Report No. 22543-ZA Zambia Public Expenditure Review Public Expenditure, Growth and Poverty: A Synthesis December 2001 Macroeconomics I Southern Africa Africa Region Document of the World Bank FISCAL YEAR OF BUDGET January I - December 31 CURRENCY EOUIVALENTS Currency unit: Zambian Kwacha (K) US$ 1 = 3425 Kwacha (Exchange rate Effective June 1, 2001) WEIGHT AND MEASURES Metric system ABBREVIATIONS AND ACRONYMS AAC Anglo-American Corporation ABB activity based budgeting AIDS acquired immunodeficiency syndrome ASIP Agriculture Sector Investment Program BESSIP Basic Education Subsector Investment Program BOP balance of payments BOZ Bank of Zambia CDF cumulative distribution function CSO Census Statistical Office CU Commercialized Utilities DFID Department for International Development (U.K.) ESW Economic and Sector Work ESAC Economic and Social Adjustment Credit (World Bank) FMU Financial Management Unit FSC Fiscal Sustainability Credit GDP Gross Domestic Product GRZ Government of the Republic of Zambia HIPC Heavily Indebted Poor Countries HIV Human immunodeficiency virus IDA International Development Association IFMIS integrated financial management information system IMF International Monetary Fund LCMS Living Conditions Monitoring Survey MAFF Ministry of Agriculture, Food and Fisheries MOE Ministry of Education MOFED Ministry of Finance and Economic Development MOH Ministry of Health MTEF Medium-Term Expenditure Framework NGO non-governmental organization NRB National Roads Board OAG Office of the Auditor General PAM Program Against Malnutrition (NGO) PE personal emoluments PER public expenditure review PETS public expenditure tracking surveys PPP purchasing power parity PREM Poverty Reduction & Economic Management Network (World Bank) PRGF Poverty Reduction and Growth Facility PRSP Poverty Reduction Strategy Paper PSCAP Public Service Capacity Building Project PTA Parents and Teachers Association QAP Quality Assurance Process (Africa Region, the World Bank) QER Quality Enhancement Review (ESW process, the World Bank) RDC recurrent department charges REF Rural Electrification Fund RIF Rural Investment Fund ROADSIP Road Sector Investment Program SIF Social Investment Fund SIP Sector Investment Program SSA Sub-Saharan Africa STD sexually transmitted disease SWAp Sector Wide Approach USAID United States Agency for International Development WSS water and sanitation services ZCCM Zambia Consolidated Copper Mines Limited ZNCB Zambia National Commercial Bank ZNOC Zambia National Oil Company Limited ZESCO Zambia Electricity Supply Corporation ZPA Zambia Privatization Agency ZRA Zambia Revenue Authority Vice President: Callisto Madavo Country Director: Yaw Ansu Sector Manager: Philippe Le Houerou Task Manager: Delfin S. Go TABLE OF CONTENTS ACKNOWLEDGEMENTS ................... ............. viii LIST OF PARTICIPANTS. ix EXECUTIVE SIUMMARY .x i INTRODUCTION .xxiv PART I: CROSS-CUTTING ISSUES OF PUBLIC EXPENDITURE. 1. Macroeconomic Constraints To Public Expenditure . 2. Poverty, Inequality And Public Expenditure .. 21 Appendix 2.1: Measures of Poverty and Inequality .34 3. Composition Of Public Expenditure .3 Appendix 3.1: Regression Analysis .50 4. Management Constraints To Public Expenditure . 5 PART 11: SECTOR SPECIFIC ISSUES OF PUBLIC EXPENDITURE: 64 5. Intersectoral Comparisions: Introduction And Criteria 64 6. Education. 68 7. Health .80 8. Water And Sanitation .88 9. Agriculture. 94 10. Infrastructure: ..104 REFERENCES ........... ............ I b ANNEX I: STATISTICAL TABLES .120 ANNEX II: SCEANARIOS FROM THE 123PRSP MODEL .127 ANNEX III: GOVERNMENT INTERIM SYSTEM FOR TRACKING HIPC RELATED EXPENDITURES .136 LIST OF TABLES Table 1.1: Zambia: Debt Service Paid and External Flows 1990-2000 ............................. 3 Table 1.2: Real GDP and Sectoral Value-Added, 1995-2000 ............................................. 4 Table 1.3: Revenue and Expenditure Trends, 1995-2000 ................................................... 5 Table 1.4: The Potential Effect on Economic Growth of a Decrease in Inflation ......... ..... 9 Table 1.5: Medium-term Macroeconomic Framework, Zambia ......................................... 13 Table 2.1: National Incidence of Poverty by Rural/Urban Classification, Stratum and Province ................................................................. 21 Table 2.2: Percentage of Households that Received and Sent Remittance ........... ............. 25 Table 2.3: Incidence of extreme Poverty by Rural/Urban Classification ............ ............... 33 Table 2.4: Incidence of Moderate Poverty by Rural/Urban Classification ........... .............. 33 Table 3.1: Zambia: Share of Social Expenditures, 1996-99 ....................... ........................ 39 Table 3.2: Percent Ratios of GRZ Actual Budget Releases Over Initial Allocation, 1995-99 .................................................................. 45 Table 4.1: Action Plan for Public Expenditure Management Reform ............. ................... 62 Table 6.1: School Enrollment of Children Aged 7-13 in 1998, by Quintile ............. ......... 69 Table 6.2: Government Expenditure on Education ............................................................. 75 Table 8.1: Percent of Population with Water and Sanitation Services ................................ 88 Table 9.1: Agriculture, Public Budget Allocation and Actual Expenditures .......... ........... 98 Table 9.2: Historical and Projected intra-sectoral Shares of Expenditures ........... .............. 99 Table 9.3: Agriculture - Expenditure Distribution ............................................................. 99 Table 10.1 Rural Electrification Releases to Ministry of Energy and Water Development (MEWD) against Remittances from ZESCO to MF .......................... 110 Table 10.2. Fuel Levy Collections and Arrears to the Road Fund, 1994-2001 .......... ......... 110 Table 10.3: Structure of Funding to the Roads Department, 1995-2000 ............................. 111 Table 10.4: Public Spending on Transport by Type, 1995-99 ............................................. 112 Table Al.1: Central Government Budget, 1990-2000 ......................................................... 120 Table A1.2: Central Government Budget, 1990-2000, % of GDP ................ ...................... 121 Table Al.3: Structure of the Central government Budget, 1990-2000 ............................... 122 Table Al.4: Zambia: Allocated and Actual Public Expenditures for Key Sectors, 1990-2000 ...................... ........................................... 123 Table Al.5: Zambia: Selected macroeconomic Indicators, 1990-2000 ............................. 125 Table A2.1: Consensus Forecast: Medium-term Macroeconomic Framework, Zambia ................................................................. 129 Table A2.2: Growth Elasticities in the First Two Years ................................... ................... 131 Table A2.3: Impact of Shocks in Government Expenditures and Copper Price .................. 133 Table A2.4: Growth Coefficients of the Get Real Model ................................ .................... 135 Table A2.5: Impact of HIV/AIDS on GDP Growth ............................................................. 136 Table A2.6: Alternative Long-Term Growth ................................................................. 137 LIST OF FIGURES Figure 1.1: Copper Price and Zambia's Per Capita Income ............................................... 2 Figure 1.2: Per Capita Revenue and Expenditure in 1994 Constant Prices ........................ 5 Figure 1.3: Deficit and Inflation .................. .............................................. 7 Figure 1.4: Zambia's Overall Public Sector Deficit ...........................................................8 Figure 1.5: More State-Owned Enterprise Is More or Less Associated with Less Income 10 Figure 1.6: Mean Income and Poverty Rates in Zambia by Province, 1998 ...................... 12 Figure 1.7: The Relationship Between Expenditure in the Poorest Quintile and Mean Expenditure in Zambia, by province in 1993, 1996, and 1998 ........................ 12 Figure 2.1: Cumulative Distribution Function (CDF) of Household per Adult .......... ....... 22 Figure 2.2: Population by Consumption Quintile and Place of Residence, 1998 ............... 23 Figure 2.3: Education Attainment of 15-19 Year Olds by Wealth Category .......... ........... 24 Figure 2.4: Child Mortality (in percent) by Decile of Wealth, Zambia 1997 ..................... 24 Figure 2.5: Overall Enrollment Rate of Primary and Secondary Schooling by Consumption Quintile by Urban/Rural Residence, Zambia, 1998 .................. 29 Figure 2.6: Percentage of Students, Ages 7-13, by Quintile and Location, Zambia ......... 29 Figure 2.7: Cost per Student by Provincial per Capita Consumption, Zambia 1998 ......... 30 Figure 2.8: Distances to Primary School by Quintile for Rural and Urban Area Zambia, 1998 ............................................................... 30 Figure 2.9: Percentage of Clients of Facilities from Each Consumption Group ......... ....... 31 Figure 2.10: Percentage Use of Each Type of Facility by Consumption Quintile ............... 31 Figure 2.11: Proportion of Visitors (under 5) to Public Health Clinics . . 32 Figure 2.12: Percentage Users of Public Primary Facilities ................................................. 32 Figure 3.1: Composition of Audited Expenditure 1991-2000 .41 Figure 3.2: Share of Committed Donor Projects in Public Expenditure 1991-2000 43 Figure 6.1: Student Flow Profile in Grade 1-7 Education, 1998 .73 Figure 6.2: Profile of Schooling from Grade 1 through Grade 12, 1998 .74 Figure 7.1: Ministry of Health - Real Expenditure in 1994 Billion Kwacha .80 Figure 7.2: Child Mortality (in percent) by Decile of Wealth, Zambia 1997 .83 Figure 7.3: Immunization Rates (in percent) by Income Quintile ................. ..................... 84 Figure 8.1: Percentage of People with Piped Water by "wealth" decile, 1993 & 1997 ............................................................... 91 Figure 8.2: Hypothetical Private and Social Demand for Water ........................... ............ 93 Figure 9.1: Source of all maize sold ........................ ....................................... 95 Figure 10.1: Electricity Consumption by Sector, 1986 & 1999 . .......................................... 1 05 Figure 10.2: Final Energy Consumption ............................................ 105 Figure A2. 1: Schematic View of the 123 PRSP Model ....................................................... 130 Map 1: Poverty by District, Zambia 1988 ............................................................... 140 Map 2: Zambia's Main Infrastructure ............................................................... 141 LIST OF BOXES Box 1.1: The Efficacy of Public Expenditure and Foreign Aid ...................... ...................... 19 Box 3.1: Public Expenditure Supported by HIPC Interim Debt Relief, 2001 ........... ........... 49 Box 4.1: Some Initial Survey Findings Regarding the Delivery of Public Services in Selected Education and Health Facilities ............................................................... 52 Box 4.2: Tracking of HIPC Expenditures - Some Trade-offs ....................... ....................... 56 Box 4.3: Sectoral Programs and Issues of Donor Coordination and Donor-Executed Projects ............................................................... 59 Box 4.4: Donor Coordination Issues - Sectoral Programs and Donor-Executed Projects .... 61 Box 6.1: Zambia's Basic Education Sub-Sector Investment Program .................................. 71 Box 7.1: Zambia's Health Sector Investment Program ......................................................... 87 Box 9.1: Zambia's Agriculture Sector Investment Program ................................................. 103 Box 10.1: Zambia's Road Sector Investment Program ......................................................... 115 ACKNOWLEDGMENTS The Zambia PER is a joint report between the Government of the Republic of Zambia and the World Bank. A short description of the participatory process is provided in the Introduction. The Government PER team (see list below) was led by Mr. Likolo Ndalamei, Director of Budget, Ministry of Finance and Economic Development (MoFED), with broad supervision from the current Permanent Secretary, Budget and Economic Affairs, MoFED, Mr. Boniface Nonde and the former Permanent Secretary, Dr. Herrick C. Mpuku. The Government team produced a draft PER report, Public Expenditure Review: 1995-2000 for their contribution. This synthesis report, which puts together key points raised in the background materials and at the PER Workshop in Kabwe, is drafted mainly by Delfin S. Go (task manager), Jeffrey S. Hammer (lead advisor) and Mushiba Nyamazana (coordination of the PER workgroups). The report and findings were reviewed by the Government PER team and its comments and recommendations were incorporated. Bruce Jones participated in all key meetings and coordinated the social sectors work. Background papers and inputs were also provided by the World Bank Zambia Country Team (see list below). Joint activities with the IMF and IMF documents on budget management, tracking of HIPC expenditures, PRSP and the PRGF were also important inputs. Deon Filmer provided several specific and substantive inputs. Leonid Koryukin helped in the implementation of the quantitative macro framework and supplied key tables. Integrating papers were additionally produced and used: Delfin Go (public expenditure, growth and poverty), Deon Filmer and Jeffrey Hammer (poverty, inequality and public expenditures), and Mushiba Nyamazana (benefit incidence analysis, infrastructure). The reviewers of the report were: Shantayanan Devarajan, William Leslie Dorotinsky and Jeni G. Klugman. Helpful comments and advice were also received from Abebe Adugna, Sudhir Chitale, William Easterly, Alan Gelb, Nicolas M. Gorjestani, Malcolm Holmes, Iraj Talai, Jose Leandro, Bernard Myers, Vinaya Swaroop, and Fahretin Yagci. The report benefited from two internal processes at entry: the Quality Enhancement Review (QER) and a brainstorming session organized by Africa region's Quality Assurance Process (QAP). Additional support was given by QAP to tap Bank network resources and by the PREM Thematic Group on Data and Tools for Economic Analysis to develop the quantitative macroeconomic framework for PRSP used in the macro chapter. Those processes and the agreement with the Government basically shaped and defined the directions and coverage of the present report. Yaw Ansu, Hinh Dinh and Laurence Clarke guided the work and process and provided various specific comments. In addition to giving substantive comments and guidance throughout the process, Philippe le Houerou supplied key management and steering support necessary in a participatory PER and for the expanded scope of the ESW. viii LIST OF GRZ PARTICIPANTS Supervision INSTITUTION Mr. Boniface Nonde Permanent Secretary-Overall Supervisor Ministry of Finance and Economic Development Mr. Likolo Ndalamei Acting Director of Budget- Coordinator and Chairperson Ministry of Finance and Economic Development Mr. James Mulungushi Chief Economist- Deputy Coordinator and Chairperson Ministry of Finance and Economic Development MACRO CHAPTER Mr. P. Mulenga Advisor - Macro Economic Department Ministry of Finance and Economic Development Mr. P. Malambo Acting Dir. - Ext. Resource Mobilization Dept. Ministry of Finance and Economic Development Mr. M. Sinkala Chief Accountant Ministry of Finance and Economic Development Mrs Stella N. Mutale Acting Senior Economist Ministry of Finance and Economic Development Mr. Boss Msimuko Principal Revenue Analyst- Budget Office Ministry of Finance and Economic Development Mr. Chiluwe Principal Project Planner Ministry of Tourism Mr. Steve Mbewe Principal Economist Ministry of Finance and Economic Development POVERTY Mr. L.Kaemba Acting Dir. - Population In Development Dept. Ministry of Finance and Economic Development Mr. John Mwansa Chief Revenue Analyst Ministry of Finance and Economic Development Mr. S.Sichone Assistant Director(Public Investment Unit) Ministry of Finance and Economic Development Ms. Jane Haboongo Chief Planner Ministry of Communications and Transport Mr. John Banda Economist Ministry of Finance and Econ omic Development Mrs Pamela K. Bwalya Economist Ministry of Finance and Economic Development Mr. Conrad Himwiinga Planner Ministry of Community Dev and Social Services Ms Elfridah Chuulu Assistant Director - Central Statistics Office Central Statistics Office BUDGET MANAGEMENT Mr. Martin Phillips Advisor Budget Ministry of Finance and Economic Development Mr. Rao Advisor Budget Ministry of Finance and Economic Development Mrs. Getrude C Ngoma Assistant Director(Public Investment Unit) Ministry of Finance and Economic Development Mr. Clifford Banda Principal Budget Analyst Ministry of Finance and Economic Development Mr. Ledson Zulu Acting Deputy Controller of Audits Ministry of Finance and Economic Development Mr. T. K. Phiri Accountant General Ministry of Finance and Economic Development Mr. P. Chanda Chief Human Resource Ministry of Finance and Economic Development Mr. K. Mpembamoto Senior Revenue Analyst Ministry of Finance and Economic Development Mrs Shirley Zulu Acting Chief Budget Analyst Ministry of Finance and Economic Development EDUCATION Mr. Emmanuel Silanda Director - Education Ministry of Education Mr. Fred Makondo Senior Analyst Ministry of Finance and Economic Development Mr. S.S. Banda Chief Economist Ministry of Finance and Economic Development Mr. James Mbewe Acting Assistant Director Ministry of Local Government and Housing HEALTH Mr. Davies Chimfwembe Acting Director Ministry of Health Mr. Bonah Chita Health Planner Ministry of Health ix Mr. Mike Goma Chief Accountant Ministry of Health Mr. Nicholas Chikwenya Health Planner Ministry of Health AGRICULTURE Mr. Julius Shawa Assistant Director Ministry of Agriculture Mr. Martin Liywali Senior Economist Ministry of Agriculture INFRASTRUCTURE Mr. Rogers Kapila Economist Ministry of Finance and Economic Development Ms Bupe Kaonga Acting Director of Planning Ministry of Works and Supply Mrs L H. L. Chandi Senior Economist Ministry of Energy and Water Development Mr. A.J.K. Muzilika Director - Physical Planning and Housing Ministry of Local Government and Housing x LIST OF CONTRIBUTORS FROM THE WORLD BANK Supervision Mr. Delfin S. Go PER Task Manager, Editor Mr. Jeffrey S. Hammer PER Lead Adviser, Editor Mr. Mushiba Nyamazana Coordinator of PER Workgroups, Editor Mr. Bruce Jones Coordinator of Social Sectors MACRO CHAPTER Mr. Delfin S. Go Paper on public expenditure, growth and poverty Mr. Andrew Dabalen Macroeconomic framework Mr. Leonid Koryukin Macroeconomic framework Mr Rene Bonnel Background paper on HIV/AIDs POVERTY Mr. Jeffrey S. Hammer Paper on poverty, inequality and public expenditures Mr. Deon Filmer Paper on poverty, inequality and public expenditures Mr. Mushiba Nyamazana Background paper on benefit incidence analysis Mr. John Ngwafon Poverty data, description and analysis COMPOSITION OF PE Mr. Delfin S. Go Background paper Mr. Jeffrey S. Hammer Background paper BUDGET MANAGEMENT Mr. Bernard Myers Background paper Mr. Hinh Dinh Paper on cash budget, inputs and recommendations Mr. Harry Garnett Inputs, review and comments EDUCATION Mr. Bruce Jones Background paper and summary Mr. Keiichi Ogawa Background paper and summary HEALTH Ms. Shiyan Chao Background paper Mr. Albertus Voetberg Inputs, review and comments WATER Mr. Fook Chuan Eng Background paper Ms. N. Jane Walker Inputs, review and comments Agriculture Mr. Tekola Dejene Background paper and summary Infrastructure Mr. Stephen Brushett Background paper on roads and transport Mr Musniba Nyamazana Summary on infrastructure Participatory Aspects Mr. Jim Edgerton Adivser xl EXECUTIVE SUMMARY I. At the heart of the growth problem, the persistence of poverty, and issues of policy reform in Zambia is the public sector reform program. The best practice in public sector reform identifies three areas in which governments can improve their perfornance and their impact on the economy and poor: (1) macroeconomic discipline (the stabilization problem); (2) strategic priority setting (the allocation problem); and (3) efficient public-service delivery (the execution problem). Zambia's problem appears to be in all three areas. In particular, the aggregate performance, allocation, and execution of the budget are vital to the success of Zambia's public sector reform program. All three areas, particularly the allocation and execution issues, as they apply to public expenditure are the subject matter of the present PER. GENERAL FINDINGS AND RECOMMENDATIONS 2. Zambia has experienced an unusually long period of economic contraction due to the decline of the resource base (copper), wrong policies in the past, as well as other factors such as severe transportation problems, several adverse terms of trade shocks, high level of external debt and droughts. Zambia's current per capita income, about US$ 300, is now half of what it was at the time of independence in 1964; and poverty incidence is very high, over 70 percent by the national poverty line and near 64 percent based on the international poverty line at $1 (PPP) a day. 3. By the time economic reforns started in the 1990s, much of the past levels of social assistance and expenditure were no longer affordable and real public expenditure had to be reduced severely over the years. However, public expenditures were also badly allocated and executed. The quasi-fiscal deficit, stemming mainly from the significant losses of key parastatals, diverted vital resources from public and pro-poor expenditures. In addition, attempts to address public expenditure piecemeal through the operation of the cash rationing system, Zambia's cash budget, however successful in initially bringing down high inflation, led to very poorly and erratically executed public expenditure and to the failure to develop effective expenditure management institutions. The combination of the budget squeeze, cash rationing, fluctuation of heavy external servicing and the protection of social expenditures by donors and by the Govemrnent also has led to severe cutbacks in infrastructure and other capital spending, hurting the income opportunities of the rural poor in the process. 1 4. While favorable external circumstances (e.g. terms-of-trade shocks, debt relief etc.) and other sources of growth may matter, public expenditure is rightly viewed as one of the vital ingredients to growth and poverty reduction. However, the efficacy of public expenditure can only be improved with good policies, good governance, and good The Government is reversing the trend in capital expenditures in recent years., but audited expenditures of recent years are not yet available to verify their relative magnitudes. xii institutions and a high-level commitment to bring about the necessary reforms. To make public expenditure the lynchpin of Zambia's growth and poverty reduction strategy, three basic measures are recommended: (i) free up or stop the wasteful leakages of resources by reducing the public sector deficit, particularly the losses of parastatals, so that these resources can be utilized more efficiently for growth and poverty reduction; (ii) improve the execution of the budget and the delivery of public services by replacing the cash rationing system with effective fiscal management and budgetary institutions; and (iii) restore rural spending, particularly rural infrastructure. Main Conclusions 5. The Stabilization Problem. The overall public sector deficit (quasi-fiscal deficit) in Zambia - which includes the deficits of the central government, the local governments, extra-budgetary accounts, state-owned enterprises, and the Central Bank - is still very high and continues to be a key threat to macro stability, growth and poverty reduction: * The best estimate available of the quasi-fiscal deficit puts it somewhere from 14 to 18 percent of GDP during 1995-97 and there is a clear need to carefully update the figure. Of this amount, losses of state-owned enterprises are the biggest drain on public revenues (36 percent). From 1998 onwards when copper prices collapsed, heavy operational losses of the largest state-owned enterprise, the mining company Zambia Consolidated Copper Mines Limited (ZCCM), reached US$lm a day - which constituted an additional 10 percent of GDP to the existing annual quasi-fiscal deficit. e The size of the quasi-fiscal deficit easily overshadows any budgetary spending for the poor. The magnitude before recent ZCCM's losses was more than half of total budgetary spending, inclusive of foreign current and capital grants; ZCCM's annualized losses just before privatization was more than a third of total spending. The sale of ZCCM in March 2000 was therefore a significant achievement towards correcting an unsustainable public sector financial situation, but much remains to be done. * Without raising GDP growth to higher than population growth on a sustainable basis, poverty reduction in Zambia would be difficult; the facts indicate that the income growth of the poorest quintile will respond to general economic growth in Zambia by close to one-to-one, which is consistent with the experience of many countries. 6. There is a continuing need to keep public expenditure at non-inflationary level and to allocate and utilize public resources carefully so that the objectives of growth and poverty reduction can begin to be realized. * Zambia's economic outlook has brightened with the sale of ZCCM, recent reforms, improved copper prices and production, HIPC interim debt relief, and good weather. GDP growth may reach near 5 percent for the first time xiii in a long while. However, that expected growth and lower inflation (of less than 20 percent a year) have yet to be realized and maintained consistently for several years. Fiscal prudence is therefore still necessary. * There are also several downside risks and factors to macro stability and a higher economic growth: a. Like in 1996, expenditure pressures during an election year in 2001 would likely lead to substantial increases in such category as "other" or in "other constitutional and statutory expenditures". Historically, the impact for such increases has been poor: Data indicates that small expansionary effect of government spending in the first year would largely be offset by a contractionary effect in the next. b. In the medium term, payments for outstanding obligations from parastatals and other public entities (US $700m for ZCCM) as well as fiscal concessions will continue to constrain resources in the budget and in the economy/banking sector. c. A possible slowdown in world demand and the price of copper (outside of a severe drought) would continue to have the biggest potential threat to growth and income of the poor, but the factors mentioned above would exacerbate the negative effects. * Until the gains of reforms are secured and additional resources are generated, total budget expenditure must be kept where they are now at 30 percent of GDP. With HIPC assistance, there is now a limited but non- trivial marginal budget to improve performance without increasing total expenditure - it should be allocated wisely and executed well. The PER suggests some directions for change, but medium-term allocation of budget resources, including HIPC expenditures should ideally follow the priority setting by government in consultation with civil society in the formulation of the country's Poverty Reduction Strategy Paper (PRSP). 