67703-GM Report No. XXX-GM THE GAMBIA FISCAL DEVELOPMENTS AND THE AGRICULTURE SECTOR Public Expenditure Review Update June 2006 PREM 4 Africa Region Document of the World Bank TABLE OF CONTENTS 1. THE MACROECONOMY AND FISCAL POLICY ............................. 1 Introduction............................................................................................................. 1 Macroeconomic Performance ................................................................................. 2 2005 Fiscal Performance......................................................................................... 4 2006 Budget .......................................................................................................... 10 PRSP Expenditures ............................................................................................... 11 Domestic Debt ...................................................................................................... 13 2. THE AGRICULTURAL SECTOR .................................................. 17 Introduction........................................................................................................... 17 Sectoral Comparison............................................................................................. 18 Economic Classification ....................................................................................... 19 Administrative Classification................................................................................ 19 Groundnut Sector .................................................................................................. 20 Development Expenditures................................................................................... 22 Research ................................................................................................................ 27 Extension Services ................................................................................................ 30 Input Supplies (Fertilizers) ................................................................................... 31 Decentralization .................................................................................................... 34 Conclusion ............................................................................................................ 35 Tables Table 1.1: Key Macroeconomic Indicators, 2001-2005 ..................................................... 2 Table 1.2: Central Government Operations by Economic Classification........................... 7 Table 1.3: Domestically Financed Expenditures by Administrative Classification ........... 8 Table 1.4: Domestically Finance PRSP Expenditures...................................................... 12 Table 2.1: Recurrent Budget Allocations of PRSP Priority Sectors ................................. 18 Table 2.2: Cross-Country Comparaison of Public Expenditures on Agriculture ............. 18 Table 2.3: DOSA Recurrent Expenditures by Economic Classifications......................... 19 Table 2.4: ANR Recurrent Expenditures by Department ................................................. 20 Table 2.5: Government Support to the Groundnut Sector ................................................ 21 Table 2.6: 2006 Agriculture Development Budget by Projects........................................ 24 Table 2.7: 2006 Agriculture Development Budget by Programs ..................................... 26 Table 2.8: 2005 NARI Research Programs....................................................................... 27 Table 2.9: Funding Shares for Public Agricultural Research in SSA Countries .............. 28 Table 2.10: Government Supplied Fertilizers................................................................... 32 Table 2.11: DOSA Staff Allocations ................................................................................ 34 i Figures Figure 1.1: Sectoral Growth................................................................................................ 2 Figure 1.2: Monetary Developments .................................................................................. 3 Figure 1.3: Current Account Balance ................................................................................. 4 Figure 1.4: Domestic Revenues, 1995-2005....................................................................... 5 Figure 1.5: Domestic Revenues, 1995-2005....................................................................... 5 Figure 1.6: Recurrent Expenditures, 1995-2005................................................................. 6 Figure 1.7: Domestic Debt................................................................................................ 13 Figure 1.8: Fiscal and Monetary Causes of Domestic Debt Increases ............................. 14 Figure 1.9: Domestic Debt Projections............................................................................. 15 Figure 2.1: DOSA Actual Expenditures .......................................................................... 23 Figure 2.2: DOSA Budget Execution Rates ..................................................................... 23 Figure 2.3: Agriculture Development Budget by Programs ............................................. 25 Boxes Box 1.1: Comparaison of Budget Outturn Between Education and Health ....................... 9 Annexes Annex 1: Estimate of Operating Costs for DAS Extension Services ............................... 37 Annex 2: Estimate of Operating Costs for DLS Extension Services................................ 38 ii Currency Equivalents Currency Unit = Dalasi (GMD) US$1 = 28.38 GMD (as of June 27, 2005) Fiscal Year January 1 – December 31 ACRONYMS AND ABBREVIATIONS ACS Administrative and Client Support ACU Aid Coordination Unit AGD Accountant General’s Department CBG Central Bank of The Gambia CED Customs and Excise Department CRD Central Revenue Authority DAS Department of Agricultural Services DLDM Directorate of Loans and Debt Management DLS Department of Livestock Services DOS Department of State DOSA Department of State for Agriculture DOSFEA Department of State for Finance and Economic Affairs DOSLG&L Department of State for Local Government and Lands DT Directorate of Treasury EC European Commission FAO Food and Agriculture Organization GIPFZA The Gambia Investment Promotion and Free Zones Agency GLF Gambia Local Fund HIPC Heavily Indebted Poor Countries ICRG International Country Risk Guide IFMIS Integrated Financial Management Information System PAC Public Accounts Committee PER Public Expenditure Review PRER Poverty Reduction Expenditure Reports PRGF Poverty Reduction and Growth Facility PRSP Poverty Reduction Strategy Paper RA Revenue Authority SSA Sub-Saharan Africa SD Spending Department SDR Special Drawing Rights Vice President: Gobin T. Nankani Country Director: Madani M. Tall Sector Manager: Robert Blake Task Team Leader: Hoon S. Soh iii Acknowledgement The PER Update exercise was jointly conducted by the authorities, the Bank and FAO. The identification mission took place in October 2005, and the main field work took place during a mission in January 2006. This PER Update is the third in an annual series of joint PERs which started in 2004. The January 2006 mission consisted of Hoon S. Soh (Task Team Leader, AFTP4), Satish Kumar (agricultural specialist, World Bank consultant), and Buddhika Samarasinghe (groundnut sector specialist, FAO consultant). Discussions were held with senior representatives of the Department of States of Finance and Economic Affairs (DOSFEA), the Department of State of Agriculture (DOSA), the Department of State of Trade and Industry (DOSTI), and various public and private agencies related to the agriculture sector. The mission undertook upcountry trips to engage local public and private representatives. Christopher Willford (MTEF technical adviser, DOSFEA) was mainly responsible for providing the budget data. The exercise for the agricultural sector was significantly participatory. A series of workshops were held in which key departmental representatives providing inputs and participated in the discussions. The chapter on agriculture was jointly prepared by the authorities and the donor team. A more detailed report on the groundnut sector, which includes sectoral policy analysis, was separately prepared. The report reflects the comments from the review meeting which took place in June 2006, particularly comments by the peer reviewers Jane Hopkins (Senior Agriculture Economist, AFTS4) and William Sutton (Agriculture Economist, ECSSD). The team would also like to thank the guidance provide by Robert Blake (Sector Manager, AFTP4), Madani Tall (Country Director, AFC14), and Iradj Alikhani (Country Program Coordinator, AFCSN). Josette Percival (ACS staff, AFTP4) provided editorial assistance. iv Executive Summary The objectives of this Public Expenditure Review (PER) Update are to analyze public expenditures of the overall budget and the agriculture sector in greater detail. This PER Update is the third in a series of PERs conducted jointly by the authorities and the donors. The first PER conducted in 2004 was a full PER which focused on analyzing overall government expenditures, public financial management capacity, and the education and health sectors. The second PER conducted in 2005 was a PER Update which limited its analysis to government expenditures and public financial management. This third PER is also an update, thus limited in scope. The authorities prepared a sectoral PER of the agriculture and natural resources sectors in 2001. The second chapter of this PER Update is a partial update of the 2001 sectoral PER. It focuses only on the agriculture sector and excludes the natural resource sectors, and it does not analyze the sector strategy or the latest sector developments. The topics were chosen through discussions with the authorities during an identification mission. The timing of the report was mainly dictated by data availability. However, the report can be easily incorporated into the budget preparation cycle which typically takes place in the second half of the calendar year. THE MACROECONOMY AND FISCAL POLICY The authorities have generally adhered to prudent fiscal and monetary policies since the economic downturn in 2002, resulting in a resumption of growth, lowering of inflation, and stabilization of the exchange rate. However, sustained fiscal consolidation is required in the long term in order to reverse adverse debt dynamics such that domestic debt is put on a sustainable path. This would expand discretionary fiscal space which allows greater flexibility in reallocating resources to PRSP related activities. From 2003 to 2005, the primary fiscal balance increased from 3.6 percent to 8.5 percent of GDP, and the annual growth of broad money decreased from 43.4 percent to 9.4 percent. Inflation fell to single digits, the exchange rate stabilized, and annual GDP growth resumed in the five to six percent range. The current account deficit widened considerably, largely due to an upsurge in imports financed by large capital inflows. In order to sterilize the capital inflows, the authorities issued treasury bill up to 8.9 percent of GDP on a net basis, further increasing the domestic debt. 2005 Fiscal Performance. Although the primary balance remained a significant surplus, the overall fiscal deficit including grants increased to 8.6 percent of GDP due to a shortfall in tax on international trade, increased domestic debt service and donor project disbursements, and extrabudgetary expenditures of 1.0 percent. In response, the authorities reduced discretionary expenditures, and increased domestic financing to 3.6 percent of GDP. v Debt service continued their upward trend in 2005, increasing to 8.6 percent of GDP. Accounting for 46.9 percent of recurrent expenditures, debt service effectively crowded out spending on wages and salaries and operations and maintenance. The increasing debt service also reduced the recurrent expenditures of most Departments of State (Ministries) and public agencies compared to their budget allocations. However, spending on general administration as a whole was largely protected, while mainly expenditures on social services and economic services were reduced. This is reflected in the decrease of the share of PRSP related expenditures. Nearly half of the domestically funded development expenditures were for roads projects. With the imminent start of a large road construction and rehabilitation project funded by the EU, the authorities could consider reallocating some of the government’s own funds to other priority areas. 2006 Budget. The 2006 budget maintains fiscal discipline. The targeted overall deficit of 3.0 percent of GDP and a basic primary surplus of 10.0 percent of GDP are both fairly ambitious. They will be achieved through higher domestic revenue, lower domestic interest payments, and lower externally funded development expenditures. In particular, the targeted domestic revenue for 2006 is 3.0 percent of GDP higher than the average budget outturn of the three previous years, and 2.0 percent of GDP higher than the previous year. Domestic debt service is expected to decrease from 6.8 percent to 5.0 percent of GDP mainly due to lower domestic interest rates. The discount rates on the 90 day treasury bills had steadily decreased from 30.0 percent in December 2004 to 8.5 percent in December 2005. Although the rate has subsequently increased to 13.0 percent in April 2006, the targeted domestic debt services of 5.0 percent of GDP in 2006 should be roughly achievable as long as the average interest rates for the year remain approximately at 15.0 percent. The key fiscal challenge is to contain expenditures for the African Union (AU) Summit meeting and the presidential election. As a result of these two expenditure items, the shares of spending on social and economic services decrease in the 2006 budget. They also result in a decrease in the share of PRSP expenditures, with the majority of the decrease coming from PRSP programs in basic social services and infrastructure, particularly health related expenditures. Domestic Debt. Domestic public debt is one of the most critical policy issues in the country. It reached 35.5 percent of GDP at the end of 2005. By contrast, the average stock of domestic debt in 27 non-CFA SSA countries is 15 percent of GDP, while the median is only 10 percent. As a percentage of GDP, the domestic public debt tripled in a little over a decade from 12.3 percent in 1994. Its explosive growth have mainly been due to government borrowing, but monetary operations have also played a major role in recent years, particularly in 2004 when the authorities sought to sterilize large capital inflow. vi Simulations of domestic debt indicate that the authorities would be able to approximately maintain the current level of debt to GDP ratio by limiting annual domestic borrowing to 1.5 percent of GDP. Without additional principal repayments, sustained domestics saving of 1.4 percent would be needed to reduce domestic debt to 15 percent of GDP by 2015, the average of non-CFA SSA countries. Hence, sustaining fiscal constraint is critical. In addition, the domestic debt can be further reduced by repaying principals using any additional sources of funding, such as from future debt relief. AGRICULTURE SECTOR Recurrent Expenditures. The Department of State for Agriculture (DOSA) receives a share of expenditures which is reasonably close to the SSA average, but significantly less than the average of all