Terms of Refer- ence for the assignment are provided in Annex 1 to Part 1 of the report. Environment and Natural Resources (ENR) are discussed in the CEM in the context of the analysis of Mining (mainly gold), Fisheries and Tourism as the main drivers of economic growth. The focus of this report is to look beyond the direct economic implications of recent growth in these sectors, and draw attention to the wider implications and externalities. For example, whereas Mining no doubt features as one of the growth poles for the Tanzanian economy and the export sector, the recent growth has not been a strong driver of local employment. Moreover, as this re- port will argue, the consolidation of large-scale Mining activities may have se- rious consequences for the environment as well as the wider social infrastruc- ture of the communities affected by Mining activities. Likewise, the report points to trade-offs associated with recent growth in the other focus sectors. There are indications for example that significant growth in freshwater Fisheries is unsustainable. The rapid growth in the Tourism sector may also have negative implications for wildlife and protected areas. This report is based on a secondary data analysis. The sources of information are statistics from the relevant line ministries, the National Bureau of Statistics, the Bank of Tanzania and various books and scholarly literature. The examined literature is contained in Annex 3. The report is structured as follows: Chapter 1 introduces the subject and the methodology. Chapters 2 to 4 describe recent growth performance in the re- spective sectors and analyse the wider implications of these growth stories. Chapter 2 focuses on Mining, 3 on Fisheries, and 4 on Tourism. Chapter 5 summarises and concludes the analysis. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 2 2. Mining Tanzania is rich in mineral resources. In recent years, the mineral industry has produced copper, gold, silver, and rolled steep products, and such industrial minerals as calcite, diamond and other gemstones, gypsum, phosphate rock, salt, silica sand, and soda ash. Deposits of cobalt, copper, iron ore, natural gas, nickel, and titanium are also known to occur in Tanzania. In strictly economic terms, mineral resources are probably the most important natural resource in Tanzania. The Mining sector in Tanzania mainland grew by 27.4% in 1998 as a reflection of massive foreign investment in the large-scale Mining sector. In subsequent years, growth has been lower, but still remains double digit. Chart A in Annex 1 demonstrates that growth in the Mining sector has been far higher than what has been observed in the main sector of the econ- omy, Agriculture. The growth in the sector reflects a significant increase in the annual production of gold in the large-scale Mining sector (see Chart B in Annex 1) that has, fol- lowing the liberalisation of the sector in the late 1990s, increased to a level of more than 45,000 kg in 2003 (EIU, 2004: 30). The growth in gold production has led to a significant increase in gold exports, which, as Chart C in Annex 1 describes, has triggered a corresponding increase in total export earnings for Tanzania. However, Mining as a sector is still small, it accounts for only 1.9 percent of GDP (2003), up from a level of 1.4 percent in 1995 (IMF, 2004). Hence, despite its rapid growth, general GDP growth has not been significantly affected by the growth in the Mining sector (see Chart A in Annex 1). This chapter will first explain a few questions arising in the context of the dis- cussion of the gold boom in the CEM. Then it will provide a brief explanation of the methodology. Then the question of sustainability is investigated, fol- lowed by the employment effect of the gold boom. This is followed by a para- graph each on the environment and social impacts. The last section summarizes conclusions. 2.1. Questions Arising in the Context of the CEM As the draft CEM points out under the heading `Gold and Foreign Aid ­ Boom or Curse?', the impact of these two developments [gold production and exports and foreign aid inflows, red.] on economic growth and on income-poverty (...), Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 3 is ambiguous (p.14). Whereas the CEM explains this ambiguity with the "Dutch disease effect" mainly in relation to the balance of payment, a number of other important ques- tions are left out in the current draft. These are: · Given that gold is a non-renewable natural resource and the existing re- serves are exhaustible, how long will the country's reserves last? Or put differently, can the current growth rates be sustained? · Has the growth in the Mining sector translated into increased employ- ment opportunities in the Mining sector per se and in the local commu- nities? If so, what kind of employment has been triggered? · What are the environmental implications of the growth in the Mining sector? · What are the wider social implications of the growth in the Mining sec- tor for HIV/AIDS and gender in the local communities? 2.2. Data and Definitions Data availability is a constraint to the detailed analysis of the four questions above due to the following reasons: · The growth and employment data, generally, does not distinguish be- tween small- and large-scale Mining operators and so makes it impossi- ble to attribute growth effects to either sub-sector. To compensate for this lack of data a number of proxies will be used; · Growth and employment data is not available at a sufficiently low geo- graphical aggregation level. Accordingly, it is difficult to observe and compare growth patterns in Mining areas with performance in non- mining areas; · For some indicators, only limited time-series data is available, which makes it difficult to determine the effects of the introduction of large- scale Mining in Tanzania; and · Very limited reliable data is available on the wider environmental and social aspects of Mining. Therefore, this chapter cannot conclude authoritatively on the raised questions. The purpose is mainly to assess the existing data and literature with a view to make provisional findings and conclusions. Further data collection and analysis will be required to corroborate the findings. The Mining sector will for the purposes of this paper include all activities di- rectly related to mineral Mining but will focus primarily on gold, which ac- counts for the bulk of the sector's export earnings. Where detailed data is not available for mineral Mining sector, aggregate data for the Mining and quarry- ing sector will be used as a proxy. A key distinction will be made between large- and small-scale operators with the latter category encompassing artisanal miners. The working definitions, based on Mwaipopo et al. (2004), are as fol- lows: Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 4 · Large-scale operators are typically foreign owned capital-intensive opera- tions applying sophisticated imported equipment; and · Small-scale operators, by contrast, are labour-intensive operations, domes- tically owned and applying less sophisticated equipment. Small-scale min- ers include a wide spectrum of users which differ by degree of capital- intensity, but they can be grouped into two main areas: - Licensed operators applying some measure of sophisticated equipment and a limited workforce; and - Artisanal miners, labour-intensive operations applying very basic equipment, and typically operating without license and employees. Hence, artisanal miners typically operate in the informal sector. 2.3. Is the Growth Sustainable? Mineral resources are non-renewable natural resources. Since their stocks are exhaustible, the question is, whether the Mining sector will remain a growth pole for the economy in the future. There is no data or authoritative estimates on the size of mineral reserves in the country. Hence the above question cannot be clearly answered. Even if new stock are discovered, the question is, whether these can be extracted on a profitable basis. The review of literature provides the following estimates or `believes' on min- eral stocks and the related question of sustainability of growth in the sector. There are some indications that for example existing tanzanite reserves are ap- proaching a stage where the remaining resources are less accessible as they re- quire increasingly difficult Mining operations (see text box below). Tanzania's production of tsavorite nearly ceased in late 2003 (Yager 2003:31.4). Box 1 Tanzanite production In Merelani (Arusha Region), the world's only source of tanzanite, officially reported production of tanzanite fell to 4,490 kg in 2003 from 6,461 kg in 2002. The value of tanzanite produced in 2003 amounted to US$13.14 million. From 2001 to 2003, tan- zanite accounted for 80 percent of the value of domestic gemstone production. Ac- cording to the Ministry of Energy and Minerals, the decrease in output in 2003 may have been due to falling tanzanite prices in 2002 or increasingly difficult Mining con- ditions. Mineshafts are getting deeper and more difficult to reach. (Source: Yager, 2003) Gold reserves are abundant along the western side of the Rift Valley in the re- gions of Mwanza, Mara, Shinyanga and Tabora. Van Campenhout (2002:5) suggests that as much as roughly 130m tons of gold reserves are still available in Tanzania. The same author argues that the growth observed so far reflects development of mineral projects in the `known' geology, and he considers it questionable whether investments will be made in more remote locations. The NSGRP is rather optimistic that improved growth in the Mining sector can be realised over the next five years, provided that productivity improvements in Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 5 the sector will be undertaken (p.33). Similarly, Mwalyosi (2004) believes that the sector will continue to see high growth rates and with time will grow to ac- count for 10 percent of GDP (compared to the current level of 1.9%). It is how- ever not clear what this estimate is based on. According to the estimates of the Economist Intelligence Unit, growth is ex- pected to slow down in the next years. They base their estimate on the expecta- tion that new investments are not expected to increase significantly and so growth in the Mining sector will primarily be driven by operations in existing mines. 2.4. Employment and Income This paragraph aims at investigating the contribution of the Mining sector to poverty reduction via the creation of employment. The question is whether and to what extent the growth in the Mining sector has translated into an increased level of employment in Tanzania. Secondly, the aim is to assess what type of employment has been created. The analysis includes both, employment effects within the Mining sector itself, and in the local communities surrounding the Mining sites. Overview Most of the mineral resources of Tanzania, in particular gold and gemstones, are exported in unprocessed form, hence associated in-country employment ef- fects are low. As the US Geological mineral Survey quotes "more than 98% of the gemstones exported from Tanzania were shipped in rough form to foreign cutting and polishing centres" (31.4). Overall, employment in the Tanzanian Mining sector accounts for only 0.2 per- cent of total employment, corresponding to roughly 15,500 men and 13,800 female (National Bureau of Statistics, 2000/01). Thus the sector's impact on the absolute level of employment is limited. As Chart 1 below demonstrates, the structure of employment in the Mining (and quarry) sector differs significantly from the general picture. Hence, four- fifths of those employed in the sector are self-employed without employees, a proportion far higher than the national average. The national composition of employment is however highly skewed towards persons `owning own farms or shamba', but even when compared to other non-agricultural sectors such as construction, it becomes evident that the composition of the Mining sector is unique (see Chart 1 below). The high proportion of self-employed in the Mining sectors is arguably indica- tive of the high proportion of artisanal miners working in the sector. The re- maining fifth of the employed persons in the Mining sector are paid employees, typically working for large-scale Mining companies and the more advanced small-scale companies. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 6 Chart 1 Employed persons by employment status 100.0 80.0 60.0 rcentageeP 40.0 20.0 0.0 Paid employee Self-employed Self-employed Unpaid family On own farm or with employee without helper (non- shamba employee agric) Mining Construction All sectors (Source: National Bureau of Statistics) Employment in the Large-Scale Mining Sector The increase in gold production originates primarily from six foreign owned large-scale Mining companies, which are listed in Table A in Annex 2. Small- scale operators, which have traditionally dominated the Mining sector, account by contrast for only a small part of total output. Although the large-scale sector is the main growth engine, the majority of scholars attribute limited local employment creation potential to it. One reason provided is the capital-intensive nature of production of the large-scale mines (Kulindwa et al., 2003: 111; Van Campenhout, 2002: 10). Capital investment corresponding to US$1bn has been exported by foreign Mining firms for use in Tanzanian large-scale mines (Mwalyosi, 2004). Another possible reason is that the recruitment of labour for large-scale mines takes place outside the locality (see for example Mwalyosi, 2004). In addition, Mwalyosi emphasizes that Mining employment fluctuates with production lev- els and hence is not a source of stable, long-term employment. The statistical analysis suggests a positive relationship between large-scale Mining and employment (see Chart 2 below). The correlation coefficient be- tween output from gold mines and employment in large-scale mines is 0.79. However, whereas this positive relationship applies to the period 1997-2003, it breaks down in the last few years, notably from 2002 to 2003 where gold pro- duction has continued to rise, while employment has dropped quite signifi- cantly. It still needs to be established whether this recent fall in employment is a single outlier or indicative of a new trend. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 7 Chart 2 Employment in large-scale mining sector and gold production 4500 50000 4000 45000 3500 40000 3000 35000 2500 30000 25000 gold 2000 20000 Kg Employment 1500 15000 1000 10000 500 5000 0 0 1997 1998 1999 2000 2001 2002 2003 Employment of Tanzanians by large-scale mines Gold production (Note: Pearson correlation coefficient 0.79. Source: Data from Ministry of Energy and Minerals) The question of the type of employment created is difficult to answer. Whereas the number of people employed in the large-scale Mining sector has clearly in- creased since 1997, the level of remuneration and type of employment has not been specified in the available statistics. Mwalyosi notes that over 50% of Min- ing employees are originally farmers or civil servants, which, he suggests, indi- cates that those employed in the large-scale Mining sector are not recruited lo- cally. This may be true for the civil servants who typically are strongly repre- sented in urban areas away from the Mining activities, but it cannot be ex- cluded that farmers are recruited locally. As Chart 3 below demonstrates, the pay-scale for paid employees in the Mining sector (here used as a proxy for employees in the large-scale Mining sector) differs significantly from other sectors of the economy: Whereas Mining and quarrying is one the best-paid sectors for paid employees below the age of 35, the picture is completely the opposite for employees above 34, who earn sig- nificantly less than their younger colleagues and their contemporaries in other sectors. This inverse relationship between pay and age is only valid in the Mining sec- tor. It may be explained by the facts that a) the Mining industry primarily de- mands strong and fit workers, and b) the Mining industry is a young industry so the majority of the (presumably) young workers recruited at the outset are still below 35 years of age. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 8 Chart 3 Average monthly income of paid employees by age and selected industries 200000 180000 160000 140000 120000 Tsh100000 80000 60000 40000 20000 0 and Total Quarry Transport Agriculture Mining Construction Manufacture Age 10-17 Age 18-34 Age 35-64 (Source: National Bureau of Statistics, 2001) Safety and security at the mines is an issue of on-going debate in Tanzania. In early 2003, the Government instituted new measured to regulate increased se- curity of tanzanite Mining, which is undertaken to a large extent by small-scale miners under increasingly difficult Mining conditions. The increased emphasis on mine safety followed a fatal accident in June 2002 that temporarily halted tanzanite production (Yager 2003: 31.4). In summary, the employment or income related contribution to poverty reduc- tion of the Mining sector is rated as limited by the authors of this paper. Al- though the large-scale Mining sector created a significant number of employ- ment opportunities in the first years of operation, there is now some indication that the rate of employment has started to decrease, as the sector grows increas- ingly capital-intensive. In addition, employment is usually not long-term as the physical demands of the jobs impose age-limits. Employment in the small-scale Mining sector The state monopoly of the Mining industry of the 70s ended in the late 1980s opening the way for any citizen to register claims and sell minerals. In conse- quence, the number of small-scale miners grew rapidly. The growth was further boosted in the early 1990s with the Government permitting exporters to use their proceeds for financing imported goods, equipment and spare parts. Con- sequently, the number of people employed in small-scale Mining operations surpassed 500,000 by 1995 (Philips et al., 2001). Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 9 Many observers argue that the subsequent introduction and consolidation of large-scale Mining operations in Tanzania has had negative implications for employment in the small-scale sector, both in terms of number of jobs and the level of pay. The two following hypotheses have been formulated: The growth in large-scale Mining has crowded out small-scale miners, par- ticularly artisanal miners; Following increased competition from capital-intensive large-scale opera- tors, the return (pay) to small-scale miners has decreased significantly. With regard to the first hypothesis, Kulindwa et al. (2003: 86) argue that there is a trade off between the livelihoods requirements of artisanal miners on the one hand and the economic objectives of large-scale operators on the other. Likewise, Phillips et al. (2001) argue that the level of employment in the ar- tisanal Mining sector has dropped since the peak in the 1990s as a result of the influx of foreign large-scale Mining operations. To substantiate this hypothesis Kulindwa et al. describe how large-scale operators, operating within the limits of the Government's Mining policy, mobilise support from local government authorities with a view to buy out small-scale operators. This is confirmed by similar anecdotal evidence from Bulyanhulu, where the Canadian operator Bar- rick allegedly forced small-scale artisanal miners to leave after buying the rights to the area (BBC, 2001). However, there is also evidence to the contrary, i.e. that large-scale miners may contribute to keeping small-scale operators in business. One example is the large-scale operator Meremeta, which has entered into agreements with a num- ber of small-scale operators providing them with cheap equipment (Kulindwa et al. 2003: 88) in return for agreements that the small-scale miners sell their production exclusively to Meremata. Very little reliable data exists to document employment levels in the small- scale sector: A large proportion of small-scale miners, notably artisanal miners, are employed in the informal sector and therefore not necessarily accounted for in official statistics. The World Bank (2001: 1) cites estimates of around 500,000 people employed in the artisanal Mining sector as of 2001 and the em- ployment level is therefore roughly the same as in the mid 1990s. Likewise, Mwaipopo et al. (2004: 113) estimate the number of artisanal miners in Tanza- nia to be around 600,000 in 2004. By contrast official Government statistics from NBS (2001) estimate the num- ber of people employed in the informal sector with Mining and quarrying as a dominant activity (an alternative proxy for the number of people employed in the small-scale sector) to approximately 20,000 people. Still, this proxy grossly underestimates the number of people employed in the small-scale sector since it a) explicitly excludes those working in the formal small-scale sector, and b) arguably only captures a small fraction of those employed in the informal sec- tor, since informal operations by definition are difficult to comprehensively Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 10 trace and estimate. Hence, the evidence is inconclusive and it cannot be ruled out that the large- scale Mining sector has in fact both positive and negative effects on the level of employment in the small-scale sector by a) crowding out some small-scale op- erators while b) keeping others in business through contribution agreements or similar instruments and finally c) leaving some unaffected. With regard to the type of employment created, some authors argue that the small-scale sector has provided very significant income opportunities to the rural population in Tanzania. Phillips et al. (2001) for example point to evi- dence that small-scale miners at one point had an income six times higher than agricultural workers. In consequence, they see the growth in the small-scale Mining sector as a very significant contribution to poverty reduction in the con- cerned areas. Recent data from NBS (2000/01) indicate however that the mean monthly in- come of self-employed in the Mining industry (here used as a proxy for small- scale workers) is now among the lowest in the country, even lower than the in- come for the self-employed in the agricultural sector. This is clearly demon- strated in Chart 4 below, where mean income for self-employed (with no em- ployees) are compared across industries. Chart 4 Mean monthly income of self-employed (without employees) by selected industries 2000/01 100000 91143 80000 60000 49933 53908 Tsh 38053 40000 36005 17079 21291 20000 0 and and Trade Total Quarry Fishing Transport Mining Manufacture Agriculture, Forestry Construction (Source: National Bureau of Statistics, 2001) The data does not take into account that Mining typically is a seasonal activity and the mean income for months actually worked may therefore be somewhat higher. Also, the data covers a wide income disparity between men working as miners (higher paid), and women, who typically work in (lower paid) auxiliary. Self-employed women in the Mining industry earn only about a third of their Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 11 male counterparts (see paragraph 2.5 below). In consequence, there is some evidence that the returns to small-scale miners have dropped concurrently to the introduction and consolidation of large-scale operators. Although this does not prove causality it may be indicative thereof. In summary, the net effect of the introduction of large scale Mining on the number of sustainable jobs in the small-scale sector is believed to be negative: The level of people employed as small-scale miners has at best remained con- stant, but the sector now features the lowest returns for self-employed in Tan- zania. Employment effects on local communities Apart from the direct employment benefits, large-scale Mining firm can create employment more indirectly through infrastructure investments, in particular in the water and roads sector. There may be infrastructure investments of large Mining firms (roads, water, power) in the surrounding area, which in turn bene- fit the wider community. For example, EIU (2004: 31) observes that rural communities have benefited `to a degree' from the development of gold mines as companies have improved rural roads and linked electricity facilities to the national grid. However, no comprehensive analyses have been made on the sector's overall impact on the level of infrastructure in the Mining communities and the em- ployment effects triggered through this investment. In fact, most of the capital equipment is imported from abroad and the minerals, once extracted, are generally exported without any significant in-country proc- essing. Likewise, most of the processes linked to exploration of potential re- serves rely on foreign laboratories. Consequently, wider employment effects are limited to mainly services industries, such as transport, food and, on a less positive note, prostitution. These observations notwithstanding, the Mining sector's contribution to the Tanzanian economy has increased significantly over the last years. Chart 5 be- low traces the level of statutory taxes and contributions paid by commercial miners to the Government in the period 1997-2002. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 12 Chart 5 Statutory taxes and other contributions paid to the Government - Commercial mines in Tanzania 40000 35000 34240 30000 25000 24416 '000 20000 18948 US$15000 10000 5000 4882 2151 2182 0 1997 1998 1999 2000 2001 2002 (Source: Chamber of Miners, www.chamberofminers.org) Hence, there is a clear positive association between the rise in gold production and the increase in revenues received by the Government from the commercial miners. It remains to be established whether part of these funds has been rein- vested in communities adjacent to mining operations to offset any disadvan- tages. Still, it is clear that there is a significant and increasing pool of funds, which the Government may legitimately consider using to the benefit of the Mining communities. Moreover, some politicians, including President Mkapa, have argued that tax incentives have been too generous towards foreign investors and so, they ar- gues, a tax raise could be justified. 2.5. Social Impacts The social impacts of large Mining operations on the surrounding communities can be both positive and negative. Some authors (see for example Kulindwa et al., 2003: 81) describe Mining op- erations as a `successful vehicle for social integration', arguing that Mining op- erations attract labour from all over the country. Mining communities are there- fore typically much more diverse than a typical Tanzanian village. Moreover, some Mining firms have launched specific social investment pro- grammes (typically in health or education) to increase the `goodwill' of the neighbouring communities. Barrick Gold Corporation, which runs the Bulyan- hulu mine through its subsidiary Kahama Mining Corporation, has for example established a fund to support various `charitable endeavours', which claims to be responsive to local needs and priorities (Barrick Gold Corporation, 2005). Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 13 Some of the negative social impacts on local communities associated with the operations of large Mining operations are related to gender and health, in par- ticular HIV/AIDS, which will be explored further below. In addition, child labour has been voiced as a particular concern in the Mining sector (see for example GoT, 2005: 10). Kulindwa et al. (2003: 95) argue that this essentially reflects the high level of poverty in the Mining areas, which in turn forces parents to send their children to work the mines. According to data collected by George (2003) in small- and large-scale Mining operations in the Geita District, 12.5 percent of the workforce were children. Moreover, according to the data from George, child labour is primarily a con- cern in small-scale Mining operations, and, by contrast, very infrequent in large-scale mines (George, 2003: 76). HIV/AIDS As noted in Drasch and Boese-O'Reilly (2004) small-scale miners constitute a high-risk group for spreading HIV/AIDS since they `are mobile men with money'. Kulindwa et al. (2003: 94) further add that the risk of spread of HIV/AIDS in Mining communities is high due to lack of awareness, a carefree attitude, widespread prostitution, and lack of access to quality health services. The question is then whether the risk of getting HIV/AIDS is significantly higher in Mining communities than elsewhere in Tanzania. Arguably, HIV/AIDS infection rates are equally high or higher in many other sectors. Road construction projects for example are also associated with high infection rates. Unfortunately, very little reliable data is available on the number of HIV/AIDS affected living in Mining communities and the hypothesis is therefore difficult to test. According, to George's analysis of local communities in Geita District, HIV/AIDS is at least perceived to be much more common in Mining communi- ties than elsewhere: Hence, 37.8% of those living in the sampled Mining com- munities mentioned HIV/AIDS as a common disease, compared to 18.9% in the sampled non-mining communities (George, 2003: 75). In the context of a wider research project on artisanal Mining in Tanzania, Mwaipopo et al. (2004: 94) have conducted a study to identify and rank threats to households in Mining communities. HIV/AIDS does not come out as a sig- nificant threat, at least not when compared to other threats such as food short- age, legal threats, Mining prices etc. Still, this does not imply that HIV/AIDS should be neglected in Mining communities; only that Mining households have more pressing and immediate concerns to deal with in their daily lives. Gender As noted by UNIDO (2001) a high number of women are involved in small- scale Mining activities worldwide. This is also the case in Tanzania, especially Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 14 so in the informal Mining sector. According to figures from the National Bu- reau of Statistics, the majority (60%) of self-employed workers without em- ployees in the Mining sector (here used as proxy for employment in small-scale Mining sector) are in fact women. Additional NBS data on persons employed in the informal sector confirms this picture: Approximately 13,600 women are employed in the informal sector with Mining and quarrying as a main activity compared to only 8,380 men (NBS, 2001: 62). Still, only a small proportion of these women are actually employed in Mining activities per se. According to information revealed during interviews with of- ficials from the Ministry of Energy and Minerals, women mostly work in re- lated service sectors. Chart 6 Income of self-employed women by selected industry, absolute and as ratio to self-employed men, 2000/01 70000 1 60000 50000 self-employed 40000 of Tsh30000 men 20000 income 10000 mean 0 0 to and and Trade Total Ratio Quarry Fishing Transport Mining Manufacture Agriculture, Forestry Construction Female income Ratio to male income (Source: National Bureau of Statistics, 2001) As Chart 6 above demonstrates, the salary of self-employed women in the Min- ing sector is very low compared to men. It is in fact only a third of that of their male colleagues, a lower proportion than in any other industry listed in the chart. This supports the assumption that women work mainly in lower-paid jobs, providing services to the small-scale Mining industry. The question is then why so few women work as miners per se. Mutagwaba et al. (1997) sug- gest for example that women face difficulties entering the Mining sector due to cultural norms and the nature of the equipment applied in small-scale Mining, which requires hard physical labour. According to the World Bank (2001: 2) women miners are also facing obstacles due to lack of formal education, lack of access to collaterals for securing loans, and time constraints. 2.6. Environmental Impacts The NSGRP (GoT, 2005: 7) emphasises that serious concerns have been raised regarding the Mining sector's impact on the environment. Negative environ- Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 15 mental impacts associated with Mining activities has been documented in sev- eral studies (see for example Kulindwa et al., 2003: 76-77; Mwaipopo et al, 2004: 50; Van Straaten, 2000; Appleton et al., 2004; Drasch and Boese- O'Reilly, 2004). The concerns generally relate to both large-scale as well as small-scale Mining. A 2001 Government Commission on `The Legal Framework for the Develop- ment of the Mining Industry'stated for example that `while it is true that small- scale Mining endangers the environment, it is also true that large scale Mining is even more damaging' (Law Reform Commission, 2001: 20). This paper does not further discuss environmental impacts of small-scale Min- ing. This is due to the focus on the environmental trade-offs related to the growth in the sector, which is based on the gold production through large-scale operations. There are a number of studies summarizing the environmental im- pacts of artisanal Mining, which are mainly, mercury pollution through inhala- tion of vapour pollution of surrounding water bodies as well as abandonment of open pit mines and associated effects on the surrounding landscape (Kulindwa et al., 2003: 76-77; Mwaipopo et al, 2004: 50; Van Straaten (2000); Appleton et al. (2004); Drasch and Boese-O'Reilly (2004)). There are a number of environmental issues associated with large-scale Mining activities. Many of the environmental concerns are linked to the breaking and exploding of rocks, which has been reported as a major nuisance to the local environment. The particular environmental issues concern inter alia land ero- sion and degradation, air pollution, water pollution, and noise pollution. George (2003) has traced and documented all of these effects in his Geita Dis- trict survey. Dust pollution in the area around the Geita Gold Mine has for ex- ample caused pollution of drinking water sources of a nearby village. It has therefore been necessary for the Mining operator to supply tap water to the lo- cal community (George, 2003: 71). Moreover, both George (2003: 72) and Ku- lindwa et al. (2003:77) have pointed to deforestation as a major issue. Alleg- edly, large-scale Mining operations have in many cases made significant land clearance necessary. George reports from Geita District that the Geita Gold Mine large-scale operation has acquired 110 square kilometres in the Geita Forest Reserve, of which a significant proportion has been cleared for plant, housing and infrastructure (2003: 73). Still, the conclusions are mainly based on case-studies and anecdotal evidence and should therefore be verified by more comprehensive studies. The scientific evidence notwithstanding, there is no doubt that Mining activities are at least perceived as having serious negative impacts on the environment. The Government has integrated several environmental safeguards in its poli- cies, notably the 1998 Mining Act, which calls for Environmental Management Plans (EMP) and Impact Assessments (EIA) as a condition for Special Mining Licenses.The EMP has to include inter alia proposals for prevention of pollu- Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 16 tion, waste treatment, protection and reclamation of land and water resources and for eliminating or reducing adverse environmental effects. According to preliminary information from Ministry of Energy and Mineral officials, 5 EIAs and EMPs (acceptable to the Ministry) have to date been pre- pared by large-scale operators and approximately 10 by medium scale opera- tors. As the 2001 Law Reform Commission remarks, implementation of these safe- guards need to be more consistent and should involve NEMC to have an objec- tive third party assisting the line Ministry. Large-scale companies are also reacting to the general perception that Mining activities are environmentally dangerous. Geita Gold Mine, for example, was recently awarded ISO14001 standard for environmental management, a first- ever for a Tanzania-based Mining operation. Still, it is worth noting that award of ISO14001 merely implies compliance with the host country's environmental legislation and standards, which in the case of Tanzania has suffered form limited coverage and enforcement. In conclusion, award of ISO14001 does not necessarily imply environmental improvement in a Tanzanian context. 2.7. Conclusion It is beyond the scope of this analysis to fully quantify the direct and indirect benefits and costs of the introduction of large-scale Mining operations in Tan- zania. Based on the evidence surveyed in this report, the Mining sector seems to have had limited influence on poverty reduction in the local economy. Employment in the large-scale Mining sector is limited although especially younger employ- ees may receive significant salaries. The majority of those employed in the Mining sector operate, mostly as self-employed, in the small-scale sector, typi- cally as artisanal miners. This category receives very low returns, especially when one considers the hardship associated with this kind of employment. It seems furthermore that an increasing income disparity is emerging between those employed in the small-scale and large-scale Mining sector. To the extent that those recruited by the large-scale Mining-sector are recruited outside the local community, the local community is thus restricted to opt for poorly paid employment opportunities in the small-scale sector. Large-scale Mining may have positive effects for local communities through the improvement of basic infrastructure. There is however no indication that the expansion in the Mining sector triggers significant growth in the local econ- omy, since Mining operations generally are detached from local supply chains and therefore primarily create employment in the services sector. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 17 The NSGRP emphasizes that policies should be designed so that benefits from high-growth sectors are transmitted to the poor in form of better livelihoods opportunities, e.g. supporting supply linkages with local producers. The ques- tion is however how to create such supply linkages. Tanzania would need to improve its human capacity and capital stock so that value-added processing can increasingly take place within the country. Mining poses a number of threats and possibilities to the local communities as well as to the miners themselves. The nature and extent of these threats and op- portunities cannot be assessed in detail due to lack of reliable data. There are, however indications of a number of negative social effects, notably child la- bour, HIV/AIDS, and gender imbalances. Most of these problems have been observed through case studies, which does not allow making conclusions about the Mining sector in general. Finally, the evidence provided about the environmental effects of large-scale Mining suggest that Mining communities may suffer a number of severe ef- fects, spanning from direct and observable noise and erosion, to longer term pollution of air, water and soil, which in turn may have serious health conse- quences. Still, the evidence does not allow for extrapolation and more rigid, comprehensive analysis is required to have a better idea of the environmental implications of large-scale Mining in Tanzania. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 18 3. Freshwater Fisheries 3.1. Questions Arising in the Context of the CEM Fisheries from freshwater resources have, like Mining and Tourism, developed considerably over the past decade to become a main driver of economic growth and a significant source of export revenue for Tanzania. This sets the freshwa- ter Fisheries sector apart from the marine Fisheries sector, which, as discussed in Part 2 of this report, still has to develop into a major driver of growth in Tan- zania. Fisheries are based on a renewable resource, which makes the question of sus- tainability pertinent, especially for Lake fisheries. As this Chapter will show, there are already indications that fisk stocks in Lake Victoria are decreasing. Although total catch is presently stagnant, the number of fishing boats has in- creased and the `Catch per unit Efforts (CPUE)' has been decreasing in recent years. Moreover, while Fisheries are a success story in pure economic terms, this is not necessarily so in terms of poverty reduction. The impact on income poverty through increased local employment in the Fisheries industry around the lakes may well have been set-off through a negative impact on livelihoods of ar- tisanal fisherfolk and on nutrition of communities whose livelihoods are based on the lake. Tanzania features water bodies corresponding to approximately 945,000 square kilometres. Freshwater resources, notably Lake Victoria, Lake Tanganyika and Lake Nyasa, make up only five percent of this area, but yield about 85 percent of all fish production. For that reason, this chapter will focus primarily on freshwater fishery. 3.2. Trends in Catches and Revenue The freshwater bodies of Tanzania harbour a variety of fish, including Nile perch, Sardine, Tilapia and Catfish. The bulk of catch comes from lakes Victo- ria and Tanganyika, together yielding 85 percent of inland production. Other potentially important water bodies are lake Rukwa, Lake Nyasa, Mtera dam and Nyumba ya Mungu. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 19 The annual catch of Tanzania freshwater fishery over the last ten years has in- creased from 294,782 tons in 1993 to more than 300,000 tons in 2003 with most of the increase coming from the Lake Victoria Nile Perch yield. The cor- responding revenues have increased from Tsh.31 million in 1993 to Tsh.141 million in 2003 as indicated in Table 1 below. Table 1.Freshwater catches and revenue collected for the year 1993 ­ 2003 Year Freshwater catch (tons) Freshwater catch value (Tsh) 1993 294,782.1 31,238,839.00 1994 228,006.6 30,949,458.00 1995 207,139.0 45,805,145.00 1996 308,600.0 38,200,000.00 1997 306,750.0 42,265,000.00 1998 300,000.0 47,486,100.00 1999 260,000.0 44,018,000.00 2000 271,000.0 45,500,000.00 2001 283,354.0 47,108,668.00 2002 273,856.0 54,771,300.00 2003 301,855.0 141,073,500.00 (Source: URT, 2004c) Chart 7 shows recent trends in revenue for freshwater and marine Fisheries as a proportion of total Fisheries revenue. Currently revenue from freshwater Fish- eries accounts to approximately 80 percent of total revenue from the sector. The proportion was roughly the same at the beginning of the 1990s, but was some- what lower, around 60 percent during the mid-late 1990s. All of the above statistics exclude fish for own consumption (subsistence), which may account for about 10 percent of total catch (Kulindwa 2001). Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 20 Chart 7 Freshwater and marine revenues as share of total fisheries revenues 90.00 80.00 70.00 60.00 50.00 40.00 30.00 20.00 10.00 - 1992 1994 1996 1998 2000 2002 2004 Proportionof freshwater catchvalueontotal revenue Proportionof marine catchvalue ontotal revenue (Source GoT 2004) 3.3. Contribution to Export Earnings The proportion of fishery exports to total exports has been increasing from 4.8 percent in 1996 to about 7.1 percent in 2003. Of this proportion, freshwater fish exports to total exports has been between 4 percent and 5.6 percent between 1996 and 2003 as indicated in Chart 8 below. Of the total revenue generated by fish exports between 1996 and 2003, an aver- age of approximately 80 percent has been accounted for by NP exports. The NP export trade figures may furthermore be underestimated. According to Ku- lindwa (2001) exporters have quoted similar FOB prices for both frozen and chilled fish fillet, while the actual price for the chilled was higher. Hence, the Government may loose additional revenue due to underreporting of export prices. Chart 8 Proportion of freshwater fish export earnings on total exports, 1996 ­ 2003 6.00 5.00 % ni 4.00 oni 3.00 ort oprP 2.00 1.00 - 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Years Proportion of Freshwater Fish exports on Total exports (Source GoT 2004) Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 21 Still there are considerable risk factors associated with Tanzania's export of fish from Lake Victoria, which is mainly destined for the EU. Previously, fol- lowing unfavourable assessment of sanitary standards in Tanzanian fish proc- essing plants, the entire production was halted for several months due to an EU import ban This in turn indicates that a) exports from the Fisheries sector relies is highly dependent upon access to a single market (EU) and b) that the quality of sani- tary standards in processing plants is a crucial determinant for the sector's con- tinued growth. 3.4. Employment and Income Apart from fishing, freshwater Fisheries give rise to many different types of employment, notably dried fish collection, small scale or traditional processing, large scale or modern processing, drying, salting, fish meal processing and fish oil processing. The available statistics for artisan fishers indicate that the number of people engaged in fishing has increased from 64,578 in 1990 to about 95,000 in 2003. Also the number of boats used for fishing has increased from 18,696 in 1996 to 31, 849 in 2003 (See Chart 9 below and Table C in Annex 2 for details). Chart 9 Fishermen and boats in freshwater Fisheries, 120,000 100,000 80,000 60,000 40,000 20,000 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Freshwater fishers Number of boats (Source: Fisheries Division) Freshwater Fisheries also generate employment related to fishing gear manu- facturing such as fishing net manufacturing, fish processing plants, manufactur- ing of fishing crafts and gears and aquarium fish. About eight fish processing plants are in Mwanza and Mara Region of the Lake Zone. Two fishing net plants are operating in Tanzania and fishnet supply has been increasing over the year but the local production of fishnet is still far behind demand. Accordingly, the proportion of imported fishnets is approximately 95 percent as depicted below in Chart 10. Accordingly, an important backward Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 22 linkage to the industry and employment opportunity remains unexploited. The employment multiplier for the Nile Perch in Lake Victoria can be com- puted at 1.561 with backward linkages of 1.37 and forward linkages of 0.18 (Kulindwa, 2001). The real multiplier is arguably much higher than the calcu- lated one using official data, which excludes many traditional processors, boat builders, fish vendors and transporters among others. We may therefore choose to work with a multiplier of 2 to represent all freshwater inland Fisheries. Ac- cordingly, the total employment contribution of freshwater Fisheries in 2003 would be close to 190,000 people.2 Chart 10 Fishnet production and import 1998 ­ 2003 2500 2000 1500 1000 500 0 1998 1999 2000 2001 2002 2003 2004 fishnet production(ton) Imports of fishnet (tons) (Source: Ministry of Natural Resources & Tourism) However, the type of employment created needs to be assessed. Some observ- ers point out that the returns to the crew members/ workers have been decreas- ing in favour of foreign investors (Kulindwa, 2001, see also Jansen, 1997). No reliable data is available to test this claim. Moreover, several observers claim that the growth in freshwater fisheries has led to a major restructuring and net decrease in the total number of employment opportunities in the lake region (see for example Kulindwa, 2001, Jansen, 1997, Jansen, Abila & Owino, 1999 Bokea & Ikiara, 2000). The argument is that the introduction of large-scale fisheries and processing methods has eroded liveli- hoods for artisanal fishermen and fishmongers. One estimate by Jansen, Abila & Owino (1999) suggests that the introduction of large-scale fishing methods (trawlers, industrial processing plants etc.) has crowded out jobs in the tradi- tional sector by a factor of 1:6. Still, this estimate is based on data form the mid-late 1990s and based on data form the Kenyan side of Lake Victoria and its relevance for Tanzania therefore needs to be more closely assessed. 1Total employment due to fish business divided by those directly employed as fishers. 2Employment multipliers for two fishing towns of Euboea and Kavala in Greece were given at about 2.4 (backward 1.2 and forward 1.2) see MacAlister Elliot et al (1999) Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 23 Finally, it is worth noting that the occupation itself is of a risky nature with lake bandits robbing fishers of their catches and fishing gear (nets, outboard en- gines). Clashes between fishing camps have also been known to occur just as fishermen have died from crocodile attacks (in Lake Rukwa for example).3 Drowning is another frequent cause of death as the wood-vessels are unstable and the fishermen typically unable to swim. 3.5. Sustainability The majority of fish products originate from Lake Victoria, which accounts for about 60 percent of Tanzania's inland fish production (Medard, 2003). The volume of freshwater catches has except for a minor drop in the late 1990s remained constant during the past ten years. Catches per unit effort (CPUE) by contrast has started to decline since 1998 (see Table 2 below). The trend in de- clining CPUE is an indication that, freshwater fish catches are generally declin- ing despite high demand. Over-fishing practices through illegal methods of fishing such as beach seine and small size nets have arguably depleted fish resources. Moreover, absence of quotas for freshwater fisheries has further encouraged over-fishing. The poverty reduction strategies would consequently need to address the sus- tainability of freshwater Fisheries in order to meet the challenge of halving poverty by the year 2015, since many households, particularly around these lakes, depend on freshwater Fisheries for their nutrition and income. Chart 11 Freshwater Fisheries, CPUE and total catch, 1995-2003 18 350,000 16 300,000 14 12 250,000 10 200,000 CPUE 8 150,000 Tons 6 100,000 4 2 50,000 0 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 Catch Per Unit Effort Freshwater Fish Catch (Source: Fisheries Department) 3Fish camps are normally established and managed by large-scale fishers or agents who receive credit from fish processors to purchase fishing boats, nets and transport boats. In return the fish camps commit to selling their catch exclusively to the creditor plants. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 24 The Lake Victoria Fisheries Research Project (LVFRP) carried out stock as- sessments for Lake Victoria's three most important commercial fish species: the Nile perch (Lates niloticus) (NP), the Nile tilapia (Oreochromis niloticus) and a sardine-like fish known locally as the dagaa (Rastrineobola argentea). The LVFRP has found that the stocks of Nile Perch amount to 530,000­ 650,000 tons per square kilometre in 2002, while those of Nile tilapia and da- gaa each amount to approximately 1.2 million tons/km². From these biomass estimates, the scientists have calculated the indicative maximum sustainable yield (MSY), or the amount of fish that can be caught each year without deplet- ing the stocks. For the Nile perch, for example, the MSY is around 212,000 tons (Kudoja 2004). Comparing this amount to the actual amount of NP landed on the Tanzania side of the lake, it becomes evident that the MSY for the lake is almost absorbed by Tanzania alone, with 60% of all inland freshwater fish production being NP. If the region's fishery is to remain sustainable, the harvested amount should be below the MSY so that the fish are able to spawn. Table 2.Number of fishermen, catches and revenues, 1995-2003 Year Number Number of Boats Fish Catch and Reve- Value (Tsh.) nues of Artisanal Fishermen Fresh CPUE Fresh Waters Water Qty (Tons) 1995 62,486 18,696 11.08 207,139 45,805,145.00 1996 62,486 18,696 16.5 308,600 38,200,000.00 1997 62,486 18,696 16.4 306,750 42,265,000.00 1998 62,486 18,696 16.05 300,000 47,486,100.00 1999 62,486 18,696 13.9 260,000 44,018,000.00 2000 92,529 25,014 10.8 271,000 45,500,000.00 2001 101,195 25,014 11.32 283,354 47,108,668.00 2002 119,856 31,849 8.60 273,856 54,771,300.00 2003 119,856 31,849 9.48 301,855 141,073,500.00 (Source: Fisheries Department) The same project also docu mented a significant decline in the number of Nile Perch fish reaching maturity, the presence of too many immature fish in catches, and low fecundity levels due to over fishing and use of illegal fishing gear (Kudoja 2004). Corrective measures need to be instituted to ensure the sustainability of the Fisheries, including strengthening and enforcing the Fisheries regulations to deal with illegal fishing in Lake Victoria, instituting surveillance and monitor- Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 25 ing of major fishing grounds. Observing MSY and putting quotas for fishing via the market (i.e. the processing plants' exports amount per year by introduc- ing tradable permits to processors) is an option. Moreover, livelihoods for ar- tisanal fishers and local communities around freshwater Fisheries should be considered within the strategies for poverty reduction. An entirely different sustainability issue concerns the rising prices of fish as a consequence of large-scale fishing in the Lake Victoria zone. The high demand from foreign markets has made the price of NP rise and thus unaffordable to many households. Fresh fish prices have generally increased to surpass prices of meat (URT 2002). This in turn may have caused food insecurity and lack of nutrition in communities around major fish producing areas such as Mwanza, Mara, Shinyanga and Kagera (Gibbon 1997). A study conducted by Bokea and Ikiara (2000) argue for example that fish have become unaffordable to local communities living in the Lake regions, notably in the Lake Victoria area. Hence, they suggest that rising prices of Nile perch and other species eventually may have negative consequences for nutrition lev- els of lake communities. 3.6. Conclusion Freshwater fisheries have in the course of little more than a decade developed into a major source of foreign exchange for Tanzania. As mentioned export of fish products now account for more than 10 percent of total export earnings. Clearly, private investors, domestic and foreign, have seized the opportunity to set up and operate large-scale operations in the Lake Zone benefiting inter alia from absence of quotas and a major export market in the EU. Still, declining catches per unit have been observed in recent years which in turn suggest that especially lake Victoria, the main source of growth, has been subject to over-fishing. The challenge for Tanzania is consequently to create the foundation for increased sustainable levels of growth in the Fisheries sector. Freshwater Fisheries in the Lakes of Tanzania, presently follows private profit maximising yields, which do not consider factors internal to the resources, in particular rate of renewal, as well as the external economy, e.g. price, discount rate and the institutional framework. In the medium and long term, sustainable growth in marine Fisheries, can only be sustained if production methods and extraction rates follow maximum sus- tainable yield calculations. This calls inter alia for introduction of quotas and increased monitoring and surveillance by the authorities to prevent illegal fish- ing methods, which are becoming increasingly prevalent, as CPUE is decreas- ing. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 26 4. Tourism Unlike Mining and Fisheries, which are typical natural resources sectors, Tour- ism is essentially a service industry. Tourism is nevertheless discussed in this Report since the sector depends to a large degree on input from natural re- sources - along with human capital and physical infrastructure. The Tourism sector's dependency on natural resources is not made explicit in the CEM and the purpose of this chapter is accordingly to highlight and substantiate this link. Tanzania boasts some of the world's finest tourist attractions, notably the Ser- engeti National Park and the Ngorongogo Wildlife Conservation. In addition, as already described in Part 2 of this Paper, the country has a notable income from hunting tourism. In addition to these wildlife-based attractions, tourism in Tanzania is mainly composed of trekking (notably Mount Kilimanjaro) and coastal tourism. The latter has developed significantly since the mid 1980s, especially on Zanzibar. Marine parks on the mainland could also be developed into major attractions if the necessary infrastructure is put in place and private finance mobilised. Some key figures related to tourism in Tanzania are presented in Box 2 below: Box 2 Key Figures on Tourism in Mainland Tanzania and Zanzibar · International arrivals amounted to 576,000 in 2003; · Contribution of tourism to GDP is roughly 5 percent; · Export earnings from tourism amounts to US$ 581,6 million; · In 2001 Tanzanian national parks drew over 100,000 international visitors. This generated receipts of almost 5 percent of GDP, equivalent to about US$ 400 million; · The number of tourists entering Zanzibar increased by more than 50 percent between 1995 and 1999; · Tourism accounts for 80 percent of all service earning on the isles. (Source: EIU, MNRT, TANAPA) Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 27 Likewise, DFID (2003) argues that the real potential for shaping tourism as a major driver of local economic development comes from effective destination level management, where government, private sector and local interests are combined to ensure marketing, infrastructure and labour, facilities and services fit together. To properly assess the potentials of tourism as a source of growth and income it is necessary to look beyond the macro-level (contribution to GDP, foreign ex- change earnings, number of visitors, the number of formal sector jobs, foreign direct investment etc.) with a view to analyse 'destination' level economic de- velopment since this is the level at which poverty impacts should be seen and measured. This is above the micro detail of a few households and businesses but below the national level where it may make little impact. Local multiplier impacts include both formal and informal sector employment, indirect impacts such as improved infrastructure and public services, and more indirect benefits such as participation, empowerment and improved govern- ance. Although this study has tried to assess these effects, due to data con- straints, the results will be patchy and not conclusive. Chapter 4.1 of this Chapter describes recent trends in tourist arrivals and earn- ings and Chapter 4.2 focuses the sectors contribution to GDP and exports. Chapter 4.3 will discuss tourism attractions by subsector, and Chapter 4.4 will assess employment and income issues. Finally, Chapter 4.6 will discuss the sec- tor's potential for poverty reduction and Chapter 4.7 will set out a number of recommendations. 4.1. Tourist Arrivals and Earnings Tourism is a commercial industry and is increasingly becoming a major foreign exchange earner in Tanzania and other developing African countries (Kulindwa et al., 2001). Tourism has shown great potential for generating revenue and employment. The number of tourists visiting Tanzania has increased by more than 200 per- cent from 186,800 tourists in 1990 to 576, 000 tourists in 2003. The number peaked in 1999 with more than 625,000 tourists coming to Tanzania. See Chart 12 below, which illustrates recent trends. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 28 Chart 12 Number of tourists, 1990 - 2003 700,000 600,000 STS 500,000 400,000 URIOT 300,000 200,000 100,000 - 1988 1990 1992 1994 1996 1998 2000 2002 2004 YEARS Number of tourists (Source: Ministry of Natural Resources & Tourism) Concurrently, the revenue collected from Tourism has increased by US$ 666 million from US$ 93.7 million in 1991 to US$ 731 million in 2003 as shown in Chart 13 below. Hence, the Tourism sector has performed remarkably well in monetary terms over the past ten years. Chart 13 Revenue trend from tourism services for 1990 - 2003 800 )li 700 m 600 $SU( 500 400 nt 300 ou mA 200 100 0 1988 1990 1992 1994 1996 1998 2000 2002 2004 Years Revenue (US$ mil) (Source: Ministry of Natural Resources & Tourism) It is clear from the above Chart 13 as well as Chart 14 below that the strong growth performance of the 1990s has lost momentum in the beginning of the new millennium. Between 1999 and 2003, the high growth rates previously registered for num- ber of tourists and revenue started to decline. The number of visitors stopped growing altogether in 2000 and fell by contrast with approximately 20 percent. Revenue showed more resilience and so the significant drop in visitors was to some degrees countered by increasing revenues per tourist. See Chart 14 below for recent changes in the revenue per tourist ratio. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 29 Chart 14 Tourism arrivals, tourism revenue and revenue per tourist ratio growth rate for 1990 ­ 2003 100.0 80.0 60.0 40.0 20.0 0.0 -20.01988 1990 1992 1994 1996 1998 2000 2002 2004 -40.0 % Change of Tourist Arrivals % Change of Earnings from Tourism % Change of revenue/tourist ratio (Source: Ministry of Natural Resources & Tourism) The sudden decline in tourist arrivals to Tanzania coincides with the 1998 Au- gust explosions in the US embassies in Dar es Salaam and Nairobi. Moreover, the general sluggish growth in the first half of the current decade arguably re- flects a global trend triggered by the 2001 September 11 attacks in New York and Washington, D.C. Table 3.Tourists arrivals and revenue collected for the year 1990 ­ 2003 Year Number of tourists Growth rate Earnings (US$ mil) Growth rate percent percent 1990 153,000 - 65 - 1991 186,800 22.09 94.73 45.74 1992 201,744 8.00 120.04 26.72 1993 230,166 14.09 146.84 22.33 1994 261,595 13.65 192.10 30.82 1995 295,312 12.89 259.44 35.05 1996 326,188 10.46 322.37 24.26 1997 359,096 10.09 392.39 21.72 1998 482,331 34.32 570.00 45.26 1999 627,325 30.06 733.28 28.65 2000 501,669 -(20.03) 739.06 0.79 2001 525,000 4.65 725.00 -(1.90) 2002 575,000 9.52 730.00 0.69 2003 576,000 0.17 731.00 0.14 Average 11.54 21.56 (Source: Ministry of Natural Resources & Tourism, National Bureau of Statistics) 4.2. Contribution to Exports and GDP The contribution of the Tourism sector to total exports (measured by the pro- portion of Tourism earnings to total exports) increased consistently throughout Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 30 the 1990s to reach a level of approximately 64 percent in the 1999. Hereafter, the proportion has dropped significantly, not only because of the general depression in the Tourism sector mentioned in the previous section, but also because of the rise of mineral exports particularly gold, which has reduced the relative importance of Tourism as an export sector. Chart 15 below depicts the situation. In absolute terms Tourism remains a very important sector for the Tanzanian economy as noted by the CEM and the NSGRP. Chart 15 Proportion of tourism earnings to total exports for the year 1990 - 2003 70.0 60.0 50.0 40.0 30.0 20.0 10.0 - 1988 1990 1992 1994 1996 1998 2000 2002 2004 Proportion of tourism eanings to total exports (Source: Ministry of Natural Resources & Tourism) Although tourism is placed within the ministry of natural resources, its contri- bution to GDP is not recorded under the 'agriculture' GDP, where the remain- ing departments of natural resources are recorded. Tourism GDP is recorded under "trade, Hotels and restaurants'. By placing tourism into natural resources, rather than industry or trade, it is moved away from broader economic objec- tives The proportion of earnings from Tourism to GDP (see Chart 16 below) follows a similar trend and the reasons are the same: Tourism earnings has stagnated in a period where the Mining sector has grown at double-digit rates. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 31 Chart 16 Proportion of tourist earnings to the GDP for the year 1990 - 2003 12.0 10.0 8.0 6.0 4.0 2.0 - 1988 1990 1992 1994 1996 1998 2000 2002 2004 Proportion of tourism eanings to GDP (Source: Ministry of Natural Resources & Tourism) The question then remains if the Tourism sector in Tanzania is poised for re- newed growth. Tanzania, and East Africa more generally, have opportunities for renewed growth as many of the strong tourist markets in Asia have been facing serious health problems and natural disasters, notably the December 2004 tsunami. At the same time, Kulindwa et al. (2001) argue that there may be potential for the Government to broaden its revenue base especially with respect to air tick- ets, ground transport and accommodation. The Government looses significant revenue due to its inability to monitor and tax payments made abroad. This ar- gument is mainly based on assumptions and therefore difficult to verify (Ku- lindwa et al 2001). 4.3. Tourism Attractions Non-consumptive wildlife resources The Tanzania National Parks Authority (TANAPA) manages the country's 12 National Parks. As a parastatal organisation, TANAPA does not receive a re- current budget from the central government and is required to pay corporation tax to the Treasury. Pressure to generate more revenue and reduce visitors crowding the northern parks has driven TANAPA to search for ways to im- prove economic performance of the parks. While overall growth in the tourism sector increased between 1995 and 2000, growth in TANAPA visitor numbers and revenues slowed down in the second half of the 1990s compared to the first half. Figures on visiting tourists vary. Some are presented in Box 3below. The proportion of revenue from national parks to the overall Tourism sector revenue has declined throughout the 1990s from 12.7 percent in 1990 to about 0.6 percent in the year 2001 (see Chart 17 below). This may reflect an increase Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 32 Box 3 Figures on tourist visiting Tanzania's National Parks · In 2002/2003 more than 50 percent of departing tourists had visited national parks during their stay in Tanzania (TANAPA no date; survey conducted at Kilimanjaro and Dar es Salaam airports) · Between 1995-1997, 10-25 percent of international visitors to Tanzania went to national parks (Tanzanian Tourist Board) · In 1999, the parks received over 260,000 total visitor days (TANAPA) · A survey among 533 tourists in 2002/2003 showed that 86 percent of the park visitors were international tourists, 10 percent citizens and 4 percent non-citizen residents. Over 90 percent of the international tourists came from Europe and North America (TANAPA no date) in the contribution of other tourist activities such as hunting, mountain climb- ing, beach Tourism, conference Tourism etc. and is therefore not necessarily indicative of a decrease in absolute revenue from Parks and Reserves. Chart 17 Proportion of revenue from National Parks to the total tourism revenue for the year 1990 ­ 2001 14.0 12.0 10.0 8.0 6.0 4.0 2.0 - 1988 1990 1992 1994 1996 1998 2000 2002 Proportion of Parks revuene to Total tourism revenue (Source: Ministry of Natural Resources & Tourism) Revenue and expenditures related to non-consumptive wildlife resource used are discussed in more detail in the paragraphs below. TANAPA Income and Expenditure As Table 4 below shows, the approved TANAPA Budget for 1999/2000 pro- vided recurrent expenditure of Tsh. 10,131 million (US$ 15,6 million), and projected total revenues of Tsh. 12,081 million (US$ 18.6 million). Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 33 Table 4.Projected Revenue and expenditure by park for 1999/2000 Revenue Expenditure Revenue as % Park US$ US$ of expenditure Arusha 814 920 88.5 Gombe 57 243 23.4 Katavi 21 563 3.8 Kilimanjaro 8,449 1,813 466.1 Manyara 1,647 805 204.7 Mahale 44 368 12.1 Mikumi 172 886 19.4 Ruaha 275 981 28.0 Rubondo 30 387 7.7 Serengeti 5,119 3,026 169.2 Tarangire 1,718 1,086 158.2 Udzungwa 11 452 2.5 HQs 230 4,059 5.7 TOTALS 18,587 15,589 119.2 (Source: TANAPA 1999) Based on TANAPA records, revenue grew at an average annual rate of 17.6 percent over the 8-year period between 1990/91 and 1998/99, from US$ 4.4 million to US$ 15.9 million. Almost three quarters of this revenue is generated at Kilimanjaro and Serengeti, while Manyara and Tarangire each account for around one tenth of the total. These four parks all generate a financial surplus, effectively subsidizing the re- maining parks, all of which run at a loss. Seven parks cover no more than around one quarter of their recurrent cost. Chart 18 below shows the trend in number of foreign visitors to the four most important parks and all parks in to- tal. In absolute terms Kilimanjaro earned around US$ 6.6 million in 1998/99, Serengeti US$ 5 million, Tarangire and Manyara each US$ 1,5 million (TANAPA 1999). The majority of TANAPA revenue originates (57 percent) from non-citizen park entry fees, while almost one third is generated by camping fees (16 per- cent) and use of TANAPA mountain huts and banda accommodation (14 per- cent). All other sources of revenue together account for 13 percent of revenue. Noteworthy is the small contribution of concessions fees (2 percent).4 TANAPA charges a rental of 10 percent on the actual net room rate but has not issued specific instructions to concessionaries on submission of their monthly returns. Actual revenues to TANAPA vary between season and establishment. Most lodge-bed-nights generate arond US$ 2 for TANAPA, with permanent tented camps generating US$ 3 to 7 per bed-night. In contrast camping fees, 4In 1997/98 this was for example equivalent to US$ 361,000 Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 34 which are particularly high in Tanzania, generate either US$ 20 or US$ 40 per night, depending on the site. Chart 18 Number of foreign visitor to each park from 1991 to 1999 250,000 200,000 Kilimanjaro visitors 150,000 Manyara Serengeti foreign Tarangire of 100,000 All parks No. 50,000 0 91 92 93 94 95 96 97 98 99 (Source: TANAPA 1999) An economic study conducted in 2002-2003 revealed that TANAPA receives less income from high-end tours using hotels and lodges than from budget tours staying in campsites. In order for TANAPA to benefit from high-end tourism as opposed to mass tourism, they need to capture more high-end revenue by re- evaluating lodging concessions fees or encouraging and developing more spe- ciality camping. Otherwise these revenues are captured by tour companies and hotels. TANAPA's park entrance fees are comparable to those of Kenya with fee to the most popular parks averaging US$ 25. Tanzania charges the highest fees for vehicles, compared to Kenya, Botswana and South Africa. On the expenditure side, between 1990/91 and 1993/94 TANAPA spend 60-70 percent of its budget on recurrent cost, leaving a surplus for investment in vehi- cle equipment, replacement and capital development. During the late 1990s re- current expenditures rose to 85-90 percent while capital development was res- cued to 10-15 percent. A revenue reserve fund of Tsh. 25 million (US$ 38,000) has been maintained since 1991. A breakdown of park management expenses of 1999/2000 shows that only 6 percent is spent on ecological monitoring. TANAPA has a total staff of 1,452 (1999) with the majority (46 percent) being Park Rangers, notably in Serengeti and Kilimanjaro national parks. Comparison with other countries in the region show that the relationship between total staff and Area under protection is low. In Tanzania, the area per staff member is 28km2, while in South Africa it is 7.5 km2, in Kenya 11km2, in Uganda and Ghana 14km2 (TANAPA 1999). Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 35 Chart 19 TANAPA 2001/2002 Primary Revenue Concession Fees Resident Entry Professional Fees Rescue Fee 2% Filming Fees 2% 2% 1% Vehicle Entry Fees 3% Huts and Bandas Fees 13% Visitor Entry Fee 56% Camping Fees 21% (Source: TANAPA, no date) From the analysis above, it appears as if TANAPA could exploit its potential for revenue generation more fully. Compared to other parks in competing des- tinations, TANAPA parks have more activities run by external tour operators, leaving substantial revenue in the hand of foreign investors, while only charg- ing 2 percent for concessions. Other parks for example in Botswana, offer addi- tional activities not available in Tanzania, such as lake boating, fishing, over- night trips, bush breakfast etc. Coastal Tourism As mentioned, coastal tourism has developed significantly in Tanzania during the past twenty years, especially on Zanzibar which received close to 90,000 tourists in 2002, mainly form Europe (Italy) and Asia. Moreover, the number of visitors to Zanzibar more than doubled during the second half of the 1990s. The tourism sector has consequently developed into one of the main sectors of the economy on Zanzibar. According to figures from the Economist Intelli- gence Unit, 80 percent of service earnings originate for example from tourism. Consequently tourism is perceived as a success sector on the islands of Zanzi- bar. The growth has been partly facilitated by infrastructure development and private sector finance. Moreover, the growth has proven relatively resilient to the political distur- bances which have troubled the island in the past five years. If the political situation should deteriorate further in view of the upcoming elections (October Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 36 2005), it is likely that the number of tourist arrivals to the isles will stagnate or decrease in the short- to medium term. Coastal tourism has developed less rapidly on the mainland. As Table 5 below demonstrates Tanzania mainland boasts a significant number of marine areas, but receives far less visitors than Zanzibar does. Many of the marine areas are remote and difficult to access. Likewise, accommodation capacity is limited in some of these areas. Hence, for mainland Tanzania to more fully develop its coastal marine resources for tourism, it needs to focus on infrastructure devel- opment. Private capital should be mobilised to the extent possible to finance this. Table 5.Protected and managed marine areas, mainland Tanzania Name Year Sub-tidal established area in km2 Dat es Salaam Marine Reserves System 1975 26.0 Maziwe Island Marine Reserve 1981 2.6 Mafia Island Marine Park 1995 615.0 Tanga collaborative fishery management areas 1996-2000 26.4 Mnazi Bay-Ruvuma Estuary Marine Park 2000 200.0 Saadani National Park 2004 Kinondoni Integrated Coastal Area Management 2000 Program Rufiji Environmental Management Program 1998 (Source: World Bank, 2005) Cultural Tourism, archives and antiquities Cultural monuments, archives and antiquities currently attract around 60,000 tourists per year, with more than 90 percent of the visitors coming from abroad. Recent data suggest that the collection of fees is rising: Collected fees increased from Tshs.96.5 million in 2002 to approximately Tshs.132.6 million in 2003.The major sites visited are presented in Table 6 below. There is a potential for increased growth in this sub-sector as various efforts are underway to improve the conditions of most historical sites. In 2003, a master plan for conserving and developing antiquity centres was charted. Centres such as Dr. Livingstone and Mbozi meteorite were surveyed and demarcated, and a project to prepare Bagamoyo town for inclusion in the world heritage was initi- ated. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 37 Table 6.Visits to selected historical sites for the year 2000 ­ 2001 Centre/Year 2000 2001 Local Foreign Total Local Foreign Total Olduvai Gorge 1,520 51,250 52,770 1,732 52,400 54,132 Headquarter (HQ) 20 46 66 27 67 94 Isimila & Kalenga 220 460 680 440 600 1,040 Bagamoyo 210 930 1,140 420 1510 1,930 Kaole 520 2000 2,520 700 2550 3,250 Kondoa 320 105 425 360 115 475 Amboni 680 320 1,000 710 400 1,110 Tongoni 80 52 132 102 70 172 Kwihara-Tabora 700 260 960 850 280 1,130 Total 4,270 55,423 59,693 5,341 57,992 63,333 (Source: Ministry of Natural Resources & Tourism, Antiquities Division) 4.4. Employment and Income The Tourism sector has, apart from generating revenue to the government, pro- vided significant although fluctuating employment opportunities. The number of employees in the sector has been increasing. It has followed the same trend in number of visitors and earnings, i.e. significant increase through- out the 1990s to be followed by stagnation from 1998 onwards. Hence, in 1991, about 45,000 people were employed in this sector, and by the end of the decade the number had increased approximately fourfold. Hence, the sector is cur- rently providing direct employment for approximately 160,000 employees (see Chart 20 below and Table F in Annex 1). However, employment in the Tourism sector fluctuates considerably between the high and low season. Chart 20 Number of employees in the tourism sector 180,000 160,000 140,000 seeyolp 120,000 100,000 80,000 Em 60,000 40,000 20,000 - 1990 1992 1994 1996 1998 2000 2002 2004 Years employees (Source: Ministry of Natural Resources & Tourism) Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 38 Moreover the sector is believed to have significant direct and indirect employ- ment effects due to its linkage to other sectors. The stronger the backward and forward linkages, the more significant the employment (and income) effects. Using an employment multiplier effect for Tourism of 5.4 (based on Kweka et al., 2003), employment generated by the Tourism sector may have been high as 866,700 in 2003.5 Chart 21 Employment rate in the tourism sector 40.0 ) 30.0 (% tear20.0 10.0 - 1990 1992 1994 1996 1998 2000 2002 2004 years for 1991 - 2003 rate (Source: Ministry of Natural Resources & Tourism) In recent years, employment has been further boosted by an increasing number of people engaged in the production of goods and services such as the Maasai Boma, curio shops, wood carving, fishing, and farming of vegetables and fruits for the tourist market (Kulindwa et al., 2001). Not all of the activities may be formally accounted for and so the positive employment contribution of the sec- tor may be even higher. Finally, the Government itself also acts as an important driver of employment in the Tourism sector. Government Tourism projects worth Tshs 252,513 mil- lion were approved in the 2001 and 2003 period (see Table H in Annex 2 for details). These projects have created employment opportunities for more than 10,000 people and include the construction of tourist accommodation facilities such as hotels, guesthouses and lodges, and tour operators. 4.5. Environmental Impacts To maintain growth in the Tourism sector, it is necessary to conserve the natu- ral environment on which the industry depends while at the same time minimiz- ing the impact on other land users who may not benefit directly from Tourism. There are several examples in Tanzania of environmentally sustainable prac- tices in the Tourism sector. Examples include the Serena Lodge in Stone Town, 5(160,500 x 5.4). The Input ­ Output values used were for 1992 implying that the linkage effects are most probably much higher today and therefore higher employment and income multipliers as well. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 39 Zanzibar, which has introduced a pioneering sewage system (Kulindwa, 2001). Likewise, a resort on Mafia Island is running on solar power systems. Still, development of tourism also involves significant environmental trade-offs and conflicts: One obvious trade off associated with coastal tourism is the need for expanding infrastructure in order to facilitate access to protected areas: Construction of accommodation facilities and infrastructures and installation of modern tourist-related facilities may lead to degradation of landscapes or sites where the style and architecture is not in harmony with the environment. Accordingly, adequate planning and coordination is required in order to man- age environmental threats from Tourism development and creating a sustain- able Tourism industry. One option is to institutionalise the requirement for EIA at all levels according to categorisation of projects with respect to type and size of anticipated impacts. This, in turn should be coupled with continuous moni- toring of environmental effects. There is a need to ensure a balance between the type and scale of Tourism de- velopment in relation to the absorption capacity of different ecosystems. As- sessment of capacity and balancing the level of Tourism activity within it is a crucial means of preventing future environmental degradation. The location of tourist hotels at national parks could be made near the periphery of park boundaries. This would decrease adverse impact on park resources as it reduces frequency of incidental contacts between animals and people. Still, this may clash with the preferences of tourists who typically prefer to stay inside the parks. There is however scope for (re)-locating administrative facilities and staff hous- ing outside the parks. Such facilities are in many cases much more extensive than tourist accommodation and their relocation will therefore significantly re- duce the impact of human settlement inside the parks. The plan to move the Ngorongoro Conservation Area Authority (NCAA) offices and accommodation facilities outside the Conservation Area is an example of how impacts and in- terference to migratory pathways can be minimised. Likewise, TANAPA has relocated large parts of its operation in the Serengeti Park to Fort Ikoma located in the park buffer zone. Moreover, it is important to keep in mind that natural resource based tourism happens at the expense of other use of the resources. Hence, promotion of coastal tourism may conflict with the interests (and livelihoods) of coastal communities, living form fishing for example. Likewise, although there is potential for making land economically viable by mixing livestock and tourism, there is at the same time conflict over land use for local needs and protection for tourism Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 40 4.6. Tourism and poverty reduction In a tourism destination the benefits for the local population can be of three broad types: Financial, non-financial livelihoods impacts and governance re- lated impacts. Financial impacts arise where local people earn cash from waged jobs or sales of goods and services by entrepreneurs or the informal sector traders. Also, there could be shared community level income. Non-financial livelihoods impacts arise through improved infrastructure, com- munication, water and energy supply, health, education and security services. Governance impacts may arise in the context of improved local institutional development. Even if direct cash income is lacking, the wider development impacts can have long-term effects throughout the local economy. However, where tourism is planned and implemented in isolation of destination linkages, these comple- mentary impacts at the local level are not maximised. Unfortunately there are few empirical studies about the local level and poverty related impact of tourism in Tanzania. Tourism is not prominent in the Tanzanian 2001 Rural Development Strategy, which acknowledges the need for growth in rural areas through non-farm eco- nomic growth but without specifying the role and potential of tourism. Still, there are examples of how tourism can be of substantial (financial) benefit to local communities. in Loliondo Division in Ngorongoro District, seven villages earn a total of over US$ 110,000 per annum from joint ventures with wildlife tour operators. In Ololosokwan village, tourism revenue totals about US$ 55,000 per annum. (Kalonga et al 2003). The income from payments by one of four tour operators in Olosokwan Village is shown in Chart 22 below. If effects of elite capture are avoided and equitably distributed within the communities, this income has large poverty reduction potential in a dryland area, which does not offer much other types of diversification of the rural economy. The potential for improving the local development impact is still not fully ex- ploited (DFID 2003). As Homewood et al (2001) describe, in the Mara- Serengeti ecosystem, the number of households earning any income from tour- ism varies from 86% in Talek (Kenya) to 12% and 3% at the Ngorongoro Con- servation Area. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 41 Chart 22 Income from payments by one of four tour operators, Olosokwan Village and Ngorongoro District Council, 1999/2000 - 2002/03 12,000,000 10,000,000 8,000,000 Village (Tsh.) 6,000,000 District (Tsh.) 4,000,000 2,000,000 0 1999/2000 2000/01 2001/02 2002/03 (Source: Reconstructed from Kallonga et al. 2003, figures not exact) Consequently, there is amble scope for further development. In the southern circuit, e.g. Selous, Ruaha, and Udzungwa, tourism is for example growing, offering scope for positive impacts on local economic development. 4.7. Recommendations Tourism is perceived as a potential driver of macro-economic growth and em- ployment, and a contributor to foreign exchange earnings. Maximising pro-poor impact of tourism would require more than supporting small community run elements. A pro-poor tourism growth programme would require attention to company practices, destination management, infrastructure development, pro- curement patterns, national training and regulation. By placing tourism into the Ministry of natural resources, rather than trade or industry, it is made more remote from the centre of economic decision-making. In order to maximise local development impacts of the growth in the tourism sector, the following is recommended: Tourism development objectives to include the objective of stimulating lo- cal development; Concession or licensing procedures that include pro poor criteria in bids; Consultative tourism planning that increase access by the poor to tourism markets, infrastructure and services; Infrastructure (transports and accommodation) needs to be further upgraded to further develop tourism. To the extent possible this should be based on private finance. Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 42 Annexes Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 43 Annex 1. Background Charts Chart A. Growth in mining sector compared to GDP growth and growth in agriculture 18 16 14 12 change 10 8 real % 6 4 2 0 1999 2000 2001 2002 2003 Agriculture Mining and Quarrying GDP (Source: Bank of Tanzania) Chart B. Gold production, 1993-2002 50,000 40,000 30,000 Kg 20,000 10,000 0 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 (Source: Ministry of Minerals and Energy; EIU, 2004) Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 44 Chart C. Value of exports 1000 900 800 700 US$ 600 500 400 Million 300 200 100 0 1995 1996 1997 1998 1999 2000 2001 2002 Minerals Total export earnings (Note: Pearson correlation coefficient 0.69; Source: National Bureau of Statistics) Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 45 Annex 2. Background tables Table A. Main commercial mines in Tanzania Gold resources in Tanzania: Main Commercial Mines Mine Location/District Company Mineral Start Resources Annual Av- Mining (tonnes) erage Pro- duction (tonnes) Golden Lusu/Nzega Resolute Limited (Australia) Gold 1998 76.82 5.7 Pride Pro- ject Geita Gold Geita Anglo Ashanti (South Afri- Gold 2000 527.02 18.43 Mine can-Ghanaian Joint venture) BulyanhuluBulyanhulu/ Ka- Barrick Gold Corporation Gold 2001 411.07 11.34 hama Limited (Canada) North Nyabigena na Placer Dome (Canada) Gold 2003 116.23 8.51 Mara Gold Nyabi- Mine rarma/Tarime Buhemba Musoma Meremeta Limited Gold 2003 21.26 2.27 Gold Mine Tulawaka Biharamulo Pangea Mineral Limited Gold 2005 20.22 3.73 Gold Pro- ject Total 1172.62 49.98 (Source: Ministry of Energy and Minerals, 2005; Economist Intelligence Unit, 2004) Table B. Area of freshwater bodies in Tanzania. Freshwater Area occupied by Tan- Proportion in percent body zania (square kilome- tre) Lake Victoria 35,088 64.57 Lake Tanga- 13,489 24.82 nyika Lake Nyasa 5,760 10.60 Total 54,337 100.00 (Source: Ministry of Natural Resources & Tourism) Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 46 Table C. Number of people employed by the freshwater fisheries sector Year 1990 1991 1992 1993 1994 1995 1996 Freshwater fishers 64,578 60,361 48,470 46,916 46,639 50,029 50,029 Number of boats - 18,696 18,696 - - - - Year 1997 1998 1999 2000 2001 2002 2003 Freshwater fishers 50,029 50,029 56,088 75,042 75,042 95,547 95,547 Number of boats 18,696 18,696 18,696 25,014 25,014 31,849 31,849 (Source: Fisheries Department) Table D. Fishnet production and import, 1999-2003 Year Fishnet produc- Imports of fish- Total Fishnet Sup- Proportion of Im- tion (tons) net (tons) ply (tons) ported Fishnets to total (%) 1999 24 447.8 471.8 95 2000 42 720.6 762.6 94.5 2001 57 1,346.1 1403.1 96 2002 30 1,940.0 1970.0 98.5 2003 41 1,860.2 1901.2 97.8 (Source: Ministry of Natural Resources & Tourism and Ministry of Finance) Table E. Proportion of freshwater fish exports on total exports Nile Perch Proportion of Proportion of Proportion of Nile NP exports on Year Total ex- Total Fish (Freshwater) total Fish ex- Perch (Freshwater) Total Fish ex- ports (mil exports exports (mil ports on Total Fish exports on Total ports (%) US$) (mil US$) US$) exports (%) exports (%) 1996 1,298.70 61.78 52.27 84.6 4.76 4.0 1997 1,235.00 70.17 54.82 78.12 5.68 4.4 1998 1,109.80 83.52 65.73 78.70 7.53 5.9 (Sourc e: 1999 1,143.60 61.79 51.99 84.14 5.40 4.5 Eco- nomic 2000 1,290.60 64.54 45.90 71.12 5.00 3.5 Sur- vey, 2001 1,455.70 95.45 77.21 80.89 6.56 5.3 2003) 2002 1,568.30 94.24 76.31 80.97 6.01 4.8 2003 1,827.60 129.61 102.37 78.98 7.09 5.6 Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 47 Table F.Number of employees in the tourism sector Year 1991 1992 1993 1994 1995 1996 1997 Employees 45,000 50,000 66,000 86,000 96,000 100,000 110,000 Rate - 11.1 32.0 30.3 11.6 4.2 10.0 Year 1998 1999 2000 2001 2002 2003 Employees 132,000 148,000 156,050 156,500 160,200 160,500 Rate 20.0 12.1 5.4 0.3 2.4 0.2 (Source: Ministry of Natural Resources & Tourism) Table G. Proportion of tourism earnings on total exports and GDP, 1990 ­ 2003 Tourism Total ex- Tourism Proportion of Proportion of earnings ports earn- GDP at current earnings Tourism earn- Tourism earnings (US$ ings million prices ( mil) mil.) ings to GDP to total exports Year million) US$ 1990 65.00 538.40 760,005.00 11,115.00 1.5 12.1 1991 94.73 504.40 989,594.0 21,115.00 2.1 18.8 1992 120.04 564.50 1,275,917.0 42,014.00 3.3 21.3 1993 146.84 750.10 1,607,763.0 70,464.23 4.4 19.6 1994 192.10 937.60 2,125,324.0 100,555.24 4.7 20.5 1995 258.14 1,265.80 2,796,642.0 144,089.62 5.2 20.4 1996 322.00 1,298.70 3,452,558.0 191,797.17 5.6 24.8 1997 392.41 1,235.00 4,281,600.0 245,087.98 5.7 31.8 1998 570.00 1,109.80 5,124,924.0 380,047.50 7.4 51.4 1999 733.30 1,143.60 5,977,699.0 584,660.09 9.8 64.1 2000 739.10 1,290.60 6,706,381.0 593,719.03 8.9 57.3 2001 725.00 1,455.70 7,624,616.0 664,317.50 8.7 49.8 2002 730.00 1,568.30 8,699,887.0 712,699.00 8.2 46.5 2003 731.00 1,827.60 9,811,553.0 777,491.60 7.9 40.0 (Source: Ministry of Natural Resources and Tourism, Bank of Tanzania) Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 48 Table H. Employment opportunities in the TIC approved projects on tourism (2001 ­ 2003) TOURISM YEAR PROJECT VALUE EMPLOYMENT 2001 39,666.00 1,967 2002 85,690.00 4,329 2003 127,157.00 4,350 Total 252,513.00 10,646 (Source: Tanzania Investment Centre) Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 49 Table I. Revenue collected from National Parks by park, 1990 - 2001 National Park 1990 1991 1992 1993 1994 1995 Serengeti 241,492 302,100 374,813 612,974 1,086,787 562,040 Manyara 139,562 166,934 202,353 253,019 492,930 273,650 Ngorongoro 462,894 444,353 519,240 762,499 1,468,983 803,487 Arusha 21,034 20,455 55,735 71,958 205,560 82,879 Mikumi 118,053 21,123 31,398 34,947 141,289 52,154 Ruaha 6,325 8,010 13,214 26,766 48,896 22,793 Tarangire 56,816 65,954 113,344 134,304 461,501 182,544 Kilimanjaro 360,708 310,128 496,189 513,804 1,691,394 743,864 Gombe 8,436 14,376 17,605 22,409 15,620 16,263 Total 1,415,320 1,353,433 1,823,891 2,432,680 5,612,960 2,739,674 National Park 1996 1997 1998 1999 2000 2001 Serengeti 606,058 650,076 694,093 738,111 782,128 826,146 Manyara 294,794 315,939 337,084 358,229 379,374 400,519 Ngorongoro 865,489 927,492 989,494 1,051,497 1,113,499 1,175,501 Arusha 89,496 96,113 102,729 109,346 115,962 122,579 Mikumi 56,268 60,382 64,497 68,611 72,726 76,840 Ruaha 24,585 26,377 28,169 29,962 31,754 33,546 Tarangire 197,127 211,711 226,294 240,878 255,461 270,044 Kilimanjaro 801,945 860,027 918,108 976,190 1,034,271 1,092,353 Gombe 17,510 18,758 20,005 21,252 22,500 23,747 Total 2,953,272 3,166,875 3,380,473 3,594,076 3,807,675 4,021,275 (Source: Ministry of Natural Resources & Tourism) Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 50 Table J.Revenue collected from National Parks by year, 1990 - 2001 Proportion of Revenue By TOURISM Parks revenue to Proportion of National Parks REVENUE (Tshs Total Tourism parks revenue YEAR (Tshs 000) GDP (Tshs 000) 000) revenue to GDP 1990 1,415,320 760,005,000.0 11,115,000.0 12.7 0.19 1991 1,353,433 989,594,000.0 21,115,000.0 6.4 0.14 1992 1,823,891 1,275,917,000.0 42,014,000.0 4.3 0.14 1993 2,432,680 1,607,763,000.0 70,464,230.0 3.5 0.15 1994 5,612,960 2,125,324,000.0 100,555,240.0 5.6 0.26 1995 2,739,674 2,796,642,000.0 144,089,620.0 1.9 0.10 1996 2,953,272 3,452,558,000.0 191,797,170.0 1.5 0.09 1997 3,166,875 4,281,600,000.0 245,087,980.0 1.3 0.07 1998 3,380,473 5,124,924,000.0 380,047,500.0 0.9 0.07 1999 3,594,076 5,977,699,000.0 584,660,090.0 0.6 0.06 2000 3,807,675 6,706,381,000.0 593,719,030.0 0.6 0.06 2001 4,021,275 7,624,616,000.0 664,317,500.0 0.6 0.05 (Source: Ministry of Natural Resources and Tourism) Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 51 Annex 3. List of References Abila, Richard O. and Jansen, Eirik G. (1997) From Local to Global Market. The Fish Exporting and Fishmeal Industries of Lake Victoria Structure, Strategies and Socio-economic Impacts in Kenya, IUCN, The World Conser- vation Union Abila, Richard O. (2000), The Development of the Lake Victoria Fishery. A Boon or Bane for Food Security?, IUCN, The World Conservation Union Appleton et al. (2004) "Assessment of the Environment in the Rwamagasa area, Tanzania", J. D. Appleton, H. Taylor, T. R. Lister, & B. Smith in Final Report for an Assessment of the Environment and Health in the Rwamagasa area, Tanzania. UNDIO Project EG/GLO/01/G34, Nottingham: British Geological Survey/ National Environment Research Council Barrick Gold Corporation (2005) Heart of Gold Fund, available at http://www.barrick.com/index.aspx?usesid=-1&sid=184. Bokea Crispin and Ikiara, (2000), The Macro economy of the Export Fishing Industry in Lake Victoria (Kenya), The IUCN, The World Conservation Un- ion. CEEST (????) Environmental impacts of small scale Mining: A case study of Merelani, Kahama, Nzega, Geita and Musoma, CEEST Research Report Se- ries no.1, Dar es Salaam: The Centre for Energy, Environment, Science and Technology Drasch, Gustav and Stephan Boese-O'Reilly (2004) "Assessment of Health in the Rwamagasa area, Tanzania"in Final Report for an Assessment of the Envi- ronment and Health in the Rwamagasa area, Tanzania. UNDIO Project EG/GLO/01/G34, Nottingham: British Geological Survey/ National Environ- ment Research Council EIU (2004) Tanzania ­ Country Profile 2004, London: Economist Intelligence Unit. George, Angelo Kitula Ndekeja (2003) Socio-Economic Impacts of Mining on the Livelihood of Local Communities in Geita District, Tanzania, M.A. Dis- sertation, Sokoine University of Agriculture. Gibbon, Peter (1997) Of saviours and punks: The political economy of the Nile Perch marketing chain in Tanzania. CDR Working Paper 97.3, June Hangi, A. Y. (2001) "Tanzania's pot of gold" news.bbc.co.uk, 22. July 2001 Hoadley, M., Limpitlaw, D. and Weaver A. (2002) Mining, Minerals and Sus- tainable Development in southern Africa ­ Volume 1: The Report of the Re- gional MMSD Process, Wits: University of the Witwatersrand. IMF (2004) Tanzania ­ Selected Issues and Statistical Appendix, Washington, D.C.: IMF, September 2004. Jambiya, G. et al. (1997) Poverty and the Environment: Informal Sand-mining, Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 52 Quarrying and Lime-making, G. Jambiya, K. Kulindwa, and Sosovele, Policy Brief No. 97.1, Dar es Salaam: Research on Poverty Alleviation. Jansen, Eirik G. (1997) Rich Fisheries Poor FisherFolk. Some Preliminary Observation About the Effects of Trade and Aid in the Lake Victoria Fisheries, IUCN, The World Conservation Union Jansen, Eirik G., Abila, Richard O. And Owino John P. (1999) Constraints and Opportunities for `Community Participation' In the Management of the Lake Victoria Fisheries. IUCN, the World Conservation Union. Kallonga, Emmanual; Rodgers, Alan; Nelson, Fred; Ndoinyo, Yannick and Rugemeleza, Nshala (2003) Reforming Environmental Governance in Tanza- nia: Natural Resource Management and the Rural Economy, non- commissioned paper presented at the Inaugural Tanzania Biennial Develop- ment Forum, 24-25th April 2003, Dar es Salaam. Kudoja, William (2004), LVFRP conducting sonar fish surveys in Lake Victo- ria. ICT update: A current awareness bulletin for ACP agriculture. Issue No. 16, March Kulindwa et al. (2003) Mining for Sustainable Development in Tanzania, Kas- sim Kulindwa, Oswald Mashindano, Fanuel Shechambo, and Hussein Sos- ovele, Dar es Salaam: Economic Research Bureau Kulindwa, K (2001), The Contribution of Lake Victoria Fisheries to the Na- tional Economy. A Report submitted to LVEMP, Fisheries Research Compo- nent, Socio-economic sub-component. Kulindwa, K., Sosovele, H and Mashindano, O. (2001), Tourism Growth for Sustainable Development in Tanzania. Dar es Salaam University Press Kweka, J., Morrissey, O. and Blake, A (2003), The Economic Potential of Tourism in Tanzania, Journal of International Development. Willey Inter- Science (www.interscience.wiley.com) Lange, G-M., R. Hassan and K.Hamilton (2003): Environmental Accounting in Action case studies from Southern Africa. Edward Edgar, Cheltenham, UK Law Reform Commission (2001) Position paper on the legal framework for the development of the Mining industry, presented to the Commission June 2001. MacAlister Elliot and Partners (1999) Regional Socio-economic Studies on Employment and the Level of Dependency on Fishing. Greece (Lot II). Final Report, Nov. European Commission DGXIV (Fisheries). MNRT (2000) Lake Victoria Frame Survey Results Mugurusi, E.K (2002), Tourism Investment Forum, Dar es Salaam Mutagwaba et al. (1997) Poverty and Technology: The Case of Artisanal Min- ing, W. Mutagwaba, R. Mwipopo-Ako, and A. Mlaki, Policy Brief No. 97-2, Dar es Salaam: Research on Poverty Alleviation. Mwaipopo et al. (2004) Increasing the Contribution of Artisanal and Small- scale Mining to Poverty Reduction in Tanzania ­ Based on an Analysis of Mining Livelihoods in Misungwi and Geita Districts, Mwanza region, Rose- Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 53 marie Mwaipopo, Wilson Mutagwaba, and David Nyange with Eleanor Fisher, prepared for DFID, October 2004. Mwalyosi, Raphael B. B. (2004) "Impact Assessment and the Mining Industry: Perspectives from Tanzania", presentation made at 2004 IAIA Vancouver Canada. National Bureau of Statistics (2001) Integrated Labour Force Survey, 2000/01 ­ Analytical Report, Dar es Salaam: NBS. Phillips et al. (2001a) Tanzania's Precious Minerals Boom ­ Issues in Mining and Marketing, Lucie Phillips, Haji Semboja, G. P. Shukla, Rogers Sezinga, Wilson Mutagwaba and Ben Mchwampaka with Goodwill Wanga and Godius Kahyarara, Dar es Salaam: Economic and Social Research Foundation and Ar- lington: International Business Initiatives Corp. Phillips et al. (2001b) Growth and Equity: Gemstone and gold Mining in Tan- zania, Lucie Phillips, Rogers Sezinga, Haji Semboja, and Godius Kahyarara, EAGER policy brief number 56. Tassell, Arthur (2004) "Tanzania's gold boom continues", available full text at http://www.mining.co.za/Tanzania.htm, Johannesburg: Brooke Patrick Publi- cations. UNIDO (2002) "FEATURE: Artisanal Gold Mining without Mercury Pollu- tion", available full text at http://www.unido.org/en/doc/4571, Vienna: UNIDO. URT (1997) Mineral Policy of Tanzania URT (1998) Mining Act URT (2002) The Masterplan Study on Fisheries Development in the United Republic of Tanzania. Main Report. JICA/MNRT. Dar es Salaam. URT (2002), The 2002/3 Tanzania Participatory Poverty Assessment, http://www.esrftz.org/ppa. URT (2004), Economic Survey for 2003, Planning and Privatisation Commis- sion, Dare s Salaam, Tanzania. URT (2005) National Strategy for Growth and Poverty Reduction (NSGRP), United Republic of Tanzania, Vice President's Office, final draft, 15. January 2005 URT (undated) "Mining", available at http://www.tanzania.go.tz/mining.html, Dar es Salaam: Government of Tanzania. Van Campenhout, Bjorn (2002) The Mining industry and the future develop- ment of Tanzania, Workshop on Globalisation and East Africa, 15-16 April 2000, Dar es Salaam: Economic and Social Research Foundation. Van Straaten P (2000), "Mercury contamination associated with small-scale gold mining in Tanzania and Zimbabwe", Abstract, Department of Land Re- source Science, University of Guelph, Ontario, Canada. Williamson R (2003), "Private Foreign Investment and the Poorest Countries", Study on Growth and Environment Links for Preparation of Country Economic Memorandum ­ Part 3 54 Wilton Park paper based on the Wilton Park Conference Number 707 on Pri- vate Foreign Investment and the Poorest Countries", Wilton House, West Sus- sex, UK World Bank (2001) Tanzania: women in the Mining sector, findings, Washing- ton, D.C.: The World Bank World Bank (2002), World Development Report 2003: Sustainable Develop- ment in a Dynamic World. Washington, DC World Bank (2005) Concept Note for the Tanzania CEM/Poverty Assessment, World Bank, unpublished, 13. September 2004. World Bank (2005b) Blueprint 2050 ­ Sustaining the Marine Environment in Mainland Tanzania and Zanzibar, edited by Ruitenbeek, Jack; Hewawasam, Indumathie and Ngolie, Magnus, Washington, D.C.: The World Bank. Yager, Thomas R. (2003) "The Mineral Industry of Tanzania" in U.S. Geologi- cal Survey Minerals Yearbook 2003. The World Bank Study on Growth and Environment Links for Preparation of Country Economic Memorandum (CEM) Natural Resource Based Growth Summary Paper May 2005 The World Bank Study on Growth and Environment Links for Preparation of Country Economic Memorandum (CEM) Natural Resource Based Growth Summary Paper May 2005 Report no. 1 Issue no. 1 Date of issue 18 May 2005 Prepared KEP Checked TNH Approved iii Table of Contents 1 The Contribution of Natural Resources to Growth 1 1.1 Accounted and Unaccounted Contributions 2 1.2 Public Investment in Natural Resource Based Growth 4 2 Untapped Growth Potentials 6 3 Sustainability of Growth 9 4 Externalities 12 5 Recommendations 14 5.1 General Recommendations 14 5.2 Sector Specific Recommendations 15 Study on Growth and Environment Links for Country Economic Memorandum ­ Summary Paper 1 1 The Contribution of Natural Resources to Growth Natural resources in Tanzania constitute a wealth asset. During the past dec- ade Mining, Fisheries and Tourism have been the most dynamic sectors in the economy. Although Tourism development is a success story of growth in macro-economic terms, local development spin-off effects could be explored more fully. Most currently known mineral deposits are being already tapped while at the same time new mineral stocks are being discovered.1 Fisheries is still a growing sector but there are signs of decline in Catch per Unit in Lake Victoria and catch of fish and prawn in the coastal zones, which points towards a deceleration of growth in these sectors in the medium and long term. Forestry, Wildlife and Marine Fisheries resources, though declining, are still relatively abundant, rendering largely un-tapped growth potentials. Although these natural resources like labour and capital contribute to the economy and subsistence base of the rural population, their value and potential is underes- timated. This underestimation is partly based on missing markets in the case of pub- lic goods, imperfect competition in the case of distorting government inter- ventions as well as pricing of natural resources below market value. Results of all these market failures lead to sub-optimal economic decision making and loss of income to the country. The National Strategy for Growth and Poverty Reduction (NSGPR 2005) subscribes to the principles of sustainable and equitable development. Op- erational starting points of these principles include that: · Renewable resources should be exploited on a profit-maximising sus- tained yield basis and not driven to extinction, regardless of the dictates of present value maximisation. Hence, harvesting rates should not ex- ceed regeneration rates and waste emissions should not exceed absorp- tive capacities. 1Daily News of 18. May 2005. Study on Growth and Environment Links for Country Economic Memorandum ­ Summary Paper 2 · Non-renewable resources should be exploited at a rate equal to the crea- tion of renewable substitutes. Revenue from the exploitation of nonre- newable resources should contain an 'income component' and a 'capital component', the latter being used to invest in building up a new renew- able asset to replace the nonrenewable one at the point of exhaustion. · Revenue generated from natural resources should be shared equitably, in particular with rural communities on whose land these resources occur. These macro-economics of sustainability would require an integration of GDP growth and qualitative development more fully, giving equal weight to need for pro-poor growth and maintenance of a sustainable natural resource base. This summary paper as well as the three detailed background papers defend the hypothesis that, due to policy failure, Tanzania's natural resource en- dowments are not harnessed in an optimal way to both economic growth and poverty reduction. On the contrary, due to weak governance regimes in revenue generating sec- tors, resources are offered below market price to the benefit of a few power- ful winners, and at the loss of the majority of the rural population. Yet, these natural resources provide substantive potential for income to communities in rural areas. The weaknesses in governance regimes in all three sectors, Forestry, Wild- life and Fisheries, include primarily lack of transparency and accountability in issuing rights of resource extraction and revenues accrued thereof, lack of equitable sharing of benefits with communities, as well as weak monitoring and surveillance of stocks. In all four principal sectors providing natural capital to the Tanzanian growth equation, i.e. Forestry, Wildlife, Fisheries and Mining, royalties are set arbitrarily and do not reflect scarcity. Royalties are hence not used as a policy instrument of intertemporal resource pricing and sustained yield man- agement. As long as these weaknesses are not addressed, a substantial base of eco- nomic growth will be slowly eroded and poverty reduction objectives are unlikely to be achieved. 1.1 Accounted and Unaccounted Contributions Commonly for Forestry, Wildlife and Fisheries, a great share of their eco- nomic contributions does not enter GDP and export statistics and is hence not taken into account in the analysis of growth. Availability and quality of data is a general problem. Study on Growth and Environment Links for Country Economic Memorandum ­ Summary Paper 3 As part of this study, various sources have been analysed as an attempt to provide an inventory of the overall contribution of natural resources sectors to the economy. The picture is patchy and mainly based on estimates. An overview of annual revenue earned by the Ministry of Natural Resources and Tourism from its key departments in the last two financial years is pre- sented in Table 1 in the Annex. Although revenue is an important measure for growth, it does not capture all contributions to economic and rural development by the respective sectors. Forestry provided over Tsh. 5 billion in Government revenue over the last two financial periods. It contributes officially 2-3 percent to GDP and a 10- 15 percent share of export earnings. Estimates taking unaccounted services and non-industrial forestry into account are accounting for a value of 10-15 percent of GDP. Forests provide around 75% of building materials and 100% of indigenous medicinal plants and supplementary food products. In addition, Forests provide an important component of value added to na- tional income through their eco-system service functions providing for in- dustrial and domestic water and energy supply. 95% of Tanzania's energy consumption is woodfuel based, which includes major inputs factors into rural industries such as for example tobacco curing and fish smoking. For- ests provide watershed functions for major rivers feeding into the national hydropower dams. Lack of reliable power and water supply can hamper growth in the longterm and are already being cited as a serious constraint to attracting private investment. Further to their 'source' functions, forests also have 'sink' functions, absorb- ing and neutralizing negative externalities of economic growth, most impor- tantly pollution. The value of carbon sequestion services provided by Tan- zanian forests is estimated to be between 700 and US$ 1,500 per ha. Addi- tional environmental service functions include inputs from land and forests into agricultural production. The revenue generated from Wildlife resources, accrued to the MNRT mainly from hunting licenses was over Tsh. 9 million during the last two financial periods. An independent study of the sector, quotes annual earn- ings of about US$ 30 million from tourist hunting and an additional US$ 9 million generated by the private companies leasing hunting concessions from the government (2001). In the year 2002 live animals exports earnings amounted to roughly 170.000 US$. The largest income earner is the non-consumptive use of wildlife resources through game viewing by international tourists. In 2001 Tanzanian national parks drew over 100,000 international visitors. This generated receipts of almost 5% of GDP, equivalent to about US$ 400 million US$. In addition wildlife provides unaccounted subsistence values. Well over two-thirds of people eat wild meat, with up to 95 per cent of the rural popu- lation claiming it is their most important meat protein source. Study on Growth and Environment Links for Country Economic Memorandum ­ Summary Paper 4 Tanzania's Fisheries sector has grown at a rate of 6 to 7 percent annually since 2000. In 2004 revenue collection from Fisheries amounted to Tsh. 9.7 billion. This represents roughly a 50 percent increase from revenue collected in 2001/02. While about 80 percent of revenue is coming from freshwater fisheries (2003), only 20 percent of revenue originates from marine fisher- ies. However, the number of foreign vessels licensed to operate in the EEZ (Mainland and Zanzibar) has increased from less than ten in 1998 to more than 170 in 2004 corresponding to a revenue of US$ 3.3 million. In terms of export earnings, Fisheries contributed 10% of total exports in 2003, which equalled US$ 130 million, the export value of Nile Perch being US$ 100 mill. Fisheries registered a revenue over-collection of roughly Tsh. 3 billion in 2003/04. A great share of the marine catch does not enter GDP and export statistics and plays an important role in livelihoods support. The official number of artisanal fishermen has doubled since 1995 and reached close to 120,000 in 2003. Although contribution to GDP is still not more than 1.9 percent, Mining is the single most important earner of foreign exchange to the country. About 50% of export earnings accrue from Minerals, predominantly from gold mining by large-scale foreign owned operators. In addition, Mineral re- sources are of importance to the artisanal mining sector. 1.2 Public Investment in Natural Resource Based Growth It is obvious from the data above that these revenue generating sectors are making an important contribution to both the formal and subsistence economies. However, out of the three natural resource sectors, only Fisher- ies is a net contributor to Treasury, while Forestry and Wildlife are subsi- dized through government allocations to cover their recurrent expenditures and through foreign grant allocations to finance their operational budgets. Table 2 in the Annex shows the government recurrent budget allocations to the respective sectors. Forestry and Wildlife each received 29% of the Budget in the last financial year, followed by Fisheries (18 percent) and Tourism (11 percent). There is a mismatch between foreign resource allocation for sectoral devel- opment activities and national funding allocation for recurrent expenditures. The large degree of under-spending of the development budgets in both sec- tors is a possible indication of capacity constraints to absorb foreign funding and institutional inefficiencies, aggravated by uncoordinated policies of the development partners. In the Forest sector, a planned SWAP is supposed to address the latter problem. There is a tendency to draw government allocation away from those sectors, as they should finance themselves and move towards privatization. However there are trade-offs to this trend. There is a need for government to control and regulate, setting and enforcing fiscal and market instruments to ensure Study on Growth and Environment Links for Country Economic Memorandum ­ Summary Paper 5 sustained growth and incorporation of externalities. While the recurrent government budget allocations to these sectors are reducing, combined with a trend towards privatizing some of the government functions, there is an increasing need for sector wide environmental management functions, such as environmental impact assessment, market based instruments for environ- mental protection (e.g. taxes, subsidies, standards, permits etc.), monitoring of stocks and legal enforcement, which are becoming increasingly more im- portant with increased economic growth. Due to institutional failures, these over-arching environmental management functions have basically been lack- ing in Tanzania for the last decade. Although the new Environmental Management Act provides the necessary environmental framework law, it is still a long way to see the effects of its implementation. Increased public investments are needed to support these broad based environmental management activities. Study on Growth and Environment Links for Country Economic Memorandum ­ Summary Paper 6 2 Untapped Growth Potentials With regard to non-consumptive wildlife utilization for game viewing tour- ism, there is untapped potential in the Southern Parks of Tanzania. While the northern circuit has supposedly reached maximum carrying capacity in terms of numbers of visitors, places like Ruaha and Katavi National Parks are still fairly unknown. Shifting marketing and infrastructure development to those areas would provide new growth potential for the Tourism sector. In addition, there is scope to increase concession fees of international tour op- erators and lodges, which currently account only for 2% of TANAPA's revenue. Marine Fisheries has recorded a sharp revenue increase due to increased li- cense revenue from foreign vessels in the EEZ (see Chart 1 in Annex). There are estimates that the presently earned revenue is not reflecting the total amount that the government could earn and that real catch is much higher than what has been assumed as a basis to set the license fees (be- tween 200 and over 400 tons a day per boat). Notably, there is no catch based license or fee, and the vessels are allowed unlimited catch once they are in possession of a valid license. Although the above revenue figure is a considerable amount, it is low compared to the estimated value of the catch by foreign vessels in Tanzanian waters sold on foreign markets. This gives rise to the assumption that there is scope of revenue increases. With regard to Freshwater Fisheries, past growth rates are mainly based on Nile Perch Exports from Lake Victoria. Other lakes, such as Lakes Tangany- ika and Nyasa, as well as harvesting of other species are so far commercially under-developed. Diversification could also be sought in terms of exploring additional export markets. Risks and vulnerability increase in a situation where export earn- ings in a sector are entirely dependent on a single market. This is the case of Fisheries export from Lake Victoria, which is mainly destined for the EU. Previously following unfavourable assessment of sanitary standards in Tan- zanian fish processing plants halted the entire production for several months due to an important ban. Despite high growth in the fisheries sector and existing local production of fishnets, the proportion of imported fishnets is 95%. Hence, an important backward linkage to the industry and employment opportunity remains un- exploited. Study on Growth and Environment Links for Country Economic Memorandum ­ Summary Paper 7 Commercial Fisheries presents an important, emerging revenue source for the country and the sector. If the sector is well managed, commercial Fisher- ies can have a positive impact on economic growth and poverty reduction at the same time. Principles of management would need to include retention and re-investment of revenue into the sector and putting up safeguards to the artisanal Fisheries to protect their rights and access to the resource. Potential for local spin-off effects Local spin-off effects are presently missing in the context of marine fisheries in the Exclusive Economic Zone. While new Fisheries Agreements are being negotiated with foreign countries, no fish is expected to be landed ashore and few supplies will be sourced from within Tanzania. If none such spin- off effects are created, the net impact of commercial fisheries on poverty reduction may be negative, provided increased completion with artisanal Fisheries over the same resource. Similarly, the fact that Tanzania is a net importer of forest products is a sign of lost opportunities for income generation for the local economy. Similarly, the Mining sector seems to have had limited influence on poverty reduction in the local economy. Employment in the large-scale Mining sec- tor is limited although especially younger employees may receive significant salaries. The majority of those employed in the Mining sector operate, mostly as self-employed, in the small-scale sector, typically as artisanal miners. This category receives very low returns, especially when one con- siders the hardship associated with this kind of employment. It seems fur- thermore that an increasing income disparity is emerging between those em- ployed in the small-scale and large-scale Mining sector. To the extent that those recruited by the large-scale Mining-sector are recruited outside the local community, the local community is thus restricted to opt for poorly paid employment opportunities in the small-scale sector. Large-scale Mining may have positive effects for local communities through the improvement of basic infrastructure. There is however no indication that the expansion in the Mining sector triggers significant growth in the local economy, since Mining operations generally are detached from local supply chains and therefore primarily create employment in the services sector. Potential for poverty reduction In addition to their potential for government revenue generation Wildlife, Fisheries and Forestry resources provide the non-agricultural subsistence base for rural communities in remote locations. Increased emphasis on natu- ral resources related enterprises has potential to create additional income opportunities for the rural population. For example in Loliondo Division in Ngorongoro District, seven villages earn a total of over US$ 110,000 per annum from joint ventures with wild- life tour operators. In Ololosokwan village, tourism revenue totals about US$ 55,000 per annum. The income from payments by one of four tour op- Study on Growth and Environment Links for Country Economic Memorandum ­ Summary Paper 8 erators in Olosokwan Village is shown in Chart 2 in the Annex. If effects of elite capture are avoided and income equitably distributed within the com- munities, this income has large poverty reduction potential in a dryland area, which does not offer many other opportunities of diversification. While the case of Olosokwan is an exceptional example, the potential for local development from wildlife related tourism has not been fully tapped in other areas. In the Mara-Serengeti ecosystem, the number of households earning any income from tourism varies from 86% in Talek (Kenya) to 12% and 3% at the Ngorongoro Conservation Area and Loliondo Game Reserve on the Tanzanian side. In the southern circuit, tourism is growing offering potential scope for posi- tive impacts on local economic development. Participatory Wildlife Man- agement in communities close to Ruaha National Park, Iringa District, cre- ated a total of Tsh. 15 million in 1999 in local income accrued through earn- ings from resident hunting quota. In addition, an additional Tsh. 4,1 million was earned from the 25% share of license fees from tourist hunting (see Chart 3 in Annex). The income from hunting quotas was sufficient to treble village level com- munal income, enabling villages to pay district-level taxes, which would otherwise be levied by households, as well as to carry out social infrastruc- ture investments. One of the success factors identified was that the project has placed emphasis on institutional capacity building at village and inter- village levels. Similarly, community based forest management has provided revenue to villages across Tanzania. The 2002 Forest Act authorises villages to sell timber from their own forest reserves, which has potential to provide a new and additional source of forest revenue, directly accruing to the communi- ties. Despite the conducive policy framework in both the Wildlife and Forestry Sectors, weak governance systems at both central and local levels have so far limited the realization of the poverty reduction potential through com- munity based natural resources management. The main focus within community wildlife management has been on institu- tions and distribution of benefits rather than enterprise opportunities at household level. Fear over inequity has led to the relative neglect of entre- preneurship in Tanzania, reflecting persistent and much broader philosophi- cal bias against private enterprise. Study on Growth and Environment Links for Country Economic Memorandum ­ Summary Paper 9 3 Sustainability of Growth In the context of sustainability of natural resources based growth, the fol- lowing constraints are emerging in the Tanzanian context, which lead to revenue loss and possible deceleration of growth in the longterm: Under-pricing of resources not allowing capturing of resource rents; Weak environmental governance systems; Limited knowledge of stocks, their values and changes over time. Under-pricing of resources Sustainable growth based on renewable resources requires that the cost of extracting a resource and the notional cost of replacing a unit of the re- source, commonly known as 'resource rent', is evaluated so that the wealth base is not eroded. While royalties are the most important source of gov- ernment revenue in the Forestry (83%), Wildlife (96% hunting licenses) and Fisheries (84% royalties, 15% export licenses) sectors, they are set arbitrar- ily and capture neither market values nor resource rents. Similarly, in the Mining sector, licenses to foreign investors do not take the 'capital component' into account. Tax incentives have been generous to- wards foreign investors to attract capital investment and to open up the mar- ket at the expense of sustainability principles. In such a scenario, acceleration of growth comes at the expense of pricing resources below market value, which leads to loss of income, erosion of critical stocks and an associated deceleration of growth in the longterm. Estimates of resource rents from Marine Fisheries computed from license fees as percentage of value of revenue through licenses for foreign vessels fishing in the Tanzanian Exclusive Economic Zone (EEZ), show that gross resource rent is approximately 2.2 percent, which is less than half of what might be expected in a western industrial fishery. While the current license fee arrangements of private fisheries agreements (PFAs) in the EEZ generate not insignificant amounts of revenue, the level is too low to result in a rea- sonable return of revenue (>5 to 7 percent of gross revenue) to capture a re- source rent. Study on Growth and Environment Links for Country Economic Memorandum ­ Summary Paper 10 Similarly in Forestry, royalties have in the past been fixed arbitrarily. The 2002 Forest Act demands the determination of royalties based on market value, profitability and principles of sustainable harvesting. Improvement of the Forest Produce Pricing System would include market-base pricing of forest produce and public auctions or tendering for timber lots. Royalties could also be used as an instrument to divert harvest from pressured species towards lesser-known species. In the Wildlife sector, the concession component of TANAPA's earnings is only 2%, which most likely under-represents the value of these concessions compared to the income they generate to the foreign investor. Loss of reve- nue and unsustainable use is also fostered through hunting quotas that does not reflect true market values and is not based on ecological monitoring to maintain critical stocks. Presently, concessions are leased at rates far below true market value irrespective of size, quality or income potential. This represents a massive loss of income to the Wildlife Division (estimated at over US$ 7 million). The system promotes subleasing to foreigners with a result that much of the income generated by the industry never enters the country and substantial tax revenue is lost. Weak environmental governance In Forestry, an undercollection of 5 to 10 percent of revenue is reported due to inefficiencies in revenue collection and corruption in the sector. As the 2004 logging scandal in Rufiji revealed, the value of the illegally harvested logs was valued at Tsh.382.65m. Illegal trade of ivory from Tanzania to the Middle East, is estimated to cause a 200 million US$ revenue loss through a single consignment of ivory, evaluating the value of those elephants with their trophy price. In this case, corruption is based on organised criminal networks involving the police and government officials. In Marine Fisheries, there is alleged information that Zanzibar licenses for foreign vessels are registered in Muscat, Oman with the fees escaping the Zanzibar authorities. In EEZ fisheries lack of transparency is attributed to a large degree through lack of catch reporting by foreign vessels. The govern- ance regime in EEZ fisheries is unique in the sense that it imposes responsi- bilities for transparency and accountability on the Distant Water Fishing Na- tions. In the Wildlife Sector, a non-transparent system of quota setting for the hunting industry by the Government leads to imperfect competition in the market. There is no competitive bidding for hunting concessions but distri- bution through autonomous government decision making. Effective market forces are hence not applied to optimize revenues. This policy intervention leads to a 'monopoly of knowledge' by the Wildlife Division and an 'oli- gopsony' in terms of access to the resource, a situation in which a small number of large buyers controls the market. Consequently, quotas are sold below market value leading to a loss in revenue. While imperfect competi- tion usually benefits a few powerful players, it is at the disadvantage to the Study on Growth and Environment Links for Country Economic Memorandum ­ Summary Paper 11 majority of the population. It leads to loss of income and livelihoods for rural communities. Limited knowledge of resources stock values and stock changes The optimal scale of natural resource based economic growth must be at a sustainable level. Hence, a general macro-level constraint of growth is then that the optimal scale is the one at which the long-run marginal cost of ex- pansion are equal to the long-term marginal benefits of expansion. This macro-level constraint of optimal growth cannot be operationalized if the true costs of resource extraction are unknown. Commonly in Fisheries, Forestry, Wildlife and Mining, there are neither in- ventories of the full availability of stocks nor full information about their value. In addition, stock changes are not monitored in a comprehensive manner. In the absence of stock and flow data, limits of extraction and quota associated with licenses can hence only be set arbitrarily and hence not based on sound ecological calculations and realistic projections. For exam- ple in Marine Fisheries, there are no catch limits at all attached to licenses, allowing vessels to take as much fish as is available while scant information on actual catch is returned by the foreign fishing vessel to the coastal state (Tanzania). Similarly, in Forestry, land coverage, deforestation and values represented in the country's forest estate are a matter of speculation. There is already government effort in the Fisheries and Forestry sectors to address some of these problems. For example the Fisheries Department has lately increased its monitoring, control and surveillance with support of the SADC regional project. The Forestry Department is in the process of devel- oping a National Forest Monitoring Facility and Database. Study on Growth and Environment Links for Country Economic Memorandum ­ Summary Paper 12 4 Externalities Consideration and efficient control of 'externalities' is important to reflect the true cost of resource utilization preventing their over-exploitation. In addition, externalities can cause trade-offs between economic growth and poverty reduction as they can negatively affect local peoples' access to natu- ral resources. Lastly, control of externalities can realize cost savings otherwise spent on pollution control. Examples are abundant, a few are described below. Economic growth is associated with increased need for energy and water supply for domestic and industrial purposes. Currently, 95% of energy supply comes from biomass energy. Due to incorrect pricing, charcoal does not represent the full value of the wood being harvested. In terms of providing value-added to growth through energy and water supply, Tanzania's Forests provide 'critical capital'. Catchment Forests are one such example and their conservation is clearly a binding constraint to be addressed. Increased agricultural production and intensification can create external- ities. Large commercial rice farming in the Usangu Plains has reduced the dry season flow of the Great Ruaha River through intensified year round irrigation, which is negatively affecting water use by small scale farmers downstream. Commercial fish production for export markets at Lake Victoria erodes a base of livelihood and food supply for local fishing communities. Simi- larly, the penetration of foreign vessels into territorial seas impacts on the catch of artisanal fisheries. Mining poses a number of threats and possibilities to the local communi- ties as well as to the miners themselves. The nature and extent of these threats and opportunities cannot be assessed in detail due to lack of reli- able data. There are concerns that large commercial mining crowds out the artisanal sector. Also, there are indications of a number of negative social effects, notably child labour, HIV/AIDS, and gender imbalances. The evidence provided about the environmental effects of large-scale Mining suggest that Mining communities may suffer a number of severe Study on Growth and Environment Links for Country Economic Memorandum ­ Summary Paper 13 effects, spanning from direct and observable noise and erosion, to longer term pollution of air, water and soil, which in turn may have seri- ous health consequences. Still, the evidence does not allow for extrapo- lation and more rigid, comprehensive analysis is required to have a bet- ter idea of the environmental implications of large-scale Mining in Tan- zania. The current policy framework in Tanzania does not provide for sound man- agement of natural resources and the mitigation of externalities. Presently applied instruments for revenue generation do not address externalities, nor are they used as instruments to capture rents from natural resources. Rather than employing fiscal instruments to steer the exploitation of resources, there is, allegedly, tax evasion within the revenue generating public sectors themselves. Hence, in the present regime of environmental governance, increased growth will come at the cost of running down the resource stocks, impeding on longterm growth opportunities. In order to ensure a positive net effect of accelerated growth on positive poverty reduction, a careful balance needs to be drawn between increasing export earnings and maintaining the resource base for the artisanal sector. In particular in Fisheries, certain safeguards need to be put in place for the ar- tisanal fisheries to protect their rights, access to the resource and livelihoods. Key policy recommendations are formulated in the next Chapter. Study on Growth and Environment Links for Country Economic Memorandum ­ Summary Paper 14 5 Recommendations The operationalisation of sustainable development presents an international political challenge. In particular in the context of globally shared resources, such as Fisheries, responsibilities apply to both `harvesting' and `host' coun- try (Tanzania). While true factor pricing and resource rent capture are policy instruments that even some western countries are grappling with, there are some basic principles of governance that are missing in Tanzania, which if rectified, could re-gain some of the lost opportunities described in this paper. The single most important recommendation to capture and maintain natural resource based growth in Tanzania is to reform environmental governance so as to achieve 'good' governance, rule of law and equity. This would include ensuring greater coherence between different national policies and instruments, particularly community based wildlife manage- ment, tourism development, rural growth strategies, investment regulations and incentives and poverty reduction strategies. In addition, Tanzania would need to make investments into the improvement of its human capacity and capital stock so that value-added processing of natural resources can increasingly take place within the country. This would be required to comply with the principle states in the NSGRP that policies should be designed so that benefits from high-growth sectors are transmitted to the poor in form of better livelihoods opportunities, e.g. supporting supply linkages with local producers. The recommendations below are divided into general recommendations that equally apply to all natural resources sectors and sector specific recommen- dations. 5.1 General Recommendations Strengthen capacity for data collection, record keeping, monitoring, con- trol and surveillance and enforce punitive measures to control illegal practices. Control of externalities through fiscal instruments, royalties and resource pricing and to increase revenue from rent capture rather than un- controlled exploitation. Study on Growth and Environment Links for Country Economic Memorandum ­ Summary Paper 15 Increase efficiency in revenue collection and administration, as well as full transparency and accountability over revenue generation and distri- bution. Promote market based principles, where appropriate, ensuring local spin-offs and allowing competition and entrepreneurial development. 