7. The Allocation Problem. Public resources appear to be allocated incorrectly, both across sectors and within sectors, relative to the government's objectives of growth and poverty reduction: * In particular, infrastructure investments of all types have fallen in the 90s due to both the inability to make long-term, relatively large commitments and the shortfall of actual to budgeted resources. While all ministries complain about the shortfall of actual to budgeted resources, the impact has been larger for infrastructure investments than it has for service sectors or non-economic institutions. * Social expenditures, which are protected in the budget and by the lending operations of the World Bank (36 percent of the domestic budget) lack impact and rationale: a. Expenditures on health and education are not benefiting the poor as they should (more below). xiv b. School enrollment rates and the retention of rural teachers are falling. c. Partly because of HIV/Aids, health outcomes are deteriorating. d. Transparency in the procurement of pharmaceuticals and fertilizers remains a serious issue. e. There is a wasteful duplication of functions and staff between parallel health agencies - the Ministry of Health and the Central Board of Health. 8. The beneficiaries of public expenditure have generally not been the poor: * Income of the poorest quintile is closely related with growth and would particularly benefit from rural infrastructure and income opportunities in the rural areas. * Spending in the rural areas and for primary education are, not surprisingly, more oriented towards the poor, if the money does ever get there. Children attending primary school are more likely to be from the poorest groups in Zambia than from better off groups; those who are not yet or able to be in school are also disproportionately poor. Secondary education shares this pattern, though the effect is not as pronounced as in primary education. The beneficiaries of the substantial subsidies to university education are not the poor. * The beneficiaries of health expenditures are generally more complex and less clear. The largest group using public primary health facilities tends to be in the middle of the income distribution, not at either end. In addition, expenditures on hospitals, while more justifiable on efficiency grounds, almost certainly benefit the non-poor more than others. 9. The Execution Problem. The lack of an effective budgetary framework and management have affected government's ability to carry out the broad objectives of growth and poverty reduction. * There is a need to unite policy, planning and budgeting, anchoring it in the strategic prioritization of the PRSP with top-level political support and discussions but with a viable financial framework. * The cash budget, while making an important contribution to the control of inflation, has introduced systematic biases in the allocation and delivery of public resources. Execution issues include: a. The pattern and levels of expenditure outturns deviate substantially from budget allocation partly because of under-budgeting and cash rationing that require substantial in-year reallocations. b. Month to month expenditure volatility has been high. c. The delivery of public services in health and education facilities has been affected seriously by the cash rationing system. xv * Zambia has several sector investment programs, which are designed to coordinate efforts, pool resources and improve the delivery of public services in key sectors such as agriculture, basic education, health, and roads. Implementation is however affected by a. weak government capacity in such areas as fiscal controls, financial and procurement management; and b. lack of transparency and expenditure wastage, which are serious in areas such as the procurement of drugs and pharmaceuticals, agriculture input supplies (e.g. fertilizers) etc. * The true cost of government is not fully identified - (i) much of the actual donor spending is executed outside the budget (because of donors' desire to improve outcomes in the short-term), making the accounting of expenditures incomplete; (ii) the budget is confined to central government; (iii) there is no functional classification and data nor regional expenditures in the in-year budget reports. v In general, the information management system for public expenditure is deficient for supervision and control or for end-year reporting. Main Recommendations 10. PER recommendations to increase the efficient and equitable use of public resources and to improve their delivery can be grouped into three general areas: (i) reducing the public sector deficit, particularly leakages from state-owned enterprises; (ii) improving the execution of the budget and the delivery of public services with better budget institutions; and (iii) restoring rural spending, particularly infrastructure. More specifically, the following measures are recommended: 11. Reduce Spending on Commercially Saleable Goods and Services. * Privatize remaining parastatals. Being engaged in private, commercial activities, parastatals do not have a legitimate claim on public resources. Privatization of these is a high priority for reducing the deficit. In the interim, all parastatals should operate without subsidy. * The focus of structural reforms continue to be state-owned enterprises, including the deregulation and privatization of the petroleum sector and privatizing the controlling interest of the largest domestic bank, the Zambia National Commercial Bank (ZNCB). Pension reform is another area requiring attention. * Charge for publicly provided private goods unless there are clear, distinct and demonstrable benefits from a subsidy on efficiency and equity grounds. While in the longer run, many goods and services that can be sold commercially should be left to the private sector, public provision may be necessary in the interim. However, they should be charged at their marginal cost of provision both to save public resources as well as to xvi maintain incentives for private sector growth. Three important examples are fertilizer, tuition and living stipends for higher education and urban water. 12. Restore Spending on Rural Infrastructure. * Rural areas are very much poorer than urban. Redistributing income to the relatively poor means increasing rural services relative to urban. However, rural areas are difficult to serve with services that require resident staff. Rural investments that require fewer government workers to serve in rural areas should be favored over others. Further, some forms of infrastructure: surface roads and network systems (such as electricity distribution) require public subsidy on efficiency grounds. * Therefore, rural infrastructure, particularly of roads, power and communication services, is the main area of spending (net of fees charged) on which there is a clear justification for increase. 13. Replace the Cash Budgeting or Rationing System with budget procedures that can incorporate longer term planning procedures. These will involve better institutions for making budgetary priorities and choices; and for executing the budget more effectively (see below). * In the short to medium term, replace the protection of social spending with one that protects poverty-oriented expenditures, including infrastructire, improve budget classification and reporting, supplement budget monitoring with PETS, enhance the realism of budget allocation and reduce significant in-year reallocations, and replace the ad hoc cash rationing scheme with well-defined rules. * In the medium to long term, develop an effective financial management system and medium-term expenditure framework (MTEF). SECTOR-SPECIFIC RECOMMENDATIONS Education 14. Zambia can and must find a way to bring higher quality to the rural schools, with supporting measures elsewhere in the education system: * Bring more trained teachers into rural schools; integrating textbooks and other instructional materials into teaching practice; * Experiment with alternative ways of bringing educational services to remote areas, such as: improve the living conditions, lighting, sanitation and water facilities of rural schools where rural teachers are bound to live; significantly raise the allowances to teachers for extended stay in rural areas; assist local communities and voluntary organizations that wish to develop their own schools in rural areas, supporting them with necessary xvii resources for education materials and rural teachers; decentralizing the monitoring, hiring and wage disbursements of rural teachers to the local communities; and considering giving back to church groups, other bodies, and local authorities under enabling guidelines the responsibility they had in the past for the management of schools in selected rural areas etc. * Learn what works from these, and other, experiments. The introduction of any policy measure should be accompanied by a means of careful evaluation in terms of test scores, student attendance and retention. Successful interventions can be replicated and all results should be publicized in order to mobilize parental involvement. 15. Government plan to gradually eliminate university subsidies should be made within a coherent framework with an institutional mechanism for regular open discussions among stakeholders of the issues and strategies for implementation. Possible options include: * Raise tuition fees gradually, starting with new incoming students; reduce aid first to non-instructional expenditures (food, dormitories, transportation); protect poor students by applying means testing to student aid; eventually, fees should cover the marginal cost of education and housing. * Ensure the growth of the economy to raise the private returns of higher education but allow for private provision and competition. * Institute flexible payment arrangements and aid in the form of loans or student employment; consider income-contingent loan systems, sometimes referred to as graduate tax, in which loan repayments are a fixed proportion of a graduate's annual income; and limit loan programs to a fixed share- of students to control fiscal exposure and to promote academic competition and performance. * Replace the traditional approach of "negotiated" budgets with alternative mechanisms such as incremental resources tied to self-generated resources and linking funding to performance measures such as the number of graduates; competitive fund to promote quality improvements and innovations. * Consider hiring private companies to manage the universities under well- defined and enabling arrangement, etc. 16. To learn quickly what works from these and other experiments, the Ministry of Education needs to be more forward looking by improving its planning, management, and monitoring capability. The introduction of any policy measure should be accompanied by a means of careful evaluation in terms of test scores, student attendance and retention. Successful interventions can be replicated and all results should be publicized in order to mobilize parental involvement. xviii Health 17. Given the history of health sector reform over the past decade, it is important to experiiment with policy options and make a concentrated effort at evaluating their outcomes. Monitor health status regularly in order establish the effect of public programs and to replicate successes. * The most pressing threat to health and a very serious development problem is HIV/AIDS, though its claim on public resources has to be carefully evaluated. The government should continue to heavily publicize the problem, its extent, its causes and the means people can take to protect themselves. A wide variety of specific interventions should be tried but their effects should carefully be evaluated so as not to waste potentially large sums of money. * Other infectious diseases are of high priority. Health education activities, particularly for HIV/AIDS but also for basic hygiene, management of water sources and sanitation, malaria protection and nutrition are likely candidates for public intervention and evaluation. - Improvements in the operation of the health care system should be carefully evaluated. Experiments in alternative ways of increasing access of rural people to health care need to be tried and evaluated. Options might include: deternining the responsiveness to increases in pay for rural service, establishing mobile clinics, "campaign"-style immunization programs or improved communication and referral to urban areas. - FFees for primary care at hospitals should be raised to reflect the marginal cost of such provision both to divert use to cheaper, more efficient places for care as well as to conserve the budget for services that require public expenditure. Exemptions for primary care fees should be based on niral location other than for vaccinations. Payments might differ for urban and rural areas or by region, except for maternal and child care services. The administrative feasibility of specific exemptions for the poor should be explored. Water 18 Safe water is a high priority for households, but direct provision of water is primarily an urban phenomenon, well justified on efficiency grounds but less on equity. * Outside of digging wells in rural areas, there are large gains to be had from health education in terrns of reduced morbidity from diarrheal disease, gains that disproportionately accrue to the poor, and from better and more hygienic management of water and sanitation. * Develop a cheaper method of providing water and sanitation to poor communities in urban and rural areas. xix * Reversing the deterioration in the provision of water and sanitati tm services in urban areas require implementing an existing policy, clearing and sorting arrears (among Cus, ZESCO, and Govemment), metering andl collecting fees, and building up the capacity to operate viably. * In particular, aggressively pursue the metering and charging for water onX the basis of water used, not hookups. * Eventually prices should reflect marginal costs of water plus a premium to recoup capital expenses over time. Prices can be raised gradually over 3 to 5 years to determine the responsiveness of demand and the appropriate long run subsidy. 4-yriculture I 9 Growth in the vital agriculture sector can be enhanced: (i) by improving the erformniance of existing individual farming units, adding more value to land and labor ihrough the application of available technologies, developing new varieties, anld ; stroducing better farming and resource conservation practices; (ii) by encouraging the exzansion of cultivated area (less than one-fifth of the estimated 10 million hectares of ultivable land is currently utilized); and (iii) by integrating smallholder farms to the myarkets and transforming their subsistence orientation to more commercial operations. (I Public expenditure for agriculture should broadly focus on: removing the bottlenecks for agriculture developments (e.g. transport costs., extension services etc.) with farmers and traders; * supporting growth without displacing the private sector through various measures - policy (sector specific and macro and trade), extension, research, infrastructure spending, promotion of farming practices and institutions etc. that all aim at increasing agricultural production and diversification and raising the output of marketed commodities and livestock products. Policy reforms should include: e FRA should divest from fertilizer marketing and refocus its mandate to maintenance of strategic food reserves. * Increase the delivery of public agricultural services to resource-poor smallholders by expanding local participation and ownership of agricultural programs. * Improve the rural road network by increased maintenance. The availability of cheap and reliable transportation is also an important factor. * Encourage small scale irrigation schemes. • Maintain good policies in the areas of macro, trade, and real exchange rate, which have great indirect effects on agriculture outputs and exports. xx The availability of cheap and reliable transportation is also important factor. Infrastructure. 22. The following are policy options to improve the effectiveness of infrastructure expenditure: * Focus on development and implementation of least cost alternatives for access provision. This means not only looking at pavement surfacing alternatives to bitumen and to gravel, but also at standards with regard to road width and other engineering specifications which could be modified to reduce cost, i.e. a given level of access goal could be attained for a far lower cost than is currently the case. * A similar argument can be used for the electricity infrastructure. More poor people can be reached through a solar power program and in mini hydro and alternative sources of energy than through the extension of the national grid. * Privatize utilities as soon as possible and minimize direct Government participation to policy and regulation of utilities. In addition, provide a conducive environment for private sector infrastructure development in areas such as hydro power generation. • Develop a balanced infrastructure network. This requires a revised public investment plan which would follow on from the adoption of a coherent transport policy framework which is expected imminently. The transport policy framework - and associated regulatory reforms to be undertaken by the government especially in the rail and air sectors - are expected to encourage greater private sector participation in management and financing. Further reforms to road user charges, in particular the introduction of economic tariffs for heavy vehicles, are expected to contribute to a more optimal balancing of freight traffic using the full transport infrastructure. * Use special funds proceeds for intended purposes. There is no point in establishing special funds for road maintenance and rural electrification and then starve them of the proceeds from such levies. Immediate attention thus needs to be given: in the short term to clearance of all arrears and to provision of more direct and transparent methods of transferring the agreed resources into these funds; in the medium term to setting in place the required legal and institutional reforms to secure the proper, independent management of these funds and the methods by which future resource requirements of these funds can be met. * The road sector institutional framework has still to be rationalized in order to: address overlapping and unclear mandates; give impetus to the commercialization of road management; provide opportunities for decentralization where appropriate. The revised framework should follow xxi on from the adoption of the transport sector policy and would involve consolidation and further restructuring of the agencies responsible for road management to improve efficiency and build capacity. * In the electricity sub-sector, however, ZESCO also needs to be unbundled, i.e., generation, transmission and distribution operations, respectively, should be undertaken by different companies in order to remove costs associated with such a large natural monopoly. However, this would also require a restructured energy sector regulatory framework. SPECIFIC MEASURES FOR BUDGET MANAGEMENT 23. Institutional issues and weaknesses in the overall and sectoral public expenditure management have affected government's ability to carry out the broad objectives of growth and poverty reduction. It was not sufficient to tilt the composition of expenditure towards health and education in order to achieve better outcomes. As in many countries, the capacity to manage, execute and monitor expenditure in a transparent way was just as critical to improving the delivery of public services. The following are short and long term measures to improve expenditure management and monitoring, several of which are currently being undertaken by the Government with strong participation and support by donors. 24. As the country formulates and implements its first PRSP, these issues and measures are likely to draw strong interest and support from various stakeholders. Thus, it may be a good time to also consider a mechanism for coordinating and reviewing various efforts for improving the delivery of public services with the Government taking the lead. Medium to Long Term Reforms: * Introduction of a basic MTEF, which should be anchored in the strategic prioritizations of PRSP to link policy, public sector priority setting and the budget. * Development of IFMIS to provide consistent financial information for expenditure management and control. * Introduction of activity based budgeting and better budgeting and reporting techniques so that budgetary information by programs, activities and regional/district levels can be generated for planning and monitoring. * Implementation of public expenditure tracking surveys (PETS) on a regular basis so that public sector performance and the flow of resources to facility and ground levels can be monitored effectively. * Explore and plan for an inter-governmental fiscal framework (decentralization), linking to local or community based participation and social funds such as ZAMSIF, in order to enhance the execution of public expenditures and their impact on the poor. xxii Short to Medium Term Reforms: * Improve the realism of budget allocation using recent past releases as starting points and avoid significant in-year reallocations. * Improve expenditure smoothing and replace the current rationing system with well-defined rules and cash management system based on realistic quarterly projections. * Replace the protection of social spending with the following possible options: (i) protect all poverty-oriented spending (including infrastructure); (ii) maximum ceilings on non-poverty-oriented expenditure; or (iii) proportionate adjustments when there is a cash shortage. * Improve expenditure control and prevent new arrears by requiring timely and accurate reporting of monthly expenditure retums, enforcing compliance of commitment controls by Controlling Officers, and giving them high-level attention and monitoring. * Improve budget reporting and classification by - (i) completing the pilot phase of ABB; (ii) identification of poverty reducing priority expenditures in a more systematic basis; and (iii) creation of a revised functional database that incorporates all expenditures. - Implement a reporting mechanism for donor financed projects (outturns) to improve the coverage of expenditure reporting. * For the same reason as above, expand budget comprehensiveness to include extra-budgetary funds, independent agencies, and local governments. * Strengthen intemal and extemal audit capacity. Pilot PETS quickly (see long term measure for why). FINAL CAVEATS: 25. The reason why additional public spending on schooling, health and infrastructure are advocated frequently is because they are closely associated with growth and poverty reduction. However, raising public expenditures in these areas will not produce better outcomes without good policies and institutions that promote production over wasteful consumption and productive investment over rent-seeking diversions. 26. The reason why so much emphasis is placed on budget management and reporting is clearly to improve the delivery of public services and their impact on growth and the poor. However, all the modern information system and technology will not amount to much without the highest commitment to good govemance and fiduciary accountability. xxiii INTRODUCTION Purpose 1. The Public Expenditure Review (PER) is intended to provide the Government of the Republic of Zambia (GRZ) with information necessary to improve the effectiveness of its public spending towards achieving the societal goals of reviving growth and reducing poverty. The thrust of the report is the link between public expenditures, growth and the reduction of poverty. As such it is expected to contribute directly to the Poverty Reduction Strategy Paper (PRSP) the GRZ will be producing. The link to the PRSP and the enhanced HIPC (Highly Indebted Poor Countries) Initiative is an important justification of the study. 2. More specifically, the Zambia PER is a systematic assessment of the allocation of public resources in two ways - (i) across broad expenditure categories or sectors and (ii) within each key expenditure category or sector - utilizing the standard PER criteria of efficiency, equity, and implementability. In so doing, the report hopes to answer key questions regarding the appropriate mix of public spending on equity and efficiency grounds and the effective delivery of public services as the Governrment pursues the two basic goals of poverty reduction and growth promotion. Participation. 3. The report is a joint product between the Government of the Republic of Zambia and the World Bank, with support and consultations of other donors, particularly the Quality of Public Expenditure Donor Working Group. During a World Bank mission in March 2000, it was agreed to transform a traditional World Bank PER into a joint and Government-owned process. Working groups were immediately convened to conduct the necessary data collection and analysis. Outlines were agreed for the report and each chapter and the Government undertook the first initial drafts of most chapters. During a second World Bank PER mission in September and October 2000, the PER workgroups met to review initial drafts and sorted out main findings and policy recommendations. A Workshop comprising all the workgroups was held in Kabwe during October 6-8, 2000. The workshop included a mock budget exercise and provided participants with opportunities to comment and discuss findings on sectors other than their own, with Bank staff providing mainly advice and clarifying issues. It was agreed that the World Bank team will gather all materials and integrate the main conclusions of the Workshop, the initial government report, and relevant Bank materials and findings into a synthesis report. During a joint Bank-Fund mission in December 2000 and subsequent discussion with the Government held in Washington, D.C. in February 2001, preliminary findings were also utilized to inform the 2001 budget, including the identification of poverty- focused public expenditures to be funded by the HIPC debt relief as well measures for xxiv tracking these HIPC expenditures. Drafts were also utilized by the PRSP working groups as background materials. 