development countries. There appears to be key structural imbalances in DOSA’s intrasectoral resource allocations: (i) The share of inputs reached 31.1 percent in the 2006 budget, the largest share in the budget. It indicates the government’s increasing involvement in the supply of inputs, particularly fertilizers which are typically considered private goods. Responsibilities for such inputs could be transferred to the private sector, thus reducing public expenditures and increasing the public good component of agricultural expenditures. (ii) The share of expenditures for extension services decreased from 73.8 percent in 2001 to 30.3 percent in the 2006 budget. Traditionally, the share of expenditures for the Department of Agricultural Services, responsible for crop extensions, was by far the largest and at times more than half of DOSA’s expenditures. This decrease raises concerns that extension services are under-funded. A more detailed analysis of extension services corroborates these concerns. (iii) Preliminary figures indicate that the government has spent up to 134 million dalasis to support the groundnut sector during the past four years, equivalent to approximately 70 percent of the recurrent expenditures of DOSA, 0.3 percent of GDP, and approximately 2,350 dalasis per groundnut farming household. The authorities support the sector through equity investments, loans, subsidized fertilizers, and producer price support. In addition, they provide indirect support to the sector through government guarantees of loans. Despite significant expenditures, it does not appear that there have been noticeable improvements in sector performance or poverty levels among groundnut farmers. Development Expenditures. Development expenditures for agriculture have been on a downward slide in recent years, decreasing from 2.1 percent of total expenditures in 2001 to 0.7 percent in 2005. The decline has been due to both lower budget allocations as well as lower budget execution rates. Further analysis would be required to identify the exact causes. They could be due to weaknesses in budget preparation, foreign aid coordination and project management. vii The composition of DOSA’s 2006 development budget is consistent with a sector strategy based on both import substitution and export promotion. The emphasis on livestock and rice cultivation reflects policies to strengthen food security through import substitution. Projects on horticulture support a sector which has long been touted for its export potential. There is a risk that a strategy based on both import substitution and export promotion would be poorly implemented given limited resources and capacity. The agriculture sector strategy should further prioritize subsectors. Research (National Agricultural Research Institute (NARI)). The major findings are: (i) research on livestock, fisheries, marketing and land tenure are relatively neglected with respect to NARI’s long term strategy; (ii) overall funding is low compared to other SSA countries; and (iii) personnel costs are excessive compared to other SSA countries. Further analysis would be required to determine the impact of NARI’s research, the effectiveness of linkages with extension services, and the potential for public and private enterprises to contribute to agricultural research. Extension Services. The analysis indicates that the level of staffing is adequate according to the ratio of farming households to extension workers targeted by the authorities. By contrast, it appears that budget allocations are less than the minimum operating expenses required for the extension workers. Based on unit cost calculations, the 2006 budget allocations cover only 17.9 percent of crop extension’s minimum requirement, and 20.0 percent for livestock extension. The authorities had long proposed consolidating the crop and livestock extension service departments, but progress has been slow. Such a consolidation has the potential for cost savings. It would have to overcome institutional rigidities in which extension services for crops and livestock remain largely separated into the two line departments headed by two national directors. Inputs (Fertilizers). The government directly supplies most fertilizers in the country. The prices on these fertilizers are subsidized by between 30 and 35 percent, according to industry estimates. The authorities directly source and distribute the fertilizers. Such heavy government interventions have undermined the ability of the private sector to develop private distribution networks and sell inputs on flexible terms, for example on the basis of informal credit arrangements or in exchange for crop. A gradual and transparent process of withdrawal would allow the private sector to develop its capacity to supply the market. The first step that can be taken immediately is for the government to subcontract private firms to import and distribute fertilizers. Decentralization. The Local Government Act (2002) and the Local Government Finance and Audit Act (2004) provide the legislative framework for extensive decentralization of the central government. DOSA has already initiated significant deconcentration of its services. As a result, approximately a third of its staffs are deconcentrated. DOSA’s eventual goal is to have 60 to 70 percent of the staffs in the field. A concrete implementation plan with specific milestones would facilitate this process. One viii bottleneck is the lack of capacity of local governments. Local councils generally lack the capacity to manage increased resources. ANALYTICAL NEXT STEPS Agriculture Sector Analysis. The authorities are currently developing the agriculture sector strategy. This should be based on a comprehensive analysis of the latest sector developments, including identification of key sector constraints. Budget Outturn Data. The lack of recent and reliable data on budget outturn continues to hamper the PER analysis. The priority is to clear the large backlog of public accounts. An Integrated Financial Management Information System (IFMIS) is currently being implemented which should provide timely public accounts. Civil Service Capacity. The recent continued decline in the shares of wages and salaries raises the prospects that public sector capacity could be under-resourced. This concern was corroborated in a preliminary study which indicated that wages and salaries of civil servants compare poorly with the private sector and civil society organizations.1 The next step is to conduct a comprehensive assessment of civil service capacity which could be the basis for a civil service reform program. Poverty Analysis. The field work for the latest household expenditure survey was conducted in 2003. However, the data preparation and basic analysis are still ongoing. Once the data files are completed, they can be used to conduct poverty impact analyses of public expenditures. 1 Glocoms, Inc., The Strategy for Improving Recruitment and Reducing Attrition in the Civil Service in The Gambia, Capacity Building for Economic Management Project, The Gambia, report no. CBEMP-04-C- 003, April 2005. ix The Macroeconomy and Fiscal Policy INTRODUCTION The macroeconomic environment continued on its stable path in recent years. The authorities have gradually tightened fiscal and monetary policies since the economic downturn in 2002, resulting in a resumption of growth, lowering of inflation, and stabilization of the exchange rate. Nevertheless, such prudent policies will have to be maintained for an extended period in order to decisively break the pattern of “stop and go� policy regimes of the past. This would create and sustain expectations for a stable macroeconomic environment which promotes a private sector led growth. The key challenge in 2006 will be to contain public expenditures in the face of major expenditure items such as the African Union summit meeting and the presidential election. The authorities have also targeted in 2006 the establishment of the Gambia Revenue Authority and the implementation of the Integrated Financial Management Information System (IFMIS). Both are major structural reforms which should contribute towards the strengthening of public financial management. In the medium term, a key objective will be to reduce domestic borrowing and the domestic debt by continuing with fiscal consolidation. The key findings of this chapter are: ♦ Recent fiscal and monetary policies have generally been prudent, but sustained fiscal consolidation is required in the long term in order to reverse adverse debt dynamics such that domestic debt is put on a sustainable path. ♦ The current account deficit has widened considerably, largely due to an increase of imports. ♦ Debt service reached a peak of 46.9 percent of recurrent expenditures in 2005, crowding out other expenditures. ♦ The shares of domestic expenditures on social services and economic services declined in 2005 as a result of the high debt service, and in the 2006 budget as a result of expenditures for the African Union (AU) Summit and the presidential election. ♦ PRSP related domestic expenditures declined in recent years, particularly expenditures for basic health services. ♦ Recent increases of the domestic debt to 35.5 percent of GDP at end-2005 were in large part due to monetary operations. 1 ♦ Without additional principal repayments, sustained domestics saving of 1.4 percent is needed to reduce domestic debt to 15 percent of GDP by 2015, the average of non-CFA SSA countries.2 MACROECONOMIC PERFORMANCE A large and unbudgeted fiscal expansion in 2001, low rainfalls in 2002 and a large expansion of the money supply in 2002 and 2003 were some of the major factors which destabilized the economy in the early years of the 2000s. In particular, the economy shrank in 2002. Table 1.1: Key Macroeconomic Indicators, 2001-2005 2001 2002 2003 2004 2005 Real GDP growth (market prices) 5.8 - 3.2 6.9 5.1 5.0 Inflation (period average) 4.5 8.6 17.0 14.2 4.3 Real effective exchange rate (period avg) - 12.2 - 17.6 - 22.5 1.6 4.6 Avg real interest rates (90 day T-bills) 8.6 3.4 13.3 19.5 4.1 Overall fiscal balance (% of GDP) - 13.9 - 4.6 - 4.7 - 5.7 - 8.6 Basic primary balance (% of GDP) - 1.3 2.7 3.6 9.6 8.5 Government NDF (% of GDP) 14.8 2.5 5.3 0.5 3.6 Domestic debt (% of GDP) 38.1 36.6 27.6 32.9 34.6 Broad Money Growth 19.4 35.2 43.4 18.3 9.4 Sources: IMF, DOSFEA In more recent years, the authorities have maintained significantly tighter fiscal and monetary policies, as evidenced by the large primary balances, substantially reduced money supply growth, and significantly increased interest rates before they were lowered towards the second half of 2005. As a result, the inflation rate fell to single digits, the exchange rate stabilized, and economic growth has resumed in the five to six percent range. Figure 1.1: Sectoral Growth (Annual Percentage Change) 40.0 30.0 20.0 10.0 0.0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 -10.0 -20.0 -30.0 -40.0 Agriculture Industry Services Source: IMF 2 “Domestic saving� refers to net principal repayment of the domestic debt. Debt projections are based on assumptions on GDP growth, interest rates and monetary operations. The details of the estimation are outlined in the subsection, “Domestic Debt,� pages 22 to 25. 2 Annual broad money growth averaged 13.4 percent from 1996 to 1999 then rose to 33.2 percent from 2000 to 2003, causing inflationary pressure. Central Bank lending to the government and monetization of Central Bank losses were the principal causes of the broad money growth. In recent years, annual broad money growth slowed to 18.3 percent in 2004 and 9.4 percent in 2005. The deceleration in broad money growth was driven by a sharp reduction in the annual growth of reserve money from 62.7 percent in 2003 to 11.0 percent in 2004. In order to control reserve money growth in the face of an upsurge of foreign capital inflows, the Central Bank issued 1,070.3 million dalasis of treasury bills on a net basis in 2004, equivalent to 8.9 percent of GDP. The large issuance of treasury bills limited monetary growth, but as a consequence interest rates remained high, credit to the private sector was crowded out, and the domestic debt burden of the government significantly increased. The discount rate on the 91 days treasury bill fell from 31 percent at the end of 2003 to 27.4 percent at the end of 2004, but in real terms the rate increased because inflation decreased by almost 10 percentage points during 2004. In 2005, a relatively reduced issuance of treasury bills and further reduction of the inflation rate allowed the discount rate to fall to 8.5 by the end of the year. Figure 1.2: Monetary Developments (Annual Percentage Change) 80 60 40 20 0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 -20 -40 broad money reserve money private credit Source: IMF The current account deficit has widened substantially in recent years. As a percentage of GDP, the current account deficit including official transfers averaged 3.0 percent up to 2003, but then sharply increased to 11.7 percent in 2004 and 14.5 percent in 2005. This sharp increase was largely the result of a significant increase of imports for the domestic market, causing a substantial widening of the trade deficit to 29.3 percent in 2005. The large current account deficit was financed by increased foreign direct investments and official loans. In particular, foreign direct investment averaged 2.9 percent of GDP up to 2003, then increased to 12.4 percent in 2004 and 10.1 percent in 2005. Most of the country’s exports are reexports of goods that had previously been imported from abroad. Exports of domestically produced goods averaged only 17.6 percent of total exports, 16.7 percent of domestic imports and 5.1 percent of GDP. These low ratios indicate the country’s lack of export competitiveness. The export of 3 groundnuts, the country’s main export product, experienced a near collapse in 2005 mainly due to a poorly implemented reform of the groundnut sector marketing arrangement. A newly implemented market licensing system resulted in the award of only one license to a new operator, the Gambian Agricultural Marketing Cooperation (GAMCO), which experienced operational difficulties. Figure 1.3: Current Account Balance (Percentage of GDP) 5.0 0.0 -5.0 -10.0 -15.0 -20.0 -25.0 -30.0 -35.0 1997 1998 1999 2000 2001 2002 2003 2004 2005 Trade balance CA excl off transf ers CA incl of f transfers Source: IMF Although export diversification is a key objective in the PRSP, there has been little indication of progress. Non-traditional exports, defined as exports of domestically produced goods other than groundnuts, has stagnated in the past decade at 2.0 percent of GDP. The planned Diagnostic Trade Integration Study (DTIS) is an opportunity to analyze and identify constraints to economic diversification. Horticulture and agro- processing would be some of the options to consider in a diversification strategy which takes advantage of the country’s existing agricultural base. Opportunities to further develop tourism, reexport and transit trade, and light manufacturing could also be explored, including through greater regional integration. 