5.2 Sector Specific Recommendations Fisheries Put in place a regulatory framework and sound governance regime for marine Fisheries, comprising the EEZ and near-shore Fisheries. Safeguard rights and livelihoods for coastal communities, through for example demarcation of a Community Territorial Sea. Conduct a Fisheries Sector review to assess the economic and social, ecological and fiscal perspectives and policy options. This could inform policy makers and influence the strengthening of the regulatory frame- work. Establish some form of EEZ inspectorate in patrol terms to build up a more accurate picture of available resources. Investigate the potential for exports of (various) marine products and value adding of these products in order to promote growth in the coastal zone. Forestry Introduce taxes for wood lot and plantation owners, in particular an in- come tax based on timber sales, as well as a property tax based on aver- age productive capacity of different land categories. Enforce the collection of royalties and fees and include exempted indus- tries such a tobacco and fishing. Improve Forest Produce Pricing System through market-based pricing of forest produce; public auctions or tendering for timber lots; cheaper royalties to lesser-known species. Increase domestic and foreign private sector investment through reduc- tion of bureaucracy in licensing system, clear investment guidelines, clearly defined ownership of all forestland, tax incentives; credit facili- ties, and technology transfer. Increase capacity utilization of the sector to reverse the trade balance to net forest product export. Study on Growth and Environment Links for Country Economic Memorandum ­ Summary Paper 16 · Introduce new revenue sources, i.e. watershed management fees from hydropower stations, sale of genetic resources and carbon credits. Wildlife Encourage attitudinal change towards wildlife at policy level, as an asset for rural development and poverty reduction rather than as something looked after by conservationists. This will include a shift in emphasis of community wildlife approaches to focus on creating enterprise opportu- nities. Ensure that local communities are the principal decision makers for allo- cation of concessions and quota setting for hunting on their land, and they receive and manage the funds generated on their land. Reform the tourist hunting industry to realize the true revenue potential of the industry. This will include the introduction of market-based com- petition in the commercial hunting industry through competitive bidding for concessions. This may have the positive side effort of naturally con- trolling sub-leasing and related revenue losses. Introduce performance based independent monitoring of the hunting in- dustry, possibly through certification, to ensure certain standards are ad- hered to. Criteria should be set to consider maximum income from the least number of animals hunted and contribution towards protection and community involvement. Revise the quota setting system based on more objective criteria, com- puterization of hunting data; monitoring of trophy quality and age. Conduct a review by the Ministry of Finance into financial management and taxation procedures of the Wildlife Division to assess the strengths and weaknesses. This would include an inventory of the true value of hunting licenses. Tourism Integrate opportunities for pro-poor tourism into tourism strategies, set objectives in terms of local development impacts not just numbers of tourists or foreign exchange earnings. Ensure a pro-poor tourism growth programme to place attention on company practices, destination management, infrastructure develop- ment, procurement patterns, national training and regulation. Mining Ensure improved data collection on externalities in the Mining Sector, for example through more rigorous and systematic enforcement of EIAs. Revise pricing system to capture `capital component' of non-renewable mining resources. Study on Growth and Environment Links for Country Economic Memorandum ­ Summary Paper 17 Annex Table 1 MNRT Annual Revenue 2003 and 2004 Revenue in billion Tshs 2003 2004 Forestry 5.29 5.82 Wildlife 9.17 9.55 Fisheries 6.99 9.70 Tourism 0.83 0.96 (Source: MNRT 2004. Note: This includes revenue collected and retained at source) Table 2 Budget of MNRT, distribution by sub-sector, 2002/03 ­ 2003/04 2002/03 2003/04 Sub-sector '000 Tsh. % 000 Tsh. % Forestry and Beekeeping 4,897,656 24 7,633,912 29 Wildlife 6,593,025 33 7,586,736 29 Fisheries 3,688,280 18 4,648,202 18 Tourism 2,208,073 11 2,880,761 11 other Total MNRT 20,243,165 26,257,352 Note: The Total includes other sub-sectors not listed here. The amounts are only recurrent expen- ditures. Based on MNRT 2004. Chart 1Annual License Revenue and foreign vessels in Tanzanian EEZ 4,000,000 180 3,500,000 160 3,000,000 140 2,500,000 120 100 vessels US$2,000,000 80 of 1,500,000 60 No. 1,000,000 40 500,000 20 0 0 1998 1999 2000 2001 2002 2003 2004 US$ No of vessels (Source: Fisheries Department) Study on Growth and Environment Links for Country Economic Memorandum ­ Summary Paper 18 Chart 2 Income to Olosokwan Village, Ngorongoro District Council, 1999 - 200303 12,000,000 10,000,000 8,000,000 Village (Tsh.) 6,000,000 District (Tsh.) 4,000,000 2,000,000 0 1999/2000 2000/01 2001/02 2002/03 (Source: Reconstructed from Kallonga et al. 2003, figures not exact) Chart 3Village Total Incomes comprised of resident hunting quota and 25% share of license fees from tourist hunting 18,000,000 16,000,000 14,000,000 12,000,000 village income from 10,000,000 tourist hunting Tshs 8,000,000 village income from 6,000,000 resident hunting 4,000,000 2,000,000 0 1996 1997 1998 1999 (Source: Walsh 2000) DRAFT REPORT COMPONENT III: CAUSES OF MALNUTRITION AND TANZANIA'S NUTRITION PROGRAMS PAST AND PRESENT Tanzania Food and Nutrition Centre Email: md@muchs.ac.tz 16 April 2004 EXECUTIVE SUMMARY Nutrition work in Tanzania started at the turn of the century but it was not until the 1940s that a nutrition unit was established under the Ministry of Health (MoH). In 1973 with Sida support the Tanzania Food and Nutrition Centre (TFNC) was created as a parastatal organization under the same Ministry. The Centre has the mandate to carry out nutrition training, research and advise the government on matters relating to food and nutrition. During the 1940s prevention and control of malnutrition focussed on providing specific nutrients to correct known nutrient deficiencies. This approach known as the Nutrient Model was inadequate in addressing nutritional problems as it lacked considerations beyond the physiological basis of malnutrition. Then came the Food Cycle Model in the 1970s, which introduced the notion that malnutrition persisted because of food losses from farm to table and into the body. This model emphasized on the food aspect of malnutrition and neglected the non-food causes. As a result it was inadequate in tackling the malnutrition problem. Guided by experience in using the food cycle model and in realization of the complex nature of the nutrition problem, TFNC and UNICEF jointly developed a more comprehensive model known as the Integrated Conceptual Framework (CF). This model sees malnutrition as a sign of other problems in the society the cause of which may be immediate, underlying or basic to the problem. The CF emphasizes the use of information to assess a situation, analyze its causes and take action to improve the situation. (Triple A cycle). As the new situation emerges the triple A cycle is repeated iteratively. On the basis of prevalence data, Tanzania has identified protein-energy malnutrition (PEM), nutritional anaemia, iodine deficiency disorders (IDD) and vitamin A deficiency (VAD) to be the main nutritional disorders of public health significance. Population groups most affected are children below 5 years of age, pregnant and lactating women. Survey data show that wasting affects 5.3 percent, stunting 44.0 percent and underweight 29.5 percent of children aged below five years. The prevalence of anaemia among pregnant and lactating women is 80.0 percent while the prevalence of VAD among underfives is 24.2 percent and IDD among school children in endemic districts is 23.5 percent. i Using the CF the major determinants of malnutrition in Tanzania can be described at three levels of causation. In the first level are immediate causes often precipitating the observable manifestations of the nutrition situation. These causes include inadequate food intake (leading to numerous nutrient deficiencies) and diseases. The second level is underlying causes that include inadequate access to food, inadequate care for children and women, insufficient health services and unhealthy environment. These causes are often many and inter-related and most interventions aimed at reducing malnutrition are implemented at this level. The third level is basic causes that include poverty, inappropriate social and cultural beliefs and political and ideological factors. Control programs have been designed to address PEM and micronutrient deficiencies. One of the better documented program is the Iringa Nutrition Program (later Joint Nutrition Support Program or JNSP) whose objectives were to reduce infant and young child mortality and morbidity, improve maternal nutrition, improve the capabilities of society to assess and analyze nutrition problems and to design appropriate actions. The key elements of this program were social mobilization and community animation and application of triple A cycle approach to community based problem solving. The JNSP was successful in reducing the prevalence of severe malnutrition from 6.3 to 1.8 percent and total underweight from 55.9 to 38.0 percent in a period of 5 years. Also the program improved the survival of children. Based on this success in 1987 the program was gradually extended to other areas (currently 57 districts in 13 regions) of the country as Child Survival and Development Program (CSD). The annual cost per child in the CSD approach was estimated at $3. The Micronutrient Deficiency Control Programs notably Vitamin A deficiency control and Iodine Deficiency Disorders control have been implemented with notable success. With regard to vitamin A supplementation coverage is over 90 percent and the cost per child per year is $0.14. The IDD program likewise has succeeded in lowering goitre prevalence from 25 percent in 1980s to 8.1 percent in 2003. In early 1990s the World Bank supported a 5-year Health and Nutrition Project within the broader context of health sector reforms. This project had three components largely focusing support to facility-based health service delivery. Component I included control of micronutrient (iron and vitamin A) deficiencies, which was spearheaded by TFNC. A mid-term review of the project noted that knowledge of ii micronutrient deficiencies had increased among health staff and that vitamin A and iron/folate supplements were more widely available at health facilities. It is clear that PEM continues to be a major nutrition problem in Tanzania. Although PEM is mostly a problem of young children, it can also occur in older children and adults. Children with PEM are more likely than other children to have deficiencies of micronutrients such as iron or vitamin A. This situation is further complicated by the synergism between PEM and infections. In order to address this complex situation Tanzania will need to adopt a strategy that focuses on prevention and control of PEM on the one hand and control and prevention of micronutrient malnutrition on the other. In addressing these two major problems there is need to integrate efforts with other key complementary interventions and strategies that have impact on nutrition. Such interventions include EPI, IMCI, malaria control, parasite control and control of diarhoeal diseases. Strategies include prevention of low birth weight; growth monitoring and promotion; community based nutrition rehabilitation; improving household food security; supplementation and fortification, among others. In practice it is important to involve the community right from the conception and planning stages through to intervention. Elements of capacity development and intersectoral linkages must be seen as important and critical consideration as well. iii CAUSES OF MALNUTRITION AND TANZANIA'S NUTRITION PROGRAMS PAST AND PRESENT 1.0 Historical Development of Nutrition Work in Tanzania Nutrition work in Tanzania started at the turn of the century. In the 1940's a nutrition unit was established under the Ministry of Health (MoH). Famine in 1953/54 and other developments resulted in formation of a Multi-sectoral Central Advisory Committee on nutrition. After independence in 1961 the Committee was succeeded by the Tanganyika National Freedom from Hunger Committee and later the Tanzania Nutrition Committee. About the same time, in 1963 the president declared the intention of the Government to fight malnutrition. He invited all assistance possible from both inside and outside Tanzania. The first comprehensive nutrition plan (1965-1969) was developed with support from UNICEF, WHO and FAO. In the 1970's nutrition units were formed under the Ministries of Agriculture (MoA) and Education (MoE) independent of the MoH. This resulted in a demand for the establishment of a national nutrition-coordinating institute. Thus with Sida support Tanzania Food and Nutrition Centre (TFNC) was created in 1973 as a parastatal organization, originally under MoA, then under Prime Minister's Office and finally under MoH where it remains todate. Nutrition work in the 1940's focused on prevention and control of malnutrition based on the Nutrient Model. In the 1970's the approach was based on Food Cycle Model. This was later replaced by the Integrated Conceptual Framework Model, which brings in the need for multi-sectoral nutrition planning. 1.1 The Nutrient Model Research carried out over time on body composition and changes occurring due to malnutrition provided information on the specific functions of nutrients singly or in groups. In this way malnutrition was understood to be lack of any one or more of these nutrients. Thus lack of protein causes kwashiorkor while lack of carbohydrates and fats causes marasmus; lack of vitamins and minerals cause various disorders such as xerophthalmia, beriberi, scurvy, anaemia and goitre. Shortcomings of the nutrient approach as a tool for analyzing nutrition problems and as a basis for action included the following: · The model has no analysis beyond the physiological considerations. · It treats specific nutrient deficiencies in isolation while in practice people eat food and not pure nutrients. · It overlooks the social, cultural, economic and political factors, which have much influence on food availability, access and dietary practices. 1.2 The Food Cycle Model Food Cycle Model evolved in the 1970s. The line of thought now was that malnutrition persisted because of food losses at various stages ­ from farming through harvesting to utilization. The approach thus stressed on following up and detecting food losses at every stage and plugging the leakages. But soon it was to be realized that the food 1 cycle model was addressing only one side of the story ­ the food aspect of malnutrition. Non-food causes of malnutrition were still not being addressed. It was at this juncture that TFNC and UNICEF, basing on the experiences of the Iringa JNSP of the 1980s worked out another model for looking into the problem of hunger and malnutrition ­ the conceptual framework. 1.3 The Integrated Conceptual Framework During 1980's nutrition work in Tanzania had increasingly been oriented towards the adoption of a new framework for analyzing the causes of malnutrition taking the maximum holistic approach to the problem. This includes depth of analysis, level of analysis and time dimension. The framework has two components: the factual analysis that identifies the causes of malnutrition by indicating their depth according to immediate, underlying and basic causes. It also shows how the causes interrelate. The second component, the operational analysis operationalizes the process of assessment of the nutrition situation, analysis of the causes at various levels and deciding and designing actions to improve the situation. After the implementation of the actions the situation is reassessed and the process repeated. This is the approach that is being used to define nutritional problems and pointing out possible areas of intervention. The operationalization of the conceptual framework requires information at all levels ­ manifestation of the problem and immediate, underlying and basic causes. 2.0 Major Nutrition Problems in Tanzania Tanzania has identified four nutritional disorders as being of public health significance. These are protein energy malnutrition (PEM), nutritional anaemia, iodine deficiency disorders (IDD) and vitamin A deficiency (VAD). Other nutritional disorders do exist, but are of less public health significance, for example, obesity, chronic diet related non- communicable diseases and deficiencies of some vitamins and minerals. Considerable development has been made in the management and control of nutrition problems. But despite these efforts levels of malnutrition in Tanzania have remained high as shown in Table 1. 2 Table 1: Nutrition situation in Tanzania Type of malnutrition Year and group surveyed Prevalence Public (%) health significance PEM1 · Stunting (ht/age < -2 SD) 1999, under-fives 44.0 moderate · Underweight (wt/age < -2 1999, under-fives 29.5 moderate SD) · Wasting (wt/ht <-2 SD) 1999, under-fives 5.3 moderate 2. Anaemia2 1987, under-fives 45.0 severe 1987, pregnant/lactating women 80.0 severe 1987, remaining groups 20.0 moderate 1987, general population 32.0 moderate 3. IDD3 2003/2004, school children 8.1 acceptable 1999/2000, school children (in endemic 23.5 moderate districts) 4. VAD 4 1997, under-fives 24.2 major problem 1997, lactating women 69.0 major problem 1Tanzania Reproductive and Child Health Survey, 1999 2Kavishe FP. Nutrition Relevant Actions in Tanzania. TFNC Monograph Series No.1, 1993 3Evaluation of IDD control program. TFNC Reports No. 2002 and 1924 4Ballart et al. Report of the National Vitamin A Survey, 1997, TFNC Report No. 1880. 2.1 Protein Energy Malnutrition The 1999 Tanzania Reproductive and Child Health Survey shows that in children aged below five years wasting affected 5.3 percent, stunting 44.0 percent and underweight 29.5 percent. These data show that PEM was a moderate problem. In pregnant women the magnitude of the problem is reflected in the proportion of children born with low birth weight (below 2.5 kg). Available information on the prevalence of LBW in Tanzania suggests that 16 percent of babies had low birth weight (LBW) (TDHS, 1991/92). A survey conducted by TFNC in seven districts of Tanzania showed that prevalence of low birth weight ranged from 7.8 to 21 percent (TFNC 2004). The survey showed also that 6.2 percent of lactating women were undernourished. LBW babies carry significant mortality and morbidity risks that start in the immediate period after birth and continue into adulthood. It is estimated that LBW babies have a ten-fold perinatal mortality and a four-fold post neo-natal mortality risk (Barker et al 1986, 1989). In deed, one possible explanation of the high neonatal mortality in Tanzania has been explained as the high prevalence of LBW (TFNC 2002). 3 Another group that is vulnerable to PEM is older people. A survey conducted in Magu district in 2001 showed that 18.3 percent of older people (aged 60 years or above) were undernourished (HelpAge 2001). 2.2 Nutritional Anaemia Occurrence of nutritional anaemia cuts across all population groups. But women and children are the most affected (Table 1). Pregnant women are particularly vulnerable to nutritional anaemia due to increased requirement of iron and folic acid. Adolescents are also at high risk of developing anaemia due to higher demand of nutrients for growth. Currently there are no data that show national picture of anaemia. Data from limited surveys show high magnitude of the problem. The 1993 survey conducted in 10 districts showed that prevalence of nutritional anaemia in pregnant women ranged from 13.9 to 79.9 percent. Prevalence among under-fives ranged from 27.2 to 94.8 percent (URT 1994). According to another survey conducted in Korogwe and Handeni districts under World Vision (T) in 1997, the prevalence in non-pregnant women, no-pregnant women and under-fives was 84.2, 73.7 and 86.8 percent, respectively (WVT 1998). 2.3 Iodine deficiency disorders IDD exhibit a geographical pattern. The prevalence of IDD is highest in highland and mountainous areas because of leaching of iodine from the soil to the sea and other lowland areas. In Tanzania the most affected regions are Iringa, Ruvuma, Mbeya, Rukwa, Kigoma, Kagera, Arusha, Kilimanjaro, Tanga, Lindi, Morogoro and Dodoma. First survey conducted in Tanzania in the 1980s showed national goitre prevalence of 25 percent (Kavishe 1993). Interventions were then initiated including distribution of iodinated oil capsules and iodation of salt. The current national prevalence of goitre is estimated at 8 percent and about 83 percent of households are consuming iodated salt (TFNC 2004). 2.4 Vitamin A deficiency In Tanzania VAD is more severe in drought prone areas due to seasonal availability of foods rich in vitamin A. However it has also been observed in other non-drought areas, probably due to factors such as food preferences, food preparations and diseases. VAD mainly affects children between six months and six years of age. The national survey on VAD conducted in 1997 showed that 24.2 percent of children aged below five years were vitamin A deficient (had serum retinol levels 20 µg/dl). The same survey showed that 69 percent of lactating women had levels of breast milk retinol below 30 µg/dl (Ballart et.al. 1997). According to WHO, the problem is of public health significance (WHO 1993). Of the four nutrition problems in Tanzania PEM is ranked first. PEM is a reflection of reduction in overall food intake as well as micronutrients such as iron and Vitamin A. Thus PEM can be used as a pointer to general malnutrition. Furthermore the consequences of PEM ­ to the individual, community and even to the nation ­ are worth noting. It has been demonstrated, for example, that long term consequences of PEM, especially stunting, can reduce the height of an adolescent by 4.6 cm; schooling by 0.7 grades and loss of 7-12 percent of lifetime earnings (Alderman et al, 2003). 4 While PEM remains a priority nutritional problem in Tanzania micronutrient deficiencies especially iodine, vitamin A and iron continue to be of great public health importance as well. It has been shown that by improving the vitamin A status of children 6 to 59 months old up to 70% lives are saved (WB Report November 2001). Control of iodine deficiency disorders has a direct impact on reduction of low birth weight by 40% and infant and young child mortality by 30% (Behrman February 2004); while reducing iron deficiency has a direct effect on reduced low birth weight, improved schooling outcomes and increased productivity in adult workers. (Behrman February 2004). 3.0 Major determinants of malnutrition In this report the identification of determinants of malnutrition is based on the conceptual framework, which is a widely used analytical tool, portraying causal factors and their interaction at three main levels: immediate, underlying and basic. 3.1 Immediate causes of malnutrition are those most proximal factors, which finally precipitate malnutrition. These factors are low food intake and diseases that operate at individual level and are more visible. · Low food intake is characterized by low feeding frequency and consumption of low energy dense foods. Information on food consumption pattern has indicated that carbohydrates provide 81 percent of energy intake, protein 11 and fat 0 - 8 percent (Lukmanji, 1982). However, overall intake is low. It has been estimated, for example, that average energy intake for Tanzanians is only 72 percent of the requirement (Annergers, 1973). Breastfeeding situation in Tanzania is good with 90 percent of women in urban and 98 percent in rural areas initiating breast-feeding. However, only 48 percent of children continue to breastfeed to 23 months or beyond. Exclusive breastfeeding for 4 ­ 6 months is 11 percent. The low rate of exclusive breastfeeding has been attributed to social, economic and cultural factors (TRCHS 1999). Complementary foods given to infants and young children are staple based porridges which are bulky and of low density. Feeding frequency is only two to three times a day (URT 1985). Consumption of fruits and vegetables, which are rich in vitamin A and iron, is generally low particularly in semi-arid and drought prone areas. Various factors may be contributing to this situation, including inadequate knowledge on their value and low supply. In Korogwe and Handeni districts, for example, only 15 percent of households had a garden of fruits or vegetables. Furthermore, in the majority (65 percent) of those households with garden, fruits were both for sale and consumption. It is also known that consumption of foods of animal origin, which are very good source of vitamins and minerals is also low (World Vision (T), 1998). · Diseases are another group of immediate causes of malnutrition. Diseases, notably malaria and measles, cause loss of appetite. Increased rate of metabolism associated with high fever increase the body's need for food. Diarrhoea, which is most often 5 accompanied by vomiting, decreases food intake and results in loss of nutrients. This worsens the nutritional state of an individual. Likewise the impact of disease is made worse by the malnourished state, since malnutrition lowers the body's ability to resist diseases. More than 18 million cases of malaria are reported in all health facilities annually. Malaria causes haemolysis and anaemia. In a survey conducted in Lindi district in 1992 presence of malaria parasites in the blood was found to correlate positively with anaemia in children and adults (Mduma et al, 1992). Malaria also causes loss of appetite and even diarrhoea particularly in children, which may lead to reduced food intake and PEM. TRCHS (1999) show the prevalence of diarrhoea in children aged below five years to be 12.4 percent. As with ARI the data show that 14 percent of children had cough and rapid breathing in the two weeks before the survey. HIV/AIDS is an important cause of malnutrition as it leads to reduced food intake mal- absorption and metabolic changes that increase nutrient requirements. It is estimated that by the year 2002 there was a cumulative total of 785,865 AIDS cases since the beginning of the epidemic in Tanzania in 1983 (NACP 2002). Intestinal parasites are associated with reduction of food intake, mal-absorption and nutrient loss. Hookworms and schistosomes contribute to anaemia through blood loss. In the Lindi study intestinal parasites had positive correlation with anaemia in children and adults (Mduma et al, 1992). 3.2 Underlying causes of malnutrition are numerous and often interrelated in any given specific context. They reflect unequal distribution of resources and services. It is the level of implementation for most interventions aimed at reducing malnutrition. This level consists of a complicity of causes, although in a given specific context it would be possible to identify which causes and relationships are most pronounced, and once that could be tackled with available resources. In Tanzania the underlying causes are inadequate household food security, inadequate caring capacity of the vulnerable groups and inadequacies in quality and quantity of the provision of basic services including health, education, shelter, water and sanitation. · Household food insecurity: Food insecurity is a situation that exists when people lack secure access to sufficient amounts of safe, nutritious and socially accepted food for normal growth and development and an active and healthy life. It may be caused by the unavailability of food, low purchasing power, inappropriate distribution of food at household level. Food security may be chronic, seasonal or transitory. It is estimated that seven million people in the country are chronically food insecure. In addition, about 40 percent of the total population lives in drought or flood prone areas and hence face transitory food insecurity (Kavishe, 1993). Groups vulnerable to food insecurity in the country include female-headed households and rural households earning income below the absolute poverty line. Others are rural 6 minimum wage earners and urban low-income workers, mostly engaged in informal sector activities. Domestic agricultural production is the main source of food supply in the country particularly in the rural areas with maize being a staple food crop in most parts. Plantains, pulses, rice, cassava, sorghum, sweet potatoes, millet, vegetable and fruits constitute a significant portion of household food supply as well as meat, milk, eggs and fish. Food accessibility and consumption in the country is still inadequate in most of the households mainly due to low food production, poor harvest and post harvest handling practices. Household food security is also associated with levels of income. Low incomes of the people deny them access to food (HBS 2001). Food distribution among family members within the households is said to be unequal because of bad cultural practices that denies vulnerable members of the household to access enough and nutritious foods. · Basic services are those that fulfill requirements that are important for an improved livelihood. The availability and accessibility of these services contribute to nutritional status of the people. These services include health, education, water, environmental sanitation, shelter and clothing. Health services are provided through public and private systems and traditional birth attendants and healers. The public health services in the country have developed rapidly since independence and about 70 percent of the population live within 5 kilometers from the nearest dispensary. This has increased coverage of reproductive and child health services (RCHS). About 85 percent of pregnant women attend antenatal clinics. Family planning acceptance in Tanzania is estimated at 44.1 percent in urban and 31.6 percent in rural areas. As regards immunization coverage, only 68 percent of children are vaccinated against the six major childhood diseases - well below the intended goal of 80 percent (TRCHS, 1999). In general, the existing health services have several constraints that include inadequate community participation, lack of constant supply of essential drugs and equipment, trained personnel and deteriorating infrastructure. Some of the health programs such as RCHS and family planning do not actively involve men. In addition, women who are the target group of these programs have many burdening and pressing activities to attend to. Tanzania has given priority to the provision of basic education - both universal primary and adult education. Enrolment in primary schools, following the implementation of the Primary Education Development Program (PEDP) in 2003, was as high as 98.6 percent. However, females were slightly disadvantaged: the ratio of girls to boys was 47:53 in public schools and 48:52 in private schools (URT, 2003). Epidemic diseases such as cholera and typhoid are common in Tanzania and are related to inadequate safe water supply and poor environmental sanitation. 7 The most commonly occurring health problems for women and children in Tanzania, malaria, diarrhoeal diseases and respiratory infections, are closely associated with problems of water, hygiene and sanitation. Malaria mosquitoes breed in water pools and in uncleared ground; diarrhoeal diseases are associated with contaminated water and food, unhygienic practices and unsanitary conditions. Supply of clean and safe water covers 53 percent of the population in rural areas and 73 percent in urban areas (URT, 2003). However, in both areas, problems of inadequate maintenance and rehabilitation of water supply system exist leading to disruptions in supply. The importance of sanitation has been emphasized since colonial times but sewerage system in urban centres is available to only 10 - 15 percent of the population. On the other hand, over 92 percent and 84 percent of urban and rural households, respectively, are reported to have latrines; although actual utilization of these latrines is at lower levels - only 30 percent in rural areas (UNICEF 2001). · Inadequate caring capacity of the vulnerable groups: Care is the provision by the household, community members and other agencies of resources in the form of time, attention, love and skills to meet the physical, mental, physiological and social needs to all people particularly the most vulnerable groups in the society. Such needs include adequate food, health care, education, water and environmental sanitation, decent shelter and clothing. However, availability, accessibility and affordability of these needs have always been marginalized to some groups in the society and hence affecting their nutritional status. These groups include; children, pregnant and lactating women, older people and the sick. (URT 2003) There are a number of factors that lead to inadequate care of the vulnerable groups, which include inadequate protection, promotion and support of breastfeeding and complementary feeding practices. Heavy women's workload and inadequate establishment of community based child day care support systems contribute to the poor nutrition of children. Bad habits and customs amongst families and communities that do not support proper care of children make them run away from their homes and live in the streets especially in the urban areas. These children are affected nutritionally, mentally, socially and emotionally due to lack of family support. Appropriate care for pregnant and lactating women include provision of antenatal and postnatal services, adequate feeding and being exempted from various household and communal tasks. Childcare includes proper feeding, protection and provision of health services. In most community settings it's the mother who stays closest to the child and takes most responsibilities for care of the child. One of the major factors contributing to inadequate childcare and therefore child malnutrition is heavy women's workload. Although men and women may work together in the field, most agricultural activities are considered women's responsibility. Due to use of traditional methods of farming a lot of hours are spent on agricultural activities. For example, a study carried out in Tanzania showed that a total of 500 hours were spent on one hectare of maize crop ­ from ploughing through planting, harvesting and shelling (TFNC, 1983). 8 Other activities consuming women's time include collection of firewood and fetching water. In Korogwe and Handeni districts, for example, most people spent up to four hours a day collecting firewood and fetching water; and in 85 percent of the households it was the women who performed these tasks (World Vision (T), 1998). A study carried out in Arusha municipality in 1984 showed that nutritional status of children improved with increase in the time spent by mothers on child care (Kisanga et al 1984). Women's heavy workload also has direct effect on the individual in that it highly increases her energy expenditure. Under conditions of inadequate food supply it leads to under-nutrition of the woman. 3.3 Basic causes: Basic causes of malnutrition are those that relate to determinants of resource allocation for addressing the underlying causes. The basic causes include the development and structure of the economy, prevailing political ideology, social and cultural factors related to gender. Economic, cultural and political changes take varying periods of time to reflect on nutrition situation. A country with strong socio-economic conditions is likely to allocate substantial infrastructure, human, financial and organizational resources to nutrition programs than poor countries. Tanzania is a poor country with per capita income of US $ 280 per year (World Bank, 2004). In Tanzania, 19 percent of the population cannot afford their basic food requirement due to poverty. Similarly 36 percent of the population earns incomes that cannot meet the basic needs of food, shelter, clothes and primary education (TSED, 2002). Agricultural production depends on the environment under which the crops are grown. Drought and floods affects food production leading to food insecurity. Destruction of the environment also causes water deficit, which affects sanitation and subsequently ill health. Prevailing political inclination of a country, on the other hand, determines the policies that influence the underlying causes of malnutrition such as policies on food, agriculture, health, income and commodity prices. In Tanzania the Arusha declaration of 1967 emphasized on human development, and therefore enabled many low-income people to access social services such as health, water and education. Basic causes of malnutrition include socio-cultural factors as well. Property relations, for example, favour men over women. Denying women right of ownership of land as it is in many communities in Tanzania today not only leads to food insecurity on her part but to low income and hence inaccessibility to health, education and other social services. 9 4.0 Malnutrition control programs in Tanzania During the 30 years or so of nutrition work in Tanzania there have been various efforts to develop and implement programs whose goals have been to reduce the high rates of malnutrition and mortality. On the basis of goals and general thrust, the programs can be grouped as those aimed at the control of PEM and general malnutrition; and those for the control of micronutrient deficiencies. 4.1 Programs addressing PEM (i) The Infant and Young Child Nutrition Program The program which has been in operation since the 1990's aims at preventing malnutrition among infants and young children through improvement in young child feeding practices. The program focuses on breastfeeding and complementary feeding. Improved breastfeeding practices; This has been implemented through the BFHI under which training of health workers and other service providers is conducted. The subject has also been incorporated into the training curricula for nurses and traditional birth attendants. A national regulation on marketing of breast milk substitutes was passed in 1994. The program is managed through a multi-sectoral National Consultative Group on Infant and Young Child Nutrition (NCGIYCN). The consultative group draws participants from government ministries, non-governmental organizations, academic institutions and private sector. Activities of the program are implemented through the existing structures up to community level. The coverage is national, focusing on health facilities and eventually community. Presently there are 59 hospitals that have been declared baby friendly out of 128 trained. Exclusive breastfeeding is 37.9% at 4 months and 11% at 6 months. The low prevalence of exclusive breastfeeding is attributed to inability of the program to reach the community level. Efforts are underway to adopt the Global Strategy on Infant and Young Child Feeding (GSIYCF) including development of a national implementation plan. (ii) Nutrition rehabilitation has addressed the problem of PEM in various ways with varying intensity in Tanzania since the 1960's. Nutrition rehabilitation units (NURU), which started in the 1960's, were health facility based units. These units were providing convalescent care at low cost to children treated in hospitals for severe and moderate PEM. An evaluation conducted in 1978/9 found that NURUs were failing to meet their objectives, in that children were returning to the NURUs soon after discharge having fallen back to PEM condition. The main constraints were that the NURUs were operating in conditions alien to the home environment and thus the rehabilitation processes could not be continued at home. For this reason a system for managing severely malnourished children in the community was developed. 10 Community Based Nutrition Rehabilitation (CBNR) was established in 1980s to manage at community level children who develop severe PEM. It stresses that a malnourished child be rehabilitated in the same environment that precipitated the condition using resources available at home and in the community. In Iringa nutrition program (INP) in which CBNR was first practiced severe underweight dropped from 6.3 percent in 1984 to 1.3 percent in 1988. Similar success was also observed in Hai and Serengeti districts. The project was however confined in districts implementing CSDP. Due to its positive results, TFNC is revitalizing the process. Plans are in place to support councils to initiate the intervention through government financing. (iii) The Iringa Nutrition Program The Iringa Nutrition Program (INP) was an integrated, community-based program whose ultimate objective was to reduce infant and young child malnutrition, morbidity, and mortality. The Program was initiated in 1982 with funds from the Italian Government through the joint WHO/UNICEF Nutrition Support Program (JNSP). Iringa, one of Tanzania's 20 regions at that time, was chosen as the site for the project for a number of reasons: first it was the only region where comprehensive nutrition surveys had been done and the prevalence of malnutrition had been found to be very high. Secondly the region possessed diverse agro-ecological zones to enable the region to develop a broad base of experiences in different settings to facilitate replicability in other areas. Thirdly, the region possessed a relatively strong institutional infrastructure, considered important to give the project a fair chance of success (Kavishe, 1993; Pelletier 1991). The main objective of the INP was improvement of nutrition. The specific objectives were: · Reduction of infant and young child mortality and morbidity, · Better child growth and development, and · Improvement of maternal nutrition. In the conceptualization of the INP, another objective was added: `improvement of the capabilities at all levels of society to assess and analyze nutrition problems and to design appropriate actions'. This was seen as an essential means by which all other objectives should be met. The INP took advantage of the existing administrative structure in Tanzania, ­ regional, district, division, ward and village levels - to the extent possible. The INP was integrated into this structure. Various implementation committees at region, district, ward and village were established expressly for the INP to provide a mechanism to redirect resources as required for nutrition improvement. Though the INP was area based the National Steering Committee convened quarterly meetings to assure that the project evolved in a manner consistent with national policies and programs. These meetings were occasionally held in villages in the project area so that participants could interact with villagers and get first hand opportunity to see the project. 11 A key element of the program approach was provision of information about children's nutritional status to critical decision makers at all levels through a system of community growth monitoring system by quarterly weighing. Child growth cards and village registers permitted follow up of individual children who were severely malnourished in a community based nutrition rehabilitation approach that had been developed by TFNC. An important feature of the MIS was its use at all levels for decision-making [Pelletier, 1991]. It permitted Village Committees to identify and target actions to at risk households. Likewise district and regional levels used the information generated to target extension services in villages with poor nutrition. Although the information system in the INP was designated primarily as a management and motivational tool to catalyze the triple ­A-cycle at household, village, and higher levels it was also used to assist in the evaluation of the impact of the program on nutritional status. Since its inception the JNSP in Iringa continuously grew and changed, and was modified in response to the application of the Triple-A-Cycle. Lessons learnt regarding the process of the intervention and infrastructure development were transferred to other geographical areas as Child Survival and Development (CSD) and, later, Child Survival Protection and Development (CSPD) retain ref. programs. There is evidence, which indicates that the Iringa "model" of the process approach has been sustainable in the local contexts of the CSPD areas as indicated by the sustainable reductions in the rates of severe malnutrition. In 2002 CSPD was transformed into Early Childhood Care and Survival Growth and development (ECC-SGD). The essential elements, which were responsible for the replication of JNSP, were the following [URT/WHO/UNICEF, 1988; Pelletier, UNICEF, 1989]: The use of the explicit integrated conceptual framework, which helped to seek solutions in a multi-disciplinary way within the comprehensive approach provided by the framework. An important strength of the framework is its lack of clearly defined boundaries, which leave room to develop different causal models of the problem of malnutrition in different circumstances. Thus though the framework is not predictive it allows its application in a variety of situations. The Triple-A-Cycle approach of Assessment ­ Analysis ­ Action led to the improvement of the capabilities at many levels to assess and analyze nutrition problems and to design appropriate actions. It also led to a fundamental change in process as well as a development of support systems for advocacy, training and monitoring. This is important for empowerment and sustainability. Social mobilization led to a high degree of active popular involvement in the program with a consequent of resources by communities for the improvement of nutrition within households. It is likely that much of the additional time and care of parents for their children was provided by women but with an increase in the participation of men. Social mobilization elevated malnutrition from the level of an individual's problem to that of a community social concern. The permissive social and political context in Tanzania was an important condition for program sustainability and replicability. Political commitment is essential for the success of nutrition intervention programs. As often stated by the regional and district officials the 12 program's emphasis on a process approach of social mobilization and the resulting inter- sectoral action, made the effects of the program go beyond the scope of "nutrition" programs per se, to contribute to the effectiveness of all other development activities which were aimed at improving the people's well-being. (iv) The Child Survival and Development (CSD) Program The Child Survival and Development (CSD) Program was initiated in 1987 through the application of the basic elements of the approach developed in Iringa JNSP. The Iringa JNSP had demonstrated that it was possible to improve the survival of children as well as promoting their development through implementation of interventions that lead to reduced severe and moderate malnutrition. CSD program began with an extension of the Iringa JNSP beyond the borders of the original project area to cover all districts in the region and later extending similar interventions in six other regions. The emphasized on community participation which involved village mobilization, the establishment of village health committees, training of village health workers and the initiation of village health days which were responsible for increased immunization rates and community based growth monitoring. The program was however not a total replica of the JNSP especially in terms of financial input. It is acknowledged that JNSP was a capital-intensive project with an average cost per child per year being approximately US$ 19 (URT/WHO/UNICEF 1988). This was by far on the higher side, which could not be maintained and sustained in other areas. CSD programme interventions and approach emphasized on minimization of financial inputs through increased community participation and contribution in terms of labour and materials. It also emphasized on resource mobilization by the districts themselves. The implementation of CSD program demonstrated that similar outcomes could be realized with a much more reduced capital investment and more importantly with more locally mobilized resources for continuity and sustainability. CSD program became increasingly famous and by 1991, the program had been extended to nine regions. The endorsement of the World Declaration on Survival, Protection and Development of children in 1990 led to the inclusion of the "protection" component in CSD program. The protection component brought in the CSD another dimension of intervention for survival and development of child. CSPD began programming for the rights of the child. The rights programming focuses on implementation of activities that promote care and psychosocial stimulation of the young child, promotion of care seeking practices by duty bearers (parents, service providers, community organizations NGO's, Government and international organizations) and provision of quality services (health education). Additional intervention within the rights approach framework is the prevention of mother to child transmission of HIV/AIDS. It is important to note that CSPD has retained most of the CSD program intervention activities with some changes in the approach and focus. Currently CSPD program is being implemented in 57 districts in 13 regions of mainland Tanzania. The focus however, has changed with more emphasis being put on programming for Early Childhood Development (ECD). ECD has three major components which include improved birth outcome, care of the sick, vulnerable child and women and lastly the prevention 13 and promotion of measures for survival growth and development. ECD basically addresses both nutrition and health intervention strategies. (v) Prevention of Low Birth Weight Major efforts for the prevention of low birth weight (LBW) in Tanzania are seen in two undertakings: projects for the prevention of (LBW) in two refugee camps in Tanzania and another conducted in seven districts of Tanzania. Prevention of LBW in Lukole Refugee Camp (Ballart et al, 2000) In 1999 a high prevalence of LBW was experienced in Lukole refugee camp in Ngara District, Western Tanzania. The problem was perceived to be associated with high rates of malaria and anaemia that prevailed among pregnant women. This prompted the initiation of a project for the prevention of LBW comprising of multiple intervention strategies targeted to pregnant women. The interventions were malaria vector control, malaria prophylaxis and treatment, de- worming, provision of food supplements, iron and folic acid supplementation and treatment of sexually transmitted infections particularly syphilis. Evaluation of the project showed positive impact. Malaria prevalence, for example dropped from 35 percent in 1998 to 8.9 percent in 2000. Similarly in the same period prevalence of LBW dropped from 35 percent to 7.1 percent. The success of the project was a result of efforts and strong participation of various institutions operating in the camp ­ UN agencies and NGOs who provided financial and technical support. In particular an NGO "Norwegian Peoples Aid" implementing a well-organized reproductive health program helped in smooth implementation of the various interventions. The project also enjoyed the services of the existing community structures and resources in the camp such as well trained "health information teams", traditional birth attendants and health care providers working in MCH and out patient departments. Also there had been an ongoing community sensitization in the camp on prevention of communicable diseases. The project was donor dependent hence bringing into question its sustainability and scaling up. With regard to replication it would have been possible to expand it to other refugees camps as camps have similar administrative and health delivery structures. But it would have been difficult to replicate it in non-refugee settings of Tanzania. Prevention LBW in Mtendeli Refugee Camp (Reed B, 2002) A similar project was implemented in Mtendeli Refugee Camp located in Kibondo district in 1998 where almost the same type of antenatal care interventions was carried out. Provision of the health and nutrition services was done by various organizations: UNICEF, UNHCR, WFP, IRC. Over a period of four years (1998 ­ 2001) mean birth weight increased from 2.65 to 3.01 kg; and prevalence of LBW dropped from 28 to 8 percent. Prevention of LBW in seven districts of Tanzania (TFNC 2002) In the year 2000, TFNC, in collaboration with UNICEF, designed an integrated project for the prevention of LBW. The project is being implemented in seven districts (Hai, Kibaha, Kilosa, Magu, Mbarali, Masasi and Mtwara rural) with support from UNICEF. Prior to the start of the intervention in 2001, a baseline survey on health and nutrition among lactating women (4 ­ 6 weeks postpartum) and newborns was conducted. 14 The findings show that the prevalence of LBW varied from 7.9 percent in Hai district to 21.1 in Masasi. Some risk factors normally associated with LBW were confirmed by this survey. These include maternal undernutrition, anaemia, malaria and parasitic infestation. The survey confirmed also the vulnerability of adolescent mothers with regard to malnutrition. Young mothers were 1.5 times more anaemic; 2 ­ 3 times more undernourished and had 3 ­ 5 times more chances of giving birth to LBW babies compared to those aged 25 to 34 years. These findings call for a new thinking ­ to direct efforts to programs addressing adolescent nutrition. Interventions for the prevention of LBW in those seven districts were the same as those for the prevention of LBW in the two refugee camps. In addition the project carried out activities related to capacity development of key actors at district and health facility levels as well as awareness creation at community level. The impact of the project cannot be described at this time, as evaluation is yet to take place. (vi) Household Food Security Until 1970s, Tanzania was largely self sufficient in food. This situation changed following economic crisis in the late 1970s and prolonged periods of drought that resulted in food shortage throughout the country (Dolan and Levinson, 2000). This situation led to the development of Food Security Policy by the Ministry of Agriculture and Food Security (MoAFS), and subsequently to program on household food security. The program covers the whole country with activities being implemented by various sectors. It is managed by the Household Food Security Consultative Group (HFSCG), which is chaired by the MoAFS. TFNC is the secretariat. Household food security program has the following components: Household food storage; Food storage work comprised of packaging and post-harvest technologies development and improvement of traditional storage structures and capacity building. This work started in 1975 under TFNC management and covered all regions. A number of outputs were realized which included; improvement of traditional storage structures, training of extension workers from all sectors in all regions and dissemination of good storage practices. Storage of cereal grains at household level has remained poor despite the training conducted. Studies carried out in Mbozi to assess post harvest losses of maize in 1994 showed that households could store maize for 3 months only (Ndunguru et al., 1991). Storage losses occurred mainly due to due to damage by rodent and weevil infestation. Household food security card (The Bag Model): This is a tool that was developed and promoted for assessing and monitoring food stocks (cereal grains) in the households in areas where cereals are grown. The household food security card has so far been promoted in Morogoro, Iringa, Mbeya, Rukwa and Ruvuma regions. Although the household food security card is a useful tool for ensuring food availability for each member of the household (3 bags of cereal and half bag of legumes for adult person) throughout the year, the card cannot be used to assess other staples such as cassava, potatoes, sweet potatoes and bananas. 15 Promotion of root and tuber crops: The program started in 1994 with the objective of promoting cassava and sweet potatoes for improving household food security and income generation. Cassava and sweet potatoes grow well in areas where cereals such as maize cannot do well. These crops can also be stored in the ground and provide food to the households throughout the year. Unfortunately these crops are highly perishable with high post harvest losses. Under this component, cassava growers in Rufiji district were trained on how to preserve fresh cassava roots. Also more than 10 groups of entrepreneurs from Dar es Salaam and Rufiji district were trained on how to process high quality cassava flour for Dar es Salaam market and home consumption. As for fresh sweet potatoes, farmers in the Eastern and Lake Zones of Tanzania have been trained on how to store fresh roots for at least 3 months after harvest. The aim is to make the crop available throughout the year. Improved processing and technology of the crops are not available for the households leading to high post harvest losses. Despite these efforts, adverse weather conditions have resulted in severe food shortage in the drought prone regions of the country. Also cereal crops are used for sale or social functions thus aggravating the problem of food shortage. (vii) Nutrition in Primary Schools In Tanzania the program for the improvement of nutrition in primary schools started in 1921 in Dar es Salaam. Schools were served with midday meal. Later the program was extended to Tanga, Lindi, Mtwara, Arusha, Iringa, Mbeya and Tabora. The program was supported by organizations like Oxfam, Catholic Relief Services and also the United States of America. During 1950s and even a few years after independence, school meals contained red palm oil for the prevention of VAD. By 1965 the whole country was almost covered and health and nutritional status of primary school children had greatly improved. The program went hand in hand with nutrition and health surveillance involving weight and height recording, screening for infections such as malaria and worm infestation and regular parades for personal hygiene and environmental sanitation. It was thus reported that many primary school children were well nourished and school performance was high (TFNC/Action Group; 1995). Following the Arusha declaration of 1967 many activities were decentralized to regional, district and village levels. This led to the withdrawal of Central Government from participation in nutrition and health in primary schools. As a result the program deteriorated in many schools as parents could not solely contribute to the school meals. By 1972 the program had been substantially cut down or stopped totally in most schools. In 1978 the MoH reestablished primary health services. The Ministry of Education and Culture (MOEC), TFNC and USAID, in 1985, supported the efforts by initiating a pilot nutrition and health program in Singida and Dodoma regions. Nutrition and health of school children was reported to have improved under this program. However, the program did not last long owing to weak monitoring mechanism. Since then there have not been other major efforts that address nutrition in primary schools. 16 (ix) Health and Nutrition Project (HNP) In early 1990s, the World Bank responded favourably to the Government's Economic and Social Action Program (ESAP) and Priority Social Action Program (PSAP), which aimed at revamping the ailing economy. Among the sectors supported was health through Health and Nutrition Project (HNP). The objective of the HNP was to reinforce efforts of the ESAP and PSAP to raise quality of nutrition and basic health services in urban and rural areas and was to be implemented for five years, 1990/91 to 1994/95. HNP comprised of three main components: · Strengthening Central "population, health and nutrition" systems. · Trial implementation of the primary health care strategy in urban areas. · Strengthening urban primary health care systems. Component I (above) comprised of four sub-components one of them being control of micronutrient (iron and vitamin A) deficiencies. A total of US$ 1.56 million was allocated to this program. Component II of the HNP included funds for 10 districts to support their district health plans on assumption that, with technical assistance from TFNC, nutrition would have been included in these interventions. The project also financed a baseline health and nutrition survey conducted in all 10 districts by TFNC. Coverage of the micronutrient deficiency control program was national. However, special emphasis was required on the 10 districts that were implementing the other HNP components. Activities implemented under the micronutrient deficiency control program included creation of public awareness through media and reader materials; and training of health care providers on the control of micronutrient deficiencies. The TFNC laboratory was strengthened with regard to analytical capacity of micronutrients and 25 district hospital laboratories were provided with machines for measuring haemoglobin level. A pilot project for promotion of horticultural foods (foods rich in iron and vitamin A) was initiated in Ilongero Division, Singida rural district under which production of fruits and vegetables was promoted at household level and in primary schools. Improved vegetable preservation techniques (use of "solar driers") was introduced and advocated in the project area. A red palm oil processing plant was procured and handed over to a cooperative union in Kyela district. The aim was for the Union to enhance production of red palm oil, then to eventually pay back the cost of the plant for creation of a revolving fund. With such a fund more similar plants were to be procured and provided to other business entities to enhance production of red palm oil in Tanzania. HNP was concluded in 1999. Although final evaluation was not conducted a mid-term review found evidence that knowledge of micronutrient deficiencies had increased among health staff. Also vitamin A and iron/folate supplements were more widely available at health facilities (World Bank, 1994). 17 4.2 Programs for Control of Micronutrient Malnutrition Tanzania has been implementing programs for the control of three micronutrient deficiencies namely anaemia, iodine deficiency disorders and vitamin A deficiency. (i) Anaemia Control Program The program addresses the major causes of anaemia such as deficiencies of iron and folic acid and diseases that relate to anaemia (malaria and helminth infestation). A consultative group "National Anaemia Consultative Group" is the policy making and steering body for the program. The Ministry of Health chairs the group, which meets twice a year, while TFNC is the Secretariat. Membership is institutional ­ all relevant institutions and ministries are represented in the group. Activities implemented under the program include supplementing pregnant women with iron and folic acid which has run since the inception of MCH services in 1974, promotion of production and consumption of iron/vitamin rich foods and public health measures for the control of malaria and helminth infestation. Health care providers and extension workers were given training on the control of anaemia. The training was conducted district-wise, involving all health facilities (hospitals, health centers and dispensaries) in the district. In total 24 districts were trained. All district and district­level hospitals were provided with haemoglobin measuring machines (haemoglobinometers). This involved also training of laboratory technicians from the respective hospitals on the use of the haemoglobinometers for the generation of data to describe the magnitude of the problem of anaemia in their catchment areas. Reader materials on anaemia were developed and distributed to health facilities throughout the country. Also a number of radio and television programs were aired for creation of public awareness on the problem. As is for all other nutrition programs the implementation of the anaemia control program is based on the existing infrastructure ­ the national and sub national administrative structures as well as the health structure. For example, the Ministry of Health trough the EDP supplies iron and folic acid tablets earmarked for pregnant women. The MSD transports them to the district (DMO's office). The DMO's office distributes the tablets to all government primary and secondary level health facilities within the district. By incorporating the iron and folic acid supplements in the health delivery system sustainability is facilitated. On the other hand the Ministry of Agriculture and Food Security, which also has a nutrition unit in its structure, promote production of foods rich in iron and vitamins. Prevention and control of malaria and helminth infestation is implemented through the Malaria Control Program and the Parasite Control Program in the Ministry of Health. Shortcomings/constraints encountered include the fact that national coverage is not achieved. Iron/folic acid tablets and drugs for treatment of malaria and helminthes are supplied to government dispensaries and health centers only as the EDP does not cover hospitals. The populations in hospital catchments, therefore, do not benefit. Similarly the population covered by private and faith- based health facilities does not access the drugs. 18 The Malaria Control Program promotes use of insecticide treated nets (ITNs) in malaria prevention but coverage is still low. The 1999 Tanzania Reproductive and Child Survey (NBS, 1999) showed that in urban areas, households using nets made up 48 percent and in rural areas 13 percent. Furthermore those using insecticide treated nets made 10 percent only. (ii) IDD Control Program The National IDD Control Program in Tanzania comprises of two approaches: distribution of iodinated oil capsules targeted to 27 highly endemic districts and universal salt iodation. While the former is a short-term stopgap measure, the latter is the long-term approach for the control of IDD in Tanzania. The program operates under the guidance of a multisectoral consultative group "National Council for Control of Iodine Deficiency Disorders ­ NCCIDD". This is the steering body of the program. The council has Ministry of Health as the chair and TFNC as the Secretariat. Other member institutions are Planning Commission, Tanzania Food and Drugs Authority (TFDA), Tanzania Bureau of Standards (TBS), Muhimbili University College of Health Sciences (MUCHS), Muhimbili National Hospital (MNH) and Tanzania Salt Producers Association (TSPA). UNICEF and WHO attend the NCCIDD meetings as development partners. Iodinated oil capsules Distribution of iodinated oil capsules started in 1986. This was made possible through financial assistance from Sida. However when it was found that the capsules (lipiodol) were very expensive it was decided to limit distribution to districts that were highly endemic. A total of 27 districts had been identified. The capsules were targeted to all people aged 1-45 years in these districts. They received two capsules (380mg elemental iodine)) as single dose every two years. It was envisaged to phase out capsules when salt iodation program was set up and had been strengthened to be able to adequately avail the people of Tanzania with iodine. The donor, through TFNC, procured the capsules and TFNC distributed them to the districts. At the district level the District Medical Officers (DMOs) arranged for distribution through various avenues such as health facilities, village governments and schools depending on what was most convenient. The distribution exercise, however, was not without shortcomings. For example financial and transport constraints sometimes led to failure of the distribution teams to reach the very remote villages. Also proportion of people turning up for the capsules were not always as high as expected owing to inadequate community sensitization. Salt iodation Universal iodation of edible salt is the long-term strategy for the control of IDD. The idea was formally adopted in 1988 when the Netherlands Government pledged to fund such a program in Tanzania. From 1991 the Netherlands Government, through UNICEF, started providing financial support for salt iodation infrastructure. Main inputs included provision of salt iodation plants, potassium iodate, packaging materials and vehicles for supervision and monitoring. Beneficiaries of the above supplies were the large and medium scale salt producers (it was decided to deal with small scale salt producers later). Reader materials on salt iodation were produced and distributed widely. Sensitization and training were conducted 19 to actors at various levels of program management: program managers, salt inspectors, salt quality control technicians, salt producers and traders and political/administrative leaders. An association "Tanzania Salt Producers Association" was formed in 1993. This has been the body that links salt producers with other stakeholders in the salt industry including the government machinery. The Tanzanian parliament endorsed universal salt iodation and enacted "Salt Acts" which prohibited trading on non-iodated salt earmarked for human consumption. The law became effective on 1st January 1995. Achievements of salt iodation program were noted as early as 1995 when spot surveys revealed significant availability of iodated salt at household level and reduction in goitre prevalence (TFNC, 2000). In 1999/2000 TFNC carried out evaluation of the IDD control program, the findings of which show that in the 27 endemic districts goitre prevalence decreased from 67.6 percent in 1980s to 23.5 percent in 1999/2000; and 83.3 percent of households were consuming iodated salt (TFNC, 2000). More recently, a national IDD survey carried out by TFNC in 2003 indicated that goitre prevalence was 8.1 percent (down from 25 percent in 1980s); and that 83.8 percent of households were consuming iodated salt. Median iodine excretion level was 203µg/l, well above the WHO cut-off point of 100µg/l (TFNC, 2004). Salt iodation in Tanzania needs scaling up. As stated above only large and medium scale producers were addressed in the initial support for salt iodation. Inventory of small-scale salt producers carried out by TFNC in 2002/2003 (TFNC, 2003) identified about 4500 producers. Only 30 percent of them were iodating salt. A plan is underway to provide small-scale salt producers with appropriate technologies for salt iodation. Salt producers will be encouraged to form cooperative societies. Through such societies producers will be able to access credit facilities more easily, for purchase of potassium iodate, equipment and other suppliers for salt iodation. It is envisaged that, the plan will result in having 90 percent or more households consuming iodated salt as per WHO goal. (iii) VAD Control Program The first VAD control program was started in the 1950s by the British colonial government when red palm oil was targeted to primary school children. This program died out after independence. In 1977 a review of causes of blindness and xerophthalmia in school children was carried out. It revealed that about 75 percent of blindness was related to measles; and post-measles blindness was related to VAD. In 1981 the program for the control of VAD was re-established. Activities were confined to the drought­stricken Shinyanga region and constituted distribution of vitamin A capsules, promotion of horticultural products and improvement of child feeding practices. The first five-year program for the control of VAD was started in 1985. The program components included targeted vitamin A supplementation, promotion of production and consumption of vitamin A-rich foods, and public health measures aimed at conserving vitamin A in the body. Vitamin A capsules supplementation Vitamin A capsules are targeted to children aged below five years. Lactating women are prescribed with a single dose of vitamin A capsules immediately after delivery or any time 20 within four weeks after delivery. The aim is to increase the concentration of Vitamin A in breast milk for the benefit of the baby. In the process the mother benefits also in that she replenishes her body's vitamin A stores. All children up to 24 months of age are supplemented with vitamin A at nine months (during measles vaccination), 15 months and 21 months. Furthermore, children suffering from diseases that deplete the body's vitamin A status (diarrhea, measles, ARI, TB and severe clinical PEM) are supplemented with vitamin A. The strategy is referred to as " disease targeted vitamin A supplementation''. In recent years, in recognition of the fact that coverage of vitamin A capsules supplementation was not going high enough, an additional strategy was adopted ­ to supplement children through campaigns. A nationwide campaign is carried out twice a year: during the commemoration of Day of African Child in June and in December on World AIDS Day. Coverage has been high - the December 2003 campaign recorded 94 percent coverage. The routine supplementation has an element of sustainability in that it is integrated in the government health delivery system (capsules included in the monthly EDP kit). But, currently the utilization of the capsules in the EDP kit is not high enough as turn up of children to health clinics is low. Also KAP of health workers with regard to dispensing and prescribing the capsules is inadequate. Secondly, even if the community and health workers were sensitized, the quantity of capsules in the kit would not meet the demand. Two tasks, therefore, would lie ahead of the program ­ sensitization to create demand and increasing quantity of capsules in the EDP kit. With regard to supplementation through campaign, the exercise is short term measure intended to fill in the gap and will be phased out gradually as routine supplementation picks up. Food Based Approach Control of Vitamin A deficiency was also implemented through promotion of production and consumption of foods rich in vitamin A and adequate breastfeeding. In 1992 the program initiated home yard and school gardens in Singida rural district (under the World Bank supported HNP described above). Capacity Building Capacity building in the program included training of health care providers in the whole country on vitamin A capsule delivery. Reader materials were also produced and distributed to complement the efforts. Like the other micronutrient malnutrition control programs, the VAD control program is steered by National Vitamin A Consultative Group and task forces. Implementation of activities is done by the existing infrastructure. For example, the Medical Stores Department (MSD) delivers the capsules to the districts while the district uses the local government administrative structure and health facilities to organize campaigns and distribute capsules. 4.3 Nutrition Related Health Programs The nutrition programs are often complemented by activities undertaken by other sectors, particularly the health sector. Health programs which closely relate to nutrition and make significant contribution to the efforts for control of malnutrition include EPI, IMCI, malaria 21 control, parasite control and control of diarrhoeal diseases. All those programs are under the MoH. 5.0 Effectiveness of Tanzanian Programs in Reducing Malnutrition The major programs posed to demonstrate effectiveness in reducing malnutrition in Tanzania are the Iringa Nutrition Program, Child Survival, Protection and Development Program and the Micronutrient Malnutrition Control Program. 5.1 The Iringa Nutrition Program At the time of evaluation of the INP in 1998 (URT/WHO/UNICEF, 1988) a marked decrease in both severe and moderate rates of underweight were observed. Severe malnutrition had been reduced from 6.3 to 1.8 percent and total underweight from 55.9 to 38.0 percent in a period of five years. 5.2 The Child Survival and Development (CSD) Program The Child Survival and Development (CSD) Program was initiated in 1987 through the application of the basic elements of the approach developed in Iringa JNSP. CSD program began with an extension of the Iringa JNSP beyond the borders of the original project area to cover all districts in the region and later extending similar interventions in six other regions. CSD program was thus basically a result of positive outcomes of the implementation of the Iringa JNSP. The Iringa JNSP had demonstrated that it was possible to improve the survival of children as well as promoting their development through implementation of interventions that lead to reduced severe and moderate malnutrition. Similar approaches as those applied by JNSP were applied in the implementation of CSD program. CSD program emphasized on community participation which involved village mobilization, the establishment of village health committees, training of village health workers and the initiation of village health days which were responsible for increased immunization rates and community based growth monitoring. 5.3 The Child Survival, Protection and Development Program The annual cost per child in the CSPD approach was $2-3 in 1987. Three quarters of the costs were for imported items like drugs, for village health workers and transportation [URT/UNICEF, 1990]. An estimate for program costs in 1990 for the whole country at about 3 US$ per child was about US$ 15 million for all children in mainland Tanzania. Part of this money ($11 million) was for imports requiring foreign exchange. Thus with financing of three dollars per year per child from community, district and regional level it is possible to reduce significantly the rate of severe malnutrition. Contributions from district and regional level is in terms of time of extension staff and supervision while that from the community is in terms of time of parents and members of the village health committee. In addition community contribution should include compensation for village health workers and child feeding post attendants. 5.4 Prevention of Low Birth Weight Project The projects for the prevention of LBW implemented in the two refugee camps in western Tanzania showed strikingly positive impacts. However it has not been able to obtain information about the costs involved. 22 Similarly information on the costs involved in the other project being implemented in seven districts of Tanzania is not available yet. The project is being implemented in four phases: 1) sensitization / advocacy 2) baseline situation analysis 3) intervention / monitoring and 4) evaluation. Presently it is in its third phase ­ intervention and monitoring. Information on the costs involved will be obtained when the project is evaluated in 2006. 5.5 IDD Control Program The Government with major support from development partners mainly Sida and JICA has invested substantial financial resources in the IDD control program since 1986 when it was initiated. The main cost component of the program has been the procurement and distribution of iodinated oil capsules. A total of 16,420,000 (8, 210,000 doses) of iodinated oil capsules were procured and provided to population in 27 IDD endemic districts in the country in a period of 12 years, from 1986 to 1998. Calculations done showed that the cost of providing one dose per person was US$ 0.25 (Kavishe et al, 1988). The total cost for the procurement and distribution of the 16,420,000 capsules was therefore US$ 2,052,500. This approach has recently been phased out. A much more sustainable and long-term strategy of universal salt iodation of edible salt took over the implementation of which began in 1988 when capsules supplementation was still going on. The financial resources input in iodation of edible salt can be said to be sizeable. It has however not been possible to come up with actual figures of the cost involved in the implementation of the strategy. Cost items included procurement and distribution of potassium iodate, iodation machines, packaging materials and vehicles for supervision and monitoring all of which were provided to large and medium scale salt producers. 5.6 Vitamin A supplementation The cost of vitamin A supplementation activities may be considered in terms of cost of supplements, freight and handling charges, in-country distribution costs, social mobilization and personnel costs. The total cost of supplementing 6 million targeted children aged 6 months to 5 years in 2002 was estimated to be US$ 855,604 per year or approximately US$ 0.14 per child per year. This estimate assumes that local councils contributed funds to match the support provided by partners i.e. UNICEF and USAID. There has not been a cost-effectiveness assessment study in relation to VAS in Tanzania, but considering the known effects in morbidity and mortality reduction it must be highly cost effective. Experiences from other countries have demonstrated as well that VAS is cost- effective. For example, studies conducted in Ghana in year 2000 indicated that VAS had the lowest cost per child death averted (i.e. $ 10) compared to those of other PHC interventions implemented in various parts of the world that had ranged from $ 190 for breastfeeding promotion to $ 3,835 for oral re-hydration therapy (Institute of Medicine, 1998). 6.0 Current thinking about nutrition interventions PEM continues to be a major nutrition problem in Tanzania and it has been shown that controlling PEM can yield positive outcome, in improving the nutritional status of micronutrients such as Vitamin A and Iron. However for accelerated improvement of overall 23 nutrition status, there is a need to design programs addressing nutritional problems holistically including all the major micronutrient deficiencies. The ongoing public sector reforms have led to the devolvement of powers to the Local Government Authorities (LGAs) providing an avenue and new impetus in redressing the top ­ down decision-making process. The reforms recognize the role of the Local Government Authorities in spearheading development agenda in the districts with emphasis on increasing efficiency, productivity and accountability. Nutrition is in human development agenda, which should be strongly supported by the Local Government Authorities. Nutrition indicators are among priority indicators incorporated in the Poverty Reduction Strategy (PRS) implementation plan and the National Poverty Monitoring Master Plan that will monitor the implementation of PRS. The Health Sector Reform (HSR) is aimed at promoting more rational planning, implementation and support of health service delivery. Service delivery in Tanzania under HSR is implemented through the National Package on Essential Health Interventions. Nutrition is one of the interventions included in the package. However the delivery of the essential interventions is health facility based, hence marginalizing the nutrition component that is better addressed at community level. Programming nutrition interventions within the reform framework would reduce the verticality of approach that has been the common practice and would also facilitate increased human and financial resource allocation at sub-regional and community levels. In addition it would promote increased community empowerment, participation and ownership. This process allows for planning and implementation of comprehensive nutrition interventions in a more focused and targeted manner. It allows for equitable access to nutrition services by the poor. The following interventions are considered as being key in combating PEM and micronutrient malnutrition. Prevention of LBW National prevalence of LBW (below 2.5 kg) in Tanzania is estimated at 16 percent. LBW is being a reflection of poor maternal health and nutritional status. Thus to address the problem of LBW we need to improve maternal care, particularly during pregnancy. In Tanzania there is an established system of antenatal care that can improve nutritional status and birth outcome among women. Iron / folic acid supplementation during pregnancy will be strengthened. Prevention of malaria through the use of ITNs and presumptive treatment of malaria with S/P combination can significantly improve the haemoglobin level and birth outcome, so is the treatment with mebendazole / albendazole during the second trimester of pregnancy. It is also known that workload reduction and syphilis screening and treatment in women have significant benefits for both the mother and the baby. Given as an integrated package, the above interventions are envisaged to lead to improvement in maternal weight gain and micronutrient status ­ with benefits to the woman herself and pregnancy outcome (weight and survival of the neonate). This, in turn, may reflect in improved perinatal mortality. 24 Infant and Young Child Feeding Exclusive breastfeeding rate is very low only 11% at 4-6 months. There is a delay in initiation of complementary feeding with a feeding frequency of only 2-3 times a day. Complementary foods given are low in nutrient density. These factors contribute significantly to the high levels of malnutrition among the under 5s in Tanzania. Infant and Young Child Feeding activities will continue to be implemented with great emphasis at sub regional and community. The promotion of optimum breastfeeding, exclusive breastfeeding up to six months and appropriate complementary feeding at six months of age are necessary to address the problem of malnutrition in under 5s. In the era of HIV/AIDS emphasis will also be on promotion of appropriate infant feeding options and replacement feeding. The BFHI status is very low (only 59 hospitals out of 112 are baby friendly). Scaling up the BFHI will be a priority and initiating of Baby Friendly Community Initiative. A monitoring mechanism will be put in place to sustain the Baby Friendly status of the health facilities and communities. With the liberalization of trade there is a need to review and enforce the National Regulation of Marketing of Breast milk Substitutes and Designated Products. Use of local knowledge and traditional technologies will be promoted in the development of appropriate complementary and replacement feeds. Growth Monitoring and Promotion (GMP) Growth Monitoring and Promotion (GMP) was one of the major components in the Iringa JNSP. It was a strategy that was used to motivate families and communities to participate fully in the program. Growth Monitoring and Promotion is the JNSP centred on the monthly health days under which households and communities were participating in assessing the growth of their children through weighing. This exercise generated enthusiasm among the households and communities to take remedial measures and action on children who were found to have faltering weights or were found to be malnourished. GMP essentially facilitated the application of Triple A cycle in the program implementation. Growth Monitoring and Promotion was thus a motivating strategy. Experience of the Iringa JNSP in the employment of GMP was extended to CSPD program. It continues to be a major component of CSPD programs in all the 57 districts although its potential is not being fully realized. Part of the reason is that there is low capacity at community and facility level in terms of personnel and skills. Also there is shortage of supplies mainly weighing scales and pants as well as growth charts. Thus deliberate and concrete measures and actions need to be directed towards revitalizing GMP at community level as well as strengthening it within the Reproductive and Child Health system at facility level. Information generated through GMP should trigger action at primary level and should be disseminated to other levels for appropriate decision making. Community Based Nutrition Rehabilitation CBNR is a system of rehabilitating malnourished children at home. Contrary to the approach of NURUs (residential nutrition rehabilitation units attached to hospitals) the CBNR approach stresses rehabilitation of malnourished children in the same environment (village, home) that precipitated the condition, using resources and infrastructure available in the community. 25 CBNR is implemented at village level. Malnourished children are identified at various contact points such as clinics, health campaign meetings or home visits by health workers. The condition of the child is categorized (based on severity and disease complication) and action is taken. Action may be to register the child for home rehabilitation or to first admit to a health facility (for management of a severe condition and / or disease) after which the child is discharged for home rehabilitation. "Home rehabilitation emphasizes on improved child feeding, growth promotion activities, general care to prevent infections and nutrition education to parents and caretakers. In this way the benefits of the CBNR approach flows also to siblings and all family members. Establishment of an effective CBNR system requires fulfillment of the following strategies: · Advocacy to leaders (from district down to community level) so that they facilitate establishment of CBNR and ensure sustainability of the system in their catchment areas. · Equipping health care providers (inducing village health workers) with knowledge and skills to implement CBNR. · Ensuring availability of necessary equipment and supplies (weighing scales, tape measures, growth monitoring forms) used for identification and categorization of malnutrition. · Sensitizing and raising awareness of parents, care takers and community leaders on home rehabilitation of malnourished children. With well-established system of CBNR it is possible to achieve the long-term goal of reducing PEM to a level whereby it is no longer a problem of public health significance in the given community. Nutrition in Primary Schools Available information shows that poor school achievement, repetition of grades and dropouts may be symptoms of poor nutrition status. Heavy parasite load combined with low food intake precipitates malnutrition in school children. It has also been shown that the school nutrition programmes provides an opportunity for catch up growth in those children who were stunted during pre-school age. An assessment conducted by TFNC in 1998 in two districts showed low levels of food intake during school time. Of those surveyed 23% had something to eat for lunch, 24.2% ate breakfast, 36.3 ate dinner, and 16.5% had something to eat after school (TFNC Report 1839). These figures indicate that primary school going children experience hungry periods at different times during the day. The policy level support on development of guidelines on Management of Community Based School Feeding Programmes together with capacity development in the management need to be scaled up. The provision of school meals should target elimination of hunger during school time and focus on schools that are in the poorest areas of the given community. This intervention would contribute in improving school performance. Nutrition Care and Support for PLWHA The interrelationship between HIV/AIDS and nutrition has been recognized since early in the AIDS era. Deficiency of macronutrients leads to negative energy balance and wasting in HIV- infected person. In many AIDS patients when wasting of lean body mass approaches 55 percent of the normal for age, sex and height death is imminent. Similar deficiency of micronutrients leads to further lowering of the immune competence. 26 In recognition of this PLWHA will be given nutrition care and support. Policy and guidelines will be developed addressing the following: · Nutrition screening for PLWHA · Nutrition education and counseling · Dietary management of HIV/AIDS and related conditions · Infant and young child feeding in the context of HIV/AIDS · Food products development. Health and nutrition planners, managers and implementers involved in both hospital and home- based care will be trained on HIV/AIDS. The guidelines will also be disseminated to these cadres. Media channels and reader materials will be used to raise public awareness on nutrition and HIV/AIDS. Food Based Approach Promotion of the production and consumption of foods rich in Vitamin A and prolonged breastfeeding are some of the approaches that are used to control vitamin A deficiency in Tanzania. In 1992, backyard and school gardens were initiated in Singida rural district on a pilot basis. The experience gained in Singida rural district can be used to replicate the program in other areas with similar ecological conditions and where appropriate incorporated in ECC-SGD. This should go hand in hand with promotion and where appropriate improvements in traditional fruits and vegetable preservation technologies. Micronutrient Supplementation Micronutrient deficiencies are major nutrition problems in Tanzania. Iron, iodine and vitamin A deficiencies affect large proportions of individuals across communities and population groups. As regards supplementation Tanzania for the time being will continue to focus on iron / folic acid and vitamin A. Iron / folic acid and vitamin A supplementation is effective in controlling the deficiencies when distribution coverage and compliance are high. Also effectiveness and impact of supplementation is easier to recognize than that of food based approach or fortification. As a result of this, supplementation is increasingly being recognized to be a medium term rather than a short-term intervention. Tanzania's policy guidelines on micronutrient supplementation identifies the following target groups: pre-term babies, low birth weight babies, normal children aged below two years, adolescents, pregnant women and people with sickle cell disease (MoH, 1997). However, currently only pregnant women and sickle cell patients are being supplemented. Resources, infrastructure and delivery systems are yet to be worked out for supplementing the remaining groups. Iron / folic acid supplementation: While Tanzania will continue to supplement pregnant women and sickle cell patients the program needs to be strengthened greatly. At present only the primary and secondary level health facilities receive monthly supplies of iron and folic acid (through the EDP kit). For increased coverage the supplements will have to be made available to hospitals as well. Health workers will be trained and sensitized to properly prescribe the 27 supplements and give advice to clients on the use of the supplements. Also a program will be developed to sensitize communities so that pregnant women book for antenatal services early where they are provided with the supplements. Vitamin A Supplementation: will continue to be targeted to children aged below five years of age. Children aged up to two years will receive vitamin A at nine months (during measles vaccination), 15 months and 21 months. Children presenting with diseases associated with depletion of vitamin A status (xerophthalmia, measles, respiratory infection, diarrhoea, severe PEM) will also be prescribed with vitamin A. Maternal supplementation entails providing vitamin A supplement to the mother immediately or any time within four weeks after delivery. While the aim is to increase the vitamin A content of the breast milk (for the benefit of the breastfeeding child), the mother's vitamin A stores are replenished in the process. Supplementation for all children aged below five years conducted in form of campaign twice a year achieves high coverage. Presently the campaigns are centrally organized which has an implication on increasing costs. Procurement of the capsules shall therefore be pegged into the existing system; and campaigns to stimulate attendance will be organized locally at health facility and village levels. That way costs will be reduced and the approach will be sustainable. As in the case of iron / folic acid supplementation training of health workers and community sensitization with regard to vitamin A supplementation will be undertaken. Food Fortification: Apart from salt iodation, fortification of common foods as a strategy to address the problem of micronutrient deficiencies in Tanzania is in its infancy. There is an attempt to fortify maize flour. The pilot projects being implemented in Iringa, Handeni and Korogwe are aimed at demonstrating the feasibility of reducing micronutrient deficiencies through fortification of maize flour at hammer mill level. If successful the activity will be scaled up to cover most parts of the country that consume maize and / or other starchy staples that go through mill processing. The second approach is to promote fortification of centrally processed foods (flour, oil and sugar). Already a company based in Arusha called International Health Food Association is fortifying cereal flour. Efforts will therefore be made to rally other manufacturers into fortifying their products. Salt Iodation: Proportion of households consuming iodated salt in Tanzania is 83 percent which is below the WHO recommended goal of 90 percent or above. In early 1990s, under the national program for the control of IDD, large and medium scale salt producers were given material and technical support to iodate salt. Small scale salt producers were not addressed during this period. There are about 4500 small scale salt producers in Tanzania and are largely responsible for the un-iodated salt in the market. 28 Small scale salt producers will be facilitated to iodate salt. In order to enable their "take­off" with regard to iodating salt they will receive material support which will include iodation equipment, potassium iodate and iodine test solution. Technical support will include skills on salt production, iodation and handling. Awareness creation will be effected through seminars and reader materials. That way all salt producers in the country will have been in position to iodate salt. Meanwhile salt inspectors (health officers and mining officers) will be trained to conduct effective inspection of all salt producing sites and training centers in their catchment areas. Local authorities (political leaders and administrators), the police and members of the judiciary will be sensitized on the need to enforce the salt regulation. The strategy of creating consumer demand for iodated salt will be employed. Open air meetings with villagers will be organized in which the value of iodated salt with regard to health will be explained to the people at the grass-root level. Household Food Security In the 1970s TFNC initiated programmes on household food storage. The activities focused on improvement and development of traditional post harvest technologies, including food storage, processing and preservation at household level. Despite these effort post harvest food loses continue to be high. There is a need to reassess the situation and identify appropriate actions for reducing post harvest loses at household level. Other efforts to address food insecurity at household include development of a monitoring tool, (The Food Security Card) for assessing food stocks (cereal grains and pulses) at household. The card has been useful in that families have been able to at least estimate amount of food required for the whole season before selling off the extra. The tool needs to be popularized among cereal / pulses users and also it needs to be developed further to include other commonly used foods such as cassava, plantains, sweet potatoes, potatoes and animal products. Promotion of roots and tuber crops has been going on since 1994. Processing, preservation and storage techniques have been developed to extend the shelf life of these products. These efforts need to be scaled up to cover all areas that produce these crops. Also there is a need to make deliberate efforts to promote consumption of other traditional foods (millet, sorghums, pumpkins, forest products) as a means of ensuring food security at household level. Capacity Development The few successes observed in programming and implementation are attributed to participation of district and regional level actors. However, the initiation, planning and management of these programs were and have mostly been originating from the central level with very minimum active participation by regional and district level actors. Little efforts have been taken to develop required capacities at regional and district levels. The outcome of this is: · Weak link between the central level actions and the regional and district level requirements, leading to uncoordinated initiation and implementation of nutrition programs; · Inadequate integration of nutrition in regional and district development plans; · Inadequate or poor follow-up of positive impacts resulting from some successful nutrition projects at the district level; · Nutrition is not a major agenda at the regional and district level; 29 In this way nutrition interventions have always assumed the top ­ down characteristic with little ground at regional and district levels. In order to reverse this situation, deliberate efforts ought to be taken to develop capacity at regional and district levels to initiate plan, and manage nutrition programs. Nutrition IEC Most of the approaches to nutrition interventions in Tanzania have an IEC component. Some of the IEC techniques used include training through seminars and workshops; short courses, meetings, mass / print media, and counseling (particularly in CSPD areas). The impact of IEC for nutritional improvement has not been evaluated. However, the Sida evaluation of 1999 noted that well formulated IEC was the success of the IDD control program. The big challenge for Tanzania is to formulate education and communication strategy with strong social marketing and behaviour change approaches. This is due to the fact that nutrition programs have established services that are underutilized or not properly utilized. For example, there is low utilization of iron/ folate and vitamin A capsules in spite of the nutrition messages that are currently used. Management of Information Systems (MIS) Any community considering commitment to solving nutrition problems will need, early in the process, to assemble some information as a basis for deciding priority problems and possible actions. During 1990s UNICEF supported a nutrition surveillance system coordinated by TFNC. The system was implemented at three levels namely national, district and community (i.e. ward and village). Through the system, at the national level it was possible to collate nutrition related data from various sources and then disseminate them to various users through an annual publication "Tanzania Nutrition Trends". The report was a useful input to decision makers at national level and other users particularly those involved in nutrition related programme development and planning. At district level TFNC provided technical support to few pilot districts on nutrition data management. The districts were Kilosa, Njombe, Masasi, Hai and Makete. These districts established own databases and are carrying out integrated planning. Similar support was provided to some villages to produce community level data for planning and decision-making. Village registers and growth monitoring exercise formed the backbone of the community-based data. In spite of these efforts there are no reliable data from which to report the nutrition situation of the country. The only reliable source of data at national level is RCHS. However, this reports on very few nutrition indicators, gives national estimates only and provides the data on a four- year cycle. Such data is not very useful for timely decisions and actions. Since nutrition indicators are among priority indicators incorporated in TSED and the PRS implementation Plan and the National Poverty Monitoring Master Plan there is a need to revamp the nutrition surveillance system that will make available reliable nutrition data. 30 Operational Research Over the years TFNC has conducted much research that can be described as applied research i.e. research addressing problems and questions relevant to Tanzania. Notable examples include: the development and testing of "power" flour as a weaning food; testing the efficacy of a multiple micronutrient fortified beverage in improving nutrient status among school children and pregnant women; use of solar drier in conserving provitamin A and other nutrients in green leafy vegetables; use of fermented beverage togwa as a means of lowering diarrhoeal incidence among children; and potential of hammer mill fortification to combat micronutrient malnutrition. It is envisaged that the research endeavour shall continue and will focus on the four major nutritional deficiencies namely PEM, nutritional anaemias, vitamin A deficiencies and iodine deficiency disorders. Other areas that may need to be researched on include infant feeding, weaning foods, nutrition in relation to HIV/AIDS, adolescent nutrition and nutritional problems of women. The key to all these is that research must address a current priority area, should be relevant to Tanzania and is well designed. It should also meet ethical standards. Institutional Arrangement Most of the comprehensive nutrition intervention programs described in this section are in progress in Tanzania. Some of them have been successful while others have not been implemented effectively and efficiently. As it has been pointed out above successful programmes have been those targeting the prevention and control of micronutrient deficiencies mainly VAD and IDD. Programmes addressing PEM and nutritional anaemia have not demonstrated significant impact during the whole period of their implementation. This situation is attributed to operational and organizational problems specifically in terms of responsibility, accountability, coordination and collaboration, support and resources both human and financial. It is a well-known fact that nutrition in Tanzania and elsewhere has no sector or ministry responsible for it. Nutrition is a cross cutting issue requiring various sectors intervention. Nutrition has largely been the mandate of the Ministry of Health and to some extent the Ministry of Agriculture and Food Security and Presidents Office, Regional Administration and Local Government. But these ministries have their major mandates and priorities that do not feature nutrition highly. As a result nutrition has been marginalized and t best seen as a TFNC issue. For effective implementation of nutrition intervention programme and for nutrition to get a place in PRS the current institutional arrangement needs to be reversed. Nutrition has to be the responsibility of all key Public Sectors whose mandate and responsibilities include nutrition components. Nutrition needs also the inputs of NGOs and the private sector. However in ensuring success and sustainability of nutrition intervention the two sectors need to develop a partnership. Currently this partnership is almost non-existent in Tanzania. Nutrition has by and large remained the responsibility of the public sector with very minimum support and collaboration from the private sector. Currently the private sector is active is in salt iodation but there is room for more involvement and participation. Strategies have to be worked out that will improve communication between the public and private sector for their increased participation in nutrition interventions. 31 Another operational concern in the implementation of nutrition intervention program is the active participation of beneficiary community. The Iringa JNSP and currently CSPD programmes have demonstrated that the success of nutrition intervention programmes have always been very much dependent on active participation and ownership of the community itself in the whole programming process. More concerted efforts will thus need to be done to promote community ownership of the proposed comprehensive nutrition interventions programmes. Proposed Operational Framework Comprehensive nutrition interventions programme can effectively and efficiently be implemented at central, district and community level. This arrangement will be in line with the current reform process, which ensures the decentralization of actions, whereby the centre remains with the responsibilities of capacity development, supervision and monitoring resource mobilization and technical back stopping. The district and community levels become the level of implementation and will facilitate targeting and management of program interventions. This arrangement will ensure community empowerment, ownership and participation. The key players at the central level will be the relevant sectors of health, agriculture and food security, community development, and education. Other key players will be TFNC and Development Partners including UNICEF, World Bank, and WHO. At the district and community levels, the accountability for nutrition programme implementation will rest with the Regional Administration and Local Government with Local Authorities playing the leading role specifically in resource allocation and mobilization of the community to fully participate in the program implementation. Local Authorities will have the responsibility of assigning a specific officer within the ranks of its staff to manage nutrition interventions as well as incorporating nutrition in the district development plans. The incorporation of nutrition in district plans will reduce the dependency of districts on central government for all interventions demanded, which is the case today. At present the district and community levels have low capacity in-terms of human and financial resources. Deploying some of the the highly experienced personnel at the central level to regions and districts will facilitate informed decision making and proper resource allocation at the local level. The government in collaboration with development partners in nutrition can explore ways to facilitate the deployment. Sectoral Responsibilities and accountability in comprehensive nutrition interventions. The comprehensive nutrition interventions proposed for implementation is the responsibility of various sectors: health, community development, agriculture, education as well as partnership of the public and private sector. Prevention of low birth weight, infant and young child feeding, community based growth monitoring and promotion, prevention, care and support for people living with HIV/AIDS, vitamin A and iron/folate supplementation are the responsibility of health and community development sectors with support from development partners and local government. The agricultural sector has the responsibility of implementing household food security interventions and food based micronutrient deficiency control measures. 32 School health and nutrition programmes and school feeding rests with the education and local government while public ­ private sector partnership should be active in food fortification and salt iodation. TFNC in its present status will provide technical support and guidance to all sectors and at all levels. This responsibility and accountability arrangement is also in line with the National Food and Nutrition Policy which is under review. 33 7.0 REFERENCES 1. Alderman H, Hoddinott J, Kirsey B. Long term consequences of early childhood malnutrition. Dalhousie University, Canada November 2003. 2. Annegers JF. Ecology of dietary patterns and nutritional status in West Africa. Distribution of starchy staples. Ecol Food Nutr. 1973. 3. Ballart et al. Report of the National Vitamin A Survey 1997. TFNC Report No. 1880; December 1998. 4. Ballart A. Prevention of low birth weight in the refugee camps of Western Tanzania. The case of Lukole Refugee Camp. UNICEF, Ngara, 2000. 5. Barker DJP, Osmond C. Infant mortality, child nutrition and ischaemic heart disease in England and Wales. Lancet I; 1977-1081 1986. 6. Barker DJP, Winter DP, Osmond C, Margells B. Weight in infancy and death from ischaemic heart disease. (Lancet II; 577-580 1989. 7. Behrman JR, Alderman H, Hoddinott J. Hunger and Malnutritin; Coperhagen Consensus ­ challenges and opportunities. World Bank February 2004. 8. HelpAge International. Sukumaland older women programme. Prevalence of malnutrition among older people in Magu district and factors affecting food intake, leading to malnutrition. HelpAge International, Dar es Salaam, 2001. 9. Household Budget Survey (HBS) 2001. Report on Household Budget Survey in Tanzania.National Bureau of Statistics; Dar es Salaam, 2001. 10. Institute of Medicine. Presentation of Micronutrient Deficiencies. Washington DC: National Academy Press; 1998. 11. Kavishe F. P. Nutrition Relevant Actions in Tanzania. TFNC Monograph Series No.1. April 1993. 12. Kavishe F.P., Mlingi N.V., Chorlton R., Report of the Second National Expert Committee meeting held on 10-11 June 1987 at TAZARA Hostel, Dar es Salaam. TFNC Report No.1129, 1988. 13. Kisanga P, Ngonyani M, Lyamuya V. An in-depth Study in child feeding practices in two Arusha urban location (Ngarenaro and Tindiga). TFNC report No. 899. Dar es Salaam, 1984. 14. Lukmanji Z. Dietary treatment for diabetes in developing countries: Tanzania experience. TFNC Report No. 763, 1982. 34 15. Mduma et al. Report of nutritional survey in 30 villages of Lindi Rural district. TFNC Report No. 1656; 1992. 16. National AIDS Control Programme (NACP). HIV/AIDS/STI surveillance report, January ­ December 2002. Ministry of Health, Dar es Salaam 2002. 17. National Bureau of Statistics. Tanzania Social Economic Data Base 2002. Dar es Salaam 2002. 18. Ndunguru, G. and Basheke, M. A Practical Appraisal of on farm storage losses and loss assessment study on maize in Tanzania. TFNC Report No. 1628; 1994. 19. Pelletier DL, Cornell Food and Nutrition Policy Program. The Uses and Limitations of Information in Iringa Nutrition Program, Tanzania. TFNC 1991 20. Reed B. Association between antenatal care and improved birth weights in Mtendeli Refugee Camp, Kibondo Tanzania. First Draft Report: UNICEF Tanzania, February 2002. 21. Tanzania Demographic and Health Survey (TDHS), 1991/92. National Bureau of Statistics, 1992. 22. Tanzania Food and Nutrition Centre (TFNC). Evaluation of the IDD Control Programme. TFNC Report No. 1924. August 2000. 23. Tanzania Food and Nutrition Centre (TFNC). Prevalence of low birth weight, risk factors and level of maternal service and care provision in 7 districts of Tanzania. TFNC, 2002. 24. Tanzania Food and Nutrition Centre (TFNC). Data report on the food and nutrition situation in Tanzania. In: Analysis of the situation of children and women, vol 2. TFNC 1983. 25. Tanzania Food and Nutrition Centre (TFNC). The National IDD Control Programme. Inventory of small-scale alt producers in Tanzania. TFNC Report No.1993, October 2003. 26. Tanzania Food and Nutrition Centre (TFNC). UNICEF (United Nations Children's Fund) and ICH (International Child Health) London. Baseline Survey on the Prevalence of Low Birth Weight Risk Factors for LBW and levels of service and care provision in seven districts in Tanzania. TFNC 2002. 27. Tanzania Food and Nutrition Centre (TFNC). The National IDD Control Programme. National IDD Survey conducted during 2003/2004. TFNC Report No. 2002, April 2004. 28. Tanzania Reproductive and Child Health Survey (TRCHS) 1999. National Bureau of Statistics. Dar es Salaam; 1999. 29. Tanzania Food and Nutrition Centre (TFNC) / Action Group (USA). Primary School, Nutrition and Health in Tanzania; 1995. 35 30. United Nations Children's Fund (UNICEF). 1990 ­ Women and Children in Tanzania: An Overview. UNICEF, Dar es Salaam; 1990. 31. United Nations Children's Fund (UNICEF). 2001 ­ Situation Analysis of Children in Tanzania. UNICEF, Dar es Salaam; 2001. 32. United Republic of Tanzania (URT). National Economic Survey for 2002. June 2003. 33. United Republic of Tanzania (URT) ­ Ministry of Health. National Food and Nutrition Policy - Draft. TFNC March 2003. 34. United Republic of Tanzania (URT) ­ Prime Ministers's office, Regional Administration and Local Government, Tanzania Health and Nutrition Project ­ Component II. Baseline Survey Report. TFNC, July 1994. 35. United Republic of Tanzania (URT) and UNICEF. Assessment of Dietary Adequacy; In Analysis of the situation of children and women, Dar es Salaam, 1985. 36. United Republic of Tanzania (URT) WHO/UNICEF (World Health Organisation and United Nations Children's Fund) 1988. The Joint WHO/UNICEF Nutrition Support Programme in Iringa, Tanzania 1983-1988 Evaluation Report. Dar es Salaam; 1988. 37. World Bank. Health, Nutrition and Population Development goals. Measuring progress using the poverty reduction strategy framework. Report of World Bank Consultation. November 2001. 38. World Bank. World Development Report: Making services work for poor people. 2004. 39. World Health Organisation (WHO). Proposed Indicators for Achieving End of Decade Goal. In: World Summit for Children ­ Mid Decade Goal: Vitamin A deficiency. WHO/UNICEF/JCHPSS/94/2.8. November 1993. 40. World Vision (T). Consumption pattern of maize flour in Mzundu Division, Handeni District. World Vision (T); 1998. 41. World Vision (T). The MICAH Programme. Report of Nutrition Baseline Survey for the ADPs of the Northern Zone. World Vision (T); September 1998. 36 Coffee Price Risk in Perspective ­ Household Vulnerability among Rural Coffee Growing Smallholders in rural Tanzania March 10, 2006 Luc Christiaensen, Vivian Hoffmann and Alexander Sarris1 Abstract: Despite the precipitous decline in coffee prices in early 2000, cash crop growing smallholders in Kilimanjaro and Ruvuma, Tanzania, identified health shocks, droughts as well as commodity price declines as their major risk factors. About one third of the rural population in Kilimanjaro suffered either from drought or health shocks in 2003, resulting in 18 percent welfare loss. Through reliance on savings and aid they reduced this loss to 8 percent on average. In Ruvuma rainfall is more reliable and drought did not affect welfare. Surprisingly, health shocks appeared not to affect welfare either, most likely related to lower observed medical expenditures in case of illness given limited access to health facilities. Coffee growers appeared not worse off in 2003 than non-coffee growers, apart from the smallest ones in Kilimanjaro, whose consumption level was on average 20 percent lower, and the largest ones in Ruvuma, whose consumption levels appear larger. Interventions to improve health conditions and reduce the effect of droughts emerge as important to reduce vulnerability in rural Tanzania. 1 Paper presented at the CSAE Conference 2006: Reducing Poverty and Inequality: How can Africa be included? held on 19-21 March 2006, at St Catherine's College, Oxford England. The authors are Senior Economist at the World Bank, Ph.D. candidate Cornell University and Director of the Commodities and Trade Division at the Food and Agriculture Organization of the United Nations respectively. The authors would like to thank Donald Mmari, Lucas Katera, Donald Sango and Professor Joseph Semboja at Research on Poverty Alleviation (REPOA) in Dar es Salaam, Tanzania, who have managed the surveys on which this research is based, Panos Varangis and Benno Ndulu for their continuous support of the project. The study has been funded by the EU Trust Fund on Commodity Management and the Dutch-Japanese Poverty Reduction Strategy Trust Fund. The findings, interpretations, and conclusions expressed are entirely those of the authors, and they do not necessarily represent the views of the World Bank, its Executive Directors, or the countries they represent, or the views of FAO and its member countries. 2 1 Introduction The precipitous decline in coffee prices since the late 1990s has attracted a lot of attention. Yet, not only are the actual welfare effects for coffee growers in Sub Saharan Africa of this price decline poorly understood, the focus on coffee prices has also distracted attention away from the wide array of risks coffee growing households face. A more holistic approach to commodity growing households' vulnerability status is called for. Moreover, while there is increasing evidence that rural households in developing countries only partially manage to smooth their consumption in the face of shocks (Dercon, 2004), the size of the immediate and long term welfare losses associated with shocks are still poorly documented. Such information is important to gauge the benefits from vulnerability reducing interventions. This paper takes a more holistic perspective on household vulnerability and examines the immediate effects of shocks on household welfare in the context of rural Kilimanjaro and Ruvuma, both coffee and cashew growing regions in Tanzania, while accounting for households' coping capacity and differential effects across livelihoods. From a directly administered shock module in a survey administered in 2003 to 900 households in each region, we learn that households identify droughts, health shocks and commodity price declines as their major risk factors both in terms of the frequency of their occurrence as well as the severity of their effects, underscoring the need for a more holistic perspective on household vulnerability even among commodity producing households. The paper addresses three broad questions in particular. First, it explores the welfare effect of the health and drought shocks and to the extent possible, it will also reflect on the welfare effects of the more systemic commodity price shocks. Second, the paper examines the effectiveness of different ex post coping and ex ante risk reducing strategies in mitigating the negative welfare effects associated with these risks. It will focus in particular on self and informal insurance schemes and irrigation respectively. Finally, the determinants of people's coping capacity are investigated. The paper proceeds by outlining the empirical methodology in section 2. A series of data considerations are addressed in section 3. Empirical results regarding the effects of the different shocks on household welfare are presented in section 4 and the effectiveness of the different coping strategy is explored in section 5. Section 6 examines the correlates of households' coping capacity, followed by concluding remarks in section 7. 3 2 Empirical Methodology From economic theory we know that, assuming households maximize inter- temporal utility with instantaneous utility concave (i.e. risk averse), households prefer smooth to volatile consumption. Given access to well functioning credit or insurance markets, these preferences will generate stable consumption paths, even when shocks occur. If credit and insurance markets are imperfect, household consumption may be susceptible to shocks (Deaton, 1992; Besley, 1995). These theoretical insights provide a practical framework to empirically explore whether and to what extent shocks and households' coping capacity affect their consumption levels. More formally, suppose households at time t maximize inter-temporal expected utility Ut. Let u (ct) be instantaneous utility derived from consumption ct (0) and u'(.)>0, u''(.)<0 such that: T Ut = Et ( 1+ )t- u(c ) (1) =t with the rate of time preference and T the end of the life-cycle. Households face risky income yt and income can be used to obtain consumption at prices pt. Define r as the rate of returns to savings between periods and At+1 as the value of assets at the beginning of period t+1. Assets evolve from one period to the next according to: At+1 = (1+ r)(At + yt - ptct) Solving (1) and (2) using the envelope condition and assuming that households have full access to credit and/or (formal or informal) insurance yields: u'(ct )= (1+ r) Et +1 pt (1+ ) u'(ct ) pt+1 Discounted marginal utilities suitably corrected for relative price changes will be equated. In the absence of uncertainty, with r equal to and prices constant over time, the optimal consumption path implies equal consumption over time. In the tradition of Hall (1978) and Morduch (1990) we assume constant relative risk aversion with instantaneous marginal utility defined at t as ct e - t with the coefficient of relative risk aversion and t a general taste shifter to parametrize (3) and obtain an empirical specification. Taking logs, and introducing subscript i and j to denote households i in location j, (3) can be written as: 4 cijt = ln(1+ r)-ln(1+)+ ln 1 ln +1 pt (4) cijt pt +1+ (ijt -ijt)+ eijt +1 +1 with eijt+1 the expectation error which has mean zero and is orthogonal to all variables known at time t given rational expectations. According to equation (4) the path of consumption over time is only affected by taste shifters and price changes, as long as there are no binding liquidity constraints over time and provided the underlying factors determining wealth (or permanent income) are not changing. In other words, under the hypothesis of perfect consumption smoothing, the optimal consumption path is not affected by idiosyncratic and/or covariate (income) shocks Sijt+1 and introduction of these shocks overidentifies equation (4). This provides an empirical framework to explore the effects of shocks on welfare. We further allow differential ability across households to cope with shocks ex post, leading to the following linear empirical specification: cijt ln +1=0 +1Zijt +2Sijt Mijt + eijt ` +1 +1 (5) c jit with Zijt comprising price changes and taste shifters (such as changes in household composition) and Mijt a vector of variables such as initial wealth, social capital, access to credit, availability of safety net programs, capturing the household's capacity to mitigate the effect of income shocks ex post. Differential ability to cope with shocks ex post is likely to condition the effect of income shocks on consumption. Alternatively, assume Xijt the comprehensive set of observable (and exogenous) household and location characteristics affecting preferences, permanent income and coping capacity (after shocks Sijt have materialized)2, such that cijt = c(Xijt, vij, j) with vij and j reflecting unobserved (time invariant) household and location heterogeneity respectively. Equation (5) can then also be written and estimated as: ln cijt+1 = 0 + 1Xijt + 2Sijt Xijt + vij +j + ijt ` +1 +1 (6) When panel data are available, equation (5) could be estimated (either as a difference or a fixed effects model) and unobserved household (and location) heterogeneity would be explicitly controlled for. Yet in practice, panel data are often not available, and when available, they tend to focus on a limited set of livelihoods/ populations and usually span relatively short time periods. This poses a particular challenge when studying the effect of slow onset, systemic shocks such as broad economic crises or a decline in commodity prices. The period covered by the panel may be too short to fully encompass the period of the shock (e.g. precipitous commodity price decline) and the shock may affect all households in the sample leaving the researcher in effect without a control group. Estimates of the welfare effect of an economy wide shock based on welfare before and after the shock will be biased, if there are secular trends. 2These include but are not limited to Zijt and Mijt. 5 Furthermore, the availability of repeated observations on a household's consumption and income, does not eliminate the need for explicit information on shocks to estimate the welfare effects of shocks. While changes in consumption are sometimes regressed on changes in income (Harrower and Hoddinott, 2005), attenuation bias due to oft observed measurement error in the latter would lead us to underestimate the effect of an income shock. At the same time, imputation errors in valuing consumption from own food production in constructing the consumption and income variables may lead to a spurious positive correlation between total household consumption and income, biasing the income coefficient upwards (Deaton, 1997). Direct information on shocks usually provides the necessary instruments to address this problem. It also enables inference on the effect of shocks on income and consumption. In the absence of panel data, but given cross sectional data on household consumption (Cijt+1), explicit information on shocks experienced during t+1 (Sijt+1) and comprehensive recall data on households' assets and their coping capacity (Xijt) the differential effect of different shocks across households could be explored through estimation of equation (6), in effect using a retrospective panel approach and assuming E(Xijtvij) = E(Sijt+1vij)=0. In practice, a comprehensive description of the household characteristics (Xijt) helps reduce the likelihood of potential bias due to unobserved household heterogeneity. Furthermore, potential endogeneity issues related to the shock variables can be avoided through the use of external shock information as opposed to self reported measures of shocks from the household questionnaire. The use of village fixed effects controls for bias due to correlation of X and S with unobserved village effects. Yet as this may cause an underestimate of the full effect of covariate shocks, it is useful to also explore models with an explicit comprehensive description of the location/village characteristics when available. Given that slow onset commodity price shocks such as the systemic coffee and cashew price shocks only directly affect producers of these crops the effect of these shocks could in principle be explored when the sample includes a sufficiently large control group of non-coffee or cashew crop growers with similar characteristics. The shock variable (Sijt) in this case becomes being a coffee (cashew) crop grower at t or not. Yet, caution is warranted in interpreting the empirical results. First, it is implicitly assumed that cash and non-cash crop growers are ceteris paribus equivalent (i.e. E(Sijt+1·vij)=0) such that the effect of being a cash crop grower only captures the effect of the systemic price shock. Second, if the overall economic activity in the region declines as a result of the price decline, the approach is likely to underestimate the direct negative effect as non-coffee growers are likely to have suffered as well, albeit indirectly. Bearing these caveats in mind and using a comprehensive specification to minimize potential bias due to unobserved differences in nature between cash crop and non-cash crop growers, the proposed approach also sheds light on the effect of the cash crop price decline on household welfare in Kilimanjaro and Ruvuma. 6 3 Data considerations To analyze the welfare effects of these different shocks, we use a primarily collected household vulnerability survey conducted in October-December 2003 in Kilimanjaro and February-March 2004 in Ruvuma. A detailed description of the survey and key characteristics of the households is given in Christiaensen and Sarris (2006). We take (the logarithm of) total household expenditures per adult equivalent excluding expenditures on health, education and functions (baptism, funerals) from the first survey round as our measure of welfare.3 To capture differences in household preferences, their permanent income potential and their coping capacity we include age of the household head (a life cycle proxy), the dependency ratio, gender of headship and the years of formal education achieved by the household head (allowing for differential effects across primary, secondary and post secondary education). As cultivation of certain cash crops may be traditionally dominated by certain ethnic groups (see below), we also control for the ethnic origins of the household head. This also helps control for people's social capital and thus their capacity to cope with shocks ex post. For example, the Chagga, which make up 74 percent of the total rural population in Kilimanjaro, are known to be highly mobile and well connected in Tanzania. To proxy households' productive capacity and thus also their permanent income potential, we include the size of their landholdings owned, the numbers of their large (cattle, oxen, horses) and small (goat, sheep, pigs) livestock owned, and the value of their agricultural equipment and vehicles (all normalized by the number of adult equivalents) as well as their squared terms to capture non-linearities in their effects on consumption. A self reported measure of ease in obtaining seasonal credit for inputs is included to proxy access to production (as opposed consumption) credit. The effect of the fall in coffee prices is explored through inclusion of the number of coffee trees owned by the household in 2000 when the coffee price decline set in. Where the data allow, we correspondingly also lag our asset variables to 2000 to be consistent. We furthermore divide the coffee growers in our sample in five quintiles based on their amount of coffee trees in 2000 to allow for differential effects among smaller and larger coffee farmers. The omitted category is the non-coffee growers, which makes up about one third of the total sample in Kilimanjaro. A similar approach is followed in Ruvuma, though we also include quintile categories for cashew growers 3A detailed description of the construction of the expenditure variable is in Christiaensen and Sarris (2006). Comparison of health expenditures among households with and without an illness shock shows that households who experienced a health shock have on average two to three times more health expenses. Similarly, we find that expenditures on functions are also larger among households who experienced a death over the past two years. Since we can't distinguish between health expenditures for preventive reasons, which may be an expression of larger household welfare, and health expenditures for curative reasons or between expenditures on functions for funerals and baptisms, we opt to exclude all expenditures on health and functions to avoid a downward bias in the estimated coefficient of the welfare effects of health shocks. 7 based on their number of cashew trees and a category for tobacco growers.4 As our data in Ruvuma allowed us only go to two years back these were based on tree ownership in early 2002.5 Table 1 and 2A and B review differences among coffee growers in the different quintiles and non-coffee growers in Kilimanjaro and Ruvuma along a series of (observed) characteristics. Consistent with the agro-ecological requirements of coffee production coffee growing households in Kilimanjaro live at higher altitudes. They are also more likely to be Chagga than Pare in Kilimanjaro and almost exclusively Matengo in Ruvuma, confirming the importance of controlling for ethnicity when exploring the effect of coffee price shocks through a retrospective panel approach. Cashew production is largely confined to the Yao. There is a large difference in the amount of coffee trees owned across the different quintiles with the amount of trees estimated at about 40 in the lowest quintile and more than doubling from quintile to quintile to about 1325 trees in the highest quintile in Kilimanjaro. In Ruvuma, coffee growing households have on average three times as many trees than in Kilimanjaro with those in the highest quintile owning on average 5 times as many trees as those in the lowest quintile. Coffee growers in the lowest quintile in Kilimanjaro tend to own less land, livestock and consumer durables compared with non-coffee growers while coffee growers in the largest quintile tend to have more land, have more valuable housing and receive more remittances compared with non-coffee growers. Further exploration does not show much difference across coffee and non-coffee growers in the likelihood of using one's savings or receiving aid from others when experiencing a shock (see Table 1). Coffee growers in higher quintiles in Ruvuma tend to own more livestock and have more valuable compounds. The larger cashew growers tend to have much more land, though they do not appear richer on other ground compared with other households in Ruvuma. They are however more likely to use savings when faced with shocks. Households' coping capacity is approximated both directly through the inclusion of reported coping through saving or receipt of aid in case of a health and drought shock and indirectly through the value of household consumer durables (per adult equivalent) in the year preceding the survey. We also control for the proportion of time in non-farming activities and the amount of remittances (per adult equivalent) received as further indirect determinants of households' coping capacity. The amount of acres irrigated (per adult equivalent) indicates exposure to drought shocks. Similarly, the proportion of time spent on non-farming activities also indirectly captures exposure to drought shocks. To mitigate potential endogeneity problems arising from the self reporting of drought shocks we use an index of a household's qualitative assessment of the rainfall amount across its plots as opposed to the self-reported occurrence of a drought shock 4Given that only 4 percent of the households in our Ruvuma sample were tobacco growers, we did not disaggregate the group of tobacco growers further. 