4. To finalize the PER, the Government team convened a workshop in early October 2001 to review the synthesis report and drafted detailed comments and inputs. In a subsequent Bank Mission, the two sides discussed the proposed changes and reached agreement in all areas. All written materials from both the Government and the Bank are integrated into this report. The Need for PER. 5. At the heart of the growth problem, the persistence of poverty, and issues of policy reform in Zambia is the public sector reform program. A public sector reform program ought to define what the government should be doing and, perhaps more importantly as in many countries, what it should NOT be doing; and of the things it should be doing, what should be the priorities and how should government do them more effectively and within reasonable cost. The best practice in public sector reform identifies three areas in which governments can improve their performance and their impact on the economy and poor: (1) macroeconomic discipline (the stabilization problem); (2) strategic priority setting (the allocation problem); and (3) efficient public- service delivery (the execution problem). Zambia's problem appears to be in all three areas, rather than just on any single one. In particular, the aggregate perforrnance, allocation, and execution of the budget are vital to the success of Zambia's public sector reform program. All three areas as they apply to public expenditure are the subject matter of the present PER. Moreover, the last PER was completed in 19952; there is therefore a need to take stock of the expenditure issues and problems. Features and Coverage 6. The Link Between Public Expenditure, Growth, and Poverty is a key feature of the PER. The sustainability and execution of Zambia's public expenditure as well as the impact of public expenditure on growth and the poor are the main topics of Part I on cross-cutting issues. While the analysis covers mainly the experience in Zambia, cross- country evidence are also utilized in order to draw lessons from a wider range of country records on growth, poverty and public expenditure, which otherwise would be difficult to derive (generalize) from the experience of a single country. Part I: Cross-cutting issues (i) Macroeconomic constraints to public expenditure (chapter 1) (ii) Poverty, inequality and public expenditure (chapter 2) (iii) Composition of public expenditures (chapter 3) (iv) Management constraint to public expenditures (chapter 4) 2 In addition to the 1995 PER, the previous PER reports were from 1992 and 1987. In 1997, there was a related Bank Report on Fiscal Management. xxv Using consistent set of criteria, major categories or sectors of public expenditure are examined in Part II: Part II: Sector-specific issues (v) Education (chapter 4) (vi) Health (chapter 5) (vii) Agriculture (chapter 6) (viii) Water (chapter 7) (ix) Roads and other infrastructure (chapter 8) 7. Full assessment of some broad public sector issues are outside the scope of the present PER. The privatization of state-owned enterprises is discussed in the macro chapter, but the specifics are handled in the adjustment operations of the Bank, such as the Fiscal Sustainability Credit. Civil service reform, ri ght sizing and capacity building of the government are issues examined in the Public Service Capacity Building Project (PSCAP). Fiscal decentralization, along with the constraints of budget institutions and local governments, is a potential future issue to be pursued after the current PER. Other issues that could be examined in more general economic or sector specific reports include: sources of growth, rural development, management of natural resources for growth and tourism, environmental issues etc. Continuing Dialoque 8. The PER is envisioned as a step in a continuing dialogue on public expenditure. The Government has expressed the interest to take the lead in producing the next PER with World Bank Staff acting mainly as advisers and to use the PER as the basis for the, preparation of future national budgets. Furthermore, several measures to improve expenditure management and budget institutions, such as those outlined in Chapter 4, are also underway with strong cooperation from Government and donors. As the country formulates and implements its first PRSP, these issues and measures are likely to draw strong interest and support; it may be a good time to consider a mechanism for coordinating efforts for improving the delivery of public services with the Government taking the lead. xxvi PART 1: CROSS-CUTTING ISSUES OF PUBLIC EXPENDITURE 1. MACROECONOMIC CONSTRAINTS TO PUBLIC EXPENDITURES The Growth Problem (Economic Context) 1.1 Even among developing countries, Zambia represents an extraordinary case of an undiversified and landlocked economy, exhibiting a very high dependence on mineral resources and exports, subject to external price shocks as well as transport problems throughout its history. Zambia is among the poorest countries in Africa. Its per capita income in 2000 was valued at US$ 300. It is also one of the most heavily indebted and recently qualified for debt relief under the Heavily Indebted Poor Countries (HIPC) Initiative. Its population, about 10 million in 2000, is highly urbanized, and large parts of the country are thinly populated. Population is concentrated along the "Line of Rail" that links the Copperbelt with Lusaka, the capital, and with the border town of Livingstone. Apart from South Africa, Zambia is the most urbanized country in Southern Africa. At 20 percent, it also has one of the highest prevalence rates of HIV/AIDS. 1.2 Zambia's lack of growth in its per capita income is not a recent problem and is widely documented.4 The long-term economic decline is closely associated with the deterioration of copper fortunes. It has brought about painful contractions to a predominantly urban-based wage economy, which was once relatively rich in the region. It also made unaffordable the previously high levels of public expenditure, size of civil service, and amount of social interventions and assistance. Outside of the copper dependence and the long-term price decline of copper, key factors to the growth problem include transportation problems and disruptions arising from regional circumstances, command and socialist-oriented economic policies plus a dominant but inefficient public sector, and the debt overhang. 1.3 The Secular Decline of Copper Price and Zambia's Per Capita Income (see Figure 1.1). The real price of copper is currently only 26 percent of the peak level in 1973. Zambia's GDP per capita has likewise fallen; at 267 constant 1985 US dollars, it is currently half (54 percent) the level at Independence in the mid-sixties.5 Not surprisingly, Zambia's poverty has remained high and mirrored the long-term economic decline. Poverty incidence is over 70 percent by the national poverty line and near 64 percent based on $1 (PPP) a day. 3GDP per capita converted to U.S. dollars using the World Bank Atlas method. 4See, for examples, World Bank (1996 and 1993), Baldwin (1966) and Aron (1999). 5Measured alternatively in constant 1985 PPP international dollars, Zambia's per capita income is $640, which is about 58 percent of the level in the mid-sixties. 1 Figure 1.1: Copper Price and Zambia's Per Capita Income 600 2000 c, 500 = 60X 500~~~~~~~~10 _ 1200 m -80 2o n'300 200 80 O- 2 0 01 100 400 0 -0__ _ _ 1960 1970 1980 1990 2000 Real price of copper (US MUV deflated) ----- GDP per capita (constant US dollars) - - --GDP per capita (constant PPP prices) Sources: World Development Indicators (World Bank); GDP per capita in constant PPP, 1960-91, is from Penn World Table (Mark 5.6); 1991-99 is approximated from World Bank" s GDP per capita in current PPP deflated by US GDP deflator. 1.4 In the sixties, Zambia was the third largest producer of copper in the world next to the United States and Russia. At the time of independence on October 24, 1964, the mineral sector (mainly copper) contributed nearly 45 percent of Zambia's GDP, earned close to 90 percent of its foreign exchange, accounted for 65 percent of public revenue, and employed 20 percent of the formal sector employment. Although the dominance of mining/copper sector has gradually declined, it remains the most significant export earner (over 50 percent of forex) and its effects on the economy remain very significant. 1.5 Severe Transportation Problems and Disruptions. Civil strife, economic sanctions, and other developments in the region have closed key routes to the coast for much of Zambia's history since independence. The 1,060-mile Tazara railway to the Tanzania port of Dar es Salaam, which connects with the older railway at Kapiri Mposhi, has not carried the projected volume of traffic, owing partly to problems with track and rolling stock and partly to congestion at the port of Dar es Salaam. The eight-inch petroleum pipeline connecting Dar es Salaam to Zambia's Indeni refinery in Ndola has been saddled with inefficiencies of state ownership in the petroleum sector. Beneficial political changes in Southern Africa have lessened the use of the northern routes in recent years. Lighter-weight, non-traditional exports like horticulture products are increasingly being transported by air to their export destinations in Europe and elsewhere. Nevertheless, the lack of cheap transportation or easy access to ports remains a natural barrier to trade and development. 1.6 Socialist-Oriented Economic Policies Plus a Dominant But Inefficient Public Sector also exacerbated economic difficulties and poverty in Zambia. State involvement became a feature in all aspects of the economy soon after Independence and created a highly centralized and bureaucratic economic structure. The fundamental switch in the structure of the country's economy came with the Mulungushi Reforms of April 1968, in 2 which the government declared its intention to acquire majority holding of key foreign- owned companies, mostly in the mining and banking sectors. Unfortunately, the impact of state-dominated economic planning, policies and controls contributed to bringing about economic inefficiencies and problems that characterized the Zambian economy. The control regime distorted sectoral growth and in fact increased the vulnerability of the economy to external shocks. State ownership and disinvestment in the mineral sector failed to ensure the long-term efficiency of the mining sector or smooth mineral revenue across trade shocks. Revenue was also channeled to propping up a large import- dependent and largely inefficient parastatal sector. Despite policy pronouncements, agriculture did not receive the necessary support and experienced a Dutch disease-like stagnation.6 Table 1.1: Zambia: Debt Service Paid and External Flows 1990-2000 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Debt service paid, US$ million' 290 655 354 326 409 1,584 319 217 147 136 188 Debt service paid in % of exports of 21.6 56 29.6 31.2 34.8 120.4 28.7 17.6 16 16.2 20.7 goods and non-factor services Debt service paid in % of GDP 7.7 19.4 10.7 10 12.2 45.7 9.7 5.6 4.6 4.3 5.5 Debt service paid in % of 38.2 103.7 58.3 63.2 60.9 230.2 47.1 27.9 24.4 24.5 23.1 Government Revenue Gross external inflows (incl. - - 1,106 795 550 1,816 510 401 297 511 776 rescheduling and other flows) Of which: IMF - - 0 0 0 1,264 0 14 0 14 27 IDA 3 213 174 174 186 209 181 169 43 156 210 Net external flows (official) - - 752 469 141 232 191 184 150 375 434 Net external flows in % of GDP - - 22.7 14.4 4.2 6.7 5.8 4.7 4.7 11.9 12.6 Source: HIPC Decision Point Document (2000). 'The hump in debt service in 1991 and 1995 is due to repayment of arrears to international multilateral institutions, much of which was offset by new disbursements. 1.7 Zambia's Debt Burden. By mid-1980s, Zambia had become one of the most indebted nations in the world relative to its GDP. The programs of nationalization, particularly of the mining industry, were not only ill advised but ill timed with hindsight. The massive increase in the price of oil in 1973 was followed by a steep rise in import bill and a collapse of copper prices and export earnings in 1975, resulting in a severe balance-of-payment crisis starting in 1976. Zambia's external debt more than doubled in 1976 and the country started building up external arrears, reaching its peak in 1990 ($2.2 billion) before reducing to $1.6 billion in 1994. At end-1999, Zambia's nominal external public and publicly guaranteed debt reached about US$6.5 billion, nearly twice the level of GDP. Two features characterized Zambia's external debt that were very relevant to the budgetary problems in the 1990s - a heavy burden of debt service and high volatility of payments from year to year. External debt service paid ranged from 16.0 to 120.4 percent of exports, 24.4 to 230.2 percent of government revenue, and 4.3 to 45.7 percent of GDP. The country's external inflows also showed wide instability and, as a result, the 6 See Aron (1999) for a detailed analysis. 3 net external inflows (net of debt service) varied from a high of 22.7 percent to a low of 4.2 percent of GDP (see Table 1.1). 1.8 The fluctuation of net external financing imposed severe restrictions on fiscal operations and public expenditures. In the 1 990s, the Zambian Government took steps to insulate the domestic budget from the fluctuations of net external flows. Domestic expenditures are covered by domestic revenues while BOP inflows mainly pay for the high external debt requirements. The recent decision by donors to grant interim HIPC debt relief to Zambia should create budgetary headroom and improve the resource outlook for poverty-related expenditure. Recent Liberalization and the Changing Role of the Public Sector 1.9 Beginning in 1991, Zambia began dismantling the unsustainable control regime and undertook fundamental changes in economic policy. In addition to liberalizing the economy and trade regime, structural reforms also redefined the role of the public sector from its previous dominant position to a more supporting one in a market economy. Key reforms included: (i) privatization of state-owned enterprises; (ii) decontrol of agricultural prices and liberalization of maize marketing; (iii) public sector downsizing, while redirecting expenditure toward the social areas; (iv) removal of exchange controls and floating of the Kwacha; (v) trade liberalization, including the removal of quantitative restrictions on imports and exports, as well as the reduction of the level and dispersion of customs tariffs; and (vi) liberalization of the banking sector. On the face of it, the list of reforms appears impressive. 1.10 But Inadequate Growth Persisted in the Nineties. During 1995-2000, the growth rate in GDP averaged 2.1 percent per annum, only slightly better than the out-turn in the 1990-1994 period when it averaged a 1.5 percent per annum. Given that population was growing by 2.9 percent per year (2001 census of population and housing preliminary report), per capita income was therefore falling by about 1.0 percent a year in the nineties. Table 1.2 shows the variability of recent GDP and sectoral growth. Table 1.2: Real GDP and Sectoral Value-Added, 1995-2000 (annual percent growth) 1995 1996 1997 1998 1999 2000 Average Gross Domestic Product (GDP) -2.5 6.6 3.3 -1.9 2.0 3.5 2.1 Agriculture 33.4 -0.6 -5.1 1.2 6.9 1.8 6.7 Mining -27.6 2.8 2.2 -25.1 -24.8 -5.1 -16.5 Manufacturing -0.4 5.5 5.5 1.8 2.8 13.5 5.7 Construction -3.4 -11.0 29.0 -9.1 10.3 1.2 2.4 Energy -1.6 -5.6 4.2 0.6 2.5 1.2 0.2 Service -1.4 -0.9 10.7 -1.5 4.7 8.1 3.8 * Geometric average Source: Central Statistical Office, National Accounts Statistics 1.11 Why the continuing inadequate growth, particularly in the second half of the nineties when the beneficial effects of liberalization should have borne fruit? Certainly, 4 negative copper price shock, adverse weather conditions (El Niflo), and the delays in the privatization of the largest state-owned enterprise, Zambia Consolidated Copper Mines Limited (ZCCM), were key factors. More significant, however, is the fact that policies and reforms did not go far - which is examined in the next sections. 1.12 Impact on Expenditure and Revenue Trends. The inadequate growth and the reduced role of the public sector implied that the budget was not expanding in real terms - real public expenditure per capita in the last 10 years continued to decline but showed signs of a slight rebound in the last few years, while real revenue per capita, after an initial decline, managed to stay roughy at the same level (Figure 1.2). Figure 1.2: Per Capita Revenue and Expenditure In 1994 constant prices 350.0 300.0 2 50.g0' 200.0o -_ 150.0 100.0 1990 1992 1994 1996 1998 2000 ! ---Per capita revenue Per capita expenditure 1.13 Table 1.3 shows the year-by-year ratios to GDP of government revenue and expenditure during the more recent years, 1995-2000. Table 1.3: Revenue and Expenditure Trends, 1995-2000 (percent of GDP) 1995 1996 1997 1998 1999 2000 Average Domestic revenue 19.9 20.7 19.9 18.8 17.9 21.6 19.7 Tax revenue* 18.2 19.0 18.8 18.2 17.4 19.2 18.4 Foreign grants 9.2 6.1 5.1 6.6 8.1 6.4 8.3 Government expenditure 29.3 24.6 23.0 26.9 25.9 27.8 25.8 Current expenditure 24.2 18.6 17.6 17.3 15.3 16.8 18.5 Capital expenditure"* 5.1 6.0 5.4 9.6 10.7 11.0 7.3 Domestic expenditure 19.3 18.5 17.3 16.9 17.2 .. 17.8 Overall fiscal balance - cash basis 0.1 -0.4 -2.0 -4.4 -4.0 -6.7 -2.1 Overall fiscal balance -4.3 -0.5 -0.2 -3.2 -2.1 -5.4 -2.0 * Tax revenue not inclusive of ZESCO tax arrears collected in 2000; ** Capital expenditure does not include Net Lending Sources: Economic Reports and Macroeconomic Indicators - Ministry of Finance and Economic Development. 1.14 Domestic revenue, domestic expenditure, expressed as a percentage of GDP, generally all declined during the review period. Zambia's tax revenue averaged 18.4 percent of GDP, which compared very well to countries in Sub-Saharan Africa. The ratio of domestic or domestically financed expenditure to GDP averaged about 17.8 percent, but total expenditure, including externally financed expenditure (foreign current and 5 project/capital grants) averaged near 26 percent for period. Foreign grants averaged 8.3 percent of GDP. This ratio also indicates the high level of dependency on extemal financing by the cooperating partners of the national budget. Note that capital expenditure has improved steadily since 1995, rising from 5.1 to l1.0 percent. Capital expenditure did not necessarily go to infrastructure; it included a large amount of project grants from donors, with a significant part going to education and health sectors. All told, the overall fiscal balance of the central government, whether on a cash or accrual basis, looked reasonable and averaged -2.0 percent. However, this is not a complete picture, as the next section on macroeconomic discipline will show. 1.15 Constraints and Welfare Cost of Raising More Public Revenue. Despite the growth problem, Zambia's tax revenue, expressed as a percentage of GDP, averaged near 18 percent in the nineties, which compared very well to countries in Sub-Saharan Africa (SSA). The SSA average, excluding South Africa, is 15.6 percent; Ethiopia, 10.9 percent; Malawi, 14.3 percent; Tanzania, 11.2 percent; Mozambique, 11.0 percent; and Uganda, 8.3 percent. Hence, raising the revenue effort any higher would likely be difficult without addressing the growth of the tax base. In addition, it is not altogether clear where taxes could be raised at the least marginal cost to the economy. Distortionary taxes like import tariffs have certainly been streamlined, but significant tax or fiscal concessions were recently granted to the mining and hotel industries so that tax distortions outside of mining and hotel sectors would increase with domestic taxation. The net cost to the economy could amount to as much as 2.5 kwacha for every additional kwacha of tax money raised, depending on whether other distortions (in the factor markets, informal sectors etc) in the economy are enhanced or minimized by the tax change.7 Like expenditures, there is a need for a comprehensive and periodic review of the tax system in Zambia. 1.16 Structure of Revenue: Moreover, total revenue (including current and capital grants) received by the Government maintained a relatively stable share of GDP, around 25 percent. The two major components of Government's income, domestic revenues and external grants, were substantially more volatile than their sum, with grants off-setting shortfalls in domestic revenues. The decline in tax revenue and the fluctuation of nontax receipts (which included income from parastatals) were always matched by off-setting and rising external grants - which ranged from 5.1 to 9.2 percent of GDP. Moreover, the extemal grants in 1999 were below the 10 percent of GDP attained in the early years of reform during 1991-92. Table Al.l in Annex I shows the details. Macroeconomic Discipline, Growth, and Poverty 1.17 Macroeconomic discipline, a basic lever of public sector reform, is closely associated with inflation, growth, and poverty as increasingly supported by empirical evidence. 7 See, for example, Devarajan, Thierfelder, and Suthiwart-Narueput (2001) for calculation of the marginal cost of public funds in developing countries. 6 1.18 Public Sector Deficit and Inflation. The link between the fiscal deficit and inflation is high in Zambia (see Figure 1.3). The period of high inflation generally corresponded to the period of rising fiscal deficit and its financing by the banking sector. Zambia's overall fiscal deficit (after grants) averaged near 14 percent of GDP from 1986- 1993 when inflation was in excess of 50 percent per year, reaching close to 190 percent in 1992. A significant break is noticeable after 1993 when a cash rationing or budgeting system was introduced in Zambia. The high inflation was contained, which, no doubt, is a key achievement of the cash budget. Nevertheless, inflation still averaged nearly 25 percent from 1995 to 1999, reaching nearly 30 percent in 2000 and showing no signs of abating or going much below 20 percent. In contrast, inflation in the entire region of Sub- Saharan Africa has been moderate and averaged only 8.5 percent for the same period. The persistence of inflation at worrisome levels is a clear sign that there are underlying problems beyond the smaller deficit of the central government in recent years. No doubt, adverse terms-of-trade shocks, BOP pressures on the exchange rate as well as rising food prices during drought were important factors in recent years. However, fiscal discipline in the broader public sector was also a key problem. Figure 1.3: Deficit and Inflation 50.0% 200.0% 40.0% 150.0% 0.~~~~~~~~~~~~~100 I 30.0% - I O ,l ,;V: \ 100.0% 0 20.0%~~~~~~~~~~~~~~~~~~~~~~~. 50.0 ai 10.0% °-20.0% - Q O0 m. Co Co 0 Co Co Co Co CD 0 0) m 0 0) 0 i -i0 . 0% -_ . - .. -- __ - __ ._-- IM....... L_ -50 .0 % I- - - Overall Fiscal Deficit in Percent of GDP -Change in Banking Sector's Net Claims on the Gov't. as Percent of GDP _|__ Inflation (Annual Percent Change) Note: Fiscal deficit is overall deficit of the central government; data are from Ministry of Finance and Economic Development, the Live Database and economic reports from the World Bank, IMF Intemational Financial Statistics, and Bank/lMF staff estimates for recent years. 1.19 The quasi-fiscal deficit (arising mainly from losses of state-owned enterprises) is the main problem. The sustainability of the overall fiscal position is still very weak because deficits in other segments of the public sector have not been brought effectively under control. The overall public sector deficit (quasi-fiscal deficit) in Zambia -- which includes the deficits of the central government, the local governments, extra-budgetary accounts, state-owned enterprises, and the Central Bank - remained very high. Although generally difficult to estimate and verify and there is a clear need to carefully account and update the figure, the best available estimate8 indicates that the public sector deficit in See World Bank (1998 and 2000) and Hinh Dinh (1999). 7 Zambia amounted to 13.6 to 18.4 percent of GDP during 1995-97. Of this amount, losses of state-owned enterprises are the biggest drain on public revenues (see Figure 1.4). In fact, the main source of additional deficit in the public sector in more recent years came from the operational losses of ZCCM when copper prices collapsed during 1997-2000; from 1998 onwards, they reached US$Im a day, which constituted an additional quasi- fiscal deficit of 10 percent of GDP. Worried donors started preconditioning new aid on its final disposal. Figure 1.4: Zambia's Overall Public Sector Deficit Zambia: Overall Public Sector Deficit by M#jor Zambia: Percentage Share in Overall Public Sector Deficit, Components, 1995-1997 1997 I . Pension Funds 7% 16 - Local Govt 15% Puiblic (6,zf- 13.8 _ _ _ Enterprises 3 2 13.8 36% 15.3 11. Z 1 p- 11.3 _ Governren t \ __ ~~~~ ~~~~~~~~~~~17% ~ \\\ r__ 4.6 _ 2.5 2.3 \ \ 1995 1998 1997 BoZ Deficit Year 25% ocntri Govrnmnt N-coven nentr |Public Enterprises E3BoZ Deficit OCentral Government E Local Govt [ Persion Ful Source: World Bank (2000). 1.20 The heavy losses of state-owned enterprises easily overshadow any budgetary spendingfor the poor - the size of the quasi-fiscal deficit estimated above (1995-97) is more than half of total budgetary spending, inclusive of foreign current and capital grants; ZCCM's recent annualized losses just before privatization is more than a third of total spending. The sale of ZCCM in March 2000 is therefore a significant event and a major policy achievement towards correcting an unsustainable public sector financial situation. For the same reason, it is important to continue the structural reforms and tackle other leakages in the public sector arising from the remaining key parastatals - such as Zambia Electricity Supply Company (ZESCO), Zambia National Oil Company Limited (ZNOC), Zambia National Commercial Bank (ZNCB) etc. - as well as other public entities like the public pension funds, local government, etc. 1.21 Inflation and Growth. While there are several factors to inflation and growth in Zambia (such as terms-of-trade shocks, bad agricultural harvests etc.), public expenditure, the problem of state-owned enterprises and their impact on the public sector deficit and its sustainability are key underlying pressures. Time-series analysis indicates that additional public expenditure in Zambia has historically been linked to episodes of severe macro imbalances and has therefore not been productive or growth enhancing. The statistical analysis of the short-term fluctuations suggests that the small expansionary effect of government spending in the first year is generally offset by a contractionary 8 effect in the next year (see Table A2.2 in Annex II). There is no indication of significant trade-offs in the short run between a worsening of the fiscal deficit and some amount of economic growth generated from an expansionary expenditure policy (controlling for such factor as terms-of-trade shocks). On the other hand, Figure 1.3 and the discussion thereabouts suggest that public sector deficit and inflation move closely together in Zambia. When inflation is added to the time-series analysis (controlling for other determinants of growth such as terms-of-trade shocks, public expenditure etc), the impact of inflation on growth is negative in the second and third year and the magnitude of the impact is consistent with the cross-country findings below. 1.22 Lessons from recent cross-country evidence, drawn from a large sample of countries with a wide range of inflation and growth records, confirms that high inflation is indeed harmful to growth. Table 1.4 illustrates the impact of reducing inflation on growth while controlling for other underlying factors (these factors vary from study to study but generally include growth determinants, structural factors, fiscal and other policy, terms-of-trade shocks etc.) Although application should be taken with care since there can be wide variation and not all results are straightforward,9 the indicative lesson for Zambia from these cross-country results is that a drop in inflation from 30 to 5 percent a year will likely increase the rate of growth of output per capita by roughly 1 percentage point (ceteris paribus); this would mean a moderate long run GDP growth of 4 (up to 4.5) percent per year, something not seen in Zambia for a long while (see Table 1.4) Moreover, polling data from 31,869 households in 38 countries indicate that inflation makes the poor worse off and that the poor are more likely to complain about inflation as a top concern than the advantaged; the inflation tax rate also tends to reduce the share of the bottom quintile and the real minimum wage, while tending to increase poverty.'0 Table 1.4: The Potential Effect on Economic Growth of a Decrease in Inflation From 30% to 5% Per Year No. of Period Date Approx. Effect on Countries per capita growth (%) Barro (1997) 80-87 1960-90 Panel data 0.8-1.1 Bruno and Easterly (1998) 97 1961-92 Panel data 0.7 Fischer (1991) 73 1970-85 Cross-section 1.3 Fischer (1993) 80 1960-89 Cross-section 1.1 Ghosh (2000) - various IMF Studies various Various Various 1.5 Gylfason (1999) 160 1985-94 Cross-section 1.4 Gylfason & Herbertsson (1996) 145-170 1960-92 Panel data 0.4-0.8 Roubini and Sala-i-Martin (1992) 98 1960-85 Cross-section 1.3 Relative to a decrease of inflation from 30 to 10%, given that a bottom kink or threshold inflation of close to 10% is found in the IMF studies. See text for discussion 1.23 State-owned enterprises and the importance of being efficient. Although there is a wide variation in cross-country experience, preliminary empirical evidence suggests a 9 Bruno and Easter (1997), for example, found that output declines sharply with "inflation crises," but output eventually reverts back to trend after the inflation crisis; hence, the elasticity of inflation on growth over the entire inflation cycle is zero. Levine and Zervos (1993), on the other hand, did not find inflation robustly related to growth. '° Easterly and Fischer (2000). 