2005 FISCAL PERFORMANCE The estimated overall fiscal deficit including grants for 2005 was 8.6 percent of GDP, substantially greater than the budget target of 4.7 percent and the actual deficit of 5.7 percent in 2004. The fiscal deficit in 2005 exceeded the budget target by 3.9 percent of GDP mainly due to: (i) shortfall in tax on international trade by 1.2 percent; (ii) overshooting of domestic interest payments by 1.3 percent; (iii) overshooting of externally funded development expenditures by 2.1 percent; and (iv) extrabudgetary expenditures of 1.0 percent. The authorities compensated for the revenue shortfalls and overspending mainly by reducing other charges by 1.4 percent and domestically funded development expenditures by 0.7 percent of GDP. Nevertheless, the authorities still had to increase domestic financing to 3.6 percent of GDP, overshooting the budget target by 1.5 percent. Government domestic borrowing had fallen to 0.5 percent of GDP in the previous year, the lowest since the mid-1990s. 4 The basic primary balance in 2005 was 8.5 percent of GDP, similar to the 8.7 percent targeted in the budget. The primary balance was 9.6 percent of GDP in 2004. The substantial primary surpluses indicate the government’s commitment to fiscal discipline. The large primary surplus in 2004 was mainly the result of a large increase of taxes on international trade. However, taxes on international trade decreased by 2.0 percent of GDP in 2005, at least partially due to the border troubles with Senegal experienced in the second half of 2005. With this decrease, taxes on international trade continued their long term decline from the 1990s. Figure 1.4: Domestic Revenues, 1995-2005 (Percentage of GDP) 14.0 12.0 10.0 8.0 6.0 4.0 2.0 0.0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Incom e tax Goods and services International trade Source: IMF, DOSFEA Taxes on international trade as a percentage of GDP have steadily declined since 1998 before increasing in 2004. In the same period, taxes on income and on goods and services gradually increased since 2001 by approximately 2.0 percent of GDP. Taxes on international trade ranged between 22 to 24 percent of imports of good up to 2001 before steadily declining to 16.7 percent in 2005. By contrast, imports of goods as a percentage of GDP increased sharply since 2001. Hence, the increase of taxes on international trade in 2004 was mainly the result of larger imports as opposed to improved tax collection efforts. Similarly, taxes on international trade declined in 2005 mainly as a result of reduced imports as a percentage of GDP. Figure 1.5: Domestic Revenues, 1995-2005 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Tax on int trade (% of imports) Imports goods (% of GDP) Source: IMF, DOSFEA 5 Interest payments continued their upward trend in 2005, recording its highest level in recent years. Interest payments in 2005 approximately doubled interest payments in 2001, both in terms of percentage of GDP which increased from 4.5 percent to 8.6 percent, and in terms of percentage of total recurrent expenditures which increased from 23.7 percent to 46.9 percent. Domestic interest payments, at almost four-fifths of total interest payments, dominate external interest payments. Figure 1.6: Recurrent Expenditures, 1995-2005 (Percentage of total recurrent expenditures) 60.0 50.0 40.0 30.0 20.0 10.0 0.0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Wages and salaries Other charges Interest payment Source: IMF, DOSFEA The authorities have responded to the increasing share of interest payments by reducing the shares of wages and salaries and operations and maintenance. The share of wages and salaries decreased from 34.6 percent of total recurrent expenditures in 2000 to 22.8 percent in 2005, while the share for other charges decreased from 40.3 percent to 30.3 percent in the same period. Other charges include goods and services, and current transfers to domestic public agencies and international organizations. Extrabudgetary expenditures amounted to 1.0 percent of GDP in 2005 and 5.4 percent of total recurrent expenditures. They consisted of purchases of two airplanes to be used against locusts, ECOWAS arrears and an emergency loan to the Gambia International Airline. The extrabudgetary expenditures approximately equal half of the reduction in domestic spending on other charges and on development expenditures. Essentially, the extrabudgetary expenditures resulted in a larger than necessary reduction in these other expenditures In terms of administrative classification, in 2005 most of the Departments of State and major public agencies had their share of domestically financed expenditures reduced compared to their budget allocations. The main cause was the higher than expected debt service.3 The estimate for debt service in the 2005 budget was 35.9 percent of 3 Note that the analysis of budget outturn in 2005 is based on preliminary figures constructed directly from the cash books. More reliable figures are not available given the large delays in producing the public accounts. Total expenditures based on the preliminary figures are less than the IMF figures by approximately 90 million dalasis. It appears that the difference is mainly due to embassy expenditures. The PER analysis is limited to domestically funded expenditures given the unreliability of available data on externally financed expenditures, such as those from donor funded projects. In The Gambia, domestic funds account for all recurrent expenditures and a minor portion of development expenditures. 6 total domestically financed expenditures, whereas actual debt service was 47.3 percent. Table 1.2: Central Government Operations by Economic Classification (Percent of GDP) 2003 2004 2005 2005 2006 Actual Actual Budget Actual Budget Revenue and grants 18.2 25.5 23.3 21.4 23.1 Domestic revenue 15.7 20.9 21.4 19.8 21.8 Tax revenue 13.8 18.6 18.3 17.2 19.5 Direct tax 4.4 5.0 5.1 5.2 5.1 Domestic tax on goods and services 2.1 2.4 2.5 2.8 4.0 Tax on international trade 7.3 11.2 10.6 9.2 10.4 Nontax revenue 1.9 2.3 3.1 2.6 2.3 Grants 2.5 4.5 1.9 1.7 1.3 Expenditures and net lending 22.9 31.2 28.0 30.1 26.1 Current expenditures 17.0 16.9 18.4 18.3 17.4 Wages and salaries 4.5 4.3 4.2 4.2 4.8 Other charges 5.9 4.9 6.9 5.5 5.9 Interest 6.1 7.2 7.2 8.6 6.7 External 1.6 1.9 1.7 1.8 1.7 Domestic 4.4 5.3 5.5 6.8 5.0 HIPC expenditures 0.5 0.6 0.1 0.0 0.0 Development expenditures and net lending 5.8 14.2 9.6 11.8 8.7 Development expenditures 6.1 14.4 9.8 11.0 8.8 External 4.7 12.6 8.1 10.2 7.6 Loans 3.4 9.3 6.6 8.6 6.3 Grants 1.3 3.3 1.5 1.5 1.3 Domestic (GLF) 0.6 0.7 1.5 0.8 1.2 HIPC expenditures 0.8 1.1 0.2 0.0 0.0 Net lending -0.2 -0.2 -0.2 -0.2 -0.1 Extrabudgetary expenditures 0.0 0.0 0.0 1.0 0.0 Contingency 0.0 0.0 0.0 0.0 0.3 Overall balance -4.7 -5.7 -4.7 -8.6 -3.3 Basic primary balance 3.6 9.6 8.7 8.5 10.0 Errors and omissions -1.2 -0.5 0.0 -0.5 0.0 Financing 5.9 6.2 4.7 9.1 3.3 External (net) 0.6 5.7 2.7 5.5 1.9 Domestic (net) 5.3 0.5 2.1 3.6 1.4 GDP (million of dalasis) 10026.0 12036.0 13174.0 13174.0 14323.8 Source: IMF, DOSFEA After debt service, general administrative Departments of State had the largest share at 27.9 percent of total expenditures. This share excludes the head item “miscellaneous.�4 Social services had a share of 17.5 percent, and economic services had 6.6 percent. 4 Miscellaneous items in 2005 budget mainly consisted of contingency funds, subvention to the Revenue Authority, “settlement of confirmed outstanding debt,� and “loss in exchange.� Spending on most of these items did not become necessary in 2005. For example, the establishment of the Revenue Authority was delayed to 2006. 7 Although all three major administrative categories experienced shortfalls in actual spending compared to their budget allocations, the shortfall for general administration was marginal when miscellaneous items are excluded. In fact, its share increased when debt services are excluded from the total. Hence, it appears that mainly spending on social services and economic services was reduced in response to larger than expected debt services, while spending on general administration as a whole was largely protected. Table 1.3: Domestically Financed Expenditures by Administrative Classification (Percentage of Total) 2004 Budget 2005 Budget 2005 Actual 2006 Budget General Administration 30.8 33.3 28.6 37.0 Office of President 3.4 3.6 3.0 3.9 Finance (DOSFEA) 6.4 7.5 6.5 4.0 Interior 4.3 4.9 4.8 4.4 Local government 0.8 0.9 0.6 0.8 Foreign Affairs 4.0 4.6 6.0 4.8 Defense 2.4 3.2 3.5 2.9 Governance agencies 1.7 2.3 2.0 1.9 Pensions and Gratuities 1.9 1.8 1.6 1.8 Miscellaneous 6.0 4.6 0.7 12.4 Economics Services 8.3 10.1 6.6 8.2 Agriculture 2.9 3.4 2.2 2.8 Water resources 1.0 0.8 0.5 0.3 Fisheries 0.0 0.2 0.2 0.6 Works, Const & Infra 3.2 3.6 2.1 2.8 Trade, Industry 0.5 1.0 0.9 0.9 Tourism 0.3 0.5 0.3 0.3 Communication, IT 0.4 0.5 0.4 0.4 Social Services 20.6 20.7 17.5 18.9 Education 10.3 10.9 10.8 10.7 Health 9.8 9.0 6.0 7.8 Youth and Sports 0.4 0.8 0.8 0.5 Debt service 40.3 35.9 47.3 35.8 Total (millions dalasis) 2,467.1 2,644.9 2,392.3 2,666.9 Source: GOTG budget reports and cash books. The Department of State (DOS) which spent the most resources was Education. Finance, Interior, Foreign Affairs and Health were also among the largest spenders. By contrast, Departments of State in economic services were among the smallest spenders. The authorities might want to reexamine the adequacy of the resources allocated to economic services given their importance in promoting private sector development. Governance related agencies such as the National Audit Office were also among the smaller spenders, although to some extent this reflects the inherent small size of the agencies. In terms of budget execution, the Departments of State which experienced the largest shortfall of actual expenditures with respect to budget allocations were Agriculture and Health. The DOS for Agriculture spent only 57.8 percent of its budget allocation, and the Department of State for Health only 59.8 percent. By contrast, the DOS of 8 Education spent 90.0 percent of its allocation. The shortfalls are particularly disconcerting given that health and agriculture are PRSP priority areas.5 Box 1.1: Comparaison of Budget Outturn Between Education and Health The large shortfall in spending by Health contrasts sharply with the nearly full spending by Education of its budget allocation. The difference mainly arises from spending on salaries and wages and on development expenditures. Education spent more on salaries and wages than it was allocated. By contrast, Health spent 69.2 percent of its allocation for salaries and wages. Given that budgets for salaries and wages are estimated on the basis of the full complement of positions, the shortfall could indicate difficulties experienced by Health in personnel recruitment. 2005 Domestically Financed Expenditures for Education and Health (Millions of dalasis) Education Health Allocation Outturn Allocation Outturn Recurrent 262.2 239.8 218.3 141.1 Salaries and wages 156.9 168.4 39.7 27.4 Other charges 95.3 70.5 178.3 113.6 Capital expenditures 10.0 0.9 0.3 0.1 Development 25.5 19.0 19.9 1.4 Salaries and wages 0.0 0.0 0.4 0.0 Other charges 8.2 0.0 5.0 1.4 Capital expenditures 17.3 19.0 14.5 0.0 The share of salaries and wages in recurrent expenditures also differed significantly between the two Departments of State. Whereas Education spent 70.2 percent of its recurrent expenditures on salaries and wages, Health spent only 19.4 percent. The difference seemed to have been mainly due to the following reasons: (i) subventions to hospitals, which account for 43.3 percent of other charges, include salaries and wages; (ii) medical inputs are inherently costly, such as pharmaceuticals and vaccines which account for 36.3 percent of other charges; and (iii) most of the medical doctors in the country working in the public sector are provided by foreign development partners and therefore do not create a burden on the government budget. With respect to domestically financed development expenditures, Education spent 74.6 percent of its allocation whereas Health spent only 7.0 percent of its allocation. All of the spending on development expenditures was for counterpart funding for donor projects. Although most Departments of State had their spending reduced from their allocations, there were some exceptions. The DOS for Foreign Affairs spent 6.0 percent of total domestically financed expenditures, whereas it was only allocated 4.6 percent. In addition, DOS for Defense spent slightly higher than its allocation. Note that both are in general administration. In complete contrast to the distribution of recurrent expenditures, Departments of State in economic services received the largest share at 59.7 percent of total domestically financed development expenditures, followed by social services at 22.4 percent and general administration at 18.8 percent. The rate of spending with respect to allocations was also highest for economic services, then social services and finally general administration. 5 Finance also experienced a large shortfall in actual expenditures with respect to its allocation, but it appears that this was mainly due to the fact that its allocation in 2005 was increased from the previous year. Finance’s share of actual expenditures in 2005 was approximately the same as its share of budget allocations in 2004. 9 In terms of the Departments of State, there were two striking features about domestically financed development expenditures: (i) the DOS of Works, Construction and Infrastructure (WC&I) spent 45.4 percent, or nearly half, of the total; and (ii) the DOS of Health and the governance agencies spent only 7.0 percent of its allocation, by far lowest spending rate. All of the development expenditures for WC&I were for road construction and maintenance. The next two largest shares were for Education and Trade and Industry, whose shares combined with WC&I represented 82.1 percent of total development expenditures. Hence, domestically financed development expenditures are highly concentrated. With the recent agreement on a large EU project on roads, the authorities would be able to consider reallocating their own funds away from road construction into other areas. However, at the same time it is critical that road maintenance is adequately funded. 