5The first survey rounds for Kilimanjaro and Ruvuma were conducted in the fall of 2003 and the spring of 2004 respectively. 8 from the directly administered shock module in the questionnaire.6 According to the former measure, 21 percent of all households in Kilimanjaro experienced rainfall much below normal on their plots in 2003 and 42 percent rainfall below normal. Drought shocks are much less frequent in Ruvuma with four percent of all households experiencing rainfall much below normal in 2003 and 34 percent rainfall somewhat below normal. To better capture actual exposure to the rainfall shock, the rainfall shock indicator is multiplied by the household's cultivated land area per adult equivalent. Our health shock variable includes both the occurrence of a death and/or an illness shock in the two years preceding the survey. While death shocks are arguably not infected by self-reporting bias, illness shocks may be. The literature on the accuracy of self-reported health shocks (Foster, 1994; Groot, 2000; Gertler and Gruber, 2002; Baker, Stabile, and Deri, 2004) suggests that the likelihood of reporting a health shock is associated with a household's reference group (the poor tending to report fewer health problems), the intensity of the problem (the more severe the illness, the more likely it is it will be reported, and the need for justification (for example to rationalize absenteeism from work). While the two latter motivations are less of a concern in the current context, the former might bias our results. Inclusion of the comprehensive vector of households' assets and consumer durables described above capturing household wealth will however substantially mitigate the potential bias from self-reporting illness shocks. We also provide robustness tests using non-self village means of self-reported illness incidence as an instrument. Being a coffee or cashew crop grower is treated as exogenous to the household's current living standards. While we use village dummies in our base models to control for unobserved heterogeneity across locations, we also present a model unbundling the village effects. This will allow us to further explore whether our shock variables underestimate the welfare effects of shocks when they cannot fully capture the covariant nature albeit at the expense of potentially introducing endogeneity related to unobserved village effects. We measure in particular how connected a village is, proxy the quality of its infrastructure through the availability of electricity at the village level, and use the altitude at which the village is located to help define its agro-ecological characteristics and thus also its agricultural potential. To capture the connectivity of the village we use information on the presence of a tarmac road in the village, the availability of a public phone and a cell phone signal, the regular organization of a market, and the availability of a bus service to the village. 6In particular, households were asked for each plot whether the rainfall was much below normal, below normal, normal, above normal, much above normal. A plot size weighted average of these rainfall assessments was calculated and rounded off to the nearest digit to obtain a qualitative assessment for each household. 9 4 Determinants of welfare and welfare effects of shocks Given the divergent nature of the economies in Kilimanjaro and Ruvuma, we run separate regressions for both regions. The results are in tables 3 and 4. The baseline model in column (1) includes the shock variables and controls for location effects through village dummies. Models incorporating interaction terms of coping strategies (aid, use of own savings, and remittances) with the different shocks are in column (2). The differential effects of the different health shocks (death and illness of an adult member) are explored in column (3). A model explicitly identifying the location effects through inclusion of village proxies of connectivity, access to electricity and agro- ecological conditions is presented in column (4). The different specifications fit the data very well and explain almost half the variation in the observed (log) expenditures (R- squared between 0.45 and 0.50). The coefficients on the household characteristics and assets are highly significant and largely consistent with predictions from theory. Households with higher dependency ratios tend to be poorer and households with better educated heads enjoy higher consumption. However, the latter effect only holds when the heads have secondary education in Kilimanjaro and only when heads have primary education in Ruvuma, possibly reflecting the fact that Kilimanjaro finds itself further on the path of structural transformation than Ruvuma. Surprisingly, household heads with post secondary education appear disadvantaged in Kilimanjaro though not in Ruvuma, which may reflect the current lack of remunerative employment opportunities for the well educated in Kilimanjaro. Once a household's possession of assets and education are controlled for, female headed households tend to be better off, though the results are only weakly significant. Households with more asset variables (landholdings, livestock ownership, total value of productive assets) tend to be richer. These effects were found to be highly statistically significant and the marginal returns were often observed to be declining as asset possessions increase. Households with easy access to credit for modern inputs were on average estimated to be about ten percent richer in Kilimanjaro underscoring the importance of access to capital and the use of modern inputs. Yet, the opposite appears to hold in Ruvuma, where those with easy access to seasonal credit appear worse off. This result warrants deserves further investigation. Consumption is also positively associated with possession of consumer durables albeit at a declining marginal rate. Access to irrigation substantially enriches households with consumption in Kilimanjaro estimated to increase on average by 19 percent per acre per adult equivalent irrigated. While irrigation also affects consumption positively in Ruvuma, it is not found to be statistically significant. This is likely related to the limited use of irrigation in Ruvuma, consistent with its more reliable rainfall pattern, and thus the limited variability in the sample. Only 2.1 percent of all households in Ruvuma irrigate (some of) their land, while 21 percent do so in Kilimanjaro. Income from remittances 10 positively contributes to consumption both in Kilimanjaro and Ruvuma, though the effect is (again) somewhat less precisely estimated in the latter sample. Also, households with a larger proportion of productive time spent in non-agricultural activities tend to be richer. Farmers in Kilimanjaro who have faced a drought shock over the past year (ie, those who reported receiving much below normal rainfall on their plots) experienced a reduction of consumption of 10 percent per acre cultivated per adult equivalent. While the corresponding reduction in gross total agricultural revenue was estimated to be much more severe at about 50 percent per acre (Sarris, Savastano, and Christiaensen, 2005), households in Kilimanjaro clearly cannot fully protect their consumption from drought shocks. The availability of (cash) savings may help offset the effect of the drought shocks, though its effect is imprecisely estimated (column 2). While access to irrigation is associated with larger overall consumption levels, it does not mitigate the effect of severe rainfall shocks. As most irrigation in Kilimanjaro is gravitation irrigation and when rainfall failure is relatively widespread as in 2003, this does not come as a surprise. More generally, rivers are reported to dry up which reduces their effectiveness in acting as an insurance device. The result should thus be seen in the particular context of Kilimanjaro and not as a statement on the ex ante risk mitigation capacity of irrigation more generally. Our results further suggest that the reception of aid may exacerbate the effect of a drought shock. While it is quite plausible that aid received is not sufficient to offset the negative effect of covariate shocks, the estimated negative relationship seems counterintuitive. It may reflect the fact that those getting aid from neighbors and relatives even in times of a covariate shock are actually the very poorest. About one fifth of all households in Kilimanjaro experienced a drought shock in 2003 with double this number reporting suffering from drought in 2004. In contrast, households who experienced on average somewhat below normal rainfall on their plots did not see their consumption decline. The 35 percent estimated average reduction in households' gross total agricultural revenues associated with somewhat below normal rainfall on their plots (Sarris, Savastano, and Christiaensen, 2005) does not translate into a reduction in households' consumption levels. Households in Kilimanjaro appear able to cope with milder rainfall shocks. Household consumption in Ruvuma appears not to be negatively affected by drought shocks. The effect of the drought shock may however be imprecisely estimated due to the small number of households who experienced a drought shock in 2003 (less than 4 percent of the sample). Somewhat surprisingly those who experienced somewhat below rainfall were even found to be slightly better off, though this result was only statistically significant at the 10 percent level. Somewhat surprisingly those who experienced somewhat below rainfall were even found to be slightly better off, though this result was only statistically significant at the 10 percent level. The results in column (1) of Tables 3 and 4 would suggest that household welfare is unaffected by death and/or illness shocks experienced over the past two years. Yet, 11 when we also control for the household's coping behavior through the self reported use of savings and/or receipt of aid when faced with an illness or death of an adult member (column (2)), health shocks are found to have a strong negative effect on consumption. In particular, households in Kilimanjaro who were unable to cope with the shocks suffered a 16 percent loss in consumption. It furthermore appears that households who used savings (often cash) to cope with health shocks managed to almost completely offset the negative effects associated with the shock. Receipt of aid from others appeared less effective. Finally, and somewhat surprisingly, health shocks appear not to affect household welfare in Ruvuma, even after controlling for households' use of coping strategies. Further decomposition of the health shock into illness and death shocks to explore whether illness and death have differential effects (results in column 3), suggest that households suffer especially from illness shocks, and less so from the death of an adult member. This is in line with the findings from Kagera, in northwest Tanzania by Beegle (2005) who reports that wage employment of adult men declines substantially in response to a future female or male adult death, but that past deaths are not associated with changes in either wage employment or non-farm self-employment. Similarly, she finds that coffee farming is reduced in households with a death within 6 months, but not for deaths after 6 months. Welfare loss from health shocks comes about through 1) increased medical expenditures and 2) foregone opportunities through a loss in labor supply (and thus earnings) and/or a decrease in the return to labor (Gertler and Gruber, 2002). While we do not have directly comparable information on the opportunity cost related to changes in labor supply and returns to labor in both regions, the survey did record expenses related to illness and death shocks on two separate occasions in the questionnaire. First, it asked the household about how many extra expenses (medical and others) it incurred in case of an illness or death shock of one of its members. Second, health expenditures (and expenditures on functions) during the last 30 days were recorded separately as part of the expenditure module. In both cases, expenditures in case of illness and death shocks are substantially larger in Kilimanjaro than in Ruvuma indicating much larger immediate welfare losses in Kilimanjaro than in Ruvuma, in line with the results of the regression analysis (Table 5). Moreover, regular (preventive) health expenditures (i.e. health expenditures when there is no illness shock) in Ruvuma (see second part Table 5) are only about half those in Kilimanjaro (when expressed in per adult equivalence)7. This is consistent with the much lower reported use of health providers in case of illness/injury.8 While this might be because illnesses/injuries are generally less severe in Ruvuma, the larger average distance to a dispensary or health centre in rural Ruvuma (4.5 km) compared with rural Kilimanjaro (2 km) would suggest that lower accessibility of health 7Health expenditures during past 30 days per adult equivalent are not reported in Table 5, but available from authors upon request. 8While about the same proportion of households reported an illness/injury over the past 4 weeks in Kilimanjaro and Ruvuma (23 and 24 percent respectively) during the 2000/01 Household Budget Survey, 74 percent of all households (includes both rural and urban) in Kilimanjaro consulted a health provider, compared with 47 percent of all households in Ruvuma (National Bureau of Statistics, 2002, Table C16). 12 care underpins this difference in health spending.9 In other words, the absence of a significant result on the health shock in Ruvuma should not be taken to mean that there is no welfare loss associated with illness and/or dealth shocks in Ruvuma. The estimated welfare loss from the regression analysis is also consistent with those emerging from the directly reported expenditures related. First, our regression results indicated that the welfare loss is much more pronounced when there is an illness shock. This is largely consistent with the results from the bi-variate analysis in Table 6 which shows that consumption among households who experienced a death shock is sometimes even larger than among those without a death shock. Second, we estimated an average welfare loss of 16 percent associated with a health and in particular an illness shock in Kilimanjaro or an average reduction of 38,800 TSH per adult equivalent given the estimated average consumption per adult equivalent of 242,500 TSH in Kilimanjaro 2003. The directly reported health expenditures during the first survey round in Kilimanjaro in case of an illness are around 37,000 TSH.10 This does not only provide confidence in the reported estimates, but also suggests that the welfare loss is largely due to medical expenses and less due to labor supply effects and income loss. The latter is also consistent with the low marginal productivity of labor (and labor surplus) reported by Sarris, Savastano and Christiaensen (2005) in Kilimanjaro. Labor seems even more abundant in Ruvuma, and when combined with the limited medical expenditures, the absence of a welfare loss in Ruvuma does no longer come as a surprise. Finally, the overall absence of welfare loss in case of the death of an adult member, despite expenditures equivalent to those in case of an illness shock suggests that 1) households don't appear to suffer major labor supply or income losses as mentioned above and especially that 2) households manage to insure themselves from such shocks both through their savings and reliance on traditional/informal insurance schemes such as group based funeral insurance as illustrated in Dercon, et al. (2004).11 Kilimanjaro coffee growers in the lowest quintile category of tree ownership are on average ceteris paribus about 20 percent poorer than rural households not growing coffee, while those in the richest quintile tend to enjoy higher consumption levels. Households in the intermediate quintiles do not appear to differ from non-coffee growers in their consumption levels, though the signs of the coefficients are all negative. As in the case of health shocks, when we include interactions with the amount of remittances received (one of the coping strategies)12, the negative effects are exacerbated, though still not statistically significant for the intermediate quintiles, and coffee growers in the richest 9National Bureau of Statistics, 2002, Table C17. 10To obtain the reported health expenditure in case of an illness shock, we divided the average health expenditures (137,000 TSH from Table 5) by the average adult equivalent per household, i.e. average households size (=5.3)*average adult equivalent per person (=0.7). 11The reported amount of contributions to (other) funerals also suggests substantial solidarity in bearing the funeral costs. Rutherford (2001) has documented the existence of insurance mechanisms for funerals across the developing world and highlights funeral insurance as one of the most popular products offered by more formalized micro-finance institutions. 12Unlike for the health and rainfall shocks, no data has been collected on the particular strategy coffee growers used to cope with the systemic coffee price shock (e.g. use of savings and/or aid). 13 quintile are no longer statistically significantly richer. Given our comprehensive controls for differences in wealth among households at the time of the onset of the coffee price shocks, these result would suggest that while most coffee growers have managed to cope with the coffee price decline, or at least did not see their consumption levels fall below those of the non-coffee growers, for example through the use of remittances and cash savings, the smallest among them experienced a substantial decline in their consumption. Given several years of high prices preceding the collapse in coffee prices starting in 2000, it is indeed plausible that coffee growers largely managed to smooth their consumption, albeit at the expense of their (cash) savings. In sum, while it cannot be excluded that coffee growers' welfare declined, most of them appear not worse off nowadays compared with non-coffee growers, apart from the smaller coffee growers who clearly suffered substantially. Similarly, coffee growers in Ruvuma appear not worse off than non-cash crop growers and the larger ones actually enjoy substantially higher consumption levels despite the decline in coffee prices since 2000.13 Again, given that we control extensively for asset holdings, though not for cash savings, this may reflect the availability of large amounts of cash savings held by the larger coffee growers following windfall earnings from coffee production during the late 1990s. This hypothesis is further supported by the fact that the likelihood of using (cash) savings in case of a drought or health shock is largely unassociated with a household's asset holdings as discussed in section 6. While cashew growers also appear better off than non-cash crop growers, this picture reverses when we replace the village dummies (column 2) by village characteristics (column 4). This follows from the fact that cashew growers live concentrated in one district in Ruvuma and that virtually all households in our sample villages have at least some cashew trees. The overall lower consumption levels among cashew crop growers are thus captured through the village dummies. As there are no reasons to believe that the cashew crop growing villages systematically differ from the non cash crop growing villages beyond the village characteristics included in the analysis, the results in column four suggest that the smaller cashew growers are substantially worse off than the non cash crop growers. This is consistent with the observed collapse in cashew prices since the late 1990s and the fact that the smaller cashew growers are likely to hold less cash savings to help smooth their consumption compared with the larger cashew farmers. Finally, the models with the village dummies unbundled are presented in columns 4 of Tables 3 and 4. Especially noteworthy is the fact that households in villages with a tarmac road are on average about 16 percent richer in Kilimanjaro and about 33 percent richer in Ruvuma. While these effects may partly reflect placement effects, the effects are sufficiently large to underscore the critical importance for overall household welfare of being connected through all weather roads. As indicated above, village dummies may also capture some of the covariant effect of shocks. This is borne out by the slight 13Given the limited number of observations receiving remittances in each of the coffee and cashew quintile categories, we did not interact these with the receipt of remittances. 14 reinforcement of the shock effects observed in the Kilimanjaro estimations when replacing the village dummies with the village characteristics. Yet, for all practical purposes the observed changes are negligible. 5 Poverty effects of shocks and the effectiveness of coping strategies To gauge the overall effects of the shocks and coping on average welfare and poverty in our sample, we perform a series of simulations. As the evidence does not reveal a negative effect of health or drought shocks on household welfare in Ruvuma, we focus on the estimated results for Kilimanjaro. In particular we estimate by how much average consumption and poverty incidence in our sample would have improved in the absence of shocks (and thus also coping) compared with the currently observed situation and by how much it would have deteriorated if there hadn't been any coping in the face of the observed shocks. To do so, we use the village fixed effect model including interaction terms with households' coping strategies (column 2, table 3). Given that our model is loglinear, we can examine the effects of the different shocks and coping strategies on log consumption directly by adding or subtracting the relevant terms ^ ^ 21Sijt +1and 22 Sijt Xijt . We focus on the use of savings and receipt of aid from others +1 when faced with a shock as coping strategies. When coping more than offsets the effect of the shock itself, the positive compensating effect of coping is set equal to the negative effect of the shock. The results of these simulations are presented in Table 7. The gross total loss among Kilimanjaro households in 2003 due to health and drought shocks is estimated at about 11,100 TSH per adult equivalent or about 6 percent of annual consumption on average. Put differently, households who experienced either one or both shocks lost on average 33,369 TSH per adult equivalent gross or about 18 percent of their annual consumption. This amounts to a total gross loss of about 8.43 billion TSH or 8.43 million US$ in 2003 among rural households in Kilimanjaro alone.14 Clearly the gross costs of shocks to the economy can be substantial. As about 12 percent of all rural households in Kilimanjaro experienced an illness or death of an adult member in the two years preceding the survey and almost twice as many households experienced a drought shock in 2003 (Table 8), drought shocks contributed more to the loss (7,000 TSH per adult equivalent) than health shocks (4,100 TSH per adult equivalent), even though the welfare loss associated with a health shock was estimated to be slightly larger than the estimated gross loss from a drought shock.15 Put differently, the total gross loss in personal consumption among rural households in Kilimanjaro attributed to drought is estimated at 5.32 billion TSH, while the loss 14From Table 8, it can be seen that 63,134 households experienced either a health or a drought shock in 2003, corresponding to 252,536 adult equivalents at an estimated average of 4 adult equivalents per household. Given an average loss of 36,707 TSH, this results in a total estimated gross loss of 8,427,000,000 TSH or about 8.43 million US$ at an exchange rate of about 1,000 TSH per US$ in 2003. 15The gross negative effect of the health and drought shock are estimated at 16 and 11 percent respectively (see column 2, Table 3). 15 associated with illness and death of adult household members is estimated at 3.11 billion TSH. Yet, some households managed to (partly) smooth their consumption in the face of these shocks. Consequently, the actual reduction in consumption experienced by the population was smaller than it would have been in the absence of coping. The difference between the observed average consumption in our sample and the average consumption in the absence of any (or the use of other) coping strategies16 provides an estimate of the effectiveness of households' coping strategies. On average about 53 percent of the loss due to health and rainfall shocks was compensated for either through use of one's own savings or reliance on aid from family and neighbors or traditional funeral insurance schemes. This could also be taken as an upper bound estimate of the potential crowding out effect of private insurance, were public insurance to be introduced either through public health or rainfall based insurance. Furthermore, households were better able to cope with health shocks than with rainfall shocks. This follows from the fact that in the former case, which is more idiosyncratic in nature, households could rely on both their own savings as well as aid from others, while in the latter case their coping strategies were confined to use of their own savings only. Finally, assuming the decline in welfare among the small coffee growers could be completely ascribed to the coffee price decline, we estimated that the coffee price decline resulted in a net average loss of about 3,900 TSH per adult equivalent. Given that larger farmers may have used their (unobserved cash) savings to cope with the coffee price decline, this is likely to be an underestimate. To further explore the distributional consequences of the different shocks, we also report the effect of these shocks on poverty incidence. While health and drought shocks would increase poverty by 1.7 percentage points from 15.0 to 16.7 in the absence of coping, private coping strategies (either through self or informal mutual or funeral insurance) substantially mitigate the poverty increasing effects. In the absence of health and drought shocks, rural poverty incidence in Kilimanjaro would have been 0.6 percentage point lower in 2003 (14.4 versus 15.0 percent currently). Furthermore, the average increase in poverty can be equally ascribed to the drought and health shocks and the associated coping strategies. It is simulated that the coffee price increased poverty by at least 0.7 percentage points, though as argued before this may underestimate the actual welfare loss experienced by coffee growers. 6 Correlates of households' ex post coping capacity Rural households in Kilimanjaro and Ruvuma largely rely on self insurance (i.e. use of their own savings) and informal mutual insurance (i.e. receipt of aid from 16In the simulations we focus on the use of savings and aid from others as coping strategies. When coping more than offset the effect of the shock, only the effect of the shock is subtracted from the actual consumption. 16 neighbors and family) to cope with shocks. From tables 3 and 4 we have seen that the use of savings is more efficient in helping farmers cope with rainfall shocks while both savings and aid are used to mitigate the effect of health shocks. To explore who is more likely to be able to cope with shocks either through savings or through aid, we run probit models of having received aid or having used savings in case of a shock on the nature of the shock (drought versus illness or death), its demographic characteristics (educational attainment, gender of head, ethnicity), the household's possession of assets (small and large livestock, land, number of cash crop trees), and a series of village characteristics. This information is critical in targeting social protection interventions. The estimated results for Kilimanjaro and Ruvuma are in tables 9 and 10 respectively. Consistent with the covariate nature of rainfall shocks, households are more likely to use their own savings to cope with droughts, though savings are also used to cope with illness and death shocks. External formal assistance (e.g. food aid or formal social protection interventions) has been rare in our study areas. When faced with a health shock (especially when it concerns the death of an adult member) which is idiosyncratic in nature, a household is more likely to receive aid. Aid appears not responsive to drought shocks. There appears no clear pattern of association between the amount of assets possessed by the household and its use of coping strategies. The Ruvuma results suggest that the more coffee trees a household had two years ago, the higher the likelihood was that it coped either through use of savings and the reception of aid. This is consistent with our earlier finding that coffee growing households in Ruvuma are not worse off than non-cash crop growers despite the decline in coffee price during the early 2000s. We do not find a positive association between the number of coffee trees owned in 2000 and the use of self-insurance or mutual insurance in Kilimanjaro, suggesting that overall their coping capacity is by now no different from the non-coffee grower. Yet, when we include the quintile categories of coffee trees owned (as opposed to the number of coffee trees and its squared term) (results not presented), we find that those in the highest quintile are more likely to use savings (though not aid), consistent with the results in table 3 indicating that this group is still better able to cope and that it might still be better off than the non-coffee growers. Cashew tree growers were not found to be different in their coping capacity than the other non-cash crop growers. While educational attainments do not affect households' coping capacity in Kilimanjaro, in Ruvuma secondary education of the head is associated with a lower probability of receiving assistance, and primary schooling negatively correlated with the use of either coping strategy. Female headed households in Kilimanjaro appear much more likely to receive aid, and much less likely to use savings to cope with shocks. A similar pattern was observed in Ruvuma, though the coefficients were imprecisely estimated. In Kilimanjaro, the probability of receiving aid decreases with the age of the household head up to 36 years, and becomes positively associated with age at 72 years. 17 In Ruvuma, we see a corresponding increase in likelihood of using savings up to the age of 43. The availability of bus service in a village positively affect households' likelihood of using savings in Ruvuma, while electirification and cell phone reception in the village, both indicators of general wealth levels, are positively associated with the use of savings in Kilimanjaro.17 7 Concluding remarks This study has explored the immediate effects of drought and health shocks on welfare and poverty in Kilimanjaro and Ruvuma and reflected on the effect of the coffee and cashew price decline since 2000. About one third of the rural population in Kilimanjaro suffered either from drought or health shocks in the survey year and those households suffered on average a direct 18 percent gross loss in their annual consumption in 2003 as a consequence. Yet, through reliance on savings and aid from others they were able to partly smooth their consumption and reduce the immediate negative welfare effect of these shocks to 8 percent loss on average. The gross joint effect of the health and drought shock in Kilimanjaro on poverty incidence was 2.3 percentage points, while the net effect was estimated at 0.6 percentage point. These seemingly limited poverty effects follow from the low initial level of poverty incidence in Kilimanjaro, estimated at 15 percent. In percentage terms, the health and drought shocks cause poverty incidence to increase by 16 percent before and 4 percent after coping. No immediate (negative) welfare effects were found from the drought and health shocks in Ruvuma. The former result is related to the generally more secure rainfall patterns and the low incidence of drought shocks in Ruvuma in 2003. The lower medical expenditures in case of illness due to limited use of health care providers which is in turn associated with lower access to health facilities, underpins the estimated absence of an immediate welfare loss in Ruvuma. This does not necessarily imply that households in Ruvuma suffer less from illness shocks, but rather that they spend less to deal with them. In addition, the potential income loss either due to reduced labor supply or reduced return to labor following illness or death, appears sufficiently small to not change this picture for Ruvuma. Also in Kilimanjaro, appears the estimated welfare loss largely associated with the medical expenses and not due to substantive income loss. This is consistent with the relative abundance of labor in both Kilimanjaro and even more so in Ruvuma. Sarris, Savastano, and Christiaensen (2005) estimate for example that the ratio of the marginal product of labor in agriculture to the agricultural wage is only 0.22 in Ruvuma (compared to 0.32 in Kilimanjaro). Finally, while the direct reported expenses 17None of the villages in Ruvuma has electricity, and only one village has cell phone reception. 18 related to death shocks are on par with those related to illness shocks, death shocks have much smaller immediate welfare effects, likely related to the existence of effective group based funeral insurance schemes (Dercon et al., 2004). How such schemes evolve as HIV/AIDs puts increasing pressure on these mechanisms must be closely followed. Ceteris paribus, coffee growers in Kilimanjaro appear no worse off than non- coffee growers in Kilimanjaro, apart from the smallest ones, whose consumption level is on average 20 percent lower. Similarly, coffee growers in Ruvuma enjoy ceteris paribus similar consumption levels on average as non-cash crop growers, with indications that the larger ones are actually even better off. This suggests that coffee growers (apart from the smallest) have managed to weather the effects of the coffee price decline, at least to the point of not falling below the welfare levels of the non-cash crop growers and most likely at the expense of a depletion of their (cash) savings. Indeed, the decline in coffee prices since 2000 came on the heels of an income windfall from coffee during the late 1990s. In addition, many coffee growers in Kilimanjaro, who have access to the market in Dar es Salaam, have also been able to switch into bananas as an alternative cash crop. Consistent with this hypothesis is the finding that even in 2003 coffee growers in Ruvuma (as well as the richer coffee growers in Kilimanjaro) tend to be more inclined to use their own savings in case of health or drought shocks compared with non-cash crop growers. Cashew crop growers on the other hand, especially the smaller ones, appear worse off than non-cash crop growers in Ruvuma. Consumption levels among the two bottom quintiles of cashew trees are 15 to 20 percent lower than those among non-cash crop growers. Several years of low cashew prices are beginning to take their toll. While there are few formal insurance or assistance schemes available to help households smooth their consumption, households rely heavily on self insurance through a depletion of their cash savings (and to a lesser extent their assets) as well as informal mutual insurance schemes including group based funeral societies. Aid from others is frequently received in case of death shocks, and to a lesser extent in case of illness, though not in case of a drought shock. Own savings on the other hand are the more important recourse in case of drought shocks, though they are also relied upon to deal with health shocks, especially illness shocks. Somewhat surprisingly, physical asset ownership and educational attainment appear to be poor predictors of the use of savings, pointing to the importance of cash savings in rural Tanzania. Coffee farmers in Ruvuma [as well as the richer coffee farmers in Kilimanjaro) are still more inclined to use their savings to cope with drought or health shocks. Female headed households tend to rely more on aid and less on their own savings. Households in Kilimanjaro in electrified villages and villages with cell phone signals, both signs of wealth, also tend to be more likely to receive aid and use their own savings to cope with shocks. While these coping strategies help households smooth consumption, not all households have equal coping capacity and as documented in the 2002/03 Tanzanian Participatory Poverty Assessment these strategies may come at the expense of future earnings. Moreover, given that this study has abstracted from estimating the long run effects of shocks on welfare, the ex ante behavioral effects (Binswanger and Rosenzweig, 1993; Dercon, 1996; Dercon and Christiaensen, 2005) as well as their effects on human 19 development outcomes (Ainsworth, Beegle, and Koda, 2005), it must be underscored that our results presented only a lower bound on the actual welfare losses associated with health, drought and price shocks. In looking for effective vulnerability reducing interventions, public policies aimed at improving health conditions and reducing the effect of droughts emerge as important, especially in Kilimanjaro. This includes the need for continued efforts to combat the HIV/AIDS epidemic, especially as this expanding epidemic may put the traditional funeral societies under increasing pressure to effectively deal with death shocks, as well as concerted efforts to prevent malaria infections. The ability to control water levels for example through irrigation emerges as an important general instrument to help enhance household consumption even though it has lost its effectiveness as an insurance instrument in Kilimanjaro which largely depends on gravitation irrigation. There are substantial uninsured welfare losses due to drought, suggesting role a for weather based insurance schemes, an innovative approach to protect consumption from drought shocks currently piloted in a series of developing countries (Ethiopia, Morocco, India). Farmers also expressed substantial demand for market based coffee price insurance schemes to help them insure against coffee price declines (Sarris, Karfakis, Christiaensen, 2006). Access to non-agricultural employment and enterprise further helps raising overall welfare levels and reduces exposure to drought shocks. Finally, the importance of connectivity in raising overall income levels and thus also households' ability to cope with shocks cannot be sufficiently underscored. Consumption levels were found to be ceteris paribus 15 to 30 percent higher in villages with a tarmac road compared with those without a tarmac road, though the potential presence of some placement effects in these latter estimates cannot be denied. 20 References Ainsworth, Martha, Kathleen, Beegle, and Godlike, Koda, 2005, The Impact of Adult Mortality and Parental Deaths on Primary Schooling in North-Western Tanzania, Journal of Development Studies, 41-3, pp. 412-439. Alderman, H., J., Hoddinott, and B. Kinsey. "Long Term Consequences of Early Childhood Malnutrition." International Food Policy Research Institute, Washington D.C., 2004. Baker, M., Stabile, M., and C., Deri, 2004, What Do Self-Reported, Objective Measures of Health Measure? Journal of Human Resources, 39-4: pp. 1067-1093. Beegle, Kathleen, 2005, Labor Effects of Adult Mortality in Tanzanian Households, Economic Development and Cultural Change, 53-3: p. 655-683. Besley, Timothy, 1995, Savings, Credit and Insurance, Handbooks in Economics Vol9, Vol 3A: pp. 2123-2207, Elsevier Science, North Holland: Amsterdam. Christiaensen, L., and A., Sarris, 2006, eds., Household Vulnerability in Tanzania ­ An Empirical Perspective, Research on Poverty Alleviation, Dar es Salaam, mimeographed. Datt, Gaurav, and Hans, Hoogeveen, 2003, El Nino or El Peso? Crisis, Poverty and Income Distribution in the Philippines, World Development, 31-7: pp. 1103-1124. Deaton, Angus, 1992, Understanding Consumption, Oxford: Clarendon Press. Deaton, Angus, 1997, The Analysis of Household Surveys: A Microeconometric Approach to Development Policy, John Hopkins University Press: Baltimore, MD. Dercon, Stefan, and Pramila, Krishnan, 2000, Vulnerability, Seasonality and Poverty in Ethiopia, Journal of Development Studies, 36-6: pp. 25-53. Dercon, Stefan, Tessa, Bold, Joachim De Weerdt, and Alula, Pankhurst, 2004, Group Based Funeral Insurance in Ethiopia and Tanzania, CSAE Working Paper 227. Dercon, Stefan, and Luc, Christiaensen, 2005, Consumption Risk, Technology Adoption and Poverty Traps: Evidence from Ethiopia, mimeographed. Dercon, Stefan., John, Hoddinott, and Tassew Woldehanna, 2005, Shocks and Consumption in 15 Ethiopian Villages, 1999-2004, Journal of African Economies, 14-4, pp. XXX Foster, Andrew, (1994), Poverty and Illness in Low-Income Rural Areas, American Economic Review, 84-2: pp. 216-220. 21 Gertler, Paul and Jonathan, Gruber, 2002, , Insuring Consumption Against Illness, American Economic Review, 92-1, pp. 51-70. Groot, W, 2000, Adaptation and Scale of Reference Bias in Self-Assessments of Quality of Life, Journal of Health Economics, 19, p. 403-420. Hall, R., 1978, Stochastic Implications of the Life Cycle-Permanent Income Hypothesis: Theory and Evidence, Journal of Political Economy, 86-6: pp. 971-87. Harrower, Sarah and John, Hoddinott, 2005, Consumption Smoothing in the Zone Lacustre, Mali, Journal of African Economies, 14-4, pp. 489-519. Morduch, Jonathan, 1991, Risk and Welfare in Developing Countries, Harvard University, Ph.D. Thesis. National Bureau of Statistics Tanzania, 2002, Household Budget Survey, 2000/01, Final Report, National Bureau of Statistics, Tanzania: Dar es Salaam. Research and Analysis Working Group, 2004, Vulnerability and Resilience to Poverty in Tanzania: Causes, Consequences and Policy Implications ­ 2002/03 Tanzania Participatory Poverty Assessment (TZPPA): Main Report, Research and Analysis Working Group: Dar es Salaam, Tanzania. Rutherford, S., 2001, The Poor and Their Money, Delhi: Oxford University Press for India. Sarris, Alexander, Sara, Savastano, and Luc Christiaensen, 2005, The Role of Agriculture in Reducing Poverty in Tanzania: A Household Perspective from Rural Kilimanjaro and Ruvuma, World Bank, mimeographed. Sarris, Alexander, Panayiotis, Karfakis and Luc Christiaensen, 2006, Producer Demand and Welfare Benefits of Minimum Price Insurance for Export Crops in Tanzania, mimeographed. 22 Table 1: Comparison of socio-economic characteristics and past coping behavior among quintile categories of coffee and non-coffee growers in Kilimanjaro. no trees lowest 2nd 3rd 4th highest total quintile1) quintile quintile quintile quintile Altitude (meters) 3578 4575** 4548** 4694** 4546** 4451** 4177 % Pare 36.2 9.4** 2.4** 8.7** 6.7** 3.4** 17.9 % Chagga 44.8 90.6** 9.7** 90.3** 92.4** 96.6** 74.3 Land (2000) 2.9 1.6** 2.2+ 2.4 2.9 3.6+ 2.7 Goats/sheep/pigs (2000) 2.8 1.1+ 1.3 1.6 1.9 1.5 2.0 Cattle/horses/oxen (2000) 4.4 2.3 3.0 3.5 3.1 2.2 3.5 Value Consumer Durables 264 140** 248 233 234 315 246 1000 TSH, (2002) Value of compound (2003) 2342 2194 4051*** 3243** 4943***+ 7851*** 3681 Remittance income (1000 24.4 25.4 38.9 42.4** 62.1** 44.1** 35.4 TSH) (2003) Coffee trees (2000) 0 39*** 109*** 235*** 502*** 1326*** 269 Of those who faced a 47.2 44.4 43.6 63.1** 55.7 60.1 50.9 shock, % who received aid Of those who faced a 67.4 78.3 69.1 81.4 68.0 67.4 78.3 shock, % who used own savings Number of households 74,593 23,287 23,034 23,301 22,974 23,123 190,312 1)based on number of coffee trees owned in 2000 2)**denotes significance at 1%, * at 5%, at 10% when comparing characteristics to the non- + coffee growers. 23 Table 2A: Comparison of socio-economic characteristics among quintile categories of coffee growers and non-coffee growers in Ruvuma. no trees lowest 2nd 3rd 4th highest total quintile1) quintile quintile quintile quintile Altitude (meters) 2658 4362*** 4531*** 4660*** 4682*** 4833*** 3275 % Matengo 16.6 91.2*** 96.5*** 96.6*** 100.0*** 95.5*** 41.6 % Ndendeule 8.8 1.8 1.9 0.0* 0.0* 2.1 6.4 % Ngoni 20.3 5.1* 0.0*** 1.6** 0.0*** 2.4** 14.5 % Yao 32.5 0.0*** 0.0*** 0.0*** 0.0*** 0.0*** 22.3 % Nyasa 0.9 0.0 0.0 0.0 0.0 0.0 0.6 Land (2002) 9.76 9.76 9.29 8.94 8.63 9.26 9.57 Goats/sheep/pigs (2002) 2.87 4.14** 5.46*** 5.24*** 4.33** 4.30** 3.44 Cattle/horses/oxen 0.16 0.49*** 1.04*** 0.91*** 0.75*** 1.39*** 0.40 (2003) Value Consumer 139.4 109.9 137.4 141.5 125.7 192.4 139.9 Durables 1000 TSH, (2002) Value of compound 433.3 399.4 931.1*** 1196.8*** 1168.3*** 1024.6*** 596.0 (2003) Remittance income 169.8 45.1 130.9 94.9 41.0 116.7 144.1 (1000 TSH) (2003) Coffee trees (2000) 0.0 432.5*** 847.7*** 1304.4*** 1592.4*** 2084.5*** 393.9 Of those who faced a 45.2 43.3 41.7 59.1 56.1 58.5 47.0 shock, % who received aid Of those who faced a 69.7 66.4 70.8 74.8 83.5 82.2 71.3 shock, % who used own savings Number of households 119,022 11,089 10,899 11,159 11,024 10,728 173,921 1)based on number of coffee trees owned in 2002 2)**denotes significance at 1%, * at 5%, at 10% when comparing characteristics to the non- + coffee growers. 24 Table 2B: Comparison of socio-economic characteristics among quintile categories of cashew growers and non-cashew growers in Ruvuma. no trees lowest 2nd 3rd 4th highest total quintile1) quintile quintile quintile quintile Altitude (meters) 3684 1997*** 1864*** 1991*** 1994*** 2054*** 3275 % Matengo 53.9 13.3*** 0.0*** 0.0*** 0.0*** 0.0*** 41.6 % Ndendeule 7.5 7.5 3.0 0.0* 1.6* 1.9 6.4 % Ngoni 16.9 5.8* 12.4*** 10.1** 4.3*** 1.7** 14.5 % Yao 9.6 36.5*** 56.1*** 62.6*** 71.2*** 86.3*** 22.3 % Nyasa 0.5 0.0 1.8 0.0 1.7 1.6 0.6 Land (2002) 8.87 7.45* 7.48 9.83 13.74*** 20.60*** 9.57 Goats/sheep/pigs (2002) 3.81 2.31** 1.47*** 2.17*** 3.32** 2.11** 3.44 Cattle/horses/oxen (2003) 0.47 0.57*** 0.00*** 0.00*** 0.18*** 0.00*** 0.40 Value Consumer Durables 147.8 139.7 72.4** 111.9 125.9 124.4 139.9 1000 TSH, (2002) Value of compound (2003) 668.9 212.4 356.1*** 317.9*** 426.1*** 539.9*** 596.0 Remittance income (1000 158.6 142.0 27.2 57.6 46.5 83.3 144.1 TSH) (2003) Cashew trees (2000) 0.0 32.9*** 92.4*** 157.7*** 336.1*** 905.9*** 72.9 Of those who faced a 45.4 50.6 54.3 38.7 60.6 56.9 47.0 shock, % who received aid Of those who faced a 69.3 78.8 53.5*** 87.9** 71.4 93.6** 71.3 shock, % who used own savings Number of households 132,195 8,440 8,307 8,328 8,362 8,289 173,921 1)based on number of cashew trees owned in 2002 2)**denotes significance at 1%, * at 5%, at 10% when comparing characteristics to the non- + coffee growers. 25 Table 3: Shocks, coping and consumption in Kilimanjaro Log consumption per adult equivalent (exclusive of health and education baseline Shocks interacted with Health shocks Village dummies expenditures, and expenditures on functions) coping strategies unbundled unbundled (1) (2) (3) (4) Shocks, exposure and coping major illness or death of adult member -0.020 -0.161 -0.205 (0.47) (1.82)+ (2.29)* used savings to cope with major illness or death of adult member 0.148 0.202 (1.55) (2.19)* received aid to cope with major illness or death of adult member 0.070 0.074 (0.87) (0.93) death of adult member last 2 yrs -0.150 (0.78) death of adult member last 2 yrs * received aid 0.025 (0.13) death of adult member last 2 yrs * used savings 0.271 (1.99)* major illness of adult member last 2 yrs -0.170 (1.63) ill adult member last 2 yrs * received aid 0.068 (0.70) ill adult member last 2 yrs * used savings 0.101 (0.90) acres/ae * very low rainfall -0.104 -0.112 -0.108 -0.116 (2.82)** (3.04)** (2.94)** (3.24)** acres/ae * very low rainfall * got aid for drought -0.243 -0.245 -0.253 (2.02)* (2.05)* (2.07)* acres/ae * very low rainfall * used savings for drought 0.131 0.131 0.137 (1.21) (1.20) (1.25) acres/ae * somewhat low rainfall 0.044 0.025 0.027 0.029 (1.19) (0.72) (0.76) (0.87) acres/ae * somewhat low rainfall * got aid for drought -0.214 -0.195 -0.136 (1.35) (1.18) (0.87) acres/ae * somewhat low rainfall * used savings for drought 0.166 0.165 0.158 (2.93)** (2.97)** (2.70)** lowest quintile coffee trees 2000 -0.205 -0.217 -0.210 -0.233 (3.45)** (3.59)** (3.47)** (3.85)** lowest quintile coffee trees 2000 * remittance income 100,000 TSH/ae -0.119 -0.124 -0.051 (1.18) (1.23) (0.53) second quintile coffee trees 2000 -0.065 -0.092 -0.093 -0.085 (1.14) (1.60) (1.61) (1.58) second quintile coffee trees 2000 * remittance income 100,000 TSH/ae 0.150 0.148 0.163 (1.08) (1.07) (1.22) third quintile coffee trees 2000 -0.043 -0.065 -0.062 -0.071 (0.72) (1.03) (1.00) (1.27) third quintile coffee trees 2000 * remittance income 100,000 TSH/ae 0.150 0.147 0.207 (1.18) (1.15) (1.75)+ fourth quintile coffee trees 2000 -0.022 -0.051 -0.051 -0.044 (0.38) (0.86) (0.86) (0.85) fourth quintile coffee trees 2000 * remittance income 100,000 TSH/ae 0.179 0.172 0.227 (1.97)* (1.89)+ (2.68)** highest quintile coffee trees 2000 0.145 0.114 0.118 0.156 (2.10)* (1.56) (1.63) (2.48)* highest quintile coffee trees 2000 * remittance income 100,000 TSH/ae 0.155 0.155 0.111 (1.15) (1.15) (0.92) irrigated acres/ae * very low rainfall 0.039 0.060 0.053 0.099 (0.42) (0.66) (0.58) (1.10) irrigated acres/ae * somewhat low rainfall -0.265 -0.241 -0.245 -0.234 (3.10)** (2.92)** (2.99)** (2.90)** irrigated acres cultivated 2003 per ae 0.188 0.188 0.188 0.195 (2.89)** (2.96)** (3.07)** (3.36)** remittance income, 100,000 TSH/ae 0.149 0.060 0.065 0.030 (2.81)** (0.68) (0.74) (0.37) 26 Log consumption per adult equivalent (exclusive of health and education baseline Shocks interacted with Health shocks Village dummies expenditures, and expenditures on functions) coping strategies unbundled unbundled (1) (2) (3) (4) Demographic characteristics dependency ratio -0.186 -0.181 -0.180 -0.178 (3.00)** (2.89)** (2.88)** (2.89)** age of head -0.028 -0.027 -0.027 -0.027 (4.07)** (4.00)** (4.02)** (4.14)** age of head squared 0.000 0.000 0.000 0.000 (3.83)** (3.74)** (3.77)** (3.98)** female-headed household 0.068 0.063 0.068 0.089 (1.51) (1.40) (1.53) (1.97)* yrs primary education of head 0.006 0.005 0.006 0.008 (0.89) (0.74) (0.83) (1.21) yrs secondary education of head 0.034 0.033 0.033 0.034 (1.68)+ (1.65)+ (1.63) (1.66)+ whether head has post-sec education -0.206 -0.222 -0.219 -0.238 (1.80)+ (1.95)+ (1.93)+ (2.09)* head is Chagga 0.149 0.158 0.152 0.132 (2.41)* (2.52)* (2.43)* (2.40)* head is Pare 0.125 0.125 0.112 0.036 (1.82)+ (1.81)+ (1.64) (0.59) proportion of time in non-agricultural activities in 2002 0.185 0.203 0.205 0.212 (2.80)** (3.05)** (3.09)** (3.12)** Productive assets and consumer durables land owned 3 years ago/ae 0.094 0.095 0.094 0.072 (2.62)** (2.64)** (2.62)** (2.21)* land owned 3 years ago/ae sqr -0.000 0.000 0.000 0.002 (0.04) (0.06) (0.03) (0.94) value of productive assets in 2002, 100,000 TSH per ae 0.043 0.042 0.042 0.040 (3.17)** (3.04)** (3.03)** (2.35)* value of productive assets in 2002 squared, 100,000 TSH -0.000 -0.000 -0.000 -0.000 (3.37)** (3.24)** (3.23)** (2.53)* relatively easy to obtain seasonal credit for inputs 0.114 0.119 0.119 0.128 (2.50)* (2.54)* (2.53)* (2.71)** head of cattle, oxen, horses 3 years ago / ae 0.088 0.091 0.091 0.105 (4.54)** (4.70)** (4.79)** (5.49)** head of cattle, oxen, horses 3 years ago / ae sqr -0.001 -0.001 -0.001 -0.001 (2.58)* (2.80)** (2.85)** (3.59)** head of goat, sheep, pigs 3 years ago / ae 0.031 0.032 0.032 0.024 (2.39)* (2.48)* (2.58)** (1.93)+ head of goat, sheep, pigs 3 years ago / ae sqr -0.001 -0.001 -0.001 -0.001 (2.21)* (2.34)* (2.41)* (1.97)* value of consumer durables in 2002, 100,000 TSH per ae 0.304 0.297 0.297 0.311 (8.49)** (8.53)** (8.57)** (9.35)** value of consumer durables in 2002 squared, 100,000 TSH -0.027 -0.024 -0.024 -0.025 (4.96)** (4.72)** (4.72)** (5.01)** Village connectivity, infrastructure and agro-ecological potential tarmac road reaches village 0.161 (2.36)* village has public phone 0.036 (0.97) village has cell phone signal 0.024 (0.35) bus service to village 0.010 (0.25) village has a market 0.040 (1.13) village has electricity 0.102 (2.14)* village has health center, dispensary, or hospital -0.084 (0.89) Altitude of village, 1000 m 0.200 (0.09) Constant 5.268 5.268 5.260 5.136 (22.61)** (22.45)** (22.53)** (24.12)** Observations 914 914 914 914 R-squared 0.49 0.50 0.50 0.47 Models (1)-(3) include village dummies which are not presented to save space. Absolute value of t statistics in parentheses; + significant at 10%; * significant at 5%; ** significant at 1% 27 Table 4: Shocks, coping and consumption in Ruvuma Log consumption per adult equivalent (exclusive of health and education baseline Shocks interacted Health shocks Village dummies expenditures, and expenditures on functions) with coping unbundled unbundled strategies (1) (2) (3) (4) Shocks, exposure and coping major illness or death of adult member -0.005 0.030 0.067 (0.11) (0.42) (0.99) used savings to cope with major illness or death of adult member -0.083 -0.074 (1.06) (0.94) received aid to cope with major illness or death of adult member -0.004 -0.024 (0.05) (0.29) death of adult member last 2 yrs 0.075 (0.93) death of adult member last 2 yrs * received aid -0.414 (2.72)** death of adult member last 2 yrs * used savings 0.164 (1.15) major illness of adult member last 2 yrs 0.003 (0.04) ill adult member last 2 yrs * received aid 0.057 (0.62) ill adult member last 2 yrs * used savings -0.021 (0.20) acres/ae * very low rainfall -0.018 -0.020 -0.019 -0.017 (0.34) (0.36) (0.36) (0.28) acres/ae * very low rainfall * got aid for drought -0.078 -0.078 0.014 (0.69) (0.69) (0.19) acres/ae * somewhat low rainfall 0.031 0.030 0.030 0.036 (1.83)+ (1.73)+ (1.74)+ (2.11)* acres/ae * somewhat low rainfall * got aid for drought -0.326 -0.325 -0.328 (1.52) (1.51) (1.61) acres/ae * somewhat low rainfall * used savings for drought 0.004 0.006 -0.010 (0.09) (0.11) (0.17) lowest quintile coffee trees 2002 0.134 0.138 0.131 0.071 (1.49) (1.54) (1.46) (0.83) second quintile coffee trees 2002 0.156 0.156 0.158 0.066 (1.81)+ (1.81)+ (1.86)+ (0.80) third quintile coffee trees 2002 0.079 0.083 0.075 0.003 (0.94) (0.97) (0.89) (0.04) fourth quintile coffee trees 2002 0.336 0.338 0.345 0.243 (3.87)** (3.86)** (3.94)** (2.85)** highest quintile coffee trees 2002 0.290 0.289 0.291 0.199 (3.21)** (3.17)** (3.22)** (2.16)* lowest quintile cashew trees 2002 0.066 0.068 0.065 -0.148 (0.79) (0.81) (0.77) (2.01)* second quintile cashew trees 2002 0.103 0.107 0.110 -0.234 (0.99) (1.02) (1.05) (3.12)** third quintile cashew trees 2002 0.312 0.312 0.304 -0.034 (2.67)** (2.67)** (2.60)** (0.39) fourth quintile cashew trees 2002 0.312 0.326 0.316 -0.042 (2.76)** (2.87)** (2.79)** (0.52) highest quintile cashew trees 2002 0.394 0.401 0.393 0.025 (3.27)** (3.32)** (3.24)** (0.29) irrigated acres/ae * somewhat low rainfall 0.039 0.036 0.052 0.037 (0.16) (0.15) (0.22) (0.16) irrigated acres cultivated 2003 per ae 0.142 0.141 0.137 0.147 (1.03) (1.03) (1.00) (1.15) cultivated tobacco in 2004 -0.160 -0.156 -0.150 -0.091 (1.46) (1.42) (1.37) (0.84) remittance income, 100,000 TSH/ae 0.184 0.183 0.186 0.184 (1.45) (1.44) (1.49) (1.35) Demographic characteristics dependency ratio -0.196 -0.195 -0.189 -0.162 (2.44)* (2.41)* (2.34)* (1.94)+ age of head -0.046 -0.046 -0.046 -0.043 (5.66)** (5.67)** (5.64)** (5.15)** age of head squared 0.000 0.000 0.000 0.000 (4.88)** (4.89)** (4.83)** (4.45)** 28 Log consumption per adult equivalent (exclusive of health and education baseline Shocks interacted Health shocks Village dummies expenditures, and expenditures on functions) with coping unbundled unbundled strategies (1) (2) (3) (4) female headed household 0.105 0.107 0.110 0.101 (1.75)+ (1.77)+ (1.83)+ (1.64) yrs primary completed by head 0.024 0.024 0.024 0.024 (2.70)** (2.68)** (2.72)** (2.61)** yrs secondary completed by head 0.015 0.013 0.015 0.007 (0.61) (0.55) (0.60) (0.29) head has post-sec education 0.209 0.207 0.182 0.261 (1.17) (1.17) (1.06) (1.56) head is Matengo -0.063 -0.061 -0.057 -0.004 (0.71) (0.68) (0.64) (0.06) head is Ndendeule -0.009 -0.011 -0.009 0.116 (0.09) (0.10) (0.09) (1.27) head is ngoni -0.132 -0.134 -0.137 -0.025 (1.64) (1.65)+ (1.70)+ (0.32) head is yao -0.062 -0.065 -0.065 -0.056 (0.78) (0.80) (0.79) (0.85) head is nyasa 0.010 0.007 0.019 0.024 (0.07) (0.05) (0.13) (0.17) proportion of time in non-agricultural activities in 2003 0.218 0.212 0.217 0.286 (2.39)* (2.33)* (2.39)* (3.05)** Productive assets and consumer durables land owned 1 year ago/ae 0.035 0.035 0.035 0.042 (3.24)** (3.19)** (3.27)** (3.87)** land owned 1 year ago/ae sqr -0.001 -0.001 -0.001 -0.001 (2.56)* (2.52)* (2.57)* (2.85)** value of productive assets in 2003, 100,000 TSH per ae 0.047 0.046 0.046 0.039 (1.76)+ (1.73)+ (1.76)+ (1.47) value of productive assets in 2003 squared, 100,000 TSH -0.001 -0.001 -0.001 -0.001 (2.12)* (2.10)* (2.10)* (1.72)+ relatively easy to obtain seasonal credit for inputs -0.070 -0.068 -0.072 -0.072 (1.76)+ (1.72)+ (1.79)+ (1.84)+ head of cattle, oxen, horses one year ago per ae 0.389 0.385 0.401 0.353 (4.43)** (4.36)** (4.55)** (3.98)** head of cattle, oxen, horses, one year ago squared per ae -0.146 -0.143 -0.150 -0.146 (3.68)** (3.54)** (3.88)** (3.98)** head of goat, sheep, one year ago per ae 0.080 0.082 0.081 0.101 (2.88)** (2.91)** (2.91)** (3.58)** head of goat, sheep, one year ago squared per ae -0.007 -0.007 -0.007 -0.009 (1.53) (1.61) (1.57) (1.94)+ value of consumer durables in 2003, 100,000 TSH per ae 0.470 0.470 0.465 0.466 (5.91)** (5.92)** (5.89)** (5.66)** value of consumer durables in 2003, 100,000 TSH per ae, sqr -0.004 -0.004 -0.004 -0.004 (4.76)** (4.72)** (4.65)** (4.70)** Village connectivity, infrastructure and agro-ecological potential tarmac road reaches village 0.331 (3.02)** village has cell phone signal -0.059 (0.75) village has a market -0.073 (1.95)+ bus service to village 0.035 (0.72) Health facility in village 0.046 (1.28) altitude 0.154 (0.06) Constant 5.235 5.491 5.477 5.446 (21.79)** (23.75)** (23.61)** (24.61)** Observations 878 878 878 878 R-squared 0.47 0.47 0.47 0.42 Models (1)-(3) include village dummies which are not presented to save space. Absolute value of t statistics in parentheses; + significant at 10%; * significant at 5%; ** significant at 1%; rainfall very low * acres cultivated/ae * used savings to cope with drought, rainfall very low * irrigated acres/ae, village electricity, village public phone, are all dropped due to collinearity 29 Table 5: Expenses incurred as result of an illness or dealth shock, 2002-2004 Average expenses (`000 TSH) incurred Illness of adult member Death of adult member per household in case of an illness or (15-64 yrs old) (15-64 yrs old) death shock over the past 5 years Kilimanjaro - Round 1 a 137 143 - Round 2 102 108 Ruvuma - Round 1 a 38 94 - Round 2 49 51 Health expenditures (`000 TSH) per Illness shock Death shock household during 30 days preceding survey No Yes No Yes Kilimanjaro - Round 1 a 33 68 34 55 - Round 2 35 117 42 40 Ruvuma - Round 1 a 21 40 23 28 - Round 2 19 55 23 12 a) While the reported expenses in case of a shock in round 1 are averaged across the 5 years preceding the survey given a shock, those in round 2 only to the year preceding the survey. Source: Authors' calculations. 30 Table 6: Household welfare with and without illness or death shocks. Expenditures Illness or # obs illness # obs death # obs per adult death equivalent1) Kilimanjaro Round 1 no shock 177.0 832 178.2 881 175.4 893 shock 171.5 115 152.5 66 193.8 54 difference 5.6 25.6 -18.3 Total 176.4 947 176.4 947 176.4 947 Round 2 no shock 167.5 806 167.4 836 164.6 883 shock 139.3 109 129.6 79 160.4 32 difference 28.2 37.9 4.2 Total 164.5 915 164.5 915 164.5 915 Ruvuma Round 1 no shock 152.2 820 153.8 843 151.6 865 shock 158.3 69 144.4 46 166.8 24 difference -6.1 9.4 -15.1 Total 152.9 889 152.9 889 152.9 889 Round 2 no shock 148.5 723 148.2 751 147.8 810 shock 142.7 115 143.3 87 138.1 28 difference 5.7 4.9 9.7 Total 147.6 838 147.6 838 147.6 838 1)Expenditures exclude expenditures on education, health and functions and have been deflated for comparison with HBS expenditures. 31 Table 7: Welfare and Poverty effect of Shocks and Coping in Kilimanjaro1) health & health rainfall coffee rainfall only only shock only Consumption per adult equivalent (`000TSH) no shock, no coping 197.0 192.9 195.8 195.7 shock and coping (=actual) 191.8 191.8 191.8 191.8 shock, no coping 185.9 189.0 188.6 191.8 Poverty incidence (%) no shock, no coping 14.4 14.8 14.5 14.3 shock and coping (=actual) 15.0 15.0 15.0 15.0 shock, no coping 16.7 16.0 15.8 15.0 1)The simulations were performed using the village fixed effect model including interaction terms with households' coping strategies (column 2, Table 3). 32 Table 8: Incidence of rainfall and health shocks in Kilimanjaro and Ruvuma in 2002- 2004 Kilimanjaro Ruvuma percent of number of percent of number of households households households households Adult health shock last 2 years 12.2 23,336 11.9 20,706 Adult illness shock last 2 years 6.9 13,172 8.1 14,105 Adult death shock last 2 years 5.8 11,194 4.0 7,035 Very low rainfall this year 20.8 39,798 3.8 6,547 Somewhat low rainfall this year 41.9 80,234 33.8 58,822 Either very low rainfall or adult 33.0 63,134 15.7 27.253 health shock 33 Table 9: Correlates of use of savings, aid and remittances in case of a shock in Kilimanjaro (1) (2) (3) received aid used savings received aid or used savings Shocks shock was any death 1.896 0.559 1.891 (8.89)** (3.05)** (6.01)** shock was any illness 0.894 0.758 1.057 (4.47)** (3.77)** (4.85)** shock was drought -0.053 0.852 0.741 (0.29) (4.98)** (4.33)** Productive assets head of cattle, oxen, horses one year ago per ae 0.365 -0.196 0.185 (1.89)+ (1.35) (0.99) head of cattle, oxen, horses, one year ago squared per ae -0.057 -0.000 -0.047 (1.63) (0.01) (1.53) head of goat, sheep, one year ago per ae -0.134 0.022 -0.033 (2.24)* (0.41) (0.55) head of goat, sheep, one year ago squared per ae 0.005 0.001 0.002 (2.27)* (0.80) (1.00) land owned 3 years ago/ae 0.037 -0.119 0.024 (0.26) (0.91) (0.17) land owned 3 years ago/ae sqr 0.001 0.009 0.000 (0.14) (1.01) (0.03) coffee trees owned in 2000, hundreds per ae 0.013 0.115 0.054 (0.16) (1.05) (0.52) coffee trees owned in 2000 per ae squared, hundreds -0.003 0.002 0.002 (0.70) (0.24) (0.45) Demographics dependency ratio -0.205 -0.020 0.071 (0.68) (0.07) (0.21) female-headed household 0.611 -0.432 0.120 (3.00)** (2.22)* (0.54) age of head -0.072 0.028 -0.016 (2.45)* (1.19) (0.63) age of head squared 0.001 -0.000 0.000 (2.89)** (1.35) (0.56) yrs primary education of head 0.021 -0.006 -0.026 (0.60) (0.20) (0.72) yrs secondary education of head 0.058 0.021 0.022 (0.56) (0.23) (0.22) whether head has post-sec education -0.424 0.170 -0.089 (0.94) (0.37) (0.19) head is Chagga -0.505 -0.385 -0.279 (1.90)+ (1.36) (0.91) head is Pare -0.195 -0.355 -0.167 (0.67) (1.17) (0.52) Village connectivity, infrastructure and agro-ecological potential tarmac road reaches village 0.045 0.070 0.183 (0.17) (0.25) (0.59) village has public phone -0.201 -0.248 -0.197 (0.97) (1.29) (0.92) village has cell phone signal 0.251 0.728 0.910 (0.68) (2.09)* (2.67)** village has a market -0.113 -0.158 -0.150 (0.60) (0.87) (0.73) village has electricty 0.294 0.520 0.631 (1.21) (2.11)* (2.30)* 34 (1) (2) (3) received aid used savings received aid or used savings bus service to village 0.076 0.192 0.046 (0.38) (1.02) (0.22) village has bank or other formal credit inst. -0.336 -0.075 -0.181 (1.52) (0.31) (0.70) altitude 0.003 -0.002 -0.002 (0.27) (0.24) (0.16) Constant 0.500 -1.273 -0.318 (0.49) (1.44) (0.34) Observations 484 484 484 F stat 5.41 2.43 3.12 Prob > F 0.000 0.000 0.000 Pseudo R-squared 1) 0.2873 . 1230 0.2249 Absolute value of t statistics in parentheses; + significant at 10%; * significant at 5%; ** significant at 1%; results presented allow for different correlation structures within districts, except for pseudo R-squared statistics, which are taken from a model which does not. 35 Table 10: Correlates of use of savings, aid and remittances in case of a shock in Ruvuma (1) (2) (3) receipt of aid use of savings receipt of aid or use of savings Shocks shock was death 1.769 0.553 1.321 (5.37)** (1.88) + (3.30)** shock was illness 0.129 0.773 1.424 (0.50) (2.58)* (3.23)** shock was drought -0.377 0.902 0.657 (0.88) (1.82) + (1.31) Productive assets head of cattle, oxen, horses one year ago per ae -0.676 -0.560 -3.012 (0.94) (0.80) (2.30)* head of cattle, oxen, horses, one year ago squared per ae 0.485 0.036 1.333 (1.67) + (0.13) (1.77) + head of goat, sheep, one year ago per ae 0.374 -0.535 -0.322 (1.53) (2.16)* (1.10) head of goat, sheep, one year ago squared per ae -0.095 0.075 0.014 (1.78) + (1.48) (0.26) land owned 1 year ago/ae 0.041 0.141 0.196 (0.36) (1.27) (1.44) land owned 1 year ago/ae sqr -0.000 -0.006 -0.008 (0.04) (0.82) (0.99) coffee trees owned in 2002, hundreds per ae 0.331 0.339 0.902 (1.29) (1.29) (2.81)** coffee trees owned in 2002 per ae squared, hundreds -0.056 -0.034 -0.146 (1.08) (0.73) (2.63)** hundreds of cashew trees owned in 2002 per ae 0.169 -0.770 -1.007 (0.55) (0.78) (0.93) hundreds of cashew trees owned in 2002 per ae, squared -0.027 0.663 0.573 (0.72) (1.31) (1.20) whether produced tobacco this year 0.577 -0.168 -0.272 (0.77) (0.24) (0.35) Demographics dependency ratio -0.279 0.442 -0.217 (0.59) (0.87) (0.39) head is female 0.466 -0.578 0.050 (1.13) (1.37) (0.11) age of head -0.038 0.086 0.108 (0.68) (1.71) + (1.82) + age of head squared 0.000 -0.001 -0.001 (0.53) (1.99)* (2.13)* yrs primary completed by head 0.038 -0.031 -0.145 (0.72) (0.54) (2.10)* yrs secondary completed by head2) -0.411 0.113 0.235 (2.30)* (0.82) (1.63) head is Matengo -0.178 0.144 -0.285 (0.45) (0.37) (0.64) head is Ndendeule -0.772 -0.379 -0.311 (1.47) (0.76) (0.60) head is Ngoni 0.265 -0.110 -0.088 (0.59) (0.26) (0.19) head is Yao 0.429 0.297 0.648 (1.16) (0.76) (1.38) head is Nyasa -0.394 (0.57) 36 (1) (2) (3) receipt of aid use of savings receipt of aid or use of savings Village connectivity, infrastructure and agro-ecological potential tarmac road reaches village -0.134 0.647 0.286 (0.17) (0.87) (0.39) village has cell phone signal -0.470 (0.89) village has a market 0.302 -0.107 -0.099 (1.14) (0.39) (0.30) bus service to village -0.007 0.709 0.689 (0.03) (2.16)* (1.96) + village has bank or other formal credit inst. -0.020 -0.187 -0.577 (0.05) (0.44) (1.12) Constant -1.087 -1.366 -0.373 (0.53) (0.69) (0.16) Observations 202 195 195 F stat 1.63 1.43 1.53 Prob > F 0.0289 0.0865 0.0550 Pseudo R-squared 0.2025 0.1686 0.2847 Absolute value of t statistics in parentheses; + significant at 10%; * significant at 5%; ** significant at 1%; rainfall very low * acres cultivated/ae * used savings to cope with drought, rainfall very low * irrigated acres/ae, village electricity, village public phone, are all dropped due to collinearity; results presented allow for different correlation structures within districts, except for pseudo R-squared statistics, which are taken from a model which does not.; post-secondary education of head predicts use of savings and no receipt of aid perfectly; differing number of observations between regressions is due to the fact that observations are dropped when a variable is perfectly collinear with the dependent variable. Reducing Child Malnutrition in Tanzania: Combined Effects of Income Growth and Program Interventions Harold Alderman The World Bank Hans Hoogeveen The World Bank Mariacristina Rossi The World Bank and University of Rome "Tor Vergata", Italy February 2005 Abstract Malnutrition is associated with an inadequate diet, poor health and sanitation services and insufficient care for young children. A combination of income growth and nutrition interventions are therefore suggested to adequately tackle this issue (Haddad et al. 2003), yet evidence to support this claim is often not available, especially for African settings. This paper evaluates the joint contribution of income growth and nutrition interventions towards the reduction of malnutrition. Using a four round panel data set from northwestern Tanzania we estimate the determinants of a child's nutritional status, including household income and the presence of nutrition interventions in the community. The results show that better nutrition is associated with higher income, and that nutrition interventions have a substantial beneficial effect. Policy simulations make clear that if one intends to halve malnutrition rates by 2015 (the MDG objective), income growth will have to be complemented by large scale program interventions. Keywords: nutrition, program evaluation, income growth, Tanzania, MDG Contacts: Please send correspondence on this paper to Harold Alderman: halderman@worldbank.org, Hans Hoogeveen: jhoogeveen@worldbank.org or Mariacristina Rossi: rossi@economia.uniroma2.it. Acknowledgements: This paper was written in preparation for the Tanzania Country Economic Memorandum. The authors would like to thank (in alphabetical order): Awudu Abdulai, Kathleen Beegle, Deon Filmer, Philippe Krynen, Valerie Leach, Mildred McLachlan, Meera Shekar and participants to seminars at the World Bank and in Dar es Salaam for useful comments, suggestions and other types of assistance. Financial support by the Italian Trust Fund for Children and Youth is gratefully acknowledged. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and do not necessarily represent the views of the World Bank, its Executive Directors, or the countries they represent. Working papers describe research in progress by the author(s) and are published to elicit comments and to further debate. 1. Introduction The Millennium Development Goals (MDGs) consider income poverty and malnutrition indicators of poverty that should, relative to their 1990 levels, be halved by 2015. This raises the question whether nutrition and income poverty measure different dimensions of the same development strategy. In particular, would a growth strategy that might achieve a significant increase in GDP as well as a reduction in income poverty, also be sufficient to attain the nutrition MDGs? Or does malnutrition respond differently to income growth than does income poverty? Increases in income are clearly important for reducing malnutrition. Greater incomes at the household level allow families to spend more on food, clean water, hygiene and preventive and curative health care. It allows them to have a more diversified diet and to obtain more effective childcare arrangements. At the community level, greater income will eventually lead to better access to and higher quality health care, improved water and sanitation systems and greater access to information. Empirical evidence supports the importance of income growth for the reduction of malnutrition. A recent study using household survey data from 12 countries (including three African countries: Kenya, Mozambique, and South Africa) estimates the magnitude of the response of weight-for-age to income growth (Haddad et al. 2003). Under a scenario of sustained per capita income growth of 2.5% per annum, the average reduction in the fraction of underweight children would be between 27% and 34%. This average hides substantial inter country variation, even within Africa. The reduction could be as high as 42% (Kenya) or as low as 14% (South Africa). Thus, Haddad et al. (2003) suggest that complementary instruments would be needed. However, while there is a body of clinical and programmatic evidence on which interventions are most promising (Allen and Gillespie, 2001), there is comparatively little data that indicates how nutrition interventions could augment income growth, especially in Africa, the one region in which malnutrition rates have not yet begun to decline (de Onis et al. 2004). Papers that do assess the impact of interventions on nutritional status, often deal with programs such as food aid and food for work in the aftermath of a drought shock (Yamano et al. forthcoming; Quisumbing 2003). Others such as Christiaensen and 1 Alderman (2004) deal with the additional impact of nutrition education over the more general impact of schooling. This paper focuses on the role of income growth in combination with two types of nutrition interventions in northwestern Tanzania, community driven supplementary feeding of young children, feeding posts, and child feeding carried out in crèches run by an international NGO, Partage. The implementation of these programs was likely to be an emergency response yet these crises were not directly issues of food scarcity. In any case, the study does not assess the impact of the provision of food per se, but of the package of services, comprising of the provision of food, awareness creation, day and medical care, that accompany the intervention. The current study looks at anthropometric measures of child nutritional status. We study the degree to which increases in resources at household level and participation in nutrition programs contribute to the reduction of child malnutrition. Other determinants of nutritional status, such as access to health care, parental human capital and environmental factors are also considered. Our study examines the impact of the availability of program interventions on the nutritional status of children rather than the impact of participation itself. While both approaches convey useful information, Heckman, Lalonde, and Smith (1999) observe in their review of econometric methodologies for evaluation that often it is the former that is of policy relevance. The remainder of the paper is as follows. Section 2 describes the data set used. Section 3 illustrates the children's health status in the Kagera region of Tanzania. Section 4 presents the econometric specification used and the estimated results according to different specifications. Section 5 continues with policy implications that the analysis implies. Section 6 concludes the paper. 2. The Data Set The data used for this study come from a longitudinal living standards survey of households conducted in Kagera, a region of Tanzania with comparatively high adult mortality in the first decade of the AIDS epidemic. The region is located to the west of Lake Victoria and borders the Rakai district of Uganda to the north, and Burundi and Rwanda to the west. 2 The dataset is a four-round panel survey and the data collection covered the years from 1991 to 1994. The primary objective of this Kagera Health and Development Survey (KHDS) was to estimate the economic impact of the death of prime-age adults on surviving household members. Thus, the research focused on the collection of detailed socio- economic information from individuals who resided in a household in which one or more adult family members indicated they were in poor health at the time of the baseline (Ainsworth and Semali, 2001). The sample of households was selected from a stratified random sample of communities from the 1988 Tanzanian census (stratified on agroclimatic zone and adult mortality rate).1 Within each community, the household sample was stratified according to the anticipated households' risk of suffering a prime-age adult death. Households were divided in "high-risk" or "low-risk" categories, based on information obtained from a house- to-house enumeration of all selected communities. Altogether, 915 households were effectively interviewed at least once over the 4 waves of the survey, and up to a maximum of four times.2 Individuals belonging to the same household were defined as a group of people living and sharing meals together for at least 3 months in the last year (Ainsworth and Semali 2001). This survey collects information that makes it particularly appropriate for the purpose of our research, in particular the extensive information on household income and consumption and on individual's health status and anthropometrics. For the present analysis, which focuses on the children's nutritional status, we focus on the height, weight and demographic information of children aged 5 years or below and the characteristics of the households and community to which the children belong. Overall, the total sample of children surveyed amounts to 1140 different children, visited between one and four times and making up an unbalanced panel of 2871 person- wave observations. 1Both the prevalence of HIV and adult mortality rates in Kagera were geographically concentrated, and, therefore are strongly correlated with different climates and cropping patterns. The region with the highest rural HIV infection rates was in the northeast (10% in Bukoba Rural and Muleba districts and 24% in the town of Bukoba), where tree crops (bananas, coffee) were predominant. The regions with the lowest HIV infection rates were in the south and west (0.4% in Ngara and Biharamulo districts), where perennial crops and livestock are more common (Killewo and others 1990). Thus, a survey design stratified only on mortality rates might confound the effects of high mortality with different agricultural, soil, and rainfall patterns. 2Some households dropped from the sample at each round and they were replaced. Out of the original sample, 6 percent of households dropped before their first interview. Out of the remaining sample of households, 10 percent dropped out of the remaining sample before the end of the survey (Ainsworth and Semali, 2001). The attrition rate compares favorably to other panel datasets. 3 3. Nutritional Status of Children in Kagera The nutritional status of a child is usually assessed with three indicators: stunting (low height for age), wasting (low weight for height) and underweight (low weight for age). All indicators are expressed in "z-scores" which are derived by comparing the child's height and weight with that of a "reference" group of well nourished children defined by the US National Center for Health Statistics (NCHS) (World Health Organization (WHO) 1995). More specifically the stunting z-score is the difference (expressed in standard deviations (SD)) of a child's height for age from the median height of children of the same age and sex in the reference population. The z-score is thus the conversion of the child's height into a standardized unit. Relatively short children have negative z score; tall children a positive z score. The commonly used cut-off point to identify severely malnourished children is a measurement of 2 SD below the median of the "reference" group. Thus, a child is "stunted" if his/her z-score- is below or equal to ­2. Table 1 reports the descriptive statistics related to the sample. On average, 28 percent of children aged 0-5 are moderately underweight (i.e. their weight-for-age is more than two standard deviations below the median of the international "reference" population). Nearly 40 percent of the children are stunted but only 6 percent are malnourished according to the weight for height indicator. This is a general pattern; low height for age or low weight for age reflect long term cumulative deprivation while low weight for height indicates an acute crisis (Alderman, 2000). The former two measures being cumulative generally reveal more children outside of normal ranges than does the latter, which typically only shows high values during periods of acute food shortage. The WHO recommends stunting as a reliable measure of overall social deprivation (WHO, 1986), whereas the MDG nutritional indicator is defined as weight for age. In the remainder of the paper we concentrate on both stunting and being underweight. Table 2 illustrates the nutrition problems for Tanzania. According to the 1999 DHS, 43 percent of children under five are stunted, 6 percent are wasted and 29 percent are underweight. Though we do not suggest that the Kagera data are representative of Tanzania, a comparison between Tables 1 and 2 shows that the nutritional pattern in Kagera is not very different from that found elsewhere in the country. Figure 1 indicates that the probability of malnutrition declines with income. At low levels of income, more than 50% of children are malnourished. While this declines with 4 income there is still substantial malnutrition over much of the range of income. Figure 2 gives a somewhat different perspective. In this figure the y axis indicates Z scores for height for age and weight for age respectively rather than the percentage below a cutoff line. While this confirms that nutrition improves with income, it also shows that children in communities with a nutrition intervention do better than children in other communities. This pattern is particularly pronounced at the lower end of the income range. 4. Econometric Analysis and Results In this paper we explain in a reduced form regression the height / weight z score of children aged 0-5 years. The reduced form is based on a utility maximization over goods and health subject to a health production function and a budget constraint (see, among others, Glewwe et al. 2004). A general representation is as follows: Mit=a+bXit+gCit+eit (1) Where: Mit is the malnutrition indicator of the child i at time t, Xit a vector of regressors comprising socio-economic variables at household level, C it is a vector of community related variables and eit an i.i.d. disturbance term. The community related explanatory variables include a dummy variable indicating whether Partage or feeding posts are present in the village where the child resides. This variable is crucial to our analysis as it permits measurement of the effect of specific nutrition interventions. As mentioned, the paper assesses the presence of such programs in each community rather than the effective child's participation in the program. Therefore the coefficient relative to Partage / feeding post in a village should be interpreted as the impact on the (standardized) height or weight of a child due to the availability of a program in the village. A description of all other variables used in the model is presented in Table 3. Plausibly, program placement is endogenous. If, for example, programs are placed in areas with greater malnutrition problems, or in areas where program success is more probable, the measured impact of programs may be biased. Moreover, the sign of any bias due to endogenous program placement is uncertain. If programs are placed where the 5 village exhibits higher malnutrition rates stemming from unobserved factors the coefficient of program impact will be biased downward (Rosenzweig and Wolpin 1986, Pitt, Rosenzweig and Gibbons 1995). But if programs are placed in more potentially responsive villages an overestimation of program impacts is plausible. Thus, while the analysis begins with an OLS estimation, this is compared to a preferred estimation strategy which allows for endogenous placement. While in many contexts the most convincing way to handle the issue of placement is to have site selection randomized. This is feasible in our case and other measures are employed to increase confidence that the results are plausible. Firstly, we show that adding community correlates to a simple model with only individual and household variables and an indicator variable reflecting the presence of a nutrition program as regressors along does not change the basic outcomes. Moreover, we show that if the model is run with the dependent variable being nutritional status of older children ­ those who passed through the most nutritionally vulnerable ages before the programs were established3 ­ there is no apparent correlation between the presence of programs and nutritional status. Were programs established in more responsive villages, or in villages where malnutrition is an especially serious problem, a significant correlation between these sites and nutritional status would be expected, even in the absence of causality. Finally, we offer an instrumental variables approach, our preferred methodology. Seldom are instruments above criticism ­ even with a range of diagnostic tests. However, the plausibility of the results is augmented by the comparison of the age specific OLS results, which confirm that there is no prexisiting relation between the sites eventually chosen for Partage and nutritional status In the remainder of this section we will first discuss the results of the height for age regression. As the results for the weight for age regression are almost identical, we will only flag significant differences at the end of the section. Note that because the dependent variable is expressed in standard deviations, coefficients of dummy variables can be interpreted as the SD change of setting the dummy to 1; for continuous variables the coefficient associated to each regressor are the SD change in the dependent variable due to a unit increase of the regressor, while for variables expressed in logs they reflect the SD change of a doubling of the variable at hand. 6 4.1 Specification I: OLS The first specification includes in addition to the nutrition intervention, household and individual specific variables to explain aged 0-5 children's nutritional status. The results are shown in column 1 of Table 4. Column 2 of Table 4 includes, in addition to household and individual variables, a set of characteristics of the community where the children reside. The coefficient associated with the program is, in both cases, significant, increasing from 0.26 to 0.29 (column 1 and 2, respectively) when the observable community characteristics are added. As an additional check of robustness of our results, we consider how older children respond to the program. As mentioned, the rationale for this is that older children were unlikely to benefit appreciably from the intervention when the program started. Column 3 of Table 4 reports the results for children aged 6 to 9. As anticipated, the programs do not correlate with nutritional status, but other variables do. Returning to the results in column 2, consistent with the general literature (Shrimpton, et al. 2001), the age of the child has a strong non-linear impact on his/her nutritional status. To capture this effect, we include dummy variables that reflect different age categories. Compared to children aged one year or less (the omitted category), older children are more affected by nutritional problems. The nadir is reached for children aged between 3 and 4 years. Gender also matters for the height of a child. Girls perform better than boys in their height z-score, a result in line with the related literature (Svedberg, 1990). The season during which a child is born also plays a role in determining his/her nutritional status. Being born during the rainy season ­when labor demands and the burden of disease (malaria) are highest, has a negative effect on nutritional status. This effect is precisely estimated, but only for children who are less than 2 years old. The impact of the season of birth becomes insignificant later in the child's life. Parental characteristics are important determinants of the child's health status. Both father's and mother's height have a positive and significant impact on the height of the child. This result suggests persistence in nutritional status across generations: improving the nutritional status of the current generation not only improves the welfare of the current generation, but also that of future generations as children born from taller parents are less likely to be malnourished themselves. 3See Hoddinott and Kinsey (2001) for a discussion of age specific vulnerability as well as Shrimpton et al.(2001). 7 Our regressors also include whether the father and/or the mother has passed away. This captures differences in nutritional status between orphaned children and children whose parents are alive. The results suggest that it matters who passed away. Paternal orphans are shorter than children with a living father ­ a deficit of 0.32 standard deviations. Maternal orphans on the other hand, do not show a significant difference in their height compared to that of other children. Likewise, father's (but not mother's) educational attainment helps improve the nutritional status of the child. An additional year of education of the father increases a child's height by 0.03 SD. Household income and the environment where the child resides are key variables in determining the resources and the attention to the physical and psychological growth of the child. With higher income levels households are able to buy additional food (Abdulai and Aubert, 2004) as well as increase their hygiene standards. This study uses per capita consumption instead of per capita income to measure the economic resources available to the household, as suggested by Deaton and Grosh (2000). The main reasons to do so are that per capita consumption is recorded more precisely than households income and it is a better proxy for permanent income than current income. The estimates of the impact of consumption on nutritional status are at the low end of the range found in the literature (Haddad et al. 2003). More specifically, a doubling of household per capita annual consumption would generate an increase of the height z-score of the child equivalent to 0.18 SD. Income composition plays an additional role. A higher incidence of farm income out of total income has negative consequences for a child's nutritional status ­this effects is not very precisely estimated in the second column of Table 4, but is in our preferred random effects regression. Community variables also influence nutrition status. For example, children living in a community near a motor road show a height z-score that is almost half a SD higher than their peers living in a village without a motor road. Vaccination contributes significantly and positively to nutritional status. Distance to the health facility and presence of a village health worker do not have an impact on nutritional status like whether or not the motor road is passable. The coefficient relative to the variable for living in urban areas is positive, but again not significant. 8 Finally, the OLS results show that children in communities that have either a feeding posts or a Partage program or both are taller; they are almost 0.3 SD higher compared to the height of the children who live in communities without these programs.4 OLS estimates have also been performed on the underweight z score of the child. In this case, the presence of a nutrition intervention in the community no longer has a significant impact on the (standardized) weight of the child (it does in the preferred specification discussed in the next section). Expenditures on the other hand, continue to have a positive and significant effect. A doubling of household per capita consumption would generate an increase of the weight z-score of about 0.15 SD. Nutritional status still worsens with age, though this effect is more tenuous than in the case of stunting. Finally in this regression maternal education has a positive effect. So whereas stunting responded positively to additional education of the father, the weight z-score is responsive to additional maternal education. An additional year of maternal education increases the weight for age z- score of the child by 0.02 SD. 4.2 Instrumenting Program Placement and Observed Consumption. In seeking instruments to address the possibility of endogenous program placement it is necessary to find variables that have predictive power on the endogenous placement variables and, at the same time, are not correlated with the child's health output and the measurement errors present in the explanatory variables. The set of instruments used to model Partage and feeding posts include the average per capita household consumption in the village and the average height of adults aged 25-32 in the village as proxies for the average nutritional status of the village before Partage or other programs were started. The set of instruments also includes the main shocks each community experienced in the previous ten years, the main causes of adult and children death in the community where the child resides and whether the village has a daily market.5, Whether and which type of disasters happened in the past ten years presents the appropriate features for being used as an instrumental variable of the program placement. While such shocks impoverish the village it is arguable whether they affect the health of the children in a direct way rather than 4 Different specifications to capture the effect of feeding posts on the nutritional status of the child were tried. In particular we explored whether the impact of the intervention on the child's health varied with different household income levels. However, the results did not provide evidence in support of this assumption 5 See notes to Table 5. 9 through the increase in the probability of having a program placed in the village as well as through the included income, education, and community variables. We include tests to ascertain the validity of the proposed instruments. The ancillary regression used to instrument the (endogenous) feeding post / Partage variable is reported in Table A1 in Appendix 1.6 The coefficient estimates show that the intervention targeted villages with worse conditions such as villages with a low ratio of vaccinated children. This suggests that, unless instrumented, the OLS coefficient associated with the presence of a program will be downward biased, a finding in line with recent empirical findings of Behrman, Cheng, and Todd (2004) and Attanasio and Vera-Hernandez (2004) who also evaluated the impact of a nutritional program participation on the health outcomes of children. Their evidence suggest that, without taking into account endogeneity, the impact of the program would be underestimated. We also instrument per capita consumption to take into account measurement errors in per capita household consumption that could lead to attenuation bias towards zero. We use per capita household income and the quality of the roof of the house where the child resides as instruments. The use of income as an instrument does not, however, address the possibility that a child's health may affect income, possibly more through adult time allocation than the direct loss of child labor given the age of the children we are considering. The approach used here presumes that this is a minor issue relative to the problem of measurement error. F-tests indicate that the proposed instruments have predictive power to explain the presence of programs and that they have no predictive power for malnutrition beyond their influence through the nutrition program variable. The latter test is based on regressing the height of the child on all regressors and instruments. This test allows to detect whether instruments have additional power on the current height of the child, beyond through the prediction of the program probability. The value associated to this F test is 1.40, suggesting that our instruments do not have a direct effect on malnutrition. Thus it is valid to exclude them from the malnutrition equation.7 The results of the instrumenting equation, which are 6 As the placement of the program is at village level, the ancillary regression has been run on the sample based on the four round panel dataset made of village level observations; the dataset, thus, being different then the one used for our main regression. 7 We also derived a chi squared test following Wooldridge (2001). According to the test result we cannot reject the overidentifying restrictions. Different set of instruments have been tried, the coefficient associated with Partage ranging from 0.3 to 1.2. 10 shown Table 4, column 4, are similar in significance and in magnitude with the results in column 2. However, the presence of feeding posts or Partage has a much higher impact on nutritional status of children, once the endogeneity of the intervention is taken into account. This suggests that the OLS estimates (column 2) are downward biased. Column 5 of Table 4 illustrates the sensitivity of the results to assumptions on the error structure. As indicated, the data set is a panel with multiple observations on individual children. The panel structure of the data allows us to perform a random-effect estimation strategy, which takes into account the unobserved individual specific component in the error term eit. The random effect estimation strategy also includes the steps to account for the endogeneity of the feeding posts and measurement errors in per capita consumption explained above. The coefficient associated with the presence of a program in the village where the child resides is now 0.44. To test which specification is the more appropriate, the pooled regression (column 2) or the regression taking into account unobserved heterogeneity through an individual random effect (column 5), we perform a Breusch and Pagan (1980) Lagrange multiplier test9which tests the plausibility of imposing the assumption of equal variance in the individual error terms. The associated chi squared statistics is 1120, and we reject the null hypothesis that the data can be pooled. Thus, the random effect model is our preferred specification. Table 5 shows the estimates related to the (standardized) weight for age of children living in Kagera. The coefficients are similar to those obtained for stunting, in size, sign and significance. It is worth noting, however, that in the weight for age regression mother's as well as father's education has a positive impact on the weight of the child, and the coefficients show an equal impact of the parents' education on the nutrition output of the child. 5. Policy Implications In this section we use the models of Tables 4 and 5 to predict the declines in malnutrition that can be expected from a sustained annual increase in per capita income. Demombynes 8A fixed effects approach is not feasible as it wipes out all information about the presence of nutrition interventions. 11 and Hoogeveen (2004) report that, if inequality remains unchanged, a 2.2% annual increase in per capita income is sufficient to attain, for Tanzania, the income poverty MDG by 2015. We also explore to which degree income growth contributes to attaining the reduction of stunting and children being malnourished. We project forward from 1993 and allow for income growth and increases in the coverage of nutrition interventions from 12% of the households in the survey to an 32% and 62% (an additional 20%, 50%) and complete coverage and consider the joint impact. As the models in Tables 4 and 5 estimate the z score and not the fraction of children that is malnourished, the impact of growth and various interventions depend on (i) the size of the intervention and (ii) the distribution of the z- scores. As shown in Figure 3, the cumulative distribution functions for both height and weight for age are steepest near the cutoff point for malnutrition of ­2 standard deviations. Consequently, malnutrition rates will be particularly sensitive to changes in z scores. Table 6 illustrates the simulation results. It presents percentage changes relative to the baseline; the MDG objective is attained if a percentage change of 50% or more is made. If this is the case it is indicated in bold in the table. The second column shows reductions in income poverty. The other columns deal with changes in malnutrition resulting from various combinations of income growth and nutrition interventions. The results for income poverty in the second column suggest that to attain the income poverty MDG in Kagera, an annual rate of income growth 1.5% is needed. This is less than the growth rate of 2.2% suggested by the simulations of Demombynes and Hoogeveen (2004). The result is consistent because their simulations are for Tanzania as a whole for which in 1992 income poverty was 38.6%, whereas ours are for Kagera where poverty was 25.9% in 1992. Table 6 also shows that even a relatively high per capita growth rate of 3% is insufficient to attain the nutrition MDG. Additional program coverage will contribute to reducing malnutrition, but only a combination of (almost) universal program coverage with per capita income growth adequate to reach the poverty goal is sufficient to halve malnutrition by 2015. The final three columns in the Table show how other changes associated with per capita income growth will contribute to attaining the nutrition MDG. A reduction in the importance of farm income to at most 75% of total household income, will contribute an additional 2% malnutrition reduction; a motor road in every village reduces 9 See also Baltagi (2001) on this test 12 malnutrition by another 5%; an additional year of education for the father will add another 8%, and increasing the ratio of vaccinated children to 95% will decrease malnutrition rates by 10%. In combination these changes are non-negligible. If, per capita income growth can be kept at 2.5% (as has been the case in most recent years), and if this growth is associated with improvements in education, roads and less dependency on farm income, interventions in at least 50% of the communities would still be required to attain the MDG. 6. Conclusions A combination of income growth and nutrition interventions is often said to be needed to adequately tackle malnutrition, yet evidence to support this claim is not generally available. This paper evaluates the joint contribution of income growth and nutrition interventions towards the reduction of malnutrition in Tanzania. The paper also considers other determinants of malnutrition. Our results confirm that parental education and access to health care ­ as proxied by the fraction of vaccinated children in the community ­ matter. It also shows that stunting is a cumulative process; thus, underscoring the importance of ensuring adequate nutrition from very early childhood onward. Most importantly we find that both income growth and the presence of nutrition programs in the community contribute positively and significantly to the reduction of malnutrition. To investigate whether income growth, nutrition interventions or a combination of both are required to attain the nutrition MDG, we carried out a series of policy simulations. We find that income growth alone is insufficient to attain the MDG benchmark for nutrition. The same holds for nutrition interventions that reach less than half the population. Only the combination of income growth at the household level with large scale nutrition interventions is shown to be sufficient to bring about the desired results.10 Finally, this study shows that nutrition interventions are able to substantially contribute to the reduction of malnutrition. In showing this, the paper fills an important gap as impact evaluations of nutrition interventions are scarce (see Mkenda 2004 for a review for Tanzania). Even the much studied Iringa project is inconclusive about the effectiveness of various strategies such as growth monitoring and promotion, integrated care and nutrition, communications for behavioral change, supplementary feeding for women and young 13 children, school feeding, health related services, micronutrient supplementation and food- based strategies (Allen and Gillespie 2001). Though our study suggests that the described interventions are effective, their (cost)-effectiveness remains an unknown. A recent review on cost effectiveness suggests that some interventions are more cost effective than others, particularly vitamin A supplementation for children under age five, iron supplementation for pregnant women, and some types of nutrition education and behavioral change (Gillespie and Haddad 2001; Behrman, Alderman and Hoddinott, 2004). In this light it is of importance to take the next step and to assess the cost effectiveness of the programs considered in this paper. References Ainsworth M. and I. Semali, (2001), "The impact of Adult Death on Children's Health in Northwestern Tanzania", World Bank working paper, n. 2266. Abdulai A. and D. Aubert. (2004). A Cross-Section Analysis of Household Demand for Food and Nutrients in Tanzania. Agricultural Economics . Alderman H. (2000), "Anthropometry", in Designing Household Survey Questionnaires for Developing Countries in Glewwe P. and M. Grosh, vol.1 Allen, L., and S. Gillespie, (2001), "What Works? A Review of Efficacy and Effectiveness of Nutrition Interventions". Asian Development Bank, Nutrition and Development Series # 5, Manila. Arellano, M. and S. Bond. (1991). Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. The Review of Economic Studies 58: 277-297. Attanasio O. and A.M. Vera-Hernandez (2004), "Medium and Long Run Effects of Nutrition and Child Care: Evaluation of a Community Nursery Programme in Rural Colombia", unpublished paper Baltagi B.H. (2001), "Econometric Analysis of Panel Data", Chichester ; New York : John Wiley, 2001. Behrman J., H. Alderman, and J, Hoddinott, (2004), "Hunger and Malnutrition". In Bjorn Lomborg (ed.) Global Crises, Global Solutions. Cambridge (UK): Cambridge University Press. Behrman J., Cheng Y. and P.E. Todd (2004), "Evaluating Preschool Programs when Length of Exposure to the Program Varies: A Non Parametric Approach", The Review of Economics and Statistics, vol.86, no1: 108-132 10There is another reason, why income growth is required to bring about a reduction in malnutrition: it provides the only sustainable basis for increased revenue collection from which nutrition interventions could be paid. 14 Breusch, T. S. and Pagan, A. R., (1980), The Lagrange Multiplier Test and Its Applications to Model Specification in Econometrics, Review of Economic Studies, vol. 47, n1, pp. 239- 53 Christiaensen L. and H. Alderman, (2004), "Child Malnutrition in Ethiopia: Can Maternal Knowledge augment the Role of Income?", Economic Development and Cultural Change, vol. 52, n.2, pp. 287-312 Deaton, A. and M. Grosh (2000); "Designing household survey questionnaires for developing countries: Lessons from fifteen years of the Living Standards Measurement Study". Volume 1, pp. 91-133, Washington, D.C.: World Bank EDI 2004. Kagera Rural CWIQ. Baseline Survey on Poverty, Welfare and Services in Kagera Rural Districts. Demombynes G. and J. G. Hoogeveen 2004. Growth, Inequality and Simulated Poverty Paths for Tanzania, 1992-2002, World Bank Policy Research Paper No. 3432. World Bank. Washington DC. de Onis M, Blössner M, Borghi E, Frongillo EA, Morris R. 2004. "Estimates of global prevalence of childhood underweight in 1990 and 2015", Journal of the American Medical Association: no.291, pp.2600-6. Gillespie and Haddad 2001, "Effective Food and Nutrition Policy Responses to HIV/AIDS: What We Know and What We Need to Know", Journal of International Development v13, n4, pp. 487-511 Glewwe P., S. Koch and Nguyen B. L., (2004), "Child Nutrition, Economic Growth and the Provision of Health Care Services in Vietnam", in Economic Growth, Poverty, and Household Welfare in Vietnam, ed. Glewwe P., N. Agrawal and D. Dollar, World Bank. Haddad, Lawrence., Harold Alderman, Simon Appleton, Lina Song, and Yisehac Yohannes. (2003), "Reducing Child Malnutrition: How Far Does Income Growth Take Us?" World Bank Economic Review, 17(1): 107-131 Heckman J., Lalonde, R., and Smith, J. (1999). "The Economics and Econometrics of Active Labor Market Programs". In O. Ashenfelter, & D. Card (Eds.), Handbook of Labor Economics (Vol. 3A, pp. 1865-2097). Amsterdam: Elsevier Science. Hoddinott, J. and B. Kinsey, 2001. "Child growth in the time of drought". Oxford Bulletin of Economics and Statistics 63: 409-436 Killewo J. and others. 1990. "Prevalence of HIV-1 Infection in the Kagera Region of Tanzania: A Population-Based Study". AIDS 4 (11): 1,081-85. Mkenda A.E. (2004), "The Benefits of Malnutrition Interventions: Empirical Evidence and Lessons to Tanzania", unpublished paper, World Bank. Pitt M., Rosenzweig M.R. and D.M. Gibbons. (1995) "The Determinants and Consequences of the Placement of Government Programs in Indonesia", in van de Walle D. and N. Kimberly, eds. Public spending and the poor: Theory and evidence pp. 114-49 REPOA 2004. "Trends in Malnutrition in Tanzania". Mimeo. 15 Rosenzweig, M. R. and W. Kenneth, (1986), "Evaluating the Effects of Optimally Distributed Public Programs: Child Health and Family Planning Interventions", American Economic Review v76, n3, pp. 470-82 Quisumbing A. R. (2003). "Food Aid and Child Nutrition in Rural Ethiopia", World Development, Vol. 31(7): 1309-1324. Shrimpton R., Victora C., de Onis M., Costa Lima R., Blössner M., and G. Clugston. 2001. Worldwide Timing of Growth Faltering: Implications for Nutritional Interventions Pediatrics, 107(5): e75-81. Svedberg, P., (1990) "Undernutrition in Sub-Saharan Africa: Is There a Gender Bias?" Journal of Development Studies, v26, n3 (April 1990): 469-86 UNICEF 2001. Women and Children in Tanzania, Dar es Salaam, 1990. United Republic of Tanzania, National Bureau of Statistics [2002], Household Budget Survey 2000/01. United Republic of Tanzania, President's Office 2003. Economic Survey 2003. Yamano T, H. Alderman and L. Christaensen, Forthcoming. "Child Growth, Shocks, and Food Aid in Rural Ethiopia", American Journal of Agricultural Economics. Wooldridge, J. Econometric Analysis of Cross Section and Panel Data. Cambridge, MA: MIT Press, 2002. WHO (World Health Organization) 1995. Physical Status: The Use and Interpretation of Anthropometry, WHO Technical Report, No.854, Geneva WHO (World Health Organization) 1986, Use and Interpretation of anthropometric indicators of nutritional status, Bulletin of the WHO, Geneva. 16 Figure 1. Household income and the fraction of malnourished children (0-5 years) Fraction of stunted children Fraction of underweight children 2- 2- ssel ssel erocs-z e orcs-z .62 agerof gearof ght ghtie .4 hei w hti hti w w .25 onitcarF noit .2 acrF 0 0 8000 15000 25000 50000 100000 8000 15000 25000 50000 100000 Per capita consumption Per capita consumption 17 Figure 2 Nutrition z-scores for children aged 0-5 and presence of program interventions in the community Weight for age z-scores Height for age z-scores Nutrition program in village Nutrition program absent Nutrition program in village Nutrition program absent -1 -1 e e orcs-z orcs-z agerof -2 agerof -2 htgie ghtie W H -3 -3 8000 15000 25000 50000 100000 8000 15000 25000 50000 100000 Per capita consumption Per capita consumption 18 Figure 3. Cumulative density functions of nutrition z-scores for children aged 0-5 years Stunted children Underweight children . . 1 1 .75 .75 yti yti .5 .5 ens ens D D .25 .25 0 0 -5 -4 -3 -2 -1 0 1 2 3 4 5 -5 -4 -3 -2 -1 0 1 2 3 4 5 hzscore wzscore 19 Table 1. Nutritional status of children in Kagera region, Tanzania11. 1991-1994 Percentage of children whose z-score is below ­1 SD Height for age Weight for age Weight for height female 0.68 0.62 0.26 Male 0.72 0.65 0.29 Total 0.70 0.63 0.28 Percentage of children whose z-score is below ­2 SD female 0.35 0.25 0.05 Male 0.43 0.30 0.060 Total 0.39 0.28 0.06 Source: Kagera Health and Development Survey (KHDS) Table 2. Nutritional Status in Tanzania for Children Under-Five Percentage of children whose Height z-score is below ­2 SD Height for age Weight for age Weight for height 1991-92 0.43 0.29 0.06 1996 0.44 0.31 0.07 1999 0.43 0.29 0.06 Source: REPOA (2004), calculated using DHS 1991/92, DHS 1996 and TRCHS 1999 11For Kagera region the DHS data in 1996 report 42% of kids under five stunting, 36% under-weight and 11% wasting. (The percentage refelect those children whose zscore are below ­2) 20 Table 3. Description of the variable used and descriptive statistics. Kagera region (1991-1994) Stan dard Variables Definitions of Variables Mean Deviation Community variables Partage or feeding posts = 1 if there is Partage and/or feeding post in the 0.120 0.325 village Urban =1 if the child lives in a rural area 0.163 0.370 Health worker =1 if there is a health worker in the village 0.818 0.386 Distance to health center Distance to the nearest health center in km 3.131 3.495 Motoroad =1 if there is a motor road in the village 0.954 0.210 Road impassable =1 if the road is often impassable 0.483 0.500 Ratio of children vaccinated Ratio of 0 to 3 year old children in the community 0.712 0.212 who have been vaccinated Individual and household characteristics Height for age z-score Height for age z score statistics -1.643 1.388 Weight for age z-score Weight for age z score statistics -1.293 1.221 Aged 12_23 months =1 if Child's age id between 12 and 23 months 0.171 0.377 Aged 24_35 months =1 if Child's age is between 24 and 35 months 0.169 0.375 Aged 36_47 months =1 if Child's age is between 36 and 47 months 0.186 0.389 Aged48or more months =1 if Child's age is 48 months or more 0.308 0.462 Born in rainy season, aged < 2 Born in a rainy season and aged less than 2 0.251 0.434 Born in rainy season, aged > 2 Born in a rainy season and aged more than 2 0.508 0.500 Farm income/total income Farm income as fraction of total household income 0.687 0.245 Log per capita consumption Log of per capita consumption 10.483 0.609 Father dead =1 if Father dead 0.125 0.331 Mother dead =1 if Mother dead 0.041 0.197 Male =1 if male 0.523 0.500 Household size Number of household's member 8.992 4.315 Father's height Father's height (cm) 168.835 5.387 Mother's height Mother's height (cm) 158.522 5.790 Father's height missing =1 if Father's height missing 0.380 0.485 Mother's height missing =1 if Mother's height missing 0.149 0.357 Father's year of education Father's year of education 6.158 2.031 Mother's year of education Mother's year of education 5.191 2.681 Instruments Mean log per capita consumption Mean log per capita consumption in the village 10.528 0.344 Crop Disaster =1 if crop damage was the main disaster in the past ten years in the village 0.080 0.268 Drought disaster =1 if drought was the main disaster in the past ten years in a village 0.356 0.479 Daily Market in Community =1 if there is a daily market where the child resides 0.616 0.486 AIDS main adults' death cause =1 if AIDS is the main cause of adults' death in the village 0.480 0.500 Mean male height Average male height of adults aged 25-32 169.284 3.985 Mean female height Average female height of adults aged 25-32 158.09 2.851 Good roof in the house =1 if the child lives in a house with good roof 0.638 0.481 Mean (log) per capita income in Log of per capita income the village 10.224 0.724 21 Table 4. Regression results. Dependent variable: Height for age z score (standard errors adjusted for sampling structure) (1) (2) (3) (4) (5) OLS OLS OLS. Instrumental Individual Individual Individual, Children 6-9 Variable Random and household Regression Effect with household and instruments variables community variables Community Variables Partage or feeding posts 0.256 0.286 -0.016 0.543 0.437 (2.30)** (2.68)** (0.16) (1.83)* (1.90)* Urban 0.169 0.060 0.170 0.247 (1.25) (0.34) (1.18) (2.68)*** Health worker in the village -0.052 -0.158 -0.008 0.000 (0.72) (1.60) (0.10) (0.00) Minimum distance to the closest 0.012 0.014 0.017 0.007 health center (0.83) (0.74) (1.09) (0.91) Is motorable road near 0.488 0.432 0.482 0.510 (4.46)*** (3.42)*** (3.85)*** (2.94)*** Road impassable -0.005 -0.114 -0.032 -0.051 (0.05) (0.89) (0.27) (0.68) Ratio of children vaccinated 0.650 0.041 0.700 0.222 (3.99)*** (0.14) (3.52)*** (2.10)** Household and individual variables Log per capita consumption 0.206 0.184 0.154 0.207 0.103 (3.95)*** (3.55)*** (1.96)* (2.52)** (2.48)** Farm income/total income -0.322 -0.215 -0.517 -0.175 -0.160 (2.11)** (1.36) (2.83)*** (1.02) (1.79)* Household size -0.008 -0.005 -0.030 -0.008 -0.006 (0.86) (0.64) (2.96)*** (0.85) (0.88) Aged 12_23 months -1.118 -1.189 -1.186 -1.156 (11.89)*** (13.06)*** (13.00)*** (19.77)*** Aged 24_35 months -1.233 -1.238 -1.236 -1.257 (7.65)*** (8.12)*** (8.14)*** (12.57)*** Aged 36_47 months -1.468 -1.479 -1.482 -1.510 (9.66)*** (9.94)*** (9.89)*** (14.76)*** Aged 48or more months -1.255 -1.257 -1.259 -1.438 (8.68)*** (8.76)*** (8.70)*** (13.84)*** Aged 8 years 0.034 (0.48) Aged 9 years -0.036 (0.51) Born in rainy season and aged -0.238 -0.197 -0.198 -0.419 less than 2 (2.24)** (1.79)* (1.78)* (4.58)*** Born in rainy season and aged -0.008 -0.035 -0.029 -0.178 more than 2 (0.07) (0.29) (0.24) (1.93)* Father dead -0.289 -0.324 -0.295 -0.316 -0.178 (2.36)** (2.65)** (2.47)** (2.49)** (2.03)** Mother dead 0.142 0.172 -0.038 0.161 0.099 (0.60) (0.74) (0.26) (0.68) (0.81) Male -0.242 -0.249 -0.118 -0.254 -0.301 (3.63)*** (3.59)*** (1.35) (3.66)*** (4.42)*** 22 Father's height 0.043 0.042 0.031 0.041 0.038 (5.61)*** (5.59)*** (3.74)*** (5.48)*** (5.74)*** Mother's height 0.039 0.038 0.034 0.038 0.039 (4.23)*** (4.03)*** (3.44)*** (3.79)*** (6.22)*** Father's height missing 0.110 0.059 0.239 0.057 -0.029 (1.31) (0.71) (1.74)* (0.68) (0.36) Mother's height missing -0.034 0.039 -0.095 0.039 0.118 (0.25) (0.30) (0.71) (0.29) (1.10) Mother's years of education 0.011 0.000 -0.014 0.002 0.000 (0.78) (0.01) (0.71) (0.14) (0.00) Father's years of education 0.032 0.029 0.032 0.032 0.036 (2.04)** (1.85)* (2.05)** (2.03)** (2.74)*** Constant -16.292 -16.122 -14.757 -16.441 -14.533 (9.43)*** (8.28)*** (7.04)*** (9.19)*** (9.60)*** Observations 2984 2871 1357 2871 2871 R-squared 0.21 0.22 0.12 0.22 Notes: Absolute value of t statistics in parentheses. * indicates significance at 10% level; ** at 5% and *** significant at 1% level of confidence. Columns 4 and 5 use predicted values of log of per capita consumption and feeding posts to deal with the endogeneity and measurement error problems associated to their variables. The Partage / feeding post variable is a community level variable. Consequently the set of instruments are community variables. They include: mean of the (log) household expenditures, drought and crop disaster in the past ten years, presence of a daily market on community, AIDS being a main cause for adult mortality and average female and male height for those aged 25-32. The Partage / feeding post variable was then regressed on the instruments listed here and other community variables in the model. Log per capita consumption is instrumented with log income and the presence of a good (wood, iron sheets, concrete, asbestos) roof in the house. Predicted per capita consumption was then obtained after regressing log per capita expenditure on its instruments and the various community and household regressors. Standard errors corresponding to T-statistics in columns 4 and 5 are obtained by using the bootstrapping method, required because of the complication of instrumenting community variables. The total number of different individuals observed between 1 and 4 times is 1206. 23 Table 5. Regression results. Dependent variable: Weight for age (standard errors adjusted for sampling structure) (1) (2) (3) (4) (5) OLS OLS OLS. Instrumental Individual Individual Individual, Children 6-9 Variable Random and household Regression Effect with household and instruments variables community variables Community Variables Partage or feeding posts 0.130 0.158 -0.102 0.553 0.456 (1.11) (1.30) (0.88) (1.88)* (2.14)** Urban -0.020 -0.002 -0.019 0.072 (0.16) (0.01) (0.15) (0.86) Health worker in the village 0.084 -0.061 0.145 0.080 (1.23) (0.57) (2.32)** (1.63) Minimum distance to the closest 0.003 0.012 0.011 0.006 health center (0.31) (0.79) (0.98) (0.76) Is motorable road near 0.496 0.415 0.478 0.553 (5.81)*** (3.91)*** (6.18)*** (3.53)*** Road impassable -0.131 -0.142 -0.167 -0.196 (1.54) (1.33) (2.04)** (2.89)*** Ratio of children vaccinated 0.310 0.002 0.398 0.166 (2.36)** (0.01) (2.67)** (1.65)* Household and individual variables Log per capita consumption 0.209 0.170 0.142 0.209 0.117 (4.19)*** (3.38)*** (2.32)** (3.59)*** (2.98)*** Farm income/total income -0.340 -0.335 -0.359 -0.281 -0.163 (2.37)** (2.40)** (2.51)** (1.91)* (1.93)* Household size 0.004 0.004 -0.016 0.002 0.006 (0.60) (0.63) (1.74)* (0.27) (0.87) Aged 12_23 months -0.965 -0.999 -0.997 -0.932 (12.05)*** (13.15)*** (12.80)*** (16.78)*** Aged 24_35 months -1.110 -1.099 -1.101 -1.174 (7.81)*** (7.65)*** (7.67)*** (12.50)*** Aged 36_47 months -1.128 -1.114 -1.119 -1.189 (7.92)*** (7.90)*** (7.87)*** (12.40)*** Aged 48or more months -1.069 -1.072 -1.076 -1.152 (8.05)*** (7.81)*** (7.81)*** (11.88)*** Aged 8 years -0.012 (0.17) Aged 9 years -0.085 (1.11) Born in rainy season and aged -0.246 -0.203 -0.208 -0.537 less than 2 (2.56)** (2.09)** (2.16)** (6.35)*** Born in rainy season and aged -0.049 -0.041 -0.039 -0.213 more than 2 (0.43) (0.38) (0.36) (2.53)** Father dead -0.095 -0.129 -0.177 -0.123 -0.102 (1.04) (1.45) (1.54) (1.37) (1.26) Mother dead -0.077 -0.060 -0.142 -0.067 -0.030 24 (0.54) (0.42) (1.12) (0.47) (0.26) Male -0.148 -0.148 -0.036 -0.153 -0.203 (2.31)** (2.24)** (0.48) (2.29)** (3.31)*** Father's height 0.033 0.032 0.009 0.031 0.032 (4.48)*** (4.50)*** (1.24) (4.45)*** (5.32)*** Mother's height 0.021 0.022 0.022 0.022 0.024 (2.87)*** (3.03)*** (3.11)*** (2.90)*** (4.25)*** Father's height missing -0.002 -0.037 0.095 -0.040 -0.093 (0.03) (0.46) (0.83) (0.49) (1.27) Mother's height missing 0.048 0.077 -0.055 0.074 0.128 (0.40) (0.63) (0.58) (0.60) (1.30) Mother's years of education 0.023 0.021 0.002 0.022 0.017 (1.88)* (1.97)* (0.12) (2.07)** (1.35) Father's years of education 0.016 0.010 0.019 0.012 0.018 (1.08) (0.73) (1.29) (0.86) (1.51) Constant Observations 3006 2922 1362 2922 2922 R-squared 0.17 0.18 0.08 0.18 Notes: See table 4. Column (5): the total number of different individuals observed between 1 and 4 times is 1214. 25 Table 6. Reduction (in %) of the fraction of children stunted and underweight Per capita Reduction in Reduction in malnutrition (%) income income growth poverty (%) (in %) ­ since No additional Interventions Intervent Intervent Farm income Motor Additional Ratio of 1993 interventions in additional ions in ions in all maximum of road in year of children 10% of additional communities 75% of the every education vaccinated communities 50% of total village to the increased communities household father to 95% income Height for age 0 0 0.0 7.0 41.3 68.7 3.3 6.2 8.1 9.6 0.5 24.7 2.6 9.1 42.7 69.2 6.0 8.9 10.2 11.2 1 44.1 5.3 11.5 44.4 70.3 7.8 11.5 12.4 13.3 1.5 55.7 7.1 14.2 46.3 71.9 9.6 13.1 14.9 14.7 2 66. 6 8.7 16.3 48.1 73.2 11.6 14.6 16.4 17.2 2.5 79.0 11.8 17.7 49.5 74.0 13.9 17.7 18.5 18.6 3 84.1 13.9 19.2 51.0 75.2 16.6 19.9 20.5 20.5 Underweight 0 0 0 11.0 31.7 53.4 1.8 2.4 3.6 9.9 0.5 24.7 3.4 14.1 34.4 55.5 5.2 5.8 7.0 12.9 1 44.1 6.8 17.2 37.0 57.6 8.5 9.1 10.3 15.9 1.5 55.7 10.1 20.2 39.5 59.6 11.8 12.4 13.5 18.8 2 66.6 13.3 23.2 42.0 61.5 15.0 15.6 16.7 21.7 2.5 79.0 16.5 26.0 44.4 63.4 18.1 18.7 19.7 24.5 3.0 84.1 19.5 28.8 46.7 65.1 21.1 21.7 22.7 27.2 Note: Simulations are based on the random effect regression model, which is our preferred estimation strategy. The simulations were done taking 1993 as the base year. Since the per capita income growth rate between 1993-2003 is known (0.7% per annum ­ URT 2004), the effective growth rates required to attain the 1993-2015 mean growth rates 0%, 1%, 2% and 3% for the 2003-2015 period are respectively: -0.5%, 1.3%, 3.1%, and 5.0%. Figures in bold show the attainment of the MDG. 26 Appendix 1. Table A1. Determinants of feeding posts placement. Dependent variable: presence of Partage / feeding post in the village Explanatory variables coefficient Mean of (log) per capita consumption in the village 0.101 (1.48) Daily Market in Community 0.121 (2.61)*** Crop disaster in past 10 years 0.273 (2.76)*** Drought disaster in past 10 years -0.054 (1.11) AIDS main adults' death cause 0.045 (0.97) Mean male height of adults aged 25-32 -0.003 (0.60) Mean female height of adults aged 25-32 0.002 (0.31) Urban -0.017 (0.27) Health worker in village -0.120 (2.08)** Distance to the closest health center -0.013 (1.96)* Is Motorable Road Near -0.041 (0.35) Road impassable 0.115 (2.46)** Ratio of children aged 0-3 that are vaccinated -0.198 (1.75)* Constant -0.612 (0.37) Observations 197 R-squared 0.21 Note: Regression based on observations at village level and not on the entire sample of children's observations. Absolute value of t statistics in parentheses. * indicates significance at 10% level; ** at 5% and *** significant at 1% level of confidence. 27 Draft: 12January 2005 Risk, Growth and Transfers. Prioritizing Policies in a Low Income Environment with Risk The Case of Tanzania JohannesG. Hoogeveen Summary: This paper investigatesfor rnral Tanzania the consequencesof risk for growth. This impact is shown to be considerable.Climatic andprice risks and human, animal and cropdiseasean- the main sourcesof risk in rnral an-as.A combination of inadequate n-turns to investment, lumpy investments,irreversibilities, illiquid assetsand imperfectfinancial markets further pn-vent householdrfrom attaining theirgrowth potential It isfound that ma'!) risks an- pn-ventable through enhancedseroicedelivery in the health, agricultun- and water sectors.Improved infrastrnctun- and betterfinancial seroiceswill further n-ducetheimpact of risk and contribute togrowth. In light of thesefindings it is proposed to n-considerthe balancebetweeninteroentions which addn-ssstrnctural determinants ofpover() (growth)and interoentions which n-ducethe consequencesof shocks (transfers), topolicies that (i) improve the investment climate, (ii) pn-vent risks and (iii)promote financial markets. 1. Introduction With a per capitaincome of $290Tanzaniabelongsto the low incomecountries.! Low per capitaincome not only suggestsapervasivecasefor growth it alsomakesthe twin problem of poverty andexposureto uninsuredrisk pressing.Risksarepervasiveespeciallyin a low income environment:economicpoliciesareinherendyunstable;exportsdependon commoditieswhosepricesfluctuateconsiderably;andincome generationis constrainedby exposureto climatic shocksanddisease.It is beginningto getrecognizedthat exposureto theserisks maybe a major causefor transitoryandevenchronicpoverty. In Tanzaniathe associationof risk andpoverty hasbeenilluminated by the participatory poverty assessmentwhich focusedon householdvulnerability (pPA 2004).This, and other studies(e.g.Ward etal. 2005)showhow riskis an important elementin explainingpoverty. Thesestudiesalsoarguein favor of safetynetsand other typesof transfersto dealwith the consequencesof exposureto uninsuredrisk. Theseviewsarebeingtranslatedinto action. TASAF II, for instance,focuseson the creationof assetsfor the chronic poor andthe provision a safetynetsto the vulnerable. The renewedattentionto directtransfersto the poor and safetynetshasled to a degreeof dualityin the debateaboutwhich economicpoliciesto follow. Someproponentsstressthe importance of growth for poverty reduction. Others preferinterventions direcrlyaimedat 'This paperwas written in preparation for the Tanzania COWltryEconomic Memorandum. The findings, interpretations and conclusionsexpressedin this paperare entirelythoseof the author. They do not necessarily representthe view of the World Bank,its Executive Directors, or the coWltries theyrepresent.Correspondence regardingthis papercanbe sendto: jhoogeveen@worldbank.org. 2003,atlasmethod (currentUS$).GNI for SubSaharaAfrica is $490. Source:SIMA improving the welfare of the poor. Both sidesof the debatehavesincereclaims.The casefor direct supportto the poor is convincinggiventhe widespreadprevalenceof abjectpoverty. But the hesitancyto spendon unproductive transfers-evenif thesewould haveapositive impact on the welfare of the poor, is equallysoundgiventhe opportunity coststhis hasfor growth. Consequendyproponents of eachapproachend up talking mosdyto their own constituencies,andvery litde to eachother. One areawherethis is evidentis in the draft PRS2document (URT 2004)which treatsgrowth and quality of life asdistinct aspectsof governmentpolicy. This neednot be. The literature on economicgrowth startsto recognizethat exposureto risk hasdirectconsequencesfor growth. Similarly,proponents of supportto the poor focus more on mechanismsthat allowthe poor to build up their assetsso asto escapethe threat of poverty in a sustainablewayandto withstand shocks.Economic growth, risk management and poverty reductionarecloselyintertwined. In this paperI will investigatethis relationand considerhow to prioritize betweengrowth enhancingpolicies,risk managementand transfersto the poor. I do so startingfrom four stylizedfactswhich I describethe Tanzanian environment but which arenot atypicalfor subSaharaAfrica: (i)poverty is mainlyarural phenomenon, (ii)risksarepervasive(iii) financialmarketsareimperfect and (iv)geographic isolationis areality. The paperis organizedasfollows. Section2 considersthe relation betweenrisk and growth from the perspectiveof arural household.It presentsa theoreticalintroduction and discussesempiricalevidencefrom Tanzaniaandelsewhere.Section3 discussesmain sources of risk for Tanzania.Section4 extendsthe discussionof sectiontwo anddiscussespoverty traps. Section5 finally discussespolicyinterventionsandproposesanagendafor research. 2. Risk and Growth in a Low Income Environment 2.1 A stylized description of roral Tanzania In this sectionI takeamicro-economicperspectiveand sketchthe processof production and accumulationof a typicalrural Tanzanianhousehold.The perspectiveis that of a householdasit allowsto abstractfrom the relation betweengrowth andinequality. In the absenceof changesin inequalityhigher growth will leadto more income andlesspoverty. SinceI aminterestedin growth, the settingis one of manyperiods,and the householdhasto decideabout how muchto consumetoday,and how muchto savefor tomorrow. Povertyin Tanzaniais mosdyarural phenomenonand I considera ruralhousehold thathas accessto one asset(livestock)which canbe usedfor productive purposesand which actsas store of wealth. Labor supplyis fixed, thereis no labor market (or migration)and total income dependsonly on the availabilityof assetsandthe realizationof shocks. Welfareis derived from consumption, suchthat the extrawelfareassociatedto additional consumptiondiminishesthe wealthierthe householdgets.The householdhasaninterestin maximizingits consumptiontoday,but alsoin assuringthat it hassufficient to consumein the nextoeriods. The householdis exposedto risk which mayaffectits assets(livestockmaydie) or its production (the harvestis low becauselocustaffectedits crop). Shocksdo not affect consumptiondirecdy. In a contextwherehouseholdsare frequendyexposedto unexpected expenses(e.g.for medicalcare)this maybe seemunrealisticbut without lossof generality, shocksto consumptionmaybe consideredlossesto disposableincome. Shocksonly haveanimpact onwelfare through their consequenceson consumption. A shockthat reducesincome will translatein lower consumptionunlessthe householddissaves and'consumes'someof its assets:risk coping.There areno transactioncostsandassetscan easilybe translatedinto consumptionitems.If assetsareconsumedthe householdwill generatelessincome nextyear.likewise, if a shockreducesassets,this will translateinto lower future income,unlessthe householddecidedto consumelesstodayto rebuild its stock of assets.I assumethat the welfarecostsof lossesin income are especiallyhigh when consumptionis alreadylow thatis whenthe householdis poor. When consumptionreaches zero,the additionallossto welfarebecomesinfinitely large,givingthe householda strong incentive to alwaysavoid sucha situation.In otherwords, the utility function isconcave. The householdoperatesin an environmentcharacterizedby imperfect capitalmarketsand by geographicisolation. An immediateconsequenceof this is thatin its decisionhow much to consumenow and how much to savefor the nextperiod, the householdis not guidedby an exogenouslygiven interestrate.Rather,how muchto savedependson the marginal return to productive assets.Sowhen decidingwhether or not to forego consumptiontoday, for the benefit of tomorrow, the householdassesseshow muchthe additional contribution to assetswill contribute to incomein the nextperiod. How much this is dependson the realizationof risksandthe assetsthat arealreadyavailable.It is realisticto assume diminishing marginalreturns (the production function is concave)so that additional capital contributeslessto income if thereis alreadya considerableamount of assets.An important implication of the latteris that for agiven levelof assets,the lossof assetsin terms of foregoneincome will alwaysbe greaterthan an equallysizedadditionto the stockof assets will bring in additionalincome. Another consequenceof geographicisolationandimperfect financialmarketsis that credit cannotbe basedoncollateral:land haslittle valuesinceit is abundant(andnot tided); animalsarevulnerableto sicknessandtheft andaremovable.As a consequenceformal creditmarketsarepoorly developed.This descriptionfits the Tanzanianenvironment. According to the HBS 2000;01, 4% of allrural householdsoperateda savingsaccountin a formal bank, 3%participatedin aninformal savingsgroup and0.4%receiveda bank loanin the previousyear. Informal loansaremore common (e.g.Kessy2004),but arealmost exclusivelytakento dealwith emergencyexpensesandarenot provided to finance productive investments.Exposureto shocksandthe absenceto accessto creditprovide the householdwith amotive for saving,soasto alwaysavoid a situationwherethereis (nextto) nothing to consume.However safeassets-or storesof wealth,do not exist.Households mayinvest their savingsin e.g.livestock,but thesecould die or get stolenor atthe time when the householdneedsmoneyno interestedbuyersmaybe present. The householddealswith shocksin isolation, through selfinsurance.There areno insurance mechanisms.The combination of moral hazard,covariant(climatic)risksandgeographic isolationprecludesthe formal pooling of risks.The absenceof informal insurance .1 mechanismsis lessrealistic.Yet it is well known (e.g.from the PPA) that manyrisksremain uninsured.Hence I only considerthoserisks thatare not insuredotherwise.Transfersby the governmentor NGO in responseto a shockarealsonot included. Thus formulated, this stylizedenvironmentcanbe formalizedasfollows. Household,h, generatesincome by putting atthe beginning of eachperiod its assets,k, to work in an income generatingprocess:af(k)thatis affectedby shocks:I (whereE 1= 1).Income does dependon the endowmentof productive assetsandthe returnsone canreap&om those assets. Not onlyincome is subjectto shocks,soareassets(I ; E 1 = 1),so that afterthe production processis finished the following endowments,w, areavailable. Wht =shtahtfh(kht) +s~t(l-/i)kht where time is indexedwith subscriptt and8 is a discountrate. Total endowments,Whl,area function of the capitalstockandthe shocksto which the householdwasexposed.The endowmentscanbe usedfor consumptionor theycanbe savedandusedasproductive assetsin the next period. ChI =Wht -kh,t+l At the time when the householdhasto decideabouthow muchto consumeand how much to savefor the nextperiod (i.e.att=O) the realizationof pastshocksandincome, the availablecapitalstockandthe distribution of future shocksareknown andthe household solvesthe following optimization problem: 00 V(w(kho' sho' sZo» = max EoLPIU(Chl) {cht,kht+l} 1=0 subjectto the constraintsmentionedabove. The previouslysetup modelis a Ramseygrowth model with risk anda singlegood,usedfor consumption,asa store of wealthandasaproductiveasset. Thesekind of stochasticmodels typicallyhaveno closedform solution, though a solution canbe derivednumerically.The keyto the solutionis the realizationthat whenthe householddecideson how muchto consumeand how much to save both ~ andthe realizationsof the shocks(II' I) areknown. Future shocksare,of course,unknown but the householdknows the distribution of theseshocks.Thus set-upthe modelcanbe written in recursiveform asa Bellmanequation.A solutionto this modelis beyond the scopeof this paperbut seeElbers,Gunning and Kinsey(2003)! Sufficeit to saythat a solutionto this modelwould mapthe currentassetendowment,given the risk environment (kl' II' I), into 2 Elbers,Gunning and Kinsey 2003,arguethat sucha show amodelis unlikely to havean analytic solution, but show how it canbe solvednumerically. 