9 general inverse relationship between public ownership or share of resources (e.g. employment) and economic growth (see Figure 1.5). The existence of a dominant public sector seems to be associated with inefficiency, the crowding out of the private sector, and low growth of the economy. Figure 1.5: More State Enterprise Is More or less Associated with Less Income IOOODO Tnidad and Tob Gibun >Mauribu5 ' zGreada ,.-CoIombia 1000 Sri iLG haa -Madagascar .1 \ B! n Zb 1i0 0.00 v1a 0.o0 0.30 D.40 0.50 a. v 0.70 0.3W - SHARE OFSTA!E-0VWNEO NTERMSES IN EMPtOYMWr 1970. 1995 Source: Gylfason, Herbertsson, and Zoega (1998). Zambia's point added. 1.24 The experience of Zambia illustrates further the important connection between privatization (structural reforms), growth and inflation. On the face of the activities of Zambia Privatization Agency (ZPA), the pace of privatization in Zambia was impressive; 247 state firms of ZPA's working portfolio of 299 had been sold and 24 more were under preparation by early 2001. However, the tendency or implicit strategy followed in Zambia was to pursue the smaller, less controversial privatizations first, leaving the larger, technically and politically difficult privatizations (like ZCCM, utilities, the oil sector etc.) for later."1 No doubt, the political economy of large privatizations also favored inaction and delay. As a result, the economic impact of privatization in terms of additional investment and output growth had so far been small. Meanwhile, the risk of accumulating further arrears in large state-owned enterprises and their potential fiscal impact could not be overestimated. In the case of ZCCM, the delay of its sale proved extremely costly from hindsight and the final deal was less favorable and involved more concessions - including exemptions from taxes for five years, a 20-year fixed and low rate for power supply, duty exemptions, and government agreement to pay for labor retrenchment. In addition, the Government agreed to assume ZCCM's substantial outstanding debt totaling US$ 770 million, which is about the size of total budgetary spending.I Nonetheless, ZCCM privatization is now leading a rapid turnaround in copper production and economic activity in the Copperbelt. It also brought about increased donor financing and was a significant factor to reaching the decision point of HIPC debt relief at end-2000. " For more details, see World Bank (2001). 2 Ibid. 10 1.25 The case of ZCCM suggests various substantial effects. First, the losses of state- owned enterprises are likely to be covered by public sector borrowings, which add to inflationary pressures. Second, the more privatization is postponed, the greater the risks for more losses that would have to be paid out of future budgets when publicly guaranteed obligations are settled at the time of sale, hence eating into the potential savings or gains of privatization. Payments for ZCCM's debt in the Government budget are covered under the item "net lending" (part of capital expenditure) and is projected to amount to about 1.5 to 2.0 percent of GDP in the next few years. A significant part of the debt is also assumed by the Bank of Zambia and the banking sector. 1.26 Growth and Income of the Poor. Recent studies also emphasize that without growth, sustainable poverty reduction is not possible. The quality of growth may matter, but there has to be growth. Cross-country evidence indicates that the income of the poor rises one-to-one relative to the overall GDP growth.'3 More detailed analysis based on household data of several countries suggest that the poor in developing countries typically do share proportionately more in the gains from rising aggregate affluence and that the estimated "growth elasticity" of poverty is as strong as -2.50. 14 More precisely, for every one percent increase in the mean (household income), the proportion of the population living below $1/day (at 1993 Purchasing Power Parity) falls by an average of 2.5 percent. 1.27 In particular, Figure 1.6 and 1.7 illustrate the link between growth and poverty reduction in Zambia: * Figure 1.6 presents the association between mean household income and poverty of each province in 1998. Provinces with less mean income, such as Western, Luapula, Northern and Eastern provinces have higher poverty rates; conversely, richer and more urbanized provinces like Lusaka and Copperbelt have much lower poverty rates. The relationship holds even more for extreme poverty, which declines more rapidly with higher income. * Figure 1.7 shows the relationship between expenditure in the poorest quintile and overall mean expenditure in constant 1998 kwacha, over households, and by province for three household surveys, 1993, 1996, and 1998. The outlier at the bottom right belongs to Lusaka province in 1993. The scatter plot clearly reveals that mean consumption (or the living standard) of the poorest quintile improves with overall mean consumption (average living standard) of the provinces. The same pattern holds when the consumption data are presented in mean per adult equivalent expenditure, which is a more comparable unit (i.e. over individuals). 13Dollar and Kraay (2000). 14 Martin Ravallion (2000). 11 Figure 1.6: Mean Income and Poverty Rates In Zambia by Province, 1998 90 80- Westerm 70- Lut'apua Northern Northwestern Eastern * 0 60 * Central Southem 3 50*- * | + poverty rat Copperbelt mlextreme poverty ID >40- 30 20 30- 0- 0 10000 20000 30000 40000 50000 eoooo 70000 0oo0 9000o 100000 Mean Houshold Income (Kwa) Figure 1.7: The Relationship Between Expenditure in the Poorest Quintile and Mean Expenditures in Zambia by province in 1993,1996, and 1998 12,000 oE AA O', 10,000-A * * D- O A 2 ~~~A 0 00 8,000 .0 6,000 w~~~~~~~~~~ c 4r000 0. x 2,000 - 20,000 40,000 60,000 80,000 100,000 Mean per adult equivalent expenditure (in 1998 Kwa) U Poorest quintile, 1993 A Poorest quintile, 1996 0 Poorest quintile, 1998 The "growth elasticity" of the consumption of the poorest quintile to overall mean consumption is very close to but slightly less than one (0.85) - which implies that the living standard in the poorest quintile in Zambia also rises almost one-to-one with the mean (and conversely with contraction of the mean). The growth elasticity rises slowly with each quintile and the richest quintile has an elasticity slightly greater than one 12 (1. 11). 15 Hence, during growth period, the rich gain slightly more than the poor; during contraction on the other hand, which is more the Zambian experience, the poor are protected slightly more than the rich. This is splitting hairs however; compared to the much higher cross-country estimate above, these numbers are all close to one. It does emphasize that the beneficial poverty response to income growth in Zambia lags behind the average of other countries. Public Expenditure, Growth, and Poverty - Prospects, Risks, and Links 1.28 Macroeconomic Framework. The privatization of the ZCCM in March 2000, has brightened economic outlook and has led to an increase in donor support and funding, including the approval of interim HIPC debt relief in December 2000. Zambia's economic outlook is expected to improve with higher investments in the mines and rising copper prices. Inflation, which reached close to 30 percent in end-2000 partly because of the recent oil price shock and monetary expansion, needs to be contained. Though expected to be temporary, the oil shock in 2000 will likely require additional external financing of around $60 million in 2000-02, or some 1.7 percent of annual GDP.'6 Moreover, much remains to be done to ensure that the government continues with economic reforms, including governance. Key aspects of the consensus forecast between GRZ, IMF and IDA include (Table 1.5): - GDP is projected to grow by 5 to 5.5 percent a year. * Public revenue and expenditure as percent of GDP will remain relatively stable at about 18 and 30 percent, respectively. TABLE 1.5: MEDIUM-TERM MACROECONOMIC FRAMEWORK, ZAMBIA Consensus Forecast from GRZ, IMF and World Bank, in % change, unless otherwise specified 1999 2000 2001 2002 2003 Real GDP 2.4 3.6 5.0 5.0 5.5 Inflation-endyearCPI 20.6 30.1 17.5 12.0 10.0 Terns of trade - 5.7 1.9 3.1 3.5 2.4 Copper price (US$/lb) 0.70 0.83 0.87 0.91 0.92 Current account balance (percent of GDP) -15.2 -18.5 -19.9 -21.1 -19.3 Revenue and grants (percent of GDP) 25.5 25.4 26.7 23.5 23.4 Revenue (percent of GDP) 17.6 19.6 17.3 17.7 18.1 Expenditure (percent of GDP) 29.2 31.2 31.3 29.7 28.2 Overall balance (cash basis - percent of GDP) -4.0 -7.1 -5.0 -6.6 -5.2 0. balance w/o grants(cash basis-percent of GDP) -11.9 -12.9 -14.5 -12.4 -10.5 -_ _ -1s i __ -1 L__: ___ _rn A A> X ~ At lw z I 15 The growth elasticities by quintile from regressions over individuals with 27 observations are as follows (stantard errors are in parentheses): qI (poorest) = 0.85 (0.0055); q2 = 0.92 (0.0069); q3 = 0.97 (0.0062); q4 = 1.01 (0.0066); q5 (richest) = 1.l 1(0.0064). R2's are all greater than 0.9. The estimates from regressions over households are very much the same. 16 IDA provided about US $30 million at end-2000 for the oil shock. 13 1.29 Savings and Gains from Privatization (reforms): A key question relates to the net savings of privatization of key parastatals or other structural reforms. There are two interrelated dimensions to the question - (i) the net benefit of the reform and (ii) its economic and fiscal implications. * The expected net benefit of a key reform like ZCCM turns out to be very difficult to determine (and has yet to be done) because it requires calculating the expected streams of future revenues and payments of Government from the complicated deal (which would entail projecting the long term future price of copper). * For the country, there are potential indirect effects that could become very significant: (i) greater long term growth arising from additional investment, greater efficiency, and output expansion of new private owners; (ii) greater macro stability and less inflation arising from the reduction of public sector deficit and borrowings; and (iii) to the extent that the reform gains donor support, additional donor financing including the HIPC debt relief. * The economic and fiscal implications for the next three years are factored into the medium-term consensus projections as follows: 1. GDP will grow by an additional 1.5 to 2.0 percentage points. Even keeping the total expenditure-GDP ratio constant, this should translate to an annual increase of 2.5 percent in real expenditure in the budget. 2. The reduction of the public sector deficit and borrowing will help bring inflation down to less than 20 percent and eventually to about 10 percent. This in turn will help keep fiscal and wage cost in the public sector reasonable. 3. In the external BOP account, a significant amount of HIPC interim debt relief, about US$270 million a year, will offset the expected rise of external debt service payments during 2001-03.'7 In addition, about $210 million of donor BOP support grants and loans are already programmed for 2001. 4. Direct extra-budgetary headroom is also created with HIPC debt relief (from reduced interest payments in the budget), amounting to about 2.7 percent of GDP in 2001 and rising to 3.2 percent in 2003. The government has identified about US$86 million (about 2.7 percent of GDP) in poverty-related activities that will be funded by HIPC resources (See Box 2.1). 17 Without HIPC assistance, Zambia's debt service payment would have reached $436 million in 2001 and stayed over $400 million till 2005 (after the traditional debt relief mechanism). With HIPC debt relief, debt service is reduced to $1 58 million, rising to slightly over $200 million in 2004 and 2005, before falling significantly in 2006. For comparison, the debt service paid in 1999 and 2000 were 136 and 188 millions, respectively. See the HIPC Decision Point Document, IMF and IDA (2000). 14 5. On the negative side, fiscal incentives to the mining and other sectors, including the fiscal concessions in the ZCCM deal, will contribute to a 2 percentage point decline of revenue-to-GDP ratio from 19.6 percent to 17.3 percent in 2001, but eventually recovering somewhat to 18.1 percent in 2002. 6. A significant part of the payments of ZCCM's debts/arrears (US$ 700 m) will come from the Bank of Zambia and be spread over several years. The portion in the budget, under the item net lending of capital expenditure, will add close to 2.0 percent of GDP. Possible payments for outstanding liabilities of other key parastatals programmed for privatization have yet to be accounted for; a viable plan for their eventual settlements or cancellation will also have to be developed. 7. Bottom line: Because there are downside risks to the prospects (see below and Annex II) as well as financial claims that need to be paid as a result of privatization, it is prudent to keep total expenditure-to-GDP ratio constant at about 30 percent, with budgetary headroom created mainly from the HIPC interim debt relief. The PER suggests some directions for change, but the medium-term allocation of budget resources, including HIPC assistance should ideally come from the priority setting among stakeholders in the formulation of the country's PRSP. 1.30 Risks Factors to the Medium-Term Projections are: * weakening of copper price because of a possible slowdown in world demand; and * expenditure pressures from the presidential election in late 2001. 1.31 Using the 123PRSP macroeconomic framework for looking at the link between public expenditure, growth and poverty,'8 alternative scenarios suggest that the two shocks above could: (i) lower GDP growth by 0.5 to 1.4 percentage points; (ii) raise fiscal deficit to GDP by 2.3 to 3.4 percentage points; and (iii) reduce household incomes in the bottom two deciles by much as 5.3 to 6.4 percent relative to the reference forecast. 1.32 Risks to Long-Term Growth. The consensus macroeconomic framework is contingent on the realization of key reform measures identified in the IMF's Poverty Reduction and Growth Facility (PRGF) and the Bank's Fiscal Sustainability Credit (FSC) with the Government: * The main priorities continue to be macroeconomic discipline and reduction of the public sector deficit, implementing key structural and public sector reforms, strengthening budget management and governance 18 See Annex II for a short description of the framework and details of alternative scenarios. See also Devarajan and Go (2001). 15 measures, and laying the basis for poverty reduction by increasing social and poverty related spending. * The overall fiscal deficit (cash basis), currently 7.3 percent of GDP, is expected to fall and remain at about 5 percent of GDP. * Inflation is targeted to fall to about 10 percent by 2003. * The focus of key structural reforms continue to be state-owned enterprises, including the deregulation and privatization of the petroleum sector and privatizing the controlling interest of the largest domestic bank, the Zambia National Commercial Bank (ZNCB). Pension reformn is another area requiring attention. * The efficient and transparent delivery of public services requires fiduciary accountability, including the implementation of a medium term expenditure framework, strengthening of the budget and financial management, and the effective control of expenditure commitments and arrears. 1.33 Altemative scenarios from the cross country-based "Get Real" module of the 123PRSP framework (see Annex II) suggests that a 5.5 percent long-term GDP growth would require substantial progress in the reforn program, such as the following: (i) no economic distortions (no black market premium and no overvaluation of the real exchange rate); (ii) HIPC debt relief that reduces interest payment as ratio to GDP by more than half; (iii) an inflation rate of less than 10 percent; (iv) a very favorable export environment; and (v) substantial increases (over 30 percent) in the school enrollment rate and infrastructure. The main point being emphasized from the cross-country findings is that concerted reform efforts are necessary to break away from a low GDP growth trend. Moreover, HIV/AIDs is also a significant factor in Zambia; in addition to its potential fiscal cost, several studies in the region suggest that GDP growth could be reduced anywhere between 0.5 to 1.5 percentage points a year. Relative to the program target, a reform program that is off-track plus HIV/AIDs could reduce household incomes in the bottom two deciles by as much as 4.5 to 5.4 percent a year.19 1.34 Public Expenditure, Growth and Poverty - the Role of Policy and Institutional Reforms (the missing link). The recent evidence on growth determinants is presumably why additional public spending on schooling, health, and infrastructure are advocated frequently. There is no evidence however that simply raising public expenditures in these areas will lead to better outcomes, growth and poverty reduction. In fact, overall decomposition of growth attributes only about a sixth to physical capital accumulation, about a third to human capital accumulation, but as much as half or more to the residual that raises factor productivity. The common underlying factors to the efficiency of human and physical capital are now widely viewed to be the set of policies and institutions, broadly grouped together as social or institutional infrastructure, that promote production over wasteful consumption and productive investment over rent- seeking diversions. To be sure, favorable extemal circumstances like HIPC debt relief, 9 See Annex II for details. 16 terms of trade, and trade expansion will matter, but the conclusions are fairly consistent across studies and have strong implications to public expenditure and foreign aid. Several World Bank studies (see Box 1.2) confirmed that the efficacy of public expenditures and foreign aid are strongly dependent on sound economic policies and good governance. Conclusions 1.35 The overall public sector deficit (quasi-fiscal deficit) in Zambia - which includes the deficits of the central government, the local governments, extra-budgetary accounts, state-owned enterprises, and the Central Bank - is still very high and continues to be a key threat to macro stability, growth and poverty reduction: * The size of the quasi-fiscal deficit easily overshadows any budgetary spending for the poor (Figure 1.4). The magnitude before recent ZCCM's losses is more than half of total budgetary spending, inclusive of foreign current and capital grants; ZCCM's annualized losses just before privatization is more than a third of total spending. The sale of ZCCM in March 2000 is therefore a significant achievement towards correcting an unsustainable public sector financial situation, but much remains to be done. - Without raising GDP growth to higher than population growth on a sustainable basis, poverty reduction in Zambia would be difficult; the facts indicate that the income growth of the poorest quintile will respond to general economic growth in Zambia by close to one-to-one, which is consistent with the experience of many countries. 1.36 There is a continuing need to keep public expenditures at a non-inflationary level and to allocate and utilize public resources carefully so that the objectives of growth and poverty reduction can begin to be realized. - Zambia's economic outlook has brightened with the sale of ZCCM, recent reforms, improved copper prices and production, HIPC interim debt relief, and good weather. GDP growth may reach near 5 percent for the first time in a long while. However, that expected growth and a lower inflation (of less than 20 percent a year) have yet to be realized and maintained so that fiscal prudence is still definitely necessary. * There are also several downside risks and factors to a higher economic growth: a. Like in 1996, expenditure pressures during an election year in 2001 would likely lead to substantial increases in such category as "other" or in "other constitutional and statutory expenditures". The record for such increases has been poor: Data indicates that small expansionary effect of government spending in the first year would largely be offset by a contractionary effect in the next. 17 b. In the medium term, payments for outstanding obligations from parastatals (US $700m for ZCCM) and other public entities as well as fiscal concessions will continue to constrain resources in the budget and in the economy/banking sector. c. In Zambia, a possible slowdown in world demand and in the price of copper are significant threats to growth and income of the poor, but the factors mentioned above would exacerbate the negative effects. Until the gains of structural reforms and privatization are secured and additional resources are generated, total budget expenditure must be kept where they are now at 30 percent of GDP. With HIPC assistance, there is now a limited but non-trivial marginal budget to improve performance without increasing total expenditure - it should be allocated wisely and executed well. 1.37 Lessons from many countries indicate that without addressing governance, fiscal accountability, and the budget execution problems, additional spending on education, health and infrastructure will not matter. 18 Box 1.1: How to Improve the Efficacy of Public Expenditure and Foreign Aid: Cross-Country Findings and Lessons from Selected World Bank Studies Lessons from recent cross-country findings emphasize good policies, good governance and good institutions as the keys to improving the efficacy of public expenditures. The following are some recent and key findings on infrastructure/investment, education, and health. Public expenditure, investment, and foreign aid: Public expenditures on infrastructure or investment in general, whether domestically or externally financed, are traditionally viewed as the keys to economic growth. Those expenditures are no longer viewed as sufficient however. * Aid does not necessarily finance investment, and investment does not necessarily promote growth; this is true in many non-performing countries in Africa or elsewhere where poor quality of public services, closed trade regimes, financial repression, and macroeconomic mismanagement explain much of the poor economic record. But the combination of good policies, private investment, and foreign aid is quite compelling in generating additional income [see a recent study by Dollar and Easterly (1999)]. * These findings echo recent studies on the effectiveness of foreign aid: In an environment of good policy, foreign aid complements private investment and leads to faster growth and poverty reduction; 1 percent of GDP in assistance translates into I percent decline in poverty and a similar decline in infant mortality. Successful reformers generally have sound macroeconomic policies, structural policies, public sector management, and social inclusion and foreign aid in that setting helped sustained reform. Conversely, aid to countries with bad policy sustained poor policies and the delay of reform. Of the various cases examined in Africa, Zambia is a mixed reformer [see, for example, Dollar and Pritchett (1998) and Devarajan, Dollar and Holmgren (2001)]. * The quality of investment is just as important as its quantity. Without improving the quality of investment with good policies and institutions, private and public capital will not be productive in African countries - in that sense, even current levels of investment may be too high [see, for example, Devaraj an, Easterly and Pack (2000)]. Public expenditure and human capital: While the quality of human capital as reflected by secondary schooling (education) and life expectancy (health) are often found to be significant factors to growth, there however appear to be weak or no evidence that government expenditure on health and education would improve the quality of human capital. The poor outcome of public expenditure is likely due to several factors - inefficient delivery of public spending and the need for appropriate policies and institutions that would improve the payoffs and demand for education and health: * The efficacy of allocated public expenditure on health and education is questionable if governance and lack of accountability are problems [see Ablo and Reinikka (1998)]. When public expenditures are tracked from budget allocation to the delivery of services on the ground in the case of Uganda, less than 30 percent of funding for non-salary public spending actually reached the schools; in the health sector, health care drugs and medical supplies are transferred in-kind without records of their values - as much as 70 percent of the drugs and supplies are diverted for personal gains. 19 Box 1.1.... continue ...... .Public expenditure and human capital: * Even if public expenditure is able to raise schooling, there is also no guarantee that additional schooling will necessarily lead to higher human capital for various reasons [Pritchett (1997)] - (i) marginal returns to education fall rapidly where demand for educated labor is stagnant, for example in many non-performing developing countries; (ii) schooling in that circumstances raises wages but not necessarily actual productivity; and (iii) poor or weak policy and institutional environments mean that human capital accumulated has been applied to counterproductive or socially wasteful activities that serve to reduce growth - such as a bloated bureaucracy or inefficient state enterprises. * Cross-country findings indicate that the impact of public spending on health is quite small and the weak links appear to be weak institutional capacity, the crowding out of private markets for health care, and the lack of economic rationalizations (i.e., the severity of market failures) that limit the impact of public interventions in health. In particular, the emphasis on equity over efficiency criteria have generally led to relatively more public interventions on less expensive curative services with lower impact on health outcomes. [see the studies by Filmer, Hammer and Pritchett (1999, 2000 & 2001)]. Preliminary cross- country findings also indicate that public spending will improve health outcomes, like child mortality, with good governance and higher bureaucratic quality [Swaroop and Razkumar (2001)]. Current vs. capital public expenditure: The uncertain outcome of public expenditure under policy and institutional constraints also extends to another area. The evidence regarding the composition of public expenditure compiled for 43 developing countries over 20 years [Devarajan, Swaroop, and Zou (1996)] shows that - (i) because of the general under-provision of 0 and M relative to capital expenditure in developing countries, an increase in the share of current expenditure (which includes 0 and M) has immediate positive effects on the productivity of existing public capital and on economic growth; (ii) the share devoted to investment or capital expenditure, on the other hand, is negatively associated with growth because of unproductive or excessive capital expenditure. Because donors typically finance capital expenditure, there is strong incentives to underspend in 0 and M. Challenges to Africa: Four crucial fronts are identified for ushering in self-reinforcing processes of economic, political, and social development in Africa [see Gelb, Ali, Dinka, Elbadawi, Soludo, and Tidrick (2000)]: (i) improving governance and resolving conflicts; (ii) investing in people; (iii) increasing competitiveness and diversifying economies; and (iv) reducing aid dependence and strengthening partnerships. 