2006 BUDGET The 2006 budget sets several ambitious objectives for the authorities. The targeted overall deficit is 3.0 percent of GDP lower than the average budget outturn of the three previous years, and 5.3 percent of GDP lower than budget outturn of the previous year. In addition, the targeted basic primary surplus is 10.0 percent of GDP, 1.5 percent of GDP higher than the previous year. The 2006 budget indicates that the lowering of the overall deficit will be achieved mainly through higher domestic revenue, lower domestic interest payments, and lower externally funded development expenditures compared to the budget outturn of the previous year. The targeted domestic revenue for 2006 is 3.0 percent of GDP higher than the average budget outturn of the three previous years, and 2.0 percent of GDP higher than the previous year. The increase in domestic revenue is expected to come entirely from higher domestic tax on goods and services and on international trade. In fact, it is expected that taxes on domestic goods and services will increase by more than 50 percent. Total expenditures and net lending are expected to decrease from 30.1 percent of GDP in the budget outturn of 2005 to 26.1 percent of GDP in the 2006 budget. This large decrease is expected to come almost entirely from the decrease in the domestic debt service from 6.8 percent to 5.0 percent of GDP, and from the decrease in externally funded development expenditures from 10.2 to 7.6 percent of GDP. The expected large reduction in domestic debt service will mainly depend on interest rates on the domestic debt remaining at the lower levels recently achieved. The discount rates on the 90 day treasury bills had steadily decreased from 30.0 percent in December 2004 to 8.5 percent in December 2005. Although the rate has subsequently increased to 13.0 percent in April 2006, the targeted domestic debt services of 5.0 percent of GDP in 2006 should be roughly achievable as long as the average interest rates for the year remain approximately at 15.0 percent, and assuming that nominal GDP grows at approximately 10.0 percent, the authorities limit net domestic financing to 1.4 percent of GDP as targeted in the budget, and monetary operations does not contribute more than 1.5 percent of GDP to the domestic debt. It is difficult to assess the likelihood of achieving the lower level of externally funded development expenditures in 2006 without analyzing disbursement projections of 10 donor projects. However, it should be noted that the authorities have traditionally experienced difficulties in projecting external development inflows, and that they had underestimated the inflows by 2.1 percent of GDP in 2005. Regardless of the reliability of the projections, by definition the external inflows are fully funded and therefore do not create any financing needs. The 2006 budget also indicates that the expected large reduction in debt services will allow the authorities to increase expenditures on salaries and wages, goods and services and current transfers by 1.0 percent of GDP, and still lower recurrent expenditures to 17.4 percent from 18.3 percent of GDP in the previous year. However, the costs of financing the planned African Union (AU) summit and the presidential election are estimated to be 1.7 percent of GDP. This is more than the expected increase in non-debt service expenditures, and therefore implies that in fact recurrent expenditures are reduced in the 2006 budget if costs related to the AU summit and the election are excluded. As a result of the need to finance the AU summit and the presidential election, the share of general administration Departments of State in total domestically financed expenditures increases from 33.3 percent in the 2005 budget to 37.0 percent in the 2006 budget. By contrast, the share of economic services decreases from 10.1 percent to 8.2 percent, and the share of social services from 20.7 percent to 18.9 percent. However, these shares increase by 1.6 and 1.4 percents respectively compared to the shares in the 2005 budget outturn due to the expected large reduction of debt service. Also as a result of the AU summit and presidential election, the share of PRSP expenditures in domestically financed expenditures is expected to decrease from 29.9 percent in the 2005 budget to 26.3 percent in the 2006 budget while the share of non- PRSP expenditures is expected to increase from 34.2 percent to 37.9 percent. The majority of the decrease comes from PRSP programs in basic social services and infrastructure, particularly health related expenditures. With respect to the 2005 budget outturn, the shares of both PRSP and non-PRSP expenditures are expected to increase as a result of the expected decrease in debt service. PRSP EXPENDITURES In contrast to progress achieved in the 2006 budget with respect to debt service, albeit mainly due to reduced domestic interest rates, preliminary evidence indicates that efforts to align public expenditures with PRSP objectives have been less successful. The share of PRSP related poverty expenditures in total domestically financed expenditures decreased from 29.9 percent in the 2005 budget to 26.3 percent in the 2006 budget.6 By contrast, non-PRSP related expenditures increased from 34.2 percent to 37.9 percent while the share of debt services remained approximately unchanged. Among the five broad policy areas identified in the PRSP, basic social services and infrastructure received the largest share of PRSP expenditures, mainly due to 6 The analysis of the poverty expenditures is based on the budget codes which identify PRSP related expenditures. This PER Update exercise organized PRSP expenditures according to the broad policy areas identified in the PRSP document. The previous PER Update exercise substantially revise the PRSP budget codes. These codes should now be further refined and aligned with the PRSP II. 11 education and health related expenditures. By contrast, funding local government and community development, and cross-cutting programs has been neglected. Cross- cutting programs cover gender, environment, nutrition and HIV/AIDS initiatives. Table 1.4: Domestically Finance PRSP Expenditures (Percentage of Total Domestically Financed Expenditures) 2004 2005 2005 2006 Budget Budget Actual Budget Total PRSP Expenditures 28.9 29.9 24.0 26.3 I. Macro stability/governance 4.3 4.9 4.1 4.2 Public sector capacity building 1.2 0.5 0.1 0.5 Governance and civil service reform 3.1 4.4 4.1 3.7 II. Private sector led growth 4.3 4.4 2.9 3.7 Agriculture and natural resources 4.3 4.4 2.9 3.7 Regulatory and control services 0.8 1.1 0.8 1.1 Extension services 1.6 1.8 1.3 1.3 Research 0.4 0.4 0.1 0.2 Water management 0.3 0.4 0.2 0.2 Natural resources management 1.2 0.8 0.5 0.9 III. Basic social services and infrastructure 17.3 17.4 14.3 15.8 Education 8.1 8.7 8.8 8.7 Improving access to basic education 7.5 8.2 8.1 8.3 Improving quality of teaching & learning 0.3 0.2 0.7 0.0 Upgrading teaching & learning materials 0.1 0.1 0.0 0.0 Increasing duration of instruction time 0.1 0.1 0.0 0.1 Increasing access to non-formal education 0.2 0.2 0.1 0.2 Health 6.4 5.8 3.9 4.7 Planning. monitoring and evaluation 1.0 0.7 0.5 0.8 Support services to health delivery 0.6 0.7 0.7 0.7 Increasing access & quality of basic health 4.6 4.2 2.6 3.1 Social welfare programs 0.1 0.1 0.1 0.1 Infrastructure (rural roads) 2.6 2.6 1.3 2.1 ICT research and development 0.3 0.4 0.3 0.4 IV. Local govt/community development 1.2 0.9 0.5 0.8 Social funds for poverty reduction 0.8 0.5 0.3 0.4 Decentralization/local govt capacity building 0.4 0.4 0.2 0.4 V. Cross-cutting programs 0.4 0.3 0.1 0.3 Gender 0.1 0.1 0.1 0.1 Environment 0.0 0.1 0.0 0.1 Nutrition, population, HIV/AIDS 0.2 0.1 0.0 0.1 Monitoring and Evaluation 1.5 1.9 2.0 1.5 Non-PRSP spending 30.7 34.2 28.7 37.9 Debt service 40.3 35.9 47.3 35.8 TOTAL 2,467.1 2,644.9 2,392.3 2,666.9 Source: DOSFEA Expenditures in the health sector experienced the largest decrease in PRSP expenditures, representing nearly a third of the overall decrease in the share of total PRSP expenditures. Among the PRSP sub-programs in health, expenditures related to increasing quality and access to basic health accounted for all of the decrease in health PRSP expenditures. Three other sub-programs which experienced large decreases in their shares were agriculture extension services, governance and civil service reform, and rural roads. In comparing budget outturn to allocations in 2005, larger than estimated debt services crowded out both PRSP and non-PRSP expenditures to similar degrees. Among the major PRSP policy areas, approximately half of the reduction came from basic social services and infrastructure, particularly health and rural roads. Agriculture and natural resources experienced the next largest reduction. 12 DOMESTIC DEBT7 Domestic public debt in The Gambia is one of the most critical policy issues in the country. It reached 35.5 percent of GDP at the end of 2005. By contrast, the average stock of domestic debt in 27 non-CFA SSA countries is 15 percent of GDP, while the median is only 10 percent. The average of 14 HIPC eligible non-CFA SSA countries is only 8 percent. Hence, The Gambia’s domestic public debt as a percentage of GDP is more than double the average for non-CFA SSA countries and four times the average of HIPC eligible non-CFA SSA countries. The central government devoted 36.9 percent of its recurrent expenditures to service the domestic debt in 2005. Figure 1.7: Domestic Debt 1400 45.0 40.0 1200 35.0 1000 percent of GDP 30.0 million dalasis 800 25.0 600 20.0 15.0 400 10.0 200 5.0 0 0.0 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 annual increase of dom debt (left axis) stock of dom debt (right axis) Source: IMF, DOSFEA Although the domestic public debt is currently at a very high level, it was only 12.3 percent of GDP in 1994. Hence, it was below the average for non-CFA SSA countries only a little over a decade ago. As a percentage of GDP, the domestic public debt tripled in a little over a decade. Its growth since 1994 can be traced to three distinct periods. (i) Continuous increase from 1994 to 2001 mostly due to government borrowing. The domestic public debt increased from 12.3 percent to 38.1 percent of GDP. Approximately three-fourths of the increase was due to government borrowing. In 2001, off-budget expenditures financed by Central Bank advances were the main causes of a large domestic debt increase. Despite the large domestic borrowing, the government’s financing requirements still exceeded domestic borrowing in 2001 by 3.3 percent of GDP. It appears that the authorities drew down its deposits with the Central Bank in order to cover the remaining financing gap. Otherwise, domestic public debt would have increased even further. (ii) Decrease in 2002 and 2003. The domestic public debt sharply decreased from 38.1 percent to 27.6 percent of GDP. In 2002, the authorities managed to limit net domestic borrowing to 2.5 percent of GDP, and domestic debt as a percentage of GDP slightly decreased as the growth of nominal GDP outpaced the 7 Partially based on Brownbridge, Martin, “Reducing domestic debt to sustainable levels,� mimeo, DOSFEA, 2005. 13 growth of the debt. In 2003, the authorities significantly increased domestic financing to 5.3 percent of GDP. However, the stock of domestic debt sharply decreased as the authorities withdrew their deposits with the Central Bank, thus limiting domestic borrowing. In both 2002 and 2003, the Central Bank increased their exposure to government debt while commercial banks reduced theirs. (iii) Increase in 2004 and 2005 largely due to monetary operations. Domestic public debt increased from 27.6 percent to 35.5 percent of GDP, nearly the highest in the past decade. Unlike in previous years, monetary operations were the main culprit as they accounted for approximately three-fourths of the increase. In particular, in 2004 government net domestic borrowing was limited to 0.5 percent of GDP, but domestic debt increased by 5.3 percent of GDP as the authorities conducted monetary operations in an attempt to sterilize large capital inflows. Figure 1.8: Fiscal and Monetary Causes of Domestic Debt Increases (Percentage of GDP) 14.0 12.0 10.0 8.0 6.0 4.0 2.0 0.0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 govt net domestic borrow ing monetary operations Source: Author’s calculations based on data from DOSFEA, IMF. The sharp decrease from 2002 to 2003 indicates that it is possible to quickly reduce the debt. The domestic public debt could be reduced through a sustained period of limited domestic borrowing and improved coordination of monetary and fiscal policies such that the cost of domestic debt is fully reflected in monetary operations. If the impact of monetary operations on domestic debt was excluded in 2004, then the stock of domestic debt would have decreased to 23.4 percent from 27.6 percent of GDP in the previous year, whereas it actually increased to 32.9 percent. Simulations of domestic debt relief indicate that the authorities would be able to approximately maintain the current level of debt to GDP ratio by limiting annual domestic borrowing to 1.5 percent of GDP. The government’s annual domestic debt service would be maintained at 5.0 percent of GDP. The projections assume an annual nominal GDP growth rate of 10.0 percent, monetary operations of 1.5 percent of GDP, and debt service rate of 15 percent. If the government’s annual domestic borrowing increased to 3.0 percent of GDP, which is approximately equal to the average in the past decade, then the domestic debt would increase to 42.8 percent and the debt service to 6.4 percent of GDP in 2015. If 14 the annual domestic borrowing was reduced to 0.5 percent of GDP, then the domestic debt decreases to 26.9 percent and debt service to 4.0 percent of GDP in 2015. As long as the macroeconomic assumptions hold in reality, the authorities would have to maintain annual domestic savings of approximately 1.4 of GDP in order to reduce the domestic debt to GDP ratio to 15 percent, the average for non-CFA SSA countries. Figure 1.9: Domestic Debt Projections (Percentage of GDP) 45.0 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 status quo borrow ing (1.5% of GDP) large borrow ing (3.0 % of GDP) small borrow ing (0.5% of GDP) domestic saving (-1.4% of GDP) Source: IMF, DOSFEA. Hence, maintaining fiscal constraint is critical to reducing the domestic debt. The authorities can accelerate the reduction of the domestic debt by using additional resources as they become available. Some of the potential sources are the HIPC and MDRI debt relief and reduction in the interest rates on government securities. CONCLUSION Fiscal and monetary policies have been tightened since the large fiscal deficit in 2001 and the ensuing money growth in 2002 and 2003. However, there is little room for complacency. The country’s domestic debt has been exploding over the past decade, and even now it is on an unsustainable path. The authorities cannot take for granted the recent successes with fiscal consolidation as slippages can easily result in a repeat of the pre-2002 macroeconomic crisis. Prudent policies will have to be sustained in the long term in order to address the structural imbalances in the government budget. Past reliance on domestic financing resulted in debt service accounting for 46.9 percent of recurrent expenditures in 2005. Combined with wages and salaries, non-discretionary spending accounted for 69.7 percent of recurrent expenditures. The large share of non-discretionary spending implies that there is limited fiscal space for reallocating funds to PRSP related expenditures. Reducing debt service is critical for expanding the discretionary fiscal space. The debt stock could be reduced in the long run through a sustained period of limited domestic borrowing and improved coordination of monetary and fiscal policies. At present, the authorities rely on the same instrument, primary issuance of treasury bills, to conduct 15 both monetary and fiscal policies. Development of secondary markets and other instruments would facilitate coordination. In addition, the domestic debt could be more rapidly reduced through principal repayments, possibly by using resources from debt relief or budget support if they become available. Finally, debt service could be quickly reduced even if the debt stock remains large if domestic interest rates fall rapidly. This is another argument for strengthening monetary and fiscal policy coordination. Although discretionary fiscal space is presently limited, it could still be used more effectively through resource reallocations. It will be critical to reverse the recent decline in the shares of PRSP expenditures. This decline occurred even after excluding debt service. It is also consistent with the decline in the shares of spending on economic and social services which are PRSP priority areas. By contrast, spending on general administration has been largely protected. It is useful to note the limitations of this PER Update exercise. Given its scope and limited availability of data, the exercise focused on analyzing the allocation of spending. By contrast, the PER Update did not address whether poverty alleviation objectives were achieved in a cost effective manner. Impact and incidence analyses, including the tracking of expenditures through the various layers of the government, would require a more comprehensive and in-depth approach. Such issues will have to be covered in future exercises. This PER Update exercise also mostly limits its analysis to domestically financed expenditures. This is mainly due to the lack of reliable data on externally funded development projects. It severely undermines any analysis of PRSP expenditures given that PRSP programs are mostly implemented through development projects. For future exercises, it will be critical to create a central database on development expenditures. This would also be a prerequisite for strengthening aid coordination within the government. Finally, the effective use of public expenditures to achieve PRSP policy objectives depends on the human resource capacity of the public sector. The recent continued decline in the shares of wages and salaries raises the prospects that public sector capacity could be under-resourced. This concern was corroborated in a recent preliminary study which indicated that wages and salaries of civil servants compare poorly with the private sector and civil society organizations.8 The next step is to conduct a comprehensive assessment of civil service capacity which could be the basis for a civil service reform program. Such a program would improve alignment of civil service capacity with poverty alleviation objectives, and possibly further expand discretionary fiscal space for resource reallocations. 