4 next periods assets(kt+t). This mapping function, tjJ,can be seenas an investment function, giving kt+tas a function of kr The baselinecaseis the deterministiccaseandis presentedin Figure 1. It showshow householdsmap capitalfrom oneperiod to the next, showinglargeincreasesin capital(i.e. growth) when the initial capitalstockis low andthe marginalreturnsto capitalhigh and a slowing down whenmore assetsareavailableandthe marginalreturn diminishes.The point k* indicatesthe steadyassetlevelwhereassetsin period t, kt, equalsthosein the next period, kt+t.In this standardRamseymodel, growth is aresponseto the difference of current capitalstock from its steadystatevalue.Povertyin this modelis only a temporary phenomenon.Through aprocessof accumulationthe household is ableto grow out of it. 2.2 Impact of Risk on Growth The advantageof presentinga canonicalgrowth modellike the Ramseymodelis that it allowsto identify why householdsmaynot be ableto grow out of poverty. One reason immediatelypresentsitself: in the absenceof good opportunities to invest,the marginal return to capitalwill be low leavinghouseholdswith only alimited incentive to save. Consequendythe steadystateassetlevel,k,* will be low and possiblyevenlaybelow the poverty line. The absenceof anattractiverateof return maybe one reasonwhy so fewrural households in Tanzania(11.1%) investin ox-ploughsto cultivatetheir land. Though the suggestionthat poverty maybe a self-chosenstatemaystrikeasstark,it is anestablishedresultthat under semi-aridconditions with few investmentopportunities and low population density, agriculturewill takethe form of autarky (BinswangerandMcIntire 1987).And while autarky maybe too strong aqualification,HBS datasuggestthat for 48% of all rural householdsthe saleof food cropsis the main sourceof cash.Figure2 suggeststhat thereis a bifurcation in 5 the distribution of those that rely on the saleof food crops for cash,with alargefraction of villageswhere only a small fraction of the householdsrely on food salesfor cash,anda considerablesetof villageswherethe majority of householdsrelieson food salesassource of cashincome. In 25% of the sampledvillages,at least75%of the householdrelied primarily on food cropsassourceof cash. Figure 2: Figure 3: Density of fraction of householdsin rural Fractionof householdsin village sellingfood villagesthat dependon saleof food asmain cropsanddistanceto the market sourceof income Lowesssmoother i -5 j'" ,. 0:- S'" 'E i"~ i 6 t'" 1- 0 2 4 .6 6 1 0 5 10 15 20 25 ~. F_bn of1o...l1oldswithsal.. of foodcropsasOBinsourceofc..h 0;,-,. to m"ket(km) Note: The distanceto marketis determinedasthe minimum of the distanceto the marketand distanceto public transport. It is not evident from thesedatathat thosehouseholdsthat rely on food crops for cashare 'autarkic' in anysense,or that theyarepoorer.The distinction betweenfood and cashcrops is not alwaysclear,andmaizecould countasfood or aswell ascashcrop, so that eventhose householdsthat specializein food crops for cashmaybe well integratedin the market.Yet someproportion of the householdsseemsto be driven by lack of good investmentoptions: onepieceof evidencepointing in this directionis that per capitamonthly consumptionof householdsrelying on the saleof food crops for cashis significandylessthan that for those who obtain cashfrom other sources:TShs 10934versusTShs12460.Furthermore,asFigure 2 suggests,whenthe distanceto the marketincreases(andthe returnsto investmentdecline), the fraction of householdsfor whom the saleof food cropsis the main sourceof cash increasessteadilyfrom 40%whenthe distanceis limited to over 60%whencommunitiesare isolated. The Ramseymodelalsopennits to investigateanotherreasonwhy householdsmaybe poor: exposureto risk. Figure 4 showsvarious policy functions under different risk scenarios.The baselinefunction is againthe deterministiccaseof Figure 1.Once risk is introduced, householdscanrespondin two ways.They candecideto savemore for precautionary reasons.Householdscould alsodecideto saveless,if the precautionarymotive is offset by the dangerthat current savingshavea low returnif the agentexperiencesapositive shock (thereare diminishingreturnsto capital).3The net effectis determinedby the characteristics 3To keepthe models comparablewhile moving from a deterministicworld to one in which shocksare anticipated (but not realized),one hasto make surethat in expectationthe samemeanvalues for capitaland income are obtained,but thereis arisk thatincome (assets)canbe less (more)than anticipated.If assetsare kept in the form of cattle,it meansthatwe introduce the risk that cattledie unexpectedly,but also that they () of the distribution of shocks(how likelyis it thatpositive or negativeshocksoccur),the curvatureof the utility function (how muchis the loss (increase)in welfareif consumptionis lower (higher)and of the production function (how muchadditional (less)incomeis earned if the capitalstockincreases(declines).In Figure 4 the effect of anincreasein risk is to shift the policy function downward:the exante effect of risk on savingis negative(k** < k*). In practicethereis anotherreasonwhy risk mayleadto fewerproductive assets:assetilliquidity. Unlike suggestedin the model,assetscan not alwayseasilybe transformedinto consumption.This is likely to inducehouseholdsto hold aportfolio of assetsincluding cash, jewelry,bicyclesand food storeswhich, evenif precautionarymotivesinducethe household to increaseits savings,mayleadto fewerproductive assetsthanin the deterministicscenario. In addition to the ex anteeffect,thereis anex post effect of risk on growth: the actual occurrenceof shocksaffectsaccumulationfor agivenpolicy function. This additional effect occursbecauseshockshaveanasymmetricimpact on production: a negativeshockto assets leadsto a greaterlossin production thanapositive shockof equalsizewould addto production. The reasonfor this hasto do with the shapeof the production function. Christiaensenetal. (2004)provide agood illustration from the Kilimanjaro region of the asymmetricimpact of shockson assetaccumulation.They showhow, facedwith low coffee prices,poor householdswith few assetsfor consumptionsmoothing uproot their coffee treesto make room for othercrops.Yet shouldcoffeepricesagainincrease-as is likely given the cyclicalnature of coffeeprices,then it will take newlyplantedtreesat leastthree yearsto mature. Consequendythesefarmersareunlikelyto be ableto quicklyexpandtheir assetbaseoncecoffeeprices,recover. Finallythe capitalstockk*** is not the one thatone can expectto observein the long run acrossthe population.4The reasonfor this is the sameasabove:becauseof diminishing marginalreturnsadditionsto the capitalstockyield lessin income than equallysizedlosses, so that evenif positive andnegativeshocksoccurwith the samefrequency,the total asset basethat is observedis lessthanthe steadystateequilibriumbecauseit takeslongerto recoverfrom anegativeshock. The presenceof antheseex anteandex post effectson risk areimportant for the understandingof poverty. If theseeffectsare substantialtheyimply that risk is a structural determinantof poverty asit reducesthe long run value of the capitalstock(and consequentlygrowth). The implication is thatpolicieswhich aredesignedto helpreducerisk aregrowth enhancingbecausetheyhelpto avoid both ex anteand ex post effects.Other risk relatedpolicies suchasthose that allowhouseholdsto mitigate or copewith risk through insuranceor by offering bettermeansof self-insuranceeliminatethe ex anteimpact of risk on growth, but maintainthe ex posteffect. The presenceof the ex posteffectsk*** and E(k) ongrowth, presentsan especiallystrong caseto opt for risk prevention.This holds evenstrongerif risksarenot symmetric(unlike give birth so that one cow todayleads(in expectation)to one cowtomorrow. In other words,we introduce a meanpreservingspread.It is important to recognizethat this is the way risk is introduced. It is something that correspondswell to how price riskis perceived,with prices sometimesbeing very high, and sometimesbeing very low. 4I.e. in the ergodic distribution. what hasbeenmodeledhere)but aremainlynegativeasis the casewith healthshocksfor instance.If suchdownsiderisks canbe avoidedit meansnot only that the ex post, but also that the ex anteimpact of risk on growth canbe avoided. Figure 4: Policy functions for transition of capitalunder risk Source: Elbers et al. 2003 2.3 Empirical evidence: how important is the effect of risk on growth The theoretical sectionhasmadeclearthat riskis likely to matter for growth through anex ante effect and anex posteffect. This sectiondiscussesthe empiricalevidenceto supportthe impact of risk on growth. Elbers etal. (2003)from which much of the theoreticalframeworkis derivedestimatethe impact of risk on farm householdsin Zimbabwe. They not only find a substantialeffect of risk on growth, theyarealsoableto distinguishthe differential consequencesof the ex ante and ex postimpacts of risk on growth. Theyestimatethat the meanof the assetdistribution is 46% lower thanit would bein the absenceof risk, and that the annualgrowth ratewould be 20-50%higherin the absenceof risk. The ex ante effectis the most important of these and explains33% of the growth shortfall: the remaining 13%is attributable to the ex post effect. Elbers etal. (2003)areone of thevery first studiesthatusemicro evidenceto considerthe impact of risk on growth. They model only one production process:fanning with the useof cattle.Yet a striking stylizedfact aboutTanzanianrural householdsis that theyhavehighly diversifiedeconomicactivities,manyof themnon-agricultural.The HBS (2001)reports that rural householdsderive only 48% of theirincome from fanning, of which 16%is obtained from livestock. T4is apart from the fact thatthe fanning operationitself is highly diversified. It is not uncommon to comeacrossfarmersthat grow at leastfive different crops, or that fragmenttheir land holding to dealwith climaticrisk. The remaining 52%is from cash employment(9%), non-farm self-employment(18%),transfersandremittances(13%)and other sources(12%).Suchhighlevelsof diversificationarecosrly.There arestaticcostssuch 8 asincreasedtravel time in the caseof land fragmentationand dynamiccostsbecausethe householdreducesits scopefor learningby doing. Diversification alsomeansthathouseholdsexchangeefficiency for securityby cultivating lessprofitable cropswith attractiverisk characteristics(e.g.droughtresistance;returns with a low or negativecorrelationwith other sourcesof income).The HBS doesnot allow to investigatethis suggestionbut Dercon (1996)discussesfor Shinyangahow farmerswith differential capacitiesfor consumptionsmoothingcopewith risk through adjustmentsin their income process.He showsthat householdswith limited options for smoothinggrow lower return, but safercrops (sweetpotatoes,sorghumand millet) thanthe richer householdswith more options for consumptionsmoothing.The lattercultivatemore of the risky, high return cropslike cotton andpaddy.The costof sucha diversificationstrategycan behigh. Dependingon the areaplantedwith sweetpotatoes,somefarmersforego up to 20% of their income asinsurancepremium. The finding of Dercon is not uniqueto Tanzania.Rosenzweigand Binswanger(1993)have found that wealthierhouseholdsallocatetheir productive assetsto riskier activityportfolios than poorer householdsin somevillagesin India. SimilarlyLarsonand Plessmann(2002)and Morduch (2002)find evidencethat farmersin respectivelythe PhilippinesandIndia choose to diversifyinto lessprofitable crops or chooseto applylessproductive technology. Consequently,householdswhich havehadbad luck for a number of yearswill havelow assetsand thereforelittle scopefor consumptionsmoothing.This maythenleadto a situationwherethe poor opt for anincomeprocessthatgivesthem alow, but safereturn (e.g.casuallabor; growing sweetpotatoes),but that further delaystheir growth out ofpoverty. Those thatwere not strickenby bad luck and that still possesssufficient assetsto smooth consumption canopt for ahigh return, high riskincome process. Which risks? The previous sectionsuggeststhatexposureto risk leadsto lower growth and that risk may be a causefor chronic poverty.This raisesthe question:which risksaremost important? Thereis substantialqualitativeandquantitativematerialavailablethat allowsthe identification of the major sourcesrisk for rural households(pPA 2004;Kessy2004;HBS 2001;Rossi2004;deWaal etal.2004;Christiaensenetat.2004).Thesecanbe summarized ashuman,animaland plant diseases,climaticrisksandprice and otherrisks. Health shocksarepervasivein Tanzania.According to the HBS, 27% of all individualswere sickduring the last fourweeks.In abouta quarterof the casesthe diseasewasso seriousthat the respondentmissedat leastone weekof work or school.Expressedin yearsof life lost dueto prematuredeath,the burden of diseaseis overwhelminglycarriedby childrenaged lessthan five for whom this burdenis more than5 timeslarger(expressedin per capita terms)than for thoseagedfive andabove (URT 2002). Usinga crosssectionfrom Kilimanjaro, Christiaensenetat. (2004)considerthe economic impact of seriousadultillnessand find that it leadsto a reductionin per capitaconsumption of 17%.Using a four yearpanel from Kagera,Rossiinvestigatesthe samefor consumption growth. Sheconfirms the resultsfound for Kilimanjaro and especiallythat chronic illness 9 3. leadsto a 6%declinein consumptiongrowth.Transitorydiseasesand adult deathaffect consumption substantiallyless,probablybecauseits consequencesare alreadycapturedby chronic illness:most adultdeaths(over80%) resultafterchronic illness.The impact of chronic illness on consumptiongrowth is not surprising,given the consequenceschronic illnesseshave for the provision of laborin the household-the ill personis unableto work while othershaveto reallocatelabor from productive activitiesto caretakingandthe consequencesof high medicalexpensesfor the household'sproductive assetbase.Rossialso finds that the poor arelessableto smooththeir consumptionthanthe wealthier. The HBS suggeststhatmalarial fever (65%)anddiarrhea(11%) comprisethe majority of the reporteddiseases.Though the HBS resultsmaynot be the mostaccurate-if onlybecause illnessis a selfreportedvariable,the patternconfinns that from the CoastalSentinelSite (URT 2002)where it is found thatcommunicablediseasesmakeup abouthalf of the total burden of disease,followed by perinataldiseases(16%),nutrition and anemia(12%),non- communicablediseases(10%)andothers.This Coastalpatternis similarto the rest of Tanzania.The importance of communicablediseaseslike malaria,acuterespiratory infections, sexuallytransmittedinfections (including HIV I AIDS) and diarrheaindicatesthat thereremainsanunfinished agendaof largelypreventableconditions to address. Malnutrition and HIV /AIDS requiresomeextra attentionin this context. Malnutrition becauseasa directcauseof mortality it is relativelyuncommon (MoH 2002).Yet it should be appreciatedthat malnutrition is a common underlyingcauseof otherillnessesandeven mortality andlimited economicopportunities laterin life dueto a higher susceptibilityto disease,stuntingandconstrainedintellectualdevelopment.Malnutrition is contractedearlyin life (typicallyduring the first 18months -Figure 5)andis largelypreventable.The condition is lessa consequenceof food insecurity,andmore one of diet,hygieneand caringpractices. Especiallythe frequencyof feedingof youngchildrenhasbeenidentified asmajor problem (UNICEF 1990). Figure 5: Malnutrition in Tanzania:Evolution of weight by agez-scoreby age ~ ~ 0.5 N !f .B t 'U ~ -2.5+ 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 Agein months Source:Kilama and Lindeboom2004.The MDG indicator for malnutrition is the z- weight for agescorebeinglessthan-2. The information in the graphis calculated from datafrom Demographicand Health Surveysfor 1991/92,1996 and 1999. 10 HIV /AIDS requiresextra attention not onlybecausedeathis precededby aprolonged illness-creating alargeeconomicburdenon the household,but alsobecauserural householdsappearto carryadisproportionateshareof the burdenbecausemanypeople infected with HIV /AIDS decideto go home to die (deWaal etal.2004),Caring for children orphanedby AIDS is a major burdenon both communitiesandhouseholds(deWaal et aI, 2004;Kessy2004)contributing to further impoverishment.HIV /AIDS hasmajor consequencesfor the nextgeneration.As orphanedchildrentend to be lesswell educated and aremore likely to be malnourished.According to the 2002Censusorphanedchildren aged7-12,whereasnon-orphanedchildren [to be included: human capital of orphans] Using datafrom Kagera,Alderman etal. (2004)showthat orphanedchildrenare significandymore likely to be stunted or underweight, Farming in Tanzania is almost exclusively rain fed making it inherendy risky. Most farming i~ done in (semi)arid zones where climatic variability is high making weather risk an important source of income variability. Studies in Kilimanjaro [RUVUMA? To be included] and in Kagera suggest that in any given year approximately one third of all farmers is affected by adverse weather. Not only farming is affected by climatic risk. Health shocks, malaria in particular, are more prominent during the rainy season-when the labor demands in agriculture are highest (URT 2002). Kessy (2004) reports how malaria epidemics resulting in large mortality rates amongst young children were associated with extensive rains. Crop disease(coffee berry disease,beans pest, banana diseaseand cassavawilt) and animal disease (Newcasde, tsetse) for chicken, goats and callie feature prominendy amongst the risks mentioned by farmers. Price fluctuationsare anodtersourceof risk. Given dte current slumpin coffeeprices,coffee price risk featuresprominently attlongstdtosegrowing coffee.Remarkably,coffeegrowing householdstend to be relativelyableto smoodt dteir consumptionandto copewidt low coffeeprices,wherebydte most specializedcoffeegrowersexhibit dte greatestresilience. They maintainahigher consumptionleveldtan more diversifiedcoffeegrowers.In part this is explainedby dte fact dtat previouslyhigh coffeepricesallowedcoffeehouseholdsto build up anassetbasewhich is now usedfor consumptionsmoothing(Bevanetal. 1989; Christiaensenetal.2004). 1 Not all price risksaredueto fluctuationsin the world market. In isolatedmarkets,climatic shockscanleadto largechangesin prices.In particularthe livestock/food terms of trade tend to deteriorateafteraclimatic shockbecauseit leadsto anexcessdemandfor food and anexcesssupplyof livestock.The PPA (2004:37)notes for instancehow the price of cows expressedin maizereducesto onethird their normalvalueduring droughtyearsandone twelfth (!) of theirvalueduring extremedroughts.Price risk are especiallypronouncedin isolatedmarkets. [to be included: analysis of price behavior in isolated and non- isolated markets] As Sen(1981)and Ravallion(1997)point out, suchlargeprice fluctuationscan easilyleadto severefood shortages,especiallyamongstnet food purchasers,saythose thatwho earntheir income with cashcrops or in non-farm activities.Considerationsof risk, thereforeinduce householdsoperatingin suchisolatedmarketsto ensuretheyremain largelyself-sufficientin food which thenmakesthem reluctantto enternewattractive (non-food) income earning opportunities. Finallytherearea number of risks thatareworth mentioning astheyfeatureprominendyin qualitativeanalyses.Theseincludegovernancerisks (pPA 2004;Kessy2004)varying from sinsof omission-substandardservicedelivery(clinicsthat run out of medication;absenceof veterinaryservices;extensionworkers that do not showup; roads that arenot maintained)to sinsof commissionsuchasharassmentby governmentofficials or inhibiting rulesand regulations.Theft -particularly of movableassets,livestock,bicyclesand cashalsofeatures relativelyprominendyin qualitativeaswell asquantitativeanalyses. 4. Poverty traps The Ramseymodelpresentedin section2 is optimistic in the sensethat povertyis only a transitoryphenomenon.In the long run everyoneconvergestoward their equilibrium steady state,andif investmentopportunities with sufficiendyhighreturns exist,this steadystatewill layabovethe poverty line. In the Ramseymodelthereis no chronic poverty and shocksdo not havepersistenteffects.Shockspushhouseholdsbackon their accumulationpath, but householdsremain on the growth path theywere onbefore. In practiceshocksdo havepersistenteffects.Whereshockspermanentlyreducehuman capitalfor instance-as it the casefor earlychildhood malnutrition which affectscognitive development,it is this clearthat a one off eventcanleadto permanentlylowerincomes.The life historiespresentedin annex1 provide variousexamplesof lives thatwere permanently changedbecauseof shocks. Though the conceptof apoverty trap is intuitively appealingit is not self evidentwhy, aftera shock,householdsarenot be in a position to climb out of poverty. After all, if investments provide sufficiendy high returns thenhouseholdshavea strongincentive to saveto self- financethe investment.And marginalreturnsto capitalarehigh especiallywhen few assets areavailable,soone expectspoor householdsto havealargeincentiveto saveandinvest. To explainthe presenceof poverty trapsone typicallyrequiresthe existenceof one or more critical wealththresholdsthatpeoplehavedifficulty with crossingfrom below. The presence of a threshold by itself is not sufficientto explainthe presenceof apoverty trap, because evenif investmentsarelumpy householdscould slowlyaccumulatewealthandpurchasethe 12 investmentgood later.Hencethe presenceof a wealththresholdhasto be compoundedby somethingelse,for instancethe lack of safesavingsinstruments,the inability to savefrom low income (becauseof minimum consumptionrequirements)or the malfunctioning of creditmarkets. A typical thresholdareminimum requttementsin nutritional, educationandnon-food consumptionitems suchasclothing dlat areneededbefore someonecanparticipatein dle labor market.Sinceis typicallynot possiblefor destitutepeopleto borrow againstfuture income, oncedestitute,poor peoplemaybe permanendyexcludedfrom participationin dle labor market. Dercon (1997)presentsagood illustration for Shinyangaof apoverty trap thatis brought aboutby a combination of a wealththreshold,imperfect creditmarketsand the inability to savefrom low income. In Shinyangacattlearea high return investment(25-30%annually). Cattleareaalsoliquid assetthat canbeusedfor consumptionsmoothing. For thesereasons, ownershipof cattleis very attractive.But cattlearealsoa lumpyinvestment. Richerrural householdstend to specializein cattlerearing,while poorer householdsderivea largershare of their income from off farm activities.Differencesin comparativeadvantagedo not offer a convincing explanationfor this phenomenon.Householdsspecializingin off farm activities tend to havemuch lowerincomes,but areunlikely not to havethe skillsrequired for cattle rearingasthis is a traditional activityin thearea. Missing creditmarketsand the indivisibility of callie, imply that householdshaveto be able to put up relativelylargeamountsof moneyto investin callie rearing.But householdswith low initial endowmentsfrom which only low incomesareearned,find it hard to saveenough to investin callie.This problemis exacerbatedby the fact thatbecauseof low endowments, the ability to smooth consumptionis limited. Consequendythe poorer householdsenterinto safe-lower return, activities,making it evenharderto save.This combination of factors explainswhy poorer householdsspecializein off farm activitiesthatrequire few skills or investmentbut which aresafe(weedingor casuallabor)andwhich effectivelytrap themintopoverty. This despitethe fact that attractiveinvestmentopportunities existin thearea. Two conclusionsmaybe drawn from the Shinyangaexample.First,poverty trapsdo not cone about easilyandrequirea combination of factors.The elimination of anyof these factors (improved creditmarkets;alternative,non lumpy investmentopportunities with a high return) eliminatesthe povertytrap. Another conclusionis thatin the presenceof entry barriersandmissingcreditmarketsone mayfind that both poor and rich households smoothconsumption.The poor do so by relying on safe-low return income activities;the rich do so by relying on savingsor assets.Suchdualityappearsto be a common featureof rural life. HBS datashowsfor instancehow amongstthe rural householdsthat dependon businessincome for aliving and thatindicatethat theyarehardlyeverfood insecureone finds relativelypoor people(charcoaldealers,sellersof fruit and vegetables)and relatively wealthypeople(maizemillers).The first grouphasa per capitaincome of aboutTShs8,500), the latter one thatis substantiallyhigher(fShs 17,500).Entry barriers(andimperfect credit markets)appearto, agajn,explainthe differencein return. Notably 95% of the millers indicatedto be food secure,whereasthis was only the casefor 75% percentof the charcoal and fruit dealers.So,not evenis milling more secure,it alsoyieldsahigher return-provided one canovercomethe entrybarrier. 13 Thresholds exist for manyreasons.In SouthernTanzaniafor instancewherethereis an abundanceof fertile land, it is exacerbatedby the tsetsefly which preventshouseholdsfrom keepingcattle.Not only doesthis closeoff a popularmeansof accumulatingwealth, it also implies that animaltractionis not an option. As a resultthe entry barrierto increasingthe scaleof the farming operationis raised,becausethe nextavailableoption, tractorplowing, requiressubstantiallymore resources,in terms of moneyandthe clearingof land. Illiquid assetspresentanotherthresholdeffect.If a high return -non lumpy assetcannot be usedto smooth consumptionbecausethereis no secondhand market for the asset,then only the wealthyhouseholdsthat havesufficientendowmentsto smooththeir consumption will be readyto investin it. This alsoexplainswhy one observesthat endowmentpoor farmersspecializein safe,low return crops (sweetpotatoes),and endowmentrich farmersin higherreturn, but riskiercrops (coffee).Again, more developedfinancialmarketscould solve this problem -for instanceif farmerscould mortgagetheir unharvestedcrop. Addressingpoverty trapsis attractivebecausetheinterventionsaretemporaryandthe benefitspermanent.It requires,however,the identification of apoverty trap -which is not alwaysevident,especiallybecausepoverty trapsarealmostalwaysthe resultof a combination of factors,typicallyincluding a threshold,imperfect creditmarketsandthe absenceof safe assets.If the causesfor apoverty trap havebeenidentified, oneneedsto decidewhetherto assistthosetrappedovercomethe threshold effect (e.g.through assettransfers)or the causes of the trap itself. A fewillustrations. Lumpinessin combination with inadequatecreditmarketsmaypreventpoor households from enteringhigh return activities.One responseis thento transferthe assets;a better option maybe to addressthe lumpinessconstraintby improving the meansof capital accumulation,stimulatingROSCAsor through the provision of credit. Lack of property rightsin combination with inadequatecollectiveaction mayleadto an irreversibledeclineof commonproperty resourcessuchasforests or fish stock. Prevention through the promotion of property right or the facilitation of collective actionis clearlythe preferredpolicy intervention here. Adults requirea threshold of physicaland humancapitalto be productive. This capitalis typicallyobtainedduring oncechildhood, whenoneis not yetableto decidefor oneselfor to borrow againstfuture income.This presentanargumentto spendon education,nutrition or assistanceto orphansand homelesschildren.5 For eachof thesepoverty trap relatedinterventionsone canarguethat theyshouldbe consideredinvestments(ratherthanunproductive transfers)becauseto allowhouseholdsto attain their productive potential. But if theyareinvestments,one shouldalso askwhy householdsdo not investin theseactivitiesthemselves.Difficulties to undertakecollective actionexplainthe limited presenceof ROSCAsandproblemswith common property resources.High opportunity costsand discountratesof (poor)parents,andthe inability of 5Remarkably,that it is wise to invest in the educationof childrenis beyond dispute,but is lessacceptedfor nutrition or carefor orphans or (street)children. 14 young children to borrow for their own capacitybuilding affect human capitalaccumulation of children. 5, Policy Implications for Growth and Poverty Reduction The Ramseymodel of sectiontwo suggeststhatall householdswill eventuallyend up in the steadystate. It doesnot suggest,however,that, steadystatecapitalis sufficient for incomes to layabovethe poverty line. In the presenceof insufficiendyrewardinginvestment opportunities, householdshave fewincentivesto saveandto accumulate.Exposureto risk may further reducethe incentivesto accumulateproductive assetsand shocksmay temporarily sethouseholdsback.The impact of risk on growth is suchthat it is worthwhile to reconsiderthe balancebetweeninterventionswhich addressstructuraldeterminantsof poverty andinterventions which reduceexposureto shocks. The discussionof poverty trapssuggeststhat temporary setbacksfollowing shocksmay actuallybe quite permanent.If productive assetsareilliquid and not easilytransformedinto consumption smoothingitems,if investmentsarelumpy or if there existirreversibilities,the accumulationprocessof householdsandindividuals maybe hindered.Consequendysome mayfind themselvesin apoverty trap,becausesomehowtheyendedup with low endowments. This analysispresentsanagendafor growth and poverty reductionin arural context that focuseson enablinghouseholdsto accumulateproductive assetsso that they can"grow out of poverty". Socialtransfersareto beusedonlywherehouseholdshaveno options for growth, for instancebecausetheyarechronicallyill (includingthe elderly)or becausetheir human capitalis low (peoplewith certaintypesof disabilitiesbelongto this group). agendacomprisesof threeelements Increasethe rate of return to rural investmentsandaddresspoverty traps; Reduceexposureto risk through risk prevention or by providing householdswith risk managementinstruments; Promote financial developmentto facilitate capitalaccumulationand consumption smoothingand to offer appropriateinsuranceinstrumentsto dealwith shocks; Belowthe various elementswill be discussedin greaterdetail, 5.1 Increase the rate of return to rural investments A necessarycondition for growth in rural areasto occuris the presenceof investment opportunities whosemarginalrateof return exceedsthe rateof time preferenceof households.6This impliesimproving the investmentclimate.RecendyTanzaniahasmade greatstridestoward attainingthis objective.Price liberalizationin combination with a floating exchangeratehassubstantiallyimproved the rural-urbantermsof tradeand 6It is likely that the poor havehigh ratesof time preference,given that there are fewluxuries to cut out of a consumption bundle thatis alreadyscant.Hence,it maybe that quite a fewprofitable investmentopportunities alreadyexistin rural areas,but that theseare not tappedinto becauseof a lack of accessto credit. 5 The 2.3. increasedthe shareof the world marketprice farmersreceivefor their crop (World Bank 2000). Inflation hasbeenbrought undercontrol (Figure7),local taxeshavebeenabolished etc. Thesepolicy measureshaveimproved the climate for doing businessand help explain the observedincreasesin rural growth (Figure8).7 Figure 7: Inflation in MainlandTanzania Figure 8: Rural GDP growth (monetaryand non-monetary) : \ «lO 8 ~5 ~ 25.0 ,,4 211.0 8 1310 p. 15.0 1<10 ~'n ~O ~~44 QO. 1~ 1m 1m 1m 1m 1m 1~ 1m 1~ 1m 3m ~ :mz :am Source:Economic Survey,2004. Complementaryevidencethat the investmentclimateimproved canbe read from Figures9 and 10which showhow, recendy,the national savingsrateincreased(Figure9),while consumption, expressedasfraction ofGDP declined(Figure10).The modelpresentedin section2 predicts preciselythat this will happenif the returnsto investmentraiseabovethe rate of time preference.Unlike what the modelsuggestsnot all funds for investmentwill haveonly to come from reducingconsumption.Work effort canalsobeincreasedwhich is why in the wakeof improved investmentopportunities,savingsratesmayincreaserapidly (Birdsall,Pinckneyand Sabot1999).The corollary of this is that whenthereare few attractiveinvestmentopportunities,the reversemayhappen.Both the PPA (2004)and Kessy(2004)provide evidenceof this aswell, when they showhow, in the wake of recurrent shocksand with few apparentwaysto improve one'sliving conditions, somepeoplegive Up8 -see alsolife history number ** in the annex. 7Not all macro reform is likely to have sucha quickimpact. It maytake awhile for instancebefore a land market evolvesin full --evenin thoseareaswhereland is alreadyscarce.The absenceof land titling but also perceptions aboutthe likelihood of a reversalof the policy that recognizesprivate ownership of land are likely to be important. 8In extremecasesit may resultin suicide,alcohol or substanceabuseor just apathy.In all instancesit leadsto reducedlabor supply. ~ 16 Figure 9: National savingsrate Figure 10:ConsumptionasFraction of GDP ..,.., !1% 22"/, 89% 18"/, r 87% 85% ~' 1';:7 130/ 83% :;:~::~~ 81% y 79% 77% I ... 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2(XXJ200120022003 Source: Economic Survey 2003 The importance of a soundinvestmentclimate for assetaccumulation-and the responsivenessof rural population to investmentincentives,canbe inferred from micro- economicevidenceobtained from the HBS.Table 1 for instancepresentsa (probit) regressionexplainingthe probability of running abusiness.Apart from the role of distance to aroad (whichwill be discussedlater),the probability thatahouseholdruns abusinessis stronglyexplainedby the fraction of otherhouseholdsin the village that alsorun abusiness Though this still doesnot explainwhy in somevillageshouseholdsrun businessesandin others theydon't, it doessuggestthatin certaincommunitiesthe investmentclimateis substantiallybetter.Wherethis is the case,more householdsarelikely to investin off farm actiVIties. ...9 Table 1: Probit regressionof running a firm --Coefficient T-stat (abs) Distanceto market(km) -0.0057 3.40 Distanceto nearestbank (km) -0.0003 1.99 Fraction of householdsin 0.8285 42.6 community with business D-rural -0.0332 3.47 Note: The dependentvariableis oneif the householdowns abusinessand zero otherwise.The coefficients give the marginal effect of an infinitely smallchangein the continuous variablesand a setting from 0 to 1 for the dummyvariable.Distancesis the mediandistancefrom the community to the market/bank To avoid endogeneity,the fraction of householdsin community with a businessexcludesthe householdof interest.Regional (20)dummiesareincluded in the regressionbut not reported.The pseudoR-squaredis 0.10.The total number of observations: 22154.Data source:fIBS 2001 Though macro economicreforms seemto havehada beneficialeffect on the profitability of rural investments,there still is alargeunfinishedagendaof micro-economicreforms that would enhancethe profitability of rural investments.This canbe illustrated by considering prices.The HBS doesnot comprisepricesof agriculturalproduce, but it doeshave information aboutprices of consumergoods.ConsiderTable2. It presentsavillagelevel 9Clearly,this observationmakesit worthwhile to explore which arethe factors driving the investmentclimate. 7 'e::/ 1 regressionof the medianprice of maize flour -which is consumedthroughout the country. The regressionshowsthat the price of maizeis highlyexplainedby the regionin which one resides,with householdsin the Lake Zone, CentralZone and SouthernHighlandspaying more than 40%more thanthoseresidingin the Northern Highlands. Within zonesprices areaffectedby the distanceto the nearestmarketbut this effectis much smallerthanthe inter-zonal price differences. Table 2: Price regressionfor maize flour Coefficient T-stat (abs) Distanceto shop (km) 0.0302 2.58 Distanceto shopsquared(km) -0.0007 2.20 D-Coastalzone 0.0511 2.86 D- Northem highlands -0.1049 3.90 D- Lake zone 0.2156 10.46 D- Central 0.2810 8.17 D-Southem highlands 0.2790 7.96 D-South -0.0629 2.88 D-rural area 0.1952 9.23 Constant 1.4659 154.2 Note: The dependentvariableis the log of the villagemedianmaize flour price. The omitted categoryis Dar esSalaam.The R-squaredis 0.25.No of observations 1154(psulevel regression).The zonal dummiescompriseof the following regions:Coastal:Tanga,Morogoro and Pwani; Northern highlands:Arusha and Kilimanjaro; Lake zone Tabora, Kigoma, Shinyanga,Kagera,Mwanza and Mara; Central:Dodoma and Singida;Southernhighlands: lringa, Mbeyaand Rukwa;South: Lindi, Mtwara andRuvuma Though someof theseprice differentialsmaybe explainedby distancefrom consumers marketsto producermarkets,the sizeof the price differential for sucha bulk good suggest highly fragmentedmarkets.[to be included: breakdown of price inequality in inter and intra regional component] The supplychainanalysisby Kopicki (2004)providesinsights in the kind of micro-economicimpedimentsthat existto a well functioning maizemarket, including: monopolistic practicesamongsttraders;the need for governmentapprovalto import and export (andthe institution of export bansin yearsof food insecurity), limited warehousecapacityto storemaize,absenceof gradingfacilitiespreventingstoredmaizeto be usedascollateral, marketdistortions throughunpredictableinterventions in the market by the StrategicGrain Reserve(SGR)[to be expanded]. Riskitself is sometimesamicro-economic impedimentto a soundinvestmentclimate.For instance,within the lendingportfolio the CRDB Bankin Tanzaniafor loansto coffee and cotton producerorganizations,cooperatives,and ginnersconstitute about $25 million. Theseorganizationshave significantprice risk exposurefor aperiod of 8-10months during the sellingseason,makingit the cotton andcoffeeportfolio the riskiestfor the Bank. Risk 18 this actsasconstraintto expandingthe loanportfolio and,hence,limits competitionin the sector.10 The micro-economicimpedimentsto a soundinvestmentclimatearecompoundedby inadequateservicedeliveryin infrastructure,health,educationor agriculturalextension. Constrainedaccessto extensionserviceslimits the adoption of new,high return, crops. Understaffed healthfacilitiesanddrug stockouts exposehouseholdsunnecessarilyto disease.The incentivesfor parentto investin educationarelimited if the quality of educationis low. Yet in numerousrural districtslessthan 20%of the Grade 7 students passedtheir schoolleaver'sexam.ttInfrastructureis anotherareawherethereis considerable underdelivery. The importance of infrastructure canbeinferred from World Bank(2000:21) which investigateswhethermaizepricesaremore affectedby domesticsupplyand demandor by world prices.The studyfinds thatmaizebehaveslike a non-tradablein isolatedmarketsand like a tradablein marketsthatarewell accessedby rail or road.This resulthasimmediate implications for growth, becauseif it is correctthat food marketsareisolated,then producerswill be constrainedin their abilityto switchto more profitable income opportunities becauselessfood production would immediatelybid up the price for food and thuserodethe gainsfrom switching. Support for the thesis that access to roads matters can be found in Tables 1 and 2.0ther evidence comes from a.o. Kessy (2004: 34) who reports in her qualitative study on income mobility in Kagera that: (~..j the banana market is also aproblem. During a village transectwalk, the researchteam obseroedsomebananas leftto rot in thefarms. This wase.pecialfythe casewith Matoke, the staple tYPeof banana that is not usedin local brewing. Furthermore, re.pondentsmentionedthat sometimes maize is usedtofeed chicken becausethereis no market. The marketfor cottonis also aproblem but the situation is improving thoughprivate traders visiting thevillage.A marketfor tobaccois readilY available." Alderman et al. (2004), who investigate the impact of program interventions on the nutritional intervention of children under five find a considerable positive effect of living in a community near a motor road. And de Weerdt (2003) in his study of the adoption of improved banana varieties reports how this is less successful in areas with limited marketing ..12 opporturutles. 5.2 Managing Risks Sectionthreearguedthat climaticrisks,exposureto diseaseandprice risksare amongstthe majorrisksaffecting rural householdsin Tanzania.An excitingimplication of this finding is 10Recendythe bank hasdevelopedthe ability to enterinto price hedgecontractson its client's behalf. This allows the bank continue lending to the two sectors-this enhancingcompetition, while alsoconsideringto lowering interestrates for borrowers who arehedged. 11In the following districts in 2003lessthan 20 percentof pupils thatwere enrolled in gradeone passed: Kiteto, Igunga,Iramba, Nzega,Sumbawanga(R),Ulanga,Ruangwa,Tabora (R),Rufiji, Liwale, Pangani, Handeno and Simanjiro (The Citizen3January2005). 12The availableevidencethusseemsto suggestthat roadsand higherincome are associated,yet it is not clear about whetherinvestments in roads alonewould be sufficient to improve the investmentclimate or whether complementarymeasuressuchasreducingmicro-economic impediments arerequired. 19 that many of theserisksarepreventable.Communicablehuman diseases,which comprise the bulk of the burden of disease,canbeprevented.The sameholds for malnutrition, animal diseasesand crop pests.This meansthat,potentially,a considerablepart of the ex ante and ex antepost costsof exposureto risk canbe avoided.Going by the resultsof Elbers etal (2004)this hasthe potentialto considerablyincreaserural growth. An implication of this finding is that healthandagriculturalsectorexpensesthatdealwith the prevention of these diseasesshouldbe consideredinvestmentsin growth, rather thanunproductive outlays (World Bank2001). Climatic risks cannot beprevented,unlessthereis alargescaletransitionto irrigation. Climatic shocksare often compoundedbylargechangesin prices.In particularthe livestock/food terms of tradetendsto deterioratebecausethereis anexcessdemandfor food and anexcesssupplyof livestock.To the extendthatprice risksarecausedby isolation, improved infrastructureand accessto marketswill reducethe impact of the combined climatic-price shocks. By enhancinghousehold'scopingmechanismslossesdueto shockswill not reduce,but their impactin termsof welfarewill diminish.This thenallowshouseholdsto capturesomeof the ex antebenefits of reducingrisk on growth. Household copingmechanismswill be enhancedby someof the measuressuggestedabove:reducedexposureto animaldiseaseand improved livestockmarketsmakesit more attractiveto keeplivestock for consumption smoothingpurposes.Other measuresmaybe considered,especiallythose thatreducethe costof assetaccumulation,for instancebyproviding householdswith accessto savings accountsor by reducingthe costof storingfood. This maybe done through improving on farm storagefacilities or throughwarehousestoragearrangements.13 Pricerisks originating from the international commodity marketscannot beprevented either. For along time theTanzaniangovernmentintervenedto reduceriskin the market for internationallytradedgoods.Thesestabilizationpolicieshavebeenfound to be not only ineffective andunsustainable,theyhavealsoactedasimpedimentsto growth (Larsonetal. 2004;Bevanetal. 1986).Currendythe thinking is on managingrural price risks through marketbasedinstruments.Thesearediscussedin the next section. 5.3 Improved financial markets Financialmarketsplaya crucialrole in facilitatinggrowth. The modelpresentedin section2 illustrates,for instance,ilie inlportance of accessto safemeansof assetaccumulationfor growth. The empiricalillustrations showiliat ilie benefits from mitigating risk through insurancemarketsarelargebecauseilie ex anteeffectsof risk on growili is avoided.And inlproved accessto creditgoesa long wayin overcomingilie threshold effectsiliat often underlaypovertytraps. Unfortunately accessto insuranceis virtually absentin rural area.The sameholds for the use of savingsaccountsandaccessto credit. At ameandistanceto a bank of more than30 km the limited useof savingsaccountsand formal creditis unsurprising.And with the 13Warehousearrangementssometimesallow farmersto enterinto sophisticatedrisk managemen arrangements.SeeLarson etal. 2004for a review. 20 restructuringof the banking sectorduring 1990s,limited accessto fonnal savingsdeclined even further andwasnot beencompensatedby anassociatedincreasein informal savings (Table3). Table 3: Accessto savingsservicesin rural areas Savingsor current account 1991 InfOmlal savingsgroup 2000 Source: HBS 1991& 2000.Standarderrorsin parentheses Where the lossin accessto formal savingsincreasedthe obstaclesto assetaccumulation,low inflation meantthat cashbecomeamore attractivestore of wealth.Kessy (2004)reports how cashsavingsareanimportant meansfor consumptionsmoothing.And Christiaensen (2004)notes that in Kilitnajaro monetarysavingsarethe most important meansof coping with the coffeeprice shock.14 Despite the greatadvantagesthat functional rural financialmarketswould bring thereare structuralreasonswhy financialmarketsin rural areasarehighlyimperfect. Geographic isolation, moral hazardandhigh costof information precludeforms of insurancethat cover actuallosses.Largefixed costsandvolumesmakefuturesmarketsunavailablefor small farmers.And the absenceof collateralpreventcreditmarketsfrom developing.The credit andinsurancethatis provided is mosdyinformal andbasedon high observabilityand repeatedinteraction. This doesnot mean that formal forms of financialservicescannot be developed.Greater accessto formal savingsmechanismsarepossiblefor instanceby relying on local institutions sucha funeralgroups that alreadymanagefinancialresources(seeDercon etal.2004). Mobile banks or cellphone technologyalsopresentpossibilitiesto increaseaccessto savings.There arenumerousotherinitiatives that try to overcomethe constraintsto financial marketsby focusingon group responsibilityto overcomecollateralconstraints (micro-creditinitiatives); by offering insurancecontractsagainstindexes(asis the caseof weather,basedinsurancecontracts),or by organizingrotating savingschemes(ROSCAs)to dealwith lumpinessin investments. The literature on thesevarious arrangementsis adequatelysurveyedelsewhere(e.g.Larson et aI.2004,for a recentreview).Sufficeit to saythat financial marketsdo matter, that great gainscanbe reapedif they canbeimproved -even marginally,and that therearevariousvery promising initiatives in this areathat warrantsolid attention. 14Both Christiaensenand Kessyalso report that cash/savingsareusedin the more affluent villages,but that in poorer communities householdslackthe monetaryreservesto do so. Presumablythe opportunity costs in terms of foregone consumption aretoo high for poorer householdsto keep (unproductive) monetarysavings. 21 6. ConclusionsandAgendafor Research Starting from the observationthat low income andexposureto risk presenta casefor growth enhancingpolicesand transfersto the poor, this paperinvestigatesthe impact of risk on growth. It is found that the impact of risk on growth is considerable.Exposureto risk hasthe potential to reducethe steadystatecapitallevelsand consequentlygrowth becauseof ex ante responses-agents opt for lower risk exposureby adaptingtheincome process,andthe ex post consequencesof shocks.The presenceof poverty trapscreatesa casefor transfersthat allow agentsto attain their growth potential.Thesetransfersaretemporaryand shouldbe consideredinvestments.Many of themarealreadyincorporatedin the healthsector (nutrition interventions), educationsector(PEDP) or agriculture(animaldiseaseprevention). Climatic andprice risksandhuman,animaland crop diseaseareidentified asdle main sourcesof risk. Widl respectto diseaseit is found dlat manyof dlesearecommunicable, suggestingconsiderablescopefor risk prevention.And asoutlaysto preventsdleserisks enhancegrowdl, dley shouldbe consideredinvestments(andappraisedassuch)radler dlan unproductive transfers.Climatic risksaredifficult to prevent, but to dle extentdlat dle consequencesof weadlerandprice risksarecompoundedbecauseof dle presenceof thin markets,improving infrastructure and marketaccessmaygo along wayto reducedle severityof weadlervariability.Priceriskscausedby externalfactors cannotbeprevented,but newlydevelopedinsuranceinstrumentsmayhelp farmersmirigatedleserisks. This paperstartedof with the observationthat thereis dualityin the debatedaboutwhich policiesto follow in a contextcharacterizedby low income,poverty andrisk. Those that addressthe structural determinantsof poverty (growth)andinterventions which reducethe consequencesof shocks(transfers).This paperproposesto reconsiderthe balancebetween theseinterventions and to focus on policies that (i)improve the investmentclimate (ii) preventrisk and (iii) promote financialmarkets. This alsosuggestsanelaborateagendafor research.For instance,what arethe important micro-economicimpedimentsto growth?What poverty traps canbeidentified and how are they bestaddressed?Which (new)approachesto credit,savingsandinsuranceare successful and readyfor scalingup? How much risk prevention canbe deliveredthrough the current water,health,educationandagriculturalsectors?And how canthe deliveryof theseservices be improved so that exposureto riskis reduced? 22 References Bevan D., P. Collier and l.W. Gunning 1989. PeasantsandGovernments:An EconomicAnafysis. Oxford. Clarendon. BinswangerH.P. andJ. McIntire 1987.Behaviouraland MaterialDetenninants of Production Relationsin Land Abundant TropicalAgriculture. EconomicDevelopmentand CulturalChange36(1):73-99. ChristiaensenL., V. Hofmann andA. Sarris2004.Coffee Price Riskin Perspective: Vulnerability Among SmallHolder Coffee Growersin Tanzania.UnpublishedManuscript. World Bank. WashingtonD.C.. Dercon S.1996.Risk, Crop Choice,andSavings:Evidence &om Tanzania.Economic DevelopmentandCulturalChange44(3):485-513. Dercon S.1998.Wealth, RiskandActivity Choice:Cattlein WesternTanzania.Journaloj DevelopmentEconomics55(1):1-42. Dercon S.andP. Krishnan 1996.Income Portfolios in Rural Ethiopia andTanzania: ChoicesandConstraints. JournalofDevelopmentStudies32(6):850-875. Dercon S.,T. Bold,J. de WeerdtandA. Pankhurst2004. ExtendingInsurance?Funeral Associationsin Ethiopia andTanzania.UnpublishedManuscript.Mimeo. Elbers C.,].W. Gunning andB. Kinsey.2003. Growth andRisk. TinbergenInstitute DiscussionPaper.TI 2003-068/2 KessyF.2004.Rural Income Dynamicsin KageraRegion,Tanzania.ESRF.Dar esSalaam. Kopicki R. 2004.SupplyChainAnalysis.UnpublishedManuscript.The World Bank. WashingtonD.C.. LarsonD. andF. Plessmann2002.Do FarmersChooseto be Inefficient? Evidence from Bicol, Philippines.Policy ResearchWorking Paper2787.World Bank, WashingtonD.C. MorduchJ.2002. ConsumptionSmoothingAcross Space:TestingTheories of RiskSharing in the ICRISAT StudyRegion of SouthIndia. WIDER DiscussionPaper2002/55/ United Nations University,World Institute for DevelopmentEconomicsResearch,Helsinki. United Nations Children's Fund (UNICEF). 1990-Women and Childrenin Tanzania:An Overview. UNICEF, Dar esSalaam. United Republicof Tanzania2002.Burden of DiseaseProfile 2002-Coastal Zone- Ministry of Health, Dar esSalaam United Republic of Tanzania 2002. HouseholdBudgetSurvey2000/01.. Final Report.Dar esSalaam: National Bureau of Statistics 23 United Republicof Tanzania2004.National Strategyfor Growth and Reduction of Poverty. Vice President'sOffice. Dar esSalaam.UnpublishedManuscript. United Republicof Tanzania2003.BasicStatisticsin Education1999-2003.National Data. Ministry of Education andCulture,Dar esSalaam. PPA 2004. Ravallion M. 1997. Famines and Economics. JournalofEconomicLiteratun 35: 1205-1242. Sen,A. PovertyandFamines1981.An Ess'!YonEntitlementandDeprivation.Clarendon Press, Oxford. Ward P., etal.2004.Socialprotection: Assuring a minimum income for the poorest. Unpublished Manuscript. Dar esSalaam. De WaalA.,J. Tumushabe,M. MamdaniandB. Kilama 2004.ChangingVulnerability to Crisisin Tanzania.Implications for Childrenand UNICEF Activities. Unpublished Manuscript.UNICEF. Dar esSalaam. World Bank2001.SocialProtection SectorStrategy:from SafetyNet to Spingboard,Sector StrategyPaper.The World Bank. WashingtonD.C.. 24 THE BENEFITS OF MALNUTRITION INTERVENTIONS: EMPIRICAL EVIDENCE AND LESSONS TO TANZANIA Adolf F. Mkenda Department of Economics University of Dar es Salaam P.O Box 35045 Dar es Salaam TANZANIA Phone:255-22-2410661 or 255-22-2410500/8 ext. 2262 Email: amkenda@udsm.ac.tz 1. Introduction This study reviews both the economic and non-economic benefits of malnutrition interventions in Tanzania. The review covers international literature, particularly literature dwelling on empirical evidence and estimations and relates these to the Tanzania situation. The review draws policy lessons to Tanzania on the rationale, relevance and types of malnutrition interventions that may be beneficial to Tanzania particularly in the quest to attain the Millennium Development Goals and in poverty reduction efforts in general. For a long time, economic analysis assumed that welfare could be adequately captured by money metric measure. As a result of this, welfare in general and poverty in particular tended to be measured in terms of consumption at the household level and by the per capita income at the national level1. The justification for this has been that most other welfare indicators tend to be positively correlated to the money metric measure: households with higher consumption tend to also have more educated members, live longer, suffer less morbidity, enjoy better nutrition and so on. Also, countries with higher per capita income tend to have citizens with a longer life expectancy, less malnutrition, higher literacy and fare better in all other indicators of human welfare. In general and in very broad terms, this has tended to remain empirically true, but there are sufficient number of exceptions to this general and broad trend to merit an investigation of the connection between the money metric measure of welfare and other widely accepted indicators of human welfare and to explore these exception cases. In this sense, there is a need to explore the connection between income growth and nutritional achievement both at the household level and across nations. Can income growth sufficiently resolve the problem of malnutrition in Tanzania? This is the question dealt with in section two of this study. Section two of the study also explores empirical evidence on the contribution of improved nutrition on economic growth. For growth enthusiasts, evidence that improved nutrition does actually contribute to higher economic growth would be a sufficient, if not the only justification, 1Dasgupta (2001) makes a distinction between the constituents of well being such as good health, freedom and happiness and the commodity determinants of well-being such as income, food, clothing shelter and portable water. He points out that using determinants to measure well-being has the advantage of indicating the institutions best suited to serve human interest. He also pointed out that it is easier to deal with mostly observable determinants than largely unobservable constituents of well-being (see pages 54 and 55). 2 to support nutrition interventions as part of macro and micro economic effort to promote growth. Existing empirical evidence on the contribution of improved nutrition on growth must thus be brought to bear in the thinking on nutrition policy particularly for a country such as Tanzania where there is such urgency for promoting growth as a long-term solution to under-development and poverty. To this end therefore, section two of this study explores the empirical evidence of the contribution of improved nutrition to economic growth through both the immediate and direct improvement in productivity, and the indirect longer-term improvement in productivity mediated through improved cognitive power and educational attainment. While there are significant advantages in urging for the improvement in nutrition as an instrument for economic growth, the significance of nutrition as an end in itself need not be ignored; it in fact needs to be prominently recognised in policy formulation. It has been argued persuasively by Sen (1999)2 that development is a removal of various forms of un-freedoms, such as malnutrition, morbidity, `premature' mortality, illiteracy and so on. The poverty reduction strategic paper of Tanzania also mentions malnutrition as one of the forms of non- income poverty (URT, 2000). This review therefore explores nutrition as a valuable achievement in itself and highlights the empirical connections between nutrition and other forms of (`non- economic') human achievements such as reduction in morbidity and illiteracy. This particular review is taken up in section three. Section four of this study summarizes the economic rationale for government interventions to reduce malnutrition. The section draws from the existing empirical evidence to estimate the costs of not tackling the malnutrition problem and to simulate the impact of economic growth on nutrition. Based on empirical evidence, the study concludes by offering some arguments in terms of cost benefit analysis of malnutrition interventions in Tanzania. 2 This view contained in Sen's booklet titled "Development as Freedom" is shared by Nyerere's view in a book titled "Freedom and Development", one of the writings that inspired policy thinking in Tanzania for the later part of 1960's and 1970's. 3 2 Income Growth and Nutrition: Some Empirical Evidence and Lesson to Tanzania The question that we seek to answer in the section is whether economic growth is related to nutrition. We also need to establish the direction of causality, if it exists. The importance of investigating this is that we need to know whether economic growth can be relied upon to sufficiently resolve the malnutrition problem in Tanzania. Moreover, we also need to know whether investing in nutrition is good for economic growth. Given the current emphasis on poverty reduction, it is interesting to know whether reduction in "income" poverty would be sufficient for reducing malnutrition. A correlation coefficient between poverty and malnutrition can give an indication of what to expect. Table 1 gives correlation coefficients between poverty and malnutrition across twenty regions of Tanzania. The coefficients are obtained from data reported in Table A1 in the appendix of this report. An interesting result here is that the correlation coefficient between moderate stunting and poverty is positive, ranging from 0.3 with respect to poverty incidence to 0.4 with respect to poverty gap and poverty severity. Though no direction of causality is implied by correlation coefficient, a general idea from Table 1 is that reduction in poverty may contribute to the reduction of stunting by up to 40%. There is of course the remaining 60% unaccounted for, suggesting that there are other factors that may explain reduction of stunting. The correlation between wasting and poverty is small but counter intuitive, highlighting the sometime tenuous relationship between income and nutrition obtained in empirical studies. In the recently released Poverty and Human Development Report, it is shown that stunting and underweight are related to income quintile; the richer the quintile the less the malnutrition suffers (URT 2004, page 38). Nevertheless, it is also reported that the lowest quintile suffers less malnutrition than the second poorest quintile, a clear indication that the relationship between income and malnutrition is not always straight forward. Notable also is the fact that even the richest quintile in Tanzania suffers significant incidence of malnutrition of slightly above 20%, suggesting that income alone may not suffice for resolving the problem of malnutrition. 4 In general, not much has been done to relate income and nutrition or poverty and nutrition in Tanzania. The regional data given in Table A1 does not offer a sample large enough to undertake econometric analysis. We will therefore review empirical evidence from other countries also to buttress the little empirical evidence that is available in Tanzania and make policy inferences for Tanzania. Table 1: Correlation Coefficient Between Malnutrition and Various Indicators Across Regions in Tanzania Index of the % Women % Adults severity % Womenwith % Adultswith Mean Index of theof Moderate Moderate with noprimary orwith noprimary orhousehold Incidence ofpoverty gappoverty stunting wasting education more education more size poverty (P0) (P1) (P2) Moderate Stunting 1.0 -0.1 0.1 -0.2 0.1 -0.2 -0.5 0.3 0.4 0.4 Moderate Wasting -0.1 1.0 0.3 -0.2 0.2 -0.1 0.2 -0.2 -0.1 0.0 % Women with no education 0.1 0.3 1.0 -0.9 1.0 -0.9 0.3 0.3 0.3 0.2 % Women with primary or more -0.2 -0.2 -0.9 1.0 -0.9 1.0 -0.3 -0.3 -0.3 -0.2 % Adults with no education 0.1 0.2 1.0 -0.9 1.0 -0.9 0.1 -0.1 -0.1 0.0 % Adults with primary or more -0.2 -0.1 -0.9 1.0 -0.9 1.0 -0.2 -0.4 -0.4 -0.3 Mean household size -0.5 0.2 0.3 -0.3 0.3 -0.2 1.0 0.0 0.0 0.0 Incidence of poverty (P0) 0.3 -0.2 0.3 -0.3 0.4 -0.4 0.0 1.0 1.0 0.9 Index of the poverty gap (P1) 0.4 -0.1 0.3 -0.3 0.4 -0.4 0.0 1.0 1.0 1.0 Index of the severity of poverty (P2) 0.4 0.0 0.2 -0.2 0.3 -0.3 0.0 0.9 1.0 1.0 In this section, the documented empirical relationship between level of income and the status of nutrition is discussed. First, the empirical evidence on the extent that level of per capita income leads to improvement in nutrition is reviewed. A review on the extent that improvement in the nutrition status leads to improved productivity and income follows. In either case, the relationship between income and nutrition is investigated both at the household level and at the national level. The econometric and other empirical problems associated with the estimation of 5 the relationship between income and nutrition are also discussed at each stage. The section concludes by drawing a lesson to Tanzania from the existing evidence of the relationship between income and nutrition. 2.1 How Much Does Income Growth Reduce Malnutrition? Growth leads to higher per capita income. There are two ways that higher per capita income can lead to better nutrition; such incomes increase demand for more and nutritionally better food. Higher incomes also mean that the government can collect more revenues to spend on nutrition- improving programs. Apart from low income, there are other underlining causes of malnutrition such as household food insecurity, inadequate public health provision and poor social and care environment (see Young, 2001). Some of these other factors have been included in the regression analysis of income and nutrition. 2.1.1 To What Extent Does Household Income Reduce Malnutrition? We will start with a review of empirical evidence between household income and nutrition. An econometric estimation of the relationship between household income and nutrition has been done by, among others, Aderman et al (2000), Bouis and Haddad (1992), Strauss and Thomas (1995), Dawson and Tiffin (1998), Tiffin and Dawson (2002), Glewwe et al (2002), Abdulai and Aubert (2004a) Abdulai and Aubert (2004b) Subramanian and Deaton (1996) and Hoogeveen (2004). The empirical analysis has ranged from the estimation of demand for food and nutrients (e.g. Abdulai and Aubert 2004a) to the estimation of the determinants of nutrition (Glewwe et al 2002). Estimation of the determinants of nutrition generally involves using an index of nutritional status of children as the dependent variable, and income and other variables as explanatory variables. In a bivariate regression between nutrition and per capita income, it has generally been found that variation in income explains about half of the variation in nutrition (see Svedberg, 2004). There are a number of estimation problems with the bivariate regression of nutrition and income. In particular, the estimated coefficient on income is likely to be biased and inconsistent due to 6 failure to take into account the two-way causation between income and nutrition. Still, the bivariate results bring forth some interesting question: what explains the remaining 50% of the variation in nutrition? This is clearly a compelling issue: it is not all that easy to increase incomes, and that makes it equally difficult to reduce malnutrition by income growth. The finding that income growth itself does not explain all improvement in nutrition must thus be rather sobering because it refutes the proposition that one can take care of growth, and then nutrition will be adequately taken care of. On the other hand, it is encouraging to know that there may be factors other than income that may contribute significantly to reducing malnutrition. It is important that the other factors that explain the remaining 50% of the change in nutrition are known and well explored to shed light on interventions that may hasten the reduction of malnutrition. We need to investigate two issues: (i) the factors that explain the remaining 50% variation in the nutrition and (ii) the steps necessary to remove the bias in the parameter estimates and deal with other estimation problems. Other possible explanatory variables for the variations in nutrition that have been used include; parental education status, social services provision (such as sanitation, basic health care, clean water), total fertility rate, dependency ratio and parental genetic influence. However, adding more explanatory variables in the regression equation has tended to reduce the estimated impact of income on nutrition. Introduction of these explanatory variables may also create the problem of multicollinearity, a problem that may make it impossible to accurately disentangle separate impacts of each explanatory variable on the dependent variable. In an on going study, Hoogeveen (2004) found that the logarithm of per capita household expenditure positively affects the nutritional status of the child in the Kagera region of Tanzania. He found that a 1% increase in per capita annual expenditure would generate an increase of the z-score of the child equivalent to 0.16, an increase that declines with the share of farm income in the household. This study also finds that feeding programs at the village level and the proximity of the household to passable road improve a children nutritional status. These are interesting results particularly since they are obtained from Tanzanian data. At this stage, the study indicates that apart from income, a number of other factors explain a child's nutritional status. Since this is 7 an on going study, it is prudent to wait the final results, particularly after econometric issues such as potential simultaneous bias are dealt with. Moreover, since the data set is panel, a scope of dealing with unobservable fixed effect exists and needs to be exploited. Glewwe et al (2002) offer an interesting empirical study on the influence of household income on child's nutrition status using Vietnamese household data. The motivation of the study is to find out why an earlier study had indicated only a weak relation between household income and child nutrition even though the rapid economic growth in Vietnam had been associated with a decline in the incidence of stunting among children. The study makes use of several approaches for dealing with econometric problems; e.g. instrumental variables to tackle the simultaneity bias and measurement errors in the variables, and fixed-effects estimation to handle unobserved fixed effects. Glewwe et al (2002) found that the impact of household expenditures on children's nutritional status (measured by height for age z-score) is not necessarily significantly different from zero, and where positive impact is found, it is still not large enough to account for even half of the measured improvement in children's nutritional status in Vietnam from 1992-93 to 1997- 98. Tentatively, the study found that community health service had a positive impact on children's nutritional status. Haddad at al (2003) examined the impact of income on nutrition using household budget surveys from 12 developing countries. Using instrumental variable approach, they control for both the simultaneity bias and the possible errors in the variables. They use the z-score measure of nutrition as the dependent variable and regress this against the log of per capita household income, educational level of the child's mother and father, age, type of drinking water and toilet used in the household and, where appropriate, dummies for ethnic background. They found that income has a significant positive impact on a child's nutrition and that correction for the simultaneity bias only serves to increase the impact of income! In general, they found that on average, doubling household income increases weight for age of a child by half a standard deviation from the median for the reference population. Another way of looking at the impact of household income on nutritional status is by estimating how changes in household income influences demand for calories and nutrients. The idea is that 8 if it can be shown that an increase in household income significantly increases household demand for calories and nutrients, we can then infer that household income positively influences nutrition, and that economic growth (one that favours the poor or at least is distributionally neutral) can significantly reduce malnutrition. Studies along this line include Behrman and Deolaliker (1987), Subramanian and Deaton (1996), Dawson and Tiffin (1998), Abdulai and Aubert (2004a) and Abdulai and Aubert (2004b). We will review two studies on the demand for food and nutrients that uses Tanzanian data. These are Abdulai and Aubert (2004a) and Abdulai and Aubert (2004b). These two studies use household budget survey data collected in Dar es Salaam and Mbeya in 1998 and 1999 covering about 500 households. The average per capita household expenditure in this data is Tshs. 27,600 per month. This figure is substantially higher than the mean monthly per capita household expenditure of Tshs. 10,120 for Tanzania Mainland found in the nationally representative household survey conducted by the National Bureau of Statistics in 2000 and 2001 (URT, 2002). In fact, even the mean per capita expenditure for Dar es Salaam estimated from the national survey is still low, at Tshs. 21,949. These marked differences between monthly consumption between the two surveys is worrying, particularly since one would in fact have expected the monthly consumption in 1998/1999 to be slightly lower than the one obtained in 2000/2001 given the positive economic growth registered from 1998 to 2001. With this in mind, we can proceed to assess empirical results obtained by Abdulai and Aubert (2004a and 2004b) In both studies the authors deal fairly well with problems of functional specification, measurement errors in the variables and the potential simultaneous bias in the estimation. Their estimation results therefore seem fairly credible with respect to the econometric theory. In Abdulai and Aubert (2004a), the calories-expenditure elasticities are found to be above 0.4. The results show that a 1% increase in per capita expenditure results in a 0.49% increase in calorie consumption. These results indicate that income growth will contribute to the reduction in calorie deficit and thus reduce malnutrition. The study also finds a negative relationship between food prices and calorie intake. This is an indication that food shortages that lead to higher prices increase malnutrition. 9 In Abdulai and Aubert (2004b), demand for food and nutrients is estimated using a Quadratic Almost Ideal Demand System. The study carefully analysed the nutrient content of food consumed by the household and thus managed to calculate the expenditure elasticities of calories, protein, iron, zinc, calcium, vitamin A and vitamin B12 .The expenditure elasticities of calories, proteins and micronutrients are presented in Table 2. Table2: Expenditure Elasticities of Calories, Protein and other Nutrients* CALORIES PROTEIN IRON ZINC CALCIUM VITAMIN VITAMIN DEPENDENT A B12 VARIABLE IN LOGARITHM Logarithm of 0.426 0.439 0.307 0.468 0.644 0.383 1.257 Household Per Capita Expenditure: Coefficients t-Statistics 5.07 6.28 4.68 7.32 7.21 2.98 10.84 R-Squired 0.236 0.267 0.241 0.311 0.208 0.127 0.393 Source: Adopted from Abdulai and Aubert (2004b). Note that other explanatory variables are not shown here for the sake of brevity. The expenditure elasticity of calories is not markedly different from that obtained in Abdulai and Aubert (2004a). The estimated expenditure elasticities for nutrients range from 0.307 for iron to 1.26 for vitamin B12. These results show that demand for calories and micronutrients significantly increases with income. The policy implication of this finding is that a well- distributed growth is one of the effective ways of tackling malnutrition in Tanzania The authors also estimated the impact of some other social economic variables on the demand for nutrients. They further found that food prices are negatively related to the demand for nutrients, indicating that price policy may still be effective in promoting nutrition. The use of food price as a policy lever to reduce malnutrition can however prove problematic given the fact that farmers who are generally the poorest also depend on higher food prices for their incomes. Two things can be said about the estimated expenditure elasticities for food in Tanzania. First, the results are based on a limited sample that is not necessarily nationally representative. The 2000/01 household budget survey data offers an opportunity for calculating these elasticities 10 using a more representative data. Deaton (1988) and Deaton (1990) has developed an approach for calculating price elasticities from cross sectional survey data; thus even price elasticity may be calculated without estimating a system of demand functions. It will be useful to obtain income and price elasticities of calories and nutrients using the nationally representative data. The second thing to note is that even though the results indicate that growth in household income would increase demand for calories and nutrients, it is important to remember that generally, it is difficult to obtain annual income growth of more than 5% and this may not be sufficient to induce the desirable reduction in malnutrition. It is also important to find out the extent that macroeconomic growth reduces malnutrition. After all, macroeconomic growth is not necessarily neutral to distribution (it can actually increase inequality and poverty) and thus may not automatically translate into increased demand for nutrients by poor households. This is the matter taken in the sub-section below. 2.1.2 How Much Does Income Growth Reduce Malnutrition? Cross Country Evidence It is quite obvious from a casual look at the data that countries with higher per capita incomes tend to have less malnutrition problem compared to countries with low per capita incomes. Table 3 gives a sample of countries with large disparities in per capita incomes. Countries such as Brazil and Costa Rica have higher per capita incomes than Tanzania and Kenya. Correspondingly, these countries have less malnutrition problem compared to Tanzania and Kenya. One can deduce from this that malnutrition is associated with low income and poverty. However, this correlation is less than perfect; one can, for example, see from Table 3 that the GDP per capita for Bangladesh and Vietnam are about the same but the incidence of malnutrition is significantly higher in Bangladesh than in Vietnam. Similarly, the GDP per capita for Brazil is higher than that of Costa Rica, but Costa Rica suffers less malnutrition than Brazil. It is these exceptional cases that are of policy interest. What policy measures make Costa Rica do better than Brazil in nutritional terms even though it is not doing that well in terms of income? 11 Table 3: Malnutrition and Other Social Indicators for Selected Countries. 1 2 3 4 GDP PER ADULT YOUTH LIFE 5 6 CAPITA ILLITERACY ILLITERACY EXPECTANCY STUNTING WASTING Bangladesh 386.11 59.448 50.942 61.637 44.7 48 Brazil 4633.5 12.7 4.5125 68.314 10.5 5.7 Costa Rica 3899.8 4.3208 1.6647 77.564 6.1 5.1 Egypt 1228.9 43.879 29.517 68.341 18.7 4 Ghana 421.02 27.308 8.365 55.949 25.9 24.9 India 477.06 41.989 26.676 62.993 51.8 53.2 Kenya 325.11 16.657 4.5367 46.256 33 22.1 Saudi Arabia 6613.8 22.945 6.8891 72.807 na Na Seychelles 5939.2 na Na 72.646 5.1 5.7 Sri Lanka 876.37 8.1396 3.0728 73.394 20.4 33 Tanzania 196.93 23.951 8.9348 43.722 43.8 29.4 Uganda 354.82 32.029 20.562 42.807 39.1 23 Vietnam 389.83 7.3188 4.6437 69.371 36.5 34 Zambia 405.14 20.955 11.29 37.46 42.4 23.5 Note: Columns refer to: 1. GDP per capita (constant 1995 US$), 2. Illiteracy rate, adult total (% of people ages 15 and above), 3. Illiteracy rate, youth total (% of people ages 15-24), 4. Life expectancy at birth, total (years), 5. Malnutrition prevalence, height for age (% of children under 5), 6. Malnutrition prevalence, weight for age (% of children under 5). Data is for 2001 except for malnutrition prevalence (height for age, and weight for age) for the following countries: Bangladesh (2000), Brazil (1996), Costa Rica (1996), Egypt (2000), Ghana (1999/1992), India (1993), Kenya (1998), Seychelles (1988), Sri Lanka (1995/2000), Tanzania (1999), Uganda (2000), Vietnam (2000), Zambia (1997). Source: Data extracted from the World Bank Development Indicators CD-ROM 2003. A closer inspection of Costa Rican data helps to dramatise the point that income may not explain all variations in nutrition over time. In 1979, GDP per capita in real terms (in 1995 prices) was USD 3168. Then the prevalence of malnutrition measured in terms of height per age was 20.4% for children under5 years. By the year 1996, real GDP per capita had grown to USD 3374.7, which constitutes an increase of 6.5% of the GDP per capita from 1979. But by 1996, malnutrition had dropped to 6.1%, a 70% drop from 1979. Clearly, growth of income of the order of 6% is unlikely to explain a 70% drop in malnutrition. It is therefore clear that at best, income explains only some variations in the incidence of malnutrition. The question then is, by how much should we expect economic growth to reduce 12 malnutrition? At this stage we are interested in the one-way causal effect, from income to nutrition, and we will review empirical literature that seek to explain this causal link. Haddad et al (2003) take two approaches for investigating the impact of income on nutrition across countries. As has been noted above, this study uses household budget survey data for twelve counties to run a regressions on the impact of income and other factors on nutrition. They also use data from 61 developing countries to estimate the impact of GDP per capita on the state of nutrition measured as the prevalence of under-weight per age for the under 5 children in the country. They collected a number of observations where the nutrition information of the countries is reported, and then related this to the corresponding GDP. Some countries had up to four observations. This way they managed to get a sample observation of 175. Haddad et al (2003) regress the observed malnutrition in each of the 61 countries against the corresponding logarithm of the GDP per capita, female secondary enrolment, access to safe water and dummy variables to capture the decade that the observations were obtained (the data was from 1970s, 1980s and 1990s). While they did not motivate their choice of explanatory variables, their choice did not differ from standard variables commonly used. Nevertheless, they did not explain the failure of including some other commonly used variables such as the proportion of government expenditure on health (see for example, Svenberge (2004)3). This study took care of the potential simultaneity bias and has an added strength of using a relatively large sample for this type of cross-country study. Haddad et al (2004) found that growth in income per capita significantly reduces malnutrition in a country. They calculated mean elasticity of malnutrition to changes in real per capita income of order -0.51, which is remarkably close to the mean arch elasticity of order -0.53 found from the household budget surveys from 12 countries. 3 Svedberge (2004) however found health expenditure as a ratio of GDP to be insignificant in explaining malnutrition. 13 2.2 How Much Does the Reduction in Malnutrition Increase Income? It has been established that a healthy population is an engine for economic growth. It is, for example, estimated that nutrition in terms of provision of adequate calories account for about 30% of the economic growth achieved by UK between 1780 and 1980 (WHO, 2002). This is a large impact and highlights the fact that poor countries like Tanzania need to pay more attention to the nutrition of her population than has been the case so far. The causal link between nutrition, health and labour market outcomes can be argued as follows; healthier workers are more energetic and robust, they are less likely to be absent from work on account of poor health and therefore are more productive and earn better wages. In the long run, improved nutrition increases the energy and cognitive power and therefore enhances the acquisition of knowledge that ultimately improves labour productivity. A more refined argument of this link can be made using economic theory (see Behrman, 1993), but the outline given above has a merit of being plain and yet compelling. Empirical literature exists on this causal link running from nutrition to productivity and earnings (WHO, 2002; Bloom et al, 2003; Thomas and Frankenberg, 2002; Behrman, 1993). We will review this literature first by exploring the evidence at the micro-level and then see how this evidence generalises at the macro level. The empirical studies are divided between experimental evidence and econometric estimations. Econometric Evidence. A vast empirical literature on the impact of nutrition on labour market outcomes exists with an excellent review of these studies in Behrman (1993), WHO (2002) and Thomas and Frankenberg (2002). It is important to note at the outset there are a number of statistical problems that empirical studies of the impact of nutrition on labour market outcomes must contend with. First, it is quite likely that there is a two-way causation between nutrition and labour market outcomes. For example, while improved nutrition may improve labour productivity and thus labour earnings, the improved labour earnings themselves may cause an improvement in nutrition. A regression analysis that fails to take this into account produces biased and inconsistent parameter 14 estimates. There is the possibility of the presence of unobservable fixed effects such as innate ability and robustness that may impact on both nutritional status and labour productivity and earnings. Such fixed effects need to be taken care of using panel data estimation, if indeed panel data exists. Let us look at the study by Strauss (1986) that investigated the impact of calorie intake on the farm workers' productivity in Sierra Leone. This study used instruments to counter the simultaneity problem. A significant result here is that the elasticity of output with respect to nutrition was found to be non-linear, indicating that increasing intake has the maximum impact in terms of increasing productivity at the lower level of calorie intake. At an intake of 1500 calories, the calorie elasticity is 0.75, while at the sample mean, the calorie elasticity is 0.33. Either way, these results suggest that interventions that increase calorie intake would have a very high and significant impact on labour productivity in the rural areas particularly among the poorest. These results are robust but do not take into account intra-household allocation of calories, possible fixed effects or the impact of investment in education. The results by Strauss (1986) are quite relevant to Tanzania for two reasons. First, the analysis is based on rural households; Tanzania is largely a rural country with poverty more widespread in the rural areas. Second, the study investigates the nutritional impact on productivity directly rather than via the observed changes in the wage rate. In Tanzania, rural households do not typically sell labour, but rather they are self-employed. At any rate, several constraints may impede wages from responding to changes in productivity, and thus a study that uses changes in wages rather than changes in productivity is likely to have more estimation problems. Other behavioural evidence of the impact of improved nutrition on labour market outcomes, particularly on wage rates, exists and are extensively reviewed in Thomas and Frankenberg (2002), Behrman (1993) and WHO (2002). We will not dwell on these studies here. Suffice to note that this evidence is mostly based on the relationship between the wage rate and some index of body stature. While the results of these empirical studies have been mixed, on balance, the body of evidence indicates that improved nutrition tends to attract higher wages. If we assume that the higher wage is due to nutritionally induced productivity improvement (rather than the 15 cosmetic attraction to a well nourished body stature), then this is a strong indication that indeed, improved nutrition increases income growth through enhancing labour productivity. At any rate, such increase in wages among the poor contributes to the reduction of poverty. Experimental Evidence There are a number of experimental studies that have been conducted to assess the impact of improved nutrition on labour market outcomes. These studies are summarized in WHO (2002) and also in Thomas and Frankenberg (2002). We will discuss this experimental evidence, their results and limitations and try to generalise the findings to glean some lessons for Tanzania. As pointed out by the WHO (2002), the link between protein and energy malnutrition and ill health has been understood for a long time. Recognition is also now given to the impact of micronutrient deficiencies such as iron, iodine, zinc, calcium and several vitamins on ill health. As has been noted, one can reasonably postulate a relationship between health and labour market outcomes, and particularly labour productivity and earnings. One approach that has been used to test this postulate is the analysis of experimental evidence. Experimental evidence is obtained through actual laboratory-like experiments, where a group of people is picked from which some are subjected to some form of intervention while the rest are not (the control group). Then the changes are observed in relation to the situation before the intervention. If the group subjected to intervention is found to register some significant changes as initially predicted, while the control group registers no such changes, then the changes are attributed to the intervention as evidence of its effectiveness. If no changes are registered as predicted, then the intervention is deemed not to have worked. Preferably, experimental evidence should be collected through longitudinal data, where each individual is observed over a stretch of time while the intervention is repeatedly administered and outcome recorded. The longer the time period the individuals are observed the better, for it allows the full impact of intervention to be realised and registered. 16 Before discussing some of the experimental evidence of the link between nutrition and labour market, it is important to highlights some of the limitations of this approach. The more obvious limitation is that interventions targeted to an individual may not actually reach the individual in totality. For example, nutrient supplements given to workers as part of the experiments may actually be allocated equally among the household members from which the worker belongs (check this again), or may encourage the worker to reduce his otherwise normal consumption at home such that the actual impact of supplementation that was intended may not be realized. Experimental data can easily miss out this fact and thus mis-report the impact of supplementation. . Another limitation is more institutional and technological. If the labour market is such that the employer is unlikely to adjust wages to reward more productive workers, nutritional intervention may still increase the capability of the worker to produce more but the worker may maintain his original level of productivity because there is no incentive to increase it. Equally, the technology used in the production process may not allow increased productivity by workers even if they were robust enough to affect such an increase. In either case, experimental evidence will not report positive productivity effect of nutritional intervention even though the ability of workers to effect higher productivity might have been increased by the intervention. As for experimental evidence that uses longitudinal data, there is a danger that some participants in the experiment may drop out and that the dropouts may not be random in terms of nutritional status and the way the intervention impacts on them. Such drop out is called attrition, and it is likely to cause a bias in the estimation of the impact of intervention on labour market outcomes. In spite of these limitations, experimental exercises may offer great opportunity to unravel the impact of intervention in a controlled way. If efforts to curb such limitations are made through collecting more information on the intra-household allocation of nutrient consumption, the institutional and technological arrangement, and randomise the experiment and collecting more information on those who choose to drop out, the experimental exercise can be very insightful. 17 Evidences of experimental studies are presented in Table 4. The existence of a positive impact on productivity by the nutritional intervention is not always established although on balance, the tendency has been for a positive impact. Moreover, in some cases, nutritional intervention did not improve output, but rather made it possible for the workers to involve themselves more with out-of-work activities, and this may suggest that with more time, such workers would likely seek employment opportunities that reward them better for their increased productivity capability. More involvement in the out-of-work activities in itself constitutes an improvement in the well- being of the worker, in itself a sufficiently valuable achievement. A notable case of course is the 20% increase in productivity induced by the administration of iron supplements. This particular study suffered substantial attrition and it is not clear what the nature of the dropouts is, even though it is suspected that this attrition might have led to over estimation of the impact of intervention. Yet, the mere possibility that intervention to reduce iron deficient anaemia could increase labour productivity by 20% is remarkable. Evidence at the Macro Level So far, we have discussed the evidence of the positive impact of improved nutrition on growth at the micro level. It is therefore interesting to explore empirical evidence on this linkage at the macro level. Our postulated channel of the impact of nutrition on income is mediated through good health, increased energy, and thus, higher productivity. One can thus use life expectancy as a proxy of good health or of good nutrition. In this sense, regression analysis that uses life expectancy as an explanatory variable of economic growth can be considered as investigating the nutrition impact on growth. 18 Table 4: A Summary of Finding from Experimental Studies. NATURE OF MALNUTRITION AND THE INTERVENTION LABOUR MARKET OTHER OUTCOME PROBLEMS WITH THE REVIEWED EVIDENCE OUTCOME EVIDENCE Iron deficiency. Iron Increase in Workers (women) None reported A study by Li et al reviewed by Thomas and Supplementation production efficiency spent more energy Frankenberg (2002). Subjects: Chinese but no changes in and time on non- female cotton mill workers. output work activities Iron deficiency. Iron No increase in output Increase in voluntary None reported A study by Edgerton et al(1979) reviewed by Supplementation activities Thomas and Frankenberg (2002). Subjects: Sri Lank female tea plantation workers. Iron deficiency. Iron Productivity of those Not reported Original subjects numbered 156 A study by Basta (1979) et al reviewed by Supplementation who were anaemic but by the end of experiment, Thomas and Frankenberg (2002). Subjects: rose by 20% some dropped out (attrition) male rubber workers in Indonesia followed leaving the number down to 77. over a period of time, 45% of them were There is likely to be attrition anaemic with productivity 20% lower than bias which over-estimates the those not anaemic. The subjects (both impact if those who dropped anaemic and non anaemic) were observed for out are the one who did not get 60 days of the experiment supplementation. Calorie deficiency. Calorie No changes observed Not reported Randomisation at village level, A study by Immick and Viteri (1981a, supplementation might have confounded the 1981b) reviewed by Thomas and results Frankenberg (2002). Subjects: Guatemalan sugar cane cutters. Calorie deficiency. Calorie Small but significant Not reported Not reported, but the sample is A study by Wolgemuth et al (1982) reviewed supplementation positive impact on only 47 subjects. by Thomas and Frankenberg (2002). the output Subjects: road construction workers in Kenya Calorie and protein deficiency. Calorie and No significant Significant gain in None, but was noted that A study by Satyanarayana et al(1977) protein improvement in weight by the production process was perhaps reviewed by Behrman (1993). Subjects: supplementation productivity participants who such productivity could not piece-work-paid coal fillers in India. received supplements have been improved anyway ­ check sentence!! Indeed, this is the view adopted by Bloom et al (2003) in which they documented twelve empirical studies that use life expectancy as an explanatory variable in the regression analysis. Except for one study, all the other eleven studies indicate that life expectancy has a positive impact on economic growth. Bloom et al (2003) noted that most of the empirical estimation of the impact of improved nutrition on economic growth have ignored the fact that life expectancy not only captures the status of health of the population, but it also captures the experience of the workforce in the population. Since experience is positively related to economic growth, the use of life expectancy as one of the explanatory variables confounds two influences that regression analysis fails to disentangle; impact of health and the impact of experience. In their preferred model, Bloom et al (2003) included a separate explanatory variable for experience to resolve this problem and found that a one year improvement in a population's life expectancy contributes to an increase of 4% in output. This is a large and significant effect, which underscores the macroeconomic rationale for public investment into health and nutrition. A review of the study by Fogel is also instructive here. This review, documented in WHO (2002) indicates that between 1780 and 1980, the provision of adequate nutrition to the workforce contributed up to 30% of the UK per capita growth. Over this period, the British economy had an average annual growth rate of 1.15 percent. Arcand (2001) provides empirical evidence of the impact of nutrition on economic growth. This study is remarkable in that it makes use of Solow's growth model and thus rests on the established theory of economic growth and attends to all econometrics problems that some similar studies suffer from. These econometric problems are (i) the endogeneity of the dependent variable (ii) failure to deal with un-observed heterogeneity and (iii) measurement errors. The fact that this study attempts to meet both the demands of both economic theory and econometric requirements make it one of the best empirical studies on the impact of nutrition on growth. The study considers two measures of nutritional status on growth; the prevalence of food inadequacy and dietary energy supply per capita. The study finds that nutrition significantly influences GDP growth per capita worldwide and that inadequate nutrition is responsible for a shortfall of between 0.23 and 4.7 percentage points in the annual growth rate of GDP! 2.3 What are the lessons to Tanzania from the Empirical Evidence Between Income and Nutrition? In general, most of the empirical studies reviewed in the preceding sections attempt as much as possible to maintain statistical probity, and thus the evidence and policy implications from these studies are relatively credible. Nevertheless, there is not much empirical evidence that is based on Tanzanian data. While data is generally a problem everywhere and most so in Tanzania, the limited available data has not been thoroughly used to assess the empirical links between nutrition and economic growth. Even more, local researchers do not seem to have done as much as they could to investigate the available data in Tanzania and elsewhere. The empirical evidences reviewed have on balance supported the proposition that income level and economic growth are important in resolving the problem of malnutrition. Policies for reducing income poverty and stimulate macroeconomic growth are therefore useful for reducing malnutrition. In the sections below, this report will use some of the estimated coefficients to simulate the impact of growth on nutrition in Tanzania, in a bid to see if growth alone would resolve sufficiently the prevailing malnutrition in Tanzania. Empirical evidence also indicates that there are other factors that contribute to the reduction in malnutrition, and that income generally does not explain all the variations in nutrition status across households or across nations. This evidence suggests that it is possible for income to increase and yet malnutrition to remain unchanged. It also suggests that malnutrition can be reduced even in cases where income remains unchanged. This is possible through government interventions. Tanzania therefore can ensure malnutrition is resolved not only by promoting growth policies, but also by undertaking public interventions to reduce malnutrition. Lastly, empirical evidence indicates that improved nutrition contributes to increased income at the household level, and macroeconomic growth at the national level. Nutrition is therefore not only a valuable achievement intrinsically but also instrumentally in terms of economic efficiency 21 and distribution. This is a remarkable point; poverty reduction efforts and macroeconomic policy for growth can benefit from improved nutrition! It is imperative that the poverty reduction strategy and macroeconomic policy in Tanzania become sensitive to, and include aspects of, nutritional needs particularly for the poor. 3 Nutritional Benefits in Terms of Non-Economic Welfare In this section, we make a brief review of other benefit of nutrition that are not generally quantified in monetary terms, and also, we review the linkages between improved nutrition and other welfare outcomes. It is not at all difficult to argue for improved nutrition as an end in itself, for it is self evident that human beings generally pursue good life and happiness of which health, long life, lower morbidity, and low infant mortality are important and all these are connected to good nutrition. Sen (1999) has made an eloquent and compelling argument that development means reduction of un-freedoms. Malnutrition is one form of un-freedom and is therefore one of the objectives of development. In this sense, one needs no defensive arguments to causally link nutrition to other welfare outcomes as a justification for the pursuit of improved nutrition. Improved nutrition is an end in itself; it is an intrinsically valuable achievement that is worth pursuing even if it did not contribute to the achievement of other valuable welfare outcomes. Improved nutrition however contributes to other welfare outcomes. We have so far reviewed the importance of nutrition to economic growth. There is a vast empirical literature that links malnutrition to increased morbidity, shorter life expectancy, reduction in cognitive power and as a result, reduction in education attainment, increased school dropouts and lower school enrolments. All these are welfare outcomes that are valuable and are also recognised in the MDGs. Literature in this area includes Caulfield et al (n.d), Alderman and Haddinott (2003), WHO (2002), Behrman (1993) and a host of other literature. The empirical links between nutrition and education attainment and improved health is very compelling. 22 It is important to also point out that it is the poor who are generally under-educated and suffer higher morbidity and shorter life expectancy. They are also the ones who generally suffer malnutrition (see for example, URT (2004) on malnutrition by income quintiles in Tanzania). The fact that improved nutrition increases education attainment and productivity and reduces morbidity implies that interventions to reduce malnutrition are pro-poor. What is more, it is clear that the MDGs can more easily be attained if malnutrition is resolved. 4 The Rationale for Government Interventions We have reviewed a number of empirical studies that indicate the existence of linkages of nutrition to income and to other welfare outcomes. Now the question is how compelling is the case for government intervention to improve nutrition? After all, as has been indicated, there is a substantial private benefit to good nutrition for the individuals themselves to invest into good nutrition with or without government intervention. This section reviews the rationale for government intervention by looking at economic arguments, government's own stated objectives (as manifested in the Millennium Development Goals that the government itself aspires to), and an outline of the benefit-cost ratio for intervention. 4.1 Why should the government intervene? One of the overriding policy goals in Tanzania is poverty reduction. It follows that any intervention that leads directly or indirectly to the reduction of poverty would need no other justification or defence for putting it in place. As has been indicated in the text, improved nutrition has tended to also lead to higher productivity and higher wages. This means that policy that improves nutrition will also contribute to a reduction in "income" poverty. Further more, malnutrition is in itself a form of poverty, a fact that is persuasively argued by Sen (1999) and recognized in Tanzania's Poverty Reduction Strategic Paper (URT, 2000). The other argument for government intervention has to do with efficiency. A number of studies have shown that improved nutrition leads to higher economic growth. This can come about directly via improved productivity or indirectly through increased cognitive power and education attainment that lead to increased human capital and thus economic growth (see Becker, 1993). 23 Economic growth is one of the major macroeconomic policy goals the world over, more so for a poor country like Tanzania. The available empirical evidence, on balance, suggests that better nutrition is good for growth, and thus, given the quest for growth in Tanzania, intervention to improve nutrition is imperative. One may suggest that improvement in nutrition should be left to private actors in the economy because rational individuals would thrive to improve their nutrition status given the substantial private benefit involved. This is true only up to a point. To start with, knowledge about the benefits of nutrition and how to improve it is not widespread and no private individual would find it beneficial to finance public generation of such knowledge. Secondly, even when such knowledge is widespread, the infrastructures and institutions necessary to improve nutrition cannot all be put in place and operated by private individuals. For example, it is the government alone that can mandate and enforce nutrient fortification of food items such as salt. Neither can private individuals find it beneficial to finance extensive research on nutrition. Nutrition is therefore, to a significant extent a public good and thus requires public intervention for its efficient provision. Most of the negative externalities of malnutrition, e.g. increasing the spread of infectious diseases demand intervention by a public authority. Private individuals will also not take into consideration the positive externalities of good nutrition, such as its contribution to economic growth and the reduction of private consumption of publicly subsidized health services. Finally, poverty contributes to malnutrition not only because of lack of knowledge, but also because of the failure by the poor to afford nutritionally rich food. As the empirical review has shown, malnutrition also exacerbates income poverty by decreasing labour productivity and undermining the human capital potential of the poor. Government interventions can play a role in this case by providing nutrient supplements and creating incentives for the production of low price nutritious food. If it makes sense for the government to intervene to reduce poverty for whatever reasons (moral, ideological, political or economics), then the same rationale would justify nutrition interventions that benefit the poor. 24 4.2 The Long Run Cost of Not Dealing With Malnutrition The economic costs of not resolving the malnutrition problems are the benefits of improved nutrition that are foregone. We divide the benefits of improved nutrition in three; the non-income benefits, the short run income benefits, and the long run benefits. We will not attempt to estimate the intrinsic value of nutrition because it is virtually impossible to do that. It is important to note that the estimates of benefits that we will undertake will under-estimate the total benefits because of excluding the intrinsic value. The other two types of benefits are instrumental in that they value nutrition as an input in the production or cost function. We divide the instrumental benefit of nutrition into two; short term and long term. The short term relates to the following components: 1. The immediate benefits that accrue in terms of improved productivity. As we have noted, a number of empirical studies have shown that improved nutrition among the poor lead to increased productivity and higher wages. 2. There is also a benefit from improved nutrition in that it frees resources that would have been otherwise committed to dealing with malnutrition-induced illness. 3. Since malnutrition makes a person be prone to infectious deceases and increases the risk of the spread of infection even to those not suffering malnutrition, there is a positive externality in improving nutrition, a benefit that accrues to everybody in the economy. These short-term benefits are used for an outline of cost-benefit in the subsequent section of this report. However, we also maintain that these short term benefit of improved nutrition collectively generate the long-term benefits of improving nutrition in terms of economic growth. We consider economic growth as the long-term benefit of improved nutrition, and we have established, in the preceding section, that there is empirical evidence that improved nutrition leads to economic growth. Since we maintain that the short run benefits of improved nutrition are embedded in the long run benefit of economic growth, we will only use economic growth as a measure of economic benefit of improved nutrition. The cost of malnutrition is therefore viewed as the economic growth that is foregone in the long run. 25 So, what is the long run cost to Tanzania of not resolving the malnutrition problem? Using Fogel's estimates for UK, we can deduce that malnutrition costs Tanzania a 30% shortfall in the annual economic growth. For an annual growth rate of 5%, Tanzania losses an opportunity to increase this growth by 1.5% to 6.5. To be sure, Fogel's estimates are based on the very long run data and therefore one cannot expect that in the very long run, the economy will still grow at an average of 5%. We can use the Arcand (2001) estimate, and particularly the conservative estimate that malnutrition causes a shortfall of 0.23 percent in the annual growth rate4 to get a conservative estimate of the cost of malnutrition in Tanzania. Considering the estimated GDP for Tanzania, for 2003 of Tshs. 929 billion (NBS, 2003), this growth shortfall is equivalent to a Tshs. 2.1 billion shortfall per year. This is a conservative estimate, and it is possible that it grossly underestimates the long run gross shortfall due to malnutrition. Yet even with this conservative estimate the loss is substantial and it indicates that investment of an amount less than Tshs. 2.1 billion per year on nutrition will guarantee greater benefits for the nation. There are a number of estimates of the cost of not tackling malnutrition in terms of income shortfall. For example, the UN/IFPRI (2000) report that iron deficiency has been found to cause a medium value of productivity losses of about 4% per capita, or 0.9 of GDP for a range of developing countries. This is a significant loss that a poor country can ill afford to incur and clearly, the cost of defraying the provision and administration of iron supplements and promoting iron rich diet can not outweigh the cost of iron deficiency as has been shown by Behrman et al (2004). 4.3 The Quest for the Millennium Development Goals An important question is whether income growth on its own can resolve the malnutrition problem at the desired level. The Millennium Development Goals, of which Tanzania also aspires, aim at reducing under-weight among the under five children by halve from 1990 to 2015. Underweight for age is a composite indicator for malnutrition and it generalizes stunting (a long term irreversible impact of malnutrition) and wasting (a short term and potentially 4The other estimate he found is that malnutrition costs a country a 4.7% shortfall in the annual growth rate. See 26 reversible impact. See for example, Alderman et al (2003), on this). The under 5 malnutrition in terms of weight by age in 1991/1992 in Tanzania was 29, and by 1999, this has reached 30 (see Table 5). Table 5: Nutritional Status in Children Under Five: Baseline and Targets NUTRITIONAL INDICATOR 1991/92 1996 1999 2003 2010 2025 Stunting (low height for age) 47 44 44 20 Wasting (low weight for height 6 7 5 2 Under weight-low weight for age 29 31 30 Infant mortality rate 99* 85 50 20 Source: URT (2003) Poverty and Human Development Report 2003, pp 31 and 38.. *This data is for 1997. To be able to gauge whether growth can lead to the attainment of the MDGs with respect to malnutrition by the year 2015, we simulate the impact of a 5% annual growth of real per capita income on the prevalence of malnutrition measured as weight per age for the under five children. It should be noted that the 5% growth in real GDP per capita is overly optimistic; the estimated population growth in Tanzania is 2.9, which means that an annual growth of GDP of 5% can at most be translated into a 2.1% growth in real per capita GDP. To sustain a 5% growth in real per capita GDP, aggregate income needs to grow by at least 7.9%, a feat that has not yet visited Tanzania, and one that can be very difficult to sustain over the years. Moreover, the growth in GDP does not necessarily translate into the same growth rate in households' incomes. There are several reasons for not expecting a growth in aggregate GDP to translate into the same growth in household incomes and thus to have an impact on nutrition at the household level. First of all, evidence has shown significant differences between household income/consumption as reported by household budget surveys and the per capita income obtained from the national income account (Deaton, 2001). Part of the explanation for this is that not all income growth in GDP finds its way into the household coffers in the country. Part of the income is repatriated as profit. There is also the fact that growth may actually be accompanied with increasing income inequality. If such growth favours the rich and have minimal impact on the poor, one cannot section 2.2 of this report. 27 expect malnutrition among the poor to be affected unless the growth induced increase in tax revenue lead the government to undertaking effective interventions to reduce malnutrition. It is worth noting that the recent high growth in Tanzania has been mainly due to mining and tourism activities both of which have minimum linkages with the majority of the poor, and thus is unlikely to lead to significant increase in the income of the poor. In spite of these sobering facts, we present simulations of the impact of a 5% growth of real GDP per capita on malnutrition from 1999 to 2015, with an assumption that the growth is distributionally neutral (i.e. it does not change the level of inequality), and that aggregate GDP growth will average about 7.9% between 1999 and 2015. We report this scenario of overly optimistic growth just to get the sobering sense of the response of malnutrition to income growth. The reason for picking the year 1999 for simulation is that we have reliable data for the prevalence of malnutrition (weight per age for under five children) for this 1999, and the rate is almost the same as the one that prevailed in the year 1991/1992, which is about the base year used in assessing progress of the MDGs, the MDGs target is consistent with halving the 30 rate of malnutrition. The simulation results are summarised in Table 6. It is assumed that the malnutrition elasticity of real per capita income is 0.51. This is the elasticity found by Haddad et al (2003), and is also just slightly higher than the elasticity of demand for nutrients calculated by Abdulai and Aubert (2004a and 2004b) using household budget survey data collected from two regions in Tanzania in 1998 and 1999. In a sense, this elasticity is perhaps a little bit too optimistic; it is estimated from cross-country data of developing countries (Tanzania not included) without controlling for the fixed effects. An alternative simulation, with control for the fixed effects, is reported in Table A2 in the appendix. 28 Table 6: Simulation of Changes in Malnutrition Rate for a 5% GDP per Capita Growth REAL PER CAPITA ELASTICITY INCOME IN USD MALNUTRITION (EXOGENOUSLY YEAR 1995 PRICES GROWTH RATE GIVEN) 1999 185.41 0.05 30 0.51 2000 194.6805 0.05 29.235 0.51 2001 204.4145 0.05 28.4895075 0.51 2002 214.6353 0.05 27.76302506 0.51 2003 225.367 0.05 27.05506792 0.51 2004 236.6354 0.05 26.36516369 0.51 2005 248.4671 0.05 25.69285201 0.51 2006 260.8905 0.05 25.03768429 0.51 2007 273.935 0.05 24.39922334 0.51 2008 287.6318 0.05 23.77704314 0.51 2009 302.0134 0.05 23.17072854 0.51 2010 317.114 0.05 22.57987496 0.51 2011 332.9697 0.05 22.00408815 0.51 2012 349.6182 0.05 21.44298391 0.51 2013 367.0991 0.05 20.89618782 0.51 2014 385.4541 0.05 20.36333503 0.51 2015 404.7268 0.05 19.84406998 0.51 Note: Elasticity is given exogenously from other studies. The GDP per capita for 1999 is from the World Bank Development Indicators. The 1999 Malnutrition rate (weight for age) is from the Government of Tanzania's Poverty and Human Development Report. Apart from the 1999 data, all other figures are simulated based on the assumed 5% growth and the 0.51 elasticity. From Table 6, it is shown that an overly optimistic 5% growth rate in GDP per capita and assuming malnutrition elasticity of per capita income of order 0.51 (also optimistic) would reduce the prevalence of malnutrition in Tanzania from 30 in 1999 to about 20 in the year 2015. The millennium development goal for Tanzania requires the prevalence of malnutrition to be reduced from 29 in 1991/92 to 14.5 in 2015. This means that while growth in real per capita income would lead to a 33% reduction in malnutrition, a very significant decrease indeed, this will not be sufficient to achieve the millennium development goal of halving the prevalence of malnutrition between 1990 and 2015 for Tanzania! Alternative simulations are reported in Tables A2 and A3 in the appendix. In Table A2, the impact of the optimistic 5% annual growth in per capita income (meaning at least a 7.9 annual GDP growth) on malnutrition is reported where malnutrition elasticity is assumed to be 0.3. Working through the results reported by Haddad et al (2003), we calculated the elasticity of 0.29 29 based on the fixed effects model for the cross-country data, and we take this as the possible lowest elasticity of distributionally neutral income growth. Note that the minimum nutrients' elasticity of demand calculated by Abdulai and Aubert (2004a and 2004b) for a sample of data from Tanzania is 0.307 (see Table 2). We find that an average growth in per capital real income of 5% from 1999 to 2015 will reduce malnutrition from 30 to 23.3, a reduction of 22.3%. While even this is a significant reduction, it is far from leading to the achievement of MDGs with respect to malnutrition in Tanzania. A simulation that uses moderately optimistic per capita income growth of 2.1% (for at least 5% annual GDP growth with 2.9% population growth) together with the moderately conservative elasticity of 0.3 is reported in Table A3. This simulation indicates that such a growth rate will reduce malnutrition from 30 in 1999 to 27 in the year 2015, a rather marginal decrease! This scenario seems to be more realistic, if a bit conservative, and thus presents a rather sobering fact that income growth alone cannot resolve the malnutrition problem in Tanzania. Whether one wishes to use the optimistic scenario or conservative scenario, the outcome with respect to the achievement of MDG on malnutrition is clear; income growth is important but not sufficient for attaining the MDG of malnutrition in Tanzania. 4.2 Benefit-Cost Analysis; A Review of Empirical Evidence. It will be interesting to carry out a benefit ratio analysis for intervention in terms of malnutrition and assess both the private and public benefits and costs ratios. Unfortunately, this is not possible without making heroic and unrealistic assumption. To carry out this exercise requires good information on the current situation of malnutrition in Tanzania, the impact of such malnutrition on productivity and other welfare outcomes, and identification of the types, magnitudes and costs of interventions required to resolve the problem. More time is needed to collect this information and make use of it. We will therefore confine ourselves with reviewing the available empirical evidence. 30 Behrman (1993), Alderman and Behrman (2003), Behrman et al(2004) review and offer estimates of costs and benefits of malnutrition intervention. The benefit ratios estimated in Behrman et al (2004) for reducing micro-nutrient deficiencies, promotion of breast feeding, integrates child care programs, intensive pre-school program with considerable nutrition for poor and improving infant and child nutrition in population with high incidence of malnutrition is quite high, some time as high as 9 and above. And yet, these estimates may not have fully captured the positive externality of reducing malnutrition. A review by Behrman (1993) notes the calculation of benefit cost ratio of reducing anaemia that is of order 6 to 71 for countries that include Kenya obtained in one study! In Tanzania, a program called Profile has been used to estimate the benefit ratios of interventions to reduce malnutrition. These estimates are reported in Table 7. The estimates provided in Table 7 clearly suggest that there are substantial gains in investing in nutrition. Table 7: Benefit Cost Ratios for Reducing IDD, PEM and Anemia in Tanzania PROBLEM BENEFIT COST B:C IDD (Iron Deficiency) 315.44 19.89 15.86 PEM (Protein and Energy Malnutrition) 189.05 59.22 3.19 ANEMIA (Iron Deficiency) 134.74 28.69 4.70 Total 639.22 107.81 5.93 It may be useful to revisit the assumption made in the estimates obtained in Table 7, particularly to revise the baseline information used. 5 Conclusion The empirical evidences reviewed indicate that on balance, income growth both at the household and at the national level contributes significantly to reduction in malnutrition. However, income does not explain all the variations in malnutrition and that it is possible to have higher income without a corresponding reduction in malnutrition. It is important therefore that interventions are used to supplement economic growth in reducing malnutrition. 31 The evidences further indicate that reducing malnutrition contributes to economic growth and that such contribution can be substantial. Thus, even if nutrition is not considered to be of importance on its own right, there is a case for improving nutrition as a way of reducing income poverty and increasing macroeconomic growth. Nutrition is linked to several other welfare outcomes, such as reduced morbidity and mortality, and increasing education attainment. Most of these other outcomes constitute MDGs to which Tanzania aspires to achieve. Reducing malnutrition therefore may make it possible, and hasten attainment of MDGs. It is argued however that nutrition is an achievement worth pursuing on its own right, without necessarily defending it as an input into other welfare outcomes. This is a fact recognised in the Tanzania's PRSP (URT, 2000) and it is consistent with the views championed by development economists such as Sen. There are several reasons for the government in Tanzania to intervene to reduce malnutrition. First of all, malnutrition constitutes a dimension of poverty and tends to afflict more those who suffer income poverty. As such, reducing malnutrition is both an implicit and explicit aim of poverty reduction. Nutrition is to some measure a public good (and malnutrition a public bad) that requires public intervention for its resolution. Furthermore, it is shown in this review that one of the indicators of MDGs of halving malnutrition by 2015 cannot be attain in Tanzania without a combination of economic growth and government interventions. The benefit cost ratios of interventions strongly support interventions to improve malnutrition. Finally, there is a lot that can still be done using Tanzania data to empirically assess the linkages between nutrition and income, and to undertake benefit-cost analysis of nutrition interventions. 32 Appendix Table A1: Malnutrition and Some Other Indicators by Regions in Tanzania Index of % Women % Adults Index ofthe % Womenwith % Adultswith Mean Incidence the severity of Moderate Moderate with noprimary with noprimary household of povertypoverty poverty stunting wasting education or more education or more size (p0) gap (p1) (p2) Mara 32.6 8.4 34 56 24 66 5.9 0.25 0.05 0.01 Mwanza 33.8 7.6 33 57 27 61 6.2 0.33 0.09 0.03 Kagera 41.6 10.8 35 50 25 57 5.1 0.22 0.05 0.02 Shinyanga 31.3 6.8 49 42 40 48 6.7 0.43 0.11 0.04 Kigoma 52.5 7.6 33 56 28 60 5.5 0.71 0.28 0.15 Rukwa 42 9.7 40 49 30 54 5.4 0.72 0.3 0.15 Tabora 25.7 4.4 39 53 31 56 4.7 0.48 0.13 0.05 Singida 38.6 7 35 51 27 56 5 0.64 0.2 0.08 Mbeya 46.9 6.2 23 64 16 68 4.2 0.25 0.07 0.03 Iringa 70.5 6.2 24 62 16 68 3.9 0.42 0.12 0.05 Ruvuma 53.5 5.2 20 67 15 71 4.6 0.55 0.19 0.09 Mtwara 58 5.9 36 53 28 56 3.8 0.69 0.21 0.09 Lindi 58.6 7 52 38 44 39 3.9 0.52 0.19 0.09 DSM 30.6 8.1 11 84 8 86 4.3 0.03 0 0 Pwani 51.7 11.2 52 38 39 48 4.9 0.22 0.06 0.02 Morogoro 52.7 4.1 35 52 26 58 4.8 0.38 0.11 0.05 Tanga 55.3 4.9 38 47 31 50 5.6 0.41 0.15 0.07 Kilimanjaro 33.5 5.6 15 61 12 67 4.4 0.25 0.06 0.02 Arusha 43.7 7.2 24 65 20 69 5.3 0.41 0.15 0.08 Dodoma 48.1 8 38 53 31 57 4.4 0.7 0.26 0.12 33 Table A2: Simulated Changes in Malnutrition for 5% Per Capita Growth and Malnutrition Elasticity of 0.3 MALNUTRITION ELASTICITY OF PER CAPITA REAL PER CAPITA INCOME (1995GROWTH OF PERMALNUTRITION CHANGE ININCOME YEAR PRICE) IN USD CAPITA INCOME RATE MALNUTRITION GROWTH 1999 185.41 0.05 30 0.765 0.3 2000 194.6805 0.05 29.235 0.438525 0.3 2001 204.4145 0.05 28.796475 0.431947125 0.3 2002 214.6353 0.05 28.36452788 0.425467918 0.3 2003 225.367 0.05 27.93905996 0.419085899 0.3 2004 236.6354 0.05 27.51997406 0.412799611 0.3 2005 248.4671 0.05 27.10717445 0.406607617 0.3 2006 260.8905 0.05 26.70056683 0.400508502 0.3 2007 273.935 0.05 26.30005833 0.394500875 0.3 2008 287.6318 0.05 25.90555745 0.388583362 0.3 2009 302.0134 0.05 25.51697409 0.382754611 0.3 2010 317.114 0.05 25.13421948 0.377013292 0.3 2011 332.9697 0.05 24.75720619 0.371358093 0.3 2012 349.6182 0.05 24.38584809 0.365787721 0.3 2013 367.0991 0.05 24.02006037 0.360300906 0.3 2014 385.4541 0.05 23.65975947 0.354896392 0.3 2015 404.7268 0.05 23.30486308 0.349572946 0.3 34 Table A3: Simulated Changes in Malnutrition for2.1% Per Capita Growth and Malnutrition Elasticity of 0.3 REAL PER CAPITA GDP (1995 PRICES)GROWTH IN REALMALNUTRITION CHANGE IN YEAR IN USD PER CAPITA GDP RATE MALNUTRITION ELASTICITY 1999 185.41 0.021 30 0.3213 0.3 2000 194.6805 0.021 29.6787 0.18697581 0.3 2001 198.7688 0.021 29.49172419 0.185797862 0.3 2002 202.9429 0.021 29.30592633 0.184627336 0.3 2003 207.2047 0.021 29.12129899 0.183464184 0.3 2004 211.556 0.021 28.93783481 0.182308359 0.3 2005 215.9987 0.021 28.75552645 0.181159817 0.3 2006 220.5347 0.021 28.57436663 0.18001851 0.3 2007 225.1659 0.021 28.39434812 0.178884393 0.3 2008 229.8944 0.021 28.21546373 0.177757421 0.3 2009 234.7222 0.021 28.03770631 0.17663755 0.3 2010 239.6513 0.021 27.86106876 0.175524733 0.3 2011 244.684 0.021 27.68554402 0.174418927 0.3 2012 249.8224 0.021 27.5111251 0.173320088 0.3 2013 255.0687 0.021 27.33780501 0.172228172 0.3 2014 260.4251 0.021 27.16557684 0.171143134 0.3 2015 265.894 0.021 26.9944337 0.170064932 0.3 35 References Abdulai, A., and Aubert, A., (2004a), "Nonparametric and Parametric Analysis of Calorie Consumption in Tanzania", Food Policy, 29, 113-129. Abdulai, A., and Aubert, D., (2004b), "A Cross-section Analysis of Household Demand for Food and Nutrients in Tanzania", Agricultural Economics, forthcoming. Aderman H, Appleton S, Haddad L, Song L and Y. Yohannes (2000) "Reducing Child Malnutrition: How Far Does Income Growth Take Us? Alderman, H., Hoddinott, J., Kinsey, B., (2003), "Long Term Consequences of Early Childhood Malnutrition", World Bank, International Food Policy Research Institute, University of Zimbabwe and Free University. Appleton, S., (2004), "Hunger and Malnutrition ­ Comments", Copenhagen Consensus Opponent Note. Arcand Jean-Louis (2001) "Under nourishment and Economic Growth: The Efficiency Cost of Hunger" FAO Economic and Social Development Paper 147. Becker Gary S (1993) Human Capital: A Theoretical and Empirical Analysis With Special Reference to Education (Third Edition)., The University of Chicago Press. Behrman, J.R and A. B Deolalikar (1987) "Will Developing Country Nutrition Improve with Income? A Case Study for Rural India., Journal of Political Economy., 95, 492-507. Behrman, J.R., (1993), "The Economic Rationale for Investing in Nutrition in Developing Countries", World Development, 21 (11), 1749-1771. Behrman, J.R, Alderman H and J. Hoddinott (2004) "Hunger and Malnutrition", Copenhagen Consensus Challenge Paper. Bouis H.E and L. Haddad (1992) "Are the Estimates of Calorie-Income Elasticities Too High? A Calibration of the Plausible Range", Journal of Development Economics, 39, 333-64. Caulifield, L.E., Richard, S.A., and Black, R.E., (), "Undernutrition as an Underlying Cause of Malaria Morbidity and Mortality", Centre for Human Nutrition, The Johns Hopkins University. Dasgupta Pratha (2001) "Human Well-Being and the Natural Environment", BEIJER Occasional Papers. Dawson, P.J and R. Tiffin (1998) "Estimating the Demand for Calories". American Journal of Agricultural Economics., 80, 474-481. 36 Deaton A (1987) "Quality, Quantity and Spatial Variation of Price: Estimating Price Elasticities From Cross Sectional Data" Living Standard Measurement Studies No 30, World Bank. Deaton A (1990) "Price Elasticities From Survey Data: Extensions and Indonesian Results," Journal of Econometrics, 44, 281-309 Deaton A (1997) The Analysis of Household Surveys: A Microeconometric Approach to Development Policy. John Hopkins. Deaton A (2001) "Counting the Poor: Problems and Possible Solutions", World Bank Research Observer 16(2):125-47 Glewwe, P., Koch, S., and Nguyen, B.L., (2002), "Child Nutrition, Economic Growth and the Provision of Health Care Services in Vietnam in the 1990s", University of Minnesota and the World Bank, Gerhard Mercator Universitaet, and General Statistical Office of Vietnam. Haddad Lawrence, Alderman Harold, Appleton Simon, Long Lina and Yisehac Yohannes (2003) "Reducing Malnutrition: How Far Does Income Growth Take Us?"., The World Bank Economic Review, Vol. 17, No. 1 107-131. Hoogeveen, H., (2004), "Children's Malnutrition in North-western Tanzania", a draft. Mbelle, A.V.Y., (2003), "The Cost of Achieving Millennium Development Goals (MDGs) and Evaluation of their Financing: Tanzania's Experience", Presentation at the Forum on the Millennium Development Goals (MDGs) in West Africa, Dakar, Senegal 26-28 February. NBS (National Bureau of Statistics), (2003) "Statistics Release" in www.Tanzania.go.tz/statistics.html Nyerere J.K (1973) Freedom and Development Oxford University Press, Oxford. Sen Amartya (1999) Development as Freedom, Anchor Books, New York. Strauss J (1986) "Does Better Nutrition Raise Farm Productivity?" Journal of Political Economy, 94:297-320. Strauss J. and D. Thomas (1995) "Human Resources: Empirical Modelling of Household and Family Decisions" in Behrman J and T.N Srinivasan (eds) (1995) Handbook of Development Economics, Amsterdam, Elsevier. Subramanian S and A. Deaton (1996) "The Demand for Food and Calories", Journal of Political Economy, 104, 133-62. 37 Svedberg Peter, (2004), "Has the Relationship Between Undernutrition and Income Changed?", Comment on "Hunger and Malnutrition", Copenhagen Consensus Opponent Note. Thomas, D., and Frankenberg, E., (2002), "Health, Nutrition and Economic Prosperity: A Micro- Economic Perspective", California Centre for Population Research, On-Line Working Paper Series. Tiffin R and P.J Dawson (2002) "The Demand for Calories: Some Further Estimates from Zimbabwe"., Journal of Agricultural Economics., 53, 221-232. Tinios, P., A. Sarris, H. Amani and W. Maro (1993). Households, Consumption, and Poverty in Tanzania: Results from the 1991 National Cornell - ERB Survey. United Republic of Tanzania (2000) Poverty Reduction Strategic Paper, Government Printer, Dar es Salaam. United Republic of Tanzania (2002) Household Budget Survey 2000/01: Final Report Dar es Salaam: National Bureau of Statistics United Republic of Tanzania (2004) "Poverty and Human Development Report 2003" Dar es Salaam. UN & IFPRI (2000) "Fourth Report in the World Nutrition Situation", 2000. WHO, (2002) Health, Economic Growth, and Poverty Reduction, The Report of Working Group 1 of the Commission on Macroeconomics and Health. Young Helen (2001) "Nutrition and Intervention Strategies" in Devereux Stephen and Simon Maxwell (eds) (2001) Food Security in Sub Saharan Africa University of Natal Press. 38 DRAFT: 04 October 2004 The Distributional Impact of the PEDP in Rural Kilimanjaro Johannes G. Hoogeveen, World Bank 1. Introduction The Primary Education Development Program (PEDP) was introduced in July 2001. The program includes the abolition of school fees, increased teacher recruitment, investment grants for new buildings, as well as community mobilization and empowerment of school committees. The reform has been well received in Tanzania and has led to a large increase in primary school enrollment. Because data collection for the HBS preceded the introduction of the PEDP, the consequences of this program could, till date, only be assessed through administrative data such as those collected by the Ministry of Education and Culture. These data suggest a large increase in enrollment in primary school from 4.8 million students in 2001 to 6.0 million in 2002 and 6.6 million in 2003 (URT 2003). Because administrative data systems do not collect information about the families from which school going children originate it has not been feasible to assess (i) which families, the rich or the poor, benefited most from the introduction of the PEDP and (ii) to which degree the PEDP has contributed to attaining MDGs like achieving universal primary education and gender equality. Using recently collected information from the rural Kilimanjaro region, this note addresses these issues. 2. Impact of the PEDP on attendance In 2003 a representative survey of 957 rural households was implemented in 45 villages in the rural Kilimanjaro region (Sarris, Karfakis and Christiaensen, 2004). This survey collected information about household consumption and education in a way that allows to make comparisons with the HBS. For instance the same questions were asked about educational attainment and both survey instruments collected detailed information about household consumption. Since the HBS is representative at the district level, in combination with the Kilimanjaro survey, the impact of the PEDP can be considered.1 Table 1 presents which fraction of children in different age categories attend primary school. It shows that the PEDP has had a considerable impact. The fraction of children aged 6-12 attending primary school increased from 52% to 82% and many start school earlier. For instance, in 2001, 13% of all 7 year olds attended primary school; with PEDP 1The 2000/01 HBS is representative at the regional level, by rural and urban areas. However, some households in the HBS have been attributed very high weights, up to 12,000 where the median weight is 200. Consequently few households are very important for outcomes. In the case of the one household with a weight of 12,000, it equals the contribution of 60 `median' households in the data set. The presence of few households with high weights require care in interpretation of the results. The trends reported here are robust to the use of different weights; individual point estimates reported in the various tables are not, as reflected by the high standard errors, which are reported in all tables. 1 this has increased more than fourfold to 68%. From the age of 9, enrollment is virtually complete. Also, whereas there may have been some ­albeit very weak, suggestion that pre-PEDP less girls attended school than boys, with PEDP this disappeared entirely. School attendance is high, irrespective of gender. Table 1 Fraction of all children attending primary school, by Age Age 2001 2003 2001 2003 All Male Female Male Female 6 2.5% 19.6% 0.3% 3.6% 15.0% 24.1% (2.6) (4.4) (0.3) (3.9) (4.9) (5.6) 7 12.7% 68.1% 18.6% 4.2% 72.4% 63.9% (7.8) (4.1) (10.7) (4.1) (5.7) (5.8) 8 51.5% 88.5% 24.6% 61.2% 85.1% 92.5% (27.5) (3.0) (9.2) (21.6) (3.8) (3.0) 9 27.5% 96.5% 24.2% 30.8% 95.0% 98.2% (10.8) (1.5) (12.3) (12.9) (2.4) (1.8) 10 73.0% 97.7% 80.9% 67.7% 96.8% 98.7% (10.8) (1.1) (7.8) (17.3) (1.8) (1.3) 11 94.3% 97.3% 96.8% 89.2% 98.5% 96.3% (3.5) (1.5) (3.2) (6.1) (1.5) (2.6) 12 93.6% 98.1% 94.9% 92.4% 100% 96.7% (3.7) (0.9) (3.8) (6.0) (1.7) Average attendance 51.8% 81.6% 54.0% 50.1% 81.1% 82.2% 6-12 year olds (9.2) (1.2) (7.2) (13.4) (1.6) (1.7) Note: Standard errors in parentheses. The largest increase in attendance can be observed amongst children aged 7-9. This is unsurprising as the PEDP was introduced in 2001 and the typical age at which primary education is started is 7.2 2The fractions for children aged 7 have to be interpreted with care, however, because there exists some unclarity about how the education questions have been interpreted by enumerators. The question asks for educational attainment (highest grade/attending) and this can be interpreted as the highest grade completed, or the current grade attended. As both surveys ask the question in the same way, there is reason to assume that the outcomes from both surveys are comparable. Yet the absolute fractions for those aged 7 may be off as some children currently attending primary school may be missed because their highest grade completed was recorded as pre-school (or none). Irrespective of the way the question is interpreted, for children aged 8-9 one can be assured that (i) if children are attending primary school this is recorded and (ii) that in 2003 these children are sufficiently young to have been PEDP beneficiaries from the start. Consequently to assess the impact of PEDP it is most prudent to focus on children aged 8-9. 2 Next consider Table 2. It shows that with PEDP the fraction of 8-9 year olds not attending school dropped from 60% to 8%. It also confirms what Table 1 already suggested, namely that with PEDP children start school younger. In this case, it shows up because for a given age, more children have reached a higher level of education (standard 2 and higher). Table 2 Fraction of children aged 8-9 attending primary school Level of education 2001 2003 No education 59.5% 7.8% (14.5) (1.8) Standard 1 32.9% 27.7% (15.1) (2.9) Standard 2 3.6% 42.5% (2.1) (3.4) Standard 3 3.7% 14.3% (3.0) (2.3) Standard 4 and up 0.3% 7.6% (0.3) (1.4) Total 100% 100% Note: Standard errors in parentheses. 3. Distributional Impact of the PEDP Having information about per adult equivalent consumption in both surveys it is possible to consider the distributional impact of the PEDP by considering how school attendance changed across different consumption quantiles.3 Table 3 does so and presents primary school attendance by consumption quintile for 2001 and 2003. The Table shows how in all quintiles school attendance increased. Especially in the lower quintiles did primary school attendance increase remarkably (and significantly). For instance whereas in 2001 only 28% of the children aged between 8 and 9 and originating from the poorest 20% of the households attended primary school, by 2003 this had increased to 86%. 