20 2. POVERTY, INEQUALITY AND PUBLIC EXPENDITURE 2.1 Reducing poverty and increasing economic efficiency20 are the two main justifications for public expenditure. While the elimination of poverty is a long term goal, any one year's budget is not capable of achieving it, in which case each expenditure item should be able to demonstrate that it contributes to either poverty reduction or enhancing efficiency (which, itself, will likely reduce poverty). The problem of poverty in Zambia is enormous and complex and cannot be comprehensively discussed in the context of a public expenditure review. However, there are a few key features of the problem that are critical in allocating public resources. 2.2 Essentially, public expenditures can either address the causes of poverty or mitigate its effects. It is necessary to identify the main sources of income of the poor and the main market failures that serve as bottlenecks to the growth of income that public action can influence. It is also necessary, particularly for social sector expenditure and deliberate redistributive transfer programs, to identify the main beneficiaries in order to assess how effectively each program gets money into the hands of the poor. Table 2.1: National Incidence of Poverty by Rural/Urban Classification, Stratum and Province 1991 1993 1996 1998 All Zambia 70 74 69 73 Rural/Urban Rural 88 92 83 83 Urban 47 45 46 56 Stratum Rural small scale farmers 90 93 84 84 -Rural medium scale farmers 79 91 65 72 Rural large scale farmers 62 35 16 Rural non-agricultural households 70 - 72 80 Urban low cost areas 56 50 51 61 Urban low cost areas 56 50 51 61 Urban medium cost areas 43 41 32 50 Urban high cost areas 36 33 24 33 Province Central 70 81 74 77 Copperbelt 61 49 56 65 Eastern 85 91 82 80 Luapula 84 88 79 81 Lusaka 31 39 38 52 Northern 84 86 84 81 North-Western 75 88 80 76 Southern 79 87 76 76 Westem 84 91 84 89 Source: PSI (1991), PSII (1993), Living Conditions Monitoring Survey, 1996 and 1998 20 In PERs, raising economic efficiency (also referred to as allocative or Pareto efficiency) may entail public interventions in market failures, the provision of public goods like roads and defense or the avoidance of private commercial activities by the government. See also Introduction to Part II. 21 Poverty and Inequality in Zambia 2.3 Before examining the effects of expenditures, however, it is important to clarify how information concerning poverty and income distribution can be used to identify beneficiaries of policies. Table 2.1 provides the basic information concerning poverty from 1993 through 1998 according to the Living Condition Monitoring Surveys (see also Map 1 on page 140). As measured by the official poverty line, poverty stood over 70 percent in 1998. Because the rate is so high, changes in poverty rates bear an unusual relation to other measures of income distribution and using it can provide a misleading picture of the relative benefits of policies. In particular, since the poverty line is higher than both average and median incomes, (i) changes in poverty rates reflect primarily changes in urban poverty and do not convey enough information about the very poor or the rural poor; and (ii) it is very plausible that increases in the differences between the relatively poor and the relatively better-off in Zambia, that is, a worsening of the income distribution, would result in a reduction in the poverty rate (see Appendix 2.1). It is therefore necessary to examine both poverty rates as well as the overall distribution of consumption in order to get a more complete picture of the very poor (the next few paragraphs explain further.) Figure 2.1: Cumulative Distribution Function (CD F) of Household per Adult Equivalent Consumption, Zambia 1993,1996, and 1998. All Households 1.0 0.8 Pr 0.6 X op or lix 0.4 1000 10000 100000 Expenditures (1996 Kwa), log _SDA 1993 . . . LCMS 1996 LCMS 1998 Urban Households Rural Households 1.0 1.0 - 0.8 0.81 Pr 0 Pr 0.6 - op ~~~~~~~~~~~~~~~~op or or tiO 0.4 t10 0.4- 0.202- 1000 10000 100000 1000 10000 100000 Expenditures (1996 Kwa), log Expenditures (1996 Kwa), log v-SDA 1993 . . .LCMS 1996 - LCMS 1998 SOA 1993 - - . -LCMS 1996 -LCMS 1998 22 2.4 The stories concerning changes in the full distribution of income and the poverty rates over the period are quite different. Figure 2.1 shows the CDF for urban and rural areas in 1993, 1996 and 1998. Poverty has increased in urban areas over time and decreased slightly in rural areas. However, when the full distribution of income is considered, very little trend is discernable. The lack of change in rural poverty rates is related to the fact that the CDF is very flat - meaning there are very few people - in the area around the poverty line. Most people are substantially below the poverty line. In urban areas, many more people are near the poverty line and so small changes in the distribution result in substantial differences in measured poverty. In neither case is there much of a trend, changes are more directly attributable to various external "shocks". In rural areas, the drought in 1992 being followed by a very good harvest for the 1996 survey and an average harvest for the 1998 survey tracks the position of the CDF quite well - showing no trend. In urban areas, income has fallen in 1998 simply reflecting historically low copper prices. If copper prices are expected to continue to decline, this does imply continuous pressure on urban incomes and poverty. The discrepancies in the conclusions that can be inferred from looking at poverty rates versus the whole distribution of income is one reason for preferring the latter on grounds of completeness. Figure 2.2: Population by Consumption Quintile and Place of Residence, Zambia 1998 2000000* 1800000 1600000 - 1400000- - 1200000- - 1000000- - - ruralI 800000 -- urban| 600000 - - 400000 200000 - 0 I ,, Poorest 2 3 4 Richest 2.5 More important for policy purposes is the relative poverty rates and differences of overall distribution of consumption between different regions, urban/rural residents and categories of earners. These are shown in Tables 2.1 and Figure 2.2. It is clear that the rural areas, particularly small scale farmers are the poorest members of society.2' Figure 2.1 quite dramatically shows the relative deprivation of rural areas across the entire distribution. While the discrepancy between urban and rural areas was less in 1998 21 See also Table 2.3 and Table 2.4 at the end of the chapter. 23 compared to 1993 due to the configuration of weather and copper prices, it was still very large. The reason the relative rates are important is that they reflect the effectiveness of transfers targeted to such groups in reaching the poor. For example, from the fact that 83% of rural and 56% of urban people were poor inl998 (Table 2.1), we can say that for each 100 Kwacha spent in rural areas and shared equally among rural residents, 83 Kwacha will reach poor people. In urban areas, this is 56 Kwacha. A transfer from urban to rural areas, then, represents a net transfer of 27 Kwacha (83 minu 56) to the poor. Figure 2.3: Educational Attainment of 15-19 Year Olds by wealth category 1.2 0.8 -___ 0.8 .. - \Richest 20% 0.6- 0.6 -. .s ---Middle 40% 0.4 0.2 ^. s - " - Poorest 30% 1 2 3 4 5 6 7 8 9 Highest Grade Completed Figure 2.4: Child Mortality (in percent) by Decile of Wealth, Zambia 1997 30 - 25 - 20 - s 15 - 0- 10- 5 0 1 2 3 4 5 6 7 8 9 10 Wealth Decile 24 2.6 Poverty is best understood as multidimensional, involving not only income but deprivation in many aspects of life.22 Figures 2.3 and 2.4 illustrate this point by presenting educational attainment and mortality rates by groups varying by wealth. Poor children do not start school as often as do better-off children and while 98% of the richest 20% complete four years of education, only two-thirds of the poorest 30% do so. Death rates of children under two are more than two and a half times higher in the poorest decile than in the richest. Given the high correlation of poverty with rural residence, these disparities are also evident between urban and rural areas. Public Expenditure and Income Generation of the Poor. 2.7 The sources of income of the poor are consistent with rural residence. Small farmers are far and away the poorest occupational group of society. The vitality of the agricultural sector is therefore critical to the livelihoods of the poor. Attention in public expenditures analysis should be paid to ways to increase farm incomes. This naturally leads to interest in rural infrastructure since this area of investment is likely to be good for the economy (much of which are public, i.e., non-excludable, goods) as well as being for the benefit of the poorest in society. However, there is substantial interaction between farmers and the rest of the economy. Table 2.2 shows the percentage of households that sent and received remittances from family members in rural and urban areas. Over 16% of rural households received remittance income in 1998 (and after the drought year, this was over a third), so it is not easy to separate the fate of the rural sector from the rest of the economy. This reinforces the point made in Figure 1.5 and 1.6 that income growth among the poorest fifth of the population (predominantly rural small holders) is closely related to income growth in the economy as a whole. Table 2.2: Percentage of Households That Received and Sent Remittances (Amounts over 12 Months for Each Year) 1993 1996 1998 Received remittances All 26.1 17.7 16.3 Rural 33.8 18.2 16.3 Urban 12.1 16.9 16.2 Sent remittances All 16.2 22.6 13.0 Rural 13.9 20.5 11.4 Urban 20.3 26.6 16.1 Source: SDA 1993, LCMS 1996, LCMS 1998 22 No single approach can possibly provide a fully satis factory measure of economic well-being or poverty. Alternatives include the basic human needs approach (ILO 1976), which adds people's access to food, clothing, shelter, health care, education etc. , and human development index (UN 1990), which include measures of life expectancy, adult literacy, and real GDP per capita. However, household consumption or income still provides the best available proxy and consistent quantitative measure for standard of living and poverty. 25 Benefit Incidence of Public Expenditure. 2.8 The other connection between public expenditure and poverty is the distribution of benefits of government services. In general it is difficult to attribute to specific beneficiaries the value of public expenditures for a variety of reasons. One reason is that true public goods don't have identifiable "customers" - which is why they are publicly provided in the first place. Those services whose beneficiaries are identifiable are, generally, social services such as education and health because they are publicly provided private goods. Even in these cases, it is difficult to assess the value of the services to the recipient since important quality dimensions of services, the valuation of these services by the people themselves and the indirect impacts (such as levels of education on the overall pattern of wages) are not easily observable. However, simpler questions such as "what are the income levels of children in public schools" or "are relatively poor people using health clinics or agricultural extension services" can be answered and can give some relevant, if incomplete, information concerning the beneficiaries of public services. In this section, information across expenditure items on distributional effects are collected from the household surveys and from information of the location of beneficiaries for comparison. 2.9 Figures 2.5 through 2.12 show patterns of usage of various public facilities by consumption quintile and urban/rural location. Several clear conclusions can be drawn. First, children attending primary school are more likely to be from the poorest groups in Zambia than from better off groups. This is not to say that poor children are more likely to attend school than others - that is not true - but that because poorer families have more children, the student body in basic education is disproportionately poor. Further, as in Figure 2.5, those who are not yet in school are also disproportionately poor. This means that further expenditures in primary education, whether they go to improved quality of those in school or in expanding the numbers in school, will disproportionately benefit the relatively poor. Secondary education shares this pattern, though the effect is not as pronounced as in primary education. Further, as in Figure 2.7, when resources spent per student are included, primary education is somewhat more oriented towards the poor (i.e., costs per student may be slightly higher in relatively poor provinces) while secondary education is somewhat less. 2.10 Beyond the national aggregates, the results also show a distinct difference in the beneficiaries of education expenditures in rural and urban areas. It is clear from Figure 2.6 that money spent in rural areas is dramatically pro-poor and in urban areas quite the opposite. 2.11 It was not possible to accurately assess the distributional impact of higher education subsidies from the LCMS survey for the simple reason that very few people attend such institutions. Proportionate to the number of university students in the population, the LCMS survey had only six households with children attending tertiary education in the stratified sample. All six however lived in Lusaka and were from the richest quintile of the country; all six were also from the richest quintile of people living in Lusaka, the richest region of the country. Nonetheless, the pattern is consistent with the conclusions of other studies. Mwikisa and Lungwangwa (1998) find that only primary 26 education is clearly pro-poor in Zambia. In a sample of 9 African countries (C6te d'Ivoire, Ghana, Guinea, Kenya, Madagascar, Malawi, South Africa, Tanzania and Uganda), 63 percent of tertiary subsidy benefit the richest quintile versus 4 percent to the poorest quintile; in contrast, the poorest quintile benefit as much as the richest quintile with respect to primary subsidy (19 versus 18 percent, respectively).23 The critical issue is not the importance of higher education but the equity impact of higher education and the continuation of public interventions in higher education (see education chapter). 2.12 On health, expenditure patterns are somewhat more ambiguous. It is clear, as is true in virtually every country2 , that lower level curative health services are used more by the poor than are hospital services. Expenditures on hospitals hardly benefit the poorest (Figure 2.9). The reasons are obvious. Hospitals must be in urban areas -those are the only places with population densities high enough to justify such facilities. Local residents make use of hospital services more often than others and urban dwellers are, as is shown above, much better off than rural dwellers. However, the pattern in lower level facilities is not as clear. Figure 2.9 to 2.10 shows that the largest group using public primary facilities tends to be in the middle of the income distribution, not at either end. There are two reasons for this, one policy induced and one simply a fact of life. The fact of life is that it is hard to reach the very poorest with medical care in general. Not only does the public sector have a hard time reaching the poorest, but facilities sponsored by religious institutions do as well (see same figures). Of course, coverage by the private sector, other than those dispensing herbs and other traditional medicines, is also minimal in rural areas. It is difficult to get medical personnel to live in rural areas, apparently more difficult than it is to get teachers out there, since this pattern of use of facilities does not show up in education. 2.13 The second problem, though, is a consequence of policy. Current policy gives exemptions to fees at health facilities to everyone younger than 5 and older than 65. As a result, people of all incomes who qualify use public facilities. Figure 2.10 shows the pattern of usage for children under 5. Clearly, the non-poor choose to go to private clinics rather than public when the exemptions are not in force. Thus, expenditures on health facilities is much more concentrated on the poor when attention is restricted to that age group. However, the overall impact is not nearly as well targeted due to the rules governing exemptions. 2.14 Details of health and education expenditures are left to their respective chapters. One main conclusion from this section is that social sector spending is not necessarily pro-poor. There are particular expenditure items within this category that are, or can be 23 See Table 8 in Castro-Leal, Dayton, Demery and Mehra (1999). The pattern in Africa is also consistent with the distribution of public spending benefits on education among developing countries in general - (i) in the case of primary education, the poorest quintile of the population capture 19 percent of the benefits versus 17 percent by the richest quintile; (ii) by contrast, at the tertiary level 66 percent of the benefits from education spending are realized by the wealthiest quintile versus 3 percent by the poorest quintile; and (iii) hardly any children from the poorest households progress through to university education. See UNESCO statistics or Oxfam (2000). 24 See Filmer et al (2000). 27 made to be, progressive but without closer attention, there is no guarantee that such spending reaches the poor. Another main conclusion is that spending in rural areas is certain to reach poor people to a much greater extent than urban. 2.15 In future, every sector should provide evidence of who the beneficiaries of their expenditures are. This is easiest to do for publicly provided "private" goods, that is, those goods and services with an identifiable recipient such as a school child or a clinic patient. As above, usage can be determined with surveys. For true "public" goods whose beneficiaries cannot be so readily identified (e.g., roads without tolls) expenditures can be categorized by province or region and compared to average income in that region. Conclusions 2.16 Using poverty information to guide expenditure allocation and public intervention should be taken with care: * The national poverty line and rate (at 70 percent) is higher than both average and median incomes, reflecting primarily changes in urban poverty.25 It is necessary to examine also relative poverty rates and differences of overall distribution of consumption between different regions, urban/rural residents and categories of earners in order to get a more complete picture of the very poor. * Poverty has increased in urban areas over time and decreased slightly in rural areas. However, when the full distribution of income is considered, very little trend is discernable. Moreover, most people who live substantially below the poverty line are in the rural areas, which makes public intervention for equity reasons difficult and probably more costly because rural population is widely dispersed. 2.17 Income of the poorest quintile is closely related with growth and would particularly benefit from rural infrastructure and income opportunities in the rural areas. 2.18 The beneficiaries of public expenditure have been very mixed: * Spending in the rural areas and primary education are, not surprisingly, more oriented towards the poor, if they do get there. * The largest group using public primary health facilities tends to be in the middle of the income distribution, not at either end., The beneficiaries of health expenditures are generally more complex and less clear. 2.19 While the PER focuses on the benefit incidence of expenditure, there is also a need to review Zambia's taxation in order to look at who pays for the beneficiaries of expenditure in order to arrive at the net benefit of government programs in the budget. 25 For comparison, population below the international poverty line ($1 a day) - estimate based on 1998 data - is 63.7 percent for Zambia. 28 Figure 2.5: Overall Enrollment Rate of Primary and Secondary Schooling by Consumption Quintile by Urban/Rural Residence, Zambia, 1998 1.0 0.9 0.8- 0.7 0.6 - - -----o--t23 ura 0.5 - 0.4 QUrban 0.3 - 0.2- 0.1 - U---- 0.0 Poorest 2 3 4 Richest Figure 2.6: Percentage of Students, Ages 7-13, by Quintile and Location, Zambia 1998 25- 20 15 _ _ _ _ _ _ _ _ 0 Rural U Urban 10 Poorest 2 3 4 Richest 29 Figure 2.7: Cost per Student by Provincial per Capita Consumption, Zambia 1998 (in logarithms) 12.5- 12.0- 11.5 5 A Primary school 0 5 Secondary school 11. 0 10.5 - A A AA 10.0 10.2 10.4 10.6 10.8 11 11.2 11.4 11.6 Figure 2.8: Distances to Primary School by Quintile for Rural and Urban Areas, Zambia, 1998 4.0 3.5 3.0 3 2.5- 2.0 -- tot- 4ual - urban 1.5 -E A-. 0.5 - -A - - - - A 0.0 Poorest 2 3 4 Richest 30 Figure 2.9: Percentage of Clients of Facilities from Each Consumption Group 100% 80% - _ Richest 60% C] 4 40% -E 20% LX Poorest 0% Religious Public Prirnary Private for Prof it Pubic Flospitals Figure 2.10: Percentage Use of Each Type of Facility by Consumption Quintile 100% _ _ _ 90% _ 80% 70% Public Hospitals 60% - 0 Private for Prof it 50% g C Religious 40% El Public Prirnary 30% _ No Care 20% 10% 0% Pborest 2 3 4 Richest 31 Figure 2.11: Proportion of Visitors (under 5) to Public Health Clinics by Quintiles for Each Province, Zambia 1998 45 - 40- 35 -_ _ _ ___ _ _ _ 30- 15 10 5 0 Poorest 2 3 4 Rchest * Lusaka o CopperbeHt Qj Northern Ca Eastern Figure 2.12: Percentage Users of Public Primary Facilities by Age Group and Quintile, Zambia 1998 35 30 25 _ _ _ _ _ _ _ _ _ _ _ _ _ _ 20 -< 15 -L6 '1 10 5 0 poorest 2 3 4 richest 32 Table 2.3: Incidence of Extreme Poverty by Rural/Urban Classification, Stratum and Province 1 QQ1 1 QQ4 1 QO 14QQR All Zambia 58 61 53 58 RuraLWUrban Rural 81 84 68 71 Urban 32 24 27 36 Stratum Rural small scale fanners 83 84 71 72 Rural medium scale farmers 68 80 50 56 Rural large scale farmers 45 - 15 14 Rural non-agricultural households 58 - 52 67 Urban low cost areas 37 28 31 41 Urban medium cost areas 28 20 16 28 Urban high cost areas 24 19 1 1 19 Province Central 56 71 59 63 Copperbelt 44 28 33 47 Eastem 76 81 70 66 Luapula 73 78 64 69 Lusaka 19 24 22 34 Northem 76 72 69 67 North-Westem 65 76 65 63 Southem 69 76 59 60 Westem 76 84 74 78 Source: PSI (1991), PSII (1993), Living Conditions Monitoring Survey, 1996 and 1998 Table 2.4: Incidence of Moderate Poverty by Rural/Urban Classification, Stratum and Province 1991 1993 1996 1998 All Zambia 12 13 16 15 RuraVUrban Rural 7 9 14 12 Urban 16 21 19 20 Stratum Rural small scale farmers 7 9 14 12 Rural medium scale farmers 11 10 15 16 Rural large scale farmers 17 - 20 2 Rural non-agricultural households 12 - 20 13 Urban low cost areas 18 22 20 20 Urban medium cost areas 15 21 17 22 Urban high cost areas 12 14 13 14 Province Central 14 10 15 14 Copperbelt 17 21 22 18 Eastern 9 10 12 14 Luapula 12 11 15 12 Lusaka 12 15 16 18 Northern 8 15 15 14 North-Western 10 13 16 13 Southern 10 10 17 16 Western 8 8 11 11 Source: PSI (1991), PSII (1993), Living Conditions Monitoring Survey, 1996 and 1998 33 Appendix 2.1: Measures of Poverty vs. Inequality A cumulative distribution function (CDF) shows on the vertical axis the percentage of the population who consume less than the level of consumption on the horizontal. The following figure shows the CDF of per capita consumption for two hypothetical years. Poverty l i- Year 1 Year 2 The poverty rate is the value of this function at the official poverty line. The more "stretched out" the S-shape of the function, the wider the distribution of consumption. The distribution in year 2 is "worse" than in year I since it shows increasing consumption for those above the median and less consumption for those below. As illustrated, this leads to a reduction of the measured poverty rate yet it reflects a worsened situation. Reported poverty rates do not convey enough information concerning changes in the conditions of the very poor relative to others nor of all of the relevant features of the distribution of income needed for policy analysis. 34 3. COMPOSITION OF PUBLIC EXPENDITURE: 3.1 At the heart of a PER is the allocation problem of public resources. Where should government put its money so as to maximize growth and poverty reduction? In Zambia, the corollary to this question is also equally or more important - which sector or activity should the government stay away from so as not to hurt growth and increase poverty? Intersectoral allocation of public expenditure in Zambia is affected by several factors such as - (i) systemic budget contraction caused by the long term economic decline; (ii) two-tier budgeting and coverage issues; (iii) cash rationing under Zambia's cash budget; and (iv) the protection of social spending. These issues and their consequences on the composition of public expenditure are now examined. KEY FEA TURES OF INTERSECTORAL ALLOCATION 3.2 The Impact of General Budgetary Contraction. While there is no exact guidance for the role and size of the public sector, the slow or lack of economic growth in Zambia has added a great deal of budgetary pressure - previous levels of public expenditures and social assistance are simply beyond the means of current public revenue. The general decline in real public expenditure per capita during the nineties was shown in Figure 1.2. One implication has been that sectoral or functional components of public expenditure have also faced the same budgetary and general contraction. Consequently, when viewing public expenditure for each sector or component in isolation, it would be hasty to argue for restoring expenditures relative to standards of some past levels or other countries with a faster growing GDP; reversing the decline of each expenditure component (in real terms) is obviously not affordable. It would also not resolve the allocation problem or relative priorities of public intervention in a tight budget. 3.3 The serious mismatch between public resources and the scope of public interventions in Zambia is rather evident from the persistence of arrears in the operations of the government. In addition to stopping the direct and indirect charges of parastatals and public agencies to the budget, the structure and size of the government inherited from the past still require serious rationalization; there may be too many ministries (27), embassies and duplication of functions. Such choices should be done in the context of a medium-term expenditure framework with high level discussion and decision making. Down sizing and restructuring also requires long-term capacity and institutional building such as the current efforts provided by PSCAP and other related donor-funded activities. The forthcoming Poverty Reduction Strategy Paper (PRSP) provides the occasion for wider discussion regarding the priorities of government and the implications of these priorities to government's role, size and organization, and future public expenditures. 