8 Glocoms, Inc., The Strategy for Improving Recruitment and Reducing Attrition in the Civil Service in The Gambia, Capacity Building for Economic Management Project, The Gambia, report no. CBEMP- 04-C-003, April 2005. 16 The AGRICULTURE SECTOR INTRODUCTION 1.1 The authorities prepared a public expenditure review (PER) of the agriculture and natural resources sectors in 2001. The 2001 PER also presented a draft sector strategy and analyzed sector developments. The 2006 PER Update is a partial update of the sectoral PER. It focuses on the agriculture sector whereas the previous PER also included the natural resources sector. Also, it does not provide an analysis of the sector strategy nor the latest sector developments. 1.2 The main findings of this chapter are: ♦ The share of expenditures for extension services decreased from 73.8 percent in 2001 to 30.3 percent in the 2006 budget, while the share for general administration increased from 16.2 percent to 57.7 percent. ♦ The share of inputs, including fertilizers, reached 31.1 percent in the 2006 budget, indicating government’s significant involvement in the supply of inputs. ♦ Government has spent up to 134 million dalasis to support the groundnut sector during the past four years, equivalent to approximately 70 percent of the recurrent expenditures of DOSA ♦ Support to the groundnut sector consists of: equity investment; loans, both direct and indirect; producer price support; and subsidized fertilizers. ♦ The development budget decreased in recent years to 0.7 percent of total expenditures, and its execution rates have been as low as 10.7 percent. ♦ The 2006 development budget emphasizes: (i) horticulture, long touted for its export potential; and (ii) rice cultivation, a core part of the country’s import substitution strategy. ♦ Agricultural research appears to be under-funded, with personnel costs accounting for an excessive share. ♦ The apparent under-funding of the crop and livestock extension service departments could be partially addressed through their consolidation into a single department based on polyvalent extension services. ♦ The government’s direct supply of subsidized fertilizers has crowded out private sector participation. 17 ♦A third of DOSA’s staffs are deconcentrated, the department for crop extension is the least deconcentrated, and the eventual goal is to have 60 to 70 percent of the staffs in the field. SECTORAL COMPARISON Agriculture, education and health are considered the highest priority sectors in the PRSP. However, the shares of budget allocations for the DOS for Agriculture have been substantially smaller than the other two priority sectors. In recent years, DOSA received on average 2.7 percent of the total budget allocations for recurrent expenditures, compared to 11.0 percent for Education and 9.4 percent for Health. In absolute amounts, Education and Health receive between three to four times more recurrent budget allocations than Agriculture.9 Table 2.1: Recurrent Budget Allocations of PRSP Priority Sectors (Millions of Dalasis and Percentage of Total Recurrent Expenditures) 2003 2004 2005 2006 Avg Mil D % Mil D % Mil D % Mil D % Agriculture 53.0 3.2 47.2 2.1 68.8 2.8 64.8 2.6 2.7 Education 208.3 12.7 224.3 10.0 262.2 10.8 266.7 10.7 11.0 Health 180.9 11.0 221.9 9.9 218.3 9.0 188.2 7.5 9.4 Source: The Gambia government budget reports. However, agricultural expenditures as a share of total government expenditures have been reasonably close to the SSA average, although there is considerable variation among these countries. The Gambia compares more unfavorably among its neighboring countries in West Africa where the average share is greater than the SSA average. This gap became even larger in recent years when DOSA’s average share fell further to 2.7 percent in recent years. Moreover, the average share based on a sample of all developing countries is even greater at 7.5 percent, although the data are outdated.10 Table 2.2: Cross-Country Comparaison of Public Expenditures on Agriculture National Government ODA Countries % of Total Exp % of AgGDP % of Total Ag Exp The Gambia 3.3 3.1 32.0 SSA average 3.5 3.8 28.0 West Africa average 5.5 4.5 34.6 Burkina Faso 14.6 10.6 11.0 Mali 10.7 8.1 33.0 Senegal 3.8 5.3 42.0 Ghana 0.5 0.4 64.0 Source: IFPRI (2005) cited in Regina Burner, “Agricultural Sector Public Expenditure Review,� mimeo for World Bank PER Clinic, IFPRI, 2006. Note: West Africa average based on Burkina Faso, Cote D’Ivoire, The Gambia, Ghana, Guinea, Guinea-Bissau, Mali and Senegal. 9 The previous PER on agriculture included fisheries, forestry and natural resources in the analysis. However, resource allocations for the DOS of Fisheries and Water Resources and the DOS of Forestry and Environment are relatively minor and therefore do not change the major findings of this analysis. 10 Blarcom, Bonni van, Odin Knudsen and John Nash, “The reform of public expenditures for agriculture,� World Bank Discussion Papers, no. 216, 1993. 18 ECONOMIC CLASSIFICATION Personnel costs are the largest component of DOSA’s recurrent expenditures. However, its share in actual expenditures sharply declined from 62.2 percent in 2001 to 36.4 percent in 2005, and it further decreases to 30.0 percent in the 2006 budget. By contrast, the share of subventions to local and international organizations increased from 6.3 percent in 2001 to 28.8 percent in 2005, making it the second largest component in DOSA recurrent expenditures. Subventions in the main budget are represented as a single line item, and further details on the expenditures are presented in a separate section of the budget. Incorporation of the more detailed expenditures would be needed for a more comprehensive analysis of expenditures by economic classification. Table 2.3: DOSA Recurrent Expenditures by Economic Classifications (Percentage of Total DOSA Recurrent Expenditures) 2001 2002 2003 2004 2005 2006 Actual Actual Actual Actual Actual Budget Personnel 62.2 56.7 37.1 40.0 36.4 30.0 Travel Expenses 2.7 1.8 3.5 3.1 3.8 3.0 Telecom & Office Expenses 2.2 2.2 3.4 3.5 4.1 2.7 Inputs (fertilizers) 19.8 0.0 7.9 0.0 0.0 0.0 O&M (Fuel & Lubricants) 3.4 2.5 3.8 4.3 5.8 3.3 Equipment Purchase/Maintenance 1.8 1.7 1.9 1.5 1.8 1.5 Buildings/Facilities Maintenance 0.6 1.0 6.0 8.6 7.2 5.4 Contributions & Subventions 6.3 19.4 18.5 30.1 28.8 19.2 Others 1.1 14.7 17.8 9.0 12.1 31.3 Total (million dalasis) 18.4 28.7 46.1 43.0 55.6 66.0 The share of the component “others� in actual expenditures increased sharply from 1.1 percent in 2001 to 12.1 percent in 2005, thus becoming the third largest component. In fact, it is the largest component in the 2006 budget. The component “others� includes animal feed, vaccines, chemicals, pesticides, fertilizers, seed stores and the handling of inputs. The rapid increase in the component “others� indicates that DOSA has become heavily involved in the supply of inputs. This trend is even more pronounced if the one time supply of inputs in 2001 and 2003 is added to the “others� component. Such heavy involvement of the authorities in the supply of inputs could have undermined private sector participation, and reduced potential spending on items such as research and extension services which are more traditionally considered public goods and services. Personnel costs, subventions and inputs accounted for 77.3 percent of actual recurrent expenditures in 2005, and 80.5 percent of the 2006 budget. Such large shares tends to crowd out expenditures for recurring operating expenses such as travel, maintenance and operation of vehicles, office expenses and maintenance of equipment, buildings and facilities. As a result, it appears that vehicles and facilities are typically in poor conditions, and key extension and research outreach programs are not adequately provided. ADMINISTRATIVE CLASSIFICATION Recurrent expenditures of DOSA by line departments indicate that the share of expenditures for general administration increased, while the shares for departments 19 responsible for extension services decreased in recent years. The share of actual expenditures for the Department of Administration increased from 16.2 percent in 2001 to 33.2 percent in 2005, and the 2006 budget indicates a further increase to 57.7 percent. By contrast, the share of actual expenditures for the Department of Agricultural Services (DAS) decreased from 56.9 percent in 2001 to 38.4 percent in 2005, and the 2006 budget indicates a further decrease to 19.8 percent. Also, the share of actual expenditures for the Department of Livestock Services (DLS) deceased from 16.9 percent in 2001 to 12.3 percent in 2005, and the share decreases further to 10.5 percent in the 2006 budget. DAS and DLS are the main departments responsible for extension services. Although traditionally the share of expenditures for DAS was by far the largest and at times more than half the expenditures, it appears that the share for Administration is now the largest as a result of recent trends. Table 2.4: ANR Recurrent Expenditures by Department (Percentage of Total DOS Recurrent Expenditures) 2001 2002 2003 2004 2005 2006 Actual Actual Actual Actual Budget Budget Agriculture Office SOS 1.3 0.7 0.1 0.0 2.7 2.1 Administration 16.2 29.9 23.4 36.1 33.2 57.7 DOP 5.7 7.4 6.2 6.4 5.5 4.5 DOCD 3.0 2.8 4.1 5.5 5.0 3.7 DLS 16.9 19.2 14.0 14.4 12.3 10.5 DAS 56.9 39.9 52.3 37.6 38.4 19.8 DAC 0.0 0.0 0.0 0.0 2.9 1.6 Total (Mil Dalasis) 18.4 25.8 53.0 47.2 68.8 64.8 Source: DOSFEA, DOSA Although the share of expenditures for the Department of Administration clearly increased in recent years, it should be noted that large parts of the expenditures for Administration are items which are not necessarily associated with general administration activities. In the 2006 budget, subventions to local and international organizations, including NARI, the Chamen Agricultural Training Center, the Kuntaur Rice Mill, and the International Trypanotolerance Center, accounts for 12.7 million dalasis or 33.9 percent of the recurrent budget allocation for Administration. The 2006 budget for Administration also includes 13.2 million dalasis for settlement of a fertilizer loan from BADEA, an item which would not normally be considered a general administration activity. If this item is excluded, the share of recurrent expenditures in the 2006 budget for Administration decreases to 37.4 percent, still the highest share among the departments. GROUNDNUT SECTOR 1.3 Groundnut is the country’s most important agricultural good, accounting for between 60 to 70 percent of domestic exports in typical years. Groundnut farmers are the largest group in the agricultural sector, and they have the highest incidence of poverty. An estimated 57,000 farm households engage in groundnut production, or over 80 percent of the agricultural households. The main objectives of government support to the sector are: (i) guaranteed price for farmers that provides an adequate livelihood; (ii) farmers paid cash on sale; (iii) 20 producers and operators are remunerated equitably; and (iv) maximized value addition. The authorities have sought to achieve these objectives through market reforms, producer price support, subsidized fertilizers, direct loans and government guarantees of loans, and direct investments in distribution and processing facilities. Government support to the sector has resulted in significant public expenditures. In terms of direct government support, preliminary figures indicate that the government has spent up to 134 million dalasis to support the sector during the past four years, equivalent to approximately 70 percent of the recurrent expenditures of DOSA, or approximately 2,350 dalasis per groundnut farming household. Direct support included Central Bank loans, equity investments in operators, and supply of subsidized fertilizers. The authorities also provided indirect support to the sector through guarantees of loans to private operators. Table 2.5: Government Support to the Groundnut Sector (Millions of Dalasis) 2003 2004 2005 2006 Direct Support 56.7 68.3 16.7 40.0 Percentage of Rec Expenditures 107.0 144.7 24.3 61.7 Premier Agro (pre-financing) 30.0 - - - GAMCO (initial capitalization) - 63.9 - - Price support - - - 40.0 Fertilizer subsidy 0.1 0.8 1.7 - Loans to GGC 26.6 3.6 15.0 - Govt Guarantees of Loans - - 150.0 77.0 GAMCO loans - - 150.0 - GAMCO pre-financing - - - 37.0 GGC pre-financing - - - 40.0 Notes: * Estimated costs based on a Central Bank of Gambia loan for working capital to the now defunct private exporter / operator. 1 Based on discussions with GAMCO, DOSA and private operators on an assumption that 40,000 MT will be exported (according to government sources) in the 2005/06 season. The facility is expected to compensate groundnut exporters for potential losses incurred. However, the plan has yet to be ratified - the Department of State for Finance and Economic Affairs (DOSFEA) raising objections to the scheme and favouring tax concessions based on export receipts. 2 Based on calculations provided by Ministry of Agriculture using the following fertilizer importation figures: 2003 - 400 MT of fertilizer imported at 20% subsidy; 2004 - 2,500 MT of fertilizer imported at 15% subsidy; and 2005 - 7,004 MT of fertilizer imported at 10% subsidy. There is a likelihood that these figures under estimate the level of subsidy - as at the end of 2005 the price differential between imported fertilizer (CIF Banjul) and government fertilizer purchase costs varied between 30 and 35% depending on the fertilizer type. 3 2003 and 2004 figures based on draft 2004 audit report, 2005 figures based on GGC and DOSFEA figures for a capital investment loan at 13% interest aimed at strengthening GGC logistics infrastructure. 