3The HBS and Kilimanjaro surveys collected consumption in different ways. The HBS collected consumption information using an open ended questionnaire administered over a period of 12 months. The Kilimanjaro survey collected consumption information over a two months period and asked for a pre-coded set of consumption items. Consequently, consumption levels may not be comparable between 2001 and 2003. Yet within a given year there is little reason to assume that the ranking of households from poorest to wealthiest has been affected by differences in the way consumption information is collected. Since our interest is in the evolution of the distribution of consumption, and not in its levels, we treat both surveys as comparable ­at least for the purposes of this paper. 3 Table 3 Primary school attendance for 8-9 years olds, by consumption quintile Quintile 2001 2003 1 27.9 85.9 (7.8) (4.1) 2 13.6 89.0 (9.8) (3.4) 3 21.5 94.6 (9.7) (3.0) 4 55.4 85.7 (16.9) (4.1) 5 74.1 90.1 (23.8) (3.7) Mean 40.2 89.1 (14.5) (2.0) Note: Standard errors in parentheses. Figure 1, finally, considers the distributional impact of PEDP graphically. It presents the concentration curves for 2001 and 2003 and the concentration incidence growth curve (CIGC).4,5 The concentration curve for 2001 shows how pre-PEDP access to primary education was unequally distributed: children from wealthier households attended school relatively more often than children from poorer households. The distribution was as unequal as that of the distribution of consumption, as represented by the Lorenz curve. With PEDP inequalities in access to education have disappeared and use of primary education is proportional across the wealth distribution (the concentration curve coincides with the 45-degree line). 4A concentration curve shows the use of a service across the consumption distribution. The curve shows on the horizontal axis cumulative per adult equivalent consumption (normalized to fit a 0 ­ 1 scale) against, on the vertical axis, (normalized) cumulative access to a particular service or facility. 5The CIGC shows the growth rate of the use of services for each quantile of the distribution of per adult equivalent consumption. If the CIGC lies above the zero growth line, it indicates that use of the service has increased; if the CIGC lies below the zero line it indicates a decline in the fraction of households that use the service. If the CIGC lies everywhere above zero then the use of benefits improved for all households and there is first order dominance of the distribution at date t over t-1. 4 Figure 1: Changes in the distribution of access to education, rural Kilimanjaro Concentration curve: Access to Primary Education 2001 Concentration curve: Access to Primary Education 2003 100 100 ) ) %( 80 Concentration-curve Concentration-curve %( 80 ecirves ecirves 60 60 of of onitubirtsdi. oni 40 Lorenz-curve butirtsdi. 40 muC Lorenz-curve muC 20 20 0 0 0 20 40 60 80 100 0 20 40 60 80 100 Cum. distribution of population Cum. distribution of population Concentration Incidence Growth Curve: Primary Education 2001-03 5 4 3 ht 2 Grow 1 0 -1 0 20 40 60 80 100 Percentile The concentration incidence growth curve, finally, confirms this. It not only lies everywhere above 0 ­suggesting that all wealth groups benefited from the introduction of the program, it also shows that the average increase was higher for the poorer than the richer households. In other words the PEDP was not only absolute pro-poor in that the poorer households experienced an increase in the use of the service, PEDP was also relatively pro-poor in that the poorest quintiles benefited more than the richest quintiles. 4. Conclusion Evidence from rural Kilimanjaro suggests that the PEDP has led to a large increase in school enrollment. In fact almost all eligible children are now enrolled in primary school. Not only does full enrollment imply gender equality, since in the pre-PEDP period primary school enrollment was lowest amongst the poorer wealth quintiles, it follows that the program is pro-poor: children originating from poor households benefited more than proportionally. This is very encouraging though it has to be kept in mind that high enrollment rates may not necessarily lead to high primary school completion rates (the MDG objective). But if the children that are currently enrolled can be kept in school, the evidence suggests that 5 Tanzania is well placed to meet the MDGs for primary education, as well as for gender equality in education. It also suggests that the time may have come to shift focus from enrollment to the quality of primary education, an area where the sector still faces major challenges (Lange 2004). References Lange S. 2004. Primary Education Since the Introduction of the Primary Education Development Plan. REPOA, Project Brief No. 8. June 2004. Sarris, A., P. Karfakis and L. Christiaensen 2004. Household Income Risks and the Willingness to Pay for Coffee Price Insurance in the Kilimanjaro Region of Tanzania. Mimeo. URT 2003. Basic Statistics in Education 1999-2003. National Data. Ministry of Education and Culture, Dar es Salaam. 6 TRENDS IN MALNUTRITION IN TANZANIA Submitted by REPOA to World Bank September, 2004 Trends in Malnutrition in Tanzania TABLE OF CONTENTS 1. Introduction................................................................................................................4 2. Time trends ................................................................................................................5 3. Spatial variation in malnutrition ................................................................................9 3.1 International comparison .....................................................................................9 3.2 Regional variation of under-five malnutrition in Tanzania ...............................10 3.3 Urban rural differences ......................................................................................13 4. Socio-economic determinants of under-five malnutrition.......................................15 4.1 Poverty impacts on under-five nutritional status ...............................................15 4.2 Socio-economic household characteristics and nutritional status......................18 5. Timing of birth and nutritional status ......................................................................22 6. Health and health related determinants of malnutrition...........................................24 6.1 Health related determinants ...............................................................................25 6.2 Water and sanitation ..........................................................................................28 7. Multivariate analysis................................................................................................30 7.1 Independent effect of exclusive breastfeeding in infants under 6 months.........31 7.2 Independent predictors of stunting in children under 36 months ......................34 7.3 Independent predictors of stunting in children between 36 and 60 months ......36 7.4 Independent predictors of stunting in all children less than 5 years of age.......39 8. Conclusion and recommendation.............................................................................43 Bibliography ................................................................................................................45 Appendix......................................................................................................................47 FIGURES Figure 2.1a Cumulative distribution of the z-scores for Height for Age, 1991-1999......... 5 Figure 2.1b Cumulative distribution of the z-scores for Weight for Age, 1991-1999........ 5 Figure 2.1c Cumulative distribution of the z-scores for Weight for Height, 1991-1999.... 6 Figure 2.2a Trends and targets of underweight .................................................................. 6 Figure 2.2b Trends and targets of stunting ......................................................................... 6 Figure 2.3a Z-scores for Height for Age by age in months, 1991-1999............................. 7 Figure 2.3b Z-scores for Weight for Age by age in months, 1991-1999............................ 7 Figure 2.3c Z-scores for Weight for Height by age in months, 1991-1999........................ 7 Figure 3.1 Height for Age in selected countries, 1991 - 2001............................................ 9 Figure 3.2 Weight for Height in selected countries, 1991 - 2001..................................... 10 Figure 3.3 Urban ­rural differentials in malnutrition ....................................................... 13 Figure 4.1 Cumulative frequency distribution of household poverty scores, 1991-1996. 15 Figure 4.2 Cumulative frequency distribution of household wealth/poverty scores for households with and without stunted children.......................................................... 16 Figure 4.3 Cumulative frequency distribution of z-score for Height for Age (stunting) by wealth/poverty quintile ............................................................................................. 17 Figure A. Zonal differences in stunting for 1991 and 1996.............................................. 47 2 Trends in Malnutrition in Tanzania TABLES Table 2.1 Trends in nutrition indicators in children under the age of five, 1991-1996 ............................ 5 Table 2.2 Trends in nutrition indicators in children under the age of 36 months and between 36 and 59 months, 1991-1996........................................................................................................... 8 Table 3.1 Regional variations in Height for Age (stunting) in Tanzania, 1991-1999 ............................ 11 Table 3.2 Regional variations in nutritional status and weekly consumption of meat, fish, eggs milk and beans........................................................................................................................... 12 Table 3.3 Differences in Height for Age (Stunting) by age group and place of residence, 1991-1999...... 13 Table 4.1 Prevalence of moderate/severe stunting by wealth/poverty status, 1991-1999....................... 16 Table 4.2 Socio-economic household characteristics, 1991-1999..................................................... 18 Table 4.3 Characteristics of parents in relation to severe to moderate stunting in children under the age of 5, 1991-1999 .................................................................................................................... 19 Table 4.4 Household characteristics in relation to severe to moderate stunting in children under the age of 5, 1991-1999................................................................................................................. 20 Table 5.1 Prevalence of moderate to severe stunting by month of birth, 1991-1999............................. 22 Table 6.1 Health and health indicators of malnutrition, 1991-1999 .................................................. 24 Table 6.2 Health and health related determinants of moderate to severe stunting, 1991-1999 ................ 25 Table 6.3 Exclusive breastfeeding by wealth/poverty status, 1991-1999............................................ 27 Table 7.1a Independent predictors of moderate to severe stunting in children under the age of 6 months: Model without birth weight.............................................................................................. 32 Table 7.1b Independent predictors of moderate to severe stunting in children under the age of 6 months: Model with birth weight.................................................................................................. 33 Table 7.2a Independent predictors of moderate to severe stunting in children under the age of 36 months: Model without birth weight.............................................................................................. 35 Table 7.2b Independent predictors of moderate to severe stunting in children under the age of 36 months: Model with birth weight.................................................................................................. 36 Table 7.3a Independent predictors of moderate to severe stunting in children between the age of 36 and 60 months: Model without birth weight .................................................................................. 36 Table 7.3b Independent predictors of moderate to severe stunting in children between the age of 36 and 60 months: Model with birth weight...................................................................................... 38 Table 7.4a Independent predictors of moderate to severe stunting in children under fives: Model with birth weight ......................................................................................................................... 40 Table 7.4b Independent predictors of moderate to severe stunting in children under fives: Model with birth weight ......................................................................................................................... 41 Table 7.5: Simulations outcome................................................................................................ 42 Table A: Regional Muslim composition ..................................................................................... 47 Table B: Distribution of religions by wealth status 1991-1999 ........................................................ 47 3 Trends in Malnutrition in Tanzania 1. Introduction The importance of adequate nutrition for general health has been a long established fact. Good nutrition, both in quantity and quality plays an important role in the prevention of infectious and other diseases on one hand, and on the other hand well nourished children do have better expectation of full recovery once they fall ill. In addition, adequate nutrition during early childhood improves mental capabilities in later life. Though child nutrition has been given an important role in the Poverty Reduction Strategy, malnutrition in under-fives has not yet received the attention needed. No thorough and evidenced based study on determinants of malnutrition in Tanzania has been made. This study will hopefully give sufficient input for the discussion on the background and determinants of child malnutrition in mainland Tanzania. This study will include analysis of the internationally accepted indicators of under-five nutritional status, height for age (stunting), weight for age (under-weight) and weight for height (wasting). Stunting is an indicator for chronic malnutrition, measuring the cumulative long-term effect of inadequate nutritional intake and repeated episodes of infectious diseases as well as chronic diseases. This indicator is less likely to be influenced by seasonal variation. Wasting (weight for height) is used as an indication of acute malnutrition, and is more likely to reflect seasonal variation as it measures current nutritional status. The main focus of this study will be on height for age, or stunting, since height is a good summary measure of the health status of children (Cole and Parkin, 1977; Younger 2004). Analysis will be based on the overall population of children aged up to 5 years as well as on the two age groups, 0 to 36 months (3 years) and from 36 months to 59 months (5 years). Observed values for height for age are compared with the values from the NCHS (National Center for Health Statistics)/WHO standard population. This standard population is globally used to assess nutritional status of children. Values that are below 2 standard deviations or more from the median of the reference population are generally considered an indication for moderate to severe malnutrition. Values that are below more than 3 standard deviations from the median of the reference population are an indication of severe malnutrition. We will limit ourselves to moderate to severe stunting. Major data sources will be three Demographic and Health Surveys held in Tanzania in 1991/92, 1996, and 1999, from which the nutrition indicators are drawn and the 2000/01 Household Budget Survey (HBS), from which regional variables on consumption will be used to associate these with regional estimates of the under-five nutritional status. In this report section two (2) presents the changes over time in Tanzania, section three (3) examines spatial variation in malnutrition, section four (4) analyses socio-economic determinants of under-five malnutrition, section five (5) studies timing of births and nutritional status, section six (6) considers health related determinants of malnutrition and section seven (7) focuses on multivariate analyses, identifying possible independent predictors of malnutrition. We conclude with a summary of the main findings in section eight (8). 4 Trends in Malnutrition in Tanzania 2. Time trends The observation that the nutritional status of children did not change significantly over the 1990s has been made in several studies (PHDR 2002 and PHDR 2003). To illustrate this once more, the levels and distribution characteristics of the various indicators are displayed in Table 2.1 and in Figure 2.1. Table 2.1 Trends in nutrition indicators in children under the age of five, 1991-1996 Height for Age (Stunting) Weight for Age (Underweight) Weight for Height (Wasting) %below [95% CI] z-score [95% CI] %below [95% CI] z-score [95% CI] %below[95% CI] z-score [95% CI] 2sd 2sd 2sd 1991/92 43 [41-46] -1.76 [-1.83--1.68] 29 [27-31] -1.28 [-1.34--1.23] 6 [5-7] -0.24 [-0.29--0.19] 1996 44 [42-46] -1.76 [-1.83--1.69] 31 [29-33] -1.35 [-1.41--1.29] 7 [6-8] -0.34 [-0.39--0.30] 1999 43 [39-47] -1.78 [-1.88--1.68] 29 [26-32] -1.34 [-1.43--1.24] 6 [4-7] -0.3 [-0.40--0.20] Sources: Authors' calculation using Demographic and Health Survey data of 1991/92, 1996 and 1999 Table 2.1 underlines the modest changes in the under-five nutritional status in Tanzania from the early to the late 1990s. In 1991/92 43% of the children were classified as moderate to severely stunted (height for age), in 1996 this was 44% and in 1999 the rate was again at 43%. The same trend can be observed in weight for age and weight for height. Figures 2.1a to 2.1c show the cumulative distributions of the z-scores for the several indicators. Figure 2.1a Cumulative distribution of the z-scores for Height for Age, 1991-1999 ) %( 100 cyneu 80 DHS 1991/92 eqrF 60 DHS 1996 40 vei DHS 1999 latu 20 muC 0 -6 -4 -2 0 2 4 6 Height for Age Z-scores Figure 2.1b Cumulative distribution of the z-scores for Weight for Age, 1991-1999 ) %( 100 cyneu 80 DHS 1991/92 eqrF 60 DHS 1996 40 vei DHS 1999 latu 20 muC 0 -6 -4 -2 0 2 4 6 Weight for Age z-scores 5 Trends in Malnutrition in Tanzania Figure 2.1c Cumulative distribution of the z-scores for Weight for Height, 1991-1999 ) %( 100 cyneu 80 60 DHS 1991/92 eqrF 40 DHS 1996 vei 20 DHS 1999 latu 0 muC -6 0 6 12 Weight for Height z-scores Sources: Authors' calculation using Demographic and Health Survey data of 1991/92, 1996 and 1999 Also from the cumulative distributions it can be concluded that no significant changes in levels or distribution of the nutritional status for the under-fives took place during the 1990s. These trends clearly demonstrate the difficulty of the task of achieving the Poverty Reduction Strategy targets set for 2015 at national level, which aims to reduce stunting by more than 20% (see Figure 2.2b). Figure 2.2a Trends and targets of underweight 35 30 25 Percent 20 15 10 5 0 1990 1995 2000 2005 2010 2015 2020 year Mainland TZ Urban Rural MDG Figure 2.2b Trends and targets of stunting 50 45 40 35 Percent 30 25 20 15 10 5 0 1990 1995 2000 2005 2010 2015 2020 year Mainland TZ PRS targets Urban Rural Source: Authors' calculation using DHS 1991/92, DHS 1996 and DHS 1999, PRSP 2000 and World Bank 2001 The indicator for tracking progress on the MDGs (Goal 1) is weight-for-age and the target is to reduce the prevalence of hunger (weight-for-age <-2 SD) by 50% from 1990 6 Trends in Malnutrition in Tanzania rates. While the PRS aim to reduce the prevalence of stunting from 43.4 to 20% and reduce the prevalence of wasting from 7.2% to 2% as its medium term target from 1999. Figure 2.2a shows the underweight from the 1990 to 1999, the graph shows that Tanzania is far from reaching the projected target of the MDG in 2015 represented by the dotted lines. As Figure 2.2b demonstrates, urban rates for stunting dropped substantially during the 1990s and the 1999 rate (24%) already passed the PRS target for mainland Tanzania set for the year 2003. Clearly, the rural rates are well above the MDG/ PRS targets. The MDG uses underweight, the combined measure of stunting and wasting, while the PRS sets its target using stunting and wasting. Figure 2.3a Z-scores for Height for Age by age in months, 1991-1999 0.5 0 -0.5 -1 (NCHS/WHO) -1.5 seroc -2 -2.5 -3 Z-s 0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 Age in months 1991/92 1996 1999 Figure 2.3b Z-scores for Weight for Age by age in months, 1991-1999 1.5 1 0.5 0 -0.5 (NCHS/WHO) seroc -1 -1.5 -2 -2.5 Z-s 0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 Age in months 1991/92 1996 1999 Figure 2.3c Z-scores for Weight for Height by age in months, 1991-1999 1.5 1 0.5 0 (NCHS/WHO) seroc -0.5 -1 -1.5 Z-s 0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 Age in months 1991/92 1996 1999 Sources: Authors' calculation using Demographic and Health Survey data of 1991/92, 1996 and 1999 7 Trends in Malnutrition in Tanzania The Figures 2.3a-c shows the incidence of the nutrition indicators, height for age, weight for age and weight for height for the three different surveys. The height for age curve by age (Figure 2.3a) clearly shows the worsening of the nutritional status up to 24 months, after which its seems to stabilize. Due to the small number of observation for the different ages, the curves show substantial random fluctuations. Weight for age seems to worsen from the age of three months to 12 months after which the situation seems to stabilize (Figure 2.3b). Again strong fluctuation can be observed due to the very small numbers, however the trends for the various years seem to be similar. Weight for height as shown in Figure 2.3c appears to be more stable with increasing age, after an increase up to 3 months weight for height slightly decreases up to the age of around 18 months after which it stabilizes at slightly higher levels. Trends for the different years are comparable. The same trend observed in all children under the age of five can also be observed in children under the age of 36 months and in children in the age group 36 to 59 months; i.e., no significant change during the 1990s (see Table 2.2). Table 2.2 Trends in nutrition indicators in children under the age of 36 months and between 36 and 59 months, 1991-1996 1991/92 1996 1999 Up to 36 36 ­ 59 Up to 36 36 - 59 Up to 36 36 ­ 59 months months months months months months % [95%CI] % [95%CI] % [95%CI] % [95%CI] % [95%CI] % [95%CI] Moderate to severe stunting 39 [37-42] 52 [48-55] 41 [38-43] 50 [47-53] 38 [34-42] 51 [47-56] Sources: Authors' calculation using Demographic and Health Survey data of 1991/92, 1996 and 1999 There was no significant changes in levels or distribution of the nutritional status for the under fives during the 1990's. The urban rates for underweight and stunting surpass the MDG and PRS targets respectively. In the youngest infants/children, the incidences of nutritional indicators stunting, wasting and underweight seem to be highly correlated with age. 8 Trends in Malnutrition in Tanzania 3. Spatial variation in malnutrition 3.1 International comparison As mentioned in the introduction, better-nourished children are less likely to fall ill and if they fall ill have a higher likelihood of a swift recovery. Areas with high rates of malnutrition are therefore more likely to have higher burden of disease and mortality rates. For international comparison of the under-five nutrition indicators, we used information from the surrounding countries Kenya, Uganda and Malawi. Figure 3.1 Height for Age in selected countries, 1991 - 2001 60 50 DS 40 30 below % 20 HA < -2 SD 10 HA < -3 SD 0 92 96 91/92 96 99 95/96 00/01 93 98 Malawi Tanzania Uganda Kenya Source: DHS country reports, 1991 - 2001 and authors' calculations Figure 3.1 shows that the prevalence of malnutrition (stunting) measured by the height for age index in Tanzania is comparable with levels found in other surrounding countries, with the exception of Kenya, which shows substantially lower stunting rates during the 1990s. The indicator for acute malnutrition, weight for height (wasting), as depicted in Figure 3.2, shows the same pattern as the selected countries. All show stagnant wasting rates over the 1990s. The nutritional situation in Malawi seems to be worse than other East African countries; the difference with Kenya is substantial 9 Trends in Malnutrition in Tanzania Figure 3.2 Weight for Height in selected countries, 1991 - 2001 8 7 6 SD 5 4 below 3 % WH < -2 SD 2 WH < -3 SD 1 0 91/92 96 99 93 98 92 96 95/96 00/01 Tanzania Kenya Malawi Uganda Source: DHS country reports, 1991 ­ 2001 and authors' calculations 3.2 Regional variation of under-five malnutrition in Tanzania As is already shown in the previous figures, under-five malnutrition prevalence showed very little variation during the 1990s. Despite several development efforts to reduce poverty and to increase the standard of living, child malnutrition hardly changed in rural areas but there seem to be improvement in the urban areas. Figure 3.2 even suggests a slight increase in wasting from 1991/92 to 1996 and back to 1991/92 levels. As the sample sizes of the subsequent surveys were too small to allow for differentiation of the malnutrition indicators by region, due to the absence of substantial change over time, pooling the three surveys became a good option to increase sample size, and allowing for higher levels of disaggregation. A total of about 12,000 observations provided an acceptable number of children to allow for the assessment of nutritional differentials by region. In order to measure the regional differences for both the youngest and the older under-fives, separate analysis was done for children under 36 months (less than 3 years of age) and children between 36 months and 60 months (between 3 and 5 years of age). Table 3.1 shows the results of this `pooled' analysis. Variations between regions are large. Children with the best nutritional status live in Dar es Salaam with an overall prevalence of stunting of 28%. The worst nutritional status can be observed in children living in Iringa (63% stunting), Mtwara (60%) and Lindi (59%). 10 Trends in Malnutrition in Tanzania Table 3.1 Regional variations in Height for Age (stunting) in Tanzania, 1991-1999 Up to 36 Months 36 to 59 months All under-fives % below 2 Lower Upper % below 2 Lower Upper % below Lower Upper Region SD Limit Limit SD Limit Limit 2 SD Limit Limit Dodoma 50 41 59 66 49 82 55 45 64 Arusha 34 26 40 49 40 58 39 32 45 Kilimanjaro 31 24 39 39 29 48 34 27 41 Tanga 45 38 52 58 50 66 50 43 56 Morogoro 50 45 55 67 58 76 56 51 61 Coastal Region 42 36 47 69 61 76 50 45 55 Dar es Salaam 27 22 32 30 24 36 28 24 32 Lindi 57 50 63 65 55 75 59 53 66 Mtwara 52 47 58 75 66 84 60 54 65 Ruvuma 48 42 55 59 52 67 52 47 58 Iringa 59 53 65 70 62 78 63 56 70 Mbeya 41 34 47 58 47 68 47 41 52 Singida 35 29 41 48 39 56 39 34 45 Tabora 33 28 38 43 32 54 36 31 41 Rukwa 41 35 46 47 38 55 43 37 48 Kigoma 47 41 53 56 49 63 50 45 55 Shinyanga 31 27 35 38 31 45 33 29 37 Kagera 39 31 47 48 40 56 42 36 48 Mwanza 32 28 37 41 35 47 35 31 39 Mara 29 25 34 38 28 48 32 27 38 Sources: Authors' calculation using pooled Demographic and Health Survey data of 1991/92, 1996 and 1999 The data suggest that regions located in the south (Figure A in the appendix shows zonal differences in malnutrition) and along the coast show substantially higher prevalence of chronic malnutrition (stunting) than those regions located in the northern part of Tanzania especially the lake zone in 1996. Using consumption and expenditure data from the 2000/01 Household Budget Survey it is possible to look into some of the background variables that may explain the observed regional differences. Table 3.2 clearly displays the high correlation between the frequent consumption of meat and milk or milk products and a favourable under-five nutritional status. Regions with a high average consumption of meat and milk not only show considerably lower levels of moderate to severe stunting, but also of severe stunting. The weight for age indicator also shows the relationship with regional differences in consumption patterns, though less pronounced. Higher levels of meat and milk consumption, however, don't seem to be associated with reduced levels of acute malnutrition, measured through weight for height. The observed positive relationship between high weekly milk consumption rates and reduced levels of stunting are clearly shown in Arusha, Kilimanjaro and Mara. These regions have the highest intake of milk per week and also show low stunting prevalence rates. Mtwara and Lindi on the other hand have the highest levels of stunting and also the lowest frequency for milk consumption levels. Although the evidence is indirect, these findings suggest that not only adequate availability of food but also specific diets may lead to an improved nutritional status. Frequent intake of animal fats through the consumption of meat and milk and milk products seems to lead to better nourished children. Multivariate analysis, using in addition to the food consumption variables, regional prevalence of fever, diarrhoea and education of the mother (all derived from the 11 Trends in Malnutrition in Tanzania DHS surveys) showed that the frequency of meat consumption and milk consumption (again at regional level) were both independent negative predictors of the prevalence of stunting, the strongest effect being shown by the consumption of meat. Consumption of eggs seems to be a positively associated with the prevalence of stunting (see table 3.2). These results underline the importance of consumption patterns and feeding practices to be included in future health surveys so they may provide valuable additional information for the assessment of the nutritional status of the children. One of the anomalies is Iringa. Within Tanzania, Iringa region is one of the main producers of food crops, and ranks 10th on the household assets score (to be discussed in later in section 4.1) and ranks 5th on the ranked Human Development Index (HDI) scores. However, as shown earlier, levels of stunting in Iringa rank among the highest within the country. One of the other outliers is Mara region. Mara shows the lowest score on the assets index and also a below average score on the HDI, but in terms of nutrition status of the under-fives Mara ranks among the best performing regions. When relating consumption patterns to wealth/poverty, it is important to note that not only does poverty play an important role when it comes to consumption, but also the regional variation in the availability of specific food items plays a role. Cattle are concentrated mainly in northern Tanzania, facilitating the relatively high consumption of milk and beef in these areas. Data suggests that high milk consumption rates are associated with reduced stunting rates. It should be noted again that the evidence presented is indirect, since we were not able to link nutritional status with daily consumption patterns. Table 3.2 Regional variations in nutritional status and weekly consumption of meat, fish, eggs milk and beans Height for Weight for Weight for MEAL MEAT FISH EGGS MILK BEANS Age Age Height Region % below % below % below average average average average average average -2SD -2SD -2SD Dodoma 55 36 4.4 2.3 1.4 0.8 0.3 2 2.7 Arusha 38.6 33.5 8.7 2.7 1.8 0.9 0.6 4.3 3.1 Kilimanjaro 34.2 26.2 6.8 2.5 1.6 1.5 0.4 2.7 2.9 Tanga 49.7 36.2 11.3 2.8 1.4 2.5 0.2 1.9 2.7 Morogoro 56 31.5 4.5 2.5 1.4 1.8 0.3 1 2.7 Coast 50.4 29.8 5.6 2.9 0.9 3.1 0.3 0.7 3.6 Dar es Salaam 28.3 20.9 7.1 2.9 2.3 2 0.8 1.7 3.1 Lindi 59.3 35.2 5.5 2.6 0.7 3.1 0.2 0.2 2.4 Mtwara 59.7 39.8 5.3 2.6 1 3 0.3 0.3 2.5 Ruvuma 52.4 32.5 4.9 2.7 1.3 2.5 0.5 0.5 3.6 Iringa 63.1 42 4.5 2.5 1.4 1.1 0.3 0.7 2.6 Mbeya 46.5 23.4 4.9 2.5 1.6 1.7 0.3 1.1 2.7 Singida 39.1 31.9 7.1 2.3 1.3 1 0.2 1.8 1.5 Tabora 36.2 20.9 4.8 2.6 1.7 2 0.2 1.9 1.9 Rukwa 42.8 28.4 6.9 2.1 1.3 2.2 0.3 0.5 3.7 Kigoma 50.4 38.5 7 2.1 0.8 1.7 0.1 0.2 4.7 Shinyanga 33.2 23.5 5.7 2.6 1.2 1.2 0.1 2.2 1.9 Kagera 42 31.1 8.2 2.1 1 2.5 0.6 1.6 5 Mwanza 35 24.4 5.9 2.4 1.2 3.4 0.2 2 1.8 Mara 32.5 18.9 5.5 2.3 1.6 3.1 0.2 2.4 1.6 Average 46.8 31.2 6.3 2.7 1.7 2.2 0.5 1.7 2.8 Correlation 0.03 0.63 0.09 -0.19 -0.65 0.30 P Values 0.916 0.003 0.696 0.413 0.002 0.195 Source: Authors' calculations using DHS 1991/92, 1996, 1999 and HBS 2000/01 12 Trends in Malnutrition in Tanzania 3.3 Urban rural differences Malnutrition is usually associated with poverty. Since poverty is generally concentrated in rural areas it is to be expected that the nutritional status of the under-fives will also be less favourable in rural areas compared to urban areas. Table 3.3 and figure 3.3 show the differences in stunting, observed over time. Table 3.3 Differences in Height for Age (Stunting) by age group and place of residence, 1991-1999 Age in months Place of Residence 1991/92 1996 1999 Urban 35 [31 - 38] 33 [29 - 36] 20 [15 - 24] 0 - 36 months Rural 41 [39 - 42] 42 [41 - 44] 42 [40 - 45] Urban 46 [41 - 51] 34 [29 - 39] 35 [27 - 43] 36 - 59 months Rural 53 [51 - 56] 53 [51 - 56] 55 [51 - 58] Urban 38 [35 - 41] 33 [30 - 36] 24 [20 - 28] 0 - 59 months Rural 45 [43 - 46] 46 [45 - 48] 47 [45 - 49] Source: Authors' calculation based on Demographic and Health Surveys, 1991/92, 1996 and 1999 Differences between rural and urban areas are substantial and seem to have increased during the 1990s. The prevalence of stunting in children under the age of five has remained constant in rural areas (46%) during the 1990s, whereas the nutritional status of the under-five population in urban areas improved significantly. The 1999 data show that the risk of having a low height for age was about twice as high in young children in rural areas than in young children in urban areas. The prevalence of stunting (in overall population of under-fives) in urban areas dropped by almost 40% from 38% in 1991/92 to 24% in 1999. Figure 3.3 Urban ­rural differentials in malnutrition 50.0 45.0 40.0 35.0 30.0 Urban 25.0 Rural 20.0 15.0 10.0 5.0 0.0 1991 1996 1999 1991 1996 1999 1991 1996 1999 1991 1996 1999 M oderate Severe M oderate Severe Height for Age Weight for Age Source: Demographic and Health Surveys, 1991/92, 1996 &1999 and authors' calculations Figure 3.3 once more shows the widening of the urban-rural gap. Moderate to severe stunting rates are stagnant in rural areas whereas these indicators show a declining trend in urban areas during the 1990s. The prevalence of stunting in urban children under the age of 3 years dropped significantly more than the prevalence of stunting in the children between the age of 3 and 5 years (43% versus 24%), which indicates that the conditions for the younger children improved more than for the older children. One of the 13 Trends in Malnutrition in Tanzania explanations might be that an improve nutritional status of mother and reduce low birth weight in urban areas, which as will be shown later, have important decreasing effect on the prevalence of malnutrition in young infants. Surrounding countries like Kenya, Uganda, Malawi show similar trends and comparable stagnant rates in stunting and wasting. Regions along the coast except for Dar es Salaam and those in the south have high prevalence of chronic malnutrition. Moreover, regions with high consumption of milk and meat have lower rates. Urban areas are better off than rural areas; while improvements can be observed in the urban areas, stagnation is shown in the rural areas. 14 Trends in Malnutrition in Tanzania 4. Socio-economic determinants of under-five malnutrition It is generally believed that direct environment, social as well as physical, has an important impact on a child's health and nutritional status. Poverty is believed to be an important determinant of malnutrition, both directly through lack of resources and inadequate living conditions and indirectly through lack of education, which translates into lack of knowledge with respect to nutrition and health. 4.1 Poverty impacts on under-five nutritional status As the DHS did not collect data on income or household expenditure, the poverty assessment was carried out using primarily household assets variables, building characteristics of the house, as well as water and sanitation facilities. As shown by Filmer and Pritchett (2001) amongst others, welfare indicators constructed on the basis of household assets and characteristics have statistical properties comparable to standard household expenditure variables. For the purpose of a poverty/welfare analysis, a welfare indicator was constructed on the basis of: · Durable consumer goods available in the household · Water and sanitation facilities to which the household has access · Size of the household To obtain comparable welfare indicators for all the Demographic and Health Surveys, the three surveys were pooled, and treated as one single database. Identical variables related to the above mentioned household characteristics and assets were selected, on the bases of which a welfare score was calculated for each household having a child under 5 years old for whom standardized nutrition information was available. Subsequently for each of the three surveys, households were ranked according to the poverty/welfare score and grouped into five more or less equal sized groups. Figure 4.1 shows the cumulative distribution of the included households on their respective poverty/welfare index score for the three years. Figure 4.1 Cumulative frequency distribution of household poverty scores, 1991-1996 100 75 Frequency 50 vetialu (%) 25 muC 0 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 Wealth/poverty scores 1991/92 1996 1999 Source: Authors' calculation using DHS 1991/92, 1996 and 1999 15 Trends in Malnutrition in Tanzania The cumulative poverty/welfare distributions suggest that only modest improvement in the welfare/poverty situation took place during the 1990s, with a slight decrease from 1991 to 1996 and no change afterwards. This improvement, however, can be observed only for those households in the upper part of the distribution; within the poorest 40% no changes seem to have taken place. The poverty/welfare estimates from the three DHS are very well in line with the expenditure based observations from the 1991 and 2000/01 Household Budget Surveys which showed a modest decrease in the basic needs headcount ratio from 39% in 1991 to 36% in 2000/01. Macro economic growth rates in Tanzania increased steadily since 1993, however these growth rates are not reflected in the change in headcount ratios nor in a significant change in the household assets based welfare indicator used in this study, showing the weak link between macro economic growth and poverty reduction. Neither the nutritional status of the children under-five, nor the poverty status of the households they live in changed substantially during the 1990s. Therefore, to assess the poverty distribution of households with stunted children under five years and of households with children under five who are not malnourished, a pooled data was used to allow for an increased number of observations. Figure 4.2 Cumulative frequency distribution of household wealth/poverty scores for households with and without stunted children 100 75 Frequency 50 vetialu (%) 25 muC 0 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 Wealth/poverty scores moderate/severe stunting no stunting Source: Authors' calculation using DHS 1991/92, 1996 and 1999 Figure 4.2 suggests that the wealth/poverty situation of the poorest 40% of the households either with stunted children or with non-stunted children is the same. From 40% and upwards, the wealth/poverty situation seems to differ between households with a stunted child and households without a stunted child. As can be observed in Figure 4.2, stunted children are more likely to live in households with lower wealth/poverty scores. Table 4.1 Prevalence of moderate/severe stunting by wealth/poverty status, 1991-1999 1991/92 1996 1999 combined poorest 20% 47[43-50] 50[46-55] 52[47-58] 49[47-52] second 20% 50[45-54] 47[42-52] 46[40-52] 48[45-51] middle 20% 51[47-55] 49[45-53] 49[42-56] 50[47-52] fourth 20% 40[36-44] 45[41-49] 40[34-46] 42[39-45] least poor 20% 31[24-37] 30[26-33] 28[21-35] 30[26-33] Note: between [] the 95% confidence intervals are displayed Source: Authors' calculation using DHS 1991/92, 1996 and 1999 16 Trends in Malnutrition in Tanzania Table 4.1 shows comparable information, combined with the distribution of three surveys separately. As the table shows, within the poorest 60% of the households, the proportion of stunted and non-stunted is about the same (51% versus 49%), however, in the least poor households 29% of the children are moderately to severely stunted. The stunting prevalences by wealth/poverty quintiles in this study differ from those calculated by (Gwatkin et al. 2000)1 based on the 1999 TRCHS. Reasons for the differences are: · In this study Zanzibar is not included; · Factors scores for calculating the poverty index were based on the pooled 1991/92, 1996 and 1999 data and not only on the 1999 data as was done by Gwatkin et al; and · For analytical purposes, only those households with children under fives for whom anthropometric measures were available were ranked according to their poverty index score. Evaluating the various years, the differences by wealth/poverty quintiles are comparable; within the least poor 20% households the prevalence of severe/moderate stunting is substantially lower than in the 20% poorest. It seems that in the highest economic status groups, a significant proportion of the children under-five can escape malnutrition, whereas in the other groups large proportions (though not the majority) remain malnourished. Table 4.1 shows that this situation has remained stagnant during the 1990s and although groups are too small to draw definite conclusions, data suggest that even in the least poor households the nutritional situation of the children did not change during the 1990s. Figure 4.3 Cumulative frequency distribution of z-score for Height for Age (stunting) by wealth/poverty quintile 100 (%) 90 cyneu 80 70 poorest 20% eqrF 60 2nd quintile 50 40 middle lative 30 4th quintile mu 20 least poor 20% Cu 10 0 -6 -4 -2 0 2 4 6 Height for Age z-scores Source: Authors' calculation using DHS 1991/92, 1996 and 1999 The cumulative frequency distributions of the height for age z-scores show almost identical distributions and levels; only the curve for the least poor 20% is shifted slightly to the right, again indicating overall lower levels of stunting. 1Moderate to severe stunting rates in 1999 from the poorest to the least poor quintiles were 50%, 53%, 45%, 37% and 23% respectively. 17 Trends in Malnutrition in Tanzania In the following sections we examine the impact of several background characteristics of children's nutritional status. Some of these variables were also used in the construction of the wealth/poverty index. In doing so the impact of sensitive issues like religion, is addressed. Before addressing the impact of socio-economic characteristics, the change over time of these characteristics is discussed. 4.2 Socio-economic household characteristics and nutritional status Table 4.2 shows that in general, the socio-economic characteristics of the household did not change significantly during the 1990s. The most important changes can be observed for the educational levels of both parents of children aged under five years. Table 4.2 Socio-economic household characteristics, 1991-1999 1991/92 1996 1999 Religion Muslim 26 [21-30] 26 [22-30] 25 [18-33] Catholic 33 [29-37] 33 [29-37] 29 [24-35] Protestant 26 [22-30] 26 [23-29] 27 [21-32] Other 0 [0-0] 0 [0-0] 0 [0-1] No Religion 15 [11-19] 15 [11-19] 18 [11-25] Female Headed Household 11 [10-13] 15 [13-17] 16 [13-19] Household Composition Total number (mean) 8.2 [7.7-8.6] 7.0 [6.8-7.2] 7.7 [7.0-8.3] Number of women aged 15-45 (mean) 1.8 [1.7-1.9] 1.5 [1.5-1.6] 1.7 [1.6-1.8] Number of under-fives (mean) 2.4 [2.2-2.5] 2.1 [2.0-2.2] 2.3 [2.1-2.5] Education of the Mother No Education 34 [31-37] 29 [25-32] 28 [22-35] Primary education 64 [61-67] 69 [66-72] 69 [63-75] Secondary Education 2 [1-3] 2 [2-3] 3 [2-4] Education of the Father No Education 24 [21-27] 19 [16-22] 17 [12-22] Primary education 69 [66-71] 75 [72-77] 78 [73-83] Secondary Education 6 [5-8] 6 [4-7] 5 [3-6] Higher Education 1 [0-1] 0 [0-1] 0 [0-1] Mother currently working 73 [70-75] 58 [55-60] 79 [71-86] Household Food Security Always enough 45 [41-48] Most of the times not enough 5 [4-7] Source: Authors' calculation using DHS 1991/2, 1996 and 1999 The proportion of mothers without any formal education dropped from 34% in 1991/92 to 28% in 1999, whereas the proportion of fathers with primary education increased from 69% in 1991/92 to 78% in 1999. Fathers tend to have higher levels of education than mothers. Household information on food security was only available from the 1996 survey and therefore, unfortunately, change in food security over time could not be assessed. Nevertheless, the 1996 data show that 46% of the households state that they always had enough food; whereas only 5% of the households indicated that most of the time they did not have enough food. To increase the explanatory power of the independent variables it was decided to continue with the analysis using the pooled data set. 18 Trends in Malnutrition in Tanzania Table 4.32 shows the relationship between characteristics of the mother and father on the nutritional status of their child. Age of the mother beyond 35 has a significant negative impact on the nutritional status of a child. Children below the age of 3 with a mother aged over 40 have a higher probability of being malnourished than children with younger mothers. Table 4.3 Characteristics of parents in relation to severe to moderate stunting in children under the age of 5, 1991-1999 0 - 36 36 - 59 0 - 59 % % % Age of the Mother 15-19 37 [33 - 40] 51 [37 - 66] 38 [34 - 41] 20-24 39 [37 - 41] 50 [47 - 54] 42 [40 - 43] 25-29 39 [37 - 41] 52 [49 - 54] 43 [42 - 45] 30-34 40 [37 - 42] 54 [51 - 57] 45 [43 - 47] 35-39 43 [40 - 46] 48 [44 - 51] 45 [42 - 47] 40-44 46 [42 - 51] 48 [44 - 53] 47 [44 - 50] 45 + 50 [43 - 57] 52 [45 - 60] 51 [46 - 56] Education of the Mother No Education 44 [42 - 46] 55 [52 - 57] 48 [47 - 50] Primary education 39 [37 - 40] 50 [48 - 52] 42 [41 - 43] Secondary Education 17 [12 - 22] 22 [14 - 30] 19 [14 - 23] Education of the Father No Education 41 [38 - 43] 52 [49 - 55] 45 [43 - 47] Primary education 40 [39 - 42] 52 [50 - 54] 45 [44 - 46] Secondary Education 22 [18 - 26] 29 [23 - 34] 24 [21 - 28] Mother currently working 40 [39 - 41] 51 [49 - 53] 44 [43 - 45] Sources: Authors' calculation using pooled Demographic and Health Survey data of 1991/92, 1996 and 1999 Education Education, especially increased maternal education, has a reducing effect on stunting prevalence. Both for the very young as well as for those between 3 and 5 years of age, the probability of being stunted decreases rapidly with a higher educational level of the mother. Of the children aged below 3 the proportion of malnourished children dropped from 44% when the mother has no education, to only 17% when the mother has secondary education. Children with mothers who have no formal education have a more than 10% higher probability of being malnourished than children whose mothers have some education. Higher levels of education increase the access to information regarding health and health related issues and proper care in general. It also increases the capability of the mother to adequately assess the health status of the child and seek care when needed. On the other hand, level of education is highly correlated with socio-economic well being. From all the under-five children, the impact of the father's education level is less pronounced. With all under-five children, any positive impact of the father's education level appears nil unless he has secondary or higher education. Only the secondary or higher education of the father seems to reduce the probability of the child being malnourished. 2There are other possible determinants not considered in this table, but maybe quite important in terms of impact on stunting ­ eg. Mother's nutritional status, maternal depletion syndrome, LBW, spacing and other children under 3 in the household, etc 19 Trends in Malnutrition in Tanzania Table 4.4 Household characteristics in relation to severe to moderate stunting in children under the age of 5, 1991-1999 Upto 36 Months 36- 59 months All Ages % below 2SD [95% CI] % below 2SD [95% CI] % below 2SD [95% CI] Religion Muslim 44 [42 - 46] 55 [52 - 58] 48 [46 - 49] Catholic 39 [38 - 41] 52 [50 - 55] 44 [42 - 45] Protestant 37 [35 - 39] 48 [45 - 51] 41 [39 - 42] No Religion 38 [35 - 40] 48 [44 - 51] 41 [39 - 43] Female Headed Households 40 [37 - 43] 54 [50 - 58] 45 [42 - 47] Number of Household Members less than 4 41 [39 - 43] 54 [50 - 58] 44 [42 - 46] 5-6 members 41 [40 - 43] 54 [51 - 56] 46 [44 - 48] 7-9 member 40 [39 - 42] 50 [48 - 53] 44 [42 - 45] 10 or more members 35 [33 - 38] 46 [43 - 49] 39 [37 - 41] Number of women aged 15-45 1 woman 41 [40 - 42] 54 [52 - 56] 46 [44 - 47] 2- 3 women 37 [36 - 39] 48 [45 - 50] 41 [39 - 42] 4 or more women 38 [34 - 42] 39 [34 - 45] 38 [35 - 42] Number of children under-five in the household 1 child 40 [38 - 41] 52 [49 - 54] 43 [41 - 45] 2 children 41 [40 - 43] 52 [50 - 54] 45 [44 - 47] 3-4 children 39 [37 - 41] 51 [48 - 54] 43 [41 - 45] 5 or more children 34 [30 - 38] 44 [38 - 50] 37 [34 - 41] Note: n.s = not significant; * significance level = 0.05; ** significance level = 0.01 Sources: Authors' calculation using pooled Demographic and Health Survey data of 1991/92, 1996 and 1999 Religion Table 4.4 relates several household characteristics to the prevalence of moderate to severe stunting. In general, children in Muslim households tend have an increased risk of being stunted. However it might very well be location factors that play a role (Table A in the appendix show regional percentages of Muslims). The Muslim community3 in Tanzania is mainly concentrated in the Coastal Regions like Pwani, Tanga and Mtwara, which are already the regions with an above average risk of stunting, as previously mentioned. In the multivariate analysis we will discuss some possible explanations for these observations. Household Members As the data shows, the nutritional status of children under-five does not significantly differ between those who are part of a male or female-headed household. The negative association between household size and the incidence of stunting was not expected. The initial hypotheses assumed that an increased household size, including a higher number of competing children under-five, would increase household crowdedness, 3The data shows that Muslims are assets rich (38%are in the least poor quintile) when compared to Catholics and Protestants (30% are in the least poor quintile). See Table B in the appendix. 20 Trends in Malnutrition in Tanzania and would reduce the well being of young children, resulting in a deprived nutritional status. A larger number of under-fives present in the household is also associated with a reduced risk of being stunted. Referring to the multivariate analysis later in section seven (7) the opposite relationship is observed. A reducing effect on malnutrition through increased presence of caregivers (adult women) was expected; the presence of more women aged 15 to 45 seems to reduce the probability of stunting in the under-five population. In particular, children 3 and 5 years old seem to profit from an increased number of household members. Several studies give a mixed picture of the relationship between household size and children's well-being or household welfare. A study of the nutritional status of pre- school children in Sri Lanka shows that the number of children living in the household is one of the independent predictors of moderate to severe stunting (Department of Census and Statistics, 2003). Others, like Kamuzora and Mkanta (2000) argue that larger households have greater economic potential and are therefore less poor, having a positive effect on the health status of the children. In this section it was shown that no substantial improvement in the welfare/poverty situation took place during the 1990s; for poorest 40% of the households no change was observed. The rates of moderate to severe stunting for the least poor 20% is substantially lower than in the 20% poorest. Children below the age of three years with mothers aged over 40 have a higher probability of being malnourished than children with young mothers. The probability of stunting reduces when level of education of the mother increases. Furthermore, children from Muslim households tend to have a higher risk of being stunted, which may be attributed to location factors. Sex of the head of household does not mater, while increase number of female (caretakers) reduce the risk of being stunted. The next sections will focus on timing of birth, followed by socio-economic characteristics and health determinants and health related care-seeking behaviour. 21 Trends in Malnutrition in Tanzania 5. Timing of birth and nutritional status As generally perceived, the month of interview has a relation with the reported nutritional status of a child. A nutritional status assessment made after a period of drought and food shortage will most likely give different results from an assessment made after a period of adequate food supply. We assume that the same effects may well apply for the period when the child was born. The seasonal fluctuation of the food supply and also health risks may very well have its effect on the health and nutritional status of the child immediately after birth and also during the first six months of life. Additionally, the same seasonal fluctuations may have impact on the health status of the mother during pregnancy, as well as during the lactation period, which both will have impact on nutritional and health status of a child. By studying 15 million death certificates in the USA, Gabriele Doblhammer (2002) found a significant relation between the mean age at death and the month of birth. These finding were related to seasonal differences in nutrition of the mother during pregnancy and seasonal differences in the exposure to infectious diseases during early childhood. It is most likely that the same seasonal fluctuations will have a direct impact on birth weight of the child and thereby on the child's nutritional status later on. A major disadvantage of the possible findings is that they are difficult to influence through policy or preventive action. Children are born throughout the year, and provided there is a willingness to do so, only adequate family planning knowledge and access to effective methods might provide a possible tool to influence the timing of a birth. However, insight and understanding of the seasonal fluctuations of risk may well lead to more focused prevention and intervention. For the youngest children, the February, March and April months of birth pose the highest risk to children of becoming malnourished. Of the children under the age of 3 years 46% and 43% were stunted respectively (see Table 5.1). July, August and September are the months of low risk. Only children born in February and September seem to differ significantly in their levels of stunting as the confidence intervals show. Table 5.1 Prevalence of moderate to severe stunting by month of birth, 1991-1999 Month 0 ­ 36 36 ­ 59 0 ­ 59 January 41[37-45] 49[43-55] 44[40-47] February 46[41-51] 60[54-65] 51[47-55] March 43[39-47] 55[49-61] 47[44-51] April 43[39-47] 51[45-57] 46[43-49] May 41[37-45] 53[47-60] 45[41-49] June 38[34-42] 57[51-63] 44[40-47] July 35[32-39] 46[40-53] 39[36-43] August 35[29-41] 52[46-58] 40[36-45] September 34[30-38] 45[39-51] 38[35-41] October 39[35-44] 49[42-56] 43[39-46] November 41[36-46] 50[43-57] 44[40-49] December 40[36-44] 45[39-50] 41[38-45] Note: 95% Confidence Intervals are displayed between [ ] Sources: Authors' calculation using pooled Demographic and Health Survey data of 1991/92, 1996 and 1999 There is little indication that the increased risk of stunting is caused by inadequate nutrition of the mother during pregnancy; birth weights of those born in February and 22 Trends in Malnutrition in Tanzania March do not differ from the birth weights of the children born in July, August or September. This may lead to the conclusion that seasonal fluctuation in environmental risk factor during the first months of life may play a major role in explaining the observed differences in stunting prevalence. 23 Trends in Malnutrition in Tanzania 6. Health and health related determinants of malnutrition Health and health related factors are closely linked to nutritional status of children, and the relations between health and nutrition work both ways. Deprived health conditions worsen the nutritional status but a deprived nutritional status has again a worsening effect on the health of the child. In this section we will focus on the impact of health and health related factors on the nutritional status. Table 6.1 Health and health indicators of malnutrition, 1991-1999 Health related indicators 1991/92 1996 1999 Diarrhoea during the past 2 weeks 14[12-15] 14[13-16] 13[11-15] Medical treatment given 61[55-68] 57[53-61] 64[57-71] ORS given with diarrhoea 59[54-64] 49[44-54] 58[50-65] Increased amount of Fluids given 28[23-32] 57[52-62] 32[24-40] Reduced amount of Fluids given 14[11-16] 17[14-20] 35[27-44] Fever during the past 2 weeks 32[30-34] 32[30-34] 37[34-41] Medical treatment given with fever 58[54-62] 63[58-67] 75[66-83] Cough during the past 2 weeks 29[28-31] 31[29-33] 37[35-40] Medical treatment given 55[51-58] 57[54-61] 71[64-77] Ever Breastfed 99[99-100] 98[98-99] 97[96-98] Exclusive Breastfeeding up to 6 months 15[12-18] 24[20-28] 29[20-37] Low Birth Weight ( <2500 grams) 22[20-24] 19[17-21] 17[13-20] Sibling child death 32[30-34] 31[28-33] 34[30-38] Private Piped water 9[7-11] 6[5-8] 10[6-14] Public Piped water 20[15-26] 25[21-29] 18[12-25] Well 32[28-37] 31[26-36] 44[36-52] Spring 10[7-13] 12[9-15] 14[9-18] River/Stream/Pond/Lake 24[19-29] 25[20-30] 14[10-18] Flush toilet 1[1-2] 1[1-2] 2[1-2] Traditional Pit Latrine 84[81-88] 84[81-87] 82[77-87] Improved Ventilated Pit Latrine 1[1-1] 1[0-1] 1[0-1] No toilet/bush 13[10-17] 14[10-17] 16[10-21] Source: Authors' calculation using DHS 1991/92, 1996 and 1999 DHS data only provides a limited set of variables related to the current health status of a child. Information available relates to diarrhoea, fever and cough, the health care seeking behaviour of the mother when a child is suffering from these illnesses. In addition we look at breastfeeding practices and also the more distant health related factors, such as source of drinking water and use of sanitation in the household. Table 6.1 displays health and health related determinants included in this study. Table 6.1 shows some changes over time in the health related indicators used in the analysis. The prevalence of cough shows an increase from 29% in 1991/92 to 37% in 1999. Medical treatment given in case of fever or cough also showed an increase during the 1990s. Other important changes are the increase in the proportion of children up to 6 months exclusively breastfed (from 15% in 1991/92 to 29% in 1999), and the modest decrease in the proportion of children born with a low birth weight (from 22% to 17%). 24 Trends in Malnutrition in Tanzania 6.1 Health related determinants Table 6.2 Health and health related determinants of moderate to severe stunting, 1991-1999 Health related indicators 0 ­ 36 36 - 59 0 - 59 Diarrhoea During past 2 weeks 43[40-46] 62[54-69] 46[43-49] No diarrhoea during past 2 weeks 39[37-41] 50[48-53] 43[42-45] Medical treatment given 43[39-48] 63[53-73] 46[42-50] No medical treatment given 43[38-47] 61[49-72] 45[41-50] ORS given 43[39-47] 66[56-76] 46[41-50] No ORS given 44[40-48] 57[45-68] 46[42-50] Increased amount of fluids given 43[38-47] 64[52-76] 46[42-50] Reduced amount of fluids given 43[36-50] 50[31-69] 44[38-50] Fever During past 2 weeks 43[40-45] 53[50-57] 45[43-47] No fever during past 2 weeks 38[36-40] 50[48-53] 43[41-44] Medical treatment given 40[37-44] 56[50-61] 44[41-47] No medical treatment given 45[41-49] 50[44-57] 47[43-50] Cough During past 2 weeks 41[39-43] 49[45-52] 43[41-45] No cough during past 2 weeks 39[37-41] 52[50-54] 44[42-46] Medical treatment given 40[37-43] 49[44-55] 42[40-45] No medical treatment given 42[39-46] 48[43-54] 44[41-47] Breastfeeding Ever Breastfed 40[38-41] 51[49-53] 43[42-45] Never Breastfed 44[33-56] 51[38-64] 47[38-56] Exclusive B.feeding up to 6 months 8[5-12] 57[51-63] 8[5-12] No exclusive B.feeding up to 6 months 15[12-18] 44[41-47] 15[12-18] Low Birth Weight ( <2500 grams) 54[48-60] 57[51-63] 55[50-60] Normal Birth Weight (>= 2500 grams) 33[30-35] 44[41-47] 36[34-39] Sibling child death 43[40-45] 52[49-56] 46[44-49] No Sibling child death 38[37-40] 50[48-53] 42[40-44] Source of Drinking Water Private piped water 23[20-27] 29[23-35] 25[22-29] Public piped water 39[36-43] 54[50-59] 44[41-48] Well 41[38-44] 52[48-55] 45[42-47] Spring 44[40-48] 52[45-58] 47[43-50] River/stream/pond/lake 42[39-45] 55[51-60] 46[44-49] Toilet facilities Flush toilet 14[7-20] 31[16-45] 18[11-25] Traditional pit latrine 40[38-42] 52[49-54] 44[42-46] Improved ventilated pit latrine 28[17-40] 31[15-47] 30[19-40] No toilet/bush 40[36-43] 49[45-54] 43[40-46] Note: 95% Confidence intervals are displayed between [ ] Source: Authors' Calculation using DHS 1991/92, DHS 1996 and DHS 1999 25 Trends in Malnutrition in Tanzania Diarrhoea Looking at Table 6.2, diarrhoea prevalence has remained quite constant during the 1990s. In all three surveys, the prevalence of diarrhoea in the first four wealth/poverty quintiles is comparable. In the 1991 and 1999 surveys there is a slight though insignificant difference in diarrhoea prevalence levels between the poorest and the least poor. This leads to the conclusion that children from all socio-economic groups are equally at risk to diarrhoea. Medical treatment in case of diarrhoea is slightly more common among the least poor than among the poorest. The same holds true for medical treatment in case of fever and in case of cough. Even though diarrhoea is an acute disease, having diarrhoea during the past 2 weeks preceding the survey may indicate the overall risk to this disease. Table 6.2 shows children who had at least one episode of diarrhoea have an increased probability of being stunted. Data also shows, the risk of becoming chronically malnourished due to diarrhoea increases with age; children under 36 months with an episode of diarrhoea have a 10% higher risk of being stunted and this increase to almost 25% for children between 3 and 5 years (relative risk). Differences in health care seeking behaviour in case of diarrhoea don't seem to have an impact on the eventual nutritional status of the child. Data even suggest a slight increase in stunting prevalence for those children who were medically treated for diarrhoea. ORS treatment shows the same, but stronger effect. It might very well be that `apparent' negative treatment effect of ORS has more to do with the severity and duration of the diarrhoea episodes than with the actual effectiveness of ORS treatment; ORS is more likely to be given when diarrhoea is persistent and lasts for a relatively long time Fever According to WHO, worldwide in 1995 "pneumonia, diarrhoea, malaria and measles caused one half of all deaths of children under five. Many of these children also suffered from malnutrition, a causative or contributory factor in an estimated 54% of child deaths, bringing the total proportion of child deaths due to these five conditions to 70%." Fever should be considered a proxy for malaria. As known malaria is one of the main causes of child mortality in Tanzania, both direct as well as indirect through the increase of the prevalence of anemia. Anemia itself is an important eroding factor of the nutritional status of children. Fever during the 2 weeks preceding the survey was frequent. 33% of the children were reported to have had fever during this period. There has been very little change in the prevalence of fever/malaria over the 1990s. The relatively high fever prevalence rates in children under five can be observed in regions situated along the coast, e.g. Coast and Morogoro (41%), Tanga (38%) and Dar es Salaam (36%). The highest fever/malaria incidence rates were observed in Mara (45%). Lowest fever prevalence rates were observed in Iringa (23%) and Ruvuma (25%). The regions situated along the Coast show the highest malaria incidence rates. Unlike diarrhoea, fever episodes seem to have the greatest impact on the nutritional status of the younger children. The effect of fever is comparable with the effects of diarrhoea in the ages up to 36 months. Medical treatment of fever/malaria is expected to have a positive effect on the nutritional status of children; 40% of children under the age of 36 26 Trends in Malnutrition in Tanzania months who experienced a fever episode during the 2 weeks preceding the survey were moderate to severely stunted compared to 50% of children from the same age group who did not receive any medical treatment for fever. Breastfeeding The World Health Organization (WHO) recommends exclusive breastfeeding, without supplementary feeding or liquids, for the first six months of the child's life. "Breast milk is the natural first food for babies, it provides all the energy and nutrients that the infant needs for the first months of life, and it continues to provide up to half or more of a child's nutritional needs during the second half of the first year, and up to one-third during the second year of life.... Breast milk promotes sensory and cognitive development, and protects the infant against infectious and chronic diseases. Exclusive breastfeeding reduces infant mortality due to common childhood illnesses such as diarrhoea or pneumonia, and helps for a quicker recovery during illness." WHO Breastfeeding is universal in Tanzania; data show high rates of breastfeeding, though the trend seems to be slightly downward from 99% of the children in 1991/92 to 96% in 1999. Due to this high rate the number of non-breastfed children becomes very small so it was not possible to show a significant benefit of breastfeeding for the nutritional status of the children (see Table 6.2). Table 6.3 Exclusive breastfeeding by wealth/poverty status, 1991-1999 1991/92 1996 1999 Combined poorest 20% 22[16-28] 30[22-39] 41[24-59] 29[23-35] second 20% 18[10-25] 30[21-40] 21[5-37] 23[17-29] middle 20% 10[4-16] 22[14-30] 31[13-50] 19[13-25] fourth 20% 13[8-19] 21[13-29] 23[6-41] 18[13-23] least poor 20% 9[4-14] 16[8-23] 20[7-34] 13[9-18] Note: 95% Confidence Intervals are displayed between [ ] Source: Authors' calculation using DHS 1991/92, 1996 and 1999 For the children up to the age of 6 months we examined the impact of exclusive breastfeeding on their nutritional status. Exclusive breastfeeding has shown a steady and significant increase over the 1990s. The poverty/wealth analysis of exclusive breastfeeding showed that it was more frequent among the poorest households. In fact, exclusive breastfeeding in the poorest quintile rose from 22% in 1991/92 to a little over 40% in 1999. For the least poor group the rate of exclusive breastfeeding rose from 9% in 1991/92 to 20% in 1999 (see Table 6.3). Exclusive breastfeeding therefore seems to be much more common in the poorest households than in the least poor households; this gives the poorest children a better protection during the first few months. Table 6.2 underlines the importance of exclusive breastfeeding during the first months of life. Of the children less than six months who were exclusively breastfed, the prevalence of stunting was 8%, compared to 15% for children who received food supplements. This indicates a risk reduction of close to 50%. 27 Trends in Malnutrition in Tanzania Maternal education with respect to breastfeeding still is of great importance. A study done by TFNC (2002) in Iringa rural district found showed that only 7% of the recent born infants were given colostrums (first yellowish milk). It was generally believed that colostrums is bad and dirty and therefore should not be given to the recent born. Furthermore, most children are introduced to any solid or liquid food between the ages of 0 to 4 months (83%), the main reason was due to insufficient breast milk. Birth Weight Low birth weight depends to a large extent on the well being of the mother before getting pregnant as well as during pregnancy. Inadequate nutritional intake, incidences of malaria and related illnesses have a negative effect on the development of the fetus. In addition, teenage pregnancies have an increased probability of low birth weight. Maternal height is related to difficult delivery, as small stature is often associated with small pelvic size and short women have the risk of bearing infants with low birth weight (DHS 1996). According to Raman Kutty, in the US teenage pregnancies and smoking have lead to an increased proportion of babies with low birth weights. While in India the main contribution to low birth weight is short stature and low pregravid weight of mothers, and also maternal infection during pregnancy, and at times inadequate coverage by antenatal services. Birth weight shows little variation over the 1990s and between the different socio- economic groups. Though absolute numbers are too small to be conclusive, the 1999 data suggests a lower incidence of birth weight among the least poor 20%, compared to the other groups. The consequences of a low birth weight on the height for age indicator are substantial. Of the children who were born with a weight below 2500 grams, 54% were considered to be moderate to severely stunted up to the age of 36 months, compared to 33% stunting in those born with a normal birth weight. Data suggest that even after the age of 3, the effects of low birth weight are still to be felt in a significant proportion of the children. Under-five mortality For each under five it was determined whether a sibling had died during early childhood. All three surveys show about a 10% difference in sibling death between the poorest and the least poor. The data does not indicate that child mortality decreased during the 1990s. Mortality levels among the first to the last quintile differ only slightly. According to UNICEF, 70% of the under-five deaths result from diarrhoea, acute respiratory infections, measles, malaria, and malnutrition. Death from malaria stated above, constitutes about 50% of the deaths. 6.2 Water and sanitation Source of drinking water Drinking water is one of the main sources of contamination that threatens the health of young children and also adults. The provision of safe drinking water is of great importance for the prevention of water-borne diseases. Access to clean water is of vital importance especially in densely populated poor urban areas. Lacking the proper means 28 Trends in Malnutrition in Tanzania to dispose waste means that diarrhoea is nearly unavoidable. Using clean drinking water can prevent diarrhoea and also cholera. As was expected, a good and safe supply of drinking water is associated with decreased risks of under-five malnutrition and this effect seems to increase with age. The prevalence of stunting in children living in households having access to piped water in their own residence is only 25%, compared to 47% in children where any kind of surface water is used as main source of drinking water. Children who only have access to surface water are almost twice as likely to become stunted than children using any other source of drinking water. Sanitation Table 6.2 shows, the flush toilet and also the access to an improved ventilated pit latrine reduce the risk of being stunted. Access to a flush toilet seems to have the strongest effect. The proportion stunted in children with access to a flush toilet is only 19%, whereas the stunting prevalence in children in households with access to a traditional pit latrine is 44%. Good and clean sanitation reduces the risk of infection, protecting the health and nutritional status of young children. In particular, hand washing after using latrines, or before touching food, is very crucial to reduce the spread of diarrhoea. Diarrhoea, fever, low birth weight and sanitation have negative impact in the nutrition status of children while those who receive treatment have a higher risk. Breast on the other hand is more frequent amongst poorest households; exclusive breast-feeding, piped water and flush toilets have risk reduction in stunting. 29 Trends in Malnutrition in Tanzania 7. Multivariate analysis In the previous sections, regional, socio-economic and health correlates were discussed in terms of their impact on the stunting (expressed as 2 standard deviation or more below the median of the standard population). As already mentioned, some of the explanatory variables used are mutually correlated. For example, better educated women tend to live in households of a higher socio-economic status, which in most cases have access to higher quality facilities. All these factors combined are most likely to have a positive effect on the nutritional status of children. In this section we look at the independent effects of the analysed indicators on the levels of the z-scores of height for age, with the aim to determine the independent effect of socio-economic, health and environmental characteristics on these z-scores. We will look into the independent effects of selected indicators for: - Infants up to 6 months to assess the effects of exclusive breastfeeding - Children under the age of 36 months - Children between the age of 36 and 60 months - All children under the age of 5. Only those explanatory variables were included for which at least 90% of the observations valid values were available. Since the previous analyses already demonstrated the importance of low birth weight on the nutritional status for each of the age groups, separate models were generated to demonstrate the impact of low birth weight. One has to bear in mind that the models using birth weight as an independent variable are not only influenced by the impact of this variable but are also driven by exclusion of observations, since birth weight measures were present in less than 50% of the observations. To compensate for possible differences in parameter estimates between the different surveys, interaction terms of the independent variables with DHS 1991/92 and DHS 1996 were calculated. In case the parameter estimates of the independent variables reached p- values of less than 0.2 these interaction terms were included in the model and were retained in the final model when the p-value of the interaction terms fell below 0.05. Standard errors of the parameter estimates were adjusted for survey design effect (clustering and stratification). The size of the parameter will indicate the impact the independent has on the height of the standardized z-scores, a positive sign indicating a positive effect on the nutritional status a negative sign therefore indicates a negative effect. The R2 displayed at the bottom of each model expresses the amount of variation in the dependent variable that is explained by the model. All models included will be discussed briefly. Finally the model for all under-fives with the variable birth weight included will be used for simulation purposes to assess the effects of change in some of the significant independent variables on the nutritional status of the children. 30 Trends in Malnutrition in Tanzania 7.1 Independent effect of exclusive breastfeeding in infants under 6 months Table 7.1a shows the independent predictors for the standardized z-scores for height for age in infants up to the age of 6 months. Interest mainly goes to testing whether exclusive breastfeeding up to six months can be considered an independent reducing factor of stunting in infants. Both models, including and excluding birth weight, show no significant independent effect of exclusive breastfeeding during the first six months on the z-scores for height for age. Water and sanitation don't seems to play a significant role within the age group up to 6 months As also demonstrated in Figure 2.3a the nutritional status deteriorates with increasing age and immediately from birth, boys seem to do significantly worse than girls. For the infants under 6 months of age, mothers with higher age seem to have infants with better nutritional status, however this relationship in not linear; the positive effect of higher age of the mother decreases when the age of the mother increases. Southern regions such as Lindi, Mtwara, Ruvuma, Iringa, Morogoro and Dodoma do have significant independent negative impact on the nutritional status of infants less than 6 months. When adding low birth weight to the model (see Table 7.1b), this variable shows a significant independent effect on height for age. Most likely due to the reduction in number of observations, most of the regional variables become insignificant. From this analysis it becomes clear that exclusive breastfeeding in infants up to 6 months does not have the independent positive impact as was observed in the univariate analysis. The R2's displayed in Tables 7.1a and b show that only 17% and 27% of variation in the standardized z-scores can be explained by the models, meaning that important variables that also determine the nutritional status are not captured in the analysis. 31 Trends in Malnutrition in Tanzania Table 7.1a Independent predictors of moderate to severe stunting in children under the age of 6 months: Model without birth weight Parameter estimate Standard error Pr>|t| Intercept -1.258 0.938 0.180 Floor: earth/sand -0.173 0.123 0.158 Electricity 0.200 0.164 0.222 RADIO -0.005 0.074 0.948 TELEV -0.107 0.404 0.792 REFRIG -0.411 0.354 0.246 BICYCLE 0.010 0.075 0.889 MOTORCY 0.188 0.352 0.593 CAR 0.160 0.256 0.533 Water: Piped into res. -0.226 0.241 0.349 Water: Public tap -0.088 0.209 0.675 Water: Well -0.121 0.207 0.559 Water: Spring -0.042 0.247 0.864 Water: river/stream/ pond/lake -0.098 0.219 0.653 Toilet: Flush 0.202 0.295 0.494 Toilet: Traditional pit -0.097 0.119 0.416 Toilet: Vent impr pit 0.126 0.240 0.600 Religion: Moslem 0.080 0.174 0.645 Religion: Protestant 0.150 0.151 0.319 Religion: Catholic 0.116 0.150 0.441 Female Headed Household 0.165 0.096 0.085 Age of head of household 0.006 0.003 0.053 Number of household members -0.004 0.018 0.844 # of women aged 15-45 in the household 0.039 0.063 0.539 # of under -5 in the household -0.044 0.042 0.298 Fever during the past 2 weeks -0.106 0.085 0.214 Diarrhoea in the past 2 weeks -0.038 0.105 0.719 Cough during the past 2 weeks -0.017 0.083 0.840 Child death experienced by mother -0.167 0.085 0.049 Ever breastfed 0.249 0.325 0.444 Exclusive breastfed -0.109 0.089 0.222 No solid food given 0.025 0.094 0.791 Sex of child 1=male 0= female -0.174 0.062 0.005 Age of child -0.288 0.069 <.0001 Age square of child 0.018 0.011 0.098 Birth order 0.001 0.031 0.974 Previous birth interval: < 24 months 0.080 0.133 0.547 Previous birth interval: 24 -47 months 0.242 0.128 0.059 Previous birth interval: 48+ months 0.273 0.145 0.061 Age of mother 0.078 0.039 0.046 Age square of mother -0.001 0.001 0.042 Mother currently working -0.013 0.078 0.870 No education 0.249 0.571 0.663 Primary education 0.213 0.571 0.709 Secondary education 0.105 0.597 0.861 RURAL -0.089 0.109 0.414 DODOMA -0.725 0.202 0.000 ARUSHA 0.200 0.254 0.432 KILIMANJARO 0.070 0.221 0.750 TANGA -0.171 0.257 0.505 MOROGORO -0.425 0.211 0.045 COAST -0.134 0.258 0.603 DAR -0.196 0.244 0.422 LINDI -0.460 0.245 0.061 MTWARA -0.634 0.232 0.007 RUVUMA -0.604 0.238 0.011 IRINGA -0.523 0.219 0.017 MBEYA -0.266 0.238 0.265 SINGIDA -0.110 0.188 0.556 TABORA -0.085 0.274 0.755 RUKWA -0.171 0.238 0.473 KIGOMA -0.083 0.235 0.725 SHINYANGA -0.161 0.174 0.356 KAGERA -0.088 0.201 0.663 MARA -0.073 0.217 0.737 Interaction of DHS 1991 with LINDI -0.768 0.264 0.004 Number of Observations Used 1548 R-square 0.166 Source: Authors' calculation using DHS 1991/92, DHS 1996 and DHS 1999 32 Trends in Malnutrition in Tanzania Table 7.1b Independent predictors of moderate to severe stunting in children under the age of 6 months: Model with birth weight Parameter estimate Standard error Pr>|t| Intercept -1.767 1.559 0.258 Floor: earth/sand 0.054 0.168 0.748 Electricity 0.238 0.165 0.150 RADIO 0.143 0.099 0.152 TELEV -0.393 0.430 0.361 REFRIG -0.319 0.284 0.261 BICYCLE 0.096 0.113 0.399 MOTORCY 0.540 0.296 0.068 CAR 0.258 0.209 0.220 Water: Piped into res. -0.145 0.267 0.587 Water: Public tap -0.112 0.230 0.626 Water: Well -0.125 0.226 0.580 Water: Spring -0.224 0.272 0.411 Water: river/stream/ pond/lake -0.063 0.242 0.795 Toilet: Flush 0.070 0.342 0.838 Toilet: Traditional pit -0.088 0.184 0.633 Toilet: Vent impr pit 0.325 0.305 0.287 Religion: Moslem 0.116 0.263 0.659 Religion: Protestant 0.169 0.243 0.488 Religion: Catholic 0.150 0.246 0.543 Female Headed Household 0.160 0.135 0.238 Age of head of household 0.004 0.004 0.333 Number of household members 0.000 0.024 0.989 # of women aged 15-45 in the household 0.108 0.098 0.273 # of under -5 in the household -0.063 0.074 0.395 Fever during the past 2 weeks 0.033 0.134 0.804 Diarrhoea in the past 2 weeks 0.247 0.155 0.112 Cough during the past 2 weeks -0.087 0.122 0.478 Child death experienced by mother -0.231 0.125 0.066 Ever breastfed 0.587 0.631 0.352 Exclusive breastfed -0.143 0.142 0.315 No solid food given 0.210 0.135 0.120 Birth weight <= 2500 gr. -0.628 0.128 <.0001 Sex of child 1=male 0= female -0.236 0.098 0.017 Age of child -0.263 0.108 0.016 Age square of child 0.010 0.016 0.530 Birth order 0.031 0.050 0.540 Previous birth interval: < 24 months -0.124 0.213 0.561 Previous birth interval: 24 -47 months 0.302 0.195 0.123 Previous birth interval: 48+ months 0.173 0.201 0.390 Age of mother 0.115 0.076 0.130 Age square of mother -0.002 0.001 0.104 Mother currently working -0.097 0.116 0.405 No education 0.115 0.577 0.842 Primary education 0.110 0.573 0.848 Secondary education -0.150 0.615 0.808 RURAL -0.123 0.135 0.363 DODOMA -0.973 0.343 0.005 ARUSHA -0.147 0.393 0.709 KILIMANJARO -0.148 0.401 0.712 TANGA -0.108 0.490 0.826 MOROGORO -0.571 0.411 0.166 COAST -0.563 0.469 0.231 DAR -0.358 0.399 0.370 LINDI -0.851 0.411 0.039 MTWARA -0.606 0.416 0.146 RUVUMA -0.940 0.407 0.021 IRINGA -0.485 0.379 0.201 MBEYA -0.644 0.415 0.121 SINGIDA -0.440 0.401 0.274 TABORA 0.014 0.452 0.975 RUKWA -0.353 0.491 0.472 KIGOMA -0.196 0.413 0.636 SHINYANGA -0.592 0.372 0.113 KAGERA -0.221 0.418 0.598 MARA -0.531 0.390 0.174 Number of Observations Used 745 R-square 0.271 Source: Authors' calculation using DHS 1991/92, DHS 1996 and DHS 1999 33 Trends in Malnutrition in Tanzania 7.2 Independent predictors of stunting in children under 36 months Tables 7.2a and b show the independent effects of regional, socio-economic as well as health related correlates on the standardized z-scores of children under the age of 36 months, again as in the previous models with and without the inclusion of low birth weight. For the comparison of the models for up to 6 months the direct environmental variables become more important for both models with and without birth weight. Poor floor quality (earth or sand) has a significant negative effect on the nutritional status. Water and hygiene show significant effects on nutritional status using spring as a source of drinking water has an independent negative effect, whereas the use of a flush toilet has a positive independent effect. When looking at household size it seems that children in larger households are better off. However, a larger number of under-fives has a negative effect, probably due to inter- child competition (this effect was not observed in the univariate analysis, probably due to the correlation between household size and number of children under-five). For children under 36 months, prevalence of fever (proxy for malaria) is more important than the prevalence of diarrhoea; fever shows in both models an independent negative effect on the nutritional status. The effect of sex of the child is similar to the effect observed in the models for infants under 6 months. Also the age of the child has a strong negative impact on the nutritional status, however, as observed before, the relationship is not entirely linear. In the model excluding low birth weight long preceding birth intervals show a positive effect on nutritional status. Looking at region of residence it becomes clear that the nutritional status in the southern regions is significantly worse. As in the model up to 6 months (Table 7.1b), also in Table 7.2b, low birth weight has a significant negative impact on the nutritional status of children under the age of 36 months. The R2's of both models show that only between 20% and 30% of the variation in the standardized z-scores can be explained by the models. 34 Trends in Malnutrition in Tanzania Table 7.2a Independent predictors of moderate to severe stunting in children under the age of 36 months: Model without birth weight Parameter estimate Standard error Pr>|t| Intercept -0.371 0.566 0.513 Floor: earth/sand -0.347 0.053 <.0001 Electricity 0.026 0.092 0.775 RADIO 0.117 0.044 0.008 TELEV 0.281 0.205 0.171 REFRIG -0.119 0.212 0.575 BICYCLE 0.007 0.043 0.871 MOTORCY -0.029 0.196 0.883 CAR 0.275 0.160 0.085 Water: Piped into res. 0.040 0.131 0.761 Water: Public tap 0.049 0.116 0.669 Water: Well 0.000 0.112 0.998 Water: Spring -0.088 0.125 0.483 Water: river/stream/ pond/lake -0.020 0.116 0.863 Toilet: Flush 0.374 0.184 0.042 Toilet: Traditional pit 0.009 0.053 0.871 Toilet: Vent impr pit 0.056 0.176 0.750 Religion: Moslem -0.025 0.083 0.761 Religion: Protestant 0.126 0.063 0.047 Religion: Catholic 0.056 0.062 0.362 Female Headed Household 0.030 0.052 0.560 Age of head of household 0.001 0.001 0.318 Number of household members 0.018 0.010 0.062 # of women aged 15-45 in the household -0.051 0.029 0.079 # of under -5 in the household -0.049 0.021 0.020 Fever during the past 2 weeks -0.146 0.039 0.000 Diarrhoea in the past 2 weeks -0.045 0.045 0.316 Cough during the past 2 weeks -0.074 0.038 0.053 Child death experienced by mother -0.111 0.040 0.005 Ever breastfed 0.242 0.166 0.145 Sex of child 1=male 0= female -0.086 0.031 0.006 Age of child -0.142 0.007 <.0001 Age square of child 0.003 0.000 <.0001 Birth order -0.025 0.017 0.157 Previous birth interval: < 24 months -0.026 0.077 0.734 Previous birth interval: 24 -47 months 0.090 0.057 0.116 Previous birth interval: 48+ months 0.151 0.063 0.017 Age of the mother 0.046 0.020 0.021 Age square of mother -0.001 0.000 0.078 Mother currently working -0.041 0.039 0.294 No education -0.406 0.458 0.376 Primary education -0.349 0.456 0.444 Secondary education -0.114 0.469 0.807 RURAL -0.053 0.059 0.370 DODOMA -0.615 0.098 <.0001 ARUSHA -0.003 0.094 0.977 KILIMANJARO -0.119 0.119 0.318 TANGA -0.035 0.150 0.814 MOROGORO -0.372 0.097 0.000 COAST -0.025 0.118 0.835 DAR -0.115 0.128 0.370 LINDI -0.621 0.118 <.0001 MTWARA -0.060 0.230 0.793 RUVUMA -0.455 0.109 <.0001 IRINGA -0.581 0.133 <.0001 MBEYA -0.293 0.098 0.003 SINGIDA -0.034 0.110 0.759 TABORA 0.058 0.104 0.577 RUKWA -0.101 0.100 0.314 KIGOMA -0.282 0.095 0.003 SHINYANGA 0.073 0.083 0.376 KAGERA -0.058 0.085 0.500 MARA 0.102 0.094 0.273 Interaction of DHS 1991 with COAST -0.289 0.140 0.039 Interaction of DHS 1996 with COAST -0.511 0.170 0.003 Interaction of DHS 1991 with MTWARA -0.605 0.255 0.018 Interaction of DHS 1996 with MTWARA -0.496 0.236 0.036 Interaction of DHS 1996 with IRINGA -0.442 0.144 0.002 Number of Observations Used 7691 R-square 0.220 Source: Authors' calculation using DHS 1991/92, DHS 1996 and DHS 1999 35 Trends in Malnutrition in Tanzania Table 7.2b Independent predictors of moderate to severe stunting in children under the age of 36 months: Model with birth weight Parameter estimate Standard error Pr>|t| Intercept -0.095 0.690 0.891 Floor: earth/sand -0.345 0.062 <.0001 Electricity -0.072 0.094 0.442 RADIO 0.110 0.063 0.083 TELEV 0.245 0.207 0.237 REFRIG -0.187 0.201 0.351 BICYCLE -0.056 0.063 0.370 MOTORCY 0.201 0.187 0.281 CAR 0.270 0.170 0.113 Water: Piped into res. -0.136 0.148 0.357 Water: Public tap -0.141 0.139 0.313 Water: Well -0.259 0.139 0.063 Water: Spring -0.342 0.150 0.023 Water: river/stream/ pond/lake -0.217 0.143 0.131 Toilet: Flush 0.418 0.197 0.034 Toilet: Traditional pit -0.019 0.087 0.825 Toilet: Vent impr pit 0.044 0.205 0.831 Religion: Moslem 0.149 0.115 0.196 Religion: Protestant 0.342 0.100 0.001 Religion: Catholic 0.232 0.102 0.024 Female Headed Household -0.010 0.062 0.870 Age of head of household 0.001 0.002 0.526 Number of household members 0.028 0.013 0.041 # of women aged 15-45 in the household -0.026 0.043 0.545 # of under -5 in the household -0.077 0.032 0.017 Fever during the past 2 weeks -0.118 0.054 0.028 Diarrhoea in the past 2 weeks -0.074 0.065 0.254 Cough during the past 2 weeks -0.050 0.048 0.301 Child death experienced by mother -0.217 0.067 0.001 Ever breastfed 0.349 0.236 0.139 Birth weight -0.511 0.058 <.0001 Sex of child 1=male 0= female -0.115 0.051 0.023 Age of child -0.141 0.009 <.0001 Age square of child 0.003 0.000 <.0001 Birth order 0.011 0.025 0.662 Previous birth interval: < 24 months -0.039 0.117 0.738 Previous birth interval: 24 -47 months 0.067 0.081 0.409 Previous birth interval: 48+ months 0.157 0.082 0.056 Age of the mother 0.043 0.031 0.158 Age square of mother -0.001 0.001 0.153 Mother currently working -0.074 0.050 0.138 No education -0.385 0.444 0.386 Primary education -0.320 0.438 0.466 Secondary education -0.101 0.452 0.823 RURAL 0.024 0.059 0.683 DODOMA -0.715 0.132 <.0001 ARUSHA -0.258 0.163 0.114 KILIMANJARO -0.231 0.163 0.157 TANGA -0.137 0.172 0.427 MOROGORO -0.386 0.140 0.006 COAST -0.416 0.168 0.014 DAR -0.246 0.152 0.105 LINDI -0.626 0.170 0.000 MTWARA -0.586 0.166 0.000 RUVUMA -0.686 0.139 <.0001 IRINGA -0.870 0.150 <.0001 MBEYA -0.458 0.140 0.001 SINGIDA -0.338 0.177 0.057 TABORA 0.013 0.157 0.936 RUKWA -0.178 0.154 0.249 KIGOMA -0.457 0.161 0.005 SHINYANGA -0.032 0.140 0.820 KAGERA -0.187 0.140 0.184 MARA -0.129 0.154 0.403 Number of observations used 3786 R-square 0.262 Source: Authors' calculation using DHS 1991/92, DHS 1996 and DHS 1999 7.3 Independent predictors of stunting in children between 36 and 60 months Table 7.3a Independent predictors of moderate to severe stunting in children between the age of 36 and 60 months: Model without birth weight 36-59 Parameter estimate Standard error Pr>|t| 36 Trends in Malnutrition in Tanzania Intercept 0.118 1.390 0.933 Floor: earth/sand -0.381 0.083 <.0001 Electricity 0.322 0.103 0.002 RADIO 0.215 0.054 <.0001 TELEV 0.103 0.167 0.540 REFRIG -0.151 0.186 0.416 BICYCLE 0.344 0.114 0.003 MOTORCY 0.626 0.230 0.007 CAR 0.534 0.229 0.020 Water: Piped into res. -0.276 0.171 0.106 Water: Public tap -0.314 0.157 0.046 Water: Well -0.254 0.154 0.099 Water: Spring -0.312 0.171 0.068 Water: river/stream/ pond/lake -0.308 0.161 0.056 Toilet: Flush -0.201 0.252 0.424 Toilet: Traditional pit -0.071 0.085 0.405 Toilet: Vent impr pit 0.112 0.362 0.758 Religion: Moslem -0.114 0.091 0.215 Religion: Protestant -0.040 0.089 0.657 Religion: Catholic -0.021 0.086 0.808 Female Headed Household 0.059 0.070 0.402 Age of head of household 0.001 0.002 0.518 Number of household members 0.011 0.014 0.458 # of women aged 15-45 in the household -0.011 0.039 0.787 # of under -5 in the household -0.019 0.025 0.445 Fever during the past 2 weeks -0.101 0.063 0.108 Diarrhoea in the past 2 weeks -0.318 0.097 0.001 Cough during the past 2 weeks 0.067 0.059 0.261 Child death experienced by mother 0.001 0.053 0.982 Ever breastfed 0.159 0.172 0.356 Sex of child 1=male 0= female -0.125 0.043 0.003 Age of child -0.049 0.050 0.329 Age square of child 0.001 0.001 0.310 Birth order -0.018 0.022 0.405 Previous birth interval: < 24 months -0.327 0.090 0.000 Previous birth interval: 24 -47 months -0.117 0.078 0.132 Previous birth interval: 48+ months 0.028 0.102 0.784 Age of the mother 0.016 0.032 0.613 Age square of mother 0.000 0.001 0.940 Mother currently working -0.174 0.089 0.049 No education -0.773 0.562 0.169 Primary education -0.708 0.560 0.206 Secondary education -0.493 0.572 0.389 RURAL -0.152 0.076 0.046 DODOMA -0.408 0.150 0.007 ARUSHA -0.008 0.133 0.953 KILIMANJARO 0.077 0.130 0.554 TANGA -0.290 0.140 0.038 MOROGORO -0.375 0.104 0.000 COAST -0.594 0.174 0.001 DAR -0.087 0.131 0.505 LINDI -0.474 0.162 0.004 MTWARA -0.685 0.147 <.0001 RUVUMA -0.339 0.110 0.002 IRINGA -0.732 0.136 <.0001 MBEYA -0.197 0.176 0.262 SIGINDA 0.027 0.132 0.841 TABORA 0.005 0.173 0.979 RUKWA -0.001 0.130 0.994 KIGOMA -0.144 0.117 0.218 SHINYANGA 0.309 0.102 0.003 KAGERA 0.028 0.111 0.800 MARA 0.356 0.140 0.011 Interaction of DHS 1991 with BICYC -0.279 0.133 0.036 Interaction of DHS 1996 with BICYC -0.222 0.129 0.086 Interaction of DHS 1991 with Mother currently working 0.240 0.088 0.007 Interaction of DHS 1996 with Mother currently working 0.218 0.091 0.017 Number of Observations Used 3966 R-square 0.143 Source: Authors' calculation using DHS 1991/92, DHS 1996 and DHS 1999 37 Trends in Malnutrition in Tanzania Table 7.3b Independent predictors of moderate to severe stunting in children between the age of 36 and 60 months: Model with birth weight Parameter estimate Standard error Pr>|t| Intercept 0.265 1.892 0.888 Floor: earth/sand -0.277 0.092 0.003 Electricity 0.250 0.118 0.035 RADIO 0.304 0.071 <.0001 TELEV 0.091 0.184 0.621 REFRIG -0.165 0.193 0.392 BICYCLE 0.238 0.085 0.005 MOTORCY 0.681 0.251 0.007 CAR 0.439 0.262 0.094 Water: Piped into res. -0.434 0.152 0.004 Water: Public tap -0.406 0.131 0.002 Water: Well -0.772 0.175 <.0001 Water: Spring -0.255 0.153 0.095 Water: river/stream/ pond/lake -0.350 0.141 0.013 Toilet: Flush 0.097 0.209 0.642 Toilet: Traditional pit -0.005 0.104 0.961 Toilet: Vent impr pit 0.341 0.395 0.388 Religion: Moslem -0.061 0.159 0.703 Religion: Protestant -0.107 0.153 0.486 Religion: Catholic 0.013 0.145 0.927 Female Headed Household 0.106 0.089 0.232 Age of head of household 0.001 0.003 0.611 Number of household members 0.003 0.016 0.874 # of women aged 15-45 in the household 0.036 0.044 0.415 # of under -5 in the household -0.030 0.036 0.405 Fever during the past 2 weeks -0.189 0.076 0.013 Diarrhoea in the past 2 weeks -0.313 0.128 0.015 Cough during the past 2 weeks 0.107 0.074 0.147 Child death experienced by mother 0.005 0.066 0.941 Ever breastfed 0.252 0.251 0.317 Birth weight -0.254 0.068 0.000 Sex of child 1=male 0= female -0.283 0.072 <.0001 Age of child -0.038 0.068 0.576 Age square of child 0.000 0.001 0.574 Birth order -0.033 0.027 0.225 Previous birth interval: < 24 months -0.372 0.121 0.002 Previous birth interval: 24 -47 months -0.160 0.104 0.124 Previous birth interval: 48+ months -0.066 0.129 0.610 Age of mother 0.011 0.048 0.821 Age square of mother 0.000 0.001 0.957 Mother currently working 0.069 0.065 0.288 No education -0.762 0.622 0.221 Primary education -0.740 0.614 0.229 Secondary education -0.513 0.620 0.408 RURAL -0.154 0.071 0.030 DODOMA -0.696 0.185 0.000 ARUSHA -0.355 0.177 0.045 KILIMANJARO -0.181 0.144 0.210 TANGA -0.134 0.200 0.502 MOROGORO -0.687 0.132 <.0001 COAST -0.411 0.176 0.020 DAR -0.319 0.150 0.034 LINDI -0.106 0.179 0.552 MTWARA -1.018 0.183 <.0001 RUVUMA -0.693 0.138 <.0001 IRINGA -1.150 0.117 <.0001 MBEYA -0.437 0.156 0.005 SIGINDA -0.054 0.178 0.763 TABORA -0.064 0.180 0.724 RUKWA 0.724 0.392 0.065 KIGOMA -0.267 0.159 0.094 SHINYANGA 0.126 0.147 0.392 KAGERA -0.234 0.139 0.094 MARA 0.181 0.150 0.228 Interaction of DHS 1991 with BICYC -0.282 0.128 0.028 Interaction of DHS 1996 with Water: Well 0.491 0.156 0.002 Interaction of DHS 1996 with Water: Well 0.531 0.156 0.001 Interaction of DHS 1991 with sex of child 0.233 0.102 0.022 Interaction of DHS 1996 with TANGA -0.738 0.259 0.005 Interaction of DHS 1991 with COAST -0.508 0.241 0.036 Interaction of DHS 1996 with COAST -0.495 0.195 0.012 Interaction of DHS 1991 with LINDI -0.600 0.284 0.035 Interaction of DHS 1996 with LINDI -1.008 0.262 0.000 Interaction of DHS 1991 with RUKWA -1.138 0.431 0.008 Interaction of DHS 1996 with RUKWA -1.228 0.413 0.003 Number of Observations Used 2030 R-square 0.239 Source: Authors' calculation using DHS 1991/92, DHS 1996 and DHS 1999 An important observation that can be made when comparing the models for the age group up to 36 months to the age group 36-60 months is that asset-variables, used as indicators for the wealth/poverty status of the household, are more important for the nutritional status of the older children than for the younger children. Though not all the parameters 38 Trends in Malnutrition in Tanzania estimated for the variables related to source of drinking water are significant, the impact of the different sources seems to differ only a little. Household hygiene, expressed by the use of different toilet facilities, has no independent significant effect on height for age in this age group. Whereas fever/malaria had a significant impact on nutritional status for children below 36 months, it has no independent impact in children between 36 and 60 months. For these children diarrhoea is of importance. Differences between boys and girls remain present even between 36 and 60 months. The significant impact of age seems to have disappeared after 36 months, which was also shown in Figure 2.3a. Also, the older children seem to benefit from longer birth intervals. Furthermore, the significant impact of the mother currently working is negative. But, when taking into account the interaction terms with the surveys of 1991/92 and 1996 the sign changes for observations from these two surveys, and this effect remains only negative for observations from the 1999 survey. The impact of region of residence is comparable with the previous analyses, only the difference between north and south become more pronounced Residing in Mara and Shinyanga has a significant positive impact on nutritional status. Low birth weight remains an important independent predictor of nutritional status, even after 3 years children were not able to compensate for the unequal start at birth. Again the R2's are low,14% and 26% respectively for the without and with low birth weight. 7.4 Independent predictors of stunting in all children less than 5 years of age Table 7.4a and b summarize the effects of independent variables on height for age for the whole population of children under-five. These models hide to a large extent the differences observed between the children under 36 months and those between 36 and 60 months. The difference between the model with and without low birth weight also seems to be larger for the overall population. Looking at the characteristics of the child, sex is important factor throughout early childhood. Age plays an important role at early age up to 36 months. Due to the dominance of the under 36 months in the overall under-five population, also at this level age is of significant importance. Earth/sand floors show throughout a negative effect on nutritional status. However, in the overall under-five population, household assets as proxies for poverty/wealth don't play a very important role. Both fever and diarrhoea are significant independent predictors of nutritional status. Again the southern regions show a significant negative impact on nutritional status. 39 Trends in Malnutrition in Tanzania Table 7.4a Independent predictors of moderate to severe stunting in children under fives: Model with birth weight Parameter estimate Standard error Pr>|t| Intercept -0.131 0.478 0.785 Floor: earth/sand -0.119 0.099 0.228 Electricity 0.117 0.070 0.093 RADIO 0.161 0.036 <.0001 TELEV 0.147 0.137 0.284 REFRIG -0.084 0.150 0.574 BICYCLE 0.056 0.039 0.149 MOTORCY 0.193 0.172 0.262 CAR 0.346 0.142 0.015 Water: Piped into res. -0.041 0.097 0.676 Water: Public tap -0.047 0.084 0.577 Water: Well -0.066 0.080 0.414 Water: Spring -0.136 0.092 0.139 Water: river/stream/ pond/lake -0.098 0.084 0.241 Toilet: Flush 0.213 0.161 0.186 Toilet: Traditional pit -0.019 0.052 0.720 Toilet: Vent impr pit 0.059 0.187 0.753 Religion: Moslem -0.061 0.071 0.391 Religion: Protestant 0.062 0.056 0.263 Religion: Catholic 0.019 0.055 0.732 Female Headed Household 0.052 0.043 0.233 Age of head of household 0.001 0.001 0.270 Number of household members 0.014 0.009 0.125 # of women aged 15-45 in the household -0.045 0.025 0.080 # of under -5 in the household -0.024 0.017 0.155 Fever during the past 2 weeks -0.151 0.033 <.0001 Diarrhoea in the past 2 weeks -0.138 0.041 0.001 Cough during the past 2 weeks -0.033 0.034 0.334 Child death experienced by mother -0.065 0.035 0.064 Ever breastfed 0.151 0.131 0.250 Sex of child 1=male 0= female -0.100 0.026 0.000 Age of child -0.086 0.003 <. 0001 Age square of child 0.001 0.000 <. 0001 Birth order -0.024 0.013 0.072 Previous birth interval: < 24 months -0.130 0.055 0.019 Previous birth interval: 24 -47 months 0.013 0.042 0.762 Previous birth interval: 48+ months 0.101 0.053 0.057 Age of the mother 0.033 0.017 0.054 Age square of the mother 0.000 0.000 0.235 Mother currently working -0.029 0.033 0.384 No education -0.603 0.374 0.107 Primary education -0.531 0.372 0.153 Secondary education -0.293 0.382 0.443 RURAL -0.393 0.100 <. 0001 DODOMA -0.517 0.094 <. 0001 ARUSHA 0.009 0.086 0.913 KILIMANJARO -0.035 0.100 0.725 TANGA -0.115 0.123 0.351 MOROGORO -0.358 0.081 <. 0001 COAST -0.366 0.097 0.000 DAR -0.111 0.098 0.259 LINDI -0.580 0.105 <. 0001 MTWARA -0.579 0.107 <. 0001 RUVUMA -0.411 0.086 <. 0001 IRINGA -0.745 0.100 <. 0001 MBEYA -0.239 0.104 0.021 SIGINDA 0.002 0.092 0.980 TABORA 0.044 0.104 0.675 RUKWA -0.054 0.088 0.537 KIGOMA -0.220 0.084 0.009 SHINYANGA 0.155 0.068 0.023 KAGERA -0.024 0.078 0.762 MARA 0.220 0.092 0.017 Interaction of DHS 1991 with Floor: earth/sand -0.319 0.103 0.002 Interaction of DHS 1996 with Floor: earth/sand -0.259 0.110 0.019 Interaction of DHS 1991 with RURAL 0.419 0.109 0.000 Interaction of DHS 1996 with RURAL 0.300 0.113 0.008 Number of Observations Used 11657 R-square 0.190 Source: Authors' calculation using DHS 1991/92, DHS 1996 and DHS 1999 40 Trends in Malnutrition in Tanzania Table 7.4b Independent predictors of moderate to severe stunting in children under fives: Model with birth weight Parameter estimate Standard error Pr>|t| Intercept 0.267 0.570 0.640 Floor: earth/sand -0.339 0.054 <. 0001 Electricity 0.053 0.073 0.470 RADIO 0.186 0.051 0.000 TELEV 0.137 0.142 0.336 REFRIG -0.167 0.146 0.255 BICYCLE 0.009 0.057 0.880 MOTORCY 0.354 0.176 0.045 CAR 0.319 0.159 0.045 Water: Piped into res. -0.259 0.115 0.025 Water: Public tap -0.181 0.113 0.109 Water: Well -0.313 0.107 0.004 Water: Spring -0.337 0.118 0.004 Water: river/stream/ pond/lake -0.285 0.113 0.012 Toilet: Flush 0.326 0.162 0.044 Toilet: Traditional pit -0.014 0.068 0.831 Toilet: Vent impr pit 0.155 0.214 0.470 Religion: Moslem 0.093 0.107 0.387 Religion: Protestant 0.192 0.090 0.034 Religion: Catholic 0.159 0.090 0.078 Female Headed Household 0.038 0.054 0.490 Age of head of household 0.001 0.002 0.401 Number of household members 0.017 0.011 0.123 # of women aged 15-45 in the household -0.011 0.033 0.742 # of under -5 in the household -0.045 0.025 0.075 Fever during the past 2 weeks -0.161 0.045 0.000 Diarrhoea in the past 2 weeks -0.162 0.057 0.005 Cough during the past 2 weeks -0.011 0.043 0.796 Child death experienced by mother -0.125 0.052 0.016 Ever breastfed 0.232 0.167 0.166 Birth weight -0.428 0.048 <.0001 Sex of child 1=male 0= female -0.161 0.036 <.0001 Age of child -0.078 0.004 <.0001 Age of child^2 0.001 0.000 <.0001 Birth order -0.012 0.018 0.489 Previous birth interval: < 24 months -0.161 0.083 0.054 Previous birth interval: 24 -47 months -0.011 0.061 0.850 Previous birth interval: 48+ months 0.100 0.067 0.132 age of the mother 0.023 0.024 0.327 age of the mother^2 0.000 0.000 0.476 Mother currently working -0.045 0.042 0.275 No education -0.631 0.390 0.106 Primary education -0.574 0.384 0.135 Secondary education -0.360 0.393 0.359 RURAL -0.009 0.052 0.861 DODOMA -0.737 0.136 <.0001 ARUSHA -0.276 0.128 0.032 KILIMANJARO -0.205 0.126 0.103 TANGA -0.047 0.158 0.766 MOROGORO -0.514 0.109 <.0001 COAST -0.565 0.136 <.0001 DAR -0.267 0.110 0.016 LINDI -0.684 0.145 <.0001 MTWARA -0.740 0.135 <.0001 RUVUMA -0.661 0.106 <.0001 IRINGA -0.986 0.110 <.0001 MBEYA -0.440 0.116 0.000 SIGINDA -0.226 0.125 0.072 TABORA 0.010 0.128 0.939 RUKWA -0.260 0.126 0.039 KIGOMA -0.374 0.129 0.004 SHINYANGA 0.046 0.102 0.651 KAGERA -0.204 0.118 0.084 MARA 0.018 0.125 0.884 DW_PT96 -0.132 0.072 0.067 TANGA96 -0.591 0.202 0.004 Number of Observations Used 5816 R-square 0..234 Source: Authors' calculation using DHS 1991/92, DHS 1996 and DHS 1999 41 Trends in Malnutrition in Tanzania The overall model including birth weight was used to measure possible effects of changes in some of the independent variables. For this purpose we assessed the impact of change in the prevalence of fever, diarrhoea, the incidence of low birth weight and the increased use of flush toilets keeping all other factors constant. Table 7.5 displays the results of this exercise. Table 7.5: Simulations outcome No change 25% change 50% change 75% 100% change change Observed stunting 40.2 Fever during the past 2 weeks 39.6 39 38.7 38.4 Diarrhoea in the past 2 weeks 39.8 39.6 39.5 39.4 Birth weight <= 2,500 gr. 39.4 38.6 38 37.1 Toilet: Flush 37.8 35.2 32.6 29.8 Source: Authors' calculation using DHS 1991/92, DHS 1996 and DHS 1999 The observed rate of moderate to severe stunting in the population used for modeling was 40.2%. A reduction in the prevalence of fever/malaria from 25% to 100% shows a drop in stunting of only 1.2%, the effects of the change in diarrhoea prevalence are even less. Reduction in the incidence of low birth weight is slightly higher, but still not impressive, given the investment needed to accomplish these reductions. The largest effect is shown by the increased use of flush toilets; where household had a flush toilet, keeping everything else constant, the prevalence of stunting is projected to go down to 29.8%. Exclusive breastfeeding does not have independent effect on nutritional status of children. For the children under 36 months, poor floor quality, spring as source of drinking water, larger number of under fives and fever have independent negative effect on nutritional status while on the other hand, children in larger households and long preceding birth interval. For children between 36 and 60 months, assets variables (proxy for poverty/wealth) and diarrhoea are important. Here birth interval, residing in the North (Mara and Shinyanga) has independent positive effect on nutritional status. Low birth weight and mother currently working have negative effect. For all under fives, residing in the southern regions, poor quality of floor, fever and diarrhoea have independent negative effect on nutritional status while access to a flush toilet has independent positive effect on nutritional status 42 Trends in Malnutrition in Tanzania 8. Conclusion and recommendation Previous analysis as well as this study shows that the nutritional status of children under- five did not improve during the 1990s. The same holds true for the poverty situation. Both head count ratio's based in the 1991 and 2000/01 Household Budget Survey as well the constructed assets based wealth/poverty index indicate little or no change during the last decennium of the 20th Century. The nutritional status of the children under the age of five in Tanzania is comparable to the status of children in surrounding East African Countries. Within Tanzania, though poverty has its impact on malnutrition, it is not the only factor of importance. The analysis of malnutrition by wealth/poverty quintile showed only marked differences between the least poor 20% and the other 4 groups. It needs to be said, that the least poor 20% is dominated by households living in urban areas, whereas the other 4 quintiles are dominated by rural households; so poverty analysis hides to a large extent urban ­ rural differences. Based on information from the three Demographic and Health Surveys a pooled analysis of malnutrition with focus on moderate to severe stunting showed the importance of the following factors: - Low birth weight - Modern and hygienic toilet facilities - Prevalence of fever/malaria in children under the age of 36 months - Prevalence of diarrhoea in children between the age of 36 and 60 months Contrary to what was expected, exclusive breastfeeding did not have an independent positive effect on the nutritional status. Unfortunately information on exclusive breastfeeding could only be derived from children currently being breastfed, so the long term effects of exclusive breastfeeding could not be assessed. Of the endogenous factors, age (up to 36 months) and sex were important determinants of nutritional status. As the data shows, the poor nutritional status at later ages finds its origin in the first two years of life. As of birth, boys do worse than girls, and even after the age of 36 months they were not able to catch up. The univariate analysis showed that Muslim religion was a significant factor contributing to a deprived nutritional status of children. In the multivariate analysis religion however did not come out as an independent predictor of nutritional status, which makes it likely to believe that differences observed in the univariate analysis with respect to religion are most likely caused by region of residence. The finding of variation in stunting caused by month of birth was one of the most surprising findings of this study. Even though survey data were pooled, this only provided sufficient power to show differences, but not enough power to look for explanations. Though information was limited on low birth weight, no seasonal fluctuation in incidence of low birth weights was observed. However, as mentioned before, there seems to be variation in exclusive breastfeeding by month of birth. Children born during July, August and September have a higher probability of exclusive breastfeeding than children born during any of the other months; a smaller proportion of children born in February are exclusively breastfed. It should be mentioned that exclusive 43 Trends in Malnutrition in Tanzania breastfeeding did not prove to be an independent explanatory variable of nutritional status. There seems to be a positive relationship between nutritional status of children under-five and the number of household members, children in larger household tend to have a reduced probability of being stunted. As the univariate analysis showed, this risk reduction can be observed in households that have 10 or more household members. Although the univariate analysis indicated that children who live together with more other children in the same age group tend to be better of, the multivariate analysis shows the opposite effect, probably caused by the high correlation with total household size. As expected, more children lead to more competition for scarce resources, such as food but also individual attention. To come up with recommendations based on this research seems to be difficult. The multivariate analyses clearly demonstrated that the major part of the variation in the nutritional status of children under the age of five could not be explained by the information available in the surveys. There are important variables, probably both endogenous as well as exogenous, that could give additional explanation of the nutritional status of the child; one could think of consumption pattern of the household and within household food distribution. It is important to stress that focused research in the area of child nutrition is needed. 44 Trends in Malnutrition in Tanzania Bibliography Blanc, Ann K., Brent Wolff, Anastasia J. Gage, Alex C. Ezeh, Stella Neema and John Ssekamatte-Ssebuliba.1996.Negotiating Reproductive Outcomes in Uganda. Calverton, Maryland: Macro International Inc. and Institute of Statistics and Applied Economics [Uganda]. 1995/96 DHS Bureau of Statistics (Tanzania) and Macro International Inc. (1997). "Tanzania Demographic and Health Survey 1996". Calveston, Maryland: Bureau of Statistics and Macro International. Cole, T.J., Parkin, J.M., (1977). "Infection and its effect on growth of young children: A comparison of the Gambia and Uganda". Transactions of the Royal Society of Tropical Medicine and Hygiene 71, 196-198. Filmer, Deon, & Pritchett, Lant, (1998), "Estimating wealth effects without expenditure data ­ or tears: An application of educational enrollment in states of India," Mimeo. The World Bank, Washington, DC Gabriele Doblhammer (1999), "Longevity and month of birth evidence from Austria and Denmark" Max Planck Institute for Demographic Research volume 1, article 3 www.demographic-research.org/Volumes/Vol1/3 Rostock, Germany Gwatkin D.R., S. Rutstein, K. Johnson, R.P. Pande and A. Wagstaff. (2000). Socio- Economic Differences in Health, Nutrition and Population: Tanzania. World Bank Howard, Mary, and Ann V. Millard. (1997)."Hunger and Shame Child Malnutrition and Poverty on Mount Kilimanjaro." New York and London: Routledge. Kamuzora, C. L. and Mkanta W., (2000), "Poverty and Family Size in Tanzania: Multiple Responses to Population Pressure?" REPOA. Research Report No. 00.4 Kutty, V. Raman, (2004) "Why low birth weight is still a problem in Kerala: A preliminary exploration." Discussion Paper No.57 KRPCDS, Kerala, India. Mgoba, C, Elizabeth Macha, Sabas Kimboka, and Wilbert Lorri (2002) `Baseline survey report for initiation of community based maise flour fortification project in Iringa rural District, Tanzania' TFNC National Bureau of Statistics, (1992), " Demographic Health Survey (DHS) 1991" Dar es Salaam National Bureau of Statistics, (2000), "Tanzania Reproductive and Child Health Survey (TRCHS) 1999." Dar es Salaam National Bureau of Statistics, (2002), "Household Budget Survey (HBS) 2000/01." Dar es Salaam 45 Trends in Malnutrition in Tanzania National Council for Population and Development (NCPD), Central Bureau of Statistics (CBS) (Office of the Vice President and Ministry of Planning and National Development [Kenya]), and Macro International Inc. (MI). (1994.) "Kenya Demographic and Health Survey 1993". Calverton, Maryland: NCPD, CBS, and MI. National Council for Population and Development (NCPD), Central Bureau of Statistics (CBS) (Office of the Vice President and Ministry of Planning and National Development) [Kenya], and Macro International Inc. (MI). (1999.) "Kenya Demographic and Health Survey 1998". Calverton, Maryland: NDPD, CBS, and MI. National Statistical Office, Macro International Inc., (1994), "1992 Malawi Demographic and Health Survey (MDHS)" Zomba, Malawi and Calveston, Maryland National Statistical Office, Macro International Inc., (1997), "1996 Malawi Demographic and Health Survey (MDHS)" Zomba, Malawi and Calveston, Maryland Uganda Bureau of Statistics (UBOS) and ORC Macro. (2001). "Uganda Demographic and Health Survey 2000-2001" Calverton, Maryland, USA: UBOS and ORC Macro. URT (2002) " Poverty and Human Development Report 2002". R&AWG Dar es Salaam URT (2003) " Poverty and Human Development Report 2003". R&AWG Dar es Salaam WHO "11 Million Child Deaths Per Year: Many Can Be Prevented" Press Release WHO/64 WHO Child and Adolescent Health Exclusive Breastfeeding http://www.who.int/child-adolescent-health/NUTRITION/infant_exclusive.htm accessed 20th April 2004. Younger, Stephen D., (2004) "Growth and Poverty Reduction in Uganda, 1992-1999: A Multidimensional Analysis of Changes in Living Standards". Mimeo. 46 Trends in Malnutrition in Tanzania Appendix Table A: Regional Muslim composition Region % Muslim 1 Coast 92.72 2 Lindi 79.76 3 Mtwara 78.61 4 Tanga 77.56 5 Dar es Salaam 69.1 6 Morogoro 52.59 7 Tabora 32.77 8 Singida 30.62 9 Ruvuma 30.57 10 Kigoma 26.91 11 Kilimanjaro 23.89 12 Dodoma 23.07 13 Arusha 12.33 14 Kagera 8.59 15 Mbeya 8.41 16 Rukwa 8.25 17 Mara 4.74 18 Shinyanga 3.34 19 Mwanza 2.3 20 Iringa 1.89 Source: Authors' calculation using DHS 1991/92, DHS 1996 and DHS 1999 Table B: Distribution of religions by wealth status 1991-1999 Poorest quintile Second quintile Middle quintile Fourth quintile Least poor quintile Muslim 21.3 19.13 27.08 25.62 38.41 Protestants 21.48 26.33 25.41 27.61 30.03 Catholic 28.83 31.27 35.13 36.22 30.12 Source: Authors' calculation using DHS 1991/92, DHS 1996 and DHS 1999 Figure A. Zonal differences in stunting for 1991 and 1996 70.0 60.0 50.0 Nothern Highlands 40.0 Lake Central 30.0 Southern Highland South 20.0 10.0 0.0 moderate severe moderate severe 1991 1996 Source: Demographic and Health Surveys, 1991/92, 1996 &1999 and authors' calculations 47 PART 5 Appendixes Statistical Tables APPENDIX A Population and Demographics Appendixes Table A.1. Summary of Population in Tanzania, 1988 and 2002 1988 census 2002 census Mainland Zanzibar Tanzania Mainland Zanzibar Tanzania Number of population (in thousands) 22,486 641 23,127 33,585 985 34,569 Population under 15 years old (%) 46 47 46 44 44 44 Sex ratio (males per 100 female) 94 95 94 96 96 96 Male (in thousands) 10,907 311 11,218 16,428 483 16,910 Female (in thousands) 11,580 329 11,909 17,157 502 17,659 Infant mortality rate (per 1,000 live births) 115 120 115 115 120 115 Crude death rate (per 1,000 population) 15 15 15 15 15 15 Life expectancy at birth 49 48 49 49 48 49 Crude birth rate (per 1,000 population) 46 49 46 46 49 46 Total fertility rate (%) 6.2 7.1 6.5 6.2 7.1 6.5 Sources: National Bureau of Statistics, Population and Housing Censuses for 1998 and 2002; World Bank World Development Indicators 2005. Note: The 1988 and 2002 census results show a smaller population size than estimates of the United Nations. 286 Population and Demographics Table A.2. Population Size by Region, 1967­2002 (thousands) Region 1967 1978 1988 2002 Mainland 11,958.7 17,036.5 22,533.8 33,584.6 Arusha 610.5 926.2 1,351.7 1,293.0 Coast 428.0 516.6 638.0 889.2 Dar es Salaam 356.3 843.1 1,360.9 2,497.9 Dodoma 709.4 972 1,237.8 1,699.0 Iringa 689.9 925.0 1,208.9 1,495.3 Kagera 658.7 1,009.8 1,326.2 2,033.9 Kigoma 473.4 648.9 854.8 1,679.1 Kilimanjaro 652.7 902.4 1,108.7 1,381.1 Lindi 419.9 527.6 646.6 791.3 Mara 544.1 723.8 970.9 1,368.6 Mbeya 753.8 1,079.9 1,476.2 2,070.0 Morogoro 682.7 939.3 1,222.7 1,759.8 Mtwara 621.3 771.8 889.5 1,128.5 Mwanza 1,055.9 1,443.4 1,878.3 2,942.1 Rukwa 276.1 451.9 695 1,141.7 Ruvuma 395.4 561.6 783.3 1,117.2 Shinyanga 899.5 1,323.5 1,772.5 2,805.6 Singida 457.9 613.9 791.8 1,090.8 Tabora 502.1 817.9 1,036.3 1,717.9 Tanga 771.1 1,037.8 1,283.6 1,642.0 Manyaraa n.a. n.a. n.a. 1,040.5 Zanzibar 354.8 476.1 640.6 984.6 Total 12,313.5 17,512.6 23,174.3 34,569.2 Sources: National Bureau of Statistics, Population and Housing Censuses for 1967, 1978, 1998 and 2002; World Bank World Development Indicators 2005. Note: n.a. = not applicable. The 1988 and 2002 census results show a smaller population size than estimates of the United Nations. a. Manyara is a newly created region that resulted from the division of the Arusha region into two parts: Arusha and Manyara. 287 Appendixes Table A.3. Dependency Ratios 1967 1978 1988 1991­ 1994 1996 2002 2004 Age group census census census 92 DHS KAPS DHS census DHS Under 15 years 43.9 46.1 45.8 46.8 49.3 47.2 44.2 46.9 15­64 years 50.5 49.9 49.9 49.3 46.4 48.5 51.9 48.9 64 years and up 5.6 4.0 4.3 3.9 4.3 4.3 3.9 4.2 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Median age U U U 16.4 15.4 16.4 16.4 16.4 Dependency ratio 98.0 100.4 100.4 102.8 115.5 106.2 92.9 104.5 Sources: National Bureau of Statistics, Population and Housing censuses for 1967, 1978, 1998 and 2002; Tanzania Demographic and Health Surveys for 1996 and 2004. Note: U = unknown; DHS = Demographic and Health Survey; KAPS = Knowledge, Attitudes, and Practice Survey. 288 APPENDIX B The Economy Table B.1. Gross Domestic Product by Sector at Current Prices, 1995­2005 millions of Tanzania shillings 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005a Sector Agriculture, for- estry, fishing, and hunting 1,318,459 1,658,275 2,003,763 2,295,027 2,694,374 3,021,157 3,406,146 3,884,521 4,417,855 5,227,457 6,025,848 290 Mining and quarrying 35,190 38,511 53,515 74,386 85,792 99,519 120,454 152,977 210,574 278,262 368,141 Manufacturing 200,525 254,326 295,272 382,901 434,544 499,726 564,689 638,663 710,951 791,416 891,135 Electricity and water 60,347 65,800 74,599 81,751 101,301 112,753 124,789 145,753 156,963 177,614 202,499 Construction 109,429 132,248 188,123 255,330 305,859 343,355 405,159 469,984 546,121 637,768 741,912 Trade, hotels, and restaurants 417,626 493,572 562,760 635,305 740,181 823,025 926,870 1,038,094 1,153,323 1,319,172 1,513,090 Transport and communication 159,771 193,946 219,393 250,081 294,180 328,259 361,558 404,945 451,281 509,948 580,754 Finance, insur- ance, and real estate 353,080 451,961 570,686 731,257 816,672 920,595 1,075,806 1,240,697 1,407,197 1,550,265 1,801,089 Public admini- stration and other services 255,401 300,876 457,087 564,779 649,553 709,351 796,930 893,083 956,209 1,044,229 1,154,682 Imputed bank services ­113,187 ­136,957 ­138,244 ­141,009 ­144,756 ­151,359 ­157,785 ­168,830 ­194,155 ­204,494 ­215,833 GDP at factor cost 2,796,641 3,452,558 4,286,954 5,129,808 5,977,700 6,706,381 7,624,616 8,699,887 9,816,319 11,331,638 13,063,317 Indirect taxes less subsidies 223,859 315,083 421,859 446,330 455,212 561,999 649,991 732,076 861,554 1,033,865 1,145,775 GDP at market prices 3,020,500 3,767,641 4,708,813 5,576,138 6,432,912 7,268,380 8,274,607 9,431,963 10,677,873 12,321,183 14,209,092 Memo items Agriculture 1,318,459 1,658,275 2,003,763 2,295,027 2,694,374 3,021,157 3,406,146 3,884,521 4,417,855 5,227,457 6,025,848 Industry 405,492 490,885 611,509 794,368 927,496 1,055,353 1,215,091 1,407,377 1,624,609 1,885,060 2,203,687 Services 1,072,691 1,303,398 1,666,328 2,035,530 2,355,830 2,629,871 3,003,379 3,407,989 3,773,855 4,219,120 4,833,782 Indirect taxes 247,745 340,999 446,673 446,330 455,212 561,999 649,991 745,595 880,867 1,057,040 1,173,637 Subsidies 23,886 25,916 24,814 .. .. .. .. 13,519 19,313 23,175 27,862 GDP at market prices (US$ million) 5,631 6,463 7,616 8,371 8,635 9,081 9,443 9,792 10,276 11,298 12,607 Sources: United Republic of Tanzania, National Accounts of Tanzania; The Economic Survey (various issues). Note: Years are calendar years. a. Preliminary estimate. 291 Table B.2. Share of Gross Domestic Product at Market Prices by Sector, 1995­2005 percentage of GDP at market prices 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005a Sector Agriculture, forestry, fishing, and hunting 43.7 44.0 42.6 41.2 41.9 41.6 41.2 41.2 41.4 42.4 42.4 Mining and quarrying 1.2 1.0 1.1 1.3 1.3 1.4 1.5 1.6 2.0 2.3 2.6 Manufacturing 6.6 6.8 6.3 6.9 6.8 6.9 6.8 6.8 6.7 6.4 6.3 Electricity and water 2.0 1.7 1.6 1.5 1.6 1.6 1.5 1.5 1.5 1.4 1.4 Construction 3.6 3.5 4.0 4.6 4.8 4.7 4.9 5.0 5.1 5.2 5.2 Trade, hotels, and restaurants 13.8 13.1 12.0 11.4 11.5 11.3 11.2 11.0 10.8 10.7 10.6 Transport and communication 5.3 5.1 4.7 4.5 4.6 4.5 4.4 4.3 4.2 4.1 4.1 Finance, insurance, and real estate 11.7 12.0 12.1 13.1 12.7 12.7 13.0 13.2 13.2 12.6 12.7 Public administration 292 and other services 8.5 8.0 9.7 10.1 10.1 9.8 9.6 9.5 9.0 8.5 8.1 Imputed bank services ­3.7 ­3.6 ­2.9 ­2.5 ­2.3 ­2.1 ­1.9 ­1.8 ­1.8 ­1.7 ­1.5 GDP at factor cost 92.6 91.6 91.0 92.0 92.9 92.3 92.1 92.2 91.9 92.0 91.9 Indirect taxes less subsidies 7.4 8.4 9.0 8.0 7.1 7.7 7.9 7.8 8.1 8.4 8.1 GDP at market prices 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Memo items Agriculture 43.7 44.0 42.6 41.2 41.9 41.6 41.2 41.2 41.4 42.4 42.4 Industry 13.4 13.0 13.0 14.2 14.4 14.5 14.7 14.9 15.2 15.3 15.5 Services 35.5 34.6 35.4 36.5 36.6 36.2 36.3 36.1 35.3 34.2 34.0 Indirect taxes 8.2 9.1 9.5 8.0 7.1 7.7 7.9 7.9 8.2 8.6 8.3 Subsidies 0.8 0.7 0.5 .. .. .. .. 0.1 0.2 0.2 0.2 Source: Table B.1. Note: Years are calendar years. a. Preliminary estimate. Table B.3. Share of Gross Domestic Product at Factor Cost by Sector, 1995­2005 percentage of GDP at factor cost 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005a Sector Agriculture, forestry, fishing, and hunting 47.1 48.0 46.7 44.7 45.1 45.0 44.7 44.7 45.0 46.1 46.1 Mining and quarrying 1.3 1.1 1.2 1.5 1.4 1.5 1.6 1.8 2.1 2.5 2.8 Manufacturing 7.2 7.4 6.9 7.5 7.3 7.5 7.4 7.3 7.2 7.0 6.8 Electricity and water 2.2 1.9 1.7 1.6 1.7 1.7 1.6 1.7 1.6 1.6 1.6 Construction 3.9 3.8 4.4 5.0 5.1 5.1 5.3 5.4 5.6 5.6 5.7 Trade, hotels, and restaurants 14.9 14.3 13.1 12.4 12.4 12.3 12.2 11.9 11.7 11.6 11.6 Transport and communication 5.7 5.6 5.1 4.9 4.9 4.9 4.7 4.7 4.6 4.5 4.4 Finance, insurance, and real estate 12.6 13.1 13.3 14.3 13.7 13.7 14.1 14.3 14.3 13.7 13.8 Public administration 293 and other services 9.1 8.7 10.7 11.0 10.9 10.6 10.5 10.3 9.7 9.2 8.8 Imputed bank services ­4.0 ­4.0 ­3.2 ­2.7 ­2.4 ­2.3 ­2.1 ­1.9 ­2.0 ­1.8 ­1.7 GDP at factor cost 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Indirect taxes less subsidies 8.0 9.1 9.8 8.7 7.6 8.4 8.5 8.4 8.8 9.1 8.8 GDP at market prices 108.0 109.1 109.8 108.7 107.6 108.4 108.5 108.4 108.8 108.7 108.8 Memo items Agriculture 47.1 48.0 46.7 44.7 45.1 45.0 44.7 44.7 45.0 46.1 46.1 Industry 14.5 14.2 14.3 15.5 15.5 15.7 15.9 16.2 16.6 16.6 16.9 Services 38.4 37.8 38.9 39.7 39.4 39.2 39.4 39.2 38.4 37.2 37.0 Source: Table B.1. Note: Years are calendar years. a. Preliminary estimate. Table B.4. Gross Domestic Product by Sector at Constant 1992 Prices, 1995­2005 millions of Tanzania shillings 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005a Sector Agriculture, for- estry, fishing, and hunting 682,338 708,741 726,098 739,942 770,510 796,513 840,275 882,107 917,395 970,378 1,020,497 294 Mining and quarrying 18,768 20,579 24,097 30,700 33,488 38,144 43,293 49,787 58,749 67,798 78,443 Manufacturing 106,750 111,894 117,489 126,887 131,491 137,809 144,647 156,219 169,653 184,218 200,797 Electricity and water 21,578 23,977 24,514 25,870 26,874 28,454 29,297 30,201 31,688 33,123 34,815 Construction 50,985 54,868 59,341 65,187 70866 76,818 83,494 92,678 102,872 113,994 127,533 Trade, hotels, and restaurants 210,813 218,105 229,010 239,830 254,114 270,567 288,718 308,928 329,009 354,726 383,814 Transport and communication 70,833 71,597 75,099 79,755 84,403 89,515 95,154 101,244 106,294 112,648 119,833 Finance, insur- ance, and real estate 138509 139,059 149,762 158,089 164,568 172,291 177,911 186,485 194,711 203,222 213,961 Public admini- stration and other services 109,763 111,473 115,007 118,114 122,207 126,567 130,987 136,307 141,880 147,950 155,490 Imputed bank services ­65,090 ­58,581 ­72,327 ­78,547 ­81,229 ­82,359 ­84,418 ­86,781 ­89,819 ­93,541 ­98,104 GDP at factor cost 1,345,247 1,401,712 1,448,090 1,505,827 1,577,292 1,654,319 1,749,358 1,857,175 1,962,432 2,094,516 2,237,079 Indirect taxes less subsidiesa 107,681 127,921 130,210 131,133 117,451 126,847 142,998 172,232 212,888 257,765 306,740 GDP at market prices 1,452,928 1,529,633 1,578,300 1,636,960 1,694,743 1,781,166 1,892,356 2,029,407 2,175,320 2,352,281 2,543,819 Memo items Agriculture 682,338 708,740 726,098 739,942 770,510 796,513 840,275 882,107 917,395 970,378 1,020,497 Industry 198,081 211,318 225,441 248,644 262,719 281,225 300,731 328,885 362,962 399,133 441,588 Services 464,827 481,653 496,551 517,241 544,063 576,581 608,352 646,183 682,075 725,005 774,994 GDP per capita at factor costb 46,576 49,615 49,727 50,162 50,971 51,861 53,199 55,299 57,299 59,389 61,803 Sources: United Republic of Tanzania, National Accounts of Tanzania; The Economic Survey (various issues). Note: Years are calendar years. a. Preliminary estimate. b. Zanzibar population is excluded. 295 Table B.5. Annual Growth of Gross Domestic Product at Constant 1992 Prices, 1995­2005 percentage 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005a Sector Agriculture, for- estry, fishing, and hunting 5.8 3.9 2.4 1.9 4.1 3.4 5.5 5.0 4.0 5.8 5.2 296 Mining and quarrying 11.7 9.6 17.1 27.4 9.1 13.9 13.5 15.0 18.0 15.4 15.7 Manufacturing 1.6 4.8 5.0 8.0 3.6 4.8 5.0 8.0 8.6 8.6 9.0 Electricity and water 6.1 11.1 2.2 5.5 3.9 5.9 3.0 3.1 4.9 4.5 5.1 Construction ­14.7 7.6 8.2 9.9 8.7 8.4 8.7 11.0 11.0 10.8 11.9 Trade, hotels, and restaurants 3.5 3.5 5.0 4.7 6.0 6.5 6.7 7.0 6.5 7.8 8.2 Transport and communication 5.9 1.1 4.9 6.2 5.8 6.1 6.3 6.4 5.0 6.0 6.4 Finance, insur- ance, and real estate 0.6 0.4 7.7 5.6 4.1 4.7 3.3 4.8 4.4 4.4 5.3 Public administra- tion and other services ­2.7 1.6 3.2 2.7 3.5 3.6 3.5 4.1 4.1 4.3 5.1 Imputed bank services ­5.4 ­10.0 23.5 8.6 3.4 1.4 2.5 2.8 3.5 4.1 4.9 GDP at factor cost 3.6 4.2 3.3 4.0 4.7 4.9 5.7 6.2 5.7 6.7 6.8 Indirect taxes less subsidies 1.5 18.8 1.8 0.7 ­10.4 8.0 12.7 20.4 23.6 21.1 19.0 GDP at market prices 3.4 5.3 3.2 3.7 3.5 5.1 6.2 7.2 7.2 8.1 8.1 Memo items Agriculture 5.8 3.9 2.4 1.9 4.1 3.4 5.5 5.0 4.0 5.8 5.2 Industry ­1.9 6.7 6.7 10.3 5.7 7.0 6.9 9.4 10.4 10.0 10.6 Services 2.8 3.6 3.1 4.2 5.2 6.0 5.5 6.2 5.6 6.3 6.9 GDP per capita at factor costb 0.6 1.3 0.2 0.9 1.6 1.7 2.6 3.9 3.6 3.6 4.1 Sources: United Republic of Tanzania, National Accounts of Tanzania; The Economic Survey (various issues). Note: Years are calendar years. a. Preliminary estimate. b. Zanzibar population is excluded. 297 Table B.6. Gross National Product by Expenditure at Current Prices, 1995­2004 millions of Tanzania shillings 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 GDP at market prices 3,020,500 3,767,641 4,708,813 5,576,138 6,432,912 7,268,380 8,274,607 9,431,963 10,677,873 12,321,184 Resource gap ­526,563 ­452,356 ­445,484 ­816,352 ­817,811 ­611,571 ­678,070 ­586,378 ­923,475 ­1,192,088 Exports, goods, and nonfactor services 727,177 751,161 762,812 748,973 885,942 1,064,773 1,284,715 1,520,522 1,887,672 2,364,731 Imports, goods, and nonfactor services 1,253,740 1,203,517 1,208,296 1,565,325 1,703,753 1,676,344 1,962,785 2,106,900 2,811,147 3,556,819 Gross domestic 298 expenditures 3,547,062 4,219,998 5,154,297 6,392,490 7,250,723 7,879,951 8,952,677 10,018,341 11,601,348 13,513,272 Consumption 2,995,156 3,565,401 4,381,635 5,343,036 6,118,574 6,552,293 7,433,905 8,098,586 9,261,803 10,835,443 Private 2,532,841 3,130,072 3,968,072 4,909,250 5,667,437 6,069,576 6,917,579 7,499,647 8,549,066 9,980,158 Public 462,315 435,329 413,563 433,786 451,137 482,717 516,326 598,939 712,737 855,285 Investment 597,792 627,237 700,803 902,614 999,650 1,281,046 1,406,302 1,807,750 2,259,019 2,593,502 Gross fixed capital formation 591,936 620,597 692,400 892,700 989,338 1,266,678 1,390,641 1,789,896 2,239,281 2,570,704 Private and non- governmental organization 494,961 493,000 557,211 736,534 820,914 875,810 978,963 1,148,465 1,430,803 1,763,138 Private 492,285 489,629 553,678 707,077 789,472 828,265 928,948 1,073,938 1,343,118 1,620,627 Nongovernmental organization 2,676 3,371 3,533 29,457 31,442 47,546 50,015 74,527 87,685 142,511 Public 99,345 127,868 134,105 156,167 168,424 390,867 411,678 641,432 808,478 807,565 Central government 23,727 16,313 25,860 128,079 136,547 354,638 371,322 580,701 737,307 712,557 Public enterprises 75,618 111,555 108,245 28,088 31,877 36,229 40,356 60,731 71,171 95,008 Changes in stocks 5,856 6,640 8,403 9,914 10,312 14,368 15,661 17,854 19,738 22,798 Errors and omissions ­45,886 27,360 71,859 146,840 132,499 46,612 112,470 112,005 80,526 84,327 Gross national product 2,961,319 3,725,682 4,633,504 5,506,198 6,362,137 7,171,932 8,199,859 9,382,166 10,621,346 12,267,160 Gross domestic savings 25,344 202,240 327,178 233,102 314,338 716,087 840,702 1,333,377 1,416,070 1,485,741 Net factor income ­59,181 ­41,959 ­75,309 ­69,941 ­70,775 ­96,448 ­74,748 ­49,797 ­56,527 ­54,024 Receipt 17,047 24,206 26,587 29,575 36,505 40,340 48,547 65,209 88,635 95,498 Payment 76,228 66,165 101,896 99,515 107,280 136,788 123,295 115,006 145,162 149,522 Net current transfers 181,512 311,730 266,821 284,625 250,767 313,677 338,778 405,026 692,872 648,182 Receipt 198,838 330,561 308,655 302,543 331,972 372,026 400,119 466,863 734,540 707,991 Payment 17,326 18,831 41,834 17,918 81,205 58,349 61,341 61,837 41,668 59,809 Gross national savings 147,675 472,012 518,690 447,786 494,330 933,316 1,104,731 1,688,605 2,052,415 2,079,899 Sources: United Republic of Tanzania, National Accounts of Tanzania; The Economic Survey (various issues). Note: Years are calendar years. 299 Table B.7. Share of Gross National Product by Expenditure at Current Prices, 1995­2004 percentage of GDP at market prices 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 GDP at factor cost 92.6 91.6 90.9 91.9 92.9 92.4 92.1 92.2 92.0 91.6 Indirect taxes 8.2 9.1 9.5 8.0 7.1 7.7 7.9 7.9 8.2 8.6 Less subsidies 0.8 0.7 0.5 .. .. .. .. 0.1 0.2 0.2 300 GDP at market prices 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Resource gap ­17.4 ­12.0 ­9.5 ­14.6 ­12.7 ­8.4 ­8.2 ­6.2 ­8.6 ­9.7 Exports, goods, and nonfac- tor services 24.1 19.9 16.2 13.4 13.8 14.6 15.5 16.1 17.7 19.2 Imports, goods, and nonfac- tor services 41.5 31.9 25.7 28.1 26.5 23.1 23.7 22.3 26.3 28.9 Gross domestic expenditures 100.0 100.0 99.9 99.9 100.0 100.0 100.0 100.0 100.1 100.0 Consumption 99.2 94.6 93.1 95.8 95.1 90.1 89.8 85.9 86.7 87.9 Private 83.9 83.1 84.3 88.0 88.1 83.5 83.6 79.5 80.1 81.0 Public 15.3 11.6 8.8 7.8 7.0 6.6 6.2 6.4 6.7 6.9 Investment 19.8 16.6 14.9 16.2 15.5 17.6 17.0 19.2 21.2 21.0 Gross fixed capital formation 19.6 16.5 14.7 16.0 15.4 17.4 16.8 19.0 21.0 20.9 Private 16.4 13.1 11.8 13.2 12.8 12.0 11.8 12.2 13.4 14.3 Public 3.3 3.4 2.8 2.8 2.6 5.4 5.0 6.8 7.6 6.6 Central government 0.8 0.4 0.5 2.3 2.1 4.9 4.5 6.2 6.9 5.8 Public enterprises 2.5 3.0 2.3 0.5 0.5 0.5 0.5 0.6 0.7 0.8 Changes in stocks 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 Errors and omissions ­1.5 0.7 1.4 2.5 2.1 0.6 1.4 1.2 0.8 0.7 Gross national product 98.0 98.9 98.4 98.7 98.9 98.7 99.1 99.5 99.5 99.6 Gross domestic savings 0.8 5.4 6.9 4.2 4.9 9.9 10.2 14.1 13.3 12.1 Net factor income ­2.0 ­1.1 ­1.6 ­1.3 ­1.1 ­1.3 ­0.9 ­0.5 ­0.5 ­0.4 Receipt 0.6 0.6 0.6 0.5 0.6 0.6 0.6 0.7 0.8 0.8 Payment 2.5 1.8 2.2 1.8 1.7 1.9 1.5 1.2 1.4 1.2 Net current transfers 6.0 8.3 5.7 5.1 3.9 4.3 4.1 4.3 6.5 5.3 Receipt 6.6 8.8 6.6 5.4 5.2 5.1 4.8 4.9 6.9 5.7 Payment 0.6 0.5 0.9 0.3 1.3 0.8 0.7 0.7 0.4 0.5 Gross national savings 4.9 12.5 11.0 8.0 7.7 12.8 13.4 17.9 19.2 16.9 Source: Table B.6. Note: Years are calendar years. 