35 3.4 The general budgetary contraction over the years certainly implies that all budget items must be well-justified in terms of basic criteria for public expenditure. While there is no precise prescription for what should be the optimal allocation of public expenditures, the broad principles or criteria for public intervention are well-defined and established. These criteria are:26 1) to correct distinct and large deficiencies of the private sector (efficiency); 2) to improve incomes of the poorest people in Zambian society (equity); and 3) to only make commitments within the capability of government to deliver (implementability). The factors that need to be considered in the intersectoral analysis are just as relevant within each sector as they are between; they are discussed in more detail in the introduction to the sectoral chapters. 3.5 Two-Tier Budgeting, Types of Public Expenditure, and Coverage Issues. Ideally, all types of public services should be included in a public expenditure review. i. One broad definition is expenditure of the entire public sector, including the central government, local government, state-owned enterprises and institutions like the central bank, state universities, pension funds etc. Estimates of the public sector deficit were discussed in the macro chapter. Outside of the central government however, regular and verifiable details of the public sector expenditures are difficult to obtain. Even in the case of the central government, audited public expenditures normally come with a lag of one and a halfto two years. 27 Furthermore, the budget is not decentralized and accounts of local government are not well-kept enough to be included. Regional and poverty related expenditures are covered in the budgets of the line agencies, not much by local governments. Much work needs to be done in the areas of budgetary reporting and capacity building before regional expenditures by program and sectoral activity can be reported with any regularity. ii. Another classification is domestically financed vs. externally financed public expenditure. The domestic budget refers to expenditures financed from domestic revenues raised by the government through its tax and non-tax revenue sources, which averaged 19.7 percent of GDP during 1995-99 (see Table 1.3). Public expenditure financed by foreign grants accounted for 6.2 percent of GDP during 1995-99 and as much as 8 percent in 1999. During budget preparation, estimates of donor financing are incorporated into the budget. Much of the aid- financed projects are however executed outside line ministries, either directly by donors or through NGOs; hence the global pattern of actual donor disbursements and outturn are not systematically verified, consolidated, and accounted for. iii. A further distinction can be made about foreign grants - project grants vs. BOP or budget support grants. Most of the forner are the ones implemented outside the budget. However, budget support or BOP grants do not necessarily buy the same expenditures as those financed domestically. As noted in the macro 26 See Chapter 5 for more discussion of the criteria. 27 However, the 2000 Financial Report has been completed well within one-year. 36 chapter, the Zambian Government insulated the domestic budget from the fluctuations of net external flows and implemented essentially a two-tier budget system - domestic expenditures are covered by domestic revenues while BOP inflows mainly pay for the high external debt requirements (see below for further discussion on the effects on sectoral expenditures). HIPC debt relief should begin to improve resource allocation towards priority areas (see Box 2.1). In fact, HIPC assistance should account for a significant portion of non- wage discretionary spending given that a generally tight budget is expected in the near future. Because of the various data constraints, coverage in the PER is limited to public expenditures reported in the budget of the central government. For the future as work on the development of a MTEF progresses, the comprehensiveness of the coverage of the budget will be increased to public sector entities outside central government. Furthermore, because actual disbursements can vary significantly from budgeted amount, an issue examined below, composition of expenditure is based mainly on audited figures in the Auditor General's financial reports for the period 1991-2000. 3.6 Cash Rationing under Zambia's Cash Budget.28 Of particular importance is the issue and impact of cash budgeting on Zambia's public finance. 3.7 History of cash budgeting. Following an initial period of unsuccessful stabilization efforts in the early nineties, the government committed itself to a very strict form of cash budgeting in 1993 - the budget was not only going to be balanced for the year but also on a month-to-month basis. In essence, public expenditure was restricted to the amount of cash in hand, i.e. positive balances held at the central bank with no recourse to deficit financing. The cash budget was remarkably successful in meeting several objectives - (i) it contributed to the fall of high inflation from three-digit to double-digit level, Figure 1. 1; (ii) by removing all discretion over aggregate fiscal stance, it signaled strong commitment to stabilization and reforms to the donor community and to business investors; and (iii) it highlighted and mitigated the effects of lax expenditure controls and allowed aid to flow through within a stable fiscal envelope and eliminated the need for specific expenditure conditionalities. 3.8 Problems and unintended consequences of cash budgeting. Not intended to be more than temporary, Zambia's cash budget has now lasted more than 7 years because of its operational simplicity. Several problems and unintended consequences are however encountered. 3.9 The efficacy of the cash budget in achieving macro stability can be questioned for several reasons: 28 Cash budgeting in public finance or accounting generally mean the basis of accounting (cash, modified accrual, accrual). However, the cash rationing system (also known as cash triage, cash flow management and sequestration) as implemented in Zambia's budget is now widely called Zambia's cash budget in the literature so the two terms are used interchangeably in the PER. 29 For detail analysis, see Adams and Bevan (2000), World Bank (2000b and 1998). 37 * Despite adherence to the cash budget, inflation, interest rates, and nominal exchange rates have tended to rise significantly after the initial success. * Fiscal discipline is still weak, because the cash budget affects only the budget of the central government and the public sector deficit is still large. Arrears in the public sector may eventually necessitate budgetary payments. * As it has evolved, the cash budget applies only to the domestic budget. If the external budget is not self-financing as assumed because donor balance-of- payments support fall short of the high external debt service payment (HIPC not withstanding), then maintenance of the cash budget is not sufficient to stabilize the growth of reserve money arising from the claims on government (i.e. because of bridge loans from the central bank). * Because of the lack of fiscal flexibility (e.g. to change the fiscal balance temporarily), the targeting of the supply of base money through the cash budget makes it difficult to offset the BOP impact of external shocks and to target or maintain a nominal exchange rate simultaneously. During adverse external shocks (e.g. in 1995, 1997-98, 2000 for varying causes such as copper price and production problems, aid shortfall, oil price shocks, etc.), the private sector anticipates devaluation during the run down of reserves and moves out of domestic assets into foreign assets, exacerbating the effects of the original shocks. 3.10 Just as important, there are several unintended and negative consequences on public expenditures: * Expenditure volatility - The cash budget does not have the flexibility to smooth expenditure, which increases expenditure volatility and hampers the efficient delivery of public services. When cash limits are breached in one period, the non-interest part of domestic discretionary expenditure has to be reduced in the next period, creating wide fluctuations in cash releases to line agencies. * In particular, within-year expenditure volatility (e.g. on a quarterly basis) is not explained by revenue volatility alone. There is substantial discretionary or arbitrary reallocations.30 * Shifts in the composition of expenditures - There is a lack of transparency in the actual allocation of public expenditure and the setting of public sector strategic priorities. Every year, budget estimates are inflated because line agencies anticipate significant reduction to the resources released from the cash budget. In essence, the mechanism for allocating public expenditure lacks a sound basis for discussion and debate. Budget commitments lack their traditional significance because of the frequency of cash shortfalls. Since cash reductions are not prorated over all types of 30 See, for example, World Bank (2000b). 38 expenditure, it also has caused unplanned discretion and shifts in the expenditure. In particular, within-year reallocations are not neutral with regard to different types of institutions: cash releases to the economic institutions are generally reduced in favor of non-economic institutions, such as Zambia Police, Defense, State House, Cabinet Office, National Assembly, Office of the President, etc. Among poverty-related expenditures, social expenditures are more protected than others (see below for analysis). * As a convenient but blunt substitute for expenditure controls and management, it prevents further development of budgetary institutions and process in a no-win circular situation. * Inefficient program execution - In general, the cash budget has led to inefficient execution of planned programs and activities becuse funding is done when cash is available and not when resources are required. The unplanned shifts in expenditures are significant allocation issues and they are further examined below, and the institutional aspects in the chapter on budget management. 3.11 Protection of Social Sectors. To protect social expenditures from budgetary contraction and wide swings in the nineties, the GRZ started to maintain a targeted budget share, which was supported by IDA adjustment credits and other donors. As a result, Zambia has made considerable steps to increase and subsequently maintain the shares of education, health, social welfare, and water and sanitation and other social expenditure in the domestically financed discretionary budget. The combined share increased from 22 percent in 1991 to about 35-36 percent during 1996-99 (see Table 3.1). The consequences of this policy on expenditure composition are further examined below. Table 3.1: Zambia: Share of Social Expenditures, 1996-991 (percent of domestically financed discretionary spending) 1996 1997 1998 1999 Education and training 18.0 18.3 17.4 18.5 Health 13.0 12.9 13.1 13.7 Social security and welfare 1.3 1.6 1.1 1.2 Water and sanitation 3.1 1.6 2.4 2.0 Other social expenditure 0.3 0.1 0.2 1.1 Total social expenditure 35.8 34.4 34.2 36.4 Discretionary expenditure in percent of total expenditure 57.0 64.3 49.6 52.9 1 Share in domestically financed discretionary government expenditure. Social sector expenditure is defined as current and capital expenditure on health, education, social safety net, water and sanitation, and disaster relief. Domestically financed discretionary expenditure is defined as total expenditure, less foreign-funded expenditure, domestic interest payments, the allocation for arrears clearance, the civil service wage adjustments, the contingency reserve, civil service retrenchment costs, payments to the Public Service Pension Fund, net lending to ZCCM, and court awards made against the government. In 1999 discretionary expenditure amounted to about 53 percent of total government expenditure. Source: MOFED. 39 Composition Of Public Expenditure In The '90's 3.12 General Pattern. Based on audited actual spending, several patterns are notable in the composition of expenditures in the 90s (see Figure 3.1). First, there is a substantial increase in the share of social expenditures after 1993. Social expenditures are made up primarily of health and education but include community development and training as well. Some of this increase is simply an accounting change. A substantial share of health and education expenditures had been included in the President's provincial allocations before 1993. These were then re-categorized into their respective line agency uses. The motivation for this change was both Govemment policy and IDA loan requirement as discussed previously. However, even with the change in accounting convention, that agreement was met and the share of the budget going towards the social sectors increased even if their share of the President's provincial allocations is taken into account. Since 1994, there have been further increases in social expenditures. 3.13 Also notable in the pattern (Figure 3.1) is the increase in debt repayments, both external and internal, as a share of the budget from 1991 to 1995 but a distinct reduction since then. This reflects the rescheduling of debt and conversion of commercial to lower- interest IDA/IMF debt (rights accumulation program) in that period and part of the quid pro quo for the increase of social expenditures in those agreements. 3.14 Apart from social expenditures, two categories - "Other" and "Internal security, defense and foreign affairs" - generally maintained or increased their shares of actual expenditures. "Other" includes all other expenditures, including other constitutional and statutory expenditures and political institutions (such as the Office of the President - State House, National Assembly, and Cabinet Office). Apart from 1993-95, the share of "Other" has tended to be maintained or increased. The share of "Internal security, defense and foreign affairs" (including Zambia police, home affairs, judiciary etc.) has also generally increased since 1993 and particularly in more recent years 1998-2000. 3.15 The election of 1996 stands out clearly in the figure as a large increase in the category "other" in that year. The increase is due entirely to a spike in "other constitutional and statutory expenditures" for that year, undoubtedly related to the election. This accounted for fully a quarter of the total budget and squeezed all other shares temporarily. The following years returned somewhat to the previous pattern with somewhat less debt repayment and somewhat more expenditures in social sectors, security-related areas, as well as "Other". 3.16 The last important point to note is the steady reduction in "infrastructure" spending that accompanied the increase in social, security, and other spending. "Infrastructure" is defined as spending in agriculture, energy and water, transport, works and supply, and all items classified as loans for both the Ministry of Finance and local government. The squeeze in infrastructure spending was the greatest during 1995-98. The long-term reduction in infrastructure spending is partially a deliberate consequence of accommodating the increase in social, security and other spending, but may be a partly unintentional consequence of the cash budgeting system. There is a slight resurgence of infrastructure expenditure in 1999-2000 attributable to a conscious policy to begin 40 reversing its downward trend. However, its share in total expenditure is still low relative to other categories of expenditure. Figure 3.1: Composition of Audited Expenditure 1991-2000 100.0 _ 90.0 80.0 70.0- 60.0- 10.0~ ~ ~~5 50.0- 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 U|1 Other E3 Internal security, defense, foreign affairs m8 External debt interest payments Cl Internal debt interest payments O Infrastructure El President's provincial allocations El Social expenditure 3.17 To analyze the pattern of expenditure further, three important aspects or factors are examined: (i) the impact of cash budgeting on expenditure allocation; (ii) unaccounted variations on the outturn of donor spending; and (iii) the effects of both on the deviations of actual from budgeted expenditures. 41 3.18 Impact of Cash Budgeting on Expenditure Allocation. The cash budgeting system, as described in the previous section, was introduced on the budgeting process to help restore macro-economic stability. In this it was quite successful and the avoidance of inflationary pressures was a significant accomplishment. However, the impact on the budget was not neutral. That is, the aggregate requirement to spend only what had been received as revenues had consequences on the composition of public expenditure and not just on its total value. Shortfalls of expenditure relative to budgeted amounts were not shared equally across budget categories or ministries. A statistical analysis of the patterns of the shortfall in expenditure by sector shows several systematic biases (see below). Under-spending was 1) smallest in wages 2) largest in capital expenditure, and 3) smallest in the social sectors, political and security institutions. These effects are all easily understood. During the course of the fiscal year, it is not generally possible to reduce personal emoluments - wage scales are set and employment levels are not easily adjusted within the budget year. The easiest item to reduce is to "postpone" capital expenditure. Given the obligation to maintain and increase the share to social sectors, reductions during the year in these sectors were likely to be treated more cautiously and, in fact, were relatively protected. It should be noted that in 1999 and 2000, measures have been taken to increase the level of actual domestically financed capital spending in the budget. 3.19 The reduction in capital expenditure, particularly infrastructure spending, was not only a result of budgeted allocations but was exacerbated by the biases inherent in the cash-budgeting system. There was no deliberate targeting of infrastructure during the fiscal year. All ministries complain about the cash budgeting system and almost all ministries had, on average, lower disbursements than budget (notable exceptions being political and security institutions as well as social spending), but the largest shortfalls were among capital or infrastructure-intensive sectors as well as other economic spending. It is simply that investments generally are easier to take from on a short-term basis and those ministries with larger shares of their budget devoted to investment suffered larger shortfalls. Over the course of several years the cumulative effects are substantial. Continual "postponement" leads to long-term secular decline (see below for further analysis). 3.20 Unaccounted Variations in the Outturn of Donor Spending. Audited expenditures, such as those in Figure 3.1, refer primarily to domestically financed expenditures. It is possible that the government budget does not adequately reflect total real investment. It is possible that donor funds that do not go through the ministry of finance may make up a substantial fraction of infrastructure investments and that these funds have propped up total investment over the period. While this possibility cannot be completely ruled out, several facts make this unlikely. The reason the possibility cannot be ruled out is due to the absence of accounting for donor funds other than those that go through the Ministry of Finance. In the course of the public expenditure review, an attempt was made to construct a consistent series of aid flows by sector and by donor. However, the Ministry of Finance has no data for any donor activity that does not pass through the budget and there is no obligation on the part of the donors to provide this information to government. Indeed, there has been resistance to the suggestion. This lack of reporting makes substantive planning as part of the budget process very difficult as 42 government can only have a picture of the overall pattern of investment in infrastructure by conducting a separate inventory of all donor activities. The donors should be encouraged to regularly report their direct investments. Figure 3.2: Share of Committed Donor Projects in Public Expenditure, 1991- 2000 1 00.0% 2 80.0%< X) 60.0% o) 40.0% c20.0% e 41 eW Al- x _ > 0.0% 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Year F t =~~~ Donor projects in capital expenditures 3.21 From the data available, it is not possible to definitively state if infrastructure has been declining. We have not been able to reconcile the budgets of the government with those of individual donors such that activities of donors included in the budget can be separated from funds directly paid to NGO's or other implementing agencies. Donor spending can account for a substantial amount of total public and capital expenditure. Figure 3.2 indicates that the sum total of committed donor funds averaged about 80 percent of capital expenditures and close to 25 percent of the overall budget. Hence, it is plausible that donor funds may be quite high relative to the infrastructure component of spending. However, it is unlikely that a full accounting will overturn the conclusion that infrastructure spending has been falling over the 1990's for the following reasons: * Partial data indicate that the share of infrastructure within the donors' portfolios has also declined. Following the definitions of "infrastructure" used here, for example, DFID spending fell from near one-third infrastructure in 1991/2 and 1992/3 to 7-10% in 1995/6 and 1996/7. This rose substantially in 1997/8 due to a large fishery project (which may explain the jump in agriculture spending in the 1997 budget) but has subsequently fallen back. Similarly, Finnish and Swedish aid have both shown a slight decline in emphasis on infrastructure over the decade. All donors seem to be increasing their own contributions to social spending, whether in addition to or simply in support of government increases is not clear. For example, the bulk of Japanese ODA inflows were provisions of equipment to the social sectors (health primarily), which accounted for 86 percent in the 90s. More than a third of Italy's assistance went to social sectors, reaching 60 percent in 2000. Institutional capacity building has also grown in importance. DFID, for example, allocated about 30 percent 43 to improving public administration in the 90s. In any case, unless all infrastructure spending has moved from inside the budget to outside, it is very unlikely that foreign assistance has compensated for the decline in infrastructure. * In addition, donor spending can be erratic and lower than the committed amounts due to problems of implementation, lack of counterpart funding, and other donor conditionalities and concerns (see Box 4.3 for discussion of issues of donor coordination). 3.22 This lack of reporting makes it difficult to determine the true cost of government or the overall priorities of the combined public expenditures of GRZ and donors. It answers some questions concerning the "ffungibility" of foreign aid however. At very broad level of economic classification, donors tend to finance capital expenditures relative to wages and recurrent department charges while the opposite pattern is true for GRZ spending (see also analysis of shortfalls of GRZ spending below). Beyond that, it is difficult to draw more specific conclusions. One hypothesis about aid is that there is substantial "fungibility" between spending on budget categories on the part of the government and that the priorities of donors may not be reflected in actual allocations since any increase on their part can be undone by compensatory reduction on the part of the government. In Zambia's case, this effect is limited to some extent by the fact that the government does not know what the donors are doing and therefore cannot be completely compensating for it. 3.23 Shortfalls of Actual GRZ Releases Relative to Budgeted Expenditures. Table 3.2 presents the ratio actual GRZ budget releases from the original budgeted allocation by ministry for the cash rationing period when detailed audited figures are available (1995- 99). These figures do not include donor spending outside the GRZ budget and pertain mainly to domestic expenditures. The following points are easily discerned: * As is apparent, wages are generally protected versus recurrent and capital expenditures. Actual wage expenditures were about 13 percent higher than their original allocation. Capital expenditures have the biggest shortfall as a broad category with a negative deviation of about 60 percent while recurrent expenditures fell short by about 16 percent. * Looking at broad categories of sectors or institutions, allocations to political institutions (such as the Office of the President - State House, National Assembly, and Cabinet Office) internal and external security (including police, defense, home affairs, and judiciary) are protected the most. Of the rest, social expenditures are protected relative to infrastructure and other expenditures. * Every ministry except for political institutions (such as the Office of the President - State House, National Assembly, and Cabinet Office) internal and external security (including police, defense, home affairs, and judiciary), and social spending (such as education and health), as well as others like labor, 44 pension, community development, and information services experienced a shortfall that averaged 38 percent over the period. The hardest hit, with cuts exceeding 50% of the original budget, were loans and investments of local government, energy and water, mines and minerals, and the President's provincial allocations. While infrastructure as a group did not fare well, loans and investments from MoFED as well as works and supply did well because of a significant rise in expenditures in 1998-99. TABLE 3.2: PERCENT RATIOS OF GRZ ACTUAL BUDGET RELEASES OVER INITIAL ALLOCATIONS 1995-99 Percent Classification Ratios By Broad Economic Classifications Wages 113.0% Recurrent Department Charges 83.5% Capital Expenditures 39.0% By Broad Sectoral/Institutional Categories Political institutions 191.4% Security institutions 178.5% Social expenditures 106.1% Infrastructure 58.9% President's provincial allocations 51.9% Others 87.4% Selected Sector/Ministry Police and Prisons Service Commission 435.7% Cabinet Office 349.3% Minister w/o Portfolio 225.6% Zambia Police 191.7% Office of the President-State House 173.3% Defense 164.6% National Assembly 154.3% Legal Affairs 147.8% Home Affairs 141.6% Pensions and Gratuity 139.2% Loans and Investments - MoFED 132.9% Commerce, Trade and Industry 118.0% Works and Supply 117.1% Education 114.2% Foreign Affairs 114.1% Labor and Social Security 103.6% Community Development and Social Services 103.6% Information and Broadcasting Services 101.7% Health 100.6% Local Government and Housing 97.4% Communications and Transport 91.2% 45 Agriculture, Food and Fisheries 73.0% Finance and Economic Development 72.9% Science, Technology, and Vocational Training 70.3% Sport, Youth and Child Development 68.8% Environment and Natural Resources 66.9% Lands 63.9% Office of the Auditor General 58.7% Judiciary 57.2% Tourism 55.8% President's provincial allocations 49.1% Mines and Minerals Development 45.5% Energy and Water Development 29.7% Loans and Investments - Local Government 10.2% Source: Ministry of Finance and Economic Development, Financial Reports (Blue Books). 