4 Based on estimates provided by GAMCO and DOSFEA. The support is linked to a Government guarantee of a Trust Bank loan 5 Based on estimates provided by DOSFEA - Government guarantee of Trust Bank loan. 6 Based on estimates provided by DOSFEA - Government guarantee of Trust Bank loan In 2003, government support to the groundnut sector mainly consisted of fertilizer subsidies and coverage of defaults on Premier Agro’s government guaranteed loans. These two items totaled 56.7 million dalasis, or 107.0 percent of DOSA’s total actual recurrent expenditures This is equivalent to 528 dalasis per groundnut farming household. Direct support to the groundnut sector essentially equaled the entire government spending for DOSA. In 2004, government support to the groundnut sector increased even further as the authorities provided 32 percent of the initial capital for GAMCO through a number of 21 public enterprises. When this cash transfer is added to the fertilizer subsidies, government’s direct support to the groundnut sector was equal to over 144.7 percent of DOSA’s actual recurrent expenditures, or 1,135 dalasis per groundnut farming household. In 2006, potential “mitigation costs,� which are support to off-set potential losses that may accrue to private sector operators in the current marketing season, has the potential to amount to 61.7 percent of DOSA’s recurrent expenditures, or 702 dalasis per groundnut farming household. From 2003 to 2005, a series of loans at below market interest rates were provided to GGC to support its operations and capital investments. GGC under government management has rarely made profits over the past five years. As long as GGC is owned by the government, it is likely that it will create a fiscal burden and a contingent liability. The authorities have stated that they plan privatizing GGC. Any proceeds from the privatization could be used to off-set some of the existing potential liabilities. The government has also supported the sector through government guarantees of loans, particularly for pre-financing of crop purchases. In 2006, large government guarantees were provided to GGC and GAMCO for pre-financing crop purchases. Government guarantees are attractive instruments because the fiscal cost is hidden unless loans are defaulted. However, they can create large liabilities when loans are actually defaulted, as was the case with Premier Agro in 2003. Given high domestic producer prices faced by operators and significant price discounts that Gambian groundnuts obtain on the international markets, there are large risks that government guarantees will be actualized. The guaranteeing of risks also creates moral hazards. In the case of Premier Agro, the government guaranteed a significant part of the underlying assets and all risks. Such a wide coverage encourages risky behavior by the private investors as they are not liable for the potential losses. It is critical that government guarantees are properly identified and assessed. Proper assessment requires valuing the expected costs of government guarantees rather than simply noting their maximum possible exposure. By calculating the expected cost of guarantees, the government can more easily compare guarantees with cash subsidies. When guarantees are properly valued, more informed decisions can be made on the basis of real costs and benefits. DEVELOPMENT EXPENDITURES It is generally more difficult to make meaningful and robust assessments of the development expenditures because the budget outturn data are less reliable than for recurrent expenditures. Also, development expenditures tend to change erratically from one year to the next due to their reliance on donor projects whose disbursements tend to be lumpy. Despite these caveats, there are some basic observations that have important implications for the development of the agricultural sector. Development expenditures in agriculture have been on a downward slide in recent years. As a percentage of total expenditures, preliminary budget outturn data indicate that development expenditures in agriculture declined from 2.2 percent in 2003 to 0.7 percent in 2005, the lowest since at least 2001. If we assume that the execution rate of 22 the development budget in 2006 equals the average of the previous three years, then development expenditures are expected to remain at 0.6 percent of total expenditures in 2006. A similar downward trend is apparent for development expenditures in agriculture as a percentage of total development expenditures. As a result of this decline, in 2005 development expenditures became significantly smaller than recurrent expenditures in agriculture. From 2001 to 2004, development expenditures accounted for on average 59.3 percent of total expenditures of agriculture, but declined to 35.1 percent in 2005. It would remain at 34.3 percent in 2006 if we again assume a budget execution rate equal to the average of the previous three years. At least in terms of aggregate spending, it appears that the level of recurrent spending could support a significantly larger development budget. Figure 2.1: DOSA Actual Expenditures (Percentage of Total Government Expenditures) 2.5 2.0 1.5 1.0 0.5 0.0 2001 2002 2003 2004 2005 2006 Recurrent Development Source: The Gambia government budget reports. Note: 2006 figures are projections. One reason the agriculture development budget is relatively small is the low budget execution rate. In the previous five years, actual development expenditures averaged only 43.2 percent of budget allocations. In addition, the execution rate for the development budget fluctuated considerably, for instance declining to 10.6 percent in 2002 then sharply increasing to 54.4 percent in 2003. Low budget execution rates could be due to deficiencies in budget preparation, project management and aid coordination. Further analysis would be needed to identify the extent to which each of these factors impact budget outturn. Figure 2.2: DOSA Budget Execution Rates (Ratio of Budget Allocation to Outturn) 23 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 2001 2002 2003 2004 2005 Recurrent Budget Development Budget Source: The Gambia government budget reports. The development budget is dominated by a small number of projects. In the 2006 budget, the three largest projects account for 65.0 percent of the total, and the six largest account for 82.9 percent of the total. Table 2.6: 2006 Agriculture Development Budget by Projects Projects Percentage of Total Peri-Urban project 39.1 Participatory Integrated Watershed Management Project 15.4 Quality Seed Production-NARI 10.5 Institutional support to DOSA 8.3 Integrated Rice Development Program 4.8 Rural Finance Project 4.8 Other projects 17.1 Source: GOTG, 2006 Budget Report. Description of the major donor projects are as follows: ♦ The Peri-Urban Smallholder Improvement Project (PSIP) is a five-year 145 million dalasis project jointly funded by the authorities and the African Development Bank (ADB). It is expected to close at the end of 2006 with a possibility of a one year extension. The project focuses on women and youths in the Western and North Bank Divisions who are involved in horticulture and small ruminant production. It invests in infrastructure and equipments, including wholesale markets, slaughterhouses, compost units, wells and pumps. It also provides training to female entrepreneurs on technical, financial and group management techniques. ♦ The Participatory Integrated Watershed Management (PIWAMP) is co- financed by the government, ADB and International Fund for Agricultural Development (IFAD). It became effective in 2006. Its main objective is to empower communities to increase their productivity of crops, livestock and forest resources at the grass-root level through the transfer of control of land and water management from the government to the local communities. ♦ The Integrated Rice Development Project (IRRIDEP) is co-financed by the government and the Kuwaiti Fund. The main objective is to increase rice productivity in the North Bank of Central River Division (CRD). The project is currently supporting the design and supervision of construction works of the 24 production schemes. The topographic survey of the target schemes and conceptual design of the pump-irrigated perimeters have been completed. ♦ The Rural Finance and Community Initiative Project (RFCIP) is funded by the government and IFAD. It is expected to close at the end of 2006 after six years of implementation. The project’s main objective has been the strengthening of microfinance institutions, particularly those associated with the Village Savings and Credit Association (VISACA). It established a Credit Line Revolving Fund of 14.3 million dalasis for the VISACAs in order to support crop and livestock production and small-scale enterprises. It also rehabilitated and expanded the number of VISACAs. The existing budget classification does not separately identify expenditures by major programs. In order to analyze development expenditures in terms of major policy areas, development project subcomponents were classified according to the following major programs: (i) extension and commercial support services; (ii) water management; (iii) research and development; (iv) regulatory and control services; and (v) planning and administration. Figure 2.3: Agriculture Development Budget by Programs (Percentage of Total Development Budget) 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 1999 2000 2001 2002 Extension serv. Water mangt R&D, regulatory/control serv. Source: GOTG, Agriculture and Natural Resources PER, 2001. In the previous sectoral PER, the allocation in the development budget among the various programs was analyzed over the recent years.11 It indicated that there was a heavy emphasis on water management projects. Their shares of the total development budget averaged 59.6 percent, although the shares substantially declined over the years. Projects for extension and commercial support services had the next largest share at 34.2 percent. At 93.8 percent of the development budget, these two largest programs accounted for almost all development expenditures. The remaining programs comprising research and development, regulatory and control services, and planning and administration accounted for the remaining 6.3 percent, indicating a relatively low priority placed on these programs. 11 Note that the analysis should be considered preliminary given the low and variable development budget execution rate. Ideally, the analysis should be based on actual expenditures, but recent data are not available. 25 The composition of the 2006 development budget differs from the past. There is less emphasis on water management projects, and a more even distribution of resources among programs. More than half, or 56.5 percent, of the development budget is allocated to extension services. Within extension services, 71.0 percent is allocated to livestock and horticulture related projects, and 22.7 percent to rice related projects. Hence, the composition reveals a relatively heavy emphasis on rice related projects. After extension services, water management has the second largest share at 23.7 percent. This is mostly due to a single project, the Participatory Integrated Watershed Management Project (PIWAMP). The National Agricultural Research Institute (NARI) is the main agriculture research institute of the country. NARI’s share of the development budget increased from an average of 2.1 percent in the past to 11.4 percent in the 2006 budget, resulting mainly from resources allocated to NARI’s Quality Seed Production program. The Rural Finance Project provided all of the resources to the microfinance program. Table 2.7: 2006 Agriculture Development Budget by Programs (Percentage of Total Development Budget) Programs and Projects (Subheads) Percentage Regulatory and Control Services 4.0 47-16 Pan-African Control Of Epizootics 0.0 47-63 Strengthening Project Coordinating Office (of Rural Finance Project) 2.7 47-86 Cooperative Development & Extension – Dept. of Cooperation 1.3 Extension Services 56.5 Horticulture and Livestock 40.1 47-32 Livestock Development Project 0.7 47-33 Peri-Urban Horticulture 30.0 47-35 Peri-Urban Livestock 0.1 47-57 Peri-Urban Small Holder Improvement Project 3.1 47-58 Peri-Urban Small Horticulture & Livestock Development Project 5.9 47-69 Rural Finance Project: Livestock Technical Support 0.4 Rice 12.8 47-03 Farmer Managed Rice Irrigation Project 4.3 47-04 Integrated Rice Development Program 4.8 47-06 Kuntaur Rice Mill 1.3 47-10 Irrigated Rice Development Project 2.4 Crops and Others 3.6 47-05 Lowlands Agricultural Development Project (LADEP) 0.3 47-07 Africa Emergency Locust Project 2.1 47-38 Special Program for Food Security (SPFS) 1.0 47-67 Rural Finance Project: Crops Technical Support, DAS 0.2 Research 11.4 47-72 Quality Seed Production (NARI) 10.5 47-95 Pest Management Program (NARI) 0.9 Water Management 23.7 47-01 Institutional Support to DOSA (irrigation) 8.3 47-02 Participatory Integrated Watershed Management Project 15.4 Microfinance 4.2 47-65 Rural Finance Project: Development of VISACAS 1.7 47-66 Rural Finance Project: Rural Finance Unit, Central Bank 0.1 47-68 Rural Finance Project: Kafo Capacity Building 0.5 47-70 Rural Finance Project: Project Support Unit, (RFCIP) 1.8 Total (Thousands of dalasis) 55940.7 Source: GOTG, 2006 Budget Report 26 The composition of the development budget indicates the dual nature of the sector’s strategy. The heavy emphasis on horticulture supports a sector which has long been touted for its export potential. By contrast, the emphasis on rice cultivation and livestock supports policies to strengthen food security through import substitution. In particular, it is not clear whether the country has a comparative advantage in rice production given the low cost of rice available from much larger international producers. RESEARCH Agricultural research plays a crucial role in underdeveloped economies where the majority of the population, particularly the poor, is engaged in agriculture. In such countries, the returns to public investment in agricultural research could far outstrip other public investments.12 In The Gambia, public research on agriculture is mainly carried out by the National Agricultural Research Institute (NARI), a semi- autonomous agency that operates under the purview of DOSA. It has two main stations at Brikama and Sapu to cover the western and eastern parts of the country respectively, and two substations, Giroba Kunda in the URD and Bakindik in the western end of NBD. Table 2.8: 2005 NARI Research Programs (Thousand Dalasis) Programs Activities Budget Funding Source • Agroforestry Natural resource mgnt 270.0 International org. (SMCRSP) • Cropping systems Soil fertility 67.5 Regional organization (ROCARIZ/WARDA) • Pest management Rice pests and diseases 229.5 Regional organization (ROCARIZ/WARDA) • Cereals Cereal and rice variety Unknown Donor projects • Grain legumes, oil Groundnuts and cowpeas Unknown Donor projects seed • Socioeconomic Participatory adoptive and 405.0 Regional organization program diffusion studies (ROCARIZ/WARDA) • Horticulture Farmer training, roots and tubers 448.8 Rural Finance Project, multiplication and dissemination Peri-Urban Project • Livestock Endemic disease control 70.0 Rural Finance Project (vaccination trials) • Agricultural Post harvest and farm implements 30.0 Rural Finance Project engineering • Seed technology Seed multiplication Unknown Donor Projects Total 1,520.8 NARI’s research strategy was developed in 1997 with support from FAO.13 The strategy’s Long Term Plan (LTP) proposes allocating resources among major research areas according to the following shares: (i) food crops at 20 percent; (ii) horticulture and perennial crops at 12.5 percent; (iii) livestock at 12.6 percent; (iv) fisheries at 7.4 12 Fan, Zhang, and Rao (2004) demonstrated that, in Uganda, a million Shilling spent on agricultural R&D reduced the number of the poor by 58.4.12 By comparison, the equivalent number was 33.8 for feeder roads, 12.8 for education and 4.6 for health. 