301 Table B.8. Gross Domestic Product Deflators by Sector, 1995­2005 1992 = 100 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005a Sector Agriculture, forestry, fishing, and hunting 193.2 234.0 276.0 310.2 349.7 379.3 405.4 440.4 481.6 538.7 590.5 Mining and quarrying 187.5 187.1 222.1 242.3 256.2 260.9 278.2 307.3 358.4 410.4 469.3 Manufacturing 187.8 227.3 251.3 301.8 330.5 362.6 390.4 408.8 419.1 429.6 443.8 Electricity and water 279.7 274.4 304.3 316.0 376.9 396.3 425.9 482.6 495.3 536.2 581.6 Construction 214.6 241.0 317.0 391.7 431.6 447.0 485.3 507.1 530.9 559.5 581.7 Trade, hotels, and restaurants 198.1 226.3 245.7 264.9 291.3 304.2 321.0 336.0 350.5 371.9 394.2 Transport and communication 225.6 270.9 292.1 313.6 348.5 366.7 380.0 400.0 424.6 452.7 484.6 Finance, insurance, and real estate 254.9 325.0 381.1 462.6 496.3 534.3 604.7 665.3 722.7 762.8 841.8 Public administration and other services 232.7 269.9 397.4 478.2 531.5 560.5 608.4 655.2 674.0 705.8 742.6 Imputed bank services 173.9 233.8 191.1 179.5 178.2 183.8 186.9 194.5 216.2 218.6 220.0 GDP at factor cost 207.9 246.3 296.0 340.7 379.0 405.4 435.9 468.4 500.2 541.0 583.9 302 Indirect taxes less subsidies 207.9 246.3 324.0 340.4 387.6 443.1 454.5 425.1 404.7 401.1 373.5 GDP at market prices 207.9 246.3 298.3 340.6 379.6 408.1 437.3 464.8 490.9 523.8 558.6 Memo items Agriculture 193.2 234.0 276.0 310.2 349.7 379.3 405.4 440.4 481.6 538.7 590.5 Industry 204.7 232.3 271.3 319.5 353.0 375.3 404.0 427.9 447.6 472.3 499.0 Services 230.8 270.6 335.6 393.5 433.0 456.1 493.7 527.4 553.3 581.9 623.7 Sources: Tables B.1 and B.4. Note: Years are calendar years. a. Preliminary estimate. Table B.9. Annual Changes in Gross Domestic Product Deflators by Sector, 1995­2005 percentage 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005a Sector Agriculture, forestry, fishing, and hunting 30.3 21.1 17.9 12.4 12.7 8.5 6.9 8.6 9.4 11.9 9.6 Mining and quarrying 20.4 ­0.2 18.7 9.1 5.7 1.8 6.6 10.4 16.7 14.5 14.3 Manufacturing 25.3 21.0 10.6 20.1 9.5 9.7 7.7 4.7 2.5 2.5 3.3 Electricity and water 44.7 ­1.9 10.9 3.8 19.3 5.1 7.5 13.3 2.6 8.3 8.5 Construction 29.9 12.3 31.5 23.6 10.2 3.6 8.6 4.5 4.7 5.4 4.0 Trade, hotels, and restaurants 26.5 14.2 8.6 7.8 10.0 4.4 5.5 4.7 4.3 6.1 6.0 Transport and communication 14.6 20.1 7.8 7.3 11.2 5.2 3.6 5.3 6.1 6.6 7.1 Finance, insurance, and real estate 13.2 27.5 17.2 21.4 7.3 7.7 13.2 10.0 8.6 5.6 10.3 Public administration and other services 28.2 16.0 47.3 20.3 11.2 5.4 8.6 7.7 2.9 4.7 5.2 Imputed bank services 1.6 34.4 ­18.2 ­6.1 ­0.7 3.1 1.7 4.1 11.1 1.1 0.6 GDP at factor cost 27.1 18.5 20.2 15.1 11.2 7.0 7.5 7.5 6.8 8.2 7.9 303 Indirect taxes less subsidies 27.1 18.5 31.5 5.1 13.9 14.3 2.6 ­6.5 ­4.8 ­0.9 ­6.9 GDP at market prices 27.1 18.5 21.1 14.2 11.4 7.5 7.2 6.3 5.6 6.7 6.6 Memo items Agriculture 30.3 21.1 17.9 12.4 12.7 8.5 6.9 8.6 9.4 11.9 9.6 Industry 28.5 13.5 16.8 17.8 10.5 6.3 7.7 5.9 4.6 5.5 5.7 Services 23.1 17.3 24.0 17.3 10.0 5.3 8.2 6.8 4.9 5.2 7.2 Sources: Tables B.1 and B.4. Note: Years are calendar years. a. Preliminary estimate. Table B.10. Investment at Current and at 1992 Constant Prices, 1995­2005 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005a Current prices (T Sh million) 304 Gross fixed capital formation 594,306 620,868 691,314 892,701 989,338 1,266,678 1,390,641 1,789,895 2,239,281 2,570,704 3,118,265 Increase in stocks 5,856 6,640 8,403 9,914 10,312 14,368 15,661 17,854 19,738 28,270 31,137 Gross capital formation 600,162 627,508 699,717 902,615 999,650 1,281,046 1,406,302 1,807,749 2,259,019 2,598,974 3,149,402 As a share of GDP at market prices (%) Gross fixed capital formation 19.7 16.5 14.7 16.0 15.4 17.4 16.8 19.0 21.0 20.9 21.9 Increase in stocks 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 Gross capital formation 19.9 16.7 14.9 16.2 15.5 17.6 17.0 19.2 21.1 21.1 22.2 1992 constant prices(T Sh million) Gross fixed capital Formation 281,766 268,419 259,155 314,264 309,328 332,588 377,556 449,475 566,035 617,512 726,194 Increase in stocks 3,764 3,794 3,825 3,855 3,886 4,003 4,123 4,739 5,240 6,860 8,986 Gross capital formation 285,530 272,213 262,980 318,119 313,214 336,591 381,679 454,214 571,275 624,372 735,180 Real growth (%) Gross fixed capital Formation ­14.9 ­4.7 ­3.5 21.3 ­1.6 7.5 13.5 19.0 25.9 9.1 17.6 Increase in stocks 0.8 0.8 0.8 0.8 0.8 3.0 3.0 14.9 10.6 30.9 31.0 Gross capital formation ­14.7 ­4.7 ­3.4 21.0 ­1.5 7.5 13.4 19.0 25.8 9.3 17.7 Investment deflator (1992 = 100) Gross fixed capital Formation 210.9 231.3 266.8 284.1 319.8 380.9 368.3 398.2 395.6 416.3 429.4 Increase in stocks 155.6 175.0 219.7 257.2 265.4 358.9 379.8 376.7 376.7 412.1 346.5 Gross capital formation 210.2 230.5 266.1 283.7 319.2 380.6 368.5 398.0 395.4 416.3 428.4 Sources: United Republic of Tanzania, National Accounts of Tanzania; The Economic Survey (various issues). Note: Years are calendar years. a. Preliminary estimate. 305 Table B.11. Fixed Capital Formation by Economic Activities at Current Prices, 1995­2004 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Sector (T Sh million) Agriculture, forestry, fishing, and hunting 22,511 20,609 19,545 20,824 27,838 35,642 37,927 47,281 59,152 60,926 306 Mining and quarrying 2,102 2,502 3,085 18,715 92,795 118,808 139,064 189,917 237,599 252,302 Manufacturing 139,134 140,116 146,171 175,662 182,815 234,063 227,560 291,856 365,132 407,691 Electricity and water 84,660 93,010 105,850 129,652 134,932 172,757 164,349 201,190 251,702 268,800 Construction 75,828 83,306 11,608 261,213 278,384 356,423 429,835 566,597 708,851 887,041 Trade, hotels, and restaurants 5,128 4,704 4,461 4,511 4,694 6,010 12,642 16,064 20,097 20,298 Transport and communication 206,795 225,295 254,260 238,718 222,707 285,138 290,770 367,403 459,645 531,846 Finance, insurance, and real estate 23,022 21,077 19,988 18,767 19,531 25,006 25,284 31,473 39,375 40,163 Public administration and other services 32,744 29,978 28,430 24,639 25,642 32,830 63,211 78,115 97,727 101,636 Gross fixed capital formation (T Sh million) 591,924 620,597 593,398 892,701 989,338 1,266,677 1,390,642 1,789,896 2,239,281 2,570,704 Memo items (T Sh million) Agriculture 22,511 20,609 19,545 20,824 27,838 35,642 37,927 47,281 59,152 60,926 Industry 301,724 318,934 266,714 585,242 688,926 882,051 960,808 1,249,560 1,563,284 1,815,834 Services 267,689 281,054 307,139 286,635 272,574 348,984 391,907 493,055 616,844 693,943 Sector (%) Agriculture, forestry, fishing, and hunting 3.8 3.3 3.3 2.3 2.8 2.8 2.7 2.6 2.6 2.4 Mining and quarrying 0.4 0.4 0.5 2.1 9.4 9.4 10.0 10.6 10.6 9.8 Manufacturing 23.5 22.6 24.6 19.7 18.5 18.5 16.4 16.3 16.3 15.9 Electricity and water 14.3 15.0 17.8 14.5 13.6 13.6 11.8 11.2 11.2 10.5 Construction 12.8 13.4 2.0 29.3 28.1 28.1 30.9 31.7 31.7 34.5 Trade, hotels, and restaurants 0.9 0.8 0.8 0.5 0.5 0.5 0.9 0.9 0.9 0.8 Transport and communication 34.9 36.3 42.8 26.7 22.5 22.5 20.9 20.5 20.5 20.7 Finance, insurance, and real estate 3.9 3.4 3.4 2.1 2.0 2.0 1.8 1.8 1.8 1.6 Public administration and other services 5.5 4.8 4.8 2.8 2.6 2.6 4.5 4.4 4.4 4.0 Gross fixed capital formation (%) 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Memo items (%) 307 Agriculture 3.8 3.3 3.3 2.3 2.8 2.8 2.7 2.6 2.6 2.4 Industry 51.0 51.4 44.9 65.6 69.6 69.6 69.1 69.8 69.8 70.6 Services 45.2 45.3 51.8 32.1 27.6 27.6 28.2 27.5 27.5 27.0 Sources: United Republic of Tanzania, National Accounts of Tanzania; The Economic Survey (various issues). Note: Years are calendar years. Table B.12. Fixed Capital Formation by Types of Assets at Current Prices, 1995­2004 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004a Buildings (T Sh million) Residential 56,105 65,560 68,289 76,111 88,923 113,043 122,739 142,063 147,044 181,725 Rural own-account 98,138 122,325 35,400 148,916 156,546 229,713 243,348 275,983 295,301 318,925 Nonresidential 57,733 71,654 78,859 68,858 79,491 180,904 193,117 245,968 431,396 480,255 Other works (T Sh million) Land improvement 12,341 18,388 29,562 4,997 5,571 4,589 6,364 29,993 11,306 27,339 Roads and bridges 13,575 7,662 7,898 62,458 70,391 86,173 129,375 258,749 337,877 360,533 Water supply 2,076 1,172 4,043 33,826 37,882 34,772 38,947 90,601 106,684 118,808 Others 25,806 39,000 46,870 23,159 25,935 40,424 44,573 49,398 23,467 45,053 Equipment (T Sh million) Transport equipment 114,788 111,290 149,001 151,970 209,729 178,390 166,301 211,002 273,493 274,376 Other equipment and machinery 211,362 183,546 173,476 322,406 314,870 398,669 445,878 486,139 612,713 763,689 Gross fixed capital formation (T Sh million) 591,924 620,597 593,398 892,701 989,338 1,266,677 1,390,642 1,789,896 2,239,281 2,570,703 Buildings (%) 308 Residential 9.5 10.6 11.5 8.5 9.0 8.9 8.8 7.9 6.6 7.1 Rural own-account 16.6 19.7 6.0 16.7 15.8 18.1 17.5 15.4 13.2 12.4 Nonresidential 9.8 11.5 13.3 7.7 8.0 14.3 13.9 13.7 19.3 18.7 Other works (%) Land improvement 2.1 3.0 5.0 0.6 0.6 0.4 0.5 1.7 0.5 1.1 Roads and bridges 2.3 1.2 1.3 7.0 7.1 6.8 9.3 14.5 15.1 14.0 Water supply 0.4 0.2 0.7 3.8 3.8 2.7 2.8 5.1 4.8 4.6 Others 4.4 6.3 7.9 2.6 2.6 3.2 3.2 2.8 1.0 1.8 Equipment (%) Transport equipment 19.4 17.9 25.1 17.0 21.2 14.1 12.0 11.8 12.2 10.7 Other equipment and machinery 35.7 29.6 29.2 36.1 31.8 31.5 32.1 27.2 27.4 29.7 Gross fixed capital formation (%) 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Sources: United Republic of Tanzania, National Accounts of Tanzania; The Economic Survey (various issues). Note: Years are calendar years. a. Preliminary estimate. Table B.13. Incremental Gross Capital to Output Ratio, 1987­2005 Gross capital (T Sh million) Incremental output (T Sh million) Incremental capital output ratio (%) 1-year 5-year GDP at 1-year 3-year 5-year moving moving factor cost moving moving moving Year average average (T Sh million) average average average 1 year 3 years 5 years 1987 328,948 1,071,540 1988 254,116 1,119,017 47,477 5.4 1989 253,145 311,314 1,147,745 28,728 49,232 8.8 5.7 1990 340,256 320,133 1,219,236 71,491 44,706 40,875 4.8 7.3 7.8 1991 380,107 335,469 1,253,134 33,898 42,724 32,398 11.2 8.5 10.4 1992 373,043 351,806 1,275,917 22,783 20,590 30,240 16.4 17.5 11.6 1993 330,795 340,861 1,281,006 5,089 15,270 25,202 65.0 22.7 13.5 1994 334,829 319,282 1,298,943 17,937 23,110 29,715 18.7 13.7 10.7 1995 285,530 297,269 1,345,246 46,303 40,235 34,435 6.2 7.4 8.6 1996 272,213 294,734 1,401,711 56,465 49,716 44,964 4.8 5.5 6.6 1997 262,980 290,411 1,448,090 46,379 53,527 55,670 5.7 5.3 5.2 309 1998 318,119 300,623 1,505,827 57,737 58,527 61,815 5.5 5.1 4.9 1999 313,214 322,517 1,577,292 71,465 68,743 69,529 4.4 4.7 4.6 2000 336,591 360,763 1,654,319 77,027 81,177 81,817 4.4 4.2 4.4 2001 381,679 411,395 1,749,358 95,039 93,294 91,321 4.0 4.2 4.5 2002 454,214 473,626 1,857,175 107,817 102,704 103,445 4.2 4.6 4.6 2003 571,275 553,344 1,962,432 105,257 115,053 116,552 5.4 4.8 4.7 2004 624,372 2,094,516 132,084 126,635 4.7 5.1 2005 735,180 2,237,079 142,563 5.2 Sources: United Republic of Tanzania, National Accounts of Tanzania; The Economic Survey (various issues). Note: At 1992 constant prices. Table B.14. Incremental Fixed Capital to Output Ratio, 1995­2005 Incremental capital output Gross investment (T Sh million) Incremental output (T Sh million) ratio (%) 1-year 3-year 5-year GDP at 1-year 3-year 5-year moving moving moving factor cost moving moving moving Year average average average (T Sh million) average average average 1 year 3 years 5 years 1995 281,766 293,760 293,505 1,345,246 46,303 40,235 34,435 6.1 7.3 8.5 1996 268,419 269,780 290,940 1,401,711 56,465 49,716 44,964 4.8 5.4 6.5 1997 259,155 280,613 286,586 1,448,090 46,379 53,527 55,670 5.6 5.2 5.1 1998 314,264 294,249 296,751 1,505,827 57,737 58,527 61,815 5.4 5.0 4.8 1999 309,328 318,727 318,578 1,577,292 71,465 68,743 69,529 4.3 4.6 4.6 2000 332,588 339,824 356,642 1,654,319 77,027 81,177 81,817 4.3 4.2 4.4 2001 377,556 386,540 406,996 1,749,358 95,039 93,294 91,321 4.0 4.1 4.5 2002 449,475 464,355 468,633 1,857,175 107,817 102,704 103,445 4.2 4.5 4.5 2003 566,035 544,341 547,354 1,962,432 105,257 115,053 116,552 5.4 4.7 4.7 2004 617,512 636,580 2,094,516 132,084 126,635 4.7 5.0 2005 726,194 2,237,079 142,563 5.1 310 Sources: United Republic of Tanzania, National Accounts of Tanzania; The Economic Survey (various issues). Note: At 1992 constant prices. APPENDIX C Exports and Imports Table C.1. Balance of Payments, 1995­2005 Items 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005a Current account (US$ million) ­646.4 ­146.1 ­403.5 ­921.3 ­855.8 ­485.4 ­209.0 22.9 ­103.3 ­486.9 ­1,078.0 Goods ­657.6 ­448.9 ­395.4 ­793.6 ­872.1 ­704.6 ­622.4 ­512.3 ­746.1 ­1,099.7 ­1,494.2 Exports (free-on-board) 682.9 763.8 752.6 588.5 543.3 663.3 746.7 948.6 1,265.2 1,597.6 1,896.6 Imports (free-on-board) 1,340.5 1,212.6 1,148.0 1,382.1 1,415.4 1,367.9 1,369.2 1,460.9 2,011.3 2,697.3 3,390.8 Services ­216.9 ­159.9 ­317.8 ­450.0 ­225.3 ­52.2 234.1 278.1 230.2 174.3 68.7 Receipts 582.9 656.0 482.4 538.8 622.0 643.8 802.8 889.3 984.9 1,232.4 1,434.0 Payments 799.8 815.9 800.2 988.8 847.3 696.0 568.7 611.2 754.7 1,058.0 1,365.3 Income ­110.3 ­72.0 ­121.8 ­105.0 ­95.0 ­120.5 ­165.2 ­144.9 ­168.6 ­194.0 ­229.0 Receipts 31.8 41.5 43.0 44.4 49.0 50.4 48.8 65.6 90.7 88.6 91.6 Payments 142.1 113.5 164.8 149.4 144.0 170.9 213.9 210.5 259.3 282.6 320.6 Current transfers 338.4 534.7 431.5 427.3 336.6 391.9 344.5 402.0 581.2 632.4 576.5 Inflows 370.7 567.0 499.2 454.2 445.6 464.8 415.0 461.3 646.7 703.3 652.7 Government 236.0 432.2 433.6 421.0 411.4 429.9 405.5 434.5 637.9 582.1 575.7 312 Private 134.7 134.8 65.6 33.2 34.2 34.9 9.5 26.8 8.8 121.2 77.0 Outflows 32.3 32.3 67.7 26.9 109.0 72.9 70.5 59.3 65.4 70.9 76.2 Capital account (US$ million) 191.0 191.0 270.9 252.4 270.6 330.4 353.3 372.6 392.4 403.4 415.1 Capital transfers 191.0 191.0 270.9 252.4 270.6 330.4 333.6 343.5 353.6 359.9 366.5 Inflows 191.0 191.0 270.9 252.4 270.6 330.4 333.6 343.5 353.6 359.9 366.5 Outflows 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Financial account (US$ million) 139.5 166.7 413.7 317.8 662.4 400.4 ­310.3 185.2 264.5 413.7 311.4 Direct investment 150.0 148.5 157.8 272.2 516.7 463.4 409.5 415.4 547.1 511.9 534.4 Abroad 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 In Tanzania 150.0 148.5 157.8 272.2 516.7 463.4 409.5 415.4 547.1 511.9 534.4 Portfolio investment 0.0 0.0 20.5 0.7 0.0 0.0 7.2 2.1 2.8 2.6 2.8 Other investment ­10.5 18.2 235.4 44.9 145.7 ­63.0 ­726.9 ­232.4 ­285.4 ­100.8 ­225.8 Inflow of financial resources 437.4 270.2 673.5 645.9 445.8 265.2 ­66.6 3.1 ­61.1 ­12.0 ­74.5 Outflow of financial resources 447.9 252.0 438.1 601.0 300.1 328.2 660.3 235.5 224.3 88.8 151.3 Errors and omissions (US$ million) ­62.2 ­429.0 ­480.8 ­144.4 ­190.5 ­188.5 ­541.5 ­687.9 ­494.3 ­331.0 28.6 Overall balance (US$ million) ­378.1 ­217.4 ­199.7 ­495.5 ­113.3 56.9 ­707.5 ­107.3 59.3 ­0.7 ­322.8 Financing (US$ million) 378.1 217.4 199.7 495.5 113.3 ­56.9 707.5 107.3 ­59.3 0.7 322.8 Net reserve assets (­ increase) 60.1 ­165.3 ­182.1 20.9 ­175.4 ­197.4 ­161.7 ­361.3 ­531.1 ­266.4 276.3 Use of fund credit ­4.0 ­13.8 77.4 11.0 51.3 49.4 13.7 25.1 ­3.0 ­7.9 ­256.7 Exceptional financing 322.0 396.5 304.4 463.6 237.4 91.1 855.5 443.4 474.8 275.0 303.2 Change in arrears 269.4 312.2 77.1 96.9 152.8 81.1 740.3 433.7 384.6 275.0 303.2 Rescheduling 0.0 0.0 227.3 366.7 84.6 10.0 115.2 9.4 90.3 0.0 0.0 Financing gap 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Memorandum items GDP at market prices (T Sh billion) 3,020.5 3,767.6 4,703.5 5,581.6 6,432.9 7,268.4 8,304.3 9,399.1 10,707.1 12,351.7 13,742.6 GDP at market prices (US$ million) 5,631.1 6,462.5 7,607.1 8,379.6 8,634.8 9,080.9 9,475.6 9,724.0 10,309.3 11,338.7 12,174.1 Current account bal- 313 ance/GDP ­11.5 ­2.3 ­5.3 ­11.0 ­9.9 ­5.3 ­2.5 0.2 ­1.0 ­3.9 ­7.8 Current account bal- ance/GDP (excluding current official transfers) ­17.5 ­10.5 ­11.0 ­16.1 ­13.8 ­9.7 ­6.9 ­4.1 ­6.4 ­9.0 ­11.8 Gross official reserves (US$ million) 255.0 240.0 461.0 502.0 605.0 752.0 697.5 1,033.3 1,631.5 1,863.7 1,918.1 Months of imports of goods and nonfactor services 1.6 2.8 3.2 3.2 4.5 5.2 4.3 6.0 7.1 6.0 4.8 Source: Tanzania authorities. a. Preliminary estimate. 314 Table C.2. Merchandise Exports: Value, 1995­2005 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005a Value of exports (US$ million) Coffee 142.6 136.1 117.4 108.7 76.6 83.7 57.1 46.2 48.9 41.4 71.7 Cotton 120.2 125.3 130.4 47.6 28.5 38.0 33.7 31.2 42.3 38.2 93.6 Sisal 6.3 5.3 9.1 6.8 7.3 5.6 6.7 6.7 6.5 7.6 7.2 Tea 23.4 22.5 31.8 30.4 24.6 32.7 29.0 29.3 29.6 27.3 27.9 Tobacco 27.1 49.2 53.6 55.4 43.4 38.4 35.7 45.6 42.5 47.9 58.9 Cashew nuts 64.0 97.8 91.1 107.3 100.9 84.4 56.6 51.6 44.1 48.6 61.6 Cloves .. .. .. .. 19.8 10.0 12.3 8.2 9.8 9.5 6.5 Subtotal 383.6 436.2 433.4 356.2 301.1 292.8 231.1 218.8 223.7 220.5 327.4 Minerals 44.9 55.9 51.1 26.3 73.3 178.2 302.3 343.0 432.0 625.0 714.0 Of which gold 45.0 56.0 1.0 3.0 35.0 113.0 254.0 298.0 383.0 583.0 656.0 Manufactured goods 109.3 127.1 111.4 35.7 30.1 43.5 56.2 61.0 75.8 94.2 129.7 Other values 145.3 148.8 147.6 170.2 138.7 148.9 187.0 216.9 278.2 358.8 422.9 Of which petroleum 11.0 11.1 12.4 7.4 .. .. .. .. .. .. .. Subtotal 299.5 331.8 310.1 232.2 242.1 370.6 545.5 620.9 786.0 1,078.0 1,266.6 Total 683.1 768.0 743.5 588.4 543.2 663.4 776.6 839.7 1,009.7 1,298.5 1,594.0 Share of total exports (%) Coffee 20.9 17.7 15.8 18.5 14.1 12.6 7.4 5.5 4.8 3.2 4.5 Cotton 17.6 16.3 17.5 8.1 5.2 5.7 4.3 3.7 4.2 2.9 5.9 Sisal 0.9 0.7 1.2 1.2 1.3 0.8 0.9 0.8 0.6 0.6 0.5 Tea 3.4 2.9 4.3 5.2 4.5 4.9 3.7 3.5 2.9 2.1 1.8 Tobacco 4.0 6.4 7.2 9.4 8.0 5.8 4.6 5.4 4.2 3.7 3.7 Cashew nuts 9.4 12.7 12.3 18.2 18.6 12.7 7.3 6.1 4.4 3.7 3.9 Cloves .. .. .. .. 3.6 1.5 1.6 1.0 1.0 0.7 0.4 Subtotal 56.2 56.8 58.3 60.5 55.4 44.1 29.8 26.1 22.2 17.0 20.5 Minerals 6.6 7.3 6.9 4.5 13.5 26.9 38.9 40.8 42.8 48.1 44.8 Of which gold 6.6 7.3 0.1 0.5 6.4 17.0 32.7 35.5 37.9 44.9 41.2 Manufactured goods 16.0 16.5 15.0 6.1 5.5 6.6 7.2 7.3 7.5 7.3 8.1 Other values 21.3 19.4 19.9 28.9 25.5 22.4 24.1 25.8 27.6 27.6 26.5 Of which petroleum 1.6 1.4 1.7 1.3 .. .. .. .. .. .. .. Subtotal 43.8 43.2 41.7 39.5 44.6 55.9 70.2 73.9 77.8 83.0 79.5 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: Tanzania authorities. Note: Data may differ from table C.1 because of different sources. 315 a. Preliminary estimate. Table C.3. Volume and Unit Price of Selected Merchandise Exports, 1995­2005 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005a Volume (thousand metric tons) Coffee 48.0 61.7 46.6 44.7 39.0 54.5 48.4 42.4 46.8 35.1 48.6 Cotton 70.7 83.5 86.9 37.3 26.3 36.9 36.8 35.1 49.2 40.4 91.8 Sisal 11.3 8.9 15.1 11.6 15.5 13.2 13.9 13.3 12.6 14.0 9.6 Tea 21.6 22.0 21.6 22.1 21.4 22.6 23.0 23.7 24.6 22.2 23.6 Tobacco 17.1 24.9 27.9 26.3 21.4 19.2 18.7 22.1 20.2 24.9 26.9 Cashew nuts 75.6 118.4 121.9 164.7 102.0 102.4 95.0 85.4 74.4 76.6 71.5 Cloves .. .. .. .. 6.2 2.9 2.5 1.9 3.8 5.6 2.1 Unit price (US$ per kilogram). Coffee 2.97 2.21 2.52 2.43 1.96 1.54 1.18 1.09 1.04 1.18 1.48 Cotton 1.70 1.50 1.50 1.28 1.08 1.03 0.92 0.89 0.86 0.94 1.02 Sisal 0.56 0.60 0.60 0.58 0.47 0.42 0.48 0.50 0.51 0.55 0.75 Tea 1.08 1.02 1.47 1.38 1.15 1.45 1.26 1.24 1.20 1.23 1.18 316 Tobacco 1.58 1.98 1.92 2.10 2.03 2.00 1.91 2.07 2.11 1.92 2.19 Cashew nuts 0.85 0.83 0.75 0.65 0.99 0.82 0.60 0.60 0.59 0.63 0.86 Cloves .. .. .. .. 3.20 3.44 5.03 4.38 2.55 1.70 3.07 Source: Tanzania authorities. a. Preliminary estimate. Table C.4. Indices of Value, Volume, and Unit Price of Selected Exports, 1995­2005 1992 = 100 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005a Value Coffee 239.7 228.7 197.3 182.7 128.7 140.7 96.0 77.6 82.2 69.6 120.5 Cotton 123.2 128.4 133.6 48.8 29.2 38.9 34.5 32.0 43.3 39.1 95.9 Sisal 700.0 588.9 1,011.1 755.6 811.1 622.2 744.4 744.4 722.2 844.4 800.0 Tea 97.1 93.4 132.0 126.1 102.1 135.7 120.3 121.6 122.8 113.3 115.8 Tobacco 99.6 180.9 197.1 203.7 159.6 141.2 131.3 167.6 156.3 176.1 216.5 Cashew nuts 272.3 416.2 387.7 456.6 429.4 359.1 240.9 219.6 187.7 206.8 262.1 Cloves .. .. .. .. 550.0 277.8 341.7 227.8 272.2 263.9 180.6 Subtotal 162.3 184.5 183.3 150.7 127.4 123.9 97.8 92.6 94.6 93.3 138.5 Minerals 109.5 136.3 124.6 64.1 178.8 434.6 737.3 836.6 1,053.7 1,524.4 1,741.5 Of which gold 109.8 136.6 2.4 7.3 85.4 275.6 619.5 726.8 934.1 1,422.0 1,600.0 Manufactured goods 179.8 209.0 183.2 58.7 49.5 71.5 92.4 100.3 124.7 154.9 213.3 Other values 240.6 246.4 244.4 281.8 229.6 246.5 309.6 359.1 460.6 594.0 700.2 Of which petroleum 137.5 138.8 155.0 92.5 .. .. .. .. .. .. .. 317 Subtotal 183.1 202.8 189.5 141.9 148.0 226.5 333.4 379.5 480.4 658.9 774.2 Total 170.8 192.0 185.9 147.1 135.8 165.9 194.2 209.9 252.4 324.6 398.5 Volume Coffee 94.1 121.0 91.4 87.6 76.5 106.9 94.9 83.1 91.8 68.8 95.3 Cotton 99.2 117.1 121.9 52.3 36.9 51.8 51.6 49.2 69.0 56.7 128.8 Sisal 240.4 189.4 321.3 246.8 329.8 280.9 295.7 283.0 268.1 297.9 204.3 Tea 105.9 107.8 105.9 108.3 104.9 110.8 112.7 116.2 120.6 108.8 115.7 Tobacco 133.6 194.5 218.0 205.5 167.2 150.0 146.1 172.7 157.8 194.5 210.2 Cashew nuts 258.0 404.1 416.0 562.1 348.1 349.5 324.2 291.5 253.9 261.4 244.0 Cloves .. .. .. .. 516.7 241.7 208.3 158.3 316.7 466.7 175.0 continued Table C.4 continued Unit price Coffee 253.8 188.5 215.3 208.0 167.8 131.4 100.8 93.0 89.1 100.8 126.1 Cotton 124.1 109.5 109.5 93.2 78.9 75.3 66.8 64.8 62.6 68.9 74.4 Sisal 293.2 313.2 316.8 307.8 246.8 223.2 253.7 263.2 269.5 287.4 394.2 Tea 91.8 86.6 124.7 116.5 97.5 122.8 106.8 104.9 101.9 104.0 100.0 Tobacco 74.4 92.7 90.2 98.7 95.5 93.8 89.6 97.1 99.0 90.2 102.7 Cashew nuts 105.8 103.3 93.4 81.4 123.5 103.0 74.4 75.5 74.1 79.3 107.6 Cloves .. .. .. .. 106.7 114.8 167.6 146.0 85.0 56.5 102.3 Source: Tanzania authorities. a. Preliminary estimate. 318 Table C.5. Merchandise Imports: Value, 1995­2005 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005a Value (US$ million) Capital goods 554.2 501.0 513.6 734.3 693.2 638.2 739.8 730.4 854.7 975.0 975.0 Transport equipment 209.7 202.7 253.1 240.2 231.8 228.5 189.8 204.0 245.3 246.3 260.4 Building and construction 49.2 42.5 35.1 120.4 121.6 128.8 144.0 139.3 159.4 175.0 235.8 Machinery 295.3 255.8 225.4 373.7 339.8 280.9 406.0 387.1 346.5 444.0 573.3 Intermediate goods 609.0 531.0 382.9 300.2 319.6 319.4 440.6 432.0 643.6 798.1 798.1 Oil 193.8 158.4 173.0 97.1 148.1 142.6 220.7 208.0 261.0 459.0 794.0 Crude 115.2 69.9 93.5 76.2 .. .. .. .. .. .. .. Products 78.6 88.5 93.5 21.0 .. .. .. .. .. .. .. Fertilizers 11.8 23.3 22.6 11.7 10.8 16.8 15.5 17.8 21.6 41.6 65.3 Industrial raw materials 403.4 349.3 187.3 191.4 160.7 159.9 204.4 206.2 233.3 255.0 247.3 Consumer goods 377.7 361.8 373.0 534.6 559.9 576.8 531.8 523.0 605.8 651.7 651.7 Food 44.2 52.7 97.0 215.5 230.7 183.0 169.4 158.3 146.2 257.0 203.7 Other 333.5 309.1 276.0 319.1 329.2 393.8 362.4 364.7 406.0 489.0 613.3 Unclassified imports 0.0 0.0 0.0 0.0 0.0 2.3 2.4 2.4 2.4 2.2 2.2 319 Total 1,540.9 1,393.8 1,269.5 1,569.0 1,572.7 1,536.7 1,714.6 1,687.8 2,106.6 2,427.0 2,427.0 Share of total imports (%) Capital goods 36.0 35.9 40.5 46.8 44.1 41.5 43.1 43.3 40.6 40.2 40.2 Transport equipment 13.6 14.5 19.9 15.3 14.7 14.9 11.1 12.1 11.6 10.1 10.7 Building and construction 3.2 3.0 2.8 7.7 7.7 8.4 8.4 8.3 7.6 7.2 9.7 Machinery 19.2 18.4 17.8 23.8 21.6 18.3 23.7 22.9 16.4 18.3 23.6 Intermediate goods 39.5 38.1 30.2 19.1 20.3 20.8 25.7 25.6 30.6 32.9 32.9 Oil 12.6 11.4 13.6 6.2 9.4 9.3 12.9 12.3 12.4 18.9 32.7 Crude 7.5 5.0 7.4 4.9 .. .. .. .. .. .. .. Products 5.1 6.3 7.4 1.3 .. .. .. .. .. .. .. Fertilizers 0.8 1.7 1.8 0.7 0.7 1.1 0.9 1.1 1.0 1.7 2.7 Industrial raw materials 26.2 25.1 14.8 12.2 10.2 10.4 11.9 12.2 11.1 10.5 10.2 Consumer goods 24.5 26.0 29.4 34.1 35.6 37.5 31.0 31.0 28.8 26.9 26.9 Food 2.9 3.8 7.6 13.7 14.7 11.9 9.9 9.4 6.9 10.6 8.4 Other 21.6 22.2 21.7 20.3 20.9 25.6 21.1 21.6 19.3 20.1 25.3 Unclassified imports 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.1 0.1 0.1 0.1 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: Tanzania authorities. a. Preliminary estimate. Table C.6. Indices of Terms of Trade, 1995­2005 1992 = 100 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005a Value Exports 172.0 193.5 187.3 148.2 136.8 167.1 195.6 211.5 254.4 327.2 401.5 Imports 112.8 102.1 93.0 114.9 115.2 112.4 125.6 123.4 133.4 173.5 219.3 Unit value Exports 119.8 112.7 122.6 113.4 128.5 123.1 114.7 120.2 122.4 133.6 139.2 Imports 114.4 114.1 104.7 96.7 87.8 88.2 104.4 91.6 98.8 117.4 136.2 Volume Exports 143.6 171.6 152.7 130.7 106.5 135.7 170.5 176.0 207.9 245.0 288.3 Imports 98.7 89.5 88.8 118.8 131.2 127.3 120.3 134.7 134.9 147.7 161.1 Terms of trade 104.8 98.8 117.1 117.3 146.5 139.5 109.9 131.2 123.8 113.7 102.3 Source: Tanzania authorities. a. Preliminary estimate. 320 Table C.7. Direction of Trade: Value, 1995­2005 millions of U.S. dollars 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005a Exports to East African Community 32.0 23.0 32.6 35.1 39.4 63.1 44.1 40.8 26.2 36.0 41.1 Kenya 23.6 13.0 25.2 28.5 28.4 38.0 38.4 35.3 16.4 23.1 26.4 Uganda 8.4 9.9 7.4 6.6 11.0 25.0 5.6 5.5 9.8 12.9 14.7 British Commonwealth countries Canada 2.6 1.0 0.9 0.6 1.0 0.8 0.5 1.4 2.1 11.0 131.7 India 57.8 88.1 73.9 116.8 113.2 98.8 82.2 64.2 95.3 112.3 111.3 Ireland 1.0 1.3 1.4 20.2 14.4 23.1 27.9 14.1 0.1 0.0 0.1 Singapore 13.9 14.4 7.6 11.9 24.7 6.5 6.0 3.9 7.8 12.0 6.2 United Kingdom 39.4 41.0 52.6 61.0 95.5 148.2 140.4 161.8 50.5 62.2 60.6 European Community 236.8 251.9 236.7 245.7 228.2 370.4 436.8 479.0 302.2 429.0 364.7 Belgium 0.0 0.0 31.5 26.7 17.0 18.7 10.6 21.2 37.4 35.7 26.1 France 11.9 12.0 6.2 4.2 2.8 36.8 138.4 154.5 6.3 14.2 13.9 Germany 66.5 65.0 67.4 49.1 36.3 66.8 38.3 27.6 50.1 48.6 65.5 321 Italy 10.6 8.9 4.9 7.6 7.4 9.4 7.8 24.4 11.6 15.8 29.2 Netherlands 35.7 38.1 32.1 45.6 32.8 46.2 51.7 53.9 78.0 80.7 78.8 Organization of Petroleum Ex- porting Countries 49.2 34.0 40.5 14.9 22.1 15.9 18.4 35.3 65.9 62.2 74.0 Saudi Arabia 7.0 3.1 1.0 1.3 4.4 3.8 4.9 15.4 17.2 2.0 3.3 United Arab Emirates 22.7 3.6 5.8 5.6 7.0 6.8 9.6 14.4 32.5 42.2 53.7 Other economies China 8.7 17.3 3.5 1.9 1.9 0.7 0.7 0.7 25.1 59.8 155.4 Hong Kong, China 15.5 18.0 9.0 5.0 7.9 8.8 8.7 11.2 16.9 18.8 17.5 Japan 58.8 59.0 40.6 44.1 44.4 35.4 68.5 96.3 89.8 71.6 68.0 South Africa 4.3 5.0 8.4 6.7 16.1 15.8 8.9 16.5 17.7 31.9 38.7 Sweden 1.6 2.8 0.4 0.6 2.4 0.7 0.1 0.2 1.7 1.4 4.4 United States 22.1 18.5 11.6 13.5 19.9 15.6 15.2 13.5 23.7 23.5 32.2 Total exports 697.1 763.5 619.4 588.4 621.5 735.1 774.4 948.6 949.6 1,228.0 1,480.4 continued Table C.7 continued Imports from East African Community 239.5 265.0 94.7 107.4 102.2 98.7 107.4 97.9 217.7 257.4 294.1 Kenya 237.7 262.9 89.3 105.1 96.1 93.1 96.1 95.2 211.3 249.0 284.5 Uganda 1.8 2.1 5.4 2.2 6.0 5.6 11.4 2.7 6.4 8.4 9.6 British Commonwealth countries Canada 10.0 13.8 14.8 16.6 26.7 32.3 23.1 17.7 21.0 25.7 29.1 India 77.8 77.9 171.7 89.0 94.6 88.8 87.5 106.8 176.8 187.2 245.8 Ireland 7.8 6.2 3.5 4.0 2.2 4.5 5.2 8.3 8.0 10.0 9.6 Singapore 48.3 31.1 44.9 12.8 9.8 6.3 9.0 5.8 12.4 13.9 14.1 United Kingdom 159.9 140.5 132.3 121.5 128.3 106.5 110.6 94.9 101.6 135.6 142.6 European Community 457.3 432.4 404.1 482.5 402.5 340.0 387.1 369.6 505.2 612.2 728.5 Belgium 0.0 0.0 25.8 24.2 23.0 14.5 22.0 23.2 31.5 42.1 48.0 France 17.8 25.8 23.0 15.7 21.5 25.6 41.6 38.9 63.8 79.6 86.1 Germany 58.7 49.6 52.9 77.2 62.5 52.0 68.1 60.0 91.4 103.2 122.5 Italy 61.9 70.8 43.4 67.5 48.9 35.6 51.0 45.0 55.5 60.2 78.7 Netherlands 35.7 35.4 35.6 89.8 36.7 29.5 30.9 27.4 47.4 67.5 92.3 Organization of Petroleum Ex- 322 porting Countries 265.2 297.2 140.0 176.3 191.3 161.3 228.3 219.1 282.2 353.2 453.1 Saudi Arabia 84.9 99.2 12.6 32.2 36.2 54.5 58.2 47.3 57.3 75.8 103.2 United Arab Emirates 57.2 62.9 110.3 59.1 51.3 56.8 109.0 97.1 117.7 155.7 211.8 Other economies China 81.5 70.6 27.4 48.3 58.0 68.0 70.5 79.0 211.0 229.5 334.2 Hong Kong, China 38.9 28.0 6.6 3.0 3.2 4.3 7.1 7.8 13.3 18.6 23.0 Japan 118.6 85.0 97.0 129.6 178.1 142.1 150.7 138.7 85.2 86.6 104.5 South Africa 190.3 141.8 105.5 129.4 170.8 174.4 203.4 188.8 272.2 373.1 460.6 Sweden 29.7 21.4 19.2 29.1 17.9 17.2 12.5 22.0 39.8 29.1 81.3 United States 72.8 55.2 66.2 80.3 99.3 58.9 65.3 91.4 72.7 140.5 106.2 Total imports 1,879.2 1,787.3 1,408.8 1,571.3 1,660.3 1,520.5 1,780.2 1,749.5 2,496.6 3,074.3 3,617.3 Source: International Monetary Fund. Note: Figures for exports and imports refer to merchandise goods. a. Preliminary estimate. Table C.8. Direction of Trade: Share of Total, 1995­2005 percentage 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Exports to East African Community 4.6 3.0 5.3 6.0 6.3 8.6 5.7 4.3 2.8 2.9 2.8 Kenya 3.4 1.7 4.1 4.8 4.6 5.2 5.0 3.7 1.7 1.9 1.8 Uganda 1.2 1.3 1.2 1.1 1.8 3.4 0.7 0.6 1.0 1.0 1.0 British Commonwealth countries Canada 0.4 0.1 0.1 0.1 0.2 0.1 0.1 0.1 0.2 0.9 8.9 India 8.3 11.5 11.9 19.8 18.2 13.4 10.6 6.8 10.0 9.1 7.5 Ireland 0.1 0.2 0.2 3.4 2.3 3.1 3.6 1.5 0.0 0.0 0.0 Singapore 2.0 1.9 1.2 2.0 4.0 0.9 0.8 0.4 0.8 1.0 0.4 United Kingdom 5.7 5.4 8.5 10.4 15.4 20.2 18.1 17.1 5.3 5.1 4.1 European Community 34.0 33.0 38.2 41.8 36.7 50.4 56.4 50.5 31.8 34.9 24.6 Belgium 0.0 0.0 5.1 4.5 2.7 2.5 1.4 2.2 3.9 2.9 1.8 France 1.7 1.6 1.0 0.7 0.4 5.0 17.9 16.3 0.7 1.2 0.9 Germany 9.5 8.5 10.9 8.3 5.8 9.1 4.9 2.9 5.3 4.0 4.4 323 Italy 1.5 1.2 0.8 1.3 1.2 1.3 1.0 2.6 1.2 1.3 2.0 Netherlands 5.1 5.0 5.2 7.8 5.3 6.3 6.7 5.7 8.2 6.6 5.3 Organization of Petroleum Exporting Countries 7.1 4.5 6.5 2.5 3.6 2.2 2.4 3.7 6.9 5.1 5.0 Saudi Arabia 1.0 0.4 0.2 0.2 0.7 0.5 0.6 1.6 1.8 0.2 0.2 United Arab Emirates 3.3 0.5 0.9 0.9 1.1 0.9 1.2 1.5 3.4 3.4 3.6 Other economies China 1.3 2.3 0.6 0.3 0.3 0.1 0.1 0.1 2.6 4.9 10.5 Hong Kong, China 2.2 2.4 1.4 0.9 1.3 1.2 1.1 1.2 1.8 1.5 1.2 Japan 8.4 7.7 6.6 7.5 7.1 4.8 8.9 10.2 9.5 5.8 4.6 South Africa 0.6 0.7 1.4 1.1 2.6 2.2 1.2 1.7 1.9 2.6 2.6 Sweden 0.2 0.4 0.1 0.1 0.4 0.1 0.0 0.0 0.2 0.1 0.3 United States 3.2 2.4 1.9 2.3 3.2 2.1 2.0 1.4 2.5 1.9 2.2 Total exports 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 continued Table C.8 continued Imports from East African Community 12.7 14.8 6.7 6.8 6.2 6.5 6.0 5.6 8.7 8.4 8.1 Kenya 12.6 14.7 6.3 6.7 5.8 6.1 5.4 5.4 8.5 8.1 7.9 Uganda 0.1 0.1 0.4 0.1 0.4 0.4 0.6 0.2 0.3 0.3 0.3 British Commonwealth countries Canada 0.5 0.8 1.1 1.1 1.6 2.1 1.3 1.0 0.8 0.8 0.8 India 4.1 4.4 12.2 5.7 5.7 5.8 4.9 6.1 7.1 6.1 6.8 Ireland 0.4 0.3 0.3 0.3 0.1 0.3 0.3 0.5 0.3 0.3 0.3 Singapore 2.6 1.7 3.2 0.8 0.6 0.4 0.5 0.3 0.5 0.5 0.4 United Kingdom 8.5 7.9 9.4 7.7 7.7 7.0 6.2 5.4 4.1 4.4 3.9 European Community 24.3 24.2 28.7 30.7 24.2 22.4 21.7 21.1 20.2 19.9 20.1 Belgium 0.0 0.0 1.8 1.5 1.4 1.0 1.2 1.3 1.3 1.4 1.3 France 0.9 1.4 1.6 1.0 1.3 1.7 2.3 2.2 2.6 2.6 2.4 Germany 3.1 2.8 3.8 4.9 3.8 3.4 3.8 3.4 3.7 3.4 3.4 Italy 3.3 4.0 3.1 4.3 2.9 2.3 2.9 2.6 2.2 2.0 2.2 Netherlands 1.9 2.0 2.5 5.7 2.2 1.9 1.7 1.6 1.9 2.2 2.6 Organization of Petroleum Exporting 324 Countries 14.1 16.6 9.9 11.2 11.5 10.6 12.8 12.5 11.3 11.5 12.5 Saudi Arabia 4.5 5.5 0.9 2.1 2.2 3.6 3.3 2.7 2.3 2.5 2.9 United Arab Emirates 3.0 3.5 7.8 3.8 3.1 3.7 6.1 5.5 4.7 5.1 5.9 Other economies China 4.3 3.9 1.9 3.1 3.5 4.5 4.0 4.5 8.5 7.5 9.2 Hong Kong, China 2.1 1.6 0.5 0.2 0.2 0.3 0.4 0.4 0.5 0.6 0.6 Japan 6.3 4.8 6.9 8.2 10.7 9.3 8.5 7.9 3.4 2.8 2.9 South Africa 10.1 7.9 7.5 8.2 10.3 11.5 11.4 10.8 10.9 12.1 12.7 Sweden 1.6 1.2 1.4 1.8 1.1 1.1 0.7 1.3 1.6 0.9 2.2 United States 3.9 3.1 4.7 5.1 6.0 3.9 3.7 5.2 2.9 4.6 2.9 Total imports 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: Table C.7. Table C.9. External Reserves, 1995­2005 millions of U.S. dollars Fiscal year 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005a Reserve level 255.0 240.0 461.0 502.0 605.0 752.0 1,059.8 1,492.7 2,167.4 2,394.8 2,387.4 Bank of Tanzania 60.1 ­165.3 ­182.1 20.9 ­175.4 ­197.4 ­261.6 ­335.8 ­598.2 ­232.2 ­54.4 International Monetary Fund, net 19.0 ­15.0 ­54.0 ­10.0 ­51.0 ­43.0 ­48.6 ­47.3 ­17.7 32.8 59.5 Purchase 0.0 ­37.0 ­85.0 ­48.0 ­80.0 ­53.0 ­51.0 ­52.0 ­25.0 8.3 0.0 Repurchase 19.0 22.0 31.0 38.0 29.0 10.0 2.4 4.7 7.3 24.5 59.5 Reserve level as months of import goods and nonfactor services 1.6 2.8 3.2 3.2 4.5 5.2 4.3 6.0 7.1 6.0 4.8 Sources: Tanzania authorities and World Bank. a. Preliminary estimate. Note: Negative signs show an increase. 325 Table C.10. Scheduled Debt Service Payments, 1995­2005 millions of U.S. dollars Fiscal year 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005a Debt serviceb 228.1 262.9 161.6 240.0 211.0 171.6 152.6 108.5 91.4 119.1 275.5 Interest 82.1 96.9 40.6 104.0 74.0 63.6 47.6 42.8 40.3 52.3 69.0 Amortization 146.0 166.0 121.0 136.0 137.0 108.0 105.0 65.7 51.1 66.8 206.5 Debt service ratio (as a % of exports of goods and nonfactor services) 18.0 18.5 13.1 21.3 18.1 13.1 9.8 5.9 4.1 4.2 8.3 Debt service (as a % of GDP) 4.1 4.1 2.1 2.9 2.4 1.8 1.6 1.1 0.9 1.1 2.3 Sources: Table C.1 and various International Monetary Fund documents. a. Preliminary estimate. b. Includes obligations to the International Monetary Fund. 326 APPENDIX D External Debt Table D.1. Summary of External Debt, 1995­2005 millions of U.S. dollars 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005a Debt outstanding and disbursed 7,421 7,387 6,911 7,309 7,654 6,932 6,245 6,842 6,989 7,800 7,732 Long-term debt 6,261 6,149 5,817 6,124 6,348 5,760 5,300 5,740 5,738 6,237 6,229 Public and publicly guaranteed 6,217 6,104 5,776 6,087 6,316 5,732 5,276 5,720 5,722 6,225 6,220 Private nonguaranteed 44 45 41 37 32 28 24 20 16 12 9 328 Use of International Monetary Fund (IMF) credit 197 206 246 268 312 324 341 400 437 423 363 Short-term debt 963 1,032 848 917 994 848 604 702 814 1,140 1,140 Of which interest arrears on long-term debt 904 904 709 769 836 645 484 541 657 811 811 Disbursements 262 246 336 264 340 261 226 240 501 426 303 Long-term debt 262 209 251 216 260 208 175 188 476 418 303 Public and publicly guaranteed 261 207 251 216 260 208 175 188 476 418 303 Private nonguaranteed 1 2 0 0 0 0 0 0 0 0 0 Use of IMF credit 0 37 85 48 80 53 51 52 25 8 0 Principal paymentsb 146 166 121 134 137 108 104 66 51 67 207 Long-term debt 127 144 90 96 108 98 102 61 44 42 147 Public and publicly guaranteed 122 144 86 92 104 94 98 57 41 38 144 Private nonguaranteed 5 0 4 4 4 4 4 4 3 4 3 IMF repurchases 19 22 31 38 29 10 2 5 7 25 60 Interest paymentsb 86 103 47 102 89 64 48 44 40 52 69 Long-term debt 81 96 40 94 80 52 40 36 33 42 67 Public and publicly guaranteed 80 95 39 93 79 51 39 35 32 42 67 Private nonguaranteed 1 1 1 1 1 1 1 1 1 0 0 IMF charges 1 1 1 1 1 2 2 2 2 2 2 Short-term debt 4 6 6 7 8 10 6 6 5 7 0 Memo items Total debt serviceb 232 269 168 236 226 172 152 110 91 119 276 Long-term debt 208 240 130 190 188 150 142 97 77 84 214 Public and publicly guaranteed 202 239 125 185 183 145 137 92 73 80 211 Private nonguaranteed 6 1 5 5 5 5 5 5 4 4 3 IMF repurchase and charges 20 23 32 39 30 12 4 7 9 27 62 Short-term debt 4 6 6 7 8 10 6 6 5 .. .. Debt outstanding and disbursed as a percentage of GDP 145 116 95 93 95 80 72 75 73 94 95 Source: World Bank. Note: Debt data in this section are on a calendar year basis, and the figures for disbursements and debt service payments may be different from those in the balance of payments. a. Preliminary estimate. b. Cash basis; see table C.10 for scheduled debt service payments. 329 Table D.2. External Public Debt, 1995­2005 millions of U.S. dollars 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005a Debt outstanding and disbursed 6,261 6,104 5,775 6,087 6,317 5,732 5,276 5,719 5,729 6,225 6,221 Multilateral 2,791 2,884 2,926 3,144 3,278 3,267 3,233 3,570 4,263 4,778 4,949 International Development Association 2,182 2,242 2,306 2,463 2,594 2,593 2,588 2,869 3,474 3,916 4,026 International Bank for Reconstruction and Development 87 56 34 22 16 11 8 6 3 0 0 Bilateral 3,005 2,828 2,588 2,692 2,800 2,259 1,955 2,076 1,374 1,353 1,166 Private creditors 465 392 261 251 239 206 88 73 92 94 106 Commercial banks 73 70 65 66 77 80 17 17 38 39 35 Other private creditors 392 322 196 185 162 126 71 56 54 55 71 Commitments 107 158 766 70 91 370 96 422 282 471 375 Multilateral 13 116 196 45 42 330 76 402 251 451 356 International Development Association 13 116 196 45 42 330 76 402 251 451 356 International Bank for Reconstruction and Development 0 0 0 0 0 0 0 0 0 0 0 Bilateral 55 39 563 20 20 20 20 20 20 20 20 Private creditors 40 3 6 5 30 20 0 0 12 0 0 330 Commercial banks 5 1 0 0 5 5 0 0 20 0 0 Other private creditors 35 2 6 5 25 15 0 0 ­8 0 0 Disbursements 262 246 336 264 340 261 226 240 545 341 341 Multilateral 187 194 244 187 251 185 147 188 452 418 262 International Development Association 160 134 183 102 199 142 119 148 397 343 175 International Bank for Reconstruction and Development 0 0 0 0 0 0 0 0 0 0 0 Bilateral 33 9 8 28 3 18 26 0 4 0 23 Private creditors 42 4 0 0 7 5 1 0 20 0 18 Commercial banks 5 1 0 0 5 5 0 0 20 0 0 Other private creditors 37 3 0 0 2 0 1 0 0 0 18 Source: World Bank. Note: Public and publicly guaranteed debt only, excluding IMF obligations. See table D.6 for obligations to the IMF. a. Preliminary estimate. Table D.3. External Public Debt Outstanding, Commitment, and Disbursement by Creditor Type, 1995­2005 percentage 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Debt outstanding and disbursed 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Multilateral 37.6 39.0 40.6 41.0 40.6 45.5 47.8 48.6 56.7 59.4 59.4 International Development Association 29.4 30.4 32.0 32.1 32.1 36.1 38.2 39.1 46.2 48.0 48.0 International Bank for Reconstruction and Development 1.2 0.8 0.5 0.3 0.2 0.2 0.1 0.1 0.0 0.0 0.0 Bilateral 40.5 38.3 40.0 39.8 39.6 35.0 36.5 34.9 24.9 22.6 22.6 Private creditors 6.3 6.2 4.2 3.8 3.3 3.3 1.7 1.3 1.5 1.7 1.7 Commercial banks 1.0 0.9 0.9 0.9 1.0 1.1 0.3 0.2 0.5 0.5 0.5 Other private creditors 5.3 5.0 3.3 2.9 2.4 2.1 1.4 1.1 1.0 1.2 1.2 Commitments 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Multilateral 11.6 73.3 25.6 64.5 45.4 89.2 79.2 95.3 88.7 95.8 95.8 International Development Association 11.6 73.3 25.6 64.5 45.4 89.2 79.2 95.3 88.7 95.8 95.8 International Bank for Reconstruction and Development 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Bilateral 51.1 24.8 73.6 28.3 21.8 5.4 20.8 4.7 7.0 4.2 4.2 Private creditors 37.2 1.9 0.8 7.1 32.8 5.4 0.0 0.0 4.2 0.0 0.0 331 Commercial banks 4.7 0.6 0.0 0.0 5.5 1.4 0.0 0.0 7.1 0.0 0.0 Other private creditors 32.6 1.3 0.8 7.1 27.4 4.1 0.0 0.0 -2.8 0.0 0.0 Disbursements 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Multilateral 71.4 78.9 72.6 70.8 73.8 70.9 65.0 78.3 82.9 83.0 83.0 International Development Association 61.1 54.5 54.5 38.6 58.5 54.4 52.7 61.7 72.8 56.6 56.6 International Bank for Reconstruction and Development 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Bilateral 12.6 3.7 2.4 10.6 0.9 6.9 11.9 0.0 8.8 6.5 6.5 Private creditors 15.6 2.4 0.0 0.0 2.1 1.9 0.4 0.0 3.7 5.3 5.3 Commercial banks 1.9 0.4 0.0 0.0 1.5 1.9 0.0 0.0 3.7 0.0 0.0 Other private creditors 14.1 2.0 0.0 0.0 0.6 0.0 0.4 0.0 0.0 5.3 5.3 Source: Table D.2. Note: Public and publicly guaranteed debt only, excluding IMF obligations. See table D.6 for obligations to the IMF. 332 Table D.4. External Public Debt Service Payments by Creditor Type, 1995­2005 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005a Principal paymentsb (US$ million) 126.0 143.0 86.2 93.3 103.9 92.9 98.6 56.7 40.7 38.0 143.3 Multilateral 78.0 73.0 64.0 64.0 52.0 29.0 44.0 22.0 21.0 28.0 91.0 International Development Association 12.0 14.0 15.0 17.0 24.0 9.0 19.0 5.0 9.0 13.0 64.0 International Bank for Reconstruction and Development 34.0 27.0 20.0 12.0 6.0 4.0 3.0 2.0 3.0 3.0 0.0 Bilateral 21.0 49.0 4.0 16.0 32.0 47.0 37.0 17.0 19.0 10.0 47.0 Private creditors 27.0 21.0 18.2 13.3 19.9 16.9 17.6 17.7 0.7 0.0 5.3 Commercial banks 0.0 4.0 2.0 1.0 1.0 0.0 7.0 1.0 0.2 0.0 4.2 Other private creditors 27.0 17.0 16.2 12.3 18.9 16.9 10.6 16.7 0.4 0.0 1.1 Interest paymentsb (US$ million) 80.0 94.6 39.6 92.9 79.2 51.0 38.4 34.7 32.5 42.0 66.7 Multilateral 56.0 46.0 35.0 36.0 29.0 23.0 24.0 23.0 26.5 36.0 41.0 International Development Association 16.0 17.0 16.0 17.0 19.0 15.0 16.0 17.0 20.0 28.0 31.0 International Bank for Reconstruction and Development 8.0 6.0 4.0 2.0 1.6 1.1 0.8 0.6 0.4 0.2 0.0 Bilateral 20.0 44.0 3.0 55.4 39.2 26.0 13.4 10.6 6.0 6.0 23.9 Private creditors 4.0 4.6 1.6 1.5 11.0 2.0 1.0 1.1 0.0 0.0 1.7 Commercial banks 0.0 1.0 0.1 0.2 0.1 0.0 0.0 0.0 0.0 0.0 1.3 Other private creditors 4.0 3.6 1.5 1.2 10.9 2.0 1.0 1.1 0.0 0.0 0.4 Memo items Average terms of new commitments All creditors Interest rate (%) 2.0 1.0 1.0 2.0 2.0 1.0 1.0 1.0 .. .. .. Maturity (years) 17.0 35.0 42.0 36.0 35.0 38.0 41.0 34.0 .. .. .. Grace period (years) 6.0 9.0 10.0 8.0 10.0 10.0 10.0 8.0 .. .. .. Grant element (%). 48.0 78.0 78.0 67.0 69.0 78.0 80.0 73.0 .. .. .. Official creditors .. .. .. Interest rate (%) 2.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 .. .. .. Maturity (years) 22.0 36.0 42.0 37.0 40.0 39.0 41.0 34.0 .. .. .. Grace period (years) 7.0 9.0 10.0 8.0 10.0 10.0 10.0 8.0 .. .. .. Grant element (%) 61.0 74.0 80.0 69.0 77.0 80.0 80.0 73.0 .. .. .. Private creditors .. .. .. Interest rate (%) 2.0 9.0 5.0 7.0 5.0 3.0 0.0 6.0 .. .. .. Maturity (years) 4.0 6.0 3.0 4.0 9.0 24.0 0.0 3.0 .. .. .. Grace period (years) 2.0 2.0 1.0 2.0 6.0 4.0 0.0 1.0 .. .. .. Grant element (%) 15.0 0.0 9.0 8.0 28.0 52.0 0.0 7.0 .. .. .. Source: World Bank. 333 Note: Public and publicly guaranteed debt only, excluding IMF obligations. See table D.6 for obligations to the IMF. a. Preliminary estimate. b. Cash basis; see table C.10 for scheduled debt service payments. Table D.5. Share of External Public Debt Service Payments by Creditor Type, 1995­2005 percentage 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005a Principal paymentsb 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Multilateral 61.9 51.0 74.2 68.6 50.0 31.2 44.6 38.8 51.6 73.7 63.5 International Development Association 9.5 9.8 17.4 18.2 23.1 9.7 19.3 8.8 22.1 34.2 44.7 International Bank for Reconstruction and Development 27.0 18.9 23.2 12.9 5.8 4.3 3.0 3.5 7.4 7.9 0.0 Bilateral 16.7 34.3 4.6 17.1 30.8 50.6 37.5 30.0 46.7 26.3 32.8 Private creditors 21.4 14.7 21.1 14.3 19.2 18.2 17.8 31.2 1.6 0.0 3.7 Commercial banks 0.0 2.8 2.3 1.1 1.0 0.0 7.1 1.8 0.6 0.0 2.9 Other private creditors 21.4 11.9 18.8 13.2 18.2 18.2 10.8 29.5 1.1 0.0 0.7 Interest paymentsb 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Multilateral 70.0 48.6 88.4 38.8 36.6 45.1 62.5 66.2 81.5 85.7 61.5 International Development Association 20.0 18.0 40.4 18.3 24.0 29.4 41.7 48.9 61.6 66.7 46.5 International Bank for Reconstruction and Development 10.0 6.3 10.1 2.2 2.1 2.1 2.1 1.7 1.2 0.4 0.0 Bilateral 25.0 46.5 7.6 59.6 49.5 51.0 34.9 30.5 18.4 14.3 35.9 Private creditors 5.0 4.9 4.0 1.6 13.8 3.9 2.7 3.3 0.0 0.0 2.6 334 Commercial banks 0.0 1.1 0.3 0.3 0.1 0.0 0.0 0.0 0.0 0.0 2.0 Other private creditors 5.0 3.8 3.7 1.3 13.7 3.9 2.7 3.3 0.0 0.0 0.6 Source: Table D.4. Note: Public and publicly guaranteed debt only, excluding IMF obligations. See table D.6 for obligations to the IMF. a. Preliminary estimate. b. Cash basis; see table C.10 for scheduled debt service payments. Table D.6. Obligations to the International Monetary Fund, 1995­2005 millions of U.S. dollars 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005a Use of IMF credit 188.2 206.0 246.0 268.0 312.0 324.0 341.0 400.0 437.0 423.0 363.0 Purchases .. 37.0 85.0 48.0 80.0 53.0 51.0 52.0 25.0 8.3 0.0 Debt service obligations 28.5 23.0 32.0 39.0 30.0 12.0 4.0 7.0 9.0 43.0 43.0 Repurchases 27.4 22.0 31.0 38.0 29.0 10.0 2.0 5.0 7.3 2.5 59.5 Charges 1.1 1.4 1.0 1.2 1.5 1.6 1.7 1.8 2.1 2.1 2.0 Net Use of IMF Credit .. 14.0 53.0 9.0 50.0 41.0 47.0 45.0 16.0 ­34.7 ­43.0 Source: World Bank. Note: The original debt figures in special drawing rights were converted into U.S. dollars; the figures may be different from those of the Debtor Reporting System. a. Preliminary estimate. 335 APPENDIX E Central Government Revenue and Expenditure Table E.1. Summary of Central Government Operations, 1995­2006 Fiscal year 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006a Government operations in T Sh million Total revenue 331,239 448,373 572,030 627,500 689,325 777,644 929,625 1,042,945 1,217,517 1,459,303 1,773,709 2,124,844 Tax revenue 299,899 383,744 505,355 586,164 616,284 685,106 827,789 938,478 1,105,746 1,342,798 1,615,247 1,946,433 Nontax revenue 31,340 64,630 66,675 41,336 73,041 92,538 101,836 104,467 111,771 136,979 158,462 178,411 Total expendi- ture and net lending 453,393 500,116 730,878 856,177 927,732 1,168,778 1,307,214 1,462,767 1,989,538 2,516,943 3,248,352 3,873,255 Recurrent expenditure 386,573 470,014 606,308 669,592 791,208 808,865 1,020,961 1,118,156 1,488,641 1,780,115 2,017,490 2,661,863 Development 338 expenditure and part of net lending 66,820 30,102 124,570 186,585 136,524 359,913 286,253 344,611 500,897 736,828 1,230,862 1,211,392 Overall bal- ance (checks issued) ­122,154 ­51,743 ­158,848 ­228,677 ­238,407 ­391,134 ­377,589 ­419,822 ­772,021 ­1,057,640 ­1,474,642 ­1,748,411 Financing 122,153 51,744 158,849 237,094 238,407 391,134 377,589 419,822 772,020 1,057,640 1,474,642 1,748,411 External grants 67,250 46,882 115,400 119,358 169,946 280,307 293,436 383,479 622,302 696,673 992,975 911,449 Foreign fi- nancing, netb 14,320 15,509 62,100 136,890 53,843 105,417 87,139 118,470 199,725 434,235 423,642 561,219 Domestic borrowing, net 60,800 57,117 ­43,351 3,669 52,659 7,900 ­2,495 ­24,159 ­36,514 ­34,496 144,954 329,885 Bank and parastatal recapitaliza- tion 0 0 0 0 0 0 0 Changes in arrears 0 ­59,107 0 0 0 0 Privatization 0 17,700 24,700 0 7,000 0 26,693 0 0 0 0 33,309 Adjustment to cash ­20,217 ­85,464 0 ­22,823 ­45,041 ­2,490 ­3,772 1,139 65,901 50,776 53,211 69,713 Expenditure float ­23,412 0 ­79,394 ­89,548 ­140,141 ­157,163 Central govern- ment operations as a % of GDP at market prices Total revenue 12.5 13.2 13.5 12.2 11.5 11.4 12.0 11.8 12.1 12.7 13.6 14.0 Tax revenue 11.3 11.3 11.9 11.4 10.3 10.0 10.7 10.6 11.0 11.7 12.4 12.8 Nontax revenue 1.2 1.9 1.6 0.8 1.2 1.4 1.3 1.2 1.1 1.2 1.2 1.2 Total expendi- ture and net lending 17.0 14.7 17.2 16.6 15.5 17.1 16.8 16.5 19.8 21.9 25.0 25.5 339 Recurrent expenditure 14.5 13.8 14.3 13.0 13.2 11.8 13.1 12.6 14.8 15.5 15.5 17.6 Development expenditure and part of net lending 2.5 0.9 2.9 3.6 2.3 5.3 3.7 3.9 5.0 6.4 9.5 8.0 Overall bal- ance (checks issued) ­4.6 ­1.5 ­3.7 ­4.4 ­4.0 ­5.7 ­4.9 ­4.7 ­7.7 ­9.2 ­11.3 ­11.5 Financing 4.6 1.5 3.7 4.6 4.0 5.7 4.9 4.7 7.7 9.2 11.3 11.5 External grants 2.5 1.4 2.7 2.3 2.8 4.1 3.8 4.3 6.2 6.1 7.6 6.0 Foreign fi- nancing, netb 0.5 0.5 1.5 2.7 0.9 1.5 1.1 1.3 2.0 3.8 3.3 3.7 Domestic borrowing, net 2.3 1.7 ­1.0 0.1 0.9 0.1 0.0 ­0.3 ­0.4 ­0.3 1.1 2.2 continued Table E.1 continued Bank and parastatal recapitaliza- tion 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Changes in arrears 0.0 0.0 0.0 0.0 0.0 0.0 0.0 ­0.7 0.0 0.0 0.0 0.0 Privatization 0.0 0.5 0.6 0.0 0.1 0.0 0.3 0.0 0.0 0.0 0.0 0.2 Adjustment to cash ­0.8 ­2.5 0.0 ­0.4 ­0.8 0.0 0.0 0.0 0.7 0.4 0.4 0.5 Expenditure float 0.0 0.0 0.0 0.0 0.0 0.0 ­0.3 0.0 ­0.8 ­0.8 ­1.1 ­1.0 Source: Tanzanian authorities. a. Preliminary estimate. b. Includes foreign grants prior to fiscal year 1987. 340 Table E.2. Central Government Revenue, 1995­2006 millions of Tanzania shillings Fiscal year 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006a Total revenue 331,240 448,339 572,053 619,104 689,518 772,459 929,624 1,042,945 1,217,517 1,459,303 1,773,709 2,124,844 Tax revenue 299,900 383,752 514,557 566,144 616,477 691,944 827,789 938,478 1,105,746 1,342,798 1,615,247 1,946,433 Import duties 51,599 61,271 78,374 81,751 87,800 87,679 95,632 88,876 106,430 136,979 111,855 190,856 Value added tax 90,921 120,639 134,612 138,203 208,222 223,758 301,983 352,263 424,338 503,758 731,597 845,790 On domestic goods and services 51,271 60,667 68,593 61,758 102,531 112,511 119,840 143,589 174,484 232,614 291,801 343,353 On imports 39,650 59,972 66,018 76,445 105,691 111,247 182,143 208,674 249,854 315,958 439,796 502,437 On petroleum 9,161 11,970 182 725 58,620 60,561 65,520 90,359 125,784 153,166 On other imports 56,857 64,475 105,509 110,522 123,523 148,113 184,334 225,599 314,012 349,271 Excises 21,372 43,983 91,685 101,251 84,194 88,508 154,765 177,614 187,265 215,341 238,677 261,550 On domestic 341 goods and services 21,372 43,983 61,925 78,785 58,805 66,960 68,999 72,837 85,263 92,870 110,336 135,043 On imports 29,760 22,466 25,390 21,547 85,766 104,777 102,002 122,471 128,341 126,507 On petroleum 8,287 10,856 18,483 16,972 82,041 98,810 97,063 117,693 115,849 116,387 On other imports 21,473 11,610 6,907 4,575 3,725 5,967 4,939 4,778 12,492 10,120 Income tax 86,645 112,261 145,372 149,788 162,904 205,076 194,013 228,377 276,050 363,976 465,455 581,244 Pay-as-you- earn tax 23,548 26,895 38,358 47,821 55,201 73,290 92,744 116,567 132,699 183,230 233,984 286,808 Corporate in- come tax 43,044 51,290 54,690 65,073 48,001 45,497 46,634 57,057 77,852 109,072 152,178 203,373 Other income tax 20,053 34,076 52,324 25,180 29,344 51,022 48,164 54,753 65,498 71,675 79,293 91,063 continued Table E.2 continued Other taxes 49,363 45,598 64,514 90,412 73,356 86,923 81,396 91,348 111,663 122,743 132,040 136,954 Business skills development levy 10,586 11,806 14,714 15,826 17,091 27,237 25,230 36,535 35,834 Others 79,826 61,550 60,205 65,486 74,257 84,426 97,472 31,129 31,159 Road toll/fuel levy 36,703 38,395 37,510 43,486 50,137 55,870 64,884 68,381 69,553 Stamp duty (excluding exports) 32,925 11,602 11,695 11,649 11,030 14,095 15,801 5,570 4,461 Other (value added tax refunds) 10,198 11,553 10,999 10,351 5,805 6,143 8,201 ­52,772 ­51,716 Departure tax 7,285 7,164 8,586 9,950 8,862 Nontax revenue 31,340 64,587 57,496 52,960 73,041 80,514 101,836 104,467 111,771 116,505 158,462 178,411 Parastatal dividends 13,376 5,512 15,489 11,732 25,590 17,190 17,170 11,145 15,587 13,232 342 Treasury 3,221 7,930 10,103 12,471 18,636 16,219 13,485 17,512 26,722 17,602 Ministries and regions 30,689 32,948 42,886 46,784 51,442 68,011 78,049 83,850 106,019 124,642 Appropriations in aid 448 446 79 65 110 62 27 19 6 6 Value Added Tax and Customs Department 9,762 6,125 4,485 9,463 6,057 2,985 3,040 3,979 10,128 22,930 Source: Tanzanian authorities. a. Preliminary estimate. Table E.3. Share of Central Government Revenue, 1995­2006 percentage Fiscal year 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Total revenue 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Tax revenue 90.5 85.6 89.9 91.4 89.4 89.6 89.0 90.0 90.8 92.0 91.1 91.6 Import duties 15.6 13.7 13.7 13.2 12.7 11.4 10.3 8.5 8.7 9.4 6.3 9.0 Value added tax 27.4 26.9 23.5 22.3 30.2 29.0 32.5 33.8 34.9 34.5 41.2 39.8 On domestic goods and services 15.5 13.5 12.0 10.0 14.9 14.6 12.9 13.8 14.3 15.9 16.5 16.2 On imports 12.0 13.4 11.5 12.3 15.3 14.4 19.6 20.0 20.5 21.7 24.8 23.6 On petroleum 0.0 0.0 1.6 1.9 0.0 0.1 6.3 5.8 5.4 6.2 7.1 7.2 On other imports 0.0 0.0 9.9 10.4 15.3 14.3 13.3 14.2 15.1 15.5 17.7 16.4 Excises 6.5 9.8 16.0 16.4 12.2 11.5 16.6 17.0 15.4 14.8 13.5 12.3 On domestic goods and services 6.5 9.8 10.8 12.7 8.5 8.7 7.4 7.0 7.0 6.4 6.2 6.4 On imports 0.0 0.0 5.2 3.6 3.7 2.8 9.2 10.0 8.4 8.4 7.2 6.0 343 On petroleum 0.0 0.0 1.4 1.8 2.7 2.2 8.8 9.5 8.0 8.1 6.5 5.5 On other imports 0.0 0.0 3.8 1.9 1.0 0.6 0.4 0.6 0.4 0.3 0.7 0.5 Income tax 26.2 25.0 25.4 24.2 23.6 26.5 20.9 21.9 22.7 24.9 26.2 27.4 Pay-as-you-earn tax 7.1 6.0 6.7 7.7 8.0 9.5 10.0 11.2 10.9 12.6 13.2 13.5 Corporate income tax 13.0 11.4 9.6 10.5 7.0 5.9 5.0 5.5 6.4 7.5 8.6 9.6 Other income tax 6.1 7.6 9.1 4.1 4.3 6.6 5.2 5.2 5.4 4.9 4.5 4.3 Other taxes 14.9 10.2 11.3 14.6 10.6 11.3 8.8 8.8 9.2 8.4 7.4 6.4 Payroll and human resources 0.0 0.0 0.0 1.7 1.7 1.9 1.7 1.6 2.2 1.7 2.1 1.7 Others 0.0 0.0 0.0 12.9 8.9 7.8 7.0 7.1 6.9 6.7 1.8 1.5 Road toll/fuel levy 0.0 0.0 0.0 5.9 5.6 4.9 4.7 4.8 4.6 4.4 3.9 3.3 Stamp duty (excluding exports) 0.0 0.0 0.0 5.3 1.7 1.5 1.3 1.1 1.2 1.1 0.3 0.2 Other (value added tax refunds) 0.0 0.0 0.0 1.6 1.7 1.4 1.1 0.6 0.5 0.6 ­3.0 ­2.4 Departure tax 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.7 0.6 0.6 0.6 0.4 continued Table E.3 continued Nontax revenue 9.5 14.4 10.1 8.6 10.6 10.4 11.0 10.0 9.2 8.0 8.9 8.4 Parastatal dividends 0.0 0.0 2.3 0.9 2.2 1.5 2.8 1.6 1.4 0.8 0.9 0.6 Treasury 0.0 0.0 0.6 1.3 1.5 1.6 2.0 1.6 1.1 1.2 1.5 0.8 Ministries and regions 0.0 0.0 5.4 5.3 6.2 6.1 5.5 6.5 6.4 5.7 6.0 5.9 Appropriations in aid 0.0 0.0 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Value Added Tax and Customs Department 0.0 0.0 1.7 1.0 0.7 1.2 0.7 0.3 0.2 0.3 0.6 1.1 Source: Table E.2. 344 Table E.4. Central Government Expenditure, 1995­2006 millions of Tanzania shillings Fiscal year 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006a Total expenditure 487,900 595,986 678,328 808,698 1,021,504 1,271,922 1,314,832 1,521,875 1,989,538 2,516,943 3,248,352 3,873,255 Recurrent expenditure 401,900 474,586 528,128 567,809 657,255 807,614 986,579 1,171,442 1,488,641 1,780,115 2,017,490 2,661,863 Wages 111,500 156,086 199,228 218,842 220,478 285,336 308,052 341,982 397,770 462,964 550,554 656,789 Interest 89,700 112,600 110,800 115,967 95,900 126,961 128,150 121,125 99,780 108,534 143,285 218,861 Domestic 57,700 77,700 73,500 50,621 38,200 81,400 77,788 64,605 57,009 70,232 87,393 163,695 Foreign 32,000 34,900 37,300 65,345 57,700 45,561 50,361 56,520 42,771 38,302 55,892 55,167 Goods, services, and transfers 200,700 205,900 218,100 233,000 340,877 395,317 550,378 708,335 991,091 1,208,617 1,323,651 1,786,213 CFS (others) 22,579 36,832 55,475 61,657 79,752 102,319 155,615 159,036 226,921 Tanzania Revenue Au- thority/Capital 345 Markets and Security Authority 16,661 19,462 21,755 27,048 31,968 34,172 39,092 43,424 67,655 Subsidies to TANESCO 0 5,000 10,000 11,500 13,000 85,277 70,640 17,400 Retrenchment costs 13,225 10,900 2,400 0 461 Parastatal wages 27,895 43,862 52,405 52,417 55,186 67,659 75,962 94,491 117,466 Other 165,865 215,994 255,682 409,255 516,704 763,041 850,272 956,060 1,356,311 Priority sector spending 64,900 96,225 108,085 168,260 300,985 427,856 356,183 432,557 .. Other charges 100,965 119,769 147,597 240,996 215,719 335,184 494,089 523,503 .. continued Table E.4 continued Development ex- penditure and net lending 86,000 121,400 110,300 197,389 247,949 359,908 286,253 291,325 500,897 736,828 1,230,862 1,211,392 Local 20,100 75,500 20,300 24,027 18,808 19,429 35,069 50,236 95,662 133,041 239,651 296,100 Foreign 65,900 45,900 90,000 173,362 229,142 340,479 251,184 241,090 405,235 603,787 991,211 915,292 Project grants 32,500 42,000 77,400 118,039 165,826 207,519 123,630 140,191 255,516 103,672 153,096 217,450 Project loans 33,400 3,900 12,600 40,300 63,315 132,960 127,554 100,899 149,719 1,902 129,208 82,302 Source: Tanzania authorities. Note: .. = data not available. a. Preliminary estimate. 346 Table E.5. Share of Central Government Expenditure, 1995­2006 percentage Fiscal year 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Total expenditure 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Recurrent expenditure 82.4 79.6 77.9 70.2 64.3 63.5 75.0 77.0 74.8 70.7 62.1 68.7 Wages 22.9 26.2 29.4 27.1 21.6 22.4 23.4 22.5 20.0 18.4 16.9 17.0 Interest 18.4 18.9 16.3 14.3 9.4 10.0 9.7 8.0 5.0 4.3 4.4 5.7 Domestic 11.8 13.0 10.8 6.3 3.7 6.4 5.9 4.2 2.9 2.8 2.7 4.2 Foreign 6.6 5.9 5.5 8.1 5.6 3.6 3.8 3.7 2.1 1.5 1.7 1.4 Goods, services, and transfers 41.1 34.5 32.2 28.8 33.4 31.1 41.9 46.5 49.8 48.0 40.7 46.1 CFS (others) 0.0 0.0 0.0 2.8 3.6 4.4 4.7 5.2 5.1 6.2 4.9 5.9 Tanzania Revenue Authority/Capital Mar- kets and Security 347 Authority 0.0 0.0 0.0 2.1 1.9 1.7 2.1 2.1 1.7 1.6 1.3 1.7 Subsidies to TANESCO 0.0 0.0 0.0 0.0 0.5 0.8 0.0 0.8 0.7 3.4 2.2 0.4 Retrenchment costs 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.9 0.5 0.1 0.0 0.0 Parastatal wages 0.0 0.0 0.0 3.4 4.3 4.1 4.0 3.6 3.4 3.0 2.9 3.0 Other 0.0 0.0 0.0 20.5 21.1 20.1 31.1 34.0 38.4 33.8 29.4 35.0 Priority sector spend- ing 0.0 0.0 0.0 8.0 9.4 8.5 12.8 19.8 21.5 14.2 13.3 .. Other charges 0.0 0.0 0.0 12.5 11.7 11.6 18.3 14.2 16.8 19.6 16.1 .. Development expenditure and net lending 17.6 20.4 16.3 24.4 24.3 28.3 21.8 19.1 25.2 29.3 37.9 31.3 Local 4.1 12.7 3.0 3.0 1.8 1.5 2.7 3.3 4.8 5.3 7.4 7.6 Foreign 13.5 7.7 13.3 21.4 22.4 26.8 19.1 15.8 20.4 24.0 30.5 23.6 Project grants 6.7 7.0 11.4 14.6 16.2 16.3 9.4 9.2 12.8 4.1 4.7 5.6 Project loans 6.8 0.7 1.9 5.0 6.2 10.5 9.7 6.6 7.5 0.1 4.0 2.1 Source: Table E.4. Table E.6. Central Government Expenditure by Economic Function, 1995­2005 millions of Tanzania shillings Fiscal year 1995 1996 1997 1998a 1999a 2000a 2001 2002 2003 2004 2005 Recurrent expenditure 362,799 470,014 606,308 661,234 733,276 937,044 1,117,959 1,439,497 1,374,963 1,737,209 2,284,567 General public servicesb 106,000 141,918 183,071 206,423 269,489 282,012 336,457 421,818 457,785 995,013 1,441,264 Defense 30,272 40,937 52,808 77,193 81,561 87,661 104,585 131,119 110,809 159,702 169,756 Social services 44,263 50,223 64,786 92,580 99,215 224,315 267,631 335,725 306,650 256,118 295,544 Education 18,608 24,122 31,117 39,001 44,874 141,651 168,998 212,068 136,657 121,745 129,962 Health 21,762 17,572 22,667 33,000 35,965 54,762 65,345 81,924 113,960 82,274 106,819 Other social servicesc 3,893 8,529 11,002 20,579 18,377 27,901 33,288 41,733 56,033 52,099 58,763 Economic services 29,122 31,406 40,513 38,543 74,542 77,399 92,342 115,320 161,279 214,077 306,541 Agriculture, fishing, and forestry 7,430 6,308 8,137 13,788 19,872 19,813 23,638 29,634 49,445 64,777 61,120 Mining, manufacturing, and construction 1,537 1,302 1,679 1,525 2,288 2,536 3,025 3,793 4,011 6,282 13,867 348 Water and electricity 2,739 410 528 2,859 2,019 1,983 2,366 2,966 12,637 22,572 75,999 Roads and bridges 12,302 .. .. .. .. .. .. .. .. .. .. Transport and communication 1,701 17,987 23,203 12,482 43,957 44,042 52,544 65,875 82,356 99,570 131,230 Other economic services 3,413 5,400 6,966 7,889 6,406 9,026 10,769 13,052 12,830 20,876 24,325 Othersd 153,142 205,531 265,130 246,495 208,469 265,657 316,944 435,515 338,441 112,299 71,462 Public debt 121,134 .. .. 178,329 169,930 240,144 286,505 397,354 262,113 3,076 853 Development 151,486 235,634 389,700 401,033 408,480 508,571 591,403 678,985 833,862 457,148 827,573 General public servicesb 9,489 3,591 14,860 76,634 34,962 44,212 26,667 23,454 44,100 231,158 296,260 Defense 2,400 0 0 0 1,004 1,000 640 562 1,034 155 198 Social services 33,238 14,606 60,440 21,512 55,237 58,740 99,318 89,437 209,629 37,021 218,165 Education 20,357 9,904 40,983 8,189 9,214 18,582 35,829 34,337 103,730 2,065 7,768 Health 10,412 920 3,805 3,705 9,548 8,797 18,051 15,876 34,901 0 53,104 Other social servicesc 2,469 3,782 15,652 9,618 36,475 31,361 45,438 39,224 70,998 34,956 157,293 Economic services 106,359 11,906 49,270 56,392 108,808 138,963 147,834 130,017 240,658 188,969 313,148 Agriculture, fishing, and forestry 14,543 1,301 5,385 10,543 22,669 26,270 18,401 16,183 31,313 56,061 36,241 Mining, manufacturing, and construction 4,463 6,901 28,557 15 7,225 4,863 49,731 43,738 79,221 26,770 47,785 Water and electricity 24,100 0 0 8,075 13,588 20,197 0 0 0 0 2,100 Transport and communication 30,716 2,566 10,617 23,297 34,758 46,054 46,781 41,143 76,894 77,886 199,441 Other economic services 5,243 1,138 4,711 14,462 30,568 41,579 32,921 28,953 53,230 28,252 27,581 Othersd .. 205,531 265,130 246,495 208,469 265,657 316,944 435,515 338,441 0 0 Total expenditure 514,285 705,648 996,007 1,062,267 1,141,756 1,445,616 1,709,362 2,118,482 2,208,825 2,194,357 3,112,140 General public servicesb 115,489 145,509 197,931 283,056 304,451 326,224 363,124 445,272 501,885 1,226,171 1,737,524 Defense 32,672 40,937 52,808 77,193 82,565 88,661 105,225 131,681 111,843 159,857 169,954 Social services 77,501 64,829 125,226 114,092 154,452 283,055 366,949 425,162 516,279 293,139 513,709 Education 38,965 34,026 72,100 47,190 54,088 160,234 204,827 246,405 240,387 123,810 137,730 Health 32,174 18,492 26,472 36,705 45,513 63,559 83,396 97,800 148,861 82,274 159,923 Other social servicesc 6,362 12,311 26,654 30,197 54,852 59,262 78,726 80,957 127,031 87,055 216,056 Economic services 135,481 43,312 89,782 94,935 183,350 216,362 240,176 245,337 401,937 403,046 619,689 Agriculture, fishing, 349 and forestry 21,973 7,609 13,521 24,331 42,542 46,082 42,039 45,817 80,758 120,838 97,361 Mining, manufacturing, and construction 6,000 8,203 30,237 1,540 9,513 7,399 52,756 47,531 83,232 33,052 61,652 Water and electricity 26,839 410 528 10,934 15,607 22,180 2,366 2,966 12,637 22,572 78,099 Transport and communication 32,417 20,552 33,820 35,779 78,715 90,096 99,325 107,018 159,250 177,456 330,671 Other economic services 8,656 6,538 11,677 22,351 36,974 50,605 43,690 42,005 66,060 49,128 51,906 Othersd 153,142 411,062 530,260 492,990 416,938 531,313 633,888 871,030 676,882 112,299 71,462 Public debt 121,134 .. .. 178,329 169,930 240,144 286,505 397,354 262,113 3,076 853 Sources: United Republic of Tanzania; The Economic Survey (various issues). Note: The expenditure figures are different from those in the preceding tables because of different sources of data. Also, totals may differ because of different sources of data. a. Estimates; otherwise actual. b. Includes general administration, external affairs, and public order and safety. c. Includes housing, community amenities, community development, and sanitary services. d. Includes public debt, financial and capital subscriptions, and pension and gratuity. APPENDIX F Monetary Situation Table F.1. Monetary Survey, 1995­2006 millions of Tanzania shillings Fiscal year 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Net foreign assets 150,179 204,147 372,876 411,235 472,342 707,352 1,005,682 1,317,096 1,871,485 2,134,317 2,134,317 2,134,317 Assets 301,048 362,083 538,730 581,912 718,936 995,745 1,373,305 1,738,150 2,407,306 2,748,846 2,748,846 2,748,846 Bank of Tanzania 153,795 148,894 287,672 334,402 445,798 600,984 872,888 1,148,306 1,748,059 2,079,550 2,079,550 2,079,550 Commercial banks 147,253 213,189 251,058 247,510 273,138 394,761 500,417 589,844 659,247 669,296 669,296 669,296 Liabilities 150,869 157,936 165,854 170,677 246,594 288,393 367,623 421,054 535,821 614,529 614,529 614,529 Bank of Tanzania 143,329 153,375 162,423 168,813 246,088 287,031 359,228 417,818 533,023 610,676 610,676 610,676 Commercial banks 7,540 4,561 3,431 1,864 506 1,362 8,395 3,236 2,798 3,853 3,853 3,853 Medium-term foreign liabilities 39,981 44,192 43,865 43,587 43,613 43,613 42,172 41,895 41,801 41,133 41,134 41,135 Net domestic assets 548,987 593,202 587,541 640,548 669,774 692,558 662,473 750,101 710,283 962,901 962,901 962,901 Domestic credit 474,221 435,285 417,549 438,344 593,606 697,759 647,342 684,920 858,204 1,123,588 1,123,588 1,123,588 352 Public sector 302,611 331,368 306,386 249,652 320,562 398,640 272,564 202,879 199,679 157,607 157,607 157,607 Government, net 212,149 305,905 287,547 239,665 314,109 384,321 268,920 202,879 199,679 157,607 157,607 157,607 Nonfinancial public enter- prises 90,462 25,463 18,839 9,987 6,453 14,319 3,644 0 0 0 0 0 Private sector 171,610 103,917 111,163 188,692 273,044 299,119 374,778 482,041 658,525 965,981 965,981 965,981 Other items, net 74,766 157,917 169,992 202,204 76,168 ­5,201 15,131 65,181 ­147,921 ­160,687 ­160,687 ­160,687 Broad money (M3) 640,148 744,177 879,963 947,394 1,034,895 1,258,976 1,472,904 1,797,890 2,205,894 2,602,935 2,602,935 2,602,935 Money 338,176 383,564 472,621 481,642 508,994 601,193 691,255 815,576 981,148 1,186,063 1,186,063 1,186,063 Quasi money 301,973 360,613 407,341 465,752 525,900 657,783 781,649 982,314 1,224,746 1,416,872 1,416,872 1,416,872 Valuation 6,891 8,579 36,590 59,833 63,570 97,322 148,808 203,962 333,873 453,015 453,015 453,015 Source: Bank of Tanzania. Note: Year's end is June 30. Table F.2. Change in Money Supply and Sources of Change, 1997­2005 percentage of previous period stock of money Fiscal year 1997 1998 1999 2000 2001 2002 2003 2004 2005 Net foreign assets 8.4 22.7 4.4 6.5 22.7 23.7 21.1 30.8 11.9 Medium-term foreign liabilities ­0.9 0.1 0.0 0.0 0.0 0.1 0.0 0.0 0.0 Net domestic assets 6.9 ­0.8 6.0 3.1 2.2 ­2.4 5.9 ­2.2 11.5 Domestic credit ­6.1 ­2.4 2.4 16.4 10.1 ­4.0 2.6 9.6 12.0 Public sector 4.5 ­3.4 ­6.4 7.5 7.5 ­10.0 ­4.7 ­0.2 ­1.9 Government, net 14.6 ­2.5 ­5.4 7.9 6.8 ­9.2 ­4.5 ­0.2 ­1.9 Nonfinancial public enterprises ­10.2 ­0.9 ­1.0 ­0.4 0.8 ­0.8 ­0.2 0.0 0.0 Private sector ­10.6 1.0 8.8 8.9 2.5 6.0 7.3 9.8 13.9 Other items, net 13.0 1.6 3.7 ­13.3 ­7.9 1.6 3.4 ­11.9 ­0.6 External arrears counterpart .. .. .. .. .. .. .. .. .. Broad money (M3) 16.3 18.2 7.7 9.2 21.7 17.0 22.1 22.7 18.0 Money 7.1 12.0 1.0 2.9 8.9 7.2 8.4 9.2 9.3 Quasi money 9.2 6.3 6.6 6.3 12.7 9.8 13.6 13.5 8.7 353 Valuation 0.3 3.8 2.6 0.4 3.3 4.1 3.7 7.2 5.4 Source: Table F.1. Note: Year's end is June 30. Table F.3. Balance Sheet of the Bank of Tanzania, 1995­2005 millions of Tanzania shillings Fiscal year 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Assets 970,667 1,059,796 975,668 964,449 1,285,780 1,443,819 1,698,665 2,178,445 2,921,192 3,328,362 3,766,190 Foreign assets 142,481 256,439 381,689 397,705 607,421 772,190 1,048,262 1,479,519 2,151,938 2,378,616 2,524,279 Foreign exchange 126,895 237,247 362,312 376,350 582,315 746,523 1,018,903 1,452,491 2,151,413 2,378,542 2,523,457 Gold reserves 15,550 18,846 19,310 21,115 25,002 25,353 29,013 26,926 0 0 0 Special drawing rights holdings 36 346 67 240 104 314 345 102 525 74 822 354 Reserve position in the International Mone- tary Fund (IMF) 0 0 0 0 0 0 0 0 0 0 0 Other foreign assets 0 0 0 0 0 0 0 0 0 0 0 Quota in IMF resources 120,141 125,595 124,504 140,534 217,803 208,448 228,839 263,041 307,705 321,091 331,294 Claims on government 326,371 299,375 269,693 234,075 302,788 296,673 296,673 201,457 202,202 199,211 234,679 Loans and advances 18,950 0 0 0 0 0 0 0 0 0 0 Government securities 0 0 0 0 0 0 0 0 0 0 0 Bills rediscounted 0 0 0 0 0 0 0 0 0 0 0 Adjustment for Struc- tural Adjustment Facility 0 0 0 0 0 0 0 0 0 0 0 Other adjustments 0 0 0 0 0 0 0 0 0 0 0 Other securities 307,421 299,375 269,693 234,075 302,788 296,673 296,673 201,457 202,202 199,211 234,679 Lending to banks 5,455 5,455 4,611 5,358 4,160 0 0 0 0 0 0 Revaluation 301,077 283,212 81,199 71,411 36,326 59,864 ­15,581 0 0 0 0 Others unclassified assets 75,143 89,721 113,972 115,366 117,282 106,644 140,471 234,428 259,347 429,444 675,937 Liabilities 970,667 1,059,796 975,668 964,449 1,285,780 1,443,819 1,698,665 2,178,445 2,921,192 3,328,362 3,766,190 Currency in circulation 264,208 280,576 314,487 337,323 427,447 443,051 456,206 546,615 606,593 727,785 932,815 Central government deposits 39,673 82,658 87,321 83,027 84,012 149,732 173,800 210,475 369,648 389,818 441,555 Banks deposits 50,677 55,192 50,453 81,411 81,227 113,380 128,164 149,086 176,717 239,991 303,265 Other deposits 32,718 21,921 47,415 8,568 24,516 ­9,015 4,231 11,710 18,224 43,526 60,808 Foreign liabilities 390,541 386,341 219,836 214,955 285,985 346,742 372,840 470,501 663,948 679,473 760,488 Use of IMF resources 118,602 109,275 112,859 122,440 180,886 200,507 212,712 231,439 271,954 305,136 315,929 Allocation of special drawing rights 25,657 26,822 26,589 30,012 34,354 32,878 36,094 41,489 48,534 50,645 52,254 Other 48,590 97,011 116,708 86,713 167,353 166,544 314,618 517,131 765,575 891,988 899,076 Source: Bank of Tanzania. Note: Year's end is June 30. 355 Table F.4. Balance Sheet of Commercial Banks, 1995­2005 millions of Tanzania shillings Fiscal year 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Assets 1,288,363 884,138 1,098,238 1,606,812 1,990,031 2,460,095 4,443,622 7,991,673 2,603,422 3,090,548 4,130,529 Domestic assets 1,064,369 644,556 850,723 1,294,005 1,642,806 2,002,139 3,847,886 7,354,214 1,869,848 2,354,823 3,234,105 Cash 19,895 22,913 26,610 29,524 42,599 50,647 44,567 51,170 53,547 63,637 89,658 Deposit with BOT 46,138 49,884 33,100 83,678 80,611 121,450 132,762 143,983 173,323 246,790 305,767 Treasury bills 48,658 89,455 78,623 58,863 65,912 89,620 102,319 192,929 176,361 147,928 459,031 Other securities 119,447 166,530 166,906 251,080 264,799 313,339 193,384 171,708 146,047 190,613 254,020 Loans and bills 260,829 147,298 184,840 251,079 312,082 341,428 405,373 570,668 817,125 1,060,077 1,425,062 Other 569,402 168,476 360,644 619,781 876,803 1,085,655 2,969,482 6,223,755 503,445 645,777 700,567 Foreign assets 170,163 181,551 233,556 267,257 301,079 410,990 539,063 571,573 685,488 683,666 835,640 Liquid 169,390 179,548 232,162 243,077 290,975 398,949 499,434 547,210 657,755 605,328 760,581 Other 773 2,003 1,394 24,180 10,104 12,041 39,629 24,362 27,733 78,337 75,059 Fixed assets 53,832 58,031 13,959 45,550 46,146 46,965 56,674 65,887 48,086 52,060 60,784 Other 0 0 0 0 0 0 0 0 0 0 0 356 Liabilities 1,288,363 884,138 1,098,238 1,606,812 1,990,031 2,460,095 4,442,762 7,990,673 2,603,422 3,090,548 4130,529 Domestic liabilities 1,301,358 805,227 1,050,250 1,543,223 1,961,676 2,342,604 4,292,854 7,814,392 2,388,574 2,828,365 3,761,135 Deposits 535,246 581,357 667,619 744,513 854,339 1,031,371 1,254,412 1,583,057 1,917,103 2,319,435 3,279,030 Liability to BOT 7,586 162 12,763 0 5,430 1 0 83 50 0 0 Liabilities to other banks 5,091 9,148 49,513 23,040 25,080 38,138 47,665 104,533 138,726 160,902 107,385 Other 753,435 214,560 320,355 775,669 1,076,827 1,273,094 2,990,777 6,126,720 332,695 348,028 374,720 Foreign liabilities 16,919 3,292 4,878 1,952 1,044 4,163 16,170 34,193 6,877 19,042 55,235 Liabilities to for- eign banks 4,171 1,757 3,933 1,873 928 4,056 3,015 5,543 6,322 19,042 27,572 Liabilities to others 12,748 1,535 945 78 116 107 13,155 28,650 555 0 27,663 Capital and reserves ­29,914 75,619 43,110 61,638 27,312 113,328 133,738 142,088 207,971 243,141 314,159 Source: Bank of Tanzania. Note: Year's end is June 30. Table F.5. Domestic Lending of Commercial Banks by Economic Sector, 1995­2005 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Lending amount (T Sh million) Production 78,421 50,089 45,779 68,452 100,595 117,759 147,271 208,019 272,750 401,896 471,780 Agricultural production 21,086 10,441 11,037 12,218 16,668 19,158 27,035 61,392 79,686 120,013 156,801 Mining and manufacturing 55,345 38,681 32,962 54,081 81,039 92,532 115,981 135,069 177,935 257,843 279,109 Tourism 1,990 968 1,780 2,153 2,888 6,070 4,255 11,558 15,128 24,041 35,871 Trade 103,310 38,682 31,718 55,610 76,525 89,198 101,843 128,302 164,742 236,317 293,253 Marketing of agricultural produce 51,294 7,290 5,233 3,518 5,503 2,949 1,242 0 0 0 0 Export of agricultural pro- duce 5,131 3,542 2,278 2,961 1,876 2,187 0 0 0 0 0 Trade in capital goods 12 8 19 17 0 175 0 0 0 0 0 Trade in noncapital goods 46,873 27,843 24,189 49,115 69,146 83,887 100,601 128,302 164,742 236,317 293,253 Other 79,099 271,649 308,715 324,401 430,442 484,653 429,808 412,748 568,199 703,272 860,263 Building and construction 3,385 3,112 4,590 5,159 6,682 8,826 15,932 24,002 33,377 38,332 54,390 Transportation sector 4,774 6,950 9,301 19,773 39,823 29,695 34,968 47,303 77,559 76,725 104,322 357 Treasury bills .. 58,787 86,274 71,047 64,370 85,324 104,755 95,666 190,446 205,112 183,633 Government bonds .. 166,728 165,353 176,005 260,603 292,847 195,744 171,360 156,720 170,388 222,358 Othera 70,940 36,073 43,199 52,417 58,964 67,960 78,410 74,418 110,098 212,716 295,561 Total 260,830 360,420 386,213 448,463 607,561 691,610 678,922 749,068 1,005,691 1,341,486 1,625,296 Lending amount as a share of total (%) Production 30 14 12 15 17 17 22 28 27 30 29 Agricultural production 8 3 3 3 3 3 4 8 8 9 10 Mining and manufacturing 21 11 9 12 13 13 17 18 18 19 17 Tourism 1 0 0 0 0 1 1 2 2 2 2 Trade 40 11 8 12 13 13 15 17 16 18 18 Marketing of agricultural produce 20 2 1 1 1 0 0 0 0 0 0 Export of agricultural pro- duce 2 1 1 1 0 0 0 0 0 0 0 Trade in capital goods 0 0 0 0 0 0 0 0 0 0 0 Trade in noncapital goods 18 8 6 11 11 12 15 17 16 18 18 continued Table F.5 continued Other 30 75 80 72 71 70 63 55 56 52 53 Building and construction 1 1 1 1 1 1 2 3 3 3 3 Transportation sector 2 2 2 4 7 4 5 6 8 6 6 Treasury bills 16 22 16 11 12 15 13 19 15 11 Government bonds 46 43 39 43 42 29 23 16 13 14 Othera 27 10 11 12 10 10 12 10 11 16 18 Total 100 100 100 100 100 100 100 100 100 100 100 Annual changes from the pre- vious year (%) Production ­24 ­36 ­9 50 47 17 25 41 31 47 17 Agricultural production ­14 ­50 6 11 36 15 41 127 30 51 31 Mining and manufacturing ­24 ­30 ­15 64 50 14 25 16 32 45 8 Tourism ­64 ­51 84 21 34 110 ­30 172 31 59 49 Trade ­35 ­63 ­18 75 38 17 14 26 28 43 24 Marketing of agricultural produce ­30 ­86 ­28 ­33 56 ­46 ­58 ­100 .. .. .. Export of agricultural pro- 358 duce ­44 ­31 ­36 30 ­37 17 ­100 .. .. .. .. Trade in capital goods ­100 ­34 149 ­12 ­100 .. ­100 .. .. .. .. Trade in noncapital goods ­36 ­41 ­13 103 41 21 20 28 28 43 24 Other 659 243 14 5 33 13 ­11 ­4 38 24 22 Building and construction ­4 ­8 47 12 30 32 81 51 39 15 42 Transportation sector ­55 46 34 113 101 ­25 18 35 64 ­1 36 Treasury bills .. 47 ­18 ­9 33 23 ­9 99 8 ­10 Government bonds .. ­1 6 48 12 ­33 ­12 ­9 9 31 Othera ­2,016 ­49 20 21 12 15 15 ­5 48 93 39 Total ­4 38 7 16 35 14 ­2 10 34 33 21 Source: United Republic of Tanzania, The Economic Survey (various issues). a. Includes public administration and financial institutions. Table F.6. Interest Rate Structure, 1995­2005 percentage 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Bank of Tanzania Discount ratea 45.0 27.9 20.5 20.7 16.6 13.1 9.8 9.4 10.6 12.6 13.9 Commercial banks Deposit rates, domestic currency Savings 21.0 16.7 12.8 12.0 7.4 7.1 4.2 3.5 2.5 2.4 2.6 Fixed (3­6 months) 25.0 21.0 29.0 13.1 8.3 9.1 5.2 4.0 3.5 4.3 4.4 Lending rates, domestic currency Short-term 36.4 37.0 24.5 28.0 21.4 19.1 20.9 14.8 16.4 15.7 16.1 Medium- and long-term 35.5 33.5 21.5 26.0 21.3 21.0 19.2 16.8 13.5 14.1 15.4 Government securities Treasury bills Auction, 91 days 48.7 14.9 15.6 9.7 7.3 4.2 4.1 6.1 7.7 8.1 14.7 Auction, 182 days 50.7 15.5 16.4 11.8 7.8 4.8 4.2 5.8 8.0 9.4 14.7 Auction, 364 days .. 17.3 17.9 13.7 10.3 5.1 4.9 5.9 8.2 10.5 15.7 359 Stocks, 10-year treasury bond 19.8 19.8 24.5 24.5 15.5 6.9 4.7 7.2 10.2 12.2 14.8 Memo item Real interest rateb ­6.4 ­4.3 ­3.3 ­0.9 ­0.4 1.1 ­0.9 ­1.1 ­1.0 ­1.8 ­1.7 Source: United Republic of Tanzania, The Economic Survey (various issues). a. Discount rate has been determined by the movement of marginal yields of 91-day treasury bills. b. Real interest rate has been derived on the basis of the deposit interest rate on savings offered by the commercial banks. Table F.7. Exchange Rate Movement, 1995­2005 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Period average T Sh/special drawing rights 791.1 837.3 849.6 900.8 1,026.6 1,071.2 1,099.3 1,202.0 1,241.3 1,300.9 1,437.6 T Sh/US$ 536.4 583.0 618.3 666.1 745.0 800.4 876.3 963.2 1,039.1 1,090.6 1,127.1 US$/T Sh 100.00 0.19 0.17 0.16 0.15 0.13 0.12 0.11 0.10 0.10 0.09 0.09 End of period T Sh/special drawing rights 818.1 856.5 842.7 958.9 1,094.3 1,048.0 1,150.5 1,253.4 1,229.1 1,372.7 1,502.5 T Sh/US$ 558.2 595.6 624.6 681.0 797.3 803.3 916.3 976.3 1,063.6 1,043.0 1,164.8 US$/T Sh 100.00 0.18 0.17 0.16 0.15 0.13 0.12 0.11 0.10 0.09 0.10 0.09 Real effective exchange rate (index: 2000 = 100) 67.9 83.2 94.1 101.6 98.3 100.0 99.9 90.8 75 67.7 65.7 Annual change (%) 10.2 22.5 13.1 8.0 ­3.2 1.7 ­0.1 ­9.1 ­17.4 ­9.7 ­3.0 Memo item Period average (in fiscal years), T Sh/US$ 523.0 559.7 600.7 642.2 705.6 772.7 838.4 919.8 1,001.2 1,064.9 1,108.9 Sources: IMF, International Finance Statistics; Tanzania authorities. 360 APPENDIX G Agricultural Production Table G.1. Production of Principal Cash Crops and Food Crops, 1994/95­2004/05 thousand tons Types of crops 1994/95 1995/96 1996/97 1997/98 1998/99 1999/2000 2000/01 2001/02 2002/03 2003/04 2004/05a Cash crops Coffee 35.8 46.9 43.5 30.8 39.7 32.8 58.0 53.5 76.4 38.7 33.9 Cottonb 123.7 245.9 249.4 207.8 103.7 99.9 125.0 149.0 188.0 140.0 342.0 Teac 138.7 15.9 90.9 120.5 105.8 105.9 118.8 111.3 132.4 128.0 133.4 Cashew nuts 63.4 81.7 63.0 99.9 106.4 121.2 122.3 79.3 95.0 80.0 71.0 Tobacco 22.7 28.6 35.4 51.1 38.0 24.4 24.8 27.7 27.9 43.5 47.5 Sisal 24.7 28.9 25.0 22.2 23.2 41.1 23.5 23.6 23.3 26.8 27.8 Pyrethrum 0.5 0.4 0.4 0.4 0.5 0.6 1.5 1.7 1.1 0.8 0.7 Food crops Maize 2,567.0 2,638.0 1,831.0 2,685.0 2,452.0 2,128.0 2,240.0 2,698.0 2,526.0 4,286.0 3,131.0 Paddy 723.0 733.8 550.8 810.8 506.0 476.0 991.7 960.6 562.0 1030.0 1077.0 Wheat 75.0 61.0 78.0 112.0 82.0 32.0 118.0 72.0 75.0 66.0 44.0 Pulses 374.0 384.0 374.0 453.0 528.0 584.0 674.0 683.0 713.0 937.0 721.0 Sources: Ministry of Agriculture, United Republic of Tanzania; United Republic of Tanzania, The Economic Survey (2005). 362 a. Preliminary estimate. b. Before seeds are removed. c. Green leaves. Table G.2. Selected Agricultural Production Indices, 1995/96­2004/05 1991/92 = 100 Types of crops 1994/95 1995/96 1996/97 1997/98 1998/99 1999/2000 2000/01 2001/02 2002/03 2003/04a 2004/05a Cash crops Coffee 68.6 89.9 83.3 59.1 76.1 62.9 111.1 102.5 146.4 74.1 64.9 Cottonb 459.7 914.2 927.3 772.3 385.7 371.4 464.7 553.9 698.9 520.3 1,271.4 Teac 757.9 86.9 496.9 658.7 578.1 578.9 648.9 608.2 723.5 699.3 729.1 Cashew nuts 153.9 198.3 152.9 242.5 258.4 294.2 296.8 192.4 230.6 194.2 172.3 Tobacco 137.3 173.3 214.4 309.7 230.0 148.1 150.3 167.9 168.9 263.9 287.6 Sisal 102.1 119.4 103.4 91.7 96.0 169.8 97.3 97.7 96.2 110.6 114.9 Pyrethrum 20.9 19.8 19.3 18.2 22.7 25.9 66.6 78.9 49.5 38.3 32.3 Food crops Maize 115.3 118.5 82.2 120.6 110.1 95.6 100.6 121.2 113.5 192.5 140.6 Paddy 183.5 186.2 139.8 205.8 128.4 120.8 251.7 243.8 142.6 261.4 273.4 Wheat 117.2 95.3 121.9 175.0 128.1 50.0 184.4 112.5 117.2 103.1 68.8 Pulses 119.9 123.1 119.9 145.2 169.2 187.2 216.0 218.9 228.5 300.3 231.1 Source: Table G.1. 363 a. Preliminary estimate. b. Before seeds are removed. c. Green leaves. Table G.3. Producer Price of Selected Cash Crops and Food Crops, 1994/95­2004/05 Tanzania shillings per kilogram Type of crop 1994/95 1995/96 1996/97 1997/98 1998/99 1999/2000 2000/01 2001/02 2002/03 2003/04 2004/05a Coffeeb 1,842 1,220 1,302 .. 1,574 1,503 .. .. .. 1,200 1,935 Cottonb 120 207 170 180 185 123 180 165 180 280 250 Tea 50 55 55 55 55 55 65 80 85 86 86 Cashew nutsb 330 380 300 330 460 600 250 240 325 416 675 Tobaccob 524 508 592 596 492 585 550 474 555 703 810 Sisal 216 371 .. .. .. .. 368 338 450 540 617 Pyrethrumc 250 300 300 300 320 320 400 420 380 360 .. Sugar cane 9 9 10 11 13 13 14 15 15 17 21 Source: United Republic of Tanzania, The Economic Survey (various issues). Note: .. not available. a. Preliminary estimate. b. Weighted average price, except the prices of cotton and cashew nuts from 1994/5 to 2001/02. c. Grade 5. 364 Table G.4. Producer Price Index of Selected Cash Crops and Food Crops, 1994/95­2004/05 1991/92 = 100 1994/95 1995/96 1996/97 1997/98 1998/99 1999/2000 2000/01 2001/02 2002/03 2003/04 2004/05a Coffeeb 945.7 626.4 668.6 .. 808.1 771.7 .. .. .. 616.1 993.3 Cottonb 171.4 295.7 242.9 257.2 264.3 175.7 257.2 235.7 257.2 399.6 357.2 Tea 125.0 137.5 137.5 137.5 137.5 137.5 162.5 200.0 212.5 215.0 215.0 Cashew nutsb 253.6 292.0 230.5 253.6 353.5 461.1 192.1 184.4 184.4 249.7 518.7 Tobaccob 214.9 208.5 242.8 244.4 201.9 239.9 .. 194.3 227.8 288.3 332.3 Sisal 246.0 422.3 .. .. .. .. 419.1 384.6 512.4 .. .. Pyrethrum 108.7 130.4 130.4 130.4 139.1 139.1 173.9 182.6 165.2 156.5 .. Sugar cane 247.4 256.0 273.2 302.0 359.5 359.5 402.7 422.8 431.4 483.2 592.5 Source: Table G.3. Note: .. not available. a. Preliminary estimate. b. Weighted average price. 365 Table G.5. Average Yield per Hectare for Tea and Sisal, 1994/95­2004/05 tons per hectare Type of crop 1994/95 1995/96 1996/97 1997/98 1998/99 1999/2000 2000/01 2001/02 2002/03 2003/04 2004/05a Tea Average 6.0 4.7 4.4 5.8 5.1 .. 5.6 5.3 6.2 5.7 5.9 Estates 9.2 7.5 7.5 9.9 8.7 .. 10.1 9.0 9.1 8.4 8.4 Smallholder 2.5 1.5 0.8 0.8 0.7 .. 0.9 1.4 3.2 3.0 3.4 Sisal/g 0.5 0.6 0.5 0.5 0.2 1.2 0.5 0.5 0.5 0.6 0.5 Source: United Republic of Tanzania, The Economic Survey (various issues). a. Preliminary estimate. b. Refers only to mature plants. 366 Table G.6. Coffee Production, 1994/95­2004/05 tons Type of coffee 1994/95 1995/96 1996/97 1997/98 1998/99 1999/2000 2000/01 2001/02 2002/03 2003/04 2004/05 Arabica 28,679 37,238 34,212 21,683 26,195 21,764 40,981 41,500 42,060 22,566 24,758 Mild 27,137 35,142 32,933 19,789 23,605 18,171 37,176 38,000 37,294 20,716 23,870 Hard 1,542 2,096 1,279 1,894 2,590 3,593 3,805 3,500 4,766 1,850 888 Robusta 7,157 9,710 9,279 9,165 13,523 1,078 17,007 12,000 34,368 16,138 9,133 Total 35,836 46,948 43,491 30,848 39,718 22,842 57,988 53,500 76,428 38,704 33,891 Source: United Republic of Tanzania, The Economic Survey (various issues). 367 APPENDIX H Employment, Labor, and Production in the Manufacturing Sector Table H.1. Estimated Employment in Manufacturing Enterprises, 2000­05 number of people engaged Industry Employed Others Total (ISIC Rev. 3 class- ification) 2000 2001a 2002a 2003b 2004b 2005b 2000 2001a 2002a 2003b 2004b 2005b 2000 2001a 2002a 2003b 2004b 2005b Food (151- 4) 34,866 41,023 40,312 41,521 43,182 45,773 264 410 409 413 421 438 35,130 41,433 40,721 41,934 43,595 46,194 Beverages (155) 4,964 4,233 4,209 4,335 4,509 4,779 23 29 29 29 30 31 4,987 4,262 4,238 4,365 4,538 4,809 370 Tobacco and ciga- rettes (160) 3,618 3,324 3,328 3,428 3,565 3,779 0 0 0 0 0 0 3,618 3,324 3,328 3,428 3,565 3,779 Textiles (171-, 181-, 191-2) 8,399 8,718 8,862 9,128 9,493 10,063 28 53 53 54 55 57 8,427 8,771 8,915 9,181 9,547 10,117 Skin and skin prod- ucts (191) 87 20 20 21 21 23 2 2 2 2 2 2 89 22 22 23 23 25 Shoes (192) 714 714 1,063 1,095 1,139 1,207 11 9 9 9 9 10 725 723 1,072 1,104 1,148 1,216 Wood products (201-2) 2,419 2,508 2,607 2,685 2,793 2,960 96 252 252 255 260 270 2,515 2,760 2,859 2,940 3,047 3,220 Paper products (210, 221- 2) 2,747 3,231 3,345 3,445 3,583 3,798 76 109 108 109 111 116 2,823 3,340 3,453 3,554 3,692 3,909 Chemicals (241-2) 3,351 4,527 4,315 4,444 4,622 4,900 50 48 47 47 48 50 3,401 4,575 4,362 4,492 4,670 4,948 Rubber products (251) 454 433 488 503 523 554 0 3 4 4 4 4 454 436 492 507 527 558 Plastic products (252) 1,224 1,564 1,585 1,633 1,698 1,800 14 18 18 18 19 19 1,238 1,582 1,603 1,651 1,716 1,818 Nonmetallic products (261-9) 1,744 1,491 1,502 1,547 1,609 1,705 22 11 11 11 11 12 1,766 1,502 1,513 1,558 1,620 1,717 Others (271­369) 4,300 5,425 5,462 5,626 5,851 6,202 191 146 150 152 155 161 4,491 5,571 5,612 5,777 6,002 6,356 Total 68,887 77,211 77,098 79,411 82,587 87,543 777 1,090 1,092 1,103 1,125 1,170 69,664 78,301 78,190 80,514 83,712 88,713 Source: United Republic of Tanzania, The Economic Survey (various issues). Note: Only industries or activities that contribute about 1 percent to value added are identified, in accordance with the 1999 and 2000 Industrial Survey. a. The data are adjusted according to the results of 1999 and 2000 Industrial Survey. b. Estimates according to 1999 and 2000 Industrial Survey. 371 Table H.2. Estimated Labor Costs in Manufacturing Enterprises, 2000­05 millions of Tanzania shillings Industry Employed Others Total (ISIC Rev. 3 classifi- cation) 2000 2001a 2002a 2003b 2004b 2005b 2000 2001a 2002a 2003b 2004b 2005b 2000 2001a 2002a 2003b 2004b 2005b Food (151- 4) 20,833 24,640 24,397 25,129 26,134 27,702 3,047 3,405 3,484 3,588 3,732 3,956 23,880 28,045 27,881 28,717 29,866 31,658 Beverages (155) 19,398 9,872 11,394 11,736 12,205 12,938 3,441 3,845 4,262 4,390 4,565 4,839 22,839 13,717 15,656 16,126 16,770 17,777 Tobacco and ciga- rettes (160) 2,059 3,698 4,211 4,337 4,511 4,781 1,204 1,346 1,389 1,431 1,488 1,577 3,263 5,044 5,600 5,768 5,999 6,359 Textiles (171-, 181-, 191-2) 4,949 3,370 4,326 4,456 4,634 4,912 540 603 653 673 699 741 5,489 3,973 4,979 5,128 5,333 5,653 Skin and skin prod- ucts (191) 23 6 6 6 6 7 1 1 1,380 1,421 1,478 1,567 24 7 1,386 1,428 1,485 1,574 Shoes (192) 616 653 738 760 791 838 176 197 216 222 231 245 792 850 954 983 1,022 1,083 Wood products (201-2) 1,707 1,210 1,293 1,332 1,385 1,468 208 232 242 249 259 275 1,915 1,442 1,535 1,581 1,644 1,743 Paper products (210, 221- 2) 3,274 4,006 4,432 4,565 4,748 5,032 876 979 1,103 1,136 1,181 1,252 4,150 4,985 5,535 5,701 5,929 6,285 Chemicals (241-2) 3,209 6,359 7,004 7,214 7,503 7,953 992 1,109 1,196 1,232 1,281 1,358 4,201 7,468 8,200 8,446 8,784 9,311 Rubber products (251) 514 545 61 68 82 84 88 93 61 68 82 84 602 638 Plastic products (252) 2,569 1,473 1,683 1,733 1,289 1,366 61 68 79 81 85 90 2,630 1,541 1,762 1,815 1,373 1,456 Nonmetallic products (261-9) 5,362 5,774 6,527 6,723 6,992 7,411 2,078 2,322 2,976 3,065 3,188 3,379 7,440 8,096 9,503 9,788 10,179 10,790 Others (271­369) 9,160 20,699 21,865 22,521 23,422 24,827 1,471 1,644 1,153 1,188 1,235 1,309 10,631 22,343 23,018 23,708 24,657 26,136 Total 73,159 81,760 87,876 90,512 94,132 99,781 14,155 15,819 18,215 18,761 19,511 20,681 87,314 97,579 106,091 109,273 113,643 120,462 Source: United Republic of Tanzania, The Economic Survey (various issues). Note: Only industries or activities that contribute about 1 percent to value added are identified, in accordance with the 1999 and 2000 Industrial Survey. a. The data are adjusted according to the results of 1999 and 2000 Industrial Survey. b. Estimates according to 1999 and 2000 Industrial Survey. 373 Table H.3. Production of Selected Manufactured Commodities, 1995­2005 Type of commodity 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004a 2005b Biscuits and macaroni (tons) 246 71 288 805 611 891 1,215 2,284 5,906 10,214 10,862 Wheat flour (thousand tons) 11,612 33,998 77,598 87,669 144,693 162,634 180,098 219,118 355,616 338,076 367,493 Konyagi (spirit) (thousand liters) 2,011 1,832 1,849 1,994 1,630 2,214 2,287 2,937 3,670 4,105 4,489 Beer (million liters) 89,301 125,074 148,340 170,700 167,478 183,003 175,649 175,870 194,100 202,628 216,604 Chibuku (million liters) 11,319 14,031 13,680 11,993 12,392 17,041 18,750 19,400 14,825 10,119 11,106 Cigarettes (million pieces) 3,699 3,733 4,710 3,933 3,371 3,745 3,491 3,778 3,920 4,219 4,435 Textiles (thousand square meters) 31,201 33,178 41,706 45,546 49,757 73,566 84,325 106,305 116,714 111,637 102,532 Sisal ropes (tons) 17,323 11,178 4,919 4,329 3,253 3,900 4,547 5,901 6,839 5,161 6,018 Fish nets (tons) 108 125 70 35 24 42 57 30 41 260 274 Plywood (thousand tons) 284 315 115 .. 1,953 568 450 304 32,609 40,248 36,509 Pyrethrum (thousand tons) 21 11 3 9 17 44 71 36 16 23 25 Paints (thousand liters) 3,228 5,205 4,986 4,943 12,903 7,085 9,034 13,564 16,842 16,621 16,608 Petroleum products (thousand 374 tons) 398 336 313 312 287 177 .. .. .. .. .. Cement (thousand tons) 739 726 621 778 833 833 900 1,026 1,186 1,281 1,375 Iron sheets (tons) 2,518 7,733 12,498 9,522 8,982 11,182 16,340 25,418 38,794 41,710 41,299 Corrugated iron (tons) 18,142 6,422 15,218 14,918 23,028 25,046 25,937 31,742 31,018 29,573 32,037 Aluminum (tons) 1,158 360 117 180 187 133 137 141 199 171 183 Radios (thousand units) 76 56 56 15 .. .. .. .. .. .. .. Battery (million units) 59 66 43 46 43 44 39 42 43 74 83 Source: United Republic of Tanzania, The Economic Survey (various issues). Note: .. = not available. a. Adjusted figure. b. Preliminary estimate. Table H.4. Mineral Production, 1995­2005 Type of mineral 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005a Quantity Diamond (thousand carats) 50 13 12 98 235 354 254 213 237 304 220 Salt (thousand tons) 105 87 73 75 36 70 65 71 59 57 135 Phosphate (thousand tons) 7 1 2 1 7 5 4 1 4 7 7 Gemstone (thousand kilograms) 111 142 509 49 95 151 97 196 1,532 1,614 1,937 Gold (thousand kilograms) 3 3 3 4 5 15 30 34 48 48 52 Gypsum (thousand tons) 42 55 46 59 21 60 72 73 33 59 63 Coal (thousand tons) 43 52 28 45 75 79 78 79 55 65 75 Limestone (thousand tons) 1,062 1,200 1,282 1,181 1,241 1,500 2,269 2,857 1,206 1,391 2,780 Pozolana (thousand tons) .. .. .. .. 2 57 41 52 106 153 163 Silver ore (thousand kilograms) .. .. .. .. .. .. 7 8 8 13 13 Copper (thousand pounds) .. .. .. .. .. .. 5,832 9,310 8,191 9,348 7,633 Bauxite (thousand tons) .. .. .. .. .. .. .. .. .. .. 2 375 Production index (1991 = 100) Diamond 68.8 17.6 17.1 135.8 326.4 492.2 353.2 296.5 328.6 422.1 305.1 Salt 340.9 281.5 235.4 243.5 116.5 227.3 211.0 231.2 191.5 185.3 439.6 Phosphate 278.6 29.9 88.3 59.6 302.1 212.5 166.7 49.3 155.8 273.8 295.7 Gemstone 0.2 0.2 0.9 0.1 0.2 0.3 0.2 0.3 2.6 2.7 3.2 Gold 0.1 0.1 0.1 0.1 0.1 0.3 0.6 0.6 0.9 0.9 1.0 Gypsum 792.5 1,045.3 873.6 1,115.1 399.9 1,132.1 1,358.5 1,377.4 627.0 1,117.6 1,195.8 Coal 130.1 156.6 85.7 135.7 225.9 238.4 234.2 237.8 164.4 195.8 225.2 Limestone 191.9 216.8 231.7 213.4 224.3 271.1 410.0 516.3 217.9 251.4 502.3 Pozolana .. .. .. .. .. .. .. .. .. .. .. Silver ore .. .. .. .. .. .. .. .. .. .. .. Copper .. .. .. .. .. .. .. .. .. .. .. Bauxite .. .. .. .. .. .. .. .. .. .. .. Source: United Republic of Tanzania, The Economic Survey (various issues). a. Preliminary estimate. APPENDIX I Consumer Prices and Cost of Living Table I.1. National Consumer Price Index, 1996­2005 Weights 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Retail price index (2001 = 100) General index 100.0 63.5 73.8 83.2 89.8 95.1 100.0 101.0 104.5 108.9 113.6 Food 55.9 60.0 70.7 81.1 88.2 94.2 100.0 101.0 105.6 111.8 118.4 Drinks and tobacco 6.9 82.0 86.6 92.0 94.8 96.4 100.0 100.5 102.0 101.5 103.7 Rents 1.4 69.3 79.9 86.9 91.2 96.3 100.0 100.8 102.8 105.1 109.2 Fuel, light, and water 8.5 60.7 72.4 82.8 90.9 97.2 100.0 101.1 107.1 109.6 117.9 Clothing and footwear 6.4 72.6 82.0 85.0 94.0 96.4 100.0 102.1 104.1 109.5 104.6 Furniture and utensils 2.1 75.8 84.8 94.9 98.0 98.9 100.0 101.0 103.1 103.3 102.2 Household operations 2.1 87.2 93.5 96.3 99.9 101.8 100.0 100.6 101.7 102.6 103.0 Personal care and health 2.1 81.0 93.1 96.4 98.6 99.5 100.0 100.9 103.8 104.4 108.0 Recreation and entertainment 0.8 76.0 83.5 89.0 95.1 98.3 100.0 100.7 102.8 104.9 103.4 Transportation 9.7 67.7 82.8 90.4 94.9 98.7 100.0 100.4 101.7 102.7 107.6 Education 2.6 82.2 91.2 95.8 98.4 99.4 100.0 100.4 102.6 103.2 103.2 Miscellaneous goods and services 1.5 86.4 94.3 98.8 97.5 98.7 100.0 100.6 102.8 102.3 100.9 Annual retail price change (%) 378 General index 16.1 12.8 7.9 5.9 5.2 1.0 3.5 4.2 4.4 Food 17.8 14.7 8.8 6.8 6.2 1.0 4.6 5.9 5.9 Drinks and tobacco 5.6 6.2 3.0 1.7 3.7 0.5 0.0 1.0 2.0 Rents 15.3 8.8 4.9 5.6 3.8 0.8 2.0 2.2 3.9 Fuel, light, and water 19.3 14.4 9.8 6.9 2.9 1.1 5.9 2.3 7.6 Clothing and footwear 12.9 3.7 10.6 2.6 3.7 2.1 2.0 5.2 ­4.5 Furniture and utensils 11.9 11.9 3.3 0.9 1.1 1.0 2.1 0.2 ­1.1 Household operations 7.2 3.0 3.7 1.9 ­1.8 0.6 1.1 0.9 0.4 Personal care and health 14.9 3.5 2.3 0.9 0.5 0.9 2.9 0.6 3.4 Recreation and entertainment 9.9 6.6 6.9 3.4 1.7 0.7 2.1 2.0 ­1.4 Transportation 22.3 9.2 5.0 4.0 1.3 0.4 1.3 1.0 4.8 Education 10.9 5.0 2.7 1.0 0.6 0.4 2.2 0.6 0.0 Miscellaneous goods and services 9.1 4.8 ­1.3 1.2 1.3 0.6 2.2 ­0.5 ­1.4 Source: Tanzania authorities. Table I.2. Dar es Salaam Cost of Living Index: Middle-Income Group, 1996­2005 Weight 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Living cost index (2001 = 100) General index 100.0 81.8 86.9 91.2 92.0 94.3 100.0 108.2 115.8 114.6 123.7 Food 55.9 90.4 94.7 98.9 97.7 97.0 100.0 110.2 117.5 116.3 127.8 Nonfood 44.1 76.9 82.6 84.9 87.4 92.4 100.0 109.0 117.2 110.6 116.1 Annual living cost change (%) General index 6.2 4.9 0.9 2.5 6.0 8.2 7.0 ­1.0 7.9 Food 4.8 4.4 ­1.2 ­0.7 3.1 10.2 6.6 ­1.0 9.9 Nonfood 7.4 2.8 2.9 5.7 8.2 9.0 7.5 ­5.6 5.0 Source: Tanzania authorities. 379 Table I.3. Dar es Salaam Retail Price Index, 1996­2005 Weights 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Retail price index (2001 = 100) General index 100.0 78.7 84.1 89.8 90.6 92.2 100.0 107.4 112.7 116.8 126.5 Food 55.9 87.7 91.9 97.8 96.6 94.6 100.0 110.5 114.2 117.1 125.7 Drinks and tobacco 6.9 76.4 82.3 85.0 89.2 96.4 100.0 107.2 120.0 123.2 120.0 Fuel, light, and water 8.5 50.7 62.6 74.1 82.1 94.0 100.0 106.6 123.1 133.2 162.3 Clothing and footwear 6.4 68.2 75.3 76.1 87.2 87.5 100.0 88.2 87.9 111.9 108.7 Rents 1.4 44.8 48.7 54.6 56.4 64.3 100.0 100.0 100.0 111.9 113.6 Furniture and utensils 2.1 74.8 85.1 89.6 92.5 92.6 100.0 107.7 114.6 113.0 115.7 Household operations 2.1 71.4 70.6 74.6 74.2 74.1 100.0 73.2 74.3 102.2 110.0 Personal care and health 2.1 84.8 88.9 83.9 83.4 84.4 100.0 92.3 71.9 118.5 113.6 Transportation 9.7 80.0 102.4 102.5 101.2 102.4 100.0 104.2 104.8 100.9 109.8 Recreation and entertainment 0.8 95.2 106.5 108.3 108.4 106.5 100.0 103.2 96.2 103.7 110.4 Education 2.6 80.7 74.6 80.9 69.8 85.0 100.0 125.6 180.2 111.2 115.0 Miscellaneous goods and services 1.5 107.9 95.1 87.9 89.3 91.2 100.0 121.9 150.3 103.4 121.1 Annual retail price changes (%) 380 General index 6.9 6.8 0.9 1.8 8.5 7.4 4.9 3.6 8.3 Food 4.8 6.4 ­1.2 ­2.1 5.7 10.5 3.3 2.5 7.3 Drinks and tobacco 7.7 3.3 4.9 8.1 3.7 7.2 11.9 2.7 ­2.6 Fuel, light, and water 23.5 18.4 10.8 14.5 6.4 6.6 15.5 8.2 21.8 Clothing and footwear 10.4 1.1 14.6 0.3 14.3 ­11.8 ­0.3 27.3 ­2.9 Rents 8.7 12.1 3.3 14.0 55.5 0.0 0.0 11.9 1.5 Household operations 13.8 5.3 3.2 0.1 8.0 7.7 6.4 ­1.4 2.4 Personal care and health ­1.1 5.7 ­0.5 ­0.1 35.0 ­26.8 1.5 37.6 7.6 Transportation 4.8 ­5.6 ­0.6 1.2 18.5 ­7.7 ­22.1 64.8 ­4.1 Recreation and entertainment 28.0 0.1 ­1.3 1.2 ­2.3 4.2 0.6 ­3.7 8.8 Education 11.9 1.7 0.1 ­1.8 ­6.1 3.2 ­6.8 7.8 6.5 Miscellaneous goods and services ­7.6 8.4 ­13.7 21.8 17.6 25.6 43.5 ­38.3 3.4 Source: Tanzania authorities. Report No.: 39021-TZ Sustaining and Sharing Economic Growth in Tanzania: Summary of Main Findings and Recommendations Country Economic Memorandum and Poverty Assessment March 1, 2007 The United Republic of Tanzania Ministry of Planning, Economy, and Empowerment The World Bank Poverty Reduction and Economic Management Unit 2 Africa Region GOVERNMENT FISCAL YEAR July 1 ­ June 30 CURRENCY EQUIVALENTS (Exchange Rate Effective as of January 18, 2007) Currency Unit Tanzania Shilling (TSh) US$1.00 TSh 1,2851.1 Weights and Measures Metric System ACRONYMS AND ABBREVIATIONS ASDP Agricultural Sector Development Program NGO Non Governmental Organization ASDS Agricultural Sector Development Strategy NIDP National Irrigation Development Plan BCI Business Competitiveness Index NLP National Land Policy CDTT Centre for the Development and Transfer of NMB National Micro Finance Bank Technology NSGRP National Strategy for Growth and Reduction of CGE Computable General Equilibrium Poverty CORMA Client-Oriented Research Management ODA Official Development Assistance Approach OECD Organization for Economic Cooperation and COSTECH Tanzania Commission for Science and Development Technology O&M Operations and Maintenance COWI Consulting Engineers and Plans PEDP Primary Education Development Program CPIA Country Policy and Institutional Assessment PFA Private Fisheries Agreement DADG District Agricultural Development Grant PPA Participatory Poverty Assessment DADP District Agricultural Development Plan PPP Public-Private Partnerships DHS Demographic and Health Survey PRSP Poverty Reduction Strategy Paper EAC East African Community PSLE Primary School Leaving Examination EEZ Exclusive Economic Zone R&D Research and Development EIA Environmental Impact Assessment REER Real Effective Exchange Rate FDI Foreign Direct Investment REPOA Research on Poverty Alleviation FSDT Financial Sector Deepening Trust ROSCAs Rotating Saving Schemes GCI Growth Competitiveness Index SACCOs Savings and Credit Cooperatives GDP Gross Domestic Product SADC South African Development Community GERs Gross Enrollment Ratios SEDP Secondary Education Development Program GLS Grey Leaf Spot SME Small and Medium Enterprise GNI Gross National Income TAI Technology Achievement Index HBS Household Budget Survey TANESCO Tanzania Electricity Supply Corporation ICT Information and Communication Technologies TBS Tanzania Bureau of Standards ID Irrigation Department TCC Tanzanian Telecommunications Commission IDE International Development Enterprises TCRA Tanzania Communications Regulatory Authority ILD Instituto de Libertad y Democracia TRCHS Tanzania Reproductive and Child Health IMF Innovation Multipurpose Facility Survey IPI Institution of Production and Innovations TTCL Tanzanian Telecommunications Company KAM Knowledge Assessment Methodology TVET Technical/Vocational Education and Training KEI Knowledge Economy Index UNCTAD United Nations Conference on Trade and KHDS Kagera Health and Demographic Survey Development MAFC Ministry of Agriculture, Food Security, and UNDP United Nations Development Programme Cooperatives URT United Republic of Tanzania MBICU Mbinga Cooperative Union VET Vocational Education and Training MDG Millennium Development Goal VETA Vocational Training and Education Authority MoH Ministry of Health WUA Water User Associations MNRT Ministry of Natural Resources and Tourism ZARDEFs Zonal Agricultural Research and Development MVIWATA Mtandao wa Vikundi vya Wakulima Tanzania Funds NDC National Development Corporation ZIUs Zonal Irrigation Units NER Net Enrollment Ratio Ag. Vice President: Hartwig Schafer Country Director: Judy O'Connor Sector Manager: Kathie Krumm Task Team Leader: Robert Johann Utz Sustaining and Sharing Growth in Tanzania i Table of Contents ACKNOWLEDGEMENTS ..................................................................................................................iv SUMMARY............................................................................................................................................vi 1. POVERTY REDUCTION AND GROWTH ­ RECENT PERFORMANCE AND PROSPECTS.....................................................................................................................................1 A Decade of Structural Reforms, Macro-economic Stability and Economic Growth ...........1 Economic Growth, Poverty, and Inequality ..............................................................................10 Outlook on Growth and Poverty Reduction ..............................................................................14 Progress Towards the Millennium Development Goals ...........................................................16 2. ENHANCING INTERNATIONAL COMPETITIVENESS AND ACCELERATING DIVERSIFICATION.......................................................................................................................19 Fostering Growth, Export Competitiveness, and Employment in Tanzania's Manufacturing Sector ...............................................................................................................................................19 Improving the Business Environment ........................................................................................20 Fostering Innovation, Productivity, and Technological Change............................................28 3. MAKING GROWTH PRO-POOR.................................................................................................38 Supporting Agriculture Sector Development ............................................................................38 Harnessing Natural Resources for Sustainable Growth ............................................................47 Supporting The Informal Sector and Micro, Small, and Medium Size Enterprises.............54 Strengthening the Capacity of the Poor to Participate and Contribute to Economic Growth ..........................................................................................................................................................60 4. MANAGING POLICIES AND RESOURCES FOR SUSTAINED SHARED GROWTH......65 Strengthening Institutions to Steer the Development and Implementation of Tanzania's Strategy for Growth and Poverty Reduction. ............................................................................65 Scaling up Public Spending for the Implementation of the NSGRP ......................................67 REFERENCES ......................................................................................................................................73 ANNEX 1: LIST OF BACKGROUND STUDIES ...........................................................................77 Sustaining and Sharing Growth in Tanzania ii LIST OF BOXES Box 1. Overview of Structural Reforms in Tanzania.....................................................................2 Box 2. Government Spending and Economic Growth..................................................................5 Box 3. Benchmarking Tanzania in the Global Context .................................................................8 Box 4. Review of Experience with Export Processing Zones (EPZs). ......................................10 Box 5. Underlying principles for the design of an agricultural technology system................41 Box 6. Examples of voluntary formalization................................................................................58 LIST OF FIGURES Figure 1. Growth rates of GDP incl. and excl. government spending, 1990 - 2005 ................4 Figure 2. Contribution of public and private expenditure to economic growth, 1990 - 2005 5 Figure 3. Decomposition of economic growth per worker into contribution of human and physical capital accumulation and total factor productivity (TFP), 1985-2005. ........................6 Figure 4. Simulated changes in poverty, 1992-2002 ..................................................................12 Figure 5. Projected changes in income poverty and malnutrition .............................................17 Figure 6. Cost of inefficiencies in business environment as percent of sales, various countries ..............................................................................................................................................................21 Figure 7. Low levels of infrastructure development constrain economic growth ...................22 Figure 8. Domestic credit and interest rates, 1990-2005.............................................................24 Figure 9. Savings and Investment (% of GDP), 1990-2005........................................................24 Figure 10. Requests for bribes are especially common during tax inspections........................27 Figure 11. Adult Literacy Rates, 1960-2000 ................................................................................29 Figure 12. Average Years of Schooling, 1970-2002 ..................................................................29 Figure 13. Marginal social returns per year of education based on integrated labor force survey...................................................................................................................................................30 Figure 14. Predicted earnings in manufacturing sector based on manufacturing firm surveys ..............................................................................................................................................................31 Figure 15. Labor productivity levels in Tanzania and comparators..........................................39 Figure 16. Cost components of marketing margins vary significantly with crop and location ..............................................................................................................................................................44 Figure 17. Size of the informal economy for selected countries (as % of GNI)......................54 Figure 18. Median value added per worker ..................................................................................57 Figure 19. Changes in the distribution of access to education, rural Kilimanjaro ...................61 Figure 20. Multilateral credit disbursements and Debt Sustainability: Net Present Value of Debt-to-Export Ratio, 2006­26 .......................................................................................................70 LIST OF TABLES Table 1. Elements of a growth strategy...................................................................................... xviii Table 2. Key Economic Indicators, 1995-2005 .............................................................................1 Table 3. Sources of Growth (production), 1990 - 2005 ...............................................................3 Table 4. Sources of Growth (expenditure), 1990 - 2005..............................................................3 Table 5. Zonal Poverty Status .........................................................................................................11 Table 6. Decomposition of change in poverty .............................................................................13 Table 7. Policy based growth projection .......................................................................................14 Table 8. Scenarios for economic growth and structural transformation ...................................15 Table 9. MDG baseline, most recent estimate and target.............................................................16 Table 10. Estimated size of the informal economy (% of GDP), various countries and regions ..............................................................................................................................................................26 Table 11. Role of the Public Sector in Fostering Innovation......................................................32 Table 12. ICT Indicators (per 1,000 people), 2002.....................................................................36 Sustaining and Sharing Growth in Tanzania iii Table 13. Land use and potential for land expansion (mid 1990s) ...........................................39 Table 14. Institutional Framework for Sustainable Development of Smallholder Irrigation Systems ................................................................................................................................................43 Table 15. Typology of forms of enterprise in Tanzania .............................................................55 Table 16. Increase in per capita consumption relative to households headed by individuals with no-education...............................................................................................................................60 Table 17. Number of days missed due to illness, by quintile .....................................................62 Table 18. Differences in health outcomes by quintile .................................................................63 Sustaining and Sharing Growth in Tanzania iv Acknowledgments This report was prepared in collaboration between the Tanzanian Government, Ministry of Planning, Economy, and Empowerment (MPEE) and the World Bank with support from the African Development Bank and various local and international researchers. It was prepared under the overall supervision of Dr. Enos Bukuku (Permanent Secretary, Ministry of Infrastructure Development, previously Permanent Secretary, MPEE), Charles Mutalemwa (Permanent Secretary, MPEE), Judy O'Connor (Country Director for Uganda and Tanzania), and Kathie Krumm (Sector Manager, Poverty Reduction and Economic Management Unit for East Africa), who provided substantive inputs, comments, and support at all stages of the preparation process. The task manager and principal author of the report is Robert Utz. The Government team was initially led by Arthur Mwakapugi (Permanent Secretar, Ministry of Energy and Mining, previously Director for Macroeconomics, MPEE), followed by Laston Msongole (Director of Macroeconomics, MPEE). Substantive inputs and background papers were prepared by Johannes Hoogeveen, (poverty, nutrition, and strengthening the capacity of the poor to contribute to growth), Louise Fox, and Marianne Simonsen (poverty assessment), Meera Shekar (nutrition), Henry Gordon (agriculture), Anuja Utz and Jean-Eric Aubert (education, innovation, and ICT), Michael Wong (business environment), Ravi Ruparel (financial sector development), Vandana Chandra, Pooja Kacker, Ying Li (manufacturing sector analysis), Philip Mpango (spatial dimensions of growth), Allister Moon (institutional reform), Jack Ruitenbeek, Cynthia Carter (natural resource management) and Emmanuel Mungunasi (data and statistical annexes). Peter Mwanakatwe (African Development Bank) contributed the analysis of infrastructure. Mary-Anne Mwakangale and Arlette Sourou provided dedicated logistical support. Arlette Sourou was also responsible for word processing and physical production of the report. The Governments of Austria, Denmark, Netherlands and Sweden financed the preparation of background papers by Annabella Skof (consultant - tourism), Kirsten Pflieger (COWI consultants ­ natural resource based growth), Meine Pieter van Dijk (UNESCO Institute for Water Education ­ Analysis of the role of urban areas), and Jorgen Levin (consultant - CGE analysis). InfoDev financed a study on growth, competitiveness, and ICT carried out by OTF consultants. Annex 1 presents the list of background studies prepared for this report by local and international researchers and by World Bank staff. Insightful and challenging comments were provided at various stages of the preparation process by peer reviewers Benno Ndulu (World Bank), Josephat Kweka (ESRF), Erik Thorbecke, Steven Younger, and David Sahn (Cornell University), and Dani Rodrik (Harvard University). During the consultations held in Dar es Salaam, Haidari Amani (ESRF), Brian van Arkadie (ESRF), Amon Mbelle (ERB), Peter Noni (BoT), Adolf Mkenda (University of Dar es Salaam), Robert Mbelle (University of Dar es Salaam) served as discussants of the draft background papers. Detailed comments were also provided by members of the World Bank Country Team, including Keith Hinchliffe (education), Dieter Schelling (transport), Karen Rasmussen, Arun Sanghvi, and Duncan Reynold (energy), Mavis Ampah (ICT), Mathew Glasser (decentralization and local government), and Denyse Morin (institutional reforms). Sustaining and Sharing Growth in Tanzania v The CEM benefited from the participation in the "growth path" project led by Roberto Zagha (World Bank) in cooperation with Harvard University that helped to sharpen the growth diagnostic. The preparation of the report included several rounds of consultations in Tanzania organized by the Ministry of Planning, Economy, and Empowerment. Initial consultations in September 2003 and July 2004 served to define and agree on the scope and focus of the study as well as on collaborative arrangements. The main mission took place in November 2004 and included field visits to Kigoma, Lindi, and Mtwara. In March 2005, a series of workshops in Dodoma, Morogoro, Moshi, and Dar es Salaam were organized to obtain feedback and inputs on the draft background papers prior to the drafting of the main report. At that stage, the Ministry of Planning, Economy, and Empowerment organized a review meeting with Permanent Secretaries and senior officials from a large number of ministries for a briefing on the consultations and the discussion of emerging main messages and recommendations. The team would like to express its sincere gratitude to all who have provided valuable comments and inputs during these consultations. The report is presented in two volumes. Volume I summarizes the main findings and recommendations. Volume II contains the main report. Note: Government is currently revising its National Accounts data, including changing the base year for constant price data series from 1992 to 2001. The revised National Accounts data are expected to be released in 2007. Sustaining and Sharing Growth in Tanzania vi Summary Tanzania's National Strategy for Growth and Reduction of Poverty (NSGRP) sets an ambitious target of 6 to 8 percent annual economic growth to achieve rapid reduction in poverty. This report focuses on three issues that are central to the success of Tanzania's poverty reduction efforts: · What factors explain Tanzania's recent acceleration in economic growth? · Has the accelerated economic growth translated into reduced poverty? · What must be done to sustain economic growth that is pro-poor? The report presents evidence from the macroeconomic, sectoral, and firm and household levels that shed light on these questions. This summary provides an overview of the main findings and recommendations. What Factors Explain Tanzania's Recent Acceleration in Economic Growth? The average annual growth of Tanzania's gross domestic product (GDP) of 6.0 percent during 2000 to 2005 has been high, not only compared with its own historical growth performance but also compared with international growth rates. Growth rates increased across all sectors, with industry growing by 8.7 percent, services by 5.9 percent, and agriculture by 4.8 percent during the same period. Mining (growth rate of 15.2 percent), construction (10 percent), manufacturing (7.0 percent), and trade hotels and restaurants (6.9 percent) were the fastest-growing sub-sectors. The contribution of the various sectors to growth, which depends on both the growth rate of the sector and its share in the economy, shows that agriculture contributed 2.3 percentage points, services 2.1 percentage points, and industry 1.6 percentage points of the average annual growth of 6.0 percent during 2000 to 2005. The analysis of the sectoral contributions to the increase in the average GDP growth rate from 2.5 percent during 1990 to 1994 to 6.0 percent during 2000 to 2005 confirms that growth accelerated in all sectors. Growth in the service sector contributed 1.4 percentage points to the increase, industry 1.3 percentage points, and agriculture 0.8 percentage points. The implementation of a comprehensive set of macroeconomic and structural reforms laid the foundation for the recent growth acceleration. These reforms enhanced the incentives for private sector activities and led to improved efficiency of resource allocation and use in the economy. The domestic and foreign private sectors as well as Tanzania's development partners reacted to the improvements in the economic and incentive regime in a variety of ways that explain the increase in economic growth. A central element of Tanzania's recent growth performance is large inflows of private and public capital that were triggered by the reforms undertaken by government. The transition of Tanzania to a market economy began in the mid-1980s with an initial focus on the liberalization of the economy through the removal of constraints on private sector activities and the abolition of controls on prices and exchange and interest rates. The reforms also included a restructuring of the public sector and an ambitious privatization program. In Sustaining and Sharing Growth in Tanzania vii the mid-1990s, the reform agenda was augmented by a strong focus on macroeconomic stability and the quality of public financial management. Initially, this effort involved sharp cuts in government expenditures to minimize the government's domestic and nonconcessional borrowing. These cuts served as the basis for a prudent monetary policy that reduced the rate of inflation to well below 10 percent. Subsequently, reform efforts focused on improving Tanzania's tax system and public financial management to improve allocative and operational efficiency of public expenditures and to minimize resource leakages. An important result of prudent monetary and fiscal policy, combined with financial sector reforms, is the recovery of credit to the private sector which grew by more than 30 percent annually in recent years. The environment for economic growth is thus vastly improved, and current government efforts are targeting higher levels of investment in human capital and physical infrastructure, improvements in the business environment, and strengthening of government capacity. The intensification of reforms since 1995 and improvements in the business environment, as well as sector-specific reforms--especially in the mining sector--have triggered an increase in foreign direct investment (FDI) and aid inflows. FDI has increased rapidly since the mid- 1990s and reached about US$542 million or 5 percent of GDP by 1999, partly driven by large investments in mining and privatization-related investments. Following the completion of major investments in the mining sector and the major privatizations, FDI declined to US$473 million or 3.9 percent of GDP by 2005, a level that is still high in comparison with that of most other African countries. The sectors that received the bulk of the FDI showed the highest growth rates, including mining, manufacturing, and trade and tourism, which together attracted about 75 percent of FDI during 1999 to 2001. The reforms implemented by the government also triggered a continuous increase in aid inflows that, together with improved domestic revenue collection, supported the increase in government spending from 16 percent of GDP in 1999/2000 to 26 percent in 2005/06. National accounting statistics suggest that this increase in government spending contributed significantly to the acceleration in economic growth. In the short term, the increased demand for goods and services by the government led to increased use of available capacity. For example, the rehabilitation and expansion of administrative, economic, and social infrastructure are reflected in the fast growth of the construction sector by about 10 percent annually during 2000 to 2005. Fast growth in the service sector is also partly related to increased government expenditures. In addition to these direct effects of increased government spending, traditional multiplier effects translate increases in government spending into increased demand for goods and services in all sectors. In the medium to long term, if government spending contributes effectively to the building of human capital and the expansion of economic infrastructure, then sustained levels of increased government spending have the potential to expand the productive capacity of the economy. A noteworthy development is the rapid growth of the informal sector--particularly in Dar es Salaam--as the result of various factors. These include the liberalization of the economy, the tolerance of many informal sector activities that were previously illegal, the need for laid-off government workers and migrants to generate new income-earning opportunities, and the increased demand for informal sector products and services as a trickle-down effect from growth in the formal economy. Another significant economic development during the past decade has been the rapid expansion of mining and gold exports, whose share in total exports increased from 4 percent in 1998 to 56 percent in 2005. However, the contribution of mining to overall growth was only 0.4 percentage points, reflecting the relatively small size of the mining sector, as well as the high import dependence of the sector for machinery and its very limited domestic backward and forward links. Aside from gold, fish, and tourism, the value of exports Sustaining and Sharing Growth in Tanzania viii remains low and volatile. Between 1995 and 2001, the real effective exchange rate appreciated by almost 50 percent and then returned to its 1995 level. The real appreciation reduced the competitiveness of Tanzania's tradables sector and merchandise exports declined between 1995 and 2001. Despite a subsequent recovery in parallel to the recent real depreciation, to date, exports other than gold and fish have played a relatively small role as a dynamic source of growth and have seen little diversification. Thus, a key challenge for the Tanzanian economy is to strengthen its export competitiveness. Doing so would ensure that, aside from the dynamic growth effects of a strong export sector, exports will provide an important demand stimulus for the economy, especially because the scope for continued increases of government spending as the primary demand stimulus is clearly limited. The analysis of factor inputs suggests that the acceleration in economic growth is primarily due to an increase in cultivated land in the agriculture sector and increased factor productivity for the other sectors. The increase in total factor productivity reflects both increased capacity use in response to increased aggregate demand and economic efficiency gains in the wake of the removal of economic distortions. Innovation and technological change have so far played only small roles in improving Tanzania's total factor productivity, mainly in the form of FDI but also as some encouraging innovations emerging from the agricultural research system. At the firm level, there is some evidence that the structural reforms have resulted in a more dynamic and competitive private sector. Increased competition in the private sector is evidenced by an increasing number of firms exiting and entering the market. The fact that firms entering the market are typically more competitive than those that exit is an important driver of the increase in total factor productivity registered at the aggregate level. Public investment has recovered from an average of about 3 percent of GDP during the late 1990s to about 8 percent of GDP in recent years. The analysis of public investment suggests, however, that only about one-third of it was used on public infrastructure such as roads or electricity, while the remainder was devoted to the rehabilitation and expansion of administrative and social infrastructure. Private sector investment had been stagnant at about 11 percent until 2002 but increased to 14 percent by 2005, reflecting increased investor confidence in response to sustained implementation of investor friendly reforms and increased demand. Although the contribution of human capital accumulation to economic growth has been relatively small, the recent increases in school enrollment can be expected to be reflected in higher economic growth in the future. Drawing on the review of Tanzania's recent growth performance, the report assesses the prospects for sustained high growth and the key challenges that need to be addressed. Policy- based growth projections suggest that growth of 6 to 8 percent per year is feasible. However, some of the factors behind the recent growth acceleration are unlikely to be sustainable in the medium to long term. The demand-side impulses of foreign aid and government spending depend on ever-increasing amounts of aid and government spending. There is also a clear limit to the extent that agricultural production can be increased solely by increasing the land under cultivation. Signs of environmental and social stress (especially between pastoralists and agriculturalists) of increased land use already exist in some areas of Tanzania. Similarly, the effect of reform-induced efficiency gains on economic growth will diminish when the higher level of efficiency has been reached. Thus, for Tanzania to achieve sustained high growth, increases in government spending and expansion of land under cultivation need to be gradually replaced by increased productivity, saving, and investment by the private sector as primary drivers of growth. Sustained economic growth will depend on the ability of the economy to diversify and to increase its international competitiveness. Diversification requires efforts both to enhance the capacity to innovate and to find new areas of economic activity where Tanzanian enterprises can Sustaining and Sharing Growth in Tanzania ix successfully compete. Enhancing international competitiveness requires measures that enhance productivity and reduce the cost of doing business at the microeconomic level and macroeconomic policies that ensure a competitive exchange rate as well as interest rates and access to capital that are not distorted by high public demand for funds. Has the Accelerated Economic Growth Translated into Reduced Poverty? Sustained economic growth is critical to achieving progress in poverty reduction. The mechanisms through which the poor contribute to and participate in economic growth include the following: · Increased incomes from the main sources of livelihood of the poor; · New income-generating opportunities for the poor; · Reduced vulnerability to shocks that affect the incomes of the poor; · Increased government revenue for pro-poor expenditures; and · Increased private transfers and strengthened social safety nets. In addition, the report examines the effectiveness of measures that support the poor in efforts to accumulate human and physical capital, which would enhance their prospects of contributing to economic growth. Modest per capita GDP growth rates during the early and mid-1990s resulted in equally modest poverty reduction. In 2001, government estimates show 35 percent of the population living in poverty. The potential effect of the recent GDP growth acceleration has not yet been captured in available poverty data. Ownership of assets such as improved housing, radios, and bicycles by the poor has also increased. The expansion of access to free primary education has also clearly benefited the poor. The analysis of growth incidence suggests that expenditures of all income groups grew at about the same pace, probably because growth in agriculture, which is the source of income for most of the poor, was similar to growth in other sectors during 1991 to 2000. Since 2000, growth in the industry and service sectors has been higher than in the agriculture sector, which may have caused an increase in inequality. The Household Budget Survey data show large regional differences in poverty reduction. Although poverty dropped from 28.1 percent to 17.6 percent in Dar es Salaam, in other urban areas poverty declined only from 28.7 percent to 26 percent and in rural areas from 40.8 percent to 38.7 percent. The faster pace of poverty reduction in Dar es Salaam reflects Tanzania's pattern of growth. In particular, Dar es Salaam accounts for about 50 percent of the FDI stock and flows, and as the seat of central government and most donor agencies, it also benefits disproportionately from the increase in aid inflows. Although central government expenditures increased from 18 percent to 25.6 percent of GDP between 2000 and 2005, transfers to local governments increased only from 2.9 percent to 3.3 percent of GDP during that period. Growth in the formal sector in Dar es Salaam also supported an increase in the size and incomes in the informal sector, which contributed significantly to poverty reduction during the period from 1991/92 to 2000/01. More than 80 percent of Tanzania's poor derive their livelihoods from agriculture. Between 1991 and 2000, the agriculture sector grew by an average of 3.5 percent, which suggests per capita growth of less than 1 percent. The increase in per capita expenditure by farm households is equally modest at 7.3 percent during the period from 1991/92 to 2000/01. Nonetheless, because most of the poor derive their livelihood from agriculture, this modest increase explains more than half of the total decline in poverty observed during that period. Sustaining and Sharing Growth in Tanzania x Between 2000 and 2005, growth in the agriculture sector accelerated to an average of 4.8 percent annually, which according to poverty simulations is likely to have generated a further drop in rural poverty. The study argues that given Tanzania's agricultural potential, there is significant scope for reducing poverty by measures that would foster growth in agriculture and thus the incomes of farmers. Another path out of poverty is the movement from agriculture to other sources of income, possibly combined with migration from rural to urban areas. Data suggest that the shift from agriculture to nonagricultural activities in rural areas has been an important contributor to poverty reduction. Informal sector activities have been an important entry point for the poor to engage in nonagricultural activities. Rural-urban migration has also contributed to poverty reduction. However its quantitative significance was less than that of the other channels, probably because most of the migrants are from households above the poverty line. However, migration is only one path in which fast urban growth can benefit the poor in rural areas. Indirect channels include higher demand for rural products, wage effects, and transfers. However, the fact that rural growth and poverty reduction lag significantly behind urban growth and poverty reduction suggests that these links are still weak. However, urban-rural links can also result in a deepening of urban-rural differences. There is evidence that rural migrants are typically better educated than the average rural population, leading to a widening of the education gap as these migrants move from rural to urban areas. A large share of financial savings collected by banks in rural areas flows toward Dar es Salaam, funding credit to the private sector and government in Dar es Salaam, as well as overseas investments by the banking sector. Although the mobility of human and financial resources toward opportunities where the returns are highest is supportive of high economic growth in Tanzania, measures that counteract an increasing marginalization of the rural poor in Tanzania's growth process are needed. Such measures would include enhanced rural access to quality education and policies that support agriculture. The report highlights instances in which the lack of integration of rural areas in the economy significantly reduced rural growth. Key among these instances is the access of rural areas to markets as well as to agricultural inputs. For example, surveys carried out in the Kilimanjaro and Ruvuma regions suggest that lack of access to agricultural inputs results in low agricultural productivity and, consequently, limited progress in rural poverty reduction. This limited access to agricultural inputs is the result of two equally important problems: (a) limited access to input credit; and (b) lack of a rural input supply infrastructure that would allow farmers to purchase these inputs. These results suggest that rural development and informal sector activities are the primary direct drivers of poverty reduction in Tanzania, where the informal sector has been an important transmission mechanism that allowed the poor to participate in economic growth opportunities originating in the formal and pubic sectors. This interpretation is reinforced by the fact that although economic growth was significantly higher in urban areas than in rural areas in the period from 1990/01 to 2000/01, modest rural growth has clearly dominated the faster urban growth with respect to its effect on poverty reduction. Furthermore, even in an environment of relatively high growth differences between rural and urban areas, the contribution of migration and other urban-rural links to poverty reduction has been relatively modest. Appropriate tax and public expenditure policies play an important role in fostering shared growth. As the report highlights, enhancement of the domestic revenue base through sustained economic growth is central to the sustainable financing of public expenditures and reduction of aid dependence in the medium to long term. In turn, tax policies have a direct influence on the level of investment and economic activities. Similarly, public expenditures Sustaining and Sharing Growth in Tanzania xi play an important role not only in improving the environment for economic growth, but also in enhancing the access and quality of public services for the poor. Overall, the Tanzanian tax policy is assessed as being sound and not inimical to growth. However, several measures could enhance the contribution of the tax system to fostering shared growth. First, there remains an urban bias in the tax system: effective tax rates are higher for farmers than for businesses, which are mostly urban. In particular, the crop cess collected by local authorities imposes a relatively heavy tax burden on agriculture. Second, the presumptive tax regime for small businesses is one of the most sophisticated in the region. Nonetheless, it is regressive for small businesses that do not keep records. Third, there are significant weaknesses in the taxation of natural resources, which result in both distortions to the sustainable exploitation of natural resources and suboptimal collection of revenue. Social sector expenditures have seen significant increases in recent years. However, incidence analysis suggests that only in the education sector have public expenditures been pro-poor. In other sectors, such as water, primarily nonpoor households benefited from improved quality and access to services. The focus on social expenditures also limited the availability of funds for growth-enhancing expenditures. What Must Be Done to Sustain Economic Growth That Is Pro-Poor? The review of Tanzania's recent growth performance suggests that enhancing the pace of structural change and diversification and increasing the international competitiveness of the economy remain the key challenges for sustaining growth. The poverty analysis highlights the importance of a productive agriculture sector and of a conducive environment for the activities of small and medium enterprises (SMEs) as key elements of a shared-growth strategy. It also emphasises that participation of the poor in the growth process requires that policies support the accumulation by the poor of primarily human capital and physical and financial capital. Finally, the report underscores the importance of appropriate policies and institutions to manage the design and implementation of a shared-growth strategy, as well as resources for their implementation. Here the report discusses not only the management of economic policies and public finances, but also the equally important management of natural resources. Improved governance of Tanzania's natural resources, strengthened capacity to ensure that tax and expenditure policies are supportive of a shared-growth agenda and better institutional coordination framework for development and implementation of a growth strategy require attention in Tanzania's quest for sustainable shared growth. To sustain and share economic growth across all income groups of society, Tanzania will need to preserve achievements, consolidate ongoing reforms, and strengthen institutional capacity for both policy advice and program implementation. It will also be important to guard against backsliding in the face of pressures from vested interests or impatience with the pace of poverty reduction. The sectoral distribution of growth has a significant effect on the pace of poverty reduction and inequality. Tanzania's comparative advantage in agricultural production and its large potential to enhance agricultural productivity provide a good basis for a focus on agriculture and agriculture-related activities as the central element of its efforts to reduce poverty. Agricultural activities are the source of livelihood for 75 percent of the population, more than 40 percent of whom are poor. Efforts to reduce poverty must focus on measures that will help the poor to (a) generate more income from their current agricultural products; (b) shift their production to more profitable agricultural products; and (c) shift to income- generating opportunities outside of agriculture in both rural and urban areas. In addition, a decline in prices attributable to productivity increases benefits poor people who are net buyers of agricultural products. Sustaining and Sharing Growth in Tanzania xii Increasing agricultural incomes requires policies that target both improvements in market access and increases in agricultural productivity. Enhancing rural infrastructure remains critical to ensure market access for farmers. It is equally important to ensure that institutional arrangements are in place that link farmers to domestic and international markets. An example of reforms in this area would be to ensure that the crop boards function efficiently, with a clear separation of public and private functions, and that they are accountable to farmers. Regulations such as mandatory auctions and single license rules for coffee potentially harm the efficiency of markets and reduce farm incomes. Scaled-up investment in agricultural research and a reform of Tanzania's extension service have important roles to play in supporting farmers in the move to raising crops that yield higher returns. The study highlights the large productivity losses attributable to human diseases that make health intervention an important element of efforts to increase agricultural productivity. In addition to improving farm-level productivity, the focus needs to be on promoting upstream activities such as agroprocessing and enhancing links to domestic and foreign markets. Efforts toward raising agricultural productivity must also take into account the general question of the impact that risks have on agricultural activity in general and the more specific consideration of the capacity of the poor to grow out of poverty. Increased income-generating opportunities in nonagricultural activities, especially in rural areas, are also important for poverty reduction and for the medium- to long-term structural transformation of the economy. Providing opportunities for Tanzanians to move out of the agriculture sector can be expected to improve the labor productivity in the sector and to provide higher incomes for those moving out of the sector. MSMEs, often in the informal economy, provide an important entry point for the poor to engage in industrial and service sector activities. Measures that support MSMEs are thus important, including and public recognition, tolerance, and support for informal sector activities. Formalization should be primarily incentive based in the case of microenterprises. The report highlights the importance of the manufacturing sector as a potential dynamic driver of diversification and growth. Analysis of firm-level data from the enterprise survey suggests that, in order of priority, the leading five factors that affect firm growth and that deserve special attention by policy makers are (a) access to and cost of financial capital; (b) access to technology to improve productivity; (c) infrastructure, especially energy; (d) skilled labor; and (e) the regulatory environment for business activities. Because firm growth is intricately tied to growth in exports, an aggressive and proactive policy stance promoting manufactured exports is likely to have the greatest effect on manufacturing growth in Tanzania and is recommended. The rationale for this selective approach is motivated by today's global reality: if a firm cannot compete in the global market (that is, if it cannot export), it is unlikely to survive too long in Tanzania's domestic or Africa's regional markets, which are flooded with cheaper imports from low-cost and high-skills producers such as those from East and South Asia. The policy implications of an export-oriented stance have several overlaps with factors that promote growth in non- exporting firms, but an aggressive focus on incentives that facilitate the expansion of existing firms and promote new entrants in the export sector is likely to yield the most benefits. Priority areas that require improved policies or scaling up of expenditures include investing in infrastructure; enhancing access to finance, notably for the rural and SME sectors; and continued efforts to streamline regulation, reduce bureaucracy, and fight corruption. Focus on these issues promises increased economic activities in areas that the private sector can readily support. This advice may seem tantamount to recommending everything--that is, redressing all barriers to