3.24 To see if there is a systematic bias against particular ministries, a multiple regression analysis was undertaken to explain the degree of shortfall from budgeted amounts throughout the 1990's. The hypothesis is that it is easier to postpone investments than operating expenditures when shortfalls occur while wages and other operational expenditures are difficult to alter in mid-year. This is due to two effects. First, contracts cannot be written when continuous funding is not guaranteed. Second, operational expenditures seem more urgent in the course of the year and the pool of money waiting for investment decisions to be made is easily raided. For both of these variables, the effect of the imposition of cash budgeting after 1994 was examined as well. Finally, the specific ministry's allocations are examined for specific biases. 3.25 The specific explanatory variables for this analysis are the following: Capital share of ministry budget Wage share of ministry budget Variables to represent specific ministries (experiments with different combinations) Variables to represent provincial allocations Effects that vary for each year A set of variables exploring the interaction of the cash budgeting years with the type of expenditure (capital, wage, ministry groups). 3.26 The results are presented in the appendix to this chapter. The equation presented is the best fit of several alternative specifications. The main conclusions (robust to all specifications) are the following: * Everything else being equal, the eventual cuts or bias against any specific ministry or sector is strongly accounted for by the investment share. That is, a higher capital share in a ministry's initial allocation leads to dramatic reduction in the amount eventually released to the ministry. Beyond the broad pattern, the results can of course vary agency by agency as shown in 46 Table 3.2. Ministries such as transport, agriculture, tourism, energy and water, loans and investments of local government etc. are generally hurt. * Conversely, a higher initial share of current expenditure, such as those in the political and security institutions as well as service sectors and general administration protects against reductions or, in some cases, to greater amount eventually released.3' * Among non-political and non-security institutions, expenditures in the social sectors are relatively well protected from shortfalls in addition to their status as low investment sectors. Provincial allocations are in addition more vulnerable than other categories. All of the above effects are exacerbated by the introduction of cash budgeting though only the effect of the capital share is significantly stronger in the post 1994 or cash rationing period. 3.27 These results indicate that not only has there been a systematic reduction in investments over the 1990's but that much of this reduction is ad hoc and determined by shortfalls of budget allocations during the year, worsened by reallocations to non- economic sectors and erratic donor releases. While all sectors complain about their inability to plan on the basis of initial budget figures, this effect has systematically hurt the infrastructure sector in terms of its actual share to total domestic expenditure (Figure 3.1) and in terms of shortfall to its original budget allocation as a group (Table 3.2). Conclusions 3.28 The general budgetary contraction over the years requires that all budget items must be well-justified in terms of basic criteria for public expenditure. These criteria are: 1) to correct distinct and large deficiencies of the private sector; 2) to improve incomes of the poorest people in Zambia; and 3) to only make commitments within the capability of government to deliver. 3.29 The cash budget, while making an important contribution to the control of inflation, has introduced systematic biases in the allocation of public resources. In particular, infrastructure investments of all types have fallen specially in the first seven years of the 90s due to the inability to make long-term, relatively large commitments. While all ministries complain about the shortfall of actual to budgeted resources, the impact has been larger for infrastructure investments than it has been for service sectors. Political and security institutions have generally benefited from within-year reallocations. Of the rest, social expenditures have relatively been protected by the Government and by donor funding. 3.30 Expenditure pressures and risks during an election year are likely to be reflected in substantial increases in the category "other" or in "other constitutional and statutory expenditures". 31Since capital share and current share adds up to 10 0 percent of total expenditure. 47 3.31 The true cost of government and where resources are actually spent are not fully identified. Much of the actual donor spending is outside the budget and not accounted for. To obtain a better picture of public expenditure, there is a need for an annual reporting mechanism for the actual expenditures made, not just estimates at the stage of budget preparation. Likewise, accounts of public entities other than the central government, such as local government and state-owned enterprises, needs to be consolidated and verified to obtain good estimation of where resources are allocated in the whole govemment or public sector, as well the financial status and deficit of its operations. 48 Box 3. 1. Public Expenditure Supported by HIPC Interim Debt Relief, 2001 (In billions of kwacha) Budget HIPC Percent of Budget with without HIPC allocation total HIPC HIPC allocation allocation Rural Development 35.3 124.4 35.3 159.7 Education 10.4 67.0 19.1 77.4 Health 43.7 50.3 14.3 94.0 Water/sanitation 21.0 33.5 9.5 54.5 HIV/AIDS 0.0 31.4 8.9 31.4 Safetv nets 7.7 30.2 8.6 37.9 Human rights 0.0 10.0 2.8 10.0 Lowcosthousing 1.5 5.1 1.5 6.6 Total 119.8 351.8 100.0 471.6 In December 2000, Zambia reached its decision point under the enhanced HIPC Initiative. While PRSP priorities and targets were still being developed, the initial allocation of HIPC resources was drawn based on consultations with the line agencies, scaling up of existing programs where funds could be disbursed relatively quickly, the initial suggestions from IDA and IMF staffs, and preliminary lessons from the PER. The list was subsequently debated and modified by the Parliament and incorporated into the 2001 budget. Measures for the tracking of HIPC expenditures were also adopted (See Annex III and Box 4.2). Programs to be supported by HIPC resources are focused on poverty reduction and social activities. A substantial proportion (over 35 percent) is targeted at rural livelihood and development programs - such as better access to markets (feeder roads), improved productivity (irrigation and electrification), control of animal diseases, and a targeted rural farmer support program intended to reduce dependency on food aid by improving rural subsistence activities of some 200,000 poor farmers. Social sector programs also have a rural focus, e.g., rural water supply, rural schools, malaria programs, and rural housing. In the health sector, the focus is to increase support for programs addressing HIV/AIDS, malaria, drugs, and under-funded public hospitals. In the education sector, increased educational materials and improvements to rural teachers' housing will complement the existing Basic Education Sub-Sector Investment Program (BESSIP). Resources will also be allocated to vocational skills programs, to water supply and sanitation in rural and peri-urban areas, improved housing and support for informal sector activities and the urban poor, and social safety nets for vulnerable groups, such as the disabled, street children, women, and the rural and urban poor. 49 Appendix 3.1: Regression Analysis of Actual Budget Releases as a Fraction of Initial Allocations32 Dependent variable = Actual budget release to ministry/ Yellow Book allocation Variable Coefficient Standard error Capital share of allocation -1.158 .277 Wage share of allocation .277 .354 Interaction: cash budget with capital -.977 .378 share Interaction: cash budget with wage .249 .562 share Social Sectors .331 .158 Provincial allocations -.453 .137 1990 .711 .320 1991 .266 .317 1992 .178 .309 1993 .290 .306 1994 .135 .163 1995 .141 .162 1996 -.209 .165 Constant .465 .241 Adjusted R2 .549 32 At the time of statistical analysis, only the data for 1990-97 were available. 50 4. MANAGEMENT CONSTRAINTS TO PUBLIC EXPENDITURE 4.1 Institutional issues and weaknesses in the overall and sectoral public expenditure management can affect the government ability to carry out the broad objectives of growth and poverty reduction. It is not sufficient to tilt the composition of expenditure towards health and education in order to achieve better outcomes as emphasized in the previous chapters. In addition, several building blocks are also necessary to improve the delivery of public services: (1) the development of effective budget management system so that public spending will achieve the intended objectives and resources will actually reach the intended beneficiaries; (2) strengthening the monitoring and evaluation of public sector performance so that progress can be determined and enhanced; and (3) reforming the incentives in the public sector so that performance and efficiency are rewarded. These are practical issues that are critical in determining outcomes. 4.2 Zambia's problem appears to be basic - the lack of effective budgetary framework and management institutions. The introduction of a cash rationing system to replace the inherited central planning in the early nineties has not been followed by the development of effective budgetary institutions. The problems of Zambia's budget and financial management system, including the cash budget, are now very well documented.33 Box 4.1 presents some preliminary findings regarding the delivery of public services in selected health and education facilities and how they are affected by weak budget execution and the cash rationing system. Several efforts are now underway by the Government, supported strongly by various donors, to develop the appropriate medium-term expenditure framework and budget management system, together with the development of an integrated financial management information system (IFMIS) and activity based budgeting (ABB) as well as tracking or monitoring measures for poverty reducing public expenditure funded by HIPC assistance or the budget in general. This chapter discusses some of these institutional issues and measures currently being developed or debated. 4.3 The focus of the PER is on expenditure management and monitoring, the first two building blocks. How to improve the delivery of public services is also increasingly important at the sectoral and local levels. Box 4.3 examines sectoral programs in Zambia and the issues of donor coordination and donor-executed projects. Box 4.4 anticipates the benefits and issues of fiscal decentralization for a future study. In both areas, the capacity to manage, execute and monitor expenditure in a transparent way is critical. The equally important third component on administrative efficiency and performance, covering, among other subjects, civil service and pay reform, procurement, human resource development and management are partly covered by PSCAP, but will also need to be examined in a systematic way in the future. 33 See for example, World Bank (1998 & 2000b); IMF (2001 & 2000); Myers (2000); Hughes (2001); Adams and Bevan (2000) 51 Box 4.1: Some Initial Survey Findings Regarding the Delivery of Public Services in Selected Education and Health Facilities Following the increase in the national incidence of poverty increased from 69% in 1996 to 73% in 1998, the Bank commissioned a one-time survey of selected facilities of education and health with the following objective - to obtain a first reading on how the cash rationing system and the budgetary system in general has affected the availability of public funds and the execution of public services at the selected facilities, the quality of public facilities, access and constraints to access, and utilization of these facilities by income groups. The plan is to pursue and follow this up with more formal and systematic public expenditure tracking surveys (PETS) as a subsequent activity of the PER In this first effort, the survey consisted of structured interviews of District Health and Education Officers, Primary or Basic School Heads and Officers in Charge of Health Centers in eighteen districts, two in each of the nine provinces. The major findings indicate the interplay of budgetary execution problems under cash rationing, lack of income and local conditions or circumstances in affecting the delivery, quality and utilization of public services particularly in the rural areas. Impact of budget execution and cash rationing. Despite the allocation of at least 34 percent of the total discretionary budget to the social sectors, beginning under the 1996 Second Economic and Social Adjustment Credit (ESACH) conditionality, there seems to be a wide disconnect between cash releases and what funds finally get to education/health facilities. Personnel emoluments and related expenditures are consistently under-budgeted in which case budgeted expenditure on key supplies and services (drugs, school desks, primary schools recurrent expenditures, etc.) are normally varied and under-funded. For example, while MOFED cash releases improved significantly during 1999, survey data showed that primary schools and health centers continued to be poorly funded (i.e. in terms of essential supplies such as drug kits, chalk, text books, etc.). Generally, primary schools do not receive any essential supplies from the government (i.e., district education offices). These are normally purchased from user fee proceeds (PTA and School Fund). Some rural schools are greatly inconvenienced because of long distances from urban centers where such supplies can be sourced. Another example - the procurement of textbooks by the Ministry Headquarters is done without either consulting district officials or primary schools. For example, under a donor funded Basic Education Project, a lot of draft textbooks were supplied in 1998 to primary schools. However, the final 1999 editions were, in most cases, markedly different from the drafts (of the same books) supplied earlier. More fundamentally, neither edition matched the existing syllabi. Another problem of inconsistent supply of textbooks was noted: in most schools visited, English books were over supplied compared to other subjects; within the same district, some schools were better supplied than others. 52 Box 4.1... continue Utilization of services was affected by many other factors, particularly in the rural population: Enrollment levels in rural schools, compared to urban ones, are confirmed to be much lower for varied reasons. Children of school going age normally register at the beginning of the year but withdraw in the course of the year either due to the failure by their parents to pay user fees or for other economic reasons such as working as laborers (in 1996, Ministry of Education data show that these two factors accounted for about 80% of primary school drop-outs). Other children are forced to withdraw on account of their parents not appreciating the value of education or famine or lack of water or engaging in other informal sector activities such as curios (Western province) or fishing (Luapula province). Examination class children (grade VII and IX) are forced to withdraw on account of their parent's failure to pay exam fees. According to some district examinations officers, examination grades pupils could not be prevented from writing an exam on account of non failure to pay the exam fees. Pupils who do not pay examination fees could write exams but the Examinations Council of Zambia would withhold results for such pupils until they settle the exam fees. It was evident that there was a communication breakdown as most primary school heads had not formally been informed of this position and hence their continued enforcement of barring pupils who do not pay exam fees from writing final examinations. Rural facilities lack permanent accommodation for teachers/health workers. Most trained personnel refuse to work in rural areas on account of the hostile environment. Rural teachers take on abnormal loads which should attract an allowance but such allowances, at the time, were not being paid Consequently, education quality indicators such as pupil-teacher ratios, progression rates, enrollment ratios, etc. are generally better in urban than rural areas. On a comparative basis, primary health facilities and services seemed much better than basic education. However, the problems pertaining to education in terms of remoteness, administration, funding etc. are equally important to the health sector. The number and frequency of rural drug kits drastically declined between 1998 and 1999. Like teachers, rural health workers also devised innovative ways (i.e., barter system) to reach more people in their catchment areas to reduce the adverse impact of introduction of user charges. Before introduction of the barter system for user fees, attendance at health centers was confined to those groups which were exempt from user fees: under 5 and over 60 years old, maternity cases, and those with STDs. With the introduction of a barter system, the other groups have increased their utilization of the facilities. Sparseness of population, poor infrastructure and distance from health centers were confirmed to be major barriers to accessing the services in rural areas. 53 Current Situation 4.4 Medium Term Prioritization and PRSP. The strategic allocation of resources to growth and poverty reduction (or any other activity) must begin with a Medium Term Expenditure Framework (MTEF). The MTEF unites policy, planning and budgeting. It translates broad policies laid out in a national vision statement into more specific financial and budget priorities. Ideally, that national vision and prioritization should be brought about by Zambia's Poverty Reduction Strategy Paper (PRSP) currently being developed. While the broad based participation at the early stage of development of the PRSP is very positive, its strategic prioritization plus a financially viable medium term expenditure framework (that needs to be derived from it) would require top-level political support and discussions. Such political engagement must also carry through to the annual budget process. This is necessary so that the PRSP and its annual or medium-term expenditure prioritization are not merely technical exercises but reflect societal policy choices. 4.5 A fully functioning MTEF may be complex and may take years to develop, but the basic institutional mechanism for budgetary choices and priorities to be debated and made can certainly be implemented with high level commitment and along with the PRSP. The goal is, then, to eventually replace the current budget framework anchored in the cash budget with a basic MTEF that improves the realism of the annual budget allocation and process. 4.6 Several issues with regard to budget preparation, execution, reporting and tracking, audit and sanctions, and other areas have also been identified and are summarized briefly next. 4.7 Budget Preparation: Under-budgeting of "mandatory" expenditures. At the heart of the problem with budget preparation is the continued over-commitment of the government. Social assistance and infrastructures that were developed during more prosperous years can no longer be supported. Resources allocated to them do not reflect the extent of the obligations that exist. A review of the structure of the government for duplication of functions and other problems is necessary. * The budget is confined to central government: local government expenditures, state-owned enterprises, and those activities of extra budgetary funds, financed by non-central government grants, are not covered in the budget documents. * Capital and current expenditure planned by donors are included in the budget but outturns of donor expenditure are sketchy and a significant amount of planned expenditures do not materialize. * The present budget classification is limited to administrative and economic dimensions; a functional classification is created but not widely used; there is no program classification, although work is underway to 54 develop this as part of an Activity Based Budgeting (ABB) initiative supported by donors. * HIPC financed spending is now separately identified in the 2001 budget, but in general all poverty reducing expenditures are not. As part of Bank conditionality, there is monitoring of the proportion of domestically financed expenditures devoted to the social sectors. * The fact that the annual budget process is not anchored in an MTEF or any multiyear projections is a key problem. 4.8 Budget Execution: * The pattern of expenditure outturns deviate substantially from budget allocation as analyzed in chapter 3. Budget outturns vary significantly from the budget figures partly because of under-budgeting and cash rationing, requiring substantial in-year reallocations. This is true in terms of administrative and economic categories as well as donor financed expenditures (for unmet donor conditionality as well as changing donor goals or objectives). There has been very little budget headroom in the domestic budget, beyond wages and fixed costs, to conduct significant poverty reducing activities. Beneficiaries of expenditures have been mixed (see chapter 2). The HIPC assistance (at close to 3 percent of GDP, 10 percent of total expenditure or 15 percent of domestic expenditure) would provide marginal resources for poverty related expenditures beginning in 2001. - Month to month expenditure volatility has been high and has affected implementation of poverty oriented programs. Plans are being made for expenditure smoothing on a quarterly basis and within the present cash budgeting system. * New expenditure arrears still accumulate over a stock of outstanding central government payment arrears. * Zambia's internal audit includes pre-audit, systems audit and inspections work; effectiveness is hampered by resource constraints, over concentration on pre-audit and lack of follow-up to the findings from investigations. * Expenditure tracking surveys are planned on a pilot basis to supplement internal control, but have not yet begun. * Fiscal and banking reconciliation is still problematic because of extra budgetary spending. 55 Box 4.2 Tracking of HIPC Expenditures - Some Trade-offs There is a tremendous interest from donors that one-oft HIPC resources are not wasted and are actually allocated to poverty-reducing public expenditures (see Box 3.1). As a result, there are incentives and pressures for governments to develop a system for tracking the HIPC-supported expenditures, such as the one outlined in Annex III for Zambia. Another strong donor desire is that public expenditures funded by the HIPC Initiative will not be inconsistent with the objectives, policy and activities of the various sector programs. While it is not the intention, there are risks that the restrictions or conditionalities of HIPC debt relief will separate HIPC-supported expenditures from the regular budget in operation, thus weakening the budgetary system. Moreover, it is all poverty-reducing spending, not just HIPC's, that need to be tracked in order to: (i) ensure the additionality of spending; (ii) tilt the overall composition of spending; and (iii) strengthen the overall public expenditure management system. When there are issues of financial management and budget reforms however, the balance between the short-term gains of special tracking and the long-term development of overall expenditure management is difficult to resolve (see similar issues about donor-executed projects in Box 4.3) 4.9 Budget Reporting and Tracking: e Audited accounts are not completed and presented to the legislature within 12 months of the year end. The delays can reach one and a half to two years for various reasons - (i) the late completion of accounts at line ministry level; (ii) problems in consolidation of ministry data at the Data Computing Center; and (iii) capacity problems at the Auditor General's office in terms of the necessary human and financial resources. - In general, the information management system is deficient for supervision and control or for end-year reporting. - There is no functional classification of expenditures in the in-year budget reports. * There are no expenditure data or they are not consolidated by regional/district by program. This makes it difficult to see if expenditures are reaching the poorest regions and areas. * Outtums of donor expenditures are not compiled and reported as noted above and in chapter 3. * Internal budget reports from the ministries are now generally received within two or three weeks after month-end. 4.10 Auditing and Sanctions. The effectiveness of internal audit and external audit, vital to discourage the misallocation or misappropriation of funds, has been reduced due to lack of corrective follow-up action after identification of weaknesses. 56 * Problems in auditing include: lack of equipment, poor funding, inadequate human resources, lack of training, the need for development of standards and procedures, and the lack of follow-up actions or sanctions. * It has been established that even when the Office of Auditor General does produce major findings of fraud or mismanagement, there is a lack of follow-up. There is seemingly ambiguity as to the one who is responsible for the follow-up and whether sanctions/punishments could be exercised. 4.11 Personal Emoluments and Human Resource Management. * General problems of turnover in the civil service have left many ministries wanting for staff in key areas. * Weak payroll control and information make it difficult to accurately account for where people are truly working, to do unit costing of activities and services and to evaluate any regional inequities. * As a result, transparency is hindered by the fact that detailed accounting for personnel is not available in an audited format of the Financial Reports (Blue Books). e The Public Sector Reform Program (PSRP) is yet to address the most chronic problems with the civil service - . shortage of skilled personnel. Recommended Measures 4.12 Several measures are being pursued by the Government with strong participation and support by donors. Table 4.1 tabulates these efforts as they are currently known, which have a definite element of capacity building and technical assistance in many areas. 4.13 Medium to Long Term Reforms * Introduction of a basic MTEF, which should be anchored in the strategic prioritizations of PRSP to link policy, public sector priortity setting and the budget. * Development of IFMIS to provide consistent financial information for expenditure management and control. * Introduction of activity based budgeting and better budgeting and reporting techniques so that budgetary information by programs, activities and regional/district levels can be generated for planning and monitoring. * Implementation of public expenditure tracking surveys (PETS) on a regular basis so that public sector performance and the flow of resources to facility and ground levels can be monitored effectively. 57 * Explore and plan for an inter-governmental fiscal framework (decentralization), linking to local or community based participation and social funds such as ZAMSIF, in order to enhance the execution of public expenditures and their impact on the poor [see Box 4.4]. 