13 FAO, “The national agricultural research system of The Gambia: Analysis and strategy for the long term,� FAO/TCP project TCP/GAM/6611, November 1997. 27 percent; (v) forestry and environment at 20 percent; and (vi) production system at 27.5 percent. Currently, the institute’s research programs cover cereals and grains, horticulture, livestock, and agroforestry. Data on actual expenditures on the research programs in 2005 are incomplete. However, the available data indicate that currently there is a greater emphasis on horticulture research than in the LTP. By contrast, livestock research received far less resources than envisioned in the LTP, and currently NARI does not have any ongoing research on fisheries. In addition, the LTP emphasizes research on business channels and marketing, modalities of funding and land tenure, areas which NARI currently does not have ongoing research programs. The exact level of funding for basic research on food crops is unknown. NARI, through its Grain Legumes and Oil Seed (GLOS) program, provides support to the groundnut sector through the introduction of new technologies and seed varieties. The institute has been able to develop a number of improved groundnut seed varieties. Contract growers are normally used for large scale seed multiplication, for which the 2006 budget allocates 10.5 percent of the total development budget. In the past, it appears that only limited quantities of the improved seeds reached the farmers. There could have been multiple causes, including inadequate control over the seed multiplication process, and limited linkages with extension services. Also, some seed nut growers responsible for bulking up the seeds neglected to undertake adequate drying of the seeds, resulting in increased aflatoxin levels and decline in vigor. NARI receives its funding from a number of sources. The government subvention is used mostly for salaries and wages, and some for utilities. The operations and maintenance of research programs is financed from regional and international research programs such as WARDA, and from donor funded projects. The latter has increasingly become a dominant source of funding for research. LADEP provided nearly half of the NARI research operating budget in 2000 and 2001, and RFCIP added substantial amounts in 2002. Table 2.9: Funding Shares for Public Agricultural Research in SSA Countries (Percentages) AgGDP Total Govt Exp Govt Ag Exp The Gambia 0.1 0.1 4.2 Ethiopia - <1 11 Kenya 2.1 - - Tanzania 0.5 - 30 Burkina Faso 0.3 - - Source: Author’s own calculations based on figures by DOSA, DOSFEA and Herz, K. O. (1993). Although agricultural research and extension have been given a priority in government policy documents, the analysis of public expenditures indicates that resources have failed to match this prioritization, and the institute is under-funded in terms of its current mandate. In 2005, NARI received a total of 3.8 million dalasis, of which government subvention accounted for 61.4 percent, donor project for 26.0 percent, and regional and international organizations for 12.6 percent. Total funding for NARI equaled 0.1 percent of Agricultural GDP and 0.1 percent of total actual government expenditures. NARI’s subvention equaled 4.2 percent of DOSA’s actual recurrent expenditures. These funding shares compare poorly with other SSA countries. The subvention of 2.3 million dalasis was only 27.1 percent of its budget 28 allocation, not enough to cover even the full complement of salaries and wages. As a result, NARI has to rely on external funding for its basic operating costs as well as its research activities. One possible reason for the difficulty in funding the operating costs is that personnel costs are excessive. Salaries, wages and allowances account for 72.1 percent of its recurrent budget allocations. We can approximately assume that in 2005 actual recurrent expenditures equaled all expenditures for salaries and wages, then the share of personnel costs would be equal to the share of actual subventions, 61.4 percent. This is high compared to the 36 percent for Ethiopia, 40 percent for Tanzania, and 30 percent for Cote D’Ivoire, although low compared to the 76 percent in Kenya.14 NARI has a total staff of 285, of which approximately 90 to 100 are professional researchers. The current level is a large increase from the staffing of 23 researchers in 1997, an annual cumulative increase of approximately 17 percent. In effect, the large increase in staffing has crowded out spending on operations and maintenance. NARI indicated that its research objective is to target farmers in regions with the highest poverty, namely Kuntaur, Kerewan, Kanifing and Basse. Client demand should drive the research agenda. The working arrangement between NARI and stakeholders should be institutionalized so as to make research more transparent and responsive to the needs of the poor farmers. NARI has implemented several arrangements in an effort to strengthen linkages with the other key stakeholders: ♦ Rapid participatory appraisal surveys of the farmers are conducted through which key constraints are identified and prioritized. These provide the basis for determining and improving the research programs. ♦ In the past, linkages between NARI and the extension services were formalized through liaison officers, but this system was replaced by a participatory approach to research. NARI now works directly with extension officers and farmers in formulating, conducting and disseminating the results of their research programs. A more detailed assessment would be needed in order to determine whether the current approaches are producing the desired results. It is critical that the surveys are properly designed such that findings provide enough guidance to the research programs. They could also be used to assess the impact of research and extension services. The working arrangement between NARI and the extension services should allow for optimal prioritization of the limited resources. This implies that NARI, DAS, and DLS should have clearly defined and non-overlapping mandates. This analysis focused on public agricultural research. However, the private sector has the potential to contribute to agricultural research in the country, although at present this appears to be limited. Public agricultural research needs to be closely linked with the private sector in order to maximize its effectiveness. Private firms often conduct their own research which complements public research, such as research on coffee and tea in Kenya. Given the size of the sector, the groundnut sector in The Gambia is 14 Herz, K. O., “Funding agricultural research in selected countries of Sub-Saharan Africa,� in Report of the expert consultation on funding of agricultural research in Sub-Saharan Africa, Kenya Agricultural Research Institute, July 1993. 29 most likely to have private and public enterprises which have the capacity to contribute to the country’s agricultural research. At present, little research is directly conducted by the enterprises. EXTENSION SERVICES This analysis indicates that the level of staffing is adequate but the level of funding is inadequate. There are approximately 220 crop extension workers assigned to the Department of Agricultural Services (DAS) under DOSA. In addition, there are approximately 100 livestock extension workers, and a small number of horticultural extension workers. Given that there are approximately 74,000 farming households, the ratio of farming households to total extension workers is 231. This ratio increases to 336 for only crop extension workers and 740 for only livestock extension workers, although the latter ratio assumes that all farming household own livestocks. The authorities consider a ratio between 500 and 800 to be adequate. Hence, it appears that there are adequate numbers of extension workers, particularly crop extension workers. By contrast, budget allocations do not meet the minimum required to maintain adequate levels of services. Excluding salaries and wages, the adequate level of recurrent operating expenditures in 2006 for DAS was estimated to be 2.8 million dalasis.15 This estimate assumes an adequate annual level of funding for each extension worker which covers expenses for travel, transportation, materials, training and supervision. The annual unit cost is then essentially multiplied by 220 crop extension workers in order to derive the estimate for the minimum required total funding for operating costs. By contrast to the minimum requirement of 2.8 million dalasis, the 2006 budget allocation for DAS for travel expenses and fuel was 0.5 million dalasis. There are no separate budget allocations specifically for training and supervision. The budget allocation for travel expenses and fuel only approximately corresponds to the estimate for adequate operating budget given that the former do not include expenses for materials. However, comparison of the two figures does seem to indicate that recurrent expenditures for crop extensions are significantly under-budgeted. In the 2006 budget, the allocation was only 17.9 percent of the required minimum. In the case of livestock extension, the gap between the required and available budgets appears to be even greater. Livestock extension workers are attached to the Department of Livestock Services (DLS) under DOSA. The operating budget requirement in 2006 is estimated to be 3.0 million dalasis, while budget allocations for travel expenses and fuel were 0.6 million dalasis in 2004, 0.8 million dalasis in 2005 and 0.6 million dalasis in 2006.16 Therefore, allocations in 2006 were only 20.0 percent of the required minimum. As a result of this apparent under-funding, DOSA estimates that approximately 60 percent of crop extension workers are not mobile and therefore unable to adequately 15 See annex I for details of estimation of operating costs for DAS extension services. 16 See annex II for details of estimation of operating costs for DLS extension services. 30 carry out their services. In particular, this can have a significant impact on the quality of extension services to groundnut farmers. With adequate funding, extension workers would be able to provide intensive support to groundnut farmers on land preparation, seeding rate, time of planting, fertilizer application, weeding, harvesting and storage. Due to the lack of funding, Village Extension Workers (VEWs) limit their visits to farmers situated close to one of the 27 mixed farming centers where they are stationed. VEWs rarely travel outside of a 5 mile radius from one of these centers, and this lack of mobility results in VEWs covering only 20 percent of the 5,000 farmers that should ideally be served.17 This under-funding also undermines structured linkages with research institutions and donor projects which support extension type activities. Research institutions are unable to translate their research into extension messages. Donor projects do not benefit from the full involvement of field extension staff, resulting in weak linkages between project investments and annual programs of extension services. Given the number of extension workers, the above analysis indicates that the recurrent operating budget for extension services should be increased. This will result in an increase in the allocations for DAS and DLS, already the second and third largest departments in DOSA in terms of their share of allocations. Another option is to reduce the number of extension workers by consolidating the various extension services into a single polyvalent extension service. The analysis of the number of staffs seems to indicate that there is some margin for reducing the total number of extension workers. The consolidation of the various extension services could also result in improved coverage for livestock extension if all polyvalent extension workers were able to provide this service. The consolidation of DAS and DLS into a single polyvalent extension service has been an objective of DOSA for many years, but progress has been slow. Institutional rigidities remain which continue to constrain the attainment of effective polyvalence of extension services at the farm level. Extension services for crops and livestock are largely separated into the two line departments headed by two national directors. This duplication is costly to DOSA which must support VEWs for both crops and livestock, including logistics and facilities. The current plan is to unify the two departments into a single new agency to be named the National Agricultural Development Agency (NADA). NADA would be responsible for both crop and livestock extension delivery to farmers, and emphasize extension programs which target the poor. Residual interventions in relation to input supply and credit would be transferred to the private and NGO sector, and resources would be focused on providing relevant knowledge to the poorest farmers. INPUT SUPPLIES (FERTILIZERS) The Gambia has a history of providing subsidized fertilizer to rural producers through the Ministry of Agriculture’s Agricultural Inputs Office (AIO). At present, almost all fertilizers are supplied by the government. Fertilizers were applied to 13 percent of 17 Loum, Badara, “Enhancing extension services delivery in The Gambia,� FAO, Activity No. 163, March 2001. 31 total cultivated areas in 2003/4 and 16 percent in 2004/5. It appears that government imported fertilizers met only a part of the demand, but there was little additional supply response from the private sector due to the fact that government supplied these inputs at below market rates. It is estimated that about 35 to 40 percent of fertilizers reach groundnut farmers. With little private supply, the country’s supply of fertilizers is constrained by the ability of the government to mobilize resources. Although donor efforts in the past have attempted to develop a viable private sector fertilizer distribution network, these have failed due to the considerable risk and uncertainty that these private suppliers face in terms of government selling “donated� fertilizers at below market prices.18 This intervention has distorted prices and led to an unreliable, high cost marketing and distribution system with a limited selection of basic fertilizers. The direct supply of fertilizers by the government has been increasing in recent years. By contrast to the volume, the official estimates of the level of subsidy decreased from 20 percent in 2003 to10 percent in 2004. However, fertilizer industry sources estimate that, at the end of 2005, the level of subsidy between imported private fertilizer costs (CIF Banjul) and government fertilizer purchase costs were 30 percent for Urea and 35 percent for 15:15:15 NPK. Irrespective of the absolute level of subsidy, what is clear is that the price of fertilizer is below border parity levels. In addition, the percentage of farmers who benefited from the fertilizers remains low. Table 2.10: Government Supplied Fertilizers Year Source of Funding Tonnage Subsidy Total Farmers Benefited (MT) (Percentage) Dalasis (Percentage) 2003 AGIDB 400 20 - 10 2004 Trust Bank 2,500 15 12,135,000 12 2005 Islamic Development 7,004 10 65,800,000 15 Bank Source: Department of Planning, DOSA Instead of integrating the donated fertilizers into the domestic fertilizer market, the authorities directly manage the sourcing and distribution of the fertilizers. This has discouraged private sector investment in fertilizer importation and marketing, and perpetuates the country’s dependence on fertilizer aid. In other countries, governments have recognized that such a practice crowds out the private sector. It increases the risk that the private sector will not be able to sell its own imports at market determined prices. In countries where fertilizer subsidies continue to be provided, at a minimum the governments subsidize the suppliers of fertilizer and allow them to handle the importation and distribution operations. Such an institutional arrangement tends to be more market enabling. The present heavy government intervention in the fertilizer market has hampered the ability of the private sector to: (i) develop private distribution networks; (ii) sell inputs on flexible terms, for example on the basis of informal credit arrangements or in exchange for crop; (iii) establish a customer base based on high quality service; and (iv) provide a full package of inputs, including fertilizers, seeds, and chemicals, backed by extension knowledge. 