production presently facing all manufacturing firms in Tanzania. But it is not. To circumvent the high financial and time costs and the government's limited Sustaining and Sharing Growth in Tanzania xiii implementation capacity, the report recommends focus and pragmatism in catering to existing and potential exporters. A sound strategy for delivering physical inputs (such as infrastructure) and financial inputs (which will make bank finance more accessible) includes the identification of spatial locations where export activity is most prevalent and where exporters most likely to locate. The government is pursuing this strategy through export processing zones and special economic zones. However, in rolling out this strategy, it will be important to sequence these activities by initially addressing problems of existing zones before new ones are created. A review of Tanzania's export processing zones highlights infrastructure weaknesses, especially reliable access to electricity and water, as a main constraint for firms located in those zones. The targeted improvement of infrastructure services to the manufacturing sector thus needs to be a priority in Tanzania's efforts to spur growth and structural transformation. Spatial targeting helps in targeting exports. This approach would render public support in a financially feasible and timely manner for fast- growing exporters and potential new entrants into the export business. Sustained economic growth will increasingly depend on the capacity of economic actors to innovate, to produce an increased array of goods and services, and to accelerate the pace of technological change. It will require foremost a greater focus on investment in human resource development (and vast improvement in secondary, technical, and tertiary education), as well as strengthening of the innovation environment and Tanzania's fledgling information and communication technology (ICT) infrastructure. Tanzania's natural resource endowment could be an important source of growth and poverty reduction. Strengthening the governance of natural resource use and the backward and forward links to other activities is critical to ensure that Tanzania benefits from the exploitation of its natural resources. In addition, improved governance arrangements are important for the sustainable exploitation of renewable resources such as fish or forests and are necessary to minimize the impact of negative externalities such as that of commercial fishing and mining on their artisanal counterparts. A shared-growth strategy requires also a focus on the capacity of the poor to contribute to-- and participate in--economic growth. This strategy includes opportunities to build human capital through measures that center on equitable access to, and improved quality of, education (primary, secondary, and technical), nutrition, and health services; to reduce the burden of communicable diseases; to improve child nutritional status; and to reduce maternal mortality by helping women achieve their desired family size. Other measures include supporting household savings and investment through development of appropriate finance and ancillary institutions, especially in rural areas. Limited social protection measures can also play a role in mitigating the impact of large shocks that may create poverty traps for the poor. Regional differences in economic performance reflect not only inherent differences in economic potential, but also factors such as past investments in infrastructure and human resources, local governance, and connectivity. Data suggest some degree of convergence in per capita incomes across regions. However, in most regions, the growth performance remains intimately linked to the fortunes of individual crops. Several policy lessons emerge from the analysis of subnational growth patterns. First is the importance of capturing local knowledge to fully exploit growth opportunities. This lesson suggests that decentralization not only is a means for improved service delivery, but also has an important role to play in the implementation of Tanzania's growth strategy. Local knowledge about growth opportunities is critical for a series of public interventions to foster growth, ranging from infrastructure investments to targeted, crop-specific interventions in the agriculture sector. Second, local governments have a significant effect on the business environment, ranging from their attitude toward the informal sector to local tax policy and administration, Sustaining and Sharing Growth in Tanzania xiv licensing, land management, and so forth. Finally, attention to subnational growth is important to identify successful strategies that can be scaled up and replicated in other parts of the country. The study argues that the regional distribution of public investment should be determined by the growth opportunities, whereas distributional objectives should be primarily pursued through targeting of access to education. As the report highlights, Tanzania has been successful in establishing a sound basis for sustained and shared economic growth through the implementation of a broad reform agenda. Tanzania's strategic frameworks for growth and poverty reduction, including the National Development Vision 2025, the Medium-Term Plan for Growth and Poverty Reduction, and the NSGRP, adequately identify core interventions that are needed to sustain economic growth. These interventions include the strengthening of economic infrastructure, scaling up of human resource development from an initial focus on primary education to secondary and higher education, and the implementation of reforms to strengthen the business environment. In many key areas, specific reform programs are in place, including the Business Environment Strengthening Program in Tanzania, the Agricultural Sector Development Program, the Primary and Secondary Education Development Program, and the Second-Generation Financial Sector Reform Program. The report broadly endorses this reform program. Table 1 summarizes key issues of the growth agenda, many of which are already being addressed by government. The report highlights three elements of the reform agenda that deserve increased attention: · Enhancing international competitiveness and accelerating diversification; · Making growth pro-poor; and · Managing policies and resources for shared growth. Enhancing International Competitiveness and Accelerating Diversification Enhancing international competitiveness and accelerating diversification with a focus on macroeconomic management, infrastructure, access and cost of credit, the regulatory environment for private sector activities, human resource development, the innovation environment, and the rollout and use of ICT involves the following: · Resolve infrastructure bottlenecks. Recurrent energy shortages are the most visible constraint to economic growth. However, general underinvestment in the development and maintenance of transport infrastructure, especially in the rail sector and for rural roads, also holds back growth, particularly in rural areas. In addition to scaled-up investment, there is an urgent need to get appropriate policy and regulatory frameworks in place, which would support a greater participation by the private sector in selected areas of infrastructure development, operation, and maintenance. · Devote greater attention to fostering structural transformation. To sustain its economic growth, Tanzania will increasingly need to broaden the range of goods and services it produces. The role of government in the process of structural transformation is not to identify new growth opportunities, but rather to support the identification and exploitation of opportunities by the private sector. This effort will require scaling up of investment in higher education, with the availability of skilled labor already being a constraint to economic growth. FDI and the import of technology are the primary sources of new technology for Tanzania. However, a strengthening of Tanzania's research and development systems, especially in agriculture, also has an important role to play in the adaptation and dissemination of new technologies. Finally, greater access to ICT is an important tool to accelerate the acquisition of technology and knowledge. In addition, microeconomic evidence Sustaining and Sharing Growth in Tanzania xv supports a direct link between greater use of ICT, especially cell phones, and the productivity of farms and rural enterprises. In addition to economywide support, there is also some scope for more direct support at the sector and firm levels. An example of such support would be the establishment of a matching grant scheme for the introduction of business activities that are new to Tanzania. · Develop an urban strategy. Urban areas play an important role in the economic growth process. Thus, an important challenge is not only to sustain the good performance of Dar es Salaam, but also to enhance the role of regional urban centers as hubs of economic activity. Such an urban strategy would define the role of urban areas in Tanzania's economy, as well as the necessary investments and an appropriate institutional and fiscal framework that would allow the implementation of such an urban strategy. Making Growth Pro-Poor Making growth pro-poor highlights reforms that would accelerate growth of the agriculture sector; expand income-generating opportunities in micro-, small-, and medium-size enterprises; and strengthen the capacity of the poor to participate and contribute to economic growth: · Ensure that reforms benefit rural areas. To date, the effect of reforms undertaken has been visible primarily in urban areas, especially Dar es Salaam, while rural areas have seen much less systemic improvements in growth performance and poverty reduction. The report highlights five areas in which reform programs should scale up the focus on rural areas. First, improved access to markets is central to increasing real incomes of the rural population. Such access requires investments in road infrastructure and in storage and market facilitation (standards and business climate).The second area is financial sector reform. Despite the overall progress made, rural areas remain largely cut off from access to financial services, which, in turn, have been identified as a key constraint to both farm and off-farm activities in rural areas. The third area is public expenditure reform. In this area, the increases in the overall resource envelope have accrued primarily to central government agencies and to a much lesser extent to local governments. Progress in the development of a sound intergovernmental fiscal framework is a necessary precondition to strengthening the development, operation, and maintenance of local infrastructure and delivery of local services, which, in turn, directly affect rural growth and poverty reduction. Fourth are government policies and expenditures to support agriculture. Here the report suggests that public expenditures on research and extension or irrigation can play an important role in supporting the sector. However, scaling up the public support for agriculture needs to be grounded in a careful analysis of the effectiveness of expenditure programs. Finally, a better integration of rural areas in the growth process will require an effort to draw effectively on local knowledge in the development and implementation of regional and district growth strategies. This effort will require an increased focus of Tanzania's decentralization program on economic growth, in addition to the provision of social services. · Encourage micro-, small-, and medium-size enterprise activities. Microenterprises, mostly in the informal economy, not only are an important source of income for many Tanzanians, but also are an important entry point for Tanzanians into entrepreneurial activities. The primary focus should be on facilitating and supporting such activities. Facilitating the transition to the formal sector is important, but the Sustaining and Sharing Growth in Tanzania xvi move should be voluntary unless the primary motive for being in the informal sector is clearly tax evasion. · Devote greater attention to the effect of social expenditures. The report highlights that the substantial increase in public expenditures since the late 1990s has directly benefited the poor to only a limited extent. This situation calls for a revision of spending priorities and targeting approaches to ensure that public expenditures do benefit the poor. Managing Policies and Resources for Shared Growth Managing policies and resources for shared growth means strengthening institutions to develop and implement a growth strategy and to harness both public finances and natural resources toward the objective of sustained shared growth: · Strengthen the capacity for the implementation of Tanzania's reform agenda. As highlighted throughout the report, Tanzania has made great strides in establishing a vastly improved economic and incentive regime for economic activities. However, in many areas where appropriate policies and regulations are in place, their effectiveness is hindered by limited implementation capacity. Enhancing implementation capacity will rely on the continued implementation of public sector reforms, but the increased use of the private sector, when appropriate, is an important way to realize reform objectives. Examples range from an increased role of the private sector in the provision of agricultural support services to the contracting in of capacity at the National Audit Office. · Strengthen the management of the growth process. The report highlights the importance of developing and strengthening institutions that are able to coordinate the formulation and implementation of a shared-growth (pro-poor) strategy in Tanzania. This effort includes the ongoing strengthening of the budget process to examine and prioritize investments in infrastructure. Given the large regional differences in productive potential, strengthening institutions that are able to respond to these differences will be critical. · Strengthen analytic underpinnings and participation in the design of growth policies. The NSGRP suggests a very welcome greater focus on economic growth, supported by a scaling up of expenditures that would support accelerated growth, ranging from increased infrastructure investments to subsidies for agricultural inputs or credits to targeted segments of the economy. In addition, the government is implementing a range of specific programs, such as Tanzania's Mini Tiger Plan 2020, the Property and Business Formalization Programme, and the National Economic Empowerment Policy. For these measures to be effective, they must be grounded in a solid analytic basis that supports the choice and prioritization of specific strategies and expenditure policies. Strong processes for stakeholder inputs, adequate governance arrangements for expenditure programs at the sectoral level, and a strong monitoring and evaluation system are of central importance. Although these issues have received significant attention in the social sectors, they are yet to be fully developed for growth-enhancing expenditure programs. · Ensure the effectiveness of government interventions through appropriate governance arrangements. The report highlights a range of government interventions and supports the government's proactive efforts that complement the focus on creating an enabling environment for private sector activities. Recent government Sustaining and Sharing Growth in Tanzania xvii interventions--such as targeted credit guarantee schemes, export processing zones, and targeted agricultural subsidies--are innovative efforts with the potential to foster growth, but they also carry significant risks. These risks include the potential for governance problems that could make such interventions ineffective or even counterproductive. The report suggests a range of measures that would improve the likelihood of success of government interventions. · Pay greater attention to the management of natural resources. Tanzania's natural resource endowment could be an important source of growth and poverty reduction. Governance of natural resource use, as well as the backward and forward links to other activities, must be strengthened if Tanzania is to benefit from the exploitation of its natural resources. Moreover, governance arrangements must be improved for the sustainable exploitation of renewable resources such as fish or forests. Strengthening governance can minimize the impact of negative externalities such as commercial fishing and mining on their artisanal counterparts. Finally, improved governance should aim at an equitable sharing of natural resource rents. Table 1 provides an overview of key issues for the implementation of a shared growth strategy: Sustaining and Sharing Growth in Tanzania xviii Table 1. Elements of a growth strategy Priority Magnitude Time Frame of Impact for Impact A. ENHANCING INTERNATIONAL COMPETITIVENESS AND ACCELERATING DIVERSIVICATION Macro-economic Management Maintaining a stable macro-economic environment. Assessment as to whether the real exchange H H Short term rate is overvalued due to high inflows of foreign exchange through gold exports and aid and its consequences for the competitiveness of the economy. Infrastructure Strengthening of the legal and regulatory framework for infrastructure. H H Medium term Prioritization of infrastructure investment - including clustering of services ­ with adequate H H Medium term consideration of projects with short gestation period and high import content, to reduce possible Dutch disease effects. Maintaining an appropriate balance between new investment, rehabilitation, and recurrent H H Short term maintenance. Strengthening of regulatory institutions and ensuring their independence to attract private sector M M Medium term participation Access and Cost of Credit Strengthening the lending environment and financial infrastructure. H H Medium term Facilitating the increase in SME and Long Term Lending H H Medium term Supporting providers of financial support for micro and small enterprises M M Medium term Regulatory environment for private sector activities Revising licensing and registration legislation and regulations to consolidate recent reforms of M M Medium term business licensing Strengthening customs and trade regulations and administration. H H Short term Accelerating legal sector reforms and strengthen capacity in the sector. H H Long term Revision of the anti-corruption legislation to provide more effective instruments for the control H H Medium term of grand and petty corruption Human Resource Development Sustaining and Sharing Growth in Tanzania xix Priority Magnitude Time Frame of Impact for Impact Improving the quality of education as enrollments increase. H H Long term Expanding secondary education. H H Medium term In higher education, strengthening of the governance and administration of the country's three M M Medium term public universities. Using distance education to expand access to education services while at the same time M M Medium term improving quality. Reforming teaching methods and the curriculum at all levels to include skills and competencies H H Medium term to meet the new needs of the economy. Increasing the interface between industry and education, and offer differentiated curricula that H H Medium term better meet the new skill demands of industry, generated by changing markets and technologies. Harmonizing technical education offered in secondary schools with that offered in the current M M Medium term technical colleges, and then link these up with the proposed zonal colleges and institutes. Devising strategies to proactively deal with the problem of skills loss through brain drain. M M Medium term Innovation Environment Supporting the design, use and dissemination of technologies through well targeted subsidies H H Medium term Supporting the provision of demand driven technical support through a network of locally M M Medium term based and owned structures to serve the needs of rural and urban communities for technical advise, information, and assistance. Enhancing the regulatory environment to facilitate the linkages between the business sector and M M Medium term the R&D infrastructure, quality control mechanisms, and related accreditation and certification procedures, and access to finance. Roll-out and use of ICT Finalization and adoption of a new Electronic Communications Bill which is key to defining the M M Medium term ground rules for sector development (including rural areas); Implementation of the new converged licensing framework which will ensure further M M Medium term liberalization of the market; Review and modernization of telecom policies and regulations to generate fair competition and M M Medium term reduce high communication and operational costs. Building capacity to undertake such reforms, including through the establishment of systems and M M Medium term processes to review the performance of the regulatory institutions. Supporting the development rural telecommunications infrastructure, such as by developing M M Long term Sustaining and Sharing Growth in Tanzania xx Priority Magnitude Time Frame of Impact for Impact universal access schemes. Enhancing technical and business related skills development among the population using ICTs M M Medium term through technical institutes and vocational centers. B. MAKING GROWTH PRO-POOR Agriculture Implementation of land tenure reforms. M H Long term Strengthening of agricultural research and extension. H H Medium term Facilitating improvements in irrigation. H H Medium term Enhancing the functioning of output and input markets. H H Short term Aligning Public Expenditure with Agriculture Sector Development Program. H H Short term Informal Sector and Micro, Small, and Medium Size Enterprises Tolerating and supporting small scale informal sector activities, and awareness raising activities among local government officials. Facilitation of the voluntary transition of small scale informal sector activities to the formal sector. Review of tax regime for micro and small enterprises. M M Medium term Scaling up the implementation of the SME policy. Strengthen the capacity of the poor to participate and contribute to economic growth Allowing poor households to benefit from the economic opportunities by building the human capital base of Tanzanians in an equitable and efficient manner. This requires increased attention to the (quality of) of health care ­including how to deal with malnutrition, (primary and secondary) education and access to clean water. Institutional changes as well as the geographic distribution of resources have to be considered. Enhancing the ability of households to benefit from improved economic opportunities by facilitating their access to information (including extension) and their ability to take risk, save and invest. Pay attention to appropriate financial sector development (savings, credit, new form of insurance) and population growth. Dealing with vulnerability primarily by paying more attention to risk management. Consider options for risk prevention and reduction by adjusting existing sector policies. Be selective in the use of safety nets as a response to vulnerability. Sustaining and Sharing Growth in Tanzania xxi Priority Magnitude Time Frame of Impact for Impact Improving the quality of and access to survey and administrative data. C. MANAGING POLICIES AND RESOURCES FOR SHARED GROWTH Institutions to Steer the Development and Implementation of Tanzania's MKUKUTA Complementing economic stabilization with greater focus on economic growth and align H H Short term institutional arrangements accordingly. Strengthening the coordination of economic policy formulation and implementation. H H Short term Creation of a platform for a national dialogue on growth and related issues. M M Medium term Redefining the government-private sector relationship. M M Medium term Strengthening private sector institutions. M M Medium term Strengthening the capacity of institutions at the regional and district level to play a greater role in H H Short term promoting growth at the local level. Public Spending for the Implementation of the MKUKUTA Review of the impact of tax policy reforms on economic growth, investment, and saving and the H H Short term priority should be to establish an equitable and fair tax system rather than to maximize government revenue. Close monitoring and management of the macro-economic impact of aid to prevent a weakening H H Short term of domestic institutions and competitiveness. Preserving enough flexibility in the structure of public expenditure to be able to react flexibly to H H Medium term fluctuations in the resource environment Further strengthening of management of public expenditures and the public expenditure review H H Short term process. Natural Resource Management Strengthening of the capacity for data collection, record keeping, monitoring, control and M M Medium term surveillance and enforcing punitive measures to control illegal practices. Controlling externalities through fiscal instruments, royalties and resource pricing and to M M Medium term increase revenue from rent capture rather than un-controlled exploitation. Increasing the efficiency in revenue collection and administration, as well as full transparency M M Medium term and accountability over revenue generation and distribution. Promoting market based principles, where appropriate, ensuring local spin-offs and allowing M M Medium term competition and entrepreneurial development. Sustaining and Sharing Growth in Tanzania xxii Priority Magnitude Time Frame of Impact for Impact Sustaining and Sharing Growth in Tanzania 1 1. POVERTY REDUCTION AND GROWTH ­ RECENT PERFORMANCE AND PROSPECTS A DECADE OF STRUCTURAL REFORMS, MACRO-ECONOMIC STABILITY AND ECONOMIC GROWTH Much progress has been made in implementing structural reforms in Tanzania (Box 1). Since the mid 1990s, the country has successfully pursued macro-economic stability as its main policy objective. Monetary and fiscal policy have been designed and implemented with the objective of price stability. The privatization process is well advanced. Government has withdrawn from most commercial activities and has encouraged private sector participation in the utility sector. Trade policy has been reformed in the context of the East African Community, where a customs union became effective in 2005. Although some improvements have already taken place, bureaucracy and corruption remain a source of concern to current and potential investors. Table 2. Key Economic Indicators, 1995-2005 Indicator 1 9 9 5 2 0 0 0 2 0 0 4 2 0 0 5 Actual Actual Actual Prel. Esst. Annual inflation (percent, period average) 27.4 6.0 4.1 4.4 National accounts, percent growth rates Agriculture 5.7 3.4 5.8 5.2 Industry -2.0 6.7 10.0 10.6 Service 2.9 6.4 6.3 6.9 GDP at market price 3.6 5.1 6.7 6.8 National accounts (percent of GDP at market price) Gross domestic investment 19.8 17.4 20.9 21.9 Public investment 3.4 6.0 7.7 7.5 Private investment 16.2 11.4 13.2 14.4 Gross domestic saving 0.8 9.2 11.4 8.4 Balance of payments (percent of GDP at market price) Current account balance -9.9 -5.3 -3.9 -7.9 Current account balance (excluding current official transfers) -19.9 -10.0 -9.0 -11.8 External Indicators Debt Service/Exporting of Goods and Services (After rescheduling 21.3 13.2 1.8 2.3 Reserves in months of imports 1.5 5.2 5.9 4.4 Government Finance (percent of GDP at market price)* Domestic Revenue 12.5 11.3 12.7 13.3 Public Expenditure 18.3 18.6 22.0 24.4 Overall Deficit (excluding grants) -5.9 -7.8 -9.3 -11.1 Overall Deficit (including grants) -4.2 -3.3 -3.2 -3.4 External borrowing (net) 0.9 1.6 3.8 3.0 Domestic borrowing (net) 3.3 0.1 -0.4 1.1 data are for fiscal years, e.g., 2006 - 2005/06 Source: URT. Economic Survey. Various years. Sustaining and Sharing Growth in Tanzania 2 Box 1. Overview of Structural Reforms in Tanzania Financial Sector. During the past decade, the financial sector has seen significant change. From being the sole preserve of state-owned financial institutions, it has gone through a process of privatization and been opened up to new entrants. The successful privatization of the two largest state-owned banks (sale of a 70 percent stake of National Bank of Commerce and a 49 percent stake of National Microfinance Bank (NMB)) has contributed to increased efficiency and competition in the banking sector as well as narrowed interest spreads and a fairly rapid increase in credit to the private sector. Financial sector reforms also included the strengthening of the legal and regulatory framework, including for micro-finance, as well as a strengthening of supervision by the Bank of Tanzania. The recent establishment of a credit rating agency is a further step in enhancing the efficiency of financial intermediation in Tanzania. The main challenges for the sector include further reduction in interest spreads and enhanced access to credit by the private sector, especially in rural areas. Government has prepared a comprehensive Action Plan for Second Generation Financial Sector Reforms (SGFSR) which is now under implementation. Parastatal Sector. Tanzania has been aggressively implementing its privatization agenda. The privatization of manufacturing and commercial parastatals was virtually complete by 2000 and a solid success. The reform and development of public private partnerships in the infrastructure sector proved much more difficult. At the end of the 1990s, Tanzania launched the privatization of its infrastructure enterprises. By 2003, five key infrastructure enterprises (TANESCO, power; DAWASA, water; TTCL, telecom; TICTS, the container terminal; and Air Tanzania) had some form of private participation. However, by early 2007, only the private sector participation in TICTS is still in place and considered as being successful, while private sector participation in DAWASA, and Air Tanzania have been dissolved and that of TANESCO was not renewed as a result of performance. For the two railroads, TRC is under negotiations while TAZARA has not yet started and so is the harbor authority (TPA). Trade Policies and Institutions. Reforms of trade policies have taken mainly place in the context of regional agreements, including SADC and EAC. Tanzania adopted the Common External Tariff (CET) of the East African Community (EAC) in January 2005, lowering its average tariff from 13.8 to 12.3 percent, but further raised the dispersion of protection. The lowering of the maximum tariff of the CET from the current 25 percent to 20 percent, as expected to happen in 5 years in accordance with the Customs Union Protocol, should help correct some of the dispersion of protection. On the export side, the main issue pertains to export taxes. International experience has shown that export taxes and bans have generally failed to achieve industrial development objectives, led to informal trade, and frequently hurt small-holders who receive lower prices as a result. Factor Markets (Labor and Land). The revision of land and labor legislation has been completed, with most of the emphasis on the reform of institutions. The main challenge is now to implement this new legislation, which in the case of land reforms is expected to be a lengthy and costly process. Infrastructure (Power Sector and Transport). The establishment of the executive agency TANROADS with responsibility for the trunk road network has been a major step forward for the transport sector. However, a clear separation of responsibility between the Ministry of Works, TANROADS, and districts has not yet been implemented and this hampers effective road maintenance and development activities. The current formulation of a new Road Act provides a real opportunity to establish a more appropriate policy and institutional framework and provide the basis for accelerated infrastructure development. Detailed work on the restructuring of the power sector has been carried out but the implementation of the restructuring has been delayed. The performance of the parastatal power utility TANESCO had improved under a performance based management contract, but remains highly vulnerable as the power crisis during 2005/06 has demonstrated. Reforms of the policy and institutional framework for the power sector are essential to ensure the effectiveness of future investments in the sector. Public Institutions Interfacing with the Private Sector. Red tape, corruption, and overly burdensome regulatory and licensing requirements are among the main constraints to private sector development in Tanzania. The government has started reviewing regulations, focusing on removing obstacles and re- organizing the most important tasks of government. In practical terms this means (1) harmonization of local government taxation to remove excessive tax burden on private enterprise; (2) streamlining of work permit procedures; (3) review and amendment of licensing legislation to reduce the cost of business establishment and continuation; (4) review and revision of export-import procedures to reduce time costs and corruption-related costs; and (5) design and implementation of a program for enhancing access to commercial courts by SMEs. Sustaining and Sharing Growth in Tanzania 3 Tanzania has reformed the legal framework for regulatory institutions. The effectiveness of these newly established regulatory institutions, especially given the current oversight arrangements, needs to be closely monitored. Reforms have resulted in significant improvements in the economic fundamentals (Table 2). Net domestic borrowing by the central government has been tightly controlled through a cash budget system since the late 1990s. This has allowed the pursuit of a monetary policy which contained inflation at about five percent during the past five years. This has also created space for a rapid expansion of credit to the private sector. Credit to the public sector declined from five percent of GDP in 1998 to 1.9 percent by the end of 2005, while credit to the private sector increased from four percent to 10.4 percent. Domestic saving increased from 1.5 percent of GDP in 1998 to 9.7 percent in 2005, and the real interest rate on savings became positive. The real effective exchange rate has depreciated by about 30 percent since 2001, which eliminated the overvaluation of the exchange rate. The current account deficit (after grants) has narrowed from around ten percent of GDP in 1998 to less than five percent in 2004 but increased subsequently to 9.3 percent of GDP in 2005/06, reflecting an increase in aid inflows and debt relief. International reserves stood at about 5 months of import coverage by the end of 2006. Table 3. Sources of Growth (production), 1990 - 2005 Avg. Ann. Growth Rate Avg. Contr. To Growth ECONOMIC ACTIVITY 1990-94 1995-99 2000-05 1990-94 1995-99 2000-05 Agriculture 3.1% 3.6% 4.8% 1.5% 1.8% 2.3% Industry 2.0% 5.4% 9.0% 0.3% 0.9% 1.6% Mining and Quarrying 11.8% 14.8% 15.2% 0.1% 0.2% 0.4% Manufacturing 0.4% 4.6% 7.3% 0.0% 0.4% 0.6% Construction 2.2% 3.5% 10.3% 0.1% 0.2% 0.5% Services 1.9% 3.8% 6.1% 0.7% 1.3% 2.1% Trade, Hotels and Restaurants 2.0% 4.5% 7.1% 0.3% 0.7% 1.2% Total GDP (factor cost) 2.5% 4.0% 6.0% 2.5% 4.0% 6.0% Source: Author's calculations based on data from URT. Economic Survey. Various Years. Macro-economic stability and structural reforms have created a favorable environment for higher economic growth, which was 6.0 percent during the period 2000 -2005 compared to 4.0 percent and 2.5 percent in the preceding two five-year periods. Table 3 shows the growth rates of the main sectors of the economy as well as the contribution of these sectors to overall growth. The increase in the growth rates occurred across all sectors, with industry growing by 9 percent, services by 6.1 percent, and agriculture by 4.8 percent during the period 2000-2005, with mining, manufacturing, and construction being the fastest growing sub-sectors. The contribution of the various sectors to growth, which depends both on the growth rate of the sector and its share in the economy, shows that agriculture contributed 2.3 percentage points, services 2.1 and industry 1.6 percentage points of the average annual growth of 6.0 percent during the period 2000-2005. Table 4. Sources of Growth (expenditure), 1990 - 2005 Average annual growth rate (%) Average contribution to growth (%) Sustaining and Sharing Growth in Tanzania 4 2000­ 2000­ Economic Activity 1990­94 1995­99 2005 1990­94 1995­99 2005 GDP (market prices) 2.5 3.8 7.0 Consumption 2.1 6.0 5.4 2.1 6.2 5.6 Private consumption 2.6 5.7 4.0 2.1 5.0 3.4 Government consumption ­0.3 7.3 10.9 0.0 1.2 2.2 Gross domestic investment 5.8 ­1.3 15.3 1.2 ­0.3 3.5 Net exports ­1.4 ­2.1 9.1 0.4 0.4 ­1.7 Exports of goods and nonfactor services 12.1 5.4 19.4 1.6 1.0 5.0 Imports of goods and nonfactor services 3.2 1.4 15.0 1.3 0.6 6.7 Statistical discrepancy ­37.4 ­285.0 4.3 ­1.2 ­2.6 ­0.4 Investment by type of asset Construction 2.1 ­4.4 18.9 0.2 ­0.4 1.8 Machinery and vehicles 9.9 1.0 12.5 1.0 0.1 1.6 Investment by public and private sector Private investment 20.5 ­0.3 11.6 2.3 0.0 1.9 Public investment ­11.5 ­5.0 25.9 ­1.0 ­0.2 1.5 Author's calculations based on data from URT. Economic Survey. Various Years. Figure 1. Growth rates of GDP incl. and excl. government spending, 1990 - 2005 8.0% 7.0% 6.0% 5.0% 4.0% 3.0% 2.0% 1.0% 0.0% 1990-94 1995-99 2000-05 GDP mp GDP mp (net of government spend Source: Author's calculations based on data from URT. Economic Survey. Various Years. Increased government spending has been an important engine of economic growth, contributing 3.8 percentage points to the overall growth of 6.8 percent of GDP at market prices during the period 2000-2005. Figure 1 shows dramatically different trajectories for the growth rates of GDP at market prices including and excluding government spending. Overall GDP at market prices shows the familiar acceleration of economic growth in Tanzania over the past fifteen years. However, GDP net of public sector expenditure (i.e., private consumption and investment) shows a markedly different growth path. There, growth during the period 1995-2005 was 4.5 percent and thus actually lower than growth during the first half of the 1990s, when it was 4.8 percent. This suggests that demand side effects of variations in government spending contributed significantly to the Sustaining and Sharing Growth in Tanzania 5 recent growth acceleration. Most of the increase in government spending is financed from increased aid inflows to Tanzania. In the National Accounts data, this is reflected in a widening of net-exports during the period 2000-2005. Box 2. Government Spending and Economic Growth Government spending affects economic activity both though its demand side and supply side effects. The magnitude and time pattern of demand side effects depends on whether there are unused capacities in the economy and whether changes in government spending are temporary or permanent. In the case of excess capacity, increases in government expenditure add to economic activity directly through the added demand in the form of government purchases as well as through multiplier effects on private sector consumption. Empirical estimates of this demand side effect for developing countries (Hemming et. al. 2002, Schclarek 2005) suggest that these demand effects are typically larger in developing than in developed countries. If the increase in government expenditure exceeds unused capacity, it is likely to exert inflationary pressures in the short run and - if increases in government expenditure are considered to be permanent - induce an expansion in the productive capacity of the economy. Public expenditures also affect aggregate supply. Government spending, especially government investment in economic infrastructure, human capital, or research and development, can increase the productive capacity of the economy. While the demand side effects typically occur in the short term, supply side effects have a longer gestation period. In addition, while demand side effects only affect the level of economic activity, supply side effects have the potential to lead to a sustained increase in the growth rate. Separate from the impact of government spending on economic activity is the impact of the way government spending is financed. In general, higher taxation and domestic borrowing have a negative impact on economic activity. The impact of donor financing­separate from the impact of the associated increase in government spending­is primarily through its impact on the exchange rate, if the associated foreign exchange inflows are not sterilized. Figure 2. Contribution of public and private expenditure to economic growth, 1990 - 2005 6.0% 4.9% 5.3% 5.0% 4.5% 3.8% 4.0% 3.0% 2.0% 1.0% 1.0% 0.0% -1.0% -2.0% -1.1%-0.9% -3.0% -2.1% -2.1% 1990-94 1995-99 2000-05 Private Consumption and Investmen Public Consumption and Investment Net exports and statistical discrepa Source: Author's calculation's based on data from URT. Economic Survey. Various Years. Figure 2 presents the contribution of private and public expenditure to economic growth and shows a similar picture as Figure 1. The contribution of private expenditure to growth increased only from 4.5 percentage points during the period 1990-94 to 5.3 percentage points during the period 2000-05. On the other hand, the direct contribution of government spending to economic growth increased dramatically during that period. Public sector Sustaining and Sharing Growth in Tanzania 6 reforms during the first half of the 1990s involved a significant reduction in the size of government and thus slowed down growth of GDP. The period 1995-1999 saw as slight recovery of government spending and the contribution to GDP growth was one percentage point. The period 2000-2005 saw a sharp increase in government spending and its direct contribution to overall GDP growth increased to 3.8 percent.1 The contribution of the sum of net exports and unrecorded trade and the statistical discrepancy has been negative and increasing during the three periods. This suggests that the demand side impulses emanating from increased government spending have only partly translated into higher domestic production, but have also contributed to increased imports and a widening current account deficit. A rapid expansion of exports of gold and fish, as well as a recovery of exports of manufactures has led to rapid export growth, while agricultural exports have been in decline since the mid 1990s and have only started to recover very recently. Even though exports of gold rose from virtually nothing to about five percent of GDP, their contribution to economic growth has been only around 0.4 percentage points. There is concern that both the gold and fishing industries are reaching the limits of expansion of natural resource extraction, with only limited prospects of future growth, coupled with concern about their environmental impact. Exports of manufactured goods have recovered in recent years, from about US$30 million in 1999 to US$156 million in 2005, which is only 23 percent higher than the value of export of manufactured goods that had already been achieved in 1996. A key challenge for the Tanzanian economy is thus to strengthen and diversify its export base. Figure 3. Decomposition of economic growth per worker into contribution of human and physical capital accumulation and total factor productivity (TFP), 1985-2005. 2.5% 2.0% 1.5% 1.0% 0.5% capital Education 0.0% TPF -0.5% -1.0% -1.5% -2.0% 1985-89 1990-94 1995-99 2000-05 Source: Author's calculations based on data from Bosworth and Collins (2003) and URT. Economic Survey. Various Years. Growth accounting calculations suggest that the recent acceleration of economic growth is primarily due to more rapid accumulation of physical capital and increased factor productivity, as shown in Figure 3. During the period 1995-1999, the ratio of investment to GDP declined from 19.9 to 15.5 percent and the contribution of capital accumulation was negative. After 1999, investment increased significantly, reaching 22.2 percent of GDP by 2005. During that period, public and private sector investment increased by 2.1 and 4.4 Sustaining and Sharing Growth in Tanzania 7 percentage points, respectively. However, national accounts data may not fully reflect the significant increase in foreign direct investment (FDI) that Tanzania has witnessed since the mid 1990s. While there was virtually no FDI until the early 1990s, by 2000 FDI was more than US$ 500 million or about 5 percent of GDP. An Investment Report prepared by the Bank of Tanzania (2004) shows that most of the FDI flows go into mining (30 percent), manufacturing (31 percent), wholesale and retail trade, including tourism (14 percent), while agriculture (7 percent) received only a relatively small share of total FDI.2 The sectors that where the main beneficiaries of FDI flows showed the highest growth rates in the past five years. The FDI survey also shows the increasing importance of regional FDI flows from SADC and EAC countries (primarily South Africa, but also Kenya) accounting for 42.2 percent of total FDI flows in 2001 compared to 52.2 percent from OECD countries. In 2005/06, FDI had declined slightly to about 3.8 percent of GDP, as the major investments in the mining sector had been completed and privatization related FDI declined. Recent investments in education have a relatively long gestation period until they lead to an effective increase in the human capital of Tanzania and the contribution of human capital to growth is still small. However if recent trends in expanding primary and secondary education are sustained, the contribution of human capital to growth is likely to increase. The share of growth that cannot be explained by human and physical capital accumulation is labeled "total factor productivity," but contains a variety of different factors that contributed to growth. Firstly, it captures the increase in land under cultivation in response to improved incentives for agricultural production, which has been a major contributor to growth in the agriculture sector. It also includes increases in capacity utilization as the result of increases in aggregate demand. Enhanced efficiency of resource allocation and utilization is another important component of the observed increase in factor productivity in Tanzania. Analysis of the manufacturing sector suggests that increased productivity in the sector is the result of accelerated exit and entry of firms in the sector (Harding et. al. 2002). As older, inefficient firms leave, new, more productive firms enter the sector. All these factors contribute towards bringing Tanzania's economy closer to its production frontier. Productivity increases that lead to an expansion of the production frontier, such as the adoption of improved technologies have been of lesser importance, but a number of cases were new technologies have been successfully introduced demonstrate the scope for future growth impulses. This includes innovations introduced through FDI, but also the results of Tanzania's agricultural research system. Various international benchmarking exercises highlight weaknesses in Tanzania's competitiveness, both in comparison to other African countries and even more pronounced, in comparison to some of Tanzania's potential key competitors in Asia (Box 3). Sustaining and Sharing Growth in Tanzania 8 Box 3. Benchmarking Tanzania in the Global Context · World Economic Forum's Global Competitiveness Report 2006-2007 highlights its new Global Competitiveness Index (GCI) that provides an overview of factors critical for driving productivity and competitiveness, grouped in 9 pillars: institutions, infrastructure, macro-economy, health and primary education, higher education and training, market efficiency, technological readiness, business sophistication, and innovation. Tanzania is ranked 104th (of 125 countries) on this new 2006 GCI, behind Kenya (94), but ahead of Uganda (113). WEF's Business Competitiveness Index (BCI) is another index that focuses on microeconomic factors that determine economies' current productivity and competitiveness. Tanzania is ranked 73rd (of 121 countries) on the BCI in 2006, ahead of Uganda (88) but again behind Kenya (68). · World Economic Forum's Africa Competitiveness Report 2004 highlights the prospects for growth and obstacles to improving competitiveness in 25 African economies. In this report, Tanzania ranks 9th out of 25 countries on the overall GCI index, surpassing Uganda, which is ranked 14th and Kenya, which is placed 15th. · World Bank Institute's Knowledge Assessment Methodology (KAM)3: Figure A compares Tanzania's performance with the African region, with neighbors, as well as with Botswana and South Africa, as well as with well performing East Asian economies, such as Malaysia and Thailand on the knowledge economy index (KEI) It shows that between 1995 and the most current period (2004-2005), Tanzania has made a substantial improvement in its overall knowledge readiness, as evidenced by positive changes in the KEI, particularly for the economic and incentive regime and the innovation pillars, as well as some improvement in ICTs,. Uganda has also made strides in improving its economic incentive regime and the ICT pillars, and Kenya has also strengthened its information infrastructure over the past decade or so. Kenya's performance also surpasses that of the African region, while Tanzania and Uganda do not. Botswana and Malaysia have slightly improved their recent performance over that in 1995, while South Africa and Thailand have not. Thus, this relative comparison shows that even though a country can make progress, it can still fall relatively behind because the world as a whole may have made a much more significant improvement in the variables that are used to track knowledge and innovation related performance. Figure A: Tanzania and Comparators, 1995 and Most Recent Period Note: In Figure A, the two bars represent the aggregate KEI score for a selected country for most recent year and for 1995, split into four pillars. Each color band represents the contribution of a particular pillar to a country's overall knowledge readiness. Sources: World Economic Forum's Global Competitiveness Report 2006-2007; World Economic Forum's Africa Competitiveness Report 2004, and the World Bank's Knowledge Assessment Methodology (KAM): http://www.worldbank.org/kam. Firm level data suggest access to financial capital, improved technology, infrastructure, and labor skills, and the investment climate as the main constraints to higher investment and firm growth. The analysis also shows that firms which export typically grow faster than firms which do not. The strong link between the proportion of sales exported and growth Sustaining and Sharing Growth in Tanzania 9 indicates that exports are a critical source of growth for globally competitive firms as they relax the demand constraint that otherwise limits the size of the market for nonexporters. Younger firms tend to grow faster than older firms, which highlights the importance of facilitating market entry, but also exit, as an important element of a strategy to raise productivity and growth. The analysis of subnational patterns of income and growth reveals large differences across regions. About 52 percent of the annual national GDP (1992-2003) was produced in only six regions, namely Dar es Salaam, Mwanza, Shinyanga, Arusha, Mbeya and Iringa out of the 20 regions. The Dar es Salaam region alone, which is home to less than eight percent of Tanzania's population, contributes about 18 percent of Tanzania's GDP, equal to the contribution to national GDP of the bottom six regions (Coast, Lindi, Kigoma, Mtwara, Mara and Dodoma) combined.4 The past decade has seen significant differences in the economic fortunes of individual regions. In terms of GDP per capita, the top-five regions for the period 1992-99 were Dar es Salaam, Arusha, Rukwa, Ruvuma and Iringa. Those with lowest GDP per capita included Kigoma, Dodoma, Tanga, Kilimanjaro and Mara regions. However, the ranking changed during 2000-03. Mtwara and Mwanza regions replaced Ruvuma and Rukwa among the top five regions in terms of per capita GDP. Similarly, Kagera, Tabora and Coast regions joined the ranks of regions with the lowest GDP per capita while Mara, Kigoma and Tanga regions made marginal gains. The reversals in the regional ranking in terms of average GDP per capita partly reflects new investment in mining, fishing and related services around Lake Victoria, pick-up of the cashew nut industry in Mtwara, and collapse of the coffee industry (Kagera, Kilimanjaro and Ruvuma). Various investment climate assessments (e.g., World Bank 2004b, World Bank 2006b) carried out for Tanzania show significant regional differences in the business environment, which are related to the economic performance of enterprises in these regions. This suggests that differences in regional growth rates are partly due to factors under the control of local authorities and not only due to external factors such as price movements for region specific products or natural resources. Policies of local authorities that affect growth include local taxation, the quality and implementation of regulatory and administrative procedure, or the quality of public expenditure decisions of local authorities for rural roads, agriculture support, and economic infrastructure such as markets. Tanzania's "Mini-Tiger Plan" (URT, 2004) presents an important effort to create regional growth poles that would capitalize on region specific economic potentials. Many elements of the Mini Tiger Plan, such as the targeting of infrastructure services based on specific economic potentials would clearly serve to enhance regional economic growth. However, some proposals, such as the roll out of Special Economic Zones with expanded tax incentives need to be considered carefully in light of the experience to date with Export Processing Zones in Tanzania and internationally (Box 4). Sustaining and Sharing Growth in Tanzania 10 Box 4. Review of Experience with Export Processing Zones (EPZs). Tanzania Since the establishment of the Mainland Tanzania EPZ Program in 2002, 7 developers and 9 operators have been granted EPZ licenses. EPZs currently provide employment for about 1,600 people, compared with 2,500 in 2004. The "Tanzania - Diagnostic Trade Integration Study" (World Bank 2005a) suggests that Tanzania's legal and regulatory regime is broadly adequate. The time and cost involved in obtaining an EPZ license are competitive with neighboring countries such as Kenya and Zambia. However, Tanzanian EPZs have higher leasing cost per square foot than Kenya, while at the same time the services provided are more limited. EPZ firms continue to face infrastructure constraints, especially limited access to power, transport and water. EPZ firms also face institutional constraints, including lack of an `on-site' customs office in the largest EPZ (Millennium Park), and cumbersome reporting requirements by the Tanzania Revenue Authority and National Development Corporation. These factors have contributed to low occupancy and limited positive impact of EPZs to date. International Experience A recent World Bank report "Free Zones: Performance, Lessons Learned and Implications for Zone Development" (FIAS, 2004) documents that Free Zones have been an important tool to create jobs, generate exports and attract foreign direct investment in many countries, especially Asia. Successful zones are characterized by some or all of the following features: · Streamlined regulatory framework · Public-private partnership approaches for zone development · Largely private sector-led; lead role for one developer · Clear zone designation and development criteria · Top level, integrated support of government e.g., Jordan, UAE · Competition on the basis of facilitation and services rather than incentives · Zone authority is autonomous, flexible, and focused on regulation, · Regulatory authority capabilities are built up · Minimization of public expenditure by locating zones carefully/using existing facilities The performance of Free Zones in Africa to date is rather disappointing, with only a few well known successes such as Mauritius and Madagascar, but a large number of schemes that are performing well below expectations, or have failed completely. This suggests that Special Economic Zones (SEZs) and EPZs are not a panacea to overcome weaknesses in the business environment that prevail in many African countries. Economic Zones cannot be a viewed as a substitute for a country's larger trade and investment reform efforts. They are one tool in a portfolio of mechanisms commonly employed to create jobs, generate exports and attract foreign direct investment-through the provision of incentives, streamlined procedures and custom built infrastructure. However, maximizing the benefits of zones depends on the extent to which they are integrated with their host economies. The static and economic impacts of zone development are suppressed when zones are operated as enclaves. They are multiplied when accompanied by economic policy and structural reforms that enhance the competitiveness of domestic enterprises and facilitate the development of backward and forward linkages. ECONOMIC GROWTH, POVERTY, AND INEQUALITY With an estimated per capita income of US$340 (Atlas method) in 2005, Tanzania is among the world's poorest countries. A household budget survey carried out in 2001 shows that 35.3 percent of the population lives below the national poverty line. However, the national poverty line (at US$0.32 per day) is low both in terms of regional and international comparisons. Using the international dollar a day poverty line increases the share of the population living in poverty dramatically to 57.5 percent. Poverty levels are highest among households that depend on agriculture for their livelihood. About 40 percent of such households are below the poverty line and they make up about 81 percent of the poor in Tanzania. Poverty levels are lowest among households whose head is a paid private or public sector employee, followed by households whose head Sustaining and Sharing Growth in Tanzania 11 is self-employed in a non-agricultural occupation (the urban informal sector). Consumption per capita in non-agricultural households is about 1.5-1.8 times higher than in agricultural households, and these households have experienced a faster growth of income over the last decade, especially in Dar es Salaam. The household budget survey revealed fairly large regional differences in poverty levels in Tanzania (Table 5). Dar es Salaam experienced a major reduction in poverty, triggered by growth of employment and incomes in the formal private sector, which spilled over to the informal sector, raising incomes there as well. Meanwhile, poverty worsened in the Northern Highlands area, especially in the rural areas and is lowest in the Southern Highlands. Thus, Singida has the highest incidence of poverty (55 percent), while it stands at only 21 percent for Mbeya. Despite these regional variations, the level of inequality in Tanzania--between regions, between rural and urban households, and for the whole mainland--is still relatively low compared to other African countries. Table 5. Zonal Poverty Status Population share Poverty headcount 1991/1992 2000/2001 1991/1992 2000/2001 Tanzania 100.0 100.0 38.6 35.3 Rural 82.1 79.0 40.8 38.6 Other urban 12.6 13.6 28.7 25.9 Dar es Salaam 5.3 7.4 28.1 17.6 Coastal 13.7 14.6 40.0 34.7 Nn Highlands 10.3 9.9 20.2 36.1 Lake 26.2 30.2 37.0 39.0 Central 7.3 10.3 48.8 42.4 Sn Highlands 10.9 14.7 46.6 25.8 South 8.3 14.8 43.9 43.2 Note: Bold indicates that the difference between the two surveys is significant at the 5% level. Source: Author's calculations using data from HBS 1991/92 and 2000/01. Poverty simulations suggest that the link between poverty reduction and economic growth is strong in Tanzania. The decline in poverty between 1992 and 2001 was limited because of modest economic growth and not because of a missing link between growth and poverty. Real GDP grew at an average by 3.6 percent, which combined with a high annual rate of population increase of 2.9 percent resulted in annual per capita growth of only 0.7 percent. The HBS data provides only two snapshots in time and fails to represent the full evolution of poverty between 1992 and 2001. By applying macroeconomic growth data to the micro-level HBS data, changes in consumption are simulated year-by-year using unit- record survey data, under varying assumptions for growth rates and inequality changes. The simulations suggest that poverty rose between 1992 and 1994 and then fell in the final years of the decade. The fall in poverty was faster in Dar than in other parts of the country. Sustaining and Sharing Growth in Tanzania 12 Figure 4. Simulated changes in poverty, 1992-2002 Simulated changes in poverty 2 survey approach, using within stratum dis .45 Fraction below poverty line .4 .35 .3 .25 .2 .15 .1 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Year Mainland Tanzania Dar es Salaam Other urban areas Rural areas Source: Author's calculations based on National Bureau of Statistics. 2002. and URT. Various Years. Growth has clearly reduced poverty but at the same time, higher inequality has worked in the opposite direction. Outside of Dar es Salaam, consumption growth was small and the increase in inequality almost negated it. In Dar es Salaam, inequality increased, but household consumption growth occurred as well, and the poor benefited substantially from this growth. Thus, despite the much larger increase in inequality in Dar, poor households gained more than in other areas where the increase in inequality was more modest. In other words, while the rising inequality in Tanzania is somewhat of an issue, increasing overall income growth across the country in the poor households is an even greater issue, and income inequality does not seem to have been a drag on growth in Dar es Salaam. Most of the poverty reduction during the 1990s has happened within areas. Table 6 shows the decomposition of the change in poverty into regional composition effects. There are inter-regional effects, population shift effects, and interaction effects.5 During the 1992- 2001 period, poverty declined substantially in Dar es Salaam and other urban areas. This attracted people from rural areas, resulting in a 4.6 percent population growth rate in urban areas, compared to 2.5 percent in rural areas. Despite these population inflows and the substantial urban poverty reduction, only 0.4 percentage points of the total drop in national poverty of three percentage points can be attributed to a shift in the population from poorer rural to wealthier urban sectors. While rural-urban migration may indirectly promote poverty reduction, e.g. by facilitating stronger rural-urban linkages, migration itself is not a major contributing factor. More generally, the spill-overs from urban growth to rural poverty reduction appear to have been modest. Sustaining and Sharing Growth in Tanzania 13 Table 6. Decomposition of change in poverty Tanzania Dar es Salaam Other urban Rural Poverty 1991 38.6% 28.1% 28.7% 40.8% Poverty 2001 35.4% 17.6% 26.0% 38.7% Total change in poverty 2001-1991 -3.2% -10.5% -2.7% -2.1% Growth impact -8.4% -18.4% -6.6% -5.3% Inequality impact 5.5% 12.4% 4.0% 2.7% Residual -0.2% -4.5% -0.2% 0.6% Urban- rural decomposition Change in poverty in Dar es Salaam -0.6% Change in poverty in other urban -0.3% Change in poverty in rural -1.7% Total intra-sectoral effect -2.6% Population-shift effect -0.4% Interaction effect -0.2% Types of employment Change in poverty farming/fishing -2.0% 1.1% -0.2% -2.5% Change in poverty paid employment -0.4% -7.9% -0.9% 0.2% Change in poverty self-employment -0.6% -5.8% -0.9% -0.2% Change in poverty family employment 0.1% -0.2% 0.2% 0.0% Change in poverty non economic activity 0.0% 0.1% -0.1% 0.1% Total intra-sectoral effect -2.9% -12.6% -1.8% -2.5% Population-shift effect -0.6% 0.6% -0.8% -0.7% Interaction effect 0.4% 1.5% -0.2% 1.2% Source: Author's calculations using data from HBS 1991/92 and 2000/01. Table 6 shows the power of a rural poverty reduction strategy for Tanzania. Even slight income growth for the rural poor has a large impact on overall poverty in Tanzania. The same holds true for households engaged in farming and fishing, as they are a large share of the population and their income is low. However, the last section of shows that the movement of households out of agriculture has also played a strong role in poverty reduction. In particular, moving from agriculture to other occupations in rural areas has contributed significantly to poverty reduction. In fact these structural changes in rural employment had a more significant impact on poverty reduction than rural urban migration. Considering that urban-rural growth and income differentials were large during the period under consideration, it is unlikely that the poverty impact of sustained high urban growth will significantly increase in the future. On the other hand, the impact of higher rural growth, both in agriculture as well as other income generating activities, is likely to be significant. This analysis underlines the importance of increasing rural incomes as the primary instrument for successful poverty reduction in Tanzania. Urban growth remains important for Tanzania, especially with respect to its role as an engine of structural transformation of the economy. However, the impact of urban growth on poverty reduction is clearly limited. Measures that support mobility, primarily mobility from agriculture to other sectors, but also geographic mobility are nonetheless important elements of the structural transformation process. Sustaining and Sharing Growth in Tanzania 14 OUTLOOK ON GROWTH ANDPOVERTY REDUCTION Tanzania's National Strategy for Growth and Poverty Reduction (NSGRP) has set ambitious goals for poverty reduction, which even exceed the MDG of halving poverty by 2015. An average growth rate of at least 6.1 and 5.0 percent would be necessary to halve poverty by 2010 and 2015, respectively. The pattern of growth and inequality is also important for Tanzania's prospects of halving poverty over the next five to ten years. If growth is biased towards urban areas, higher average growth rates will be necessary to reach the overall poverty reduction targets. Achieving such poverty reduction targets will also require policies and investments that ensure that economic growth creates income earning opportunities for the poor and enables them to take advantage of these. Within the framework of current policies and institutions, sustained per-capita GDP growth rates of at least four percent per annum seem to be achievable, which would translate into overall GDP growth rates between 6-7 percent. This assessment is based on policy based growth projections, where improvements in the institutional and policy framework translate into an improved outlook for economic growth. Measures of the quality of policies and institutions are provided by four independent sources: the World Bank's Country Policy and Institutional Assessment, the Institutional Investor assessment, the ICRG assessment and the Euro money assessment, which all suggest significant improvements between 1999 and 2006 (Table 7). Based on the country ratings for 2006, the projections of per-capita growth range between 3.8 and 5.0 percent, with an average of all four projections of 4.4 percent. If there are further improvements in institutions and policies, the projections suggest that even per capita growth rates of over 5 percent per annum are feasible. Table 7. Policy based growth projection Per capita Growth Projections Rating 1999 Rating 2006 Constant Improve+0.5 CPIA 3.5 3.9 5.0 6.1 EUROMONEY 2.2 2.8 4.1 4.8 ICRG 3.8 4.1 4.6 5.1 INSTITUTIONAL 2.0 2.3 3.8 5.0 AVERAGE 2.9 3.3 4.4 5.3 Note: Ratings are on a scale of 1 ­ 6 Source: Author's calculations based on data from Euromoney. Various Years. Institutional Investor. Various Years. ICRG. Various Years, and World Bank. Various Years. The input requirements for such growth rates in terms of investment, increases in the level of education in the work force, and sustained productivity increases are large, but seem to be achievable. For example, sustaining a growth rate of output per worker of four percent could be achieved with an investment to GDP ratio of around 18 percent, an increase in the average years of schooling by one year by 2011, and sustained increases in factor productivity. The expansion in the average years of schooling per worker would require sustaining further progress in primary school enrolments and especially completion rates, as well as a significant expansion in secondary school enrolments. Development and economic growth is characterized by structural transformation of the economy. The share of agriculture typically declines, while that of industry and services increases, and higher economic growth is typically associated with faster structural transformation. Table 8 illustrates the likely development of the sectoral composition of GDP under various growth assumptions, reflecting cross country experience on the association of economic growth and structural change. The baseline scenario projects current sectoral growth rates into the future. This scenario is compared to one with higher aggregate Sustaining and Sharing Growth in Tanzania 15 growth and faster structural transformation and one with lower growth and limited structural transformation. Table 8. Scenarios for economic growth and structural transformation Slow Growth Medium Growth Fast Growth Share in Avg. real Share in Avg. real Share in Avg. real Share in GDP growth rate GDP growth rate GDP growth rate GDP 2005 2006-2025 2025 2006-2025 2025 2006-2025 2025 Agriculture 45.6% 3.3% 40% 4.8% 34% 5.0% 26% Industry 19.7% 4.8% 23% 9.0% 32% 10.7% 32% Services 34.8% 4.4% 37% 6.1% 33% 9.0% 42% Total 100.0% 4.0% 100% 6.3% 100% 8.0% 100% Source: Author's calculations. Under the medium growth scenario, the share of agriculture in GDP drops to 34 percent by 2025 while the shares of industry and services increase to 32 and 33 percent respectively. Comparing these projections with international experience suggests that these sectoral projections are consistent with the pattern of structural transformation observed in other economies. In the high growth sceniario, annual average growth of 8 percent is the result of accelerated growth in the industry and services sector. Under this scenario, agriculture would only account for 26 percent of GDP in 2025, while the share of the industry and services sector would have increased to 32 percent and 42 percent respectively. In order for the economy to achieve the NSGRP growth targets, growth will have to accelerate in all sectors. In the medium term, the agriculture sector (including agro- industry) will remain central to Tanzania's economy and growth performance. However, as discussed below, growth in the agriculture sector should be primarily driven by productivity increases and labor saving technologies, which would increase per capita incomes in the agriculture sector. At the same time, an accelerated growth and poverty reduction scenario will require a gradual movement of labor from the agriculture sector to the more remunerative secondary and tertiary sectors. The combination of these factors suggests that growth in agriculture is likely to be lower than that of other sectors. Nonetheless, sustaining average agriculture sector growth of about 5 percent is both necessary and achievable over the medium term and will constitute the foundation of overall accelerated growth of the Tanzanian economy. A continuation of the strong growth performance of the manufacturing sector is critical. Growth in the manufacturing sector is important to realize Tanzania's poverty reduction objectives, as it provides access to higher paying jobs and strengthens the demand for primary products. The contribution of the manufacturing sector to GDP has been stagnant at around eight percent since independence. But, more recently, with an average growth rate of nine percent during the past five years, it has outperformed the rest of the economy. The growth potential of the manufacturing sector derives primarily from linkages to Tanzania's agriculture sector as well as the country's natural resource base, especially forestry, minerals and fish. Strengthening the backward linkages of the manufacturing sector with its resource base as well as the forward linkages, especially to export markets, is therefore essential for the development of the manufacturing sector. Analysis of enterprise survey data covering growth, investment, exports, and employment identifies the key constraints to more rapid growth. In 2002, in order of importance the four leading constraints were (i) access and cost of finance; (ii) access to Sustaining and Sharing Growth in Tanzania 16 superior technology; (iii) infrastructure; and (iv) supply of skilled labor and its productivity. Other constraints such as demand and the business environment were significantly less important for firm growth. The service sector, especially tourism, also has the potential to sustain high growth rates in the medium term. Official statistics suggests that Tanzania's strategy of marketing itself as an upmarket destination has been relatively successful. Tanzania's vast natural resources provide ample potential for further expansion, such as through the development of the Southern Circuit which includes the Selous Reserve. However, in order to sustain Tanzania's up-market image, investments in supportive infrastructure and human resources will be critical. PROGRESS TOWARDS THE MILLENNIUM DEVELOPMENT GOALS It will be challenging for Tanzania to meet all Millennium Development Goals (MDGs), yet there is room for cautious optimism in some areas. Data from the 2004 Demographic and Health Survey suggest that considerable progress was made in the reduction of malnutrition and child mortality. Policy simulations show that the income poverty and hunger MDGs may be attainable if Tanzania is able to sustain of medium high growth and with its improvements in the social sectors. Table 9. MDG baseline, most recent estimate and target. Year of Baselin Most baselin Year most e recent Target e recent Goal 1: eradicate extreme poverty and hunger Reduce extreme poverty by half ­ national poverty line 38.6% 35.6% 19.3% 1991 2000 ­ dollar a day poverty line 61.1% 57.5% 30.6% 1991 2000 Reduce hunger by half 29% 22% 14.5% 1991 2004 Goal 2: Achieve universal primary education Net enrollment in primary school 51% 91% 100% 1990 2004 Goal 3: Promote gender equality and empower women Equal girls' enrollment in primary school 1.01 0.99 1 1990 2004 Equal girls' enrollment in secondary school 0.70 0.81 1 1990 2000 Goal 4: Reduce child mortality Reduce child mortality of under fives by two thirds 141 112 47 1991 2004 Goal 5: Improve maternal health Reduce maternal mortality by three quarters* 529 578 132 1996 2004 Goal 6: Combat HIV/AIDS, malaria and other diseases Halt and reverse spread of AIDS n.a. 7.0% 2003 Halt and reverse spread of malaria 21% 36% 1999 2004 Goal 7: Ensure environmental sustainability Halve proportion without improved drinking water in urban areas 13% 10% 7% 1991 2000 Halve proportion without improved drinking water in rural areas 65% 54% 33% 1991 2000 Halve proportion without sanitation in urban areas 2% 4% 1% 1991 2000 Halve proportion without sanitation in rural areas 9% 8% 5% 1991 2000 * Maternal mortality is a so called low frequency event. Low frequency events are difficult to measure accurately in surveys like the DHS which serve as source for these estimates, as few cases of maternal mortality are registered. Consequently, maternal mortality rates are associated with large confidence intervals and the observed maternal mortality rates of 529 in 1996 and 578 in 2004 are statistically not different. Sustaining and Sharing Growth in Tanzania 17 With respect to the eradication of extreme poverty and hunger, the latest household survey carried out in 2000/01 shows that income poverty has declined from 38.6 percent in 1991 to 35.6 percent in 2000/01, using the national poverty line. If the international dollar a day poverty line is used, income poverty declined from 61.1 percent to 57.5 percent. Halving poverty by 2015 would thus require poverty to decline to 19.3 and 30.6 percent respectively. With the limited progress made during the 1990s (Table 9) this appears unattainable, but the data obscure that the early 1990s were characterized by low growth. The period of positive growth only materialized from 1995 onward so that by the end of the 1990s most of the damage from the earlier period was undone. The period of high growth since 2000/01 is not captured in the survey data. Yet poverty simulations (Figure 5), suggest that the income poverty MDG is attainable. Much will depend on the pace of rural income growth as most poor are found in rural areas. Figure 5. Projected changes in income poverty and malnutrition Prevalence of income poverty ($ day poverty line) Prevalence of underweight children under 5 65.0 30.0 61.1 . 29.0 55.0 25.0 45.0 45.7 (%) (%) 20.0 20.7 35.0 19.1 30.6 34.8 16.5 25.0 15.0 14.5 Poverty Slow growth 15.0 19.0 Slow growth Medium growth Underweight10.0 Medium growth 5.0 Fast growth Fast growth 5.0 -5.0 2001 2003 2005 2007 2009 2011 2013 2015 2004 2006 2008 2010 2012 2014 Note: Growth scenarios allow for different rural-urban growth rates. In the slow growth scenario GDP growth is 4.0 percent per annum (3.8 percent in rural areas, 4.6 percent in urban ones). In the medium and fast growth scenarios this is 5.4 percent and 8.0 percent respectively with rural growth rates of 5.0 percent and 6.5 percent and urban growth rates of 7.4 percent and 9.7 percent. Population growth is assumed to be 2.9 percent per annum. The income poverty projections are determined by applying GDP growth rates to unit record consumption information in the HBS. The nutrition projections are based on an income-nutrition elasticity of 0.51. The horizontal lines in both graphs present the baseline and the MDG objective, respectively. After a decade of stagnancy in which the share of underweight children remained unchanged at 29 percent, it declined to 22 percent between 1999 and 2004, so that Tanzania is on track with respect to attaining the hunger MDG. Income growth can explain at most a third of the decline. Malnutrition appears more responsive to changes in non-income factors such as a more effective management of malaria and vitamin A supplementation. Policy simulations (Figure 5) suggest that if income growth and especially improvements in nutrition related interventions can be sustained, the hunger MDG may be attained by 2015. Progress towards attaining the MDG on universal primary education has been impressive. As soon as universal primary education was introduced in 2001, net enrollment rates jumped and are now close (91 percent) to the target of 100 percent enrollment. In primary education gender inequalities are small (Table 9), but in secondary education gender inequalities persist. However, the recent increase in secondary education was much more gender balanced than previous enrolment patterns: in 2004 the girl-boy ratio in Form I was 0.98 while it was 0.55 in Form IV. Other non-MDG indicators such as the ratio of female to male prevalence of HIV/AIDS (of 1.22) show that gender inequality persists. Sustaining and Sharing Growth in Tanzania 18 Another encouraging fact is the observed reduction in child mortality from 141 per 1000 in 1991 to 112 in 2004. This drop occurs after stagnancy in child mortality rates during the 1990s (it was 137 in 1996 and 147 in 1999) and appears to be a reflection of improved malaria management (improved treatment as well as increased use of bed nets) and stronger health systems in general. The MDG target of 47 deaths per 1000 is ambitious however, and it will remain a challenge to meet this objective. With respect to the reduction of maternal mortality no progress has been made. The data in Table 9 even suggest an increase in maternal mortality rates but the difference is not statistically significant. The lack of progress is reason for concern as the MDG is ambitious (a 75 percent reduction) and maternal mortality rates high (578 per 100,000 births). A recent blood sample survey suggests that the prevalence rate of HIV/AIDS is 7 percent which is less than was expected based on tests performed on blood donors. There is much variation in infection rates by age group, gender and location. Women (7.7 percent) are more affected than men (6.3 percent). Prevalence is 5.2 percent in the 20-24 age cohort, but 10.9 percent in the 30-44 age cohort. Prevalence rates vary from a low of 2.0 percent in Manyara and Kigoma to a high of 13.5 percent in Mbeya and Iringa. Finally some limited progress was made in improving access to drinking water mainly due to increased access to protected wells and springs. Still some 54 percent of all rural households do not have access to improved drinking water and much needs to be done to attain the MDG objective of 33 percent. As far as sanitation is concerned almost all urban (96 percent) and rural (92 percent) households have access to a toilet. In urban areas, one observes an increase in the share of households without a toilet from 2 percent on 1991 to 4 percent in 2000, possibly as a result from rapid urbanization. Sustaining and Sharing Growth in Tanzania 19 2. ENHANCING INTERNATIONAL COMPETITIVENESS AND ACCELERATING DIVERSIFICATION The analysis of macro-economic developments clearly indicated that sustaining high economic growth will require a strengthening of Tanzania's international competitiveness and accelerated diversification of the range of products and services produced by the Tanzanian economy. This part of the report first presents an analysis of the factors that determine growth in Tanzania's manufacturing sector. Expansion of this sector is considered a central driver of sustained growth and structural diversification of the Tanzanian economy. This discussion of the manufacturing sector provides the basis for an in depth discussion of policy priorities for enhancing international competitiveness and for accelerating diversification. To achieve increased international competitiveness, improvements in infrastructure, access to finance, and the business environment are central. Accelerated diversification will require a strengthening of post-primary education, improvements in Tanzania's innovation environment, and increased access and use of ICT. FOSTERING GROWTH, EXPORT COMPETITIVENESS, AND EMPLOYMENT IN TANZANIA'S MANUFACTURING SECTOR A growth strategy based on supporting the manufacturing sector, and within it larger firms, especially exporters, has non-trivial poverty implications for the large proportion of unemployed and relatively less skilled Tanzanians. Faster rates of growth in the sector contributed to higher growth in the overall economy and the largest proportion of better paying jobs relative to employment in the agricultural sector where incomes typically fall below the poverty line. For example, in 2001, manufacturing generated about one-third of nonagricultural private employment. How can policy makers maximize the sector's growth potential? Empirically, larger firms are the drivers of growth and employment within the manufacturing sector. In absolute terms, in a sample of 264 firms, larger firms with more than 30 employees generated 29,000 jobs relative to only 1700 in smaller firms. Employment in larger firms grew at 8.5 percent per year relative to only 3.0 percent in smaller firms. Unsurprisingly, larger firms systematically paid significantly higher wages. The median wage for professionals in larger firms was Tsh. 240,000 relative to Tsh. 100,000 in smaller firms; for skilled workers, it was Tsh. 90,000 in larger firms compared to Tsh. 60,000 in smaller firms; and for unskilled workers, it was Tsh. 60,000 in larger firms relative to Tsh. 50,000 in smaller firms. Empirical analysis of the determinants of manufacturing firm growth shows that the two leading contributors to firm growth are a technically skilled labor force and growth of large firms who also represent the majority of the exporters. A 1 percent increase in the technical skills of the workforce increases firm growth by 0.70 percent and a 5 percent increase in the output exported delivers almost a 1 percent increase in firm growth, which, in turn, raises employment growth. The third important determinant of growth in manufacturing firms is investment growth (growth elasticity is 0.08). As the jump in firm growth is intricately tied to growth in exports because of limited purchasing power in the domestic market, an aggressive and proactive policy stance for promoting manufactured exports is likely to have the biggest effect on manufacturing growth in Tanzania. The rationale for that selective approach is motivated by today's global reality: if a firm cannot compete in the global market (that is, export), it is unlikely to survive for too long in Sustaining and Sharing Growth in Tanzania 20 Tanzania's domestic or Africa's regional markets, which are flooded with cheaper imports from low-cost, high-skills producers such as those from East and South Asia. Firm size is a critical determinant of firm growth in Tanzania. As many larger firms are also exporters, and most exporters are larger firms, policies targeting larger firms should have large payoffs in helping to expand existing firms and promote new entrants into the export sector. Although policies that favor large firms also favor exporters, export promotion strategies are important in their own right, given the limited purchasing power in the Tanzanian market. This recommendation requires a strategic, two-pronged approach: one that targets large firms, and another that targets existing and potential exporters. Policies need to address domestic supply constraints associated with disproportionately higher transaction costs of investment faced by large firms and exporters, and reduce obstacles associated with finance, infrastructure, technology, and skills, which emerged as the key constraints to growth from the analysis of the investment climate assessment. All of that may seem tantamount to recommending everything--that is, redressing all barriers to production currently facing all manufacturing firms in Tanzania. It is not. To circumvent the high financial and time costs and the government's limited implementation capacity requires focus and pragmatism in catering to larger firms. The targeted delivery of physical inputs such as infrastructure would be facilitated by the identification of spatial locations where manufacturing and export activity is most prevalent. Spatial identification also helps in targeting exports. Similarly, to attract potential exporters, the government could identify special areas, such as industrial districts and export processing zones, where larger firms interested in exporting would be able to benefit from improved infrastructure and service delivery. That approach would render public support in a financially feasible and timely manner for fast-growing exporters and potential new entrants into the export business. The manufacturing sector cannot afford to wait until the constraints to investment are resolved economy-wide. Although the entry of firms into the export sector can be facilitated, much more is needed to sustain them in the business. Creating and nurturing firms' ability to export implies grappling with the challenge of improving productivity and spurring competitiveness to export to non-African markets, where competitiveness is best tested. That challenge requires a set of externally oriented policies that exploit Tanzania's latecomer advantage to leapfrog into the global marketplace. The starting point could be direct public support to facilitate the delivery of the two key public goods: (a) superior technologies of production through adaptation and (b) development of technical, tertiary, and managerial skills needed to apply them. Public-private partnerships have served as the best mechanism to deliver those two public goods. The ability of Fundación Chile to promote technology transfer and adaptation and that of the Indian Institutes of Technology and Management to deliver critical skills offer useful lessons. Additionally, financial incentives are needed to reduce the high fixed cost of entry into export markets and to attract firms and sustain them in the export business. Special incentives to promote exports out of Africa are likely to have the highest payoff and sustainability. Lessons from East and South Asia are good starting points, especially in the area of agroprocessing and light manufacturing. IMPROVING THE BUSINESS ENVIRONMENT The quality of the business environment affects the cost of doing business, and is an important determinant of a country's attractiveness for investors and its international competitiveness. For Tanzania, the cost of an inefficient business environment are estimated to be very high, amounting to 25 percent of sales, including the cost of contract enforcement difficulties, regulation, bribes, crime, and unreliable infrastructure (Figure 6). Sustaining and Sharing Growth in Tanzania 21 Figure 6. Cost of inefficiencies in business environment as percent of sales, various countries Source: World Bank, 2004a. The report focuses on three aspects of the business environment in Tanzania that promise the highest gains in terms of accelerated economic growth if appropriate actions are taken. These include the provision of complementary infrastructure, the cost and access to finance, and the cost arising from bureaucracy and corruption in the interaction of the public with the private sector. Infrastructure Tanzania's weak infrastructure, especially in transport and the power sector, impedes the competitiveness of the private sector and limits economic expansion. The levels of infrastructure provision in Tanzania are among the lowest in Africa (Figure 7) and returns to infrastructure investment are generally high, although they vary considerably by region, type of investment, and time period of the investment. The highest rates of return on selected projects are 104 percent in the road sector, 23 percent in the power sector and 50 percent in telecommunication. Sustaining and Sharing Growth in Tanzania 22 Figure 7. Low levels of infrastructure development constrain economic growth Roads, paved (% of total roads) Electric power consumpption (kwh per capita) 100 90 9000 80 8000 70 7000 60 6000 50 5000 40 4000 30 3000 20 2000 10 1000 0 0 income income income Tanzania income income income Low TanzaniaLow High Middle High Middle Source: World Bank. Development Data Platform. A reliable power supply is of particular importance for economic growth in industry and the service sector and priority should be given to enhancing power supply in urban areas. Businesses in Tanzania suffer high costs from an unreliable power supply due to the need to have their own back-up facilities and interruptions of production due to power outages. Due to the changed situation in international energy markets, the current energy sector policy may require revision to be more focused on a long term vision for the growth of the sector, including rural electrification and connection to regional grids. This vision should be underpinned by a realistic investment plan based on the financing expected to be available from various sources (Government, development partners and the private sector). It would be reasonable for the new policy to include a much less complex plan for the restructuring and possible divestiture of TANESCO. An improved road network is critical for enhancing agricultural incomes by increasing access to local markets and reducing costs. In addition, any efforts to promote agro- processing will require enhanced linkages between rural areas where agricultural production and processing takes place, and national and international markets. The maintenance and rehabilitation of existing roads based on economic criteria needs to be a priority. The passing of an effective Roads Act, defining the various responsibilities in the sector and the financing mechanisms, and the transformation of TANROADS into a Road Authority, are critical to establish an appropriate framework for increased expenditures in the road sector and ensuring its effectiveness. While financing for the road sector will primarily be in the form of public and donor financing, rehabilitation and expansion of the rail and port infrastructure offers greater potential for private sector involvement. Recommendations. In order to enhance infrastructure services in Tanzania to meet the long- term growth targets, increased spending for both investment and maintenance of infrastructure is required. The following principles are important to ensure that increased infrastructure expenditure yields the needed impact on economic growth: · Establishment of an appropriate policy and institutional framework is critical to ensure the efficacy of infrastructure expenditure. In particular, the enactment of appropriate electricity and road acts are critical for the development of infrastructure in Tanzania and improving efficiency. Sustaining and Sharing Growth in Tanzania 23 · Tanzania's infrastructure requirements are large and must be properly prioritized. This prioritization should be primarily based on the expected rates of return to infrastructure and poverty impacts. · Infrastructure needs and constraints in rural areas vary, reflecting the diversity in types of economic activities, population densities, existing infrastructure endowments and remoteness: The policy implication of this diversity is the need for regionally targeted infrastructure investments that respond to local priorities. Part of the strategy could be to cluster infrastructure investment in specific areas which would accelerate growth by enabling producers to exploit economies of scale. · Appropriate balance needs to be maintained between new investment, rehabilitation, and recurrent maintenance. Currently, road sector spending is still insufficient to maintain the existing road network. Building new roads is expensive and, thus, appropriate maintenance of existing roads is necessary to ensure cost effectiveness and sustainability of the growth and poverty impacts. · Strengthening regulatory institutions and ensuring their independence is critical to attract private sector participation in the infrastructure sectors. · Scaling up of investment in infrastructure may have adverse short-term macro- economic consequences: The high domestic content of large scale infrastructure construction such as roads, may affect relative prices between tradables and non- tradables. In the longer run, however, the impact is likely to be more than compensated by the economy wide productivity gains from the investment. Nonetheless, this is a potential down side risk that would need to be managed carefully. To avoid and mitigate Dutch disease effects in the short term, attention needs to be paid to the import content of infrastructure expenditure and its gestation period. A higher import content reduces the likelihood and magnitude of Dutch disease effects while expenditures with a shorter gestation period would accelerate the compensating productivity gains expected from infrastructure investment. Expanding access and reducing the cost of credit The current depth and efficiency of Tanzania's financial system falls well short of what is needed to help support economic growth and ensure the efficient allocation of resources. The econometric analysis of the enterprise survey data suggests that access to finance is the main constraint for growth and investment of firms. Analysis of the agriculture sector similarly suggests that access to credit is an important constraint for agricultural production. Access to credit has seen dramatic changes during the past decade. Credit as a share of GDP has declined dramatically from about 35 percent to GDP in 1993 to only 7 percent of GDP in 2002, has, however, recovered subsequently and reached 12 percent in 2005 (Figure 8). Most of this decline was due to the fiscal consolidation, which resulted in a reduction in credit to the public sector from 23 percent of GDP in 1993 to 0.2 percent in 2004. Credit to the private sector contracted from about 15 percent of GDP to only three percent in 1996. However, since then it has steadily recovered and stands now at ten percent of GDP. Sustaining and Sharing Growth in Tanzania 24 Figure 8. Domestic credit and interest rates, 1990-2005 40 20 35 (%) 15 30 rate 10 25 GDP 5 of 20 % 15 interest 0 10 Real -5 5 -10 0 1990 1992 1994 1996 1998 2000 2002 2004 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Central government Private sector Other T-bills Lending Saving Source: Bank of Tanzania. Quarterly Economic Review. Various Issues. Figure 9. Savings and Investment (% of GDP), 1990-2005 30 25 20 15 GDP of 10 % 5 0 -5 -10 1990 1991 1992 19931994 19951996 19971998 19992000 20012002 2003 2004 2005 Investment Domestic Saving Source: URT. Economic Survey. Various Years. Access to formal credit is, however, still mainly limited to enterprises in Dar es Salaam and other urban areas and collateral requirements are high. A weak legal system and the lack of a developed credit information system make it difficult for banks to assess and price risk and there may be thus a tendency to limit credit to enterprises with a relatively well established credit history and strong collateral rather than moving into the high risk, high interest segments of the market. Formal credit in rural areas is very limited. For commercial banks operating in rural areas, most of the collected savings are transferred to Dar es Salaam to be invested in assets that have a more attractive return risk profile than rural investment opportunities. Credit by traders and marketing organization plays an important role in agriculture, with interest rates significantly higher than commercial lending rates. Domestic saving has shown a dramatic increase from minus five percent of GDP in 1993 to about eleven percent in 2005 (Figure 9). The increase in saving is almost entirely the result of increased public sector saving. Between 1990 and 2005, public sector consumption declined from 18 percent of GDP to seven percent. Private sector consumption increased initially from 83 percent of GDP in 1990 to 88 percent of GDP by 1999 but has declined subsequently to about 80 percent in 2005. The difference between saving and investment, which corresponds to the current account deficit, declined dramatically since the early 1990s. A key issue is the distribution of liquidity. Although aggregate indicators suggest excess liquidity, this is not the case for all the banks. Ninety percent of the excess liquidity is Sustaining and Sharing Growth in Tanzania 25 concentrated in three banks, and a large part of this liquidity comes from government deposits. Moreover, a large portion of this liquidity is invested in government securities. Financial markets are affected by the sterilization of aid inflows by issuing T-bills. The interest rate appears to be quite sensitive to changes in the supply of T-bills. At the same time, a fairly high rate of monetary expansion has been consistent with declining rate of inflation. This suggests that the economy may be able to absorb aid inflows without the need to fully sterilize them. Recommendations. Enhancing the access to finance requires (a) policies that foster saving; (b) monetary and fiscal policies that support growth in credit to the private sector; and (c) policies for the development of the financial sector. With regard to the latter, government-- in collaboration with the Bank and Fund--has prepared a Second Generation Financial Sector Reform Program. Implementation of the reform program involves the following elements: · Strengthening the lending environment and financial infrastructure. This includes completing the task of divestiture of state-controlled entities in banking and insurance; strengthening the legal and judicial framework supporting lending; clarifying and deepening the regulatory, information and technology infrastructure for households and micro enterprises, and encouraging long-term pension and insurance funds to finance longer-term private investments. · Facilitating the increase in SME and Long Term Lending. Government is taking initiatives in five areas: (a) establish a SME Credit Guarantee Scheme to encourage commercial bank lending for SMEs; (b) launch a Development Finance Guarantee Facility (DFGF), to be managed initially by the BOT, which will provide partial government guarantees to commercial banks for their loans to development- and export- oriented projects; (c) facilitate the creation of a privately owned and managed long-term financing facility (LTFF) that would channel funds from non-banks or banks and potentially development partners (without government guarantee) to be on-lent to commercial banks on a long-term basis; (d) introduce a Development Finance Institution (DFI), most likely incorporating the Tanzania Investment Bank. The DFI would channel multilateral and bilateral donor funds and perhaps utilize government seed money from the budget, but not take any new deposits from the public; and (e) advance reforms in the pension fund sector in order to have a unified legal and regulatory framework for all pension funds, along with investment guidelines. It is expected that this effort, particularly the development of investment guidelines, will facilitate the channeling of pension funds' resources into longer term lending through commercial banks. · Direct support to providers of financial services for micro and small enterprises. The Second Generation Financial Sector Reform Plan includes initiatives related to micro and rural finance. These initiatives are designed to respond to the government's vision for the development of pro-poor finance in Tanzania, as articulated in the National Microfinance Policy. Some of these initiatives, such as the strengthening of the regulatory framework will be addressed as part of the strengthening the lending environment and financial infrastructure. However others will require direct support to the providers of financial services. It is expected that a large part of this support will be provided by the Financial Sector Deepening Trust (FSDT) funded by four bilateral development partners.6 The FSDT will provide assistance for the transformation of microfinance NGOs, the strengthening of networks of Savings and Credit Cooperatives (SACCOs), and the development of links between banks and microfinance institutions. Sustaining and Sharing Growth in Tanzania 26 Achieving the objectives of the Second Generation Reform Program requires the timely implementation of the action plan. Considerable efforts have gone into developing the strategies and building consensus. The challenge is now to begin implementing the plan. Enhancing the Public-Private Sector Interface Regulatory agencies, tax revenue authorities including customs, business and land registries, and the judicial system all form part of the public interface with the private sector that have an important bearing on the cost, risks and barriers to business in Tanzania. These costs will influence the range of opportunities that are profitable, because investments are forward looking, risks and uncertainty will determine the types and nature of investments, and entry restrictions will limit innovation and the efficient provision of goods and services. Aside from the impact on the cost of doing business, the quality of the public private interface is critical. This relates to the flow of information between the private and the public sector, which allows government to play a supportive role to the private sector by removing obstacles, collaborating in the identification of growth opportunities, and ensuring that the provision of public goods and services (especially infrastructure) is well aligned with private sector needs. An efficient public private interface is part of the second generation reforms that will determine the private sector's response to opportunities that are productivity based. The size of the informal economy is estimated to be large in Tanzania, suggesting high barriers to entry (Table 10). However, it also suggests the existence of entrepreneurial potential which could become an important driver of growth, if an environment for growth and formalization of informal activities is established. Among the main causes for informality are relatively high cost of registration and licensing as well as tax obligations implied by formal registration. Recent reforms of business licensing which abolish fees for small businesses are an important step to reduce barriers to entry. Table 10. Estimated size of the informal economy (% of GDP), various countries and regions Country Size of Shadow Economy Region/Country Size of Shadow Economy (% (% of off. GDP) Group of off. GDP) Tanzania 60% Africa 42% Kenya 36% Asia 31% Uganda 45% South America 43% Mozambique 42% Central Europe and 38% FSU South Africa 30% OECD 16% Source: Schneider 2004. Poor customs administration and overly restrictive trade and customs regulations discourage enterprises from exporting. Trade and customs regulation partially explain why enterprises in Tanzania export less than similar enterprises in Kenya. Since exporting has been linked to improved productivity, these delays and restrictions can have a real impact on enterprise performance. The 2003 enterprise survey revealed that among enterprises that engaged in foreign trade, the median firm reported that it took 14 days on average for imports and 7 days for exports to clear customs, once they had reached the point of entry or exit.7 Managers also reported that clearance times were unpredictable, forcing them to hold additional inventory in anticipation of worst-case scenarios. Port and customs delays are considerably longer in Tanzania than in any of the comparator countries. The median delays for imports and exports in China were five and three days respectively--less than half the delays facing firms in Tanzania. Similarly, reported delays for imports and exports were seven and four days respectively in Kenya and seven and three days respectively in India. Sustaining and Sharing Growth in Tanzania 27 Reform of the country's legal sector to support and enhance private sector led economic growth and efficiency has not taken place at the same pace as economic reform. Problems affecting the legal sector identified by the Legal Sector Task Force in 1996 include inordinate delays in resolving disputes and dispensing justice in the justice system, inaccessibility of the justice system for the majority of poor and disadvantaged Tanzanians, low levels of public trust in the justice system and excessive prevalence of unethical behaviour.8 The key challenges derive from the rapid social, political, economic and technological changes taking place in the country and internationally, and the need to for the legal sector to catch up with these changes to facilitate the efficient development of a private- sector and market-led economy, and for due protection of consumer rights. Reforms of the legal system to date have been slow and piecemeal. The establishment of commercial courts in 2000 initially resulted in faster dispute resolution for businesses. However, even the commercial courts became quickly overburdened with large caseloads, which suggest the need for speeding up overall reforms and strengthening of the judicial system. Grand and petty corruption have a negative impact on competition and impose high cost on businesses. There has been some progress in reducing corruption since 1996, when a presidential commission led by Judge Warioba produced the Warioba Report. Government is implementing a National Anti-Corruption Strategy and has strengthened the institutional framework--notably through the Finance Act of 2001 and the Public Procurement Act of 2004--and adopted a clear zero-tolerance position on corruption. Nonetheless, corruption continues to have an important impact on businesses. Enterprise managers who participated in a recent enterprise survey (World Bank, 2004b) see grand corruption--payments made to policymakers or senior bureaucrats in order to win government contracts and influence lawmaking--as a serious problem. Enterprises are also affected by petty corruption-- payments made to lower-level government officials to "get things done" in connection with customs, taxes, licenses, and other services. About 35 percent of enterprise managers said that informal payments were typically needed for firms like theirs. Figure 10. Requests for bribes are especially common during tax inspections 30% 375 25% 400 21% 20% 18% 300 13% enterprises) 200 110 8%115 thousands) of 10% 7%75 50 80 100 (% (TSh 0% 0 Tax security police social Fire/building Sanitation & Municipal Environmental Labor Percentage reporting bribe requestsMedian amount Note: "Percentage reporting bribe requests" only includes enterprises that were inspected; "median account" only includes enterprises that reported positive amounts Source: World Bank 2004b. Sustaining and Sharing Growth in Tanzania 28 Recommendations. Establishing a business environment which encourages the entry of businesses into the formal sector and which supports competitiveness by minimizing transaction cost is critical for Tanzania's economic growth. Priorities of reform include: · Revision of licensing and registration legislation to consolidate recent reforms of business licensing; · Review of tax regime for micro and small enterprises; · Strengthening of customs and trade regulations and administration; · Accelerating legal sector reforms and strengthen capacity in the sector; and · Revision of anti-corruption legislation to provide more effective instruments for the control of grand and petty corruption. FOSTERING INNOVATION, PRODUCTIVITY, AND TECHNOLOGICAL CHANGE Tanzania's prospects for sustained economic growth depend significantly on the capacity of the private sector to innovate and to translate innovation into new business activities. Innovation is the key driver of economic growth rather than a by product of the growth process. In the Tanzanian context, innovation covers the expansion of the range and quality of products and services and the adoption of technologies that enhance the competitiveness of the Tanzanian economy, at the firm, industry, and national level. At the micro level, there is the adoption and adaptation of available technologies for use by firms, individuals or households which helps to improve their productivity, welfare, living conditions, and so on; at the sectoral level, there is the development of new industries, generally based on foreign technologies that can be a source of new jobs, income and exports. There are four pillars that are generally considered to be important for countries to support the process of structural change: · An economic and institutional regime that provides incentives for the efficient use of existing knowledge, creation of new knowledge, and the flourishing of entrepreneurship. · An educated and skilled population that can create, share, and use knowledge well. · An efficient innovation system of firms, science/research centers, universities, think tanks, consultants and other organizations that can tap into the growing stock of global knowledge, assimilate and adapt it to local needs, and create new knowledge. · A dynamic information infrastructure that can facilitate the effective communication, dissemination, and processing of information. The following sections discuss these key drivers of innovation and productivity and provide recommendations aimed at enhancing the pace of innovation and productivity in Tanzania. Investing in education Tanzania's economy does not have a sound base of adequately qualified and trained work force that is essential for rapid economic growth and effective diversification of its production and export bases. But Tanzania has made some progress in education. Figure 11 shows that in 2001, Tanzania's adult literacy rate at 77 percent was higher than that of Sustaining and Sharing Growth in Tanzania 29 Uganda (69 percent), but lower than that of Kenya (84 percent), Botswana (79 percent), or South Africa (86 percent). In addition, according to Cohen and Soto, in 2000 (Figure 12), Tanzania's average years of schooling (3.4) was higher than that of Uganda (3.22), but lower than in Kenya (5.08), and far from that in South Africa (7.22). Figure 11. Adult Literacy Rates, 1960-2000 Adult Literacy Rates 85 Mauritius 75 South Africa Kenya 65 Malaysia Botswana 55 Uganda Tanzania 45 Ghana Tanzania Ghana 35 Kenya Uganda Botswana Mauritius Malaysia South Africa 25 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 Source: World Bank. Development Data Platform. Figure 12. Average Years of Schooling, 1970-2002 Average Years of Schooling (Cohen-Soto) 9 Tanzania Ghana Kenya Uganda Mauritius Malaysia Malaysia 8 South Africa 7 Mauritius 6 Kenya 5 South Africa Ghana 4 Tanzania 3 Uganda 2 1 1960 1970 1980 1990 2000 Source: Cohen and Soto, 2001. Tanzania's gross enrollment ratio (GER) for primary education increased from 78 percent in 2000 to 113 percent in 2006, following the abolition of primary school fees and the roll out of the primary education development program. A recent assessment (SACMEQ II) suggests that the quality of primary education is not worse than in other African countries. A key challenge is to maintain and enhance the quality of primary education while the sector expands. Tanzania has still one of the lowest secondary net enrollment ratios in Sub-Saharan Africa, despite rapid progress in recent years. Only about 13 percent of the relevant age group attended secondary education in 2006, compared with an average of 27 percent for Sustaining and Sharing Growth in Tanzania 30 Sub-Saharan Africa. To improve access and quality of secondary education, the government has launched the Secondary Education Development Program (SEDP) in 2004. Levels of tertiary education are also very low in Tanzania. The tertiary GER stood at 1.2 percent in 2006, as compared to 3.24 percent for Uganda and 3.52 percent for Kenya in 2001, while Botswana and South Africa had tertiary GERs of 4.69 and 15.05 percent, respectively in 2002. The government has been the major financier of technical/vocational education and training (TVET), but the system in Tanzania faces several problems in terms of inefficient resource utilization, inequitable distribution of educational opportunities, poor labor market linkages, and a lack of coordination between donors and the government. Analysis of the Integrated Labor Force Survey data suggest that marginal returns are around eight percent for every year of primary education, but increase sharply for additional years of lower secondary education. This data provides strong support for government's investment in expanding primary education and the current focus on secondary education (Figure 13). Figure 13. Marginal social returns per year of education based on integrated labor force survey 25.0% 20.0% 15.0% 10.0% 5.0% 0.0% Primary Low Secondary Up Secondary Univ Source: Author's calculation based on ILFS data Analysis of rates of return to education in the manufacturing sector suggests that marginal returns to education increase significantly for higher education compared to primary and secondary education (Figure 14). The difference in the profile of marginal social returns to education for workers in the manufacturing sector and the overall laborforce may suggest that at the tertiary level there are certain degree programs that are well rewarded by the manufacturing sector, but that many other degree programs result in lower paying jobs. The implication from this would be to shift the supply of higher education to those programs that seem to be in demand by the manufacturing sector. The difference in the profiles of social returns to education also suggests that a limited supply of workers with relevant tertiary education is a constraint for manufacturing, while for other sectors the limited supply of Tanzanian's with secondary skills may be more of a constraint. Sustaining and Sharing Growth in Tanzania 31 Figure 14. Predicted earnings in manufacturing sector based on manufacturing firm surveys Source: Soderbom et al. (2004) Tanzania is no stranger to the brain drain phenomenon. The most vulnerable occupations at the national level include medicine, accountancy, law, engineering, and science based occupations. Low salaries for doctors are the principle reason driving the brain drain--and those that remain seek higher wages in private hospitals in large urban centers, leading to a lack of doctors in some of the country's district hospitals. In addition, HIV/AIDS poses significant risks for human resource development in Tanzania. Recommendations. Key challenges facing Tanzania in education and human resource development include: · Sustaining and improving the quality of education as enrollments increase by recruiting teachers, constructing classrooms, increasing preservice teacher training, and providing subsidies for purchasing teaching and learning materials. · Ramping up secondary education, including improving its quality and relevance to the needs of the economy. · In higher education, strengthening the governance and administration of the country's three public universities in terms of financial sustainability, up-to-date content, and teacher training. · Using the potential of distance education to expand access to education services while improving equity. The Open University of Tanzania offers degree programs by correspondence and also in regional centers. The costs are low, because the state covers tuition. But enrollments are low, partly because of lack of content and partly because of a dearth of partnerships with international academic institutions that could provide degree programs online. Combining distance-education modalities with extended face-to-face interactions with Tanzania's other public universities may be one way to boost enrollments and increase access to higher education. Sustaining and Sharing Growth in Tanzania 32 · Reforming teaching methods and the curriculum at all levels to include skills and competencies (communication skills, problem-solving skills, creativity, and teamwork) to meet the new needs of the economy. · Increasing the interface between industry and education and offering differentiated curricula that better meet the new skill demands of industry, generated by changing markets and technologies. · Harmonizing the technical education offered in secondary schools with that offered in technical colleges and then linking these schools with the proposed zonal and regional institutes and colleges. These institutes and colleges should offer differentiated products to meet the differing needs of industries, such as mining, fisheries, major cash and food crops, external trade, and metal. · Devising strategies to proactively deal with problems of skills lost through brain drain. Strengthening the innovation environment A recent study by the World Bank (Chandra 2006) identifies a host of government actions that are essential to the development of competitive industries (Figure 15). In order of importance, firstly, these include actions that make available appropriate skills and provide support for technology acquisition and development. This is followed by actions of a regulatory nature, such as relating to standards and quality control. Then come various types of support measures that can be provided to enterprises and industry organizations for export promotion, investment, and so on. Table 11. Role of the Public Sector in Fostering Innovation Nile Cut Electronic Electronic IT Maize Grapes Oil palm Salmon Wine Perch flowers - ­ ­ - ­ ­ ­ - ­ - Malaysia Taiwan India India India Malaysia Chile Chile Uganda Kenya Spin-offs X X X X X Export & inv X X X X prom. Tech X X X X X X X X X acq/development Regulation/ X X X X X X X compliance Support to X X X X X Industry organization Tech skills X X X X X X X X X X development Source: Based on information in Chandra (2006). In Tanzania, government institutions in support of innovation lack adequate resources, infrastructure, equipment and trained personnel to respond to the increased needs of the local entrepreneurial society. The low level of R&D as percentage of GDP--only 0.2 percent (comparable to other African countries), reflects the modest nature of Tanzania's research and innovative effort. There are currently about 62 R&D institutions in Tanzania covering agriculture, including livestock and forestry (28), industry (11), medical (11), Sustaining and Sharing Growth in Tanzania 33 wildlife and fisheries (4) and Universities and other higher learning institutions (9). In addition to the lack of resources for R&D institutes, mechanisms for technology diffusion are modestly developed, with a quasi absence of decentralized structures. FDI is an important source of technological upgrading in Tanzania. FDI--which is relatively high in Tanzania compared to other African countries--has played a key role in the modernization of important sectors of the economy such as trade (retail), banking, tourism and the telecom networks. It has also been crucial in the take-off and growth of new industries such as the fishing and the gold mining industries. However, international experience suggests that it takes time to build an indigenous innovative capability through foreign investment and that this requires explicit mechanisms such the employment of large contingents of local cadres at managerial positions as well as programs to closely link local suppliers of components and materials to upgrade their equipment and the quality of their products. Mechanisms such as these do not exist in Tanzania, and consequently, the transfer of knowledge and technology from foreign sources remains modest. Recommendations. In addition to improving the overall business environment and the upgrading the education system, specific actions for the promotion of innovation and technology diffusion are needed in order to put Tanzania's R&D infrastructure in the service of the country's development. There is a need for a systemic approach that provides complementary support on three basic aspects: financial, technical and regulatory. Financial support: There is a crucial lack of resources on the demand side for accelerating the design, testing, use and dissemination of technologies in the Tanzanian context. It is suggested that Tanzania establish two complementary schemes, based on simple matching fund principles, that provide--in grant form--50 percent of the funding required for the development (R&D phase) of small and medium sized projects (for example, up to US$20,000): · User scheme: This scheme would allow particular groups and communities to buy needed technologies, and provide, if appropriate, complementary in-kind funding (labor for community purpose). · Developer scheme: This scheme would fund 50 percent of technical services or R&D projects undertaken by SMEs with R&D institutes (public, academic, etc), thereby helping indirectly, but more effectively, the R&D institutes use their competences to serve communities (note that this type of scheme is in place in a number of advanced countries). Technical support: On the basis of experience accumulated in industrial countries, it is suggested that a network of locally based and owned structures be established to serve the needs of rural and urban communities for technical advice, information, and assistance (in design, marketing, and so forth). These structures should be adapted to different sectors (for example, extension services for agriculture and design and manufacturing workshops for industry). They should also be conceived and operated as antennas of central bodies to which they would be strongly connected through information technology (IT), databases, and the like. They should be established on a clearly expressed demand from local communities and funded on a 50/50 cost-sharing basis, with local organizations (municipalities, business or farmer associations, and so forth) matching the resources put in by the central government. Regulatory support: Regulatory-related actions should address several issues: Sustaining and Sharing Growth in Tanzania 34 · Proposed actions that aim to stimulate service-based contracts and formalize new links between the business sector and the R&D infrastructure require the establishment of clear legal and administrative procedures. We therefore recommend reviewing, adjusting, and standardizing appropriate models for such relationships. Many issues may be involved, including the use of public or university laboratory equipment and personnel by firms, the temporary employment of university researchers by business enterprises, and intellectual property rights. · In many sectors, there is a need to develop quality awareness and quality control, as well as related accreditation and certification procedures. A program should be implemented to raise awareness of these issues, because doing so could yield important results in a short time span. · Firms or individuals that are first producers face major financial problems and frequently do not get access to credit from the banking system. The government's recently established credit guarantee mechanism, which is supposed to mobilize the banking and financial sector, does not seem to be working well. An audit needs to be undertaken to examine in detail the mechanisms that can be put in place to complement this incentive, such as microcredit schemes; equity investment procedures (such as the Dutch Program for Cooperation with Emerging Markets, which supports 30 percent of the investment of individual firms in the flower industry); and the like. These reforms could be supported through a multipurpose innovation facility that would target specific priority industries, including the tourism sector and the agro-food industry. The first is already an important source of income for Tanzania; the second would take advantage of the large agriculture base of the country. Systemic, well focused action would be needed in both cases. The involvement of foreign enterprises is crucial. These enterprises provide access to foreign markets and management competencies, and introduce up to date technologies. It is important to establish in liaison with Tanzanian Enterprise Associations, efficient organizations for negotiating and partnering with foreign businesses. Clear contracts regarding technology licensing, personnel training, and access to export markets should be developed. On the whole, a two pronged action strategy combining both the upgrading and developing domestic capabilities and involving foreign actors would be essential for a successful innovation intervention to help improve Tanzania's growth prospects. Enhancing the roll-out and use of ICTs Rapid advances in ICTs are dramatically affecting economic and social activities, as well as the acquisition, creation, dissemination, and use of knowledge. These advances are affecting the way in which manufacturers, service providers, and governments are organized and how they perform their functions. As knowledge and innovation become increasingly important elements of competitiveness, the use of ICTs is reducing transaction costs, and time and space barriers, thereby allowing the mass production of customized goods and services and substituting for limited factors of production. Compelling evidence suggests that strengthening telecommunications infrastructure and service is pivotal in promoting trade and economic growth. Fink et al. (2002) estimate that a 10 percent decrease in the bilateral price of phone calls is associated with an 8 percent increase in bilateral trade. In Africa, a telephone growth rate of 10 percent instead of 5 percent (and growth in electricity generation of 6 percent instead of 2 percent) could generate an increase in Africa's growth by at least 0.9 percentage points. Sustaining and Sharing Growth in Tanzania 35 The information infrastructure in a country consists of telecommunications networks, strategic information systems, policy and legal frameworks affecting their deployment, as well as skilled human resources needed to develop and use it. In the ICT domain, Tanzania is still at a very nascent stage of ICT application and use. Progress has, however, been made under the 1997 National Telecommunications Policy, and this trend is expected to continue. The mobile telephone market for one is fully competitive in Tanzania. Significant liberalization has taken place in various segments: private operators are now providing basic, mobile, data, paging, internet, pay-phone and other value-added services. The mobile- telephone market involves a number of operators, and is growing rapidly. Table 12 shows that overall teledensity--mainlines plus mobile phones--has increased to 24 per 1,000 in 2003. Anecdotal evidence suggests that mobile phones are increasingly being used in Tanzania to get business-related information and to reduce transaction costs. For example, traders in Dar es Salaam now can place orders with producers of bananas throughout the country--thus linking demand and supply in real time and enhancing the efficiency of markets. Sustaining and Sharing Growth in Tanzania 36 Table 12. ICT Indicators (per 1,000 people), 2002 1996-2003 2002 2000 2003 Mobile Personal Internet Internet Radios Mainlines telephones Televisions Computers Hosts Users (per 1000) (per 1000) (per 1000) (per 1000) (per 1000) (per 10,000) (per 1000) Botswana 87 241 150 44 40.7 13.99 50* Kenya 10 42 221 26 6.4 .32 200* S. Africa 107 304 336 177 72.6 41.94 2,890* Tanzania 5 19 406 4 4.18 .16 240 Uganda 2 16 122 18 3.32 .07 125 Sub-Saharan 15 37 198 69 11.90 3.10 6,233 Africa average * denotes data for 2001 Source: World Bank. African Development Indicators 2005. Despite increasing competition, tariffs remain high and teledensity is one of the lowest in the SADC region. This is mainly due to the poor interconnection framework, lack of regulatory independence, and other issues such as lack of infrastructure sharing. Thus, in general, the country's postal and telecommunications services are weak, and the provision of telephone lines (fixed lines) has been meager. An inadequate regulatory framework persists, and competition has been hampered by various issues, such as inadequate interconnection agreements/directives, high level of fees and royalties levied by the TCRA, and absence, or non-transparency of regulatory oversight. Recommendations. In order to strengthen its information infrastructure, the following issues are critical: · Finalizing and adopting the new electronic communications bill, which is key to defining the ground rules for sector development (including rural areas). · Implementing the new converged licensing framework, which will ensure further liberalization of the market. · Reviewing and modernizing telecommunications policies and regulations to generate fair competition and reduce high communication and operational costs. · Building capacity to undertake such reforms, including through the establishment of systems and processes to review the performance of the regulatory institutions. For example, given the great demands and expectations placed on the regulator (TCRA) by telecommunications sector reforms, the Swedish government, through the Swedish International Development Cooperation Agency, is helping TCRA create capacity to meet its existing and future challenges and learn from its experiences in operating in a more competitive market. · Supporting the development rural telecommunications infrastructure, such as by developing universal access schemes. Rural areas lack telecommunications services or have only limited access in areas adjacent to main towns and on major trunk roads. This effort requires developing content in local languages (such as Swahili). · Enhancing technical and business-related skills development among the population using ICTs through technical institutes and vocational centers. For example, the University of Dar es Salaam is offering IT training in its computer center to the public. Sustaining and Sharing Growth in Tanzania 37 · Continuing to use global experiences to enhance the efficiency of the telecommunications sector. In many areas of telecommunications reform, Tanzania has benefited by adopting best practices from both industrial and developing countries. The functions and roles of the national regulator (TCRA) are the best example. Further benefits from global experience and best practices depend on the capacity of TCRA and other institutions to learn from the experiences of other countries. Sustaining and Sharing Growth in Tanzania 38 3. MAKING GROWTH PRO-POOR In order to ensure that Tanzania's growth process includes rather than marginalizes the poor, the report highlights four areas that require particular attention. First are reforms in the agriculture sector which are likely to have a direct impact on the poor, as most of them derive their livelihoods from this sector and their transition to other sources of income is likely to be gradual and slow. Second are micro and small enterprises, most of which are presently in the informal sector. This part of the economy has been an important path out of poverty in the past and is likely to continue to do so in future. Third is a focus on Tanzania's wealth of natural resources, which if properly managed can provide an important source of income and employment for the poor, but also of revenue for government for the financing of Tanzania's Strategy for Growth and Reduction of Poverty. Finally, and probably most important, is a focus on strengthening the capacity of the poor to contribute to and benefit from economic growth. This includes not only access of the poor to education, nutrition, and health services, but also issues related to vulnerability and risk as well as population policies. SUPPORTING AGRICULTURE SECTOR DEVELOPMENT As primary agriculture contributes more than 40 percent of GDP and employs 63 percent of the labor force, growth of this sector is central to Tanzania's overall growth performance and its success in reducing poverty. Recent agricultural growth of 4.8 percent annually between 2000 and 2005 is moderately high by regional standards, but needs to increase in the future to have a significant impact on poverty reduction, especially in rural areas. This growth has occurred in a very challenging external price environment. As prices of traditional exports fell, farmers quickly increased production of exportable food crops such as maize, beans, pulses and, horticulture, and successfully increased production of import competing goods such as milk and dairy products, sugar, rice, and oilseeds. The yield trend has been positive for rice, some horticulture, potatoes, bananas, and probably for milk and dairy, but disappointing for other agricultural products. Livestock growth was slower, but shows significant potential due to increasing demand for meat and related products at home and in Asia and the Middle East. Tanzania's past growth has depended on the expansion of area cultivated, and, as is the case in much of Africa, has been accompanied by only modest increases in labor productivity. As shown in Figure 15, labor productivity in Africa has trended up only since 1994. Tanzania's average annual increase in labor productivity of 1.1 percent is close to the norm for its neighbors. The figure also highlights the emerging disparity in labor productivity between Tanzania and Asian countries. The gap in labor productivity levels increased from less than 10 percent in 1990 to more than 40 percent in 2002. Sustaining and Sharing Growth in Tanzania 39 Figure 15. Labor productivity levels in Tanzania and comparators Agricultural Labor Producctivity (ag value added per ag workerr) 380 360 340 Asia Developing 320 300 Sub-Saharan Africa 280 Tanzania 260 240 220 1990 1992 1994 1996 1998 2000 2002 Notes: (1) countries with incomplete data for series excluded; (2) Asia Deve includes India & China; & (3) Sub-Saharan Africa excludes South Africa. Source: Author's calculations based on data from World Bank Development Data Platform With very modest increases in labor productivity and yields almost stagnant, Tanzania's growth derived largely from land expansion by farms using traditional technology, with little evidence of improved technical inputs and management