4.14 Short to Medium Term Reforms nImprove the realism of budget allocation using recent past releases as starting points and avoid significant in-year reallocations. * Improve expenditure smoothing and replace the current rationing system with well-defined rules and cash management system that are based on realistic quarterly projections. * Replace the protection of social spending with the following possible options: (i) protect all poverty-oriented spending (including infrastructure); (ii) maximum ceilings on non-poverty-oriented expenditure; or (iii) proportionate adjustments when there is a cash shortage. * Improve expenditure control and prevent new arrears by requiring timely and accurate reporting of monthly expenditure returns, enforcing compliance of commitment controls by Controlling Officers, and giving them high-level attention and monitoring. * Improve budget reporting and classification by - (i) completing the pilot phase of ABB; (ii) identification of poverty reducing priority expenditures in a more systematic basis; and (iii) creation of a revised functional database that incorporates all expenditures. * Implement a reporting mechanism for donor financed projects (outtums) to improve the coverage of expenditure reporting. * For the same reason as above, expand budget comprehensiveness to include extra-budgetary funds, independent agencies, and local governments. * Strengthen internal and external audit capacity. * Pilot PETS quickly (see long term measure for why). 4.15 Final Caveat: Without the highest commitment to good governance and fiduciary accountability, all the budget management measures, information and technology will not bring about better execution and outcomes of public expenditure. 58 Box 4.3 Sectoral Programs and Issues of Donor Coordination and Donor-Executed Projects Sectoral programs. Zambia has a number of innovative Sector hivestment Programs (SIPs) or Sector Wide Approaches (SWAps) that attempt to improve aid effectiveness through less fragmentation of donor support and increased local participation in project preparation and implementation. Sectoral programs that are currently in operation include: agriculture (ASIP), basic education (BESSIP), health (Health SIP or SWAp), road (ROADSIP) etc. Several potential benefits were greatly anticipated from these programs: * Better policy: Because the sectoral programs are sector-wide in coverage, a shared vision of sectoral priorities was envisaged to eventually lead to better policy and a cohesive framework for implementing the combined resources of Government and donors. * Improved execution and delivery. While it was recognized early or at the preparation stage that institutional issues such as financial management, procurement and capacity of line ministries and local governments were still weak, it was also felt that these problems could be overcome and capacity could be developed as the programs are implemented. By replacing numerous separate projects in the public expenditure program with one single operation, improvements in the effectiveness of public expenditure program and in the delivery of public services were projected. * Greater ownership and capacity building: Because local stakeholders are involved from the program inception and are supposed to be in the "driver seat", improvements in ownership, capacity and sustainability were expected through - (i) programs supported by all donors and no separate donor investments and expenditures made outside programs; and (ii) minimal use of long-term expatriate technical assistance. * Larger pool of resources and harmonization of procedures: The sectoral programs and their potential benefits were anticipated to lead to more donor financing and greater harmonization of donors' implementation procedures such as procurement, reporting, accounting, and auditing. While considerable efforts have gone into SIPs to bring about a coordinated and integrated approach in sectoral programs, the experience in Zambia indicates both positive signs as well as realities that are still below the original expectations: * Sectoral programs have brought about greater coordination within the implementing agency and more open communication with the donors. This is true, for example in BESSIP and elsewhere. * Innovative and flexiblc financing approaches have been introduced. Problems remain however. In ROADSIP for example, putting emphasis on incremental funding from road user charges for maintenance have had less impact than expected - not enough money, not disbursed or allocated efficiently enough. * Sectoral policies are evolving, but they still need to be defined better in critical areas in roads, health or agriculture (e.g. fertilizer marketing). 59 . .. Box 4.3 Donor Coordination Issues . Cross-cutting or intersectoral issues are harder to resolve: relative priorities and amount of sectoral expenditures in the government budget; issues of decentralization, government capacity, fiscal controls, financial and procurement management could all benefit from a less sectorally fragmented or a more economy-wide concerted efforts. The formulation of the PRSP and its implementation is a step towards a more comprehensive approach for effective service delivery for poverty reduction. * Issues of financial management, transparency, and expenditure wastage are serious in many areas such as the procurement of drugs and pharmaceuticals, agriculture input supplies (e.g. fertilizers) etc. The original assumption that better financial management and procurement could be attained during implementation has proven to be more difficult to realize. * Some amount of decentralization and participatory structures have strengthened capacity at the district level to manage health services. However, the lack of an overall framework on decentralization and the lack of decentralized fiscal and financial management capacity have limited more improvements in various areas. * Capacity is sometime inadequate at line ministries to manage and coordinate the tasks of some sectoral programs; and Govemment's role in donor coordination and in providing leadership has been less than expected. * To improve execution, the tendency of donors to execute their own projects or force the pace of preparation and implementation has intensified in some areas, thereby undermining ownership, capacity building, and sustainability. As a result, not all donor projects are dovetailed into the sectoral programs. * Not all government-funded projects are also dovetailed into sectoral programs. * There are often inadequate funding of the government share of the local costs as a result of the budget squeeze and cash rationing. * Conflicts and tension between old structures (line ministries) and new (e.g. Roads Board) have meant economies that could have been gained from coordinated programming have not always been realized. * Also dependence on outside TA has not been reduced, very high proportion of external funding still goes to fund short term extemal "experts". This is true, for example in ROADSIP. Donor-executed projects- short- and long-term trade offs: . Donor-executed programs are usually aimed at improving expenditure outcomes and avoiding expenditure wastage and irregularities in the short term. * The trade-offs are the long-term development of institutional capacity and many of the potential benefits of sectoral programs mentioned above. . In particular, the very objective of better delivery of pubic services from better coordination and management is constrained. • To the extent that actual donor spending are not reported and differ from plans, the full coverage of public expenditure is also not known. 60 Box 4.4 Fiscal Decentralization, Local Participation, and Social Funds Some Findings and Issues from Case Studies The attraction of decentralization, local participation (e.g. community based development), and social investment funds lie in their promise to improving governance and public service delivery. It is argued that decentralization would improve - (i) allocative efficiency because of better matching of public expenditures/services to local demands and preferences; and (ii) productive efficiency because of better accountability of local governments to citizens and better knowledge of local costs. However, recent case studies (Philippines and Uganda) and analysis of social investment funds (SIPs) indicate that a key constraint is the capacity and institutional constraints of local governnents. In Zambia, as the Government prepares a policy on decentralization with devolution of service delivery to district government accountable to local people, issues of fiscal decentralization and the capacity of local governments will need to be examined carefully as follow-up activities to the PER and PSCAP. Below are some findings and factors from recent studies that are relevant to future efforts to improve public service delivery with fiscal decentralization [see, for example, World Bank (2001c) and Jack (2001)]: * Local governments have weak capacity that may limit the effectiveness of decentralized service delivery (e.g. Uganda). Like other developing countries, Zambia has weak financial management; an adequate mechanism for decentralizing the budget and extending expenditure control/accountability along program activities and districts/regions must accompany the granting of authority to local levels. * Local governments have limited authority to adjust resource allocation because funding and grants come from the center with significant restrictions due to weak financial management at the local levels. This is true in Philippines and Uganda. * Limited information may limit citizen influence at the local level and media coverage of local events and projects are very inadequate. To ability to post budget releases at the district facility level is an important feature in Uganda. * In the case of Uganda and Philippines, the threat of migration (voting with the feet) does not appear to discipline local governments and improve public service delivery. * The emphasis on micro-level participation can inhibit the delivery of public goods with widespread benefits or inter-jurisdictional spillovers - immunization and communicable disease control as well as economic infrastructure on a scale that is required for far-reaching or extensive generation of income opportunities and poverty reduction. * Local projects also tend to be in social sectors, where outputs are difficult to measure and effective incentives for better delivery are sometimes harder to design. * In the case of SIFs, there are possible trade offs between better service delivery to targeted groups in the short-term and some long-term issues: (i) capacity substitution - although it is not their aim, SIFs may have impeded the development of and substitute for local government agencies; and (ii) lack of cost recovery and sustainability - while this is expected because SIFs are redistributive programs to target groups, it is more acute because the primary sources of funds are often donors. On the other hand, SIFs, like the Zambia Social Investment Fund (ZAMSIF), are often the only effective sources of resources flowing into local communities, given budgetary problems and institutional constraints of line agencies at the local level. 61 Table 4.1: Action Plan For Public Expenditure Management Reform Major Objectives Priority Possible donor/TA Ranking provider (H, M, L) Improve functional classification of the budget, i.e., adopt functional classification to allow for gradual extension of the H IDA, IMF, EU,DFID number of ministries piloting activity-based budgeting; identification of HIPC expenditures and other priority pro-poor programs u Produce/document formal guidelines and procedures on budget formulation (i.e., revise the Financial Regulations) M IDA,IMF < Improve budget comprehensiveness by including extra-budgetary funds, grant-aided institutions and local authorities. H IDA,IMF,EU Periodic Government led public expenditure reviews. M IDA,DFID c :s Involve Parliament/political process/wider stakeholders consultation in budget design M IDA,IMF,USAID,DFID o Accelerate implementation public sector reforms to make Government operations cost-effective and affordable within M IDA,DFID,UNDP - existing hard budget constraint v Explore the possibility of Parliament presentation of budget from end of January to end September of the preceding L IDA,IMF fiscal year to obtain formal warrant by end of December to allow on time cash allocations for first quarter to line ministries Link policy priorities to budget estimates through activity based budgeting (ABB), integrated financial and H IDA,IMF,EU,USAID, E management information system (IFMIS) and a medium term expenditure framework (MTEF) (to replace cash budget Sida C e system) S Establish a sustainable inter-governmental fiscal framework in order to enhance the delivery, efficiency, equity, and H IDA,IMF,EU coverage of public goods, especially in health, education, and water supply to the poor34 Move from discretion to rules based budget execution to improve fiscal transparency and accountability i.e., full H IDA,IMF implementation of the new quarterly cash allocation system with a minimum of 80% of Yellow Book estimates c released on a regular and timely monthly basis (Stick to Yellow Book estimates and minimize virement and enhance predictability of resource flows to spending units). 5 Y Establish a properly staffed unit within the Budget Office to implement a more effective cash management system" H IDA,IMF E Integrate cash management system with commitment control system to ensure that the cash plans guide the H IDA,IMF determination of quarterly projections and monthly cash releases ; Improve the system of commitment controls to avoid the build-up of new arrears. H IDA,IMF o Accountant-General's Office to assume joint responsibility of effecting and monitoring actual payments as opposed to H GRZ Initiative V Budget Office alone 34 A necessary condition for this is the long-awaited decentralization policy statement. 35 Proposed MOFED restructuring envisages transforming the Budget Office into a Treasury to manage cash more optimally, i.e., release cash according to programs needs and not when it is available. 62 Major Objectives Priority Identified donor/TA Ranking provider (if known (H, M, L) already) Implement mid-year budget reviews. H IDA |Improve reconciliation of fiscal and banking accounts M IDA,IMF c Review structure of government accounts at BOZ: consolidate accounts into Treasury Single Account (TSA) to M IMF, IDA o to capture all inflows and outflows; close all dormant accounts; discontinuing use of "X" accounts 8 Improve banking arrangements with commercial banks: negotiate banking arrangements centrally to achieve greater M IMF, IDA economy and efficiency; rationalize and reduce number of commercial bank accounts; and such accounts be monitored by the Accountant General (AG) o Implement an Integrated Financial Management Information System (IFMIS) H IDA,IMF,Norway,EU, DFID Ensure Controlling Officers (COs) submit complete and accurate data on monthly expenditure and arrears by the H IMF,IDA,EU prescribed date. Enforce sanctions for those who do not observe this rule X Enforce compliance with commitment controls: COs to ensure that commitments are consistent with monthly cash H IMF,IDA a releases, to avoid accumulation of arrears Move towards cash flow statements as prescribed by GFS and IPSAS, as well as the Budget/Actual Outturn M IMF ELi E Comparative Statement . Strengthen institutional capacities of the Auditor- and Accountant-General's Offices to provide timely financial M Norway, IDA, EU o o reports as mandated constitutionally X Encourage co-operating partners to provide comprehensive and timely information on their disbursements to the H IDA Treasury in order to capture such data in consolidated Government accounts Provide detailed information on public expenditures and costs of providing public services H IDA,IMF,DFID Review the legal framework to enhance oversight/sanctions powers of Parliament on budget management issues M IDA,USAID,IMF, c 0 _ Norway,DFID _ Implement an Integrated Financial Management Information System (IFMIS) H IDA,EU,IMF,Norway Ensure flow of monies into HIPC sub-account 49 at BOZ H IMF,IDA 0 c Develop HlPC/other poverty reducing expenditures classification for new PRSP programs HIMF,IDA U Initiate public expenditure tracking surveys (PETS) and then conduct them periodically H DFID,IDA C.) ~ Implement further improvements to the present manual expenditure control regime H IMF Reform and augment internal audit capacity 63 PART II: SECTOR SPECIFIC ISSUES OF PUBLIC EXPENDITURE 5. INTRODUCTION Criteria for Sector Analysis and Intersectoral Comparisons 5.1 The most difficult choices to answer concerning public expenditures are those that require making judgments across sectoral boundaries. How much is a school worth relative to a road that reaches a remote area? Which benefits the poor more: rural electrification or better health care? While these problems may seem unsolvable, the fact is that the choices need to be made. Each ministry's proposals need to be compared head- to-head with the others' in order to determine appropriate priorities within the overall budget required by fiscal balance. Given the enormous needs of the country due to its poverty, it is inevitable that important expenditures will not be possible. How best to make these difficult choices? The principle used in this report is based on the need to get the most benefit for the amount of public resources put into the economy. This, in turn, requires judgments on three dimensions: efficiency, equity and implementability. 5.2 The criteria of efficiency implies finding those areas of spending which make the greatest impact relative to the counterfactual: what would happen if the government did not spend the money to undertake this activity? The answers to this question range from: actual improvement or, at worst, very little change in the case of commercial activities to quite severe consequences such as the complete absence of an essential good or service. The standard way to approach this problem in the economics literature is to identify and, to the extent possible, quantify the welfare cost of a market failure that prevents the private market from adequately supplying the good or service. 5.3 The most extreme case of a market failure is the existence of a pure public good. The characteristics of these goods are that they are non-excludable, meaning that people cannot be prevented from benefiting from them if they don't pay for it, and non-rival, meaning that one person using it does not prevent anyone else from doing so. These are items for which there cannot, even in principle, be a private market because no private business would be able to charge for them. There are very few such goods but include systems of justice, national defense (from external attack) and several specific programs such as malaria control via swamp drainage. 5.4 Other common market failures are externalities, where the production or use of a good either benefits (or harms) people other than the buyer or seller of the item. Pollution is an example of a harmful externality, proper sanitation facilities (in urban areas at least) or immunization against disease are examples of beneficial externalities. These can range from cases where the effect on others is a large proportion of the value to the individual 64 (in the extreme, a public good) to cases where the effect is quite small (and possibly working only through envy or offense to personal taste which may not be considered legitimate areas for correction by public authorities anyway). 5.5 Monopolistic or otherwise non-competitive markets are another failure that provides a justification of public intervention. The usual reason for there to be a monopoly is the existence of substantial economies of scale relative to market demand. Examples would be electricity distribution via grids (not necessarily electricity production) or the infrastructure necessary to distribute piped water (not the water itself). Frequently a monopoly is observed only because of public policy that limits entry of competitors and therefore is not a market failure at all. 5.6 Many activities of government are clearly suspect on the basis of this criterion. A large proportion of parastatal enterprises do not have much justification on these grounds and it is an unambiguous accomplishment that they have been subject to privatization efforts in recent years. Note that this criterion of identifying market failures does not mention the "importance" of the activity or product to the economy. Most of the economy will be based on private goods, all of which are "important" but none of which require public support. For example, copper production, unquestionably the largest single export earner for the country and therefore of vital "importance" need have no direct involvement by government in its operation. It is a completely private good operating in a competitive international market and can be expected to run most efficiently when managers are primarily concemed with the profitability of the operation. Government activity would be limited to monitoring and regulating environmental impacts of the industry and labor standards. 5.7 Since the issues involving most of the parastatals are relatively simple, they will not be included in this review. Discussion will be reserved for activities that have at least some claim on public resources. 5.8 Even if a market failure is observed (and it can be argued that no market is perfect) the question facing those making the public budget is: why spend money to fix this problem? Why not find some other (cheaper) way to do it besides providing the service directly? And even if the solution' must be public provision, why not charge full (or full marginal) cost to the consumer and not require spending scarce public resources at all? 5.9 In Zambia, there are several major areas of public spending where this issue is relevant, that is, where there are some legitimate fears of the ability of the private sector (at least currently, not necessarily in principle) to handle some critical service but which need not absorb substantial public money. Subsidies on things as diverse as fertilizer, water, electricity and university education fall into this category. While it may be imprudent to eliminate public provision of these services, at least suddenly, there is substantial scope for reducing the net cost to government by charging appropriate prices, in some cases, full marginal costs. An advantage of charging for these products is that competitive prices may encourage private entry if the private sector can provide at lower costs. This will help determine whether public provision is necessary without having to 65 make that judgment based only on current information. Maintaining heavy subsidies ensures that private participation will be minimal since it is impossible to compete against free or unprofitably low prices unrelated to costs. 5.10 The second major consideration for public expenditure is to promote equity. Private market mechanisms cannot guarantee that the resulting situation is fair or accords with social norms of justice. A legitimate role of government is to correct such imbalances and improve the distribution of the benefits of the economy. The previous sections have gone into some detail in comparing the effect of public spending on various income groups and do not need to be rehearsed here. The important point to keep in mind is that poor people are in need of many things and each of the ministries can correctly claim that poor people do not have access to their particular services, especially if prices are charged. What they cannot all claim is that subsidies to their particular service will disproportionately help the poor relative to all other alternative uses of the money. Since resources in aggregate are limited and the extent of poverty in the country so wide, it is not possible to provide all of the needs of the poor via public programs and decisions to subsidize particular items have to be based on evidence that it is those particular items that are most effective in reaching the poor. In the Zambian context, these will almost certainly be those that are directed toward rural areas, particularly the relatively remote areas. 5.11 The third criterion on which to compare the claims of different spending categories is more prosaic than the previous two. This is the degree of implementability of the proposed program, or, simply, can the government actually do what it promises? Expenditure programs differ considerably in terms of their complexity of administration, their proneness to abuse or corruption, their reliance on difficult monitoring mechanisms or on exceptionally dedicated behavior on the part of civil servants. It is important from the point of view of maintaining credible government as well as simply avoiding wasting money that programs are compared on the grounds of whether they can actually be implemented with current human and administrative resources or, if such resources can realistically be gained through training or reform of incentives. 5.12 While the administrative difficulties of various programs are quite specific to the individual sectors, a few generalizations may be possible. First, we might expect systematic difficulties in some types of programs to accomplish the equity goal, requiring some difficult trade-offs. Services that we might like to be able to provide to poor people are labor intensive and intensive in skilled labor at that. Health and education are good examples. Medical personnel and teachers are well educated relative to the large majority of Zambians and can expect to earn relatively high wages. They are also likely to have been raised for substantial portions of their lives in urban areas and to want urban amenities for themselves and their families. These are normal, predictable preferences of qualified professionals. However, for the sake of equity, we might prefer that services paid for by public funds be directed toward rural areas, the more remote, the higher the priority. It is very difficult to achieve both. Either much higher pay must be paid to such 66 providers to induce locating in poor areas36 or much more control than seems plausible to expect must be exercised by administrators. 5.13 Second, in the ability to administer different programs may require alternative delivery techniques than "best practice" would dictate in the absence of the implementability constraint. For example, while comprehensive primary health care in rural areas may be an ideal to strive for, difficulties in implementing that strategy may require alternatives such as dedicated immunization programs done by visiting teams rather than by resident staff. 5.14 Perhaps most important, given how much is unknown about the differing degrees of difficulty of administering programs, as well as their actual efficiency and equity effects, is to find out what works. Policy makers should be focused on the actual impacts of public policy. Impact evaluation of policies should be taken seriously with more regular measurement of outcomes of core public interest. Incomes and consumption levels, enrollment rates, health indicators and the availability and use of public facilities and services can be monitored regularly and related to policy efforts. Successful experiments can be duplicated, unsuccessful ones can be halted before they absorb too much money and those with mixed results can be studied to determine what makes the difference between success and failure. 5.15 For substantial parts of the budget, at least under current governmental arrangements, expenditure priorities will still need to be made centrally. The remainder of this report summarizes the claims for several ministries' spending priorities and tries to contrast them in terms of their efficiency, equity and implementability implications. 36 In Indonesia, research has indicated that the increase in pay needed to induce sufficient number of doctors to relocate was outside of the range of any feasible level . See K. Chomitz et. Al. (1998) 67 6. EDUCATION Introduction 6.1 As in the case of health, the government has