18 Donations are predominantly from Japanese KR2 fertilizer grants and, to a lesser extent, Italian grant fertilizer. 32 Despite the government taking responsibility for the delivery of fertilizer, it has been noted that the fertilizers frequently arrive late in the country and therefore timely distribution is a problem. The primary source of delay is the complicated diplomatic and bureaucratic process that is involved in donor funding. Fertilizers supplied late in the growing season reduce their impact. Furthermore, fertilizers frequently arrive in the middle of the hunger season when few farmers have the necessary cash. Although the public sector dominates the fertilizer trade, two private sector fertilizer traders, GHE and Sangol Farm Ltd, have been able to operate within the sector. However, the quantity they import is extremely small, each importing less than 100 MT of fertilizer per year. Given the current structure of the fertilizer market, these private enterprises do not compete with imported government fertilizer but primarily stock fertilizers of different varieties that are predominantly used for horticultural purposes. Both companies indicated that they are interested in becoming more actively involved in the fertilizer business. One trader apparently ships significant quantities of fertilizers through Banjul to neighboring countries, thus indicating it has at least some capacity to increase imports. However, private firms are reluctant to commit until there are clear signals from the government that it will withdraw from the fertilizer supply business. There appears to be two arguments that justify government’s direct intervention in the fertilizer market. First, the authorities argue that the private sector was not able to increase supply in the past when the government withdrew from the market. Second, the lack of access to rural credit means that farmers are not able to finance the purchase of major inputs, particularly fertilizers. Private sector representatives counter that they have the capacity to adequately supply the market if they were allowed enough time to develop appropriate linkages for mobilizing resources and accessing overseas markets, and if administrative steps and bureaucratic overhead were reduced. Withdrawing direct supply of subsidized fertilizers by the government would be consistent with international practices. Many countries in West Africa no longer subsidize fertilizers. A roadmap would be required which outlines a gradual and transparent process of withdrawal which allows the private sector to develop its capacity to supply the market. The first step that could be taken immediately is for the government to subcontract private firms to import and distribute fertilizers, as opposed to carrying out these activities themselves. The roadmap should be prepared jointly among representatives of the government, private sector input suppliers, financial institutions and farmers associations. It should be comprehensive, addressing issues such as access to rural credit which could affect the demand for fertilizers, and the business climate which could affect the supply. The withdrawal of government supply would be consistent with the country’s main medium term development strategy, the PRGS (PRSP II), which emphasizes the promotion of private sector in agricultural input marketing. The long term objective is to liberalize markets such that private firms would be the dominant supplier of fertilizers, seeds and other inputs, and prices would be determined by the market. This should be feasible given that the quantities of fertilizer imported into the country are modest by international standards. As the government gradually withdraws from the market, it will be critical that it does not replace its current direct interventions with indirect forms of support, 33 particularly government guarantees of loans to operators. In the past, withdrawal of government supply in an attempt to promote private participation in the markets initially resulted in a low private supply response. Citing high financing costs, private firms lobbied the government and succeeded in receiving government guarantees of their loan. Subsequently, the guarantees were called in when the private firms defaulted on their loans. DECENTRALIZATION The Local Government Act (2002) and the Local Government Finance and Audit Act (2004) provide the legislative framework for extensive decentralization, including decentralization of agricultural services. The authorities have so far opted to gradually build the capacity of relevant local institutions before fully implementing decentralization of public services. For decentralization to be fully operational, the Local Government Service Scheme has to be in place which would allow the transfer of staff from central agencies to local government agencies. Concrete steps would have to be taken in terms of public sector institutional restructuring in order for Departments of State to transfer to local institutions major responsibilities and resource for planning, implementation, and monitoring and evaluation. Table 2.11: DOSA Staff Allocations Departments Total Total Regional Divisional District Village Percentage HQ Local Local Office of SOS 2 0 0 0 0 0 0 Dept of Admin 94 0 0 0 0 0 0 DAS 341 128 6 26 29 67 27.3 DLS 113 88 0 6 82 0 43.8 DOP 38 49 3 6 12 28 56.3 DOCD 24 13 4 2 7 0 35.1 Total 612 278 13 40 130 95 31.2 Tech Depts 478 229 10 34 118 67 32.4 Source: DOSA, The Gambia. DOSA initiated significant deconcentration of its services even before the enactment of the decentralization legislations. As a result, all technical departments under DOSA have at least two to three levels of representation in the field, at the divisional, district and village levels. The levels and functions by department are as follows: ♦ DAS at divisional level is represented by DAC, who is assisted by SMS to oversee, supervise, assist and coordinate the implementation of extension activities conducted by subordinate agencies DES at the district level and VEW at the village level. ♦ DLS has a similar structure with DLO on top followed by Livestock Assistant I and II which are at district and village levels respectively. ♦ DOP has three levels, Regional or Divisional Supervisors, District Supervisors and Enumerators at the village level. The multiple layers are designed to ensure that the crop data are appropriately collected. 34 ♦ DOCD has three levels, consisting of Assistant Registrars at divisional level assisted by Cooperative Assistant with Inspectors at district and Secretaries at Seccos (CPMS) level. At present, DOSA has 31.2 percent of its staff outside of headquarters at the regional, divisional, district and village levels. The percentage is 32.4 percent for the technical departments encompassing DAS, DLS and DOCD. Of the three line departments, DAS has the lowest share of staff placed outside of Banjul headquarter. However, it does have the largest number of staffs located at the village level. DOSA’s ultimate objective is to place 60 to 75 percent of the staffs at the field level, along with greater responsibilities and resources. However, at present there is no implementation plan with specific milestones for implementing deconcentration or decentralization. One bottleneck is the lack of capacity of local governments. The EC and other donors have been providing projects to building local capacity that would enable local councils to take full charge of staffs within their respective LGAs. An education and sensitization campaign is needed for government officials and the general public to be aware of the role and functions of local governments. This would help ensure transparency and accountability in the dispensation of duties by the local councils. CONCLUSION Given the limited discretionary fiscal space in the overall budget, it is unlikely that public expenditures for the agriculture sector will significantly increase in the immediate future. However, public spending on agriculture could increase if the country reaches the HIPC Completion Point and thereby benefit from substantial debt relief. Regardless of whether the discretionary fiscal space is eventually expanded through debt relief, the public expenditure analysis indicates that the effectiveness of spending in agriculture could be improved if structural imbalances could be addressed. The main structural imbalances identified in the expenditure analysis are: (i) large shares of expenditures on inputs, particularly subsidized fertilizers; (ii) significant spending to support the groundnut sector, including through direct investments and loans; and (iii) under-funded research and extension services. One option to consider is to increase spending on research and extension services by reducing resources currently allocated to fertilizers and the groundnut sector. The substantial spending on subsidized fertilizers and the groundnut sector sharply contrasts with the relative under-funding of research and extension services. In addition, it contrasts with the government’s PRSP which advocates the promotion of private sector participation and the strengthening of research and extension services. It also does not appear that the spending on subsidized fertilizers and direct investments and loans in the groundnut sector have resulted in noticeably improved sector performance or poverty levels. By contrast, at present the sector is characterized by: (i) low usage and limited private supply of fertilizers; (ii) increased dependence on government investments and loans in the groundnuts sector; and (iii) 35 continued high poverty levels among groundnut farmers, exacerbated by a sharp decrease in exports in 2005. Government’s withdrawal from direct supply of subsidized fertilizers would have to be conducted in manner that ensures that the private sector would be able to increase its supply. The exact modalities would have to be identified through further analyses. Similarly, any significant reduction of direct government interventions in the groundnut sector should be conducted through a comprehensive sector reform program. Withdrawing direct government support could lead to market instability without a program which addresses potential problems with the producer price support system, export marketing schemes, and quality control of groundnuts. Also, simply increasing funding for research and extension services would not necessarily result in enhanced output and outcomes if resources within the relevant institutions are not optimally allocated, and if performance objectives and indicators are not properly identified and monitored. Development expenditures for agriculture have been on a downward slide in recent years. This decline has been due to both lower budget allocations as well as lower budget execution rates. Further analysis would be required to identify the exact causes. They could be due to weaknesses in budget preparation, foreign aid coordination and project management. The 2006 development budget emphasizes horticulture, livestock and rice cultivation. The composition of the budget is consistent with a sector strategy based on both import substitution and export promotion. The emphasis on livestock and rice cultivation reflects policies to strengthen food security through import substitution. The heavy emphasis on horticulture supports a sector which has long been touted for its export potential. The agriculture sector strategy currently being developed should further clarify subsector priorities. There is a risk that a strategy which attempts to emphasize both import substitution and export promotion would be poorly implemented given limited resources and capacity. This PER Update does not analyze whether public expenditures in the agriculture sector were effective in achieving sector policy objectives, particularly poverty reduction objectives. This will have to be covered in future analyses. Fiduciary and governance issues could be relevant in such analyses given the substantial spending on agricultural inputs and general administration. Larger spending on inputs implies greater usage of public procurement. An open and transparent public procurement process would facilitate effective use of resources. Increased spending on general administration could result in less resources actually reaching the communities in the absence of strong governance arrangements. A public expenditure tracking survey could be used to assess the situation. 36 Annex 1: Estimate of Operating Costs for DAS Extension Services Activity Item No. Cost/unit Total Annual Cost Per Extension Worker 1. Visits to contact farmers and groups Fuel lts. 30 15.1 454.0 DSA 10 75.7 756.7 2. Kaffo farm visits Fuel lts. 10 15.1 151.3 3. DEC farms Urea bags 4 44.4 177.6 NPK bags 4 60.5 242.2 Seed G/nut bags 2 40.4 80.7 Seed Millet kg 2 0.4 0.8 Seed Maize kg 7 0.2 1.4 4. In-service training Allowance 1 423.8 423.8 Transport 1 181.6 181.6 Food 1 264.9 264.9 Materials 1 75.7 75.7 5. On-farm trials/demonstrations Fuel lts. 10 15.1 151.3 6. Micro plots 10 15.1 151.3 7. Village based farmer training -3 sites DSA 30 75.7 2,270.2 Fuel lts. 10 15.1 151.3 Food/person 90 22.7 2,043.1 8. Striga control campaign - 3 sites Fuel lts. 10 15.1 151.3 DSA 3 75.7 227.0 9. Meetings Fuel lts. 15 15.1 227.0 DSA 4 75.7 302.7 10. Field Days - 2 days Fuel lts. 20 15.1 302.7 DSA 3 75.7 227.0 Lunch 3 37.8 113.5 Materials 1 605.4 605.4 Total Costs Per Extension Worker 9,734.7 Total Annual Cost Unit cost * 220 extension workers 2,141,626.9 Training of trainers 90,806.6 Supervision (25% of op. cost) 535,961.7 Total Annual Cost 2,768,395.2 37 Annex 2: Estimate of Operating Costs for DLS Extension Services Activity Item No. Cost/unit Total Annual Cost Per Extension Worker Fuel lts. 600 15.1 9,080.7 Maintenance 1 9,080.7 9,080.7 DSA 120 45.4 5,448.4 Total Cost Per Extension Worker 23,609.7 Total Annual Cost Unit cost * 100 extension workers 2,360,971.2 Training at Demo and Training Center trainees 140 1,135.1 158,911.5 On-farm training Fuel 1 45,403.3 45,403.3 DSA 1 31,782.3 31,782.3 Materials 1 45,403.3 45,403.3 Supervision (total) 326,903.7 Supervision (per division) Fuel lts 1800 15.1 27,242.0 DSA 120 151.3 18,161.3 Driver 120 75.7 9,080.7 Total Annual Cost 2,969,375.4 38 39