practices. Examples of land expansion through labor saving innovation have been localized (for example, expansion via animal traction in the Rukwa region), and undoubtedly account for much of the land expansion in excess of labor force. But for the sector as a whole, the growth pattern of the past replicates poverty, rather than reducing it, since households do not experience income growth. Moreover, land expansion using existing techniques carries environmental costs, as forests are encroached upon and increasingly marginal land comes into cultivation. The 2002/2003 Agricultural Sample Survey suggests that there has been no further land expansion since the previous survey carried out in 1998/1999. Table 13. Land use and potential for land expansion (mid 1990s) Land use `000 ha % total land area · Urban 65 0.1% Rural protected land · protected forest/woodland 13,838 15% · other protected (wildlife, nat. park) 13,291 14% Agricultural land currently used (10.8 million ha total) · temporary crops 3,700 4% · pasture 6,150 6.5% · permanent crops 950 1.5% Rural unprotected land ­ available but not used · unprotected forest/woodland 26,321 28% · suitable for cropping but unused 7,000 7% · grassland/bushland not suitable for cropping; may be suitable for grazing9 (may include some water area) 23,221 25% Total land area 94,536 100% Source: FAO and World Bank 2001. Table 13, based on aerial surveys done in the 1990s, suggests that Tanzania still has significant potential to expand agricultural production through land expansion. However, in order to improve agricultural incomes, land expansion needs to go hand in Sustaining and Sharing Growth in Tanzania 40 hand with improvements in agricultural productivity through greater investment in land improvements, both on existing and newly cultivated areas, by adopting yield increasing technologies, and by shifting the composition of output toward products with higher value added. Empirical analysis of two representative surveys of farm households in Kilimanjaro and Ruvuma - two cash crop growing regions in Tanzania - shows that poorer households do not only possess fewer assets, but are also much less productive.10 Agricultural productivity directly affects household consumption and hence overall poverty and welfare. Stochastic production frontier analysis indicates that many farmers are farming well below best practice in the region. Holding inputs constant, they attain on average only 60 percent of the output obtained by their most productive counterparts. Analysis of allocative efficiency suggests that family labor is substantially over utilized, a sign of considerable excess labor supply. Use of intermediate inputs on the other hand is well below what is commensurate with the estimated value of their marginal productivities. An important reason for low input use is lack of credit to purchase inputs, but difficult access to the inputs themselves and being connected to the economy more broadly are also important impediments. Easy access to credit is positively associated with being a member of a savings association or being in a contractual arrangement with a cooperative or firm. Irrigation infrastructure facilitates access to credit. Together these findings support a continuing emphasis on increasing agricultural productivity in Tanzania's poverty reduction efforts. Better agronomic practices and increased input use will be crucial in this strategy. Better access to inputs and improved roads and transport services will further help boost input application. Financial constraints might be relieved through fostering institutional arrangements facilitating contract enforcement (e.g. contract farming, marketing cooperatives) and institutions that facilitate saving by the households themselves. They may also be relieved by the provision of more adequate consumption safety nets. The overall results suggest that a pro-poor rural development strategy needs to be anchored around improvements in agricultural productivity. Owing to the many farm and market level constraints on smallholder producers, there is a vital, positive role for national government and local institutions in enabling agricultural growth and rural poverty reduction. Removal of constraints on agricultural marketing, processing, and farm productivity will require: · Improved implementation of land tenure reforms; · Expansion of agricultural research effort, and continued research and extension focus on client responsiveness and engagement of farmers in the research process, and strong emphasis on sustainable use of land and water resources; · Irrigation improvements; and · Support for improved functioning of output and input markets, and for associated rural services, including finance. Improved implementation of land tenure reforms. Although land appears to be ample, rural areas suffer from a frequency of disputes over land and insecurity of tenure that one would expect to see only in a country with higher density of settlement and hence greater competition for land. To address the growing problems, the country passed the national land policy (NLP) of 1995 and subsequently revised the land legislation, to provide more expeditious access, enhanced security of tenure, and better management of land as a natural resource. A plan of actions and investments to implement the laws was drawn up in 2005. Appropriate funding and implementation of this plan will be critical to its successs. Technological change to foster growth: generation, dissemination, and adoption. Three elements must be present in the technology system to support increased productivity. Researchers must have developed new varieties that can be grown with Sustaining and Sharing Growth in Tanzania 41 success and profit under conditions that smallholders face; and efforts must be made to improve the environment for investment in private input supply networks. Producers need information about the availability of the varieties and guidance on how to use them. And money must be available to cover the costs of early adoption of the new varieties and techniques, while ensuring that a broad range of technical options, including low cost ones, is available to farmers. All elements of the technology system; i.e., the knowledge, and the financing must be linked in a way that provides smallholders access. The contribution of agricultural research, extension, and empowerment of farmers in the 1990s and early 2000's was positive, but limited by several factors. Services generally focused on increasing production through short-term technical packages, without paying attention to farmers' circumstances, markets, and sustainability. Despite various attempts to strengthen them, the linkages between research, extension and training were weak, and collaboration between public and private partners limited. Disproportionate emphasis was placed on generation and dissemination of technology, and less on empowerment (both intellectually and financially) of farmers to adopt the technology. As a consequence of weak linkages within the system, research did not always address priorities of greatest potential impact to the production systems. Technical breakthroughs did not yield good economic returns. And promising technologies remained on the shelf due to lack of knowledge or financing for adoption. Ruptured links in the technology chain reduced returns to investments in each of the elements. Moreover, the system was underfinanced, but given the somewhat depressed returns resulting from the institutional deficiencies, underfinancing was an appropriate response. Under the Agriculture Sector Development Program (ASDP) which was launched in 2003, a major reform of the institutional structure of the agricultural technology system has been designed and is under implementation. This provides a foundation for future growth that merits increased public financial support (Box 5). Box 5. Underlying principles for the design of an agricultural technology system Demand-driven and market-led technology development and adaptation: Farmers select, test, compare and adapt appropriate technological, service and market options. Pluralism of providers of services and approaches: diverse methodologies, processes and funding, as well as service providers are supported. Public funding for the system remains important, but services can be provided by public extension workers, NGOs, and private advisors. Subsidiarity: A constructive division of labor between national, district, and local levels is maintained. At the national level, the extension service feeds knowledge into the system, though training, identification of new approaches and technologies, and preparation of materials. Service standards are also defined and enforced at the national level. At the local level, organizations are able to contract the most suitable service providers, both public and private. Focus on economics and natural resource management, HIV/AIDS and malaria, as well as technical solutions: Agricultural service providers assist on issues of economic decision making and management of soil, forests, and water. Transparency and accountability: Accountability is built in through performance contracts, monitoring of performance, and ability of farmers to choose their providers. Farmers' feedback on services is integrated into the periodic evaluation of service providers. Irrigation to raise productivity and incomes. Tanzania is well endowed with water, both on the surface and below ground, but suffers nevertheless from water shortages due to insufficient capacity to store and access it. Cumulatively, the lakes, wetlands and aquifers provide huge natural storage capacity. The country also has 2.7 million hectares of wetlands (Usangu, Malagarasi). The total renewable water resource in Tanzania is Sustaining and Sharing Growth in Tanzania 42 estimated to be around 80 cubic km/year, of which 30 cubic km/year are ground water (FAO, 2004). Tanzania's ample water is matched by ample land suitable for irrigation. Of the total 44 million hectares suitable for agricultural production in Tanzania, only 10 million ha is under cultivation and out of this only 200,000 ha is irrigated. This represents a mere 2 percent of total cultivated area in the country. The estimated irrigation potential is up to 2 million hectares. Approximately three quarters of the presently irrigated area is farmed by smallholders in about six hundred small-scale irrigation schemes, usually using diversions and/or furrows in one of the nine major river basins. Very little irrigation is at present based on abstraction of ground water, which provides a promising area for future development with direct and affordable benefits to the poor. Rice is by far the most important crop irrigated in Tanzania, but sugarcane is also irrigated. A suitably designed ground water irrigation system (open wells, borehole tubewells, rainwater harvesting) could reduce the reliance on large bodies of water, including rivers and lakes, and promote more sustainable use of locally sourced and managed irrigation systems. Since the surface water available varies with rainfall, open wells and borehole or tube-wells can be constructed to spread the availability of water throughout the growing season. Compared with large surface irrigation schemes, the design of which is driven by topography and hydraulics, ground water development is often much more amenable to poverty targeting and is generally less capital intensive. Ground water irrigation can complement that based on surface water. Integrating ground water abstraction with rain water harvesting and watershed management, along with efficient water distribution systems, will lead to reliable, cost-effective irrigation systems. Irrigation at present is constrained by the high cost of investments required and by limited profitability of their use. Equipment, even the relatively modest implements needed for localized access to ground water, is more expensive in Tanzania than in, for example, India, by a factor of about three (FAO, 1997). To compound the adverse impact of high initial costs, producers face difficulties accessing high yielding varieties and moving products to market. Irrigation and agricultural productivity are clearly intimately linked, and neither can advance substantially independently from the other. Policy reforms in several key areas are needed to underpin growth in irrigation. Administrative regulations and restrictions on marketing and trading of irrigation equipment need to be simplified or removed. Procedures for the import of irrigation equipment such as drilling machines, pumps, etc. need to be simplified. Instruments such as micro-finance lending, matching grants, and joint ventures should include among eligible projects the smallholder groundwater irrigation projects (micro finance, joint venture financing). Development partners and government could join forces with NGOs such as International Development Enterprises (IDE) to enhance the supply of low-cost irrigation equipment (e.g. treadle pumps, drip irrigation). Finally, actions that affect the profitability of irrigated agriculture, such as the ongoing adjustment in the real exchange rate, continued liberalization of trade, and reforms of crop boards, will support an expansion of the area under irrigation in Tanzania.. International experience also highlights the importance of an appropriate institutional and incentive framework for the sustainable development of smallholder irrigation (Table 14). Sustaining and Sharing Growth in Tanzania 43 Table 14. Institutional Framework for Sustainable Development of Smallholder Irrigation Systems Required Targets Conditions Technical Self- - Capacity of Irrigation staff, LGA Staff and Extension Workers reliance - Farmers knowledgeable about water management and O & M - Appropriate choice of technology - Attention to environmental issues Financial Self- - Rationalization of the tax regime that small farmers face reliance - Better access to financial services, especially savings - Active private firms in supplying equipment and implements Institutional and - Clarity on roles and responsibility of public servants at district and national levels organizational - Strengthening/reform of Irrigation Section, ZIUs and LGAs support - Legal attention to land tenure, water rights, ownership and responsibility of irrigation infrastructure - Improved access to advisory services - Capacity to collect water fees and pay O & M cost - Investment climate to support growing constellation of small firms manufacturing equipment and providing services Source: ASDP Working Group 2 Report on Irrigation Development in Tanzania. Agricultural marketing and producer prices. Agricultural marketing and intermediary costs have decreased over time as policy reforms during the 1990s reduced price interventions and eliminated monopoly purchases by government bodies, thereby allowing greater scope for private sector trade and investment. However, a recent study (Nyange, 2005) suggests that marketing margins are still significant in relation to final prices, accounting for 30 to 50 percent of the border value (fob). The components of marketing margins vary across crops and locations (Figure 16). Large items can include taxes and fees, the trader's margin, the processor's margin, transport, finance, or packaging materials. Is there scope for reduction in these? In some cases, perhaps not: for example, packaging material costs may simply reflect the full import cost of the required materials. But for other items it may be possible to reduce costs via public sector investments (in transport, or in power and water, which are particularly important for processing) or policy (tax and regulatory). Clearly, the public sector role in improving prices for farmers via cost reducing measures, and the scope for such reductions, depends on local conditions and the specific crop under discussion. In many cases, the specific remedy will require knowledge of the local cost constraint and of tradeoffs between removal of the constraint and other objectives (e.g. tax revenue needs). Thus, there is an important role for district level growth strategies in identifying marketing constraints and possible solutions. In some cases, local bodies will be able to address the problems on their own. In many others, they will need to collaborate with outside service providers, national or regional government bodies, and private sector groups. Sustaining and Sharing Growth in Tanzania 44 Figure 16. Cost components of marketing margins vary significantly with crop and location Arabica coffeeRobusta coffeeT o b a c c o C a s h e w s C o t t o n Kilimanjaro Kagera T a b o r a C o a s t M t w a r a S h i n y a n g a Lyamrakana M o w oChonyonyo KibweraI t u n d u i p u n g u S Kizapala C h i g u g u Kishapu Nyenze 100% Kitomondo Chikundi Traders' Margin 90% Taxes 80% 70% Fees and commissions 60% Finance 50% 40% Processing 30% Packaging 20% material 10% Transport 0% Source: Nyange 2005. Producer margins are also crucially affected by labor costs ­ especially harvesting and weeding. Producer margins are sometimes very high; but as with traders there may be non-pecuniary marketing costs that are hard to observe (e.g. search for a buyer). Furthermore, since the farmer typically travels a considerable distance to sell at market, transport costs between farm and market are an important unaccounted-for component of the producer's margin. Crop boards exert a significant influence on the environment for production of and investment in Tanzania's key agricultural exports. A recent study (Kolavalli and Beddies 2004) covering coffee, tea, cotton, and cashew nuts, highlights a number of problems attributed to the current activities of the boards, including disruption of marketing and exports, costs in excess of services received, and interference in growth of the private sector in marketing. At the same time, the boards are recognized to offer services that might not be provided by the private sector, particularly input supply for smallholders, and regulation of monopsonistic behavior by private marketing agents. Reform of crop boards is critical to enhance the export competitiveness of Tanzania's agriculture. Key priorities for reform are a clear separation of the public and private functions of the crop boards and the appropriate alignment of financing, in order to ensure accountability to stakeholders. Key recommendations with regard to coffee are to make the coffee auction voluntary (including reviewing the relevance of the `one license rule'), reduce license fees, and discontinue the input voucher scheme. Other reforms include opening of coffee cherry buying to those who invest in common pulping units; ensuring accurate weights and measures; and ensuring an effective legal framework for branding and appellation. In cotton there is a need to deal with coordination to overcome problems associated with Sustaining and Sharing Growth in Tanzania 45 quality as well as supply of inputs and credit more effectively. Innovative pilots are needed in the cashew sector as well to learn more about institutional arrangements conducive to quality improvement. The forced saving programs that seem to be critical to making inputs available to producers (e.g. the passbook program) can be made more effective by enhancing stakeholder ownership and control over these programs. The public sector also has an important role in supporting agricultural development. It not only needs to spend enough public money, but must also spend it well. This requires developing responsive local government and national institutions that support local growth initiatives; improving local government technical capacity and finance to plan for growth; and increasing accountability and transparency. In order to enhance the efficacy of public spending in agriculture, it is necessary to focus on core functions of public institutions in the agriculture sector, carry out results-based assessments of public expenditure, with expenditure increases focused on areas where there is a clear labor productivity payoff for smallholders. There is also a need to learn about the impact on technology adoption and smallholder marketing of recent "smart subsidy" pilots, as for example, the matching grants scheme under the District Agricultural Development Grant. Spending on agriculture is increasing, and this is a positive indication of willingness to rebalance the budget in favor of competitiveness. At the same time, the efficiency of expenditure in the current allocation relative to goals of growth and competitiveness appears to below. Too little is spent on research, extension, irrigation, livestock services, roads, land administration, and energy. Too much is spent on input subsidies and probably on the strategic grain reserve. Expenditures on "Crop Development" in the budget of the Ministry of Agriculture and Food Security, and Cooperatives warrant careful evaluation, since they are quite large. The consistency of present expenditures with ASDP understandings is not readily apparent in the budget as it is now presented. This hinders assessment of the consistency of public expenditures with growth and poverty reduction aims. Alignment of expenditure around core and complementary activities that support the ASDP would contribute to efficiency of expenditure. The core items would be those production and research items currently included under the MAFC mandate, livestock under the Ministry of Livestock Development, and marketing and finance under the Ministry of Industry, Trade and Marketing. Activities supported under these mandates presently take place at both national and at the local levels, and spending at the local level will increasingly be channeled through the Prime Minister's Office­Regional Administration and Local Governments (PMO-RALG). As local governments increase their capacity and as District Agricultural Development Grants becomes operational, public institutional responsibilities for agricultural development will not be defined by traditional ministerial prerogatives, but rather by local capacity, i.e., that national level institutions only undertake activities that districts and local bodies cannot handle. By 2007/08 three quarters of expenditures to support the ASDP are expected to flow through district governments. As participants in the political process become more familiar with the ASDP and the implications for organization of public support, a reconfiguration of responsibility at the national level may be appropriate and feasible. Such a reconfiguration would imply merger of the departments and entities that presently have responsibility for elements of the ASDP at the national level into one ministry, with close coordination with the ministries responsible for complementary activities (roads, energy, and water), particularly at the planning stage. Such a reconfiguration of responsibility at the national level and improved coordination with both local and complementary counterparts would facilitate adequate funding of exciting elements of the ASDP agenda, such as the following: Sustaining and Sharing Growth in Tanzania 46 · Share costs of adopting improved technologies, rather than subsidizing a single input. Tanzania is already using a public-private cost-sharing mechanism (through a matching grant) for adopting new technology under the Participatory Agricultural Development and Empowerment Project within the ASDP. The selected technology is one that is profitable in the longer run, and the matching grant makes fertilizer, seeds, seedlings, plant protection agents, implements, and any other needed inputs affordable during the period of adoption (in most cases, two years). Inputs are purchased from the private sector, and farmers save increased earnings while receiving the matching grant so that they can continue to purchase inputs on their own. This approach can be scaled up through the DADPs. · Continued reform of taxes. The government has already undertaken a number of commendable reforms to improve the tax regime for agriculture, including setting limits on local taxes for traded agricultural commodities and reducing customs fees and value added tax. Enforcement of those measures should be pursued and their effects monitored, because they can improve profitability if well implemented. The value added tax on port charges and transport costs adds to the cost of fertilizer and could be reevaluated. · Reduce cumbersome importation procedures for fertilizer. Such procedures include double inspection of consignments, preshipment inspection by Cotecna at 1.2 percent of the free-on-board price, and the Tanzania Bureau of Standards (TBS) quality inspection at 0.25 percent of the cost, insurance, and freight price. Any delay by the TBS means port charges accumulate, thereby increasing the cost of fertilizers to farmers. In addition, importers are required to produce certificates of quality from manufacturers. Abolishing inspection requirements and retaining the certificate of quality from manufacturers, together with spot-checks by the TBS, would significantly reduce the cost of fertilizers to farmers. · Improve the road network. As much as 40 percent of the cost of grains in the major urban markets is attributed to transportation costs. Reducing this cost would increase farm profitability. · Develop and disseminate more profitable technologies. This effort includes ensuring that agricultural research and extension focuses on developing more profitable technologies and varieties more responsive to the application of inputs. It also means ensuring that farmers have access to information and recommendations specific to the technologies (rather than blanket prescriptions) and that profitability is part of the calculus in deciding which technologies are recommended. · Ensure consistency in the policy environment. The costs of inputs decline as the volume of transactions increases. The number of private input dealers and distributors is more likely to expand when the policy environment is consistent and the government stays away from direct intervention. · Improve the opportunities for viable systems to finance technology and inputs. Vertically integrated private systems of extension, credit, and inputs linked to output markets (as for tobacco) are functioning in some areas. They can be encouraged by reducing inefficiencies in the banking systems (reducing interest on borrowing), strengthening institutions for contract enforcement, improving taxation (as noted above), and improving the flow of information. For example, small farmers in Msowero and Sonjo villages were assisted under the PASS (Private Agricultural Sector Support) program--supported by DANIDA (the Danish International Development Agency)--in forming groups and accessing loans that were repayable over a three-year period. Initial reports on the program are positive. Sustaining and Sharing Growth in Tanzania 47 · Reduce risks. Periodic weather shocks and external price shifts, together with household-level production risk, significantly affect technology adoption, profitability, and incomes. New methods to manage external shocks, such as weather insurance, forward contracts, and options for price risk, are being piloted and should be evaluated carefully to determine the scope for scaling them up. · Improve market access and product quality. Investment in, awareness of, and compliance with sanitary and phytosanitary standards for high-value products and stronger linkage of groups of smallholders with supermarket chains will improve marketability and profitability. The public sector can play a very constructive role here, even though the market transactions are between private parties. For example, through an agreement with TechnoServe, the U.S. Agency for International Development (USAID) has helped coffee growers respond to October 2003 changes in the policy regime, whereby Tanzanian premium coffee producers and specialty roasters can export high-quality coffee directly. This change in the marketing regulations (the contribution of the public sector) creates new opportunities for Tanzanian farmers to supply roasters with specialty coffee for premium prices. TechnoServe's conservative estimate is that farmers' incomes will be boosted by US$20 million over the next 10 years as a result of the combined effect of direct export of premium coffee and a new value added tax reclamation policy. Concerted pursuit of those steps and others through adequate funding of activities under the ASDP will provide a reasonable likelihood that the growth targets can be met. Five percent annual growth is ambitious but not impossible. Meeting this target, however, requires moving ahead decisively on the agendas of policy reform and alignment of public expenditure. Half-measures and compromises to accommodate special interests require sober reassessment of the growth targets or admission from the outset that they will not be met. HARNESSING NATURAL RESOURCES FOR SUSTAINABLE GROWTH Tanzania is rich in natural resources, but their contribution to growth and poverty reduction could be significantly improved through better governance and management. Gemstones, gold, fish, forests, and wildlife are among Tanzania's most important natural resources. Their exploitation and use provides a tremendous potential for the generation of income and employment as well as revenue for government. Contribution of natural resources to growth. During the past decade, mining, fisheries, and tourism have been the most dynamic sectors in the economy. The fishing industry is currently concentrated around Lake Victoria, which is already displaying signs of depletion of the fish stock. However, Tanzania's other lake resources and coastal areas provide significant potential for expansion of the fish industry. Most currently known mineral deposits are being already tapped which could slow down the growth in this sector in the near future. Tourism expanded rapidly during the 1990s; however, the growth of the sector slowed down in recent years, particularly after the terrorist attacks in Nairobi and Dar es Salaam in 1998. Forestry, wildlife and marine fisheries resources, though declining, are still relatively abundant, rendering largely untapped growth potentials. Although these natural resources contribute to the economy and subsistence base of the rural population, their value and potential is underestimated. This underestimation is partly based on missing markets in the case of public goods, imperfect competition in the case of distorting government interventions, as well as pricing of natural resources below market value. Results of all these market failures lead to sub-optimal economic decision making and loss of income to the country. Sustaining and Sharing Growth in Tanzania 48 Due to weak governance and capacity in revenue generating sectors, resources are offered below market price to the benefit of a few powerful winners, and at the loss of the majority of the rural population. Yet, these natural resources provide substantive potential for income to communities in r ural areas. The weaknesses in governance regimes in all three sectors, forestry, wildlife and fisheries, include primarily lack of transparency and accountability in issuing rights of resource extraction and revenues accrued thereof, lack of equitable sharing of benefits with communities, as well as weak monitoring and surveillance of stocks. In all four principal sectors providing natural capital to the Tanzanian growth equation, that is, forestry, wildlife, fisheries and mining, royalties are set arbitrarily and do not reflect scarcity. Royalties are hence not used as a policy i ns t r u m e nt of inter-temporal resource pricing and sustained yield management. As long as these weaknesses are not addressed, a substantial base of economic growth will be slowly eroded and poverty reduction objectives are unlikely to be achieved. Accounted and unaccounted for growth contributions. In general, a great share of the economic contributions of forestry, wildlife and fisheries, does not enter the GDP and export statistics, and is hence not taken into account in the analysis of growth. Availability and quality of data is a problem that needs to be tackled urgently. Forestry provided over US$5 million in government revenue between 2003/04 and 2004/05. It contributes officially 2-3 percent to GDP and a 10-15 percent share of export earnings. Taking unaccounted services and non-industrial forestry into account, estimates suggest that the value added in forestry may be as much as 10-15 percent of GDP. Forests provide around 75 percent of building materials and 100 percent of indigenous medicinal plants and supplementary food products. In addition, forests provide an important component of value added to national income through their eco-system wide service functions, by providing for industrial and domestic water and energy supply. Wildlife resources generate significant income through game viewing by international tourists. In 2001, Tanzania's national parks drew over 100,000 international visitors. This generated receipts of almost 5 percent of GDP, equivalent to about US$400 million. About US$30 million are generated from tourist hunting and an additional US$9 million generated by the private companies leasing hunting concessions from the government. In addition, wildlife provides substantial unaccounted for subsistence values. Well over two-thirds of Tanzanians eat wild meat, with up to 95 percent of the rural population claiming it as their most important meat protein source. Tanzania's fisheries sector has grown at a rate of 6 to 7 percent annually since 2000. In terms of export earnings, fisheries contributed 10 percent of total exports in 2003, which equaled US$130 million, the export value of Nile Perch being US$100 million. A great share of the marine catch does not enter GDP and export statistics but plays an important role in livelihood support. The official number of artisanal fishermen has doubled since 1995 and was close to 120,000 in 2 0 0 3. Although its contribution to GDP is still small at 3.5 percent, mining is the single most important earner of foreign exchange for Tanzania. About 50 percent of export earnings accrue from minerals, predominantly from gold mining by large-scale foreign owned operators. In addition, mineral resources are of importance to the artisanal mining sector. Untapped growth potential. With regard to game viewing tourism, untapped potential exists in the southern parks of Tanzania. While the northern tourism circuit has supposedly reached maximum carrying capacity in terms of numbers of visitors, places like Ruaha and Katavi National Parks in the south are still fairly unknown. Shifting marketing and infrastructure development to these areas would provide new growth potential for Tanzania's growing tourism sector. Sustaining and Sharing Growth in Tanzania 49 Marine fisheries have recorded a sharp revenue increase due to increased license revenues from foreign vessels in the Exclusive Economic Zone (EEZ). There are estimates that the presently earned revenue does not reflect the total amount that the government could earn, and that real catch is much higher than what has been assumed as the basis to set the license fees. Notably, there is no catch based license or fee, and the vessels are allowed unlimited catch once they are in possession of a valid license. Resource rent estimates from marine fisheries captured by Private Fisheries Agreements (PFAs) show that license fees as a percentage of revenue generate a gross resource rent of approximately 2.2 percent. This is slightly less than half of what might be expected in a western industrial fishery. Calculations for tuna seiners vary somewhat more depending on different catch scenarios. Although the current license fee arrangements of PFAs in the EEZ generate substantial revenue, the level is too low to capture a reasonable resource rent (>5 to 7 percent of gross revenue) for the coastal state. The PFAs as currently offered to foreign fishing enterprises are hence considered untenable. With regard to freshwater fisheries, past growth rates are mainly based on Nile Perch exports from Lake Victoria. Other lakes, such as Lakes Tanganyika and Nyasa, as well as harvesting of other species have so far been commercially underdeveloped. Diversification could also be sought in terms of exploring additional export markets, as risks and vulnerability increase in a situation where export earnings in a sector are entirely dependent on a single market. This is the case of fisheries exports from Lake Victoria, which are mainly destined for the EU. Commercial fisheries present an important emerging revenue source for the country and the sector. If the sector is well managed, commercial fisheries can have a positive impact on economic growth and poverty reduction at the same time. Principles of management would need to include retention and re-investment of revenue into the sector and putting up safeguards for artisanal fisheries to protect their rights and access to the resource. Potential for local spin-off effects. Local spin-off effects are presently missing in the context of mari ne fisheries in the EEZ. While new fisheries agreements are being negotiated with foreign countries, no fish is expected to be landed ashore and few supplies will be sourced from within Tanzania. If no such spin-off effects are created, the net impact of commercial fisheries on poverty reduction may be negative, given increased competition with artisanal fisheries over the same resource. The fact that Tanzania is a net importer of forest products is a sign of lost opportunities for income generation for the local economy. Similarly, the mining sector seems to have had limited influence on poverty reduction in the local economy. Employment in the large-scale mining sector is limited, although younger employees may receive significant salaries. The majority of those employed in the mining sector operate as self-employed, small- scale sector, artisanal miners. Their activities generate very low returns, especially when one considers the hardship associated with this kind of employment. In addition, an increasing income disparity seems to be emerging between those employed in the small- scale and large-scale mining sector. To the extent that those recruited by the large-scale mining sector are recruited outside the local community, the local community is thus restricted to opt for poorly paid employment opportunities in the small-scale sector. Large-scale mining may have positive effects for local communities through the improvement of basic infrastructure. There is, however, no indication that the expansion of the mining sector triggers significant growth in the local economy, since mining operations generally are detached from local supply chains and therefore create employment opportunities primarily in the services sector. Potential for poverty reduction. In addition to their potential for government revenue generation, wildlife, fisheries and forestry resources provide the non-agricultural subsistence base for rural communities in remote locations. Increased emphasis on natural resources related enterprises has the potential to create additional income opportunities for the rural population. This study highlights several examples of successful participation by villages in wildlife tourism, hunting, and forestry. However, despite the Sustaining and Sharing Growth in Tanzania 50 conducive policy framework in both the wildlife and forestry sectors, weak governance systems at both central and local levels have so far limited the realization of the poverty reduction potential through community based natural resources management. The main focus within community wildlife management has been on institutions and distribution of benefits rather than enterprise opportunities at household level. Sustainability of growth. In the context of sustainability of natural resource based growth, the following are the major constraints that can lead to revenue loss and possible deceleration of growth in the long term: · Under-pricing of resources not allowing capturing of resource rents; · Weak environmental governance systems; and · Limited knowledge of stocks, their values and changes over time. Sustainable growth based on renewable resources requires that the cost of extracting a resource and the notional cost of replacing a unit of the resource, commonly known as 'resource rent', is evaluated so that the wealth base is not eroded. While royalties are the most important source of government revenue in the forestry (83 percent), wildlife (96 percent hunting licenses) and fisheries (84 percent royalties, 15 percent export licenses) sectors, they are set arbitrarily and capture neither market values nor resource rents. Similarly, in the mining sector, licenses to foreign investors do not take the `capital component' into account. Tax incentives have been generous towards foreign investors to attract capital investment and to open up the market at the expense of sustainability principles. In such a scenario, acceleration of growth comes at the expense of pricing resources below market value, which leads to loss of income, erosion of critical stocks and an associated deceleration of growth in the long-term. The under-pricing of natural resources is compounded by weak environmental governance, including inefficiencies in revenue collection, limited transparency and accountability, as well as outright corruption. This leads to a loss of income and livelihoods for rural communities. Finally, limited knowledge of resource stock values and stock changes implies that limits of extraction and quota associated with licenses can only be set arbitrarily and cannot be based on sound ecological calculations and realistic projections. Externalities. Natural resource extraction has potentially important externalities such as the impact of commercial fishing and large scale mining on atisanal fishing and mining. Unless such externalities are properly addressed, natural resource extraction may have negative economic, social, and environmental impacts that may interfere with the poverty reduction objectives of the NSGRP. Summary. This review has generally argued that the natural resource sector in Tanzania is dynamic and a potentially important contributor to overall economic growth and poverty alleviation. It does, however, remain fragile and the use of well-placed interventions can serve to secure its role as an engine of growth. In particular, much of sectoral output remains hidden, either because it is not recorded in national statistics, or because potential economic values are either destroyed or squandered as a result of various weaknesses fall into the categories of poor governance; inefficient management; and, vulnerability. There is substantial difference among the sectors considered and efforts are underway by government in all cases to correct some existing weaknesses. Poor governance is reflected in persistent corruption and lack of transparency in how resources are administered by government. It can arise from the lack of a responsible institution, or from poor record keeping and blatant illegal activities. Governance issues are least serious in freshwater fisheries, tourism, and large-scale mining, and most serious in marine fisheries, forestry, wildlife management, and artisanal mining. Trade in carbon resources suffers from the newness of the activity and the lack of government capacity in Sustaining and Sharing Growth in Tanzania 51 the area, but there is no evidence of poor governance through corruption or illegal activity. Inefficient management relates to situations where market or policy failures have reduced the value of the resource base or reduced the rent that can be captured effectively by existing mechanisms. The most common inefficiencies arise from unclear property rights, incorrect resource pricing, tax system distortions, and inequitable benefit sharing (which creates conflicts and reduces management effectiveness). Losses associated with inefficient management are currently most significant with marine fisheries and wildlife resources because of resource pricing and property rights issues; similar inefficiencies persist in artisanal mining but these are compounded by social and environmental externalities relating to human health and environmental impacts. Vulnerability is a system-wide effect that speaks to the ability of the resource base, and of those dependent on the resource, to withstand unforeseen shocks. These shocks are becoming more frequent, and are related both to internal factors (such as general poverty or dependence on single resources) as well as external factors (such as global security, climate change, and international commodity prices). All sectors are vulnerable, in various ways, to such shocks. Recommendations and Priority Actions. Addressing these weaknesses requires an integrated approach that addresses the different classes of weakness. The basic foundation (or first pillar) is associated with reforms that promote good governance. The elimination of corruption and improved transparency is the foundation for any type of subsequent improvement in resource management. An ideal rent collection system or an optimal fiscal regime is of no use if the overall system is corrupt or lacks basic administrative capacity to monitor and assess activities. While Tanzania has made progress in overall governance, some additional interventions are appropriate. Reforms that could be considered relate to the establishment of responsible authorities, training and human resource development, and overall transparency and accountability. The second pillar involves management interventions that improve efficiency of resource development and exploitation. These presuppose the existence of a regime and administrative structure that reflects good governance. Interventions in this area generally include the creation or enforcement of property rights, correct resource pricing, elimination of distorting taxes and subsidies, and improvement in benefit sharing. The third pillar relates to the protection of the resource base through safety nets and is intended to reduce vulnerability of the resource base and the populations that depend on the resource. The instruments that could be considered include revenue retention schemes, adaptive co-management models, education and awareness, vulnerability monitoring systems, and diversification. All of these tactics, however, must be coordinated by an over-arching strategy of vulnerability management that ­ at this stage ­ does not yet exist. A key recommendation is thus to define and institutionalize such a strategy within a central coordinating mechanism in the President's Office or Vice President's Office. A summary of the recommendations is provided in Table 15 which shows the key actions within each of these three pillars. The table is presented as a simplified log-frame or results matrix, including some potential indicative indicators for tracking results and impacts. Sustaining and Sharing Growth in Tanzania 52 Table 15. Summary of Objectives, Priority Recommended Actions and Indicative Results Indicators for Enhancing the Contribution of the Natural Resources Sector to Economic Growth Reform: Management: Protection: Promoting Good Governance Improving Efficiency Providing Safety Nets Objective: Objective: Objective: Eliminate corruption and improve Eliminate market and policy Reduce vulnerability of natural transparency of governance in the distortions, improve rent capture, resources to shocks, and enhance sector. and achieve a more equitable the resilience of populations distribution of benefits. dependent on these resources. Priority Reform Actions for Priority Management Actions for Priority Coordinating Specific Sectors: Specific Sectors: Mechanism: · Operationalize Deep Sea · Strengthen community · Establish a high level Fishing Authority. property rights in wildlife and coordinating unit to address · Create and Operationalize forestry through vulnerability issues at a Executive Agency for implementing more national level, reporting to Forestry. community managed areas. the President or Vice · Review institutional · Establishment of a President. arrangements affecting community territorial sea · Development of vulnerability Wildlife Division and along coast. profiles based on social and TANAPA, with a view to · Operationalize cadastral natural resource mapping establishing a new executive system for artisanal mining. (building on poverty agency in the wildlife sector. · Establish a legal framework mapping). · Review capacity and training for ownership of carbon Priority Protection Actions: needs within MEM to improve resources. · Increase use of revenue modalities of artisanal mining · Reform and harmonize retention schemes with governance. licensing system for offshore preference for those · Training and human resource fisheries, with net increases in involving uncorrelated risks development through use of fee structures. Consider and pooling of revenues anti-corruption training implementation of bidding (such as the Marine Legacy programs for lead agencies in system. Fund). all sectors (fisheries, forestry, · Implement an open tendering · Increase reliance on adaptive wildlife, tourism, mining, scheme for wildlife licensing co-management arrangements carbon). with transparent mechanisms with local communities, · Training in anti-corruption for an appropriate level of private sector, and methods at District level and domestic preference. government. sensitization for · Review of taxation · Development of norms and implementing partners in instruments to encourage standards, and support for NGOs and private sector, effective value-added education and awareness particularly in fisheries, processing in marine fisheries around vulnerable resources. forestry, and wildlife. and wildlife products. · Implementation of · Enhancements to accounting, · Adopt uniform taxation vulnerability monitoring monitoring and reporting provisions for mining sector systems, including early systems in all sectors. in lieu of individually warning systems for known · Accelerated implementation of negotiated agreements. threats. MCS database for offshore · Increase direct involvement of · Implement diversification fisheries. private sector in management and protection mechanisms · Accelerated establishment of a of fisheries, wildlife, tourism for vulnerable groups, GPS/GIS mineral cadastral and mining. focusing on protection of system and database for the critical ecosystems and mining sector. guarantied access to traditional uses of resources. Key Reform Indicators: Key Management Indicators: Key Protection Indicators: · Operational agencies with · Increased proportion of · An emergency response transparent monitoring and resource subject to managed network and plans in place reporting systems in place. access (as opposed to open or nationally. · Reduced leakages from unmanaged access). · Poverty and vulnerability corruption and illegal · Increased rent generation and mapping being used at Sustaining and Sharing Growth in Tanzania 53 activities. capture. District level. · Improved awareness regarding · Increased participation by · Monitoring systems in place good governance among local communities in resource and delivering timely reports. government staff and other management, with increased Demonstrated positive outcomes stakeholders. share of benefits. in aftermath of unanticipated · Reduced resource use shocks. conflicts. Sustaining and Sharing Growth in Tanzania 54 SUPPORTING THE INFORMAL SECTOR ANDMICRO, SMALL, ANDMEDIUM SIZE ENTERPRISES The informal economy plays an important role in income generation, employment and skills development and often serves as an entry point for widespread participation in the private sector. International estimates (Schneider 2004) --which include informal activities in agriculture -- suggest that Tanzania's informal economy accounts for about 60 percent of the Tanzanian gross national income (GNI). The informal sector is thus relatively large, both in regional and international comparison (Figure 17). Data from the Tanzanian Integrated Labour Force Survey (ILFS) 2000/01 (URT 2001) --which exclude informal sector activities in rural agriculture-- suggest that the informal sector employs about 16 percent of the total labor force. The Institute for Liberty and Democracy (ILD 2005) found that about 98 percent of economic activities in Tanzania were within extralegal boundaries in the informal economy, and 89 percent of all proprieties are held extra-legally. According to the ILD, the Tanzanian informal economy has assets worth US$29 billion. Figure 17. Size of the informal economy for selected countries (as % of GNI) 80 70 60 50 40 Percentage 30 20 10 0 USA UK China Africa Kenya Lanka Malaysia Botswana Ethiopia Malawi Uganda Sri Thailand Nigeria Tanzania Panama Georgia Switzerland South Bangladesh Mozambique Source: Schneider 2004. Informal economy operations can be found in most sectors in Tanzania. According to the ILFS 2000/01, the industries that employ most people whose informal economy activity is their main activity are retail trade of agricultural products, meat and chicken (20.7 percent); stationary, photography and general retail (18.8 percent); retail trade of processed food (10.5 percent); and restaurants and hotels (12.4 percent).11 The Tanzanian informal economy thrives because it provides opportunities of income generation to the poor and unemployed, and because it offers a low cost ground for experimentation with business ideas that can lead to growth and formal enterprises. According to the ILFS 2000/01, one of three households was active in the informal economy in 2000/2001, as opposed to one of four households in 1990/1991. The survey shows that the number of households with informal economy activities grew during the 90s from 42 percent of all households in urban areas to 61 percent. In rural areas, 27 percent out of all Sustaining and Sharing Growth in Tanzania 55 households had informal economy activities in 2000/01 as compared to 21 percent in 1990/91. Also, in terms of employment, the informal economy holds a bigger share in urban areas as compared to rural areas. In Dar es Salaam, for instance, 36 percent of the labor force is employed in the informal economy. Generally, for households engaged in informal economic activities in urban areas, this tends to be their main activity, while it is more likely to be the secondary activity in rural areas. Crop growing is the sector employing the majority (94.4 percent) of the people whose informal activity is a secondary activity. The informal sector--especially urban self employment--has been an important path out of poverty for many Tanzanians. Not only did the share of self-employed persons in the adult population increase from 4.8 percent to 8 percent,12 but expenditures by households headed by a self-employed individuals grew by 18 percent between 1991/92 and 2000/01, compared with a much more modest growth of expenditure of agricultural households, which grew by only 7 percent during that period. In Dar es Salaam, the growth of expenditures by households headed by a self-employed person was even more dramatic at 65 percent, which was even higher than the average growth in expenditures of 60 percent experienced by households headed by a person in paid employment. Many small enterprises in Tanzania operate under a semi-formal legal status without the necessity of registration with state authorities. Semi-formal operators appear on a list of operators at the local authority. They pay taxes that are collected by local authorities. How long informal operators may remain at any level of informal or formal status differs widely in Tanzania. An operator may progress via a semi-formal stage, or move directly from informal status to the official registration (Table 15). Table 16. Typology of forms of enterprise in Tanzania Formal status Legal form Description and subcategories Illicit None An enterprise is illicit when there is no legally permitted licensed or registered counterpart. Informal None An activity (for which there is a formal counterpart) that does not comply with requirements of the regulatory system regarding licenses, permits, certificates, notification or registration of the activity is licit but informal. Semi-formal Local An activity carried on by an operator who appears on a authority local authority list of licensed operators of enterprises but licensed is not registered with the state registrar is semi-formal. enterprise Existing forms: (i) Hawkers license, (ii) Business license, and (iii) Market stalls. Formal State Person(s) registered to conduct business activities under a registered registered business name; for example, enumerated on a enterprise state register as a sole ownership enterprise, a private limited company, or a joint stock company open to public subscription shareholdings. Source: Nelson and de Burijn 2005. Informal operators state a number of constraints to the growth of their businesses. Most importantly, they are often not in a position to afford a permanent premise for their business. However, there are a number of success stories of informal operators formalizing their business after the municipal authorities provided or assisted in the access to permanent premises. Second, they lack access to credit. A lack of business management skills as well as very limited access to new technology is detrimental to the growth of informal businesses. There is also concern about reported harassment of informal entrepreneurs by local authorities -- some businesses are demolished, property is taken away and in the worst cases, Sustaining and Sharing Growth in Tanzania 56 entrepreneurs face charges. Although the situation is reported as having improved considerably since 2003, continued attention to removing obstacles to economic activities-- both in the formal and informal sectors -- and facilitating the transition from informal to formal status, is important. The cost of starting and operating a formal business are high in Tanzania. According to the World Bank's Doing Business Survey (World Bank, 2006a), entrepreneurs can expect to go through 13 steps to launch a business over 35 days on average, at a cost equal to 161.3 percent of the $330 per capita income, compared to 10.8 percent, 48.2 percent and 117.8 percent of per capita income in Botswana, Kenya and Uganda respectively or 6.5 percent in OECD countries. However, the minimum capital required to obtain a business registration number in Tanzania (6 percent of income per capita) is considerably lower than the regional average (297.2 percent) or the OECD average (28.9 percent). The ILD found costs and burdens of the procedure to legally incorporate a private Bureau of Change in Mbeya include 10 stages, 103 steps, 379 days and US$5,506, the respective figures for Dar es Salaam are 95 steps, 283 days and US$3,816. The ILO/UNIDO/UNDP Roadmap Study calculated that the cost of coping with regulatory and non-regulatory constraints would amount to as much as 75 percent of monthly sales of informal operators in the case of the firewood and charcoal sector, which explains the decision to remain informal.13 In order to register a business, entrepreneurs have to travel to Dar es Salaam, since the Business Registration and Licensing Agency (BRELA) is only located there. The ILO/UNIDO/UNDP study shows that business licensing is equally cumbersome, because it requires various procedures in different offices at the regional and district level. Both processes, licensing and registration, are required and often involve bribing officials further increasing the cost of starting a formal business. Costs of formalization are often further increased because some officers are reported to be unhelpful, obstructive and uncaring.14 The Business Activities Registration Act, which was passed by Parliament in early 2007, contains important provisions for the decentralization of BRELA and the streamlining of registration of registration and licensing requirements and procedures. Implementation of this new act is thus an important step towards facilitating the establishment of new businesses and the transition of businesses from the informal to the formal sector. Registering property is an expensive and time-consuming undertaking in Tanzania. It takes 12 steps and 61 days to register property compared to 4 procedures and 33 days in OECD countries. The cost of registering property amounts to 12.2 percent of the property value, which is considerably higher than in OECD countries (4.7 percent). The ILD found that the procedure to allocate land for urban purposes and to obtain a building permit takes 13 stages including 68 steps, which take 8 years and cost US$2,252. De Soto (2001) explains that a major problem the informal economy faces is its inability to convert what he refers to as "dead capital" in form of untitled assets into capital that can be used as for example collateral for loans. Also, enforcing formal contracts costs on average 35.3 percent of the debt, compared to the OECD average of 10.9 percent. According to the ILD, resolving claims for a debt at the Commercial Court Division take 9 stages, 96 steps and 390 days and cost US$11,964. Likewise, it takes 1,286 days and costs US$1,022 to collect a debt by executing a court decree. In addition it is comparatively difficult to hire or fire employees in Tanzania. However, the ILO/UNIDO/UNDP Roadmap Study of the informal economy in Mainland Tanzania did not identify employment regulations as a severe constraint, since they are not enforced. Tax rates, which are perceived as being too high, often act as a disincentive to formalize. A recent study (FIAS 2006) shows a number of anomalies with the tax regime for small Sustaining and Sharing Growth in Tanzania 57 businesses, in particular that the system is regressive for non-record keepers. Moreover, the study states that the "tax environment encourages expansion of the informal economy," as "small businesses face a proportionately higher time and financial costs to comply with administrative requirements and therefore may not see any benefit of joining the tax net." Furthermore, newly registered businesses have to pay taxes up-front, which further increase the required starting capital. The marginal effective tax rate is higher for small businesses than for large firms. Local governments still levy a large number of taxes, fees, licenses and charges. Local taxation is seen as a major constraint to formalization. In addition, bribes are estimated to add up to more than 5 percent of total sales. Comparative investment climate survey data sets collected by the World Bank suggest, however, that the share of revenue that Tanzanian businesses deliberately fail to report for tax purposes is more than 30 percent, compared with around or less than 20 percent in Kenya, Uganda and Zambia. It should, however, be mentioned that informal entrepreneurs do pay taxes to local authorities. For example, in Dar es Salaam the local tax is TSh. 100 per day, which is collected from every informal operator. Benefits of Increasing Formalization. Formalization creates the basis for formal transactions with other entities, including financial intermediaries for access to credit, the public sector for the provision of public services, and clients and suppliers for contract based transactions. Given the rigidities, costs and attitudes of formal sector regulators and service providers, the access of informal entrepreneurs to important public goods such as utilities like electricity or water, and other inputs and services is limited. From the perspective of business operators, formalization of their businesses increases the trustworthiness for customers--a benefit that has been mentioned by several entrepreneurs who formalized their business. Furthermore, operators in the informal economy cannot convert "dead capital" in the form of untitled assets into "productive economic currency" such as collateral for loans to start or expand a business. Moreover, their access to the capital market is very limited. Formal registration is generally a prerequisite for access to credit and small business loans (US$15,000--US$30,000). Therefore, informal sector operators are deemed to remain small without being able to exploit potential economies of scale. As a result, productivity tends to be low. In general, the labor productivity is lower in Tanzania than in many other African countries--the median value added per worker is US$2,061. In the case of micro enterprises (of which most are informal), labor productivity is even lower (Figure 18). Figure 18. Median value added per worker Median ValueAddedper worker in US dollars Median valueaddedper worker bysize(in US dollars) 4,397 $4,500 $4,500 $4,011 3,457 $3,000 2,680 $3,000 $2,354 2,061 $2,145 $1,500 1,085 $1,500 $1,036 $0 $0 Micro (1-9) Small (10-49) Medium(50- Large (100+) Tanzania Uganda Zambia Kenya China 100) Source: World Bank 2004b The most obvious benefit to the government of increasing formalization is higher tax revenue. Furthermore, the government's ability to implement policies and the effectiveness Sustaining and Sharing Growth in Tanzania 58 of government programs aimed at the private sector will rise since informal enterprises operate outside of the government system of regulation. Box 6. Examples of voluntary formalization Alan Mungo experimented with informal wine production, insurance agency, clothes shop, and eventually progressed to wine production and safari tour companies. Having developed a marketable product and identified market opportunities for expansion, he decided that it would be in his interest to legitimize his enterprise. However, after preliminary enquires he found that the premises he used, being a motorcar garage on the grounds of his dwelling house, would not be approved by health inspectors, which meant he would not be able to apply for an operator's license. At that time he attended a meeting between representatives of the Association of Food Processors (TAFOPA), the local trade licensing officer and a representative of the Ministry for Agriculture and Food Security. There it was agreed that small-scale food processors who could not comply with regulations would not be forced to cease their operations, instead they could continue to operate without licenses, but still be subject to supervision by health inspectors. This incidentally implied the advantage of being off the database of operators liable to pay formal taxes. The acknowledged reason for the Ministry's decision was that Tanzania's food processing industry was important, relied mainly on small-scale producers, most whom were not licensed and would be forced to close if regulations were enforced. Mungo was able to continue operating and licensing costs were deferred until after his premises reached the standard about one year later. At this point, his formalization costs were very low. Dan Himba formalized his business after an activity period of 29 years. He conducted informal enterprise activities as a supplement to his salaried employment. During this period he enjoyed a long run of starting informal enterprises, maximizing profits while he could and abandoning them when they showed signs of having run their most profitable life cycle. When the conditions of his employment deteriorated, a full-time entrepreneurial career became more attractive to him and increased the stimulus to formalize. He subsequently abandoned his job for a career in private enterprise. Source: Nelson and de Burijn 2005. In sum, the informal economy is large in Tanzania since for micro enterprises the benefits of formality are dwarfed by its costs. "Whether in monetary terms (direct cost or income foregone), or in terms of time and energy, the cost of compliance turns out to be too high for most starting businesses, who are therefore obliged to start informally."15 Government is thus aiming to establish an institutional environment that is conducive to investment in the formalization of informal enterprises, since according to ILO/UNIDO/UNDP (2002) the "current regulatory set up a) fails to meet objective of ensuring quality control for the majority; and b) traps the entrepreneurs in low quality settings, puts their upgrading and growth too far out of reach and limits the contribution of the subsector to poverty reduction and national growth." In order to allow for a low-cost arena for business experimentation, as well as a means of income generation for poor and unemployed people in the absence of state-provided safety net such as unemployment benefits, small scale informal economic activities should be tolerated and supported. Forced formalization risks could damage fragile enterprises and livelihoods for very little benefits, and suppressing business experimentation and development. The decision of a small scale informal operator to formalize should be a voluntary one. In addition, it would be important to provide for an adequate institutional framework that is conducive to and provides incentives for voluntary formalization of informal businesses. Possible incentives are the abolishment of up-front payment of taxes or free training on procedures of formalization. Moreover, local governments could provide more permanent premises at low rents (such as for example the market stalls in Ilala, Dar es Salaam) for informal entrepreneurs. The Small and Medium Enterprise Development Policy of 2002 addresses the problem of infrastructural requirements and calls upon local authorities Sustaining and Sharing Growth in Tanzania 59 to allocate and develop land for SMEs, to develop industrial clusters and trade centers, and to identify and allocate underutilized public buildings to SMEs. A higher degree of realization of this policy would be very beneficial. The focus of government has been primarily on encouraging formalization of businesses, while less attention has been paid to supporting small scale, informal sector activities. In addition to the review of laws and regulations that may unduly impinge on small scale informal activities, this requires also a change in mind set of public officials recognizing the value and importance of informal sector activities for Tanzania's economic development. In particular local authorities, who deal with the informal sector on a daily basis, would benefit from capacity building in this area. Access to credit poses severe constraints to informal entrepreneurs. Often, even microfinance schemes are hard to comply with for informal businesses. Alternative credit schemes are for example rotating savings and credit societies, such as SACCOS, savings and credit cooperatives. Developing and improving occupational SACCOS can be an effective alternative to formal banks. In the long run SACCOS could function as a way to connect informal savings with the formal financial sector. Government has already started to implement measures that aim at reducing the cost of doing business and facilitating the formalization of businesses and property. In October 2004, the Property and Business Formalisation Programme was launched. The objective of the program is to provide legal titles for assets and guarantee property rights. Concluding the Diagnosis Phase of the program, the ILD presented a comprehensive report in October 2005 commenting on legislation governing property ownership and compiling data on the patterns of movable and fixed asset ownership. Formalization reforms will be designed on the basis of the findings.16 Government has also taken important measures to reduce the burden of business licensing and registration and a business activities registration bill was passed by the Tanzanian Parliament in 2007. The goal is to create a business licensing system which is transparent, and efficient, aiming at registration rather than revenue collection and control in place. Furthermore, the license fee for small businesses was abolished in 2004/05. Further to simplifying and harmonizing legislation and streamlining regulations of business and property registration, reducing the costs of formalization is important to facilitate the transition from informal to formal status. For example, the opening of branches throughout the country by BRELA, would reduce travel expenses of entrepreneurs who wish to license a business. Officers of local and national government agencies need to be well informed and trained to assist and help potential licensed business owners. They should pass on correct and useful information in order to facilitate business licensing. In particular, officers should be aware of legislative changes such as the new taxation schedule and comply with it. In addition to government officers, potential and operating informal entrepreneurs should have a fair understanding of the formalization process as well as benefits and opportunities of running a formal business. Informal workers should be encouraged to form cooperatives. Training for informal business operators in managerial skills through seminars and workshops is of utmost importance and seen as very beneficial to fostering private sector development. Furthermore, there is a need of advocacy for the informal economy. Currently, there are three associations representing the interests of informal economy operators: VIBINDO, an umbrella organization; TASISO (Tanzania Small Industrialists Society); and TAFOPA Sustaining and Sharing Growth in Tanzania 60 (Tanzania Food Processors Association). However, sectoral associations are still weak. The ILO SYNDICOOP project facilitated the formation of a national steering committee that includes the Trade Union Congress of Tanzania, Tanzania Federation of Cooperatives, Savings and Credit Cooperative union, the government and individual informal economy groups. STRENGTHENING THE CAPACITY OF THE POOR TO PARTICIPATE AND CONTRIBUTE TO ECONOMIC GROWTH To achieve its poverty reduction objectives, Tanzania needs to devote particular attention to providing the poor with the capacity to participate and contribute to the growth process. Interventions that would enhance the capacity of the poor to participate in economic growth opportunities include the upgrading of their human resources (health, nutrition, and education), facilitating the accumulation of productive assets, and mitigating the effects of risks. To benefit from growth, the human capital base of the poor must be strengthened. There is a strong relation between the level of education and poverty showing the importance of investing in well educated individuals. The productivity of labor is also constrained by malnutrition and disease. The household budget survey revealed that more than 25 percent of the population was unable to work for at least one week in the four weeks prior to the survey because of illness. Except for education, access to services and human welfare outcomes has not improved in the past years. Major progress has been made in access to primary education, though concerns about quality of education remain. Greater access to secondary education remains important. Poor households also did not benefit from improvements in access to clean water (only non-poor household did). Health remains a key problem, as most health outcomes did not improve over the 1990s. An exception may hold for infant and child mortality for which census data ­but not DHS data, suggests a decline. Twenty nine percent of those aged 15 and above are illiterate (HBS 2001) and according to the 2002 census the average years of education of the working population (those aged 20-64) is 5.1 years. Not only is this one of the factors contributing to Tanzania being one of the poorest economies in the world, but within Tanzania differences in education are strongly associated with income levels. In rural areas, where poverty is highest, the average level of education of heads of household is 4.3 years as opposed to 7.8 years in Dar es Salaam where poverty incidence is lowest (HBS 2001). Consumption regressions show that individuals living in Dar es Salaam in households headed by someone who completed secondary education have a per capita income that is 49 to 68 percent higher than that of households headed by someone with no education. In rural areas the difference is even larger: 70 to 84 percent (Table 17). Table 17. Increase in per capita consumption relative to households headed by individuals with no- education Head of household has... Dar es Salaam Other urban areas Rural Some primary education 25%-36% 19%-23% 17%-22% Completed primary education 57%-80% 30%-38% 42%-48% Some secondary education 36%-52% 42%-57% 48%-68% Completed secondary education 49%-68% 52%-67% 70%-84% Post secondary education 90%-134% 73%-93% 90%-112% Adult education only 43%-58% -1%-2% -2%-3% Sustaining and Sharing Growth in Tanzania 61 Note: Increases in consumption are determined from consumption regressions (Tables 62-68). The first number in the range is from a regression that excludes consumption information, the second number from a regression that includes this information Source: Staff calculations based on Fox and Simonsen 2005 Education is associated not only with income but also with non-income dimensions of poverty. Low levels of education lead to higher total fertility, lower levels of child nutrition (REPOA 2005), higher child mortality (Rafalimanana & Westoff 2001) and an intergenerational transfer of poverty as children from poor households are less likely to attend school themselves. The total fertility rate of women aged 40-49 is 6.5 if they do not have any education but drops to 4.9 if they completed at least primary education (TRCHS 1999). Evidence from Kagera shows that children with educated parents have better nutritional outcomes (Alderman et al. 2005). And according to the HBS, 52 percent of Tanzanian children aged 7-10 years attended school, but only 44 percent from the first two quintiles did. The Primary Education Development Program has not only resulted in a significant increase in enrollment, but evidence from Kilimajaro and Ruvuma suggests that inequalities in access to primary education that existed previously have disappeared. Figure 19 presents, for rural Kilimanjaro, concentration curves for 2001 and 2003. The concentration curve for 2001 shows how pre-PEDP access to primary education was unequally distributed: children from wealthier households attended school relatively more often than children from poorer households. The distribution was as unequal as the distribution of consumption, here represented by the Lorenz curve. With PEDP inequalities in access to education disappeared and the concentration curve coincides with the 45-degree line. Though not shown in the Figure, results for Ruvuma are comparable. Figure 19. Changes in the distribution of access to education, rural Kilimanjaro Concentration curve: Access to Primary Education 2003 100 80 Concentration- curve 60 40 Lorenz-curve 20 0 0 20 40 60 80 100 Cum. distribution of population Source: Staff calculations based on HBS 2000/01 and Sarris, Karfakis and Christiaensen, 2004. Despite major progress in enrollment and addressing aspects of the gender gap, there remain regions, particularly the poorer and more isolated ones, where the implementation of PEDP lags behind. This is evident from net enrollment figures. It is 90 percent in 2004 for the country as a whole but in Tabora it was only 68 percent and in Kigoma 77.2 percent whereas in Dar es Salaam it was 93.1 percent. Interestingly, some poor regions do particularly well, e.g. Ruvuma has a net enrollment of 99.3 percent so that poverty is not the only explanatory factor for the divergence in performance. Sustaining and Sharing Growth in Tanzania 62 In secondary education there have been improvements in recent years, but the poor remain largely excluded. The number of primary school leavers entering secondary school increased from a low of 3.4 percent in the mid eighties to 22 percent at the start of the millennium. In the past years this has increased further and in 2006, 40 percent of those that finish primary school were able to obtain a place in a government secondary school. The probability that a child aged between 14 and 18 attends secondary school is 2 percent if the child is from a household in the first consumption quintile and 13 percent if the child comes from a household in the top quintile. The prevalence of illness in Tanzania remains high. According to the 1999 TRCHS, 35 percent of children under 5 reported being affected by fever in the past 2 weeks, 12 percent were affected by diarrhea and 14 percent experienced acute respiratory infections. The HBS 2000/01 reports that 27 percent of its respondents (adults and children) indicated to have experienced illness in the preceding four weeks. Malaria/fever and diarrhea are, again, the most common causes of illness. Such high levels of illness have economic consequences. Almost one in four missed at least one week of school or work a consequence of illness that is evenly distributed across consumption quintiles (Table 18). Table 18. Number of days missed due to illness, by quintile Consumption Quintile Days missed Lowest Second Middle Fourth Highest Average None 0.30 0.31 0.29 0.30 0.32 0.30 Week or less 0.45 0.46 0.46 0.46 0.43 0.45 1 ­ 2 weeks 0.13 0.11 0.11 0.12 0.13 0.12 More than two weeks 0.13 0.11 0.11 0.12 0.13 0.12 Source: Staff calculations based on HBS 2001. In recent years, progress has been made in improving the health of Tanzanians. Life expectancy increased from 44 years in 1978 to 49 years in 1988 to 54 years for males and 56 years for females in 2002 (Census 2002). Infant mortality dropped in all regions and nationally from 115 per 1000 in 1988 to 95 per 1000 in 2002.17 In addition, child malnutrition, which remained unchanged over the course of the 1990s, declined rapidly between 1999 and 2004. Other health indicators saw less progress or even a deterioration. Maternal mortality which was 529 per 100,000 births in 1996 did not decline, and may even have increased though the increase to 578 in 2004 is not statistically significant. The share of blood donors infected by HIV/AIDS increased from 7 percent in 1994 to 12 percent in 2003 for women and from 5 percent to 8 percent for men.18 Though there are few differences between poor and non-poor households in their exposure to major diseases like malaria, diarrhea or respiratory infections or the share of budgetary expenses on health services, health outcomes differ considerably by wealth status (Table 18). Infant and child mortality is 15-20 percent higher amongst the poor than amongst those in the top quintile. The difference for nutritional indicators is even larger: 30- 50 percent. The one indicator in which poor households do substantially better than non- poor households is in HIV prevalence, which is 3.4 percent amongst those from the poorest quintile and 10.5 percent amongst individuals from the wealthiest quintile. Part of these differential health outcomes may be explained by differences in coverage of preventive health services. Poor children, for instance, are less reached by vaccination services and women from poor families and are almost three times less likely to have their birth attended by trained medical personnel than non-poor women. Sustaining and Sharing Growth in Tanzania 63 Table 19. Differences in health outcomes by quintile Quintile Indicator Lowest Second Middle Fourth Highest Average Infant mortality rate, death per 1000 114.8 107.5 115.4 106.8 91.9 107.8 Under-five mortality rate, death per 1000 160.0 159.3 192.7 155.0 135.2 161.1 % of stunted children under 5 49.5 52.5 45.0 36.6 23.4 42.7 % of underweight children under 5 32.2 35.1 28.8 23.9 21.7 28.8 HIV prevalence 3.4 4.5 5.6 9.4 10.5 7.0 Source: Gwatkin et al. 2004 from 1999 TDHS data ; HIV prevalence, THIS 2004. Differences in behavior or economic circumstance also explain difference in health outcomes between poor and non-poor households. For instance, poor individuals have less access to clean water and are less well educated. Children from the poorest quintile are five times less likely to sleep under a bed net than children from the top quintile. Individuals living in poor households are less likely to consume iodized salt, and women in the poorest quintile are almost three times more likely to have experienced female genital cutting than those from the wealthiest quintile. In acknowledging that non-health factors play an important role in determining health outcomes, the role of the health sector in improving the health of the poor (and the population at large) is put in perspective. Improving the health outcomes of Tanzanians requires a broadly shared effort across different sectors, including health, education19 and water. There are several challenges that need to be addressed in the health sector. The first is the funding situation. Though funding has improved ­the total per capita allocation of public expenditure to health increased from TShs 5,100 in 2001 to Tshs 7,374 in 2004 (MoH, 2005)20 it remains extremely low. And, with rising costs of drugs the observed increase in the budget (of 45 percent) overstates the possibility to provide additional care. Drug resistance to antimalarials and increasingly to TB treatment demands new, expensive drugs, inflating the cost of health care, without offering new services. Another challenge is the high cost of treatment for HIV/AIDS. Apart from the pressure this will create on the overall budget, due to the pattern of HIV/AIDS which affects non-poor households much more than poor households, increasing the share of financing that supports treatment will make the health budget less pro-poor..21 The health sector is affected by a serious human resource crisis. Only one third of the positions for medical officers (going by the staffing norms) is filled, and only 23 percent of the assistant medical officer and public health nurse positions are filled (MoH 2005). This human resource crisis goes back to the mid 1990's when the total health workforce decreased from around 67,000 to 49,000 in 2002, with the population increasing at the same time from 25 million to 33 million inhabitants. This affects people living in poor areas especially as, in the absence of additional incentives, it is hard to motivate medical personnel to take up positions in areas with lack of houses, no good schools, and low quality medical facilities. Malnutrition is high (over 40 percent) and is closely associated to a reduced ability for learning, earning an income and bad health. Improved nutrition is an important goal in itself which, like education, will contribute to higher future growth. High income growth is not sufficient to reduce malnutrition to acceptable levels but cost effective interventions do exist. To sustainably attack malnutrition, greater policy focus on nutrition is needed not only in the health sector, but also in other sectors, especially in agriculture and education, requiring institutional reforms. Sustaining and Sharing Growth in Tanzania 64 Nutrition rates are worst amongst the poor. According to the 1999 TRCHS 50 percent (34 percent) of children in the bottom two quintiles are stunted (respectively underweight) as opposed to 23 percent (22 percent) of children from the top quintile. Tanzanians are not only affected by protein-energy malnutrition, but many also suffer from micro-nutrient deficiencies such as iodine deficiency, iron deficiency and vitamin A deficiency. According to the 2004 TDHS only 43 percent of households use adequately iodized salt and 46 percent of children aged 6-59 months benefited from vitamin A supplementation in the six months preceding the survey. Approximately two third of children and 43 percent of women are anemic. An active population policy could support Tanzania's efforts to reduce poverty. Fertility is high and is not declining in Tanzania. In combination with the decline in infant mortality, this suggests that population growth is increasing. While the average population growth rate between 1988 and 2001 was 2.9 percent, projections carried out by the National Bureau of Statistics (URT 2006) show a population growth rate of 3.3 percent in 2003 (with a population of 34,876,231) which is projected to decline to 2.9 percent in 2025 (with a population of 68,794,180), without HIV/AIDS assumptions. Taking into account HIV/AIDS assumptions, the projections indicate that Tanzania population growth rate will increase at a rate of 2.9 percent in 2003 (with total population of 34,719,999) while in 2025 the increase will be 2.6 percent (with population of 63,516,735). High fertility leads to high dependency ratios which inhibits the ability of households to save and invest. Children from large families are worse educated and are less well nourished. High fertility contributes to (gender) inequalities and puts a budgetary strain on the ability of the public sector to deliver quality services in key sectors such as health and education. To strengthen the capacity of the poor to participate in growth, exposure to risk must also be reduced. Risks are pervasive and these disproportionately affect the poor. Risk is a structural determinant of poverty and reduces growth, because it inhibits people to take up the most remunerative activities, and because its presence reduces the long run value of the capital stock. Much can be done to reduce risk through sector interventions, including more effective health services, preventive care, agricultural research into drought resistance varieties, financial market and rural infrastructure development and enhanced access to clean water. Vulnerability is best addressed through a mix of economic growth, attention to risk reduction and a selective use of safety nets. Vulnerability results from a combination of low income and exposure to shocks and reducing vulnerability requires income growth and less exposure to risk. Safety nets should be relied upon selectively and need to be well-targeted and based on a sound understanding of the problem at hand, including the distortions safety nets may cause to the business environment. An understanding of poverty traps would help to identify whether safety nets are needed or whether bottlenecks, once addressed, would allow vulnerable groups to participate in growth. The quality of, and access to, data for monitoring and evaluation of poverty and growth need to be improved. The debate around poverty and growth is hindered by a lack of data, uneven data quality, and a restrictive provision of data to researchers and other interested parties. Entire surveys have been lost, regional GDP data appear to be inconsistent, inadequate quality control contributes to unreliable survey data, and the long time span between data collection and reporting slows the policy debate. Knowledge about causes for poverty and growth or effective interventions will remain restricted as long as leading research institutions in Tanzania do not have access to complete data sets like the HBS. Sustaining and Sharing Growth in Tanzania 65 4. MANAGING POLICIES AND RESOURCES FOR SUSTAINED SHARED GROWTH STRENGTHENING INSTITUTIONS TO STEER THE DEVELOPMENT AND IMPLEMENTATION OF TANZANIA'S STRATEGY FOR GROWTH AND POVERTY REDUCTION The focus of economic policy during the past decade has been primarily on macro- economic stabilization and reform of the public sector and more recently on poverty reduction with relatively little emphasis on the quality of economic growth and structural transformation. In addition, economic management has been fragmented among a variety of institutions which has resulted in a lack of adequate coordination of policy formulation, implementation and monitoring. A successful growth process requires institutions that ensure that government policy and its implementation is fully aligned with its growth objectives and that its implementation is monitored accordingly. In addition to coordination within government, a regular dialogue between the public and the private sector is essential for the identification of growth opportunities and concerted action by the public and private sector to exploit these opportunities. There are six key challenges that need to be addressed urgently in order to ensure that economic policy making is able to support and sustain high economic growth and to react appropriately to the evolution of the domestic and international economy. Complement economic stabilization with greater focus on economic growth and align institutional arrangements accordingly. The National Strategy for Growth and Reduction of Poverty rightly puts greater emphasis on economic growth as a key mechanism contributing to reducing poverty and suggests a more proactive role of government in the pursuit of economic growth. This will require a strengthening of institutions related to the earlier narrower scope of the PRSP and may also require evolution in the respective roles of VPO, MPEE and other central agencies in the NSGRP coordination and implementation. The government's ongoing crosscutting reforms provide an opportunity for aligning the goals and objectives of these reforms (public service, public finance, local government, and legal/judicial reforms) to support the growth agenda. Strengthen coordination of economic policy formulation and implementation. In addition, while the Ministry of Finance and the Bank of Tanzania are doing a commendable job in managing the Tanzanian economy for stability, coordinating and formulating the broader growth agenda exceeds their institutional mandates and capacities and would dilute their focus on their core responsibilities. In discussions with a variety of stakeholders, there was a clearly perceived lack of effective policy coordination. The main responsibility for the coordination of the formulation, implementation, and monitoring of a growth strategy that would rely on contributions by all parts of government needs to be located at the centre of government. Clarifying the respective roles of a number of ministries, departments and agencies (MDAs) and committees with the aim of strengthening the policy management process in Tanzania needs to be a priority. This will also include greater attention to ensuring adequate capacity to effectively carry out the coordination function for improved policy implementation in Tanzania. Create a platform for a National Dialogue on Growth related issues. The process and content of policy formulation and coordination clearly needs to be an open one, which Sustaining and Sharing Growth in Tanzania 66 encourages dialogue and participation by stakeholders. In various areas, Tanzania has been very successful in establishing processes that allow for continuous dialogue between the government and stakeholders and which draw on resources outside of government for the formulation of plans and strategies. Such processes include the Public Expenditure Review/MTEF process, where the Ministry of Finance has opened up the budget process while at the same time establishing a framework in which analytic work by various parties is coordinated and brought to bear on government processes. Similarly, the Vice President's Office has complemented the broad consultative processes on the PRSP that take place during the preparation of the PRSP or progress report with the establishment of the Research and Analysis Working Group (RAWG) which on the one hand serves as a platform for an ongoing dialogue between government and stakeholders on poverty related issues, but which also serves as an instrument to coordinate analytic work in this area. In the area of economic growth and structural transformation such a platform is currently not available. This results in a situation where a variety of initiatives proceed in parallel with minimal interaction or coordination and potential overlap and duplication. A similar set-up to the PER or the RAWG processes could be envisaged in the area of economic growth, with a focus on establishing a platform for the regular dialogue on growth related issues among the various stakeholders, including government, the private sector, civil society and development partners. Ensure adequate governance arrangements for growth enhancing government interventions. In addition to government's responsibility for creating an enabling environment for private sector activities, specific market failures need to be addressed through targeted government interventions. In the past, limited government capacity, rent seeking, bureaucratic inertia, or corruption have often resulted in ineffective programs and the waste of resources. Thus to increase the likelihood of success, it will be important to carefully select and design such interventions, and to put in place appropriate monitoring and governance arrangements that would allow the evaluation of whether a program is indeed reaching its objectives. In particular, programs should have clear criteria for success and failure, contain a sunset clause to prevent that ineffective interventions persist, focus primarily on new activities to foster new areas of comparative advantage, target activities rather than sectors, and support activities with the potential for positive spill-overs and demonstration effects. In considering such interventions, an important question has to be whether the capacity to monitor and implement such a program is indeed available. The implementation should be assigned to autonomous agencies with demonstrated capacity and clearly defined accountability to a principal who has a clear stake in the outcomes and political authority at the highest level. Redefine government­private sector relationship. Recent research on economic growth (e.g., Hausmann and Rodrik 2004) highlights the importance of public­private sector interactions in finding appropriate approaches that will result in higher growth, both at the micro- and at the macro-economic level. The annual Investor Round Table, which serves as a high level platform for dialogue between the government and the private sector is a laudable initiative which should be continued and strengthened. This needs to be supported by revisiting the private­public sector interface that takes place on an ongoing basis through central government institutions such as the Tanzania Investment Center, the Tanzania Revenue Authority, the Ministry for Industry, Trade and Marketing or the agriculture sector ministries with a view towards redefining the interaction from a purely regulatory or administrative interaction to a problem solving interaction. Strengthen private sector institutions. For the public-private sector relationship to be productive, the private sector needs to develop appropriate institutions to make its views heard. A variety of such institutions exists, including the TPSF, Confederation of Tanzania Industries (CTI), Tanzania Bankers Association (TBA), Tanzania Chamber of Mines (TCM), Tourism Council of Tanzania (TCT), Tanzania Oil Marketing Companies (TOMC), Tanzania Sustaining and Sharing Growth in Tanzania 67 Chambers of Commerce, Industry and Agriculture (TCCIA), Tanzania Association of Consultants (TACO), Tanzania Chamber of Agriculture and Livestock (TCAL). An important issue would be to survey these institutions with respect to their role in the government­ private sector dialogue, their satisfaction with the dialogue, and their capacity to effectively represent their constituents. Strengthen the capacity of institutions at the regional and district level to play a greater role in promoting growth at the local level. The study highlights the regional diversity of Tanzania's economy with respect to potential sources of growth, access to infrastructure, and natural resource endowment. A shared-growth strategy thus requires not only strong institutions at the national level, but also at the local level. Tanzania's ongoing decentralization process touches on important elements of the growth agenda. In particular, responsibility for service delivery in agriculture and infrastructure (district roads, water) is being shifted to the local authorities, while the role of the central government in these areas is limited to policy formulation and monitoring. Funding for these activities is primarily through earmarked transfers from the central government. A formula based system for budgetary transfers to the districts has been adopted, based on demographic and social indicators. In order to ensure that local authorities play a supportive role for regional growth, it will be important that the use of these resources is guided by regional growth strategies developed in partnership between local authorities, the private sector, and other stakeholders. Strengthening of accountability arrangements at the local level needs to accompany increased resource flows to the local authorities. Regional or district level growth strategies combined with strong accountability arrangements form also the basis for a successful switch from conditional to unconditional transfers, which is envisaged in the medium term. Unconditional transfers will provide local authorities with greater scope to implement a growth strategy that is tailored to the specifics of the district or region. In this context, the division of revenue sources between central government and local governments is of importance. At present, the revenue sources for local authorities are limited to a closed list and typically provide only 10 to 20 percent of a district's revenue. The current system provides scope for redistribution of resources across districts. However, a system where resource availability at the local level is more closely linked to revenue generation at the local level might provide more incentives for a greater focus on economic growth at the local level. SCALING UP PUBLIC SPENDING FOR THE IMPLEMENTATION OF THE NSGRP Public expenditures play a critical role in Tanzania's efforts to achieve sustained growth and to reduce poverty. Central government expenditures have increased from 16 percent of GDP in 1998/99 to 28 percent in 2005/06. The allocation of expenditures was guided by Tanzania's Poverty Reduction Strategy and growth related and social expenditures were the primary contributors to the overall increase in public expenditures. The annual Public Expenditure Review Process serves as an important platform for the open dialogue on public expenditure issues and the Annual Public Expenditure and Financial Assessment report provide detailed analysis of aggregate fiscal discipline and allocative and operational efficiency. Estimates of the cost of achieving the MDGs suggest that these by far exceed the currently available resources. This calls for developing an appropriate strategy for mobilizing and managing the resources required to finance the implementation of Tanzania's strategy for growth and poverty reduction. Sustaining and Sharing Growth in Tanzania 68 Potential sources of financing of the NSGRP (aside from contributions by households, the private sector, and NGOs) are taxation, seignorage, domestic and foreign borrowing, and foreign aid. Each source of financing has different macro-economic cost. In theory, as long as the marginal benefits of government spending are higher than the marginal cost of funds, it would be advisable to expand spending. The marginal benefit of additional government spending is likely to be declining while the marginal cost of finance is increasing. In Tanzania, absorptive constraints such as the shortage of qualified nurses, secondary school teachers, or road engineers are likely to result in rapidly declining marginal benefits of government spending. Regarding the marginal cost of funds, policy making in Tanzania is typically based on the assumption that these are relatively low for foreign aid inflows (but constrained by the amount donors are willing to provide) and domestic revenue. Implicitly, the cost of domestic borrowing (including the crowding out effect on private credit) and foreign, non-concessional borrowing are considered to be exceedingly high and thus excluded from the financing package. The following sections examine the potential for expanding domestic resources and aid inflows. Domestic revenue currently stands at 14 percent of GDP and is the main source of government financing. Tanzania has made substantial progress in strengthening its tax administration, which underlies most of the recent increase in revenue mobilization. Overall, the Tanzanian tax policy is sound and not inimical to growth. (FIAS, 2006) However, there are several issues which could enhance the contribution of the tax system to fostering shared growth. First, there remains an "urban bias" in the tax system, where effective tax rates are higher for farmers as for ­ mostly urban ­ businesses. In particular, the crop cess collected by local authorities imposes a relatively heavy tax burden on agriculture. A recent study (FIAS, 2006) estimates the average effective tax rate for small farmers as about 5 percent, while that for small urban businesses is estimated at around 1. 4 percent. Similarly, large businesses benefit, by paying 7.5 percent average effective tax compared to 12.5 percent for large farmers. Second, the presumptive tax regime for small businesses is one of the more sophisticatedly designed in the region. Different groups of taxpayers are categorized in the four bands depending on estimates of average turn over for particular trades or professions. The system is based on generally low tax rates and provides flexibility to small businesses of different record-keeping abilities. It provides incentives to keep records, which is good management practice, and is rewarded by a lower average tax burden. However, it is regressive for small businesses that do not keep records.22 As non-record-keepers are more likely to lack the capacity to keep proper accounts and not understand the benefits of maintaining records. While the system is thus designed to provide an incentive to keep records, its success hinges on proper outreach and education so that small businesses can better understand the benefits of keeping records and graduating to the record-keepers system. The system is inconsistent and inequitable with respect to Personal Income Tax, which has an initial tax-exempt band for total income less than TSH 960,000. Under the presumptive regime, businesses in the bottom band (turnover below TSH 3 million) that do not maintain records are required to pay TSH 35,000 irrespective of their actual turnover. Third, there are significant weaknesses in the taxation of natural resources, discussed elsewhere in this report. These weaknesses in collecting natural resource rents result both in distortions to the sustainable exploitation of natural resources and suboptimal collection of revenue for the use of natural resources. Further reforms of tax policy need to carefully consider their impact on economic growth, investment and saving. This is particularly important in the presence of large aid Sustaining and Sharing Growth in Tanzania 69 inflows and possible absorptive constraints on the overall level of government spending. Here it is important to recognize that in the medium to long run, the primary driver of enhanced revenue is an expanding revenue base through a combination of economic growth and structural transformation, which would expand and broaden the revenue base. For example, per capita economic growth of four percent would more than double per capita revenue by 2015. Foreign aid finances more than 40 percent of government expenditure and commitments at the G-8 summit in Gleneagles hold the prospect of a significant scaling up of aid. In addition to the absorptive constraints mentioned before, the macro-economic effects of aid need to be monitored closely. Our analysis of the sources of economic growth suggests that aid financed expenditures have provided an important demand side stimulus to the economy in recent years in Tanzania. In the medium to long run, aid financed investments in human and physical capital are intended to strengthen the supply side of the economy as the basis for sustained growth. However, higher aid flows do not necessarily lead to higher economic growth and the achievement of targeted development objectives, with many examples of countries that despite high inflows of foreign aid saw economic stagnation or decline.23 International experience suggests that it is critical to manage two separate, but related sets of problems that could undermine the positive impact of aid: · Weakening of institutions; and · Weakening of competitiveness through Dutch disease effects. A weakening of institutions may occur if aid undermines accountability to domestic stakeholders, distorts incentives for public sector performance, or removes the pressure for an efficient revenue collection system. It is thus important, that scaling up of aid goes hand in hand with reforms that improve governance and domestic accountability. In addition, the design of aid delivery mechanism should support domestic accountability rather than replacing it with accountability to the donor agencies. In this respect, processes such as the Tanzania Assistance Strategy and the Joint Assistance Strategy have an important role to play in proposing appropriate accountability mechanisms. Efforts to fully integrate foreign aid into Tanzania's budget system and thus make them subject to the same accountability process as domestic resources are also important. The impact on competitiveness occurs through a spending effect and a resource movement effect, which are commonly labeled "Dutch disease" effects. A "spending effect" arises from the fact that unless all inflows are spent on imports or exportable goods, the real exchange rate will appreciate which implies a deterioration of a country's competitiveness. In addition, there is a "resource movement effect," as resources would move into the booming sector (i.e. the government and development bureaucracy) and the non-traded sector. Analysis of real exchange rate movements and developments in the tradables sector do not suggest a negative impact of ODA flows in Tanzania. Simulations using a computable general equilibrium model of Tanzania also suggest that the impact of aid-inflows on the real exchange rate is likely to be modest and more than offset by aid induced productivity increases. Nonetheless, to ensure that increased aid inflows enhance rather than reduce Tanzania's international competitiveness, scaling up of aid should be accompanied by the following set of measures (Foster et. al., 2005): · ensuring that increased aid-financed spending is accompanied by increased absorption of the foreign exchange, which (assuming that government spending continues to have a high local content) will probably require acceptance of exchange rate appreciation. spending the aid without absorbing the foreign exchange does Sustaining and Sharing Growth in Tanzania 70 nothing to increase the real resources available to the economy, but makes it likely that government will crowd out the private sector; · further liberalizing imports to help increase absorption and reduce the need for real exchange appreciation or reserve accumulation; · focusing on expenditures that will quickly release supply constraints and have a higher import content, including transport investments; · continuing to improve the efficiency of the banking sector to ameliorate the need for high real interest rates; · considering adopting a more relaxed monetary policy; and · exchange rate, monetary, and fiscal policy need to be coordinated with the implications of aid inflows in mind. Following the implementation of the Enhanced Highly Indebted Poor Countries (HIPC) and the Multilateral Debt Relief (MDRI) initiatives, Tanzania's debt sustainability indicators are well below the debt sustainability thresholds. The net present value of debt- to-export ratio declined to 64 percent in 2006. Panel (a) of Figure 20 shows the baseline projections of the net present value of exports-to-GDP ratio. The ratio is projected to increase initially to 71 percent by 2010, but decline subsequently to 63 percent by 2026, assuming average export growth of 8.5 percent, and annual growth of concessional credit disbursements (primarily from World Bank and African Development Bank credits) by 2 percent. A doubling of multilateral credit disbursements in 2011 and subsequent annual growth by 2 percent would result in a gradually increasing NPV of debt-to-export ratio which would reach 101 percent by 2021 and decline thereafter. Contingent on strong GDP and export growth, significant increases in concessional borrowing seem thus to be feasible, without leading to unsustainable levels of debt. Figure 20. Multilateral credit disbursements and Debt Sustainability: Net Present Value of Debt-to- Export Ratio, 2006­26 (a) Growth of multilateral credit disbursements by 2 % 800 72 ratio 700 70 600 68 500 million 66 400 64 US$300 debt-to-export 200 62 of 100 60 0 58 NPV 2006 2008 2010 2012 2014 2016 2018 2020 2022 2024 2026 MLT disbursements (US$ millio NPV of debt-to-export ratio ( (b) Doubling of multilateral credit disbursements in 2011 Sustaining and Sharing Growth in Tanzania 71 1600 120 ratio 1400 100 1200 1000 80 million 800 60 US$ 600 40 debt-to-export 400 of 200 20 0 0 NPV 2006 2008 2010 2012 2014 2016 2018 2020 2022 2024 2026 MLT disbursements (US$ millio NPV of debt-to-export ratio ( Source: IMF and World Bank staff estimates Given Tanzania's low per capita income, aid flows that are large in relation to Tanzania's own resources are likely to remain a reality for the foreseeable future. The key to reduced aid-dependence is sustained economic growth. Changes in the revenue to GDP ratio are unlikely to significantly reduce the degree of aid dependence. However, efforts to increase efficiency and reduce transaction cost through donor harmonization require continued attention in order to maximize the impact of aid and to sustain the confidence of the donor community in supporting Tanzania's growth and reform agenda. The structure of public expenditure and revenue should preserve enough flexibility to deal with fluctuations in foreign aid. This implies avoiding a situation where spending that is difficult to scale back, such as wages and salaries, dominate. Similarly, the recurrent cost implication of donor financed projects need to be evaluated carefully. Outsourcing and public private partnerships may also be useful instruments to facilitate expenditure side adjustments, in case there is a shortfall in aid. In addition, this would be a means for developing the private sector in Tanzania by using aid inflows to provide business opportunities for the private sector. On the revenue side, an appropriate strategy may be not to fully exploit the revenue potential and to focus on a strengthening of tax administration in order to be able to scale up revenue collection to compensate for a permanent decline in aid. Finally, it should be mentioned that the implementation of the NSGRP is not only constrained by the lack of resources, but also by absorptive constraints. Absorptive constraints in Tanzania relate to the availability of trained manpower, the quality of public administration, the capacity to manage an expanded expenditure program, or the quality of the institutional and policy frameworks at the sectoral level. The limited supply of specialized skills and unattractive compensation schemes in the public sector are a serious absorptive constraint. Even at current levels of financing, a shortage of nurses, qualified teachers, road engineers, and the like, constrains the implementation of the NGSRP. In order to preserve the competitiveness of the private sector, bidding away human resources from the private sector through more attractive compensation and benefit packages should be avoided. Instead, emphasis needs to be placed on cooperation with the private sector and expansion of the skill pool through training. As a low income developing economy, the policy and institutional capacity of Tanzania is modest. While in some of the several sectors the institutional capacity is reasonable, in others Sustaining and Sharing Growth in Tanzania 72 it remains weak. Sector development programs are an important instrument for putting appropriate policy and institutional frameworks in place. Sectors that already have operational sector development programs in place, such as the education and health sectors, are able to use available resources reasonably effectively, and could absorb additional resources. In other sectors such as roads, power, or water, major increases in budgetary funding will only be effective as policy, institutional, and capacity constraints are addressed. With the change in the role of government in many sectors from implementer to regulator, the capacity of the private sector is a factor. In the water and roads sectors, for example, there are not enough private contractors with adequate expertise and necessary equipment who can handle tasks or deliver services on time. The weakness is again acute in rural areas. However, experience from other countries indicates that the private sector is quick to develop adequate capacities once business opportunities become available. While public expenditure and financial management at the central government level is fairly well developed in Tanzania, capacity in local authorities is generally much weaker and shows much variation across districts. At present, most government expenditure, including spending by local authorities funded through central government transfers, is under fairly tight control of the central government with only limited scope for leakages at the district level. The public finance management reform program as well as the local government reform program are the principal mechanisms for strengthening financial management systems and capacities at the local level to ensure that as decentralization progresses adequate capacities are being put in place. Tanzania's public sector reform program has contributed significantly to the rationalization and strengthening of the public administration. Nonetheless, various factors still constrain the quality of public administration, including low pay levels and a generally low level of education in the population which is reflected in the level of education found among civil servants. Capacity constraints are generally larger at the district than at the central government level. The Local Government Reform Program is intended to strengthen the capacity of local authorities to plan and execute programs as under the fiscal decentralization program responsibilities are shifted to them. Increasing pay levels in the public sector is seen as one of the pre-conditions for creating an environment in which other reforms to enhance public sector performance can become effective. Government is implementing a Medium Term Pay Policy that seeks to increase and decompress government wages. Significant wage increases were granted in 2006, raising the government wage bill to 5.9 percent of GDP. A comprehensive review of the MTPP has been undertaken by the Presidential Commission on Public Service Pay, with a view to examine options to accelerate the implementation of the pay policy by implementing the policy in relation to the employment levels for each ministry, department, and agency and their budgetary implications. Sustaining and Sharing Growth in Tanzania 73 References Anand S., and J. Morduch. 1996. Poverty and the `Population Problem'. The Harvard Center for Population and Development Studies and Department of Economics, Harvard University. Bank of Tanzania. 2004. Tanzania Investment Report 2004. Dar es Salaam. Barro R. 1991. Economic Growth in a Cross Section of Countries. Quarterly Journal of Economics 106(2): 407-441. Behrman J., H. Alderman and J. Hoddinott 2004. Hunger and Malnutrition. Copenhagen Consensus: Challenges and Opportunities. Bosworth, Barry, and Susan Collins. 2003. The Empirics of Growth: An Update. Brookings Institution. Washington. D.C. Christiaensen L., V. Hofmann and A. Sarris. 2004. Coffee Price Risk in Perspective: Vulnerability Among Small Holder Coffee Growers in Tanzania. Unpublished Manuscript. World Bank. Washington D.C. Christiaensen, L. and Lionel Demery. 2006. The Role of Agriculture in Poverty Reduction in Sub Saharan Africa Revisited. Processed. May, 2006. Collins, Susan and Barry Bosworth. 1996. Economic Growth in East Asia: Accumulation versus Assimilation. Brookings Papers on Economic Activity, 2:1996. De Soto, Hernando. 2001. The Mystery of Capital: Why Capitalism Triumphs in the West and Fails Everywhere Else. Basic Books. Deaton A. and S. Zaidi 2002. Guidelines for Constructing Consumption Aggregates for Welfare Analysis. World Bank. LSMS Working Paper No. 135. Dercon S. 1996. Risk, Crop Choice, and Savings: Evidence from Tanzania. Economic Development and Cultural Change 44(3): 485-513. Dercon S. 1998. Wealth, Risk and Activity Choice: Cattle in Western Tanzania. Journal of Development Economics 55(1): 1-42. Dercon S., T. Bold, J. de Weerdt and A. Pankhurst. 2004. Extending Insurance? Funeral Associations in Ethiopia and Tanzania. Unpublished Manuscript. Mimeo. Economic and Social Research Foundation. 2005. Vulnerability and Resilience to Poverty in Tanzania: Causes, Consequences and Policy Implications. Dar es Salaam. Elbers C., J.W. Gunning and B. Kinsey. 2003. Growth and Risk. Tinbergen Institute Discussion Paper. TI 2003-068/2. Sustaining and Sharing Growth in Tanzania 74 Foreign Investment Advisory Service (FIAS). 2004. Free Zones: Performance, Lessons Learnt, and Implications for Zone Development. Final Draft. November 2004. Foreign Investment Advisory Service (FIAS). 2006. Sector Study of the Effective Tax Burden ­ Tanzania. Processed. April 2006. Fink, Carsten, Aaditya Mattoo and Ileana Cristina Neagu. 2002. Assessing the Role of Communications Costs in International Trade. Working Paper #2929, 2002. The World Bank: Washington, D.C. Food and Agriculture Organization (FAO). 1997. Irrigation Technology in Support of Food Security. Proceedings of a Regional Workshop, Rome, 1997. Food and Agriculture Organization (FAO). 2004. AQUASTAT. FAO's global information system on water and agriculture. FAO. Rome. http://www.fao.org/AG/agl/aglw/aquastat/water_res/index.stm Food and Agriculture Organization (FAO) and World Bank. 2001. Tanzania Agriculture: Reaping the Benefits of Reform. Foster, M. et al. 2005. The Macroeconomic Impact of Scaling up Aid Flows: The Case Study of Tanzania. Final report prepared for the ODI ­ London. Mimeo. September 9, 2005. Harding, A., G. Kahyarara, and Neil Rankin. 2002. Firm Growth, Productivity and Earnings in Tanzania 1992-1999. Center for the Study of African Economies (CSAE). University of Oxford. Hausmann, Ricardo, Dani Rodrik, Andreas Velasco. 2004. Growth Diagnostics. September 2004. Hausmann, Ricardo, and Dani Rodrick. 2005. Self-Discovery in a Development Strategy for El Salvador. Economía - Journal of the Latin American and Caribbean Economic Association 6(1): 43-101. Hemming, Richard, Michael Kell, and Selma Mahfouz. 2002. The Effectiveness of Fiscal Policy in Stimulating Economic Activity ­ A review of the Literature. IMF Working Paper WP/02/08. December 2002. Horton S. 1999. Opportunities for Investment in Low Income Asia. Asian Development Review 17(1,2): 246-273. Horton S. and J. Ross. 2003. The Economics of Iron Efficiency. Food Policy 28(1):51-75. Huppi, Monica, and Martin Ravallion. 1991. The Sectoral Structure of Poverty in Indonesia during an Adjustment Period: Evidence for Indonesia in the Mid-1980s. World Development 19 (12): 1653-78. IMF. 2004. Tanzania ­ Selected Issues and Statistical Appendix. July 22, 2004. Kessy F. 2004. Rural Income Dynamics in Kagera Region, Tanzania. ESRF. Dar es Salaam. Background paper to CEM. Kolavalli, S. and S. Beddies. 2004. Report on Phase One: Institutional Mapping of Coffee Sustaining and Sharing Growth in Tanzania 75 and Cotton, Reform of the Tanzania Cotton, Coffee, Cashew and Tea Boards: Options and Implications for Other Crop Boards. World Bank, Washington, DC. Kweka Josaphat P. and Oliver Morrissey. 1999. Government Spending and Economic Growth Empirical Evidence from Tanzania (1965-1996). Draft prepared for DSA Annual Conference. University of Bath. September 12-14, 1999. Lanjouw P and M. Ravallion.(1995). Poverty and Household Size. The Economic Journal, 105 (November), 1415-1434. Lorenz N. and C. Mpemba. 2005. Review of the State of Health in Tanzania 2004. Mimeo. Klasen S. 2003. In search of the holy grail: how to achieve pro-poor growth? Annual World Bank Conference on Development Economics ­Europe. Toward Pro-Poor Policies: Aid Institutions and Globalization. Edited by B. Tungodden, N. Stern and I. Kolstad. World Bank / Oxford University Press: New York. pp: 63-94. Larson D. and F. Plessmann. 2002. Do Farmers Choose to be Inefficient? Evidence from Bicol, Philippines. Policy Research Working Paper 2787. World Bank, Washington D.C. Lindeboom W. and B. Kilama. 2005. Trends and Determinants of Malnutrition in Tanzania. Paper Presented at the 10th Annual REPOA Research Workshop. Mankiw N.G., J. Roemer and P. Weil 1992. A Contribution to the Empirics of Economic Growth. Quarterly Journal of Economics 107(2): 407-37. Ministry of Education and Culture, 2004 & 2005. Basic Statistics in Education 1999-2003. National Data. Ministry of Education and Culture, Dar es Salaam. Ministry of Health. 2004. National Aids Control Program. HIV/AIDS/STI Surveillance Report. January-December 2003. National Bureau of Statistics. 2002. Household Budget Survey 2000/01. National Bureau of Statistics. Dar es Salaam. Nelson, Edwin G. and Erik J. de Burijn. 2005. The Voluntary Formalization of Enterprises in a Developing Economy ­ The Case of Tanzania. Journal of International Development, Issue 17, 2005. Nyange, David. 2005. Survey of Agricultural Marketing Logistics Costs in Rural Tanzania. background paper for the Tanzania DTIS. Owens, Trudy, and Francis Teal. 2005. Paths out of Poverty in Ghana and Tanzania in the 1990s. Mimeo. Rafalimanana H. and C. Westoff 2001. Gap Between Preferred and Actual Birth Intervals in Sub-Saharan Africa: Implications for Fertility and Child Health. DHS Analytical Studies No. 2. Calverton Maryland: ORC Macro. Rajan, Raghuram G., and Arvind Subramanian. 2005. What Undermines Aid's Impact on Growth? IMF Working Paper 05/126. International Monetary Fund. Washington, D.C. Sustaining and Sharing Growth in Tanzania 76 Rodrik, Dani. 1998. Saving Transitions. Draft. July 1998. Rosenzweig M.R. and K. Wolpin 1993. Credit Market Constraints, Consumption Smoothing and the Accumulation of Durable Production Assets in Low Income Countries: Investments in Bullocks in India. Journal of Political Economy 101(2): 223-224. Sarris, A., Sara Savastano, and Luc Christiaensen. 2006. The role of Agriculture in Reducing Poverty in Tanzania : A Household Perspective from Rural Kilimanjaro and Ruvuma. Mimeo. May 2006. Schclarek, Alfredo. 2005. Fiscal Policy and Private Consumption in Industrial and Developing Countries. Draft. September 30, 2005. Schneider, Friedrich. 2004. The Size of the Shadow Economies of 145 Countries all over the World: First Results over the Period 1999 to 2003. IZA Discussion Paper Series. IZA DP No. 1431. December 2004. Soderbom, Mans, Francis Teal, Anthony Wambugu and Godius Kahyarar. 2004. The Dynamics of Returns to Education in Kenyan and Tanzanian Manufacturing. Centre for the Study of African Economies (CSAE), University of Oxford, CSAE WPS/2003-17, January 2004. UN Millennium Project. 2005. Investing in Development: A Practical Plan to Achieve the Millennium Development Goals. Earthscan. London. UNICEF. 1990. Women and Children in Tanzania: An Overview. UNICEF, Dar es Salaam. United Republic of Tanzania. 2004. Tanzania Mini-Tiger Plan 2020. December 2004. United Republic of Tanzania. 2006. 2002 Population and Housing Census: Projections. Volume VI. National Bureau of Statistics. Ministry of Planning, Economy, and Empowerment. Dar es Salaam. World Bank. 2001. Social Protection Sector Strategy: from Safety Net to Springboard. Sector Strategy Paper. The World Bank. Washington D.C. World Bank. 2004a. World Development Report 2005. A Better Investment Climate for Everyone. Oxford University Press. New York. World Bank. 2004b. Investment Climate Assessment: Improving Enterprise Performance and Growth in Tanzania. Washington, DC: World Bank. World Bank. 2005a. Tanzania ­ Diagnostic Trade Integration Study. June 2005. Washington, DC: World Bank. World Bank. 2005b. Global Monitoring Report 2005. Washington, DC: World Bank. World Bank. United Republic of Tanzania ­ Public Expenditure Review FY03. June 2003. World Bank. 2006a. Doing Business 2007: How to Reform. Washington D.C. World Bank. 2006b. Tanzania Pilot Rural Investment Climate Assessment. Stimulating Non- farm Microenterprise Growth. Mimeo. June 27, 2006. Sustaining and Sharing Growth in Tanzania 77 ANNEX 1: LIST OF BACKGROUND STUDIES Alderman, Harold, J. G. M. Hoogeveen, and Mariacristina Rossi. 2005. "Reducing Child Malnutrition in Tanzania: Combined Effects of Income Growth and Program Interventions." World Bank, Washington, DC. Aubert, Jean-Eric, and Godwill Wanga. 2005. "Innovation in Tanzania: Insights, Issues, and Policies." World Bank, Washington, DC. Chandra, Vandana, Pooja Kacker, and Ying Li. 2005. "Tanzania: Growth , Exports, and Employment in the Manufacturing Sector." World Bank, Washington, DC. Christiaensen Luc, Vivian Hofmann, and Alexander Sarris. 2004. "Coffee Price Risk in Perspective: Vulnerability among Small Holder Coffee Growers in Tanzania." World Bank, Washington, DC. COWI. 2005. "Natural Resource Based Growth: Summary Paper." World Bank, Washington, DC. Demombynes, Gabriel, and J. G. M. Hoogeveen. 2004. "Growth, Inequality, and Simulated Poverty Paths for Tanzania, 1992­2002." Policy Research Working Paper 3432, World Bank, Washington, DC. Hoogeveen, J. G. M. 2004. "The Distributional Impact of the PEDP in Rural Kilimanjaro." World Bank, Washington, DC. ------. 2005. "Risk, Growth, and Transfers: Prioritizing Policies in a Low-Income Environment with Risk--The Case of Tanzania." World Bank, Washington, DC. infoDev. 2005. Improving Competitiveness in Tanzania: The Role of Information and Communication Technologies. October 2005. Kessy, Flora. 2004. "Rural Income Dynamics in Kagera Region, Tanzania." Economic and Social Research Foundation, Dar es Salaam. Kessy, Flora, and Oswald Mashindano. 2005. "Moving Out of Poverty: Understanding Growth and Democracy from the Bottom Up--The Case of Ruvuma Region, Tanzania." Economic and Social Research Foundation, Dar es Salaam. Kilama, Blandina, and Wietze Lindenboom. 2004. "Trends in Malnutrition in Tanzania." Research on Poverty Alleviation, Dar es Salaam. Kopicki, Ronald. 2004. "Supply Chain Development in Tanzania. World Bank. Washington D.C. Mahamba, Robert, and Jorgen Levin. 2005. "Economic Growth, Sectoral Linkages, and Poverty Reduction in Tanzania" World Bank, Washington, DC. Makki Shiva S., IJsbrand de Jong, and Henry Mahoo. 2005. Smallholder Ground Water Irrigation in Tanzania. World Bank. Washington D.C. Mkenda, Adolf. 2004. "The Benefits of Malnutrition Interventions: Empirical Evidence and Lessons to Tanzania." University of Dar es Salaam, Dar es Salaam. ------. 2005. Population Growth, Economic Growth and Welfare Distribution: An Overview of Theory, Empirical Evidence and Implications to Tanzania. University of Dar es Salaam, Dar es Salaam. Mpango, Philip. 2005. "Spatial Dimensions of Growth and Poverty Reduction in Tanzania Mainland." World Bank, Washington, DC. Simonsen, Marianne., and Louise. Fox. 2005. "A Profile of Poverty in Tanzania." World Bank, Washington, DC. Skof, Annabella. 2006. "Constraints to Technology Access in Tanzanian Horticulture: A Case Study of Barriers to the Introduction of Improved Seed and Pest Control Technologies." World Bank, Washington, DC. Sustaining and Sharing Growth in Tanzania 78 Tanzania Food and Nutrition Centre. 2004. "Causes of Malnutrition and Tanzania's Nutrition Programs: Past and Present." World Bank, Washington, DC. Utz, Anuja. 2004. "Fostering Innovation, Productivity, and Technological Change: Tanzania in the Knowledge Economy." Knowledge for Development Program, World Bank Institute, Washington, DC. Utz, Robert. 2005. "Review of Growth Performance and Prospects." World Bank, Washington, DC. van Dijk, Meine Pieter. 2006. Urban Rural Dynamics in Tanzania, through informal redistribution mechanisms. Mimeo. 10/15/2006. Sustaining and Sharing Growth in Tanzania 79 Notes 1 To the extent that increased government spending fell on imports, the effect on growth is reduced. 2 Shares represent the average distribution of FDI by Sector, 1999-2001 (Bank of Tanzania. 2004) The KAM (http://www.worldbank.org/kam) includes 80 quantitative and qualitative 3 variables that help to benchmark a country's position, vis a vis 128 countries, on key elements of the four pillars of the KE framework (economic and institutional regime, education, innovation, and ICTs). 4 Regional GDP data for Tanzania could be overstating regional income disparities partly because production in the regions which contribute little to national GDP is largely for subsistence and is not fully captured in market-based GDP numbers. The purchasing power of the shilling also tends to be higher in the poorer regions. 5See Huppi and Ravallion (1991). 6Britain, Canada, Netherlands, Sweden. 7 The median wait was considerably shorter than the average typical wait because several enterprises reported exceptionally long delays. For example, nine enterprises reported typical delays of over 50 days. 8 United Republic of Tanzania, Legal Sector Reform Programme Medium Term Strategy FYs 2005/6-2007/8. 9 As much as 25% of potentially suitable additional pasture land is affected by tse tse fly and cannot be used for cattle at present. 10 Christiaensen, L. and Lionel Demery. 2006. The Role of Agriculture in Poverty Reduction in Sub Saharan Africa Revisited. Processed. May, 2006. 11These are figures calculated on the basis of the national definition of employment. 12 Data from the Integrated Labour Force Survey (National Bureau of Statistics 2001) and the Household Budget Survey (National Bureau of Statistics 2002) provide different measures of informal sector activities, but the magnitude and trends provided by these two surveys present broadly similar pictures. 13 Regulatory constraints are complicated, lengthy and unpredictable procedures, inadequate institutional arrangements, rent-seeking civil servants, unreasonable specifications and standards or a multiplicity of taxes and levies etc. Non-regulatory constraints are for example poor clients, lack of access to finance, poor infrastructure, unfair competition. The 75 percent is the highest average cost for coping with constraints that occurred in the firewood and charcoal sector. In the cloth making sector, for example, the highest average cost calculated was 5.8% of monthly sales. 14This has been shown ­ among others ­ by the ILO/UNIDO/UNDP study. Sustaining and Sharing Growth in Tanzania 80 15ILO/UNIDO/UNDP (2002): "Roadmap Study of the Informal Sector in Mainland Tanzania", Dar es Salaam, April 2002, p. 3. 16Instituto Libertad y Democracia (2005): "Program to Formalize the Assets of the Poor of Tanzania and Strengthen the Rule of Law: Final Diagnosis Report." Lima, September 2005. 17The decline in infant mortality is contested. The DHS and TRSCH surveys carried out during the 1990s show a stagnation of mortality rates up to 1996 and slight increase thereafter. Data from surveillance sites, on the other hand, supports the decline in infant and child mortality. 18There is some evidence of a decline in the infection rate amongst blood donors. It is not clear whether this is a statistical artifact or a reflection of a decline in the infection rate. 19 Apart from learning students proper care practices, one low-cost nutrition intervention to reduce anemia would be deworming of all children at first day of school. 20In constant 2001 TShs. The nominal amount for 2004 is TShs 8,815. The 2001- 2004 price index is calculated from the Economic Survey, 2004, Table 11. 21This is not to say that the use of ARVs should not be promoted, or that there are no pro-poor elements to ARV treatment. Apart from the life saving benefits, HIV/AIDS and the prolonged illness that precedes it creates a large economic burden. And because many people infected decide to go home (poor) rural households carry a disproportionate share of this burden (de Waal et al. 2004). 22About 70 percent of presumptive tax payers registered with TRA do not keep proper records, despite the fact that this results in a higher average effective tax rate. 23Rajan and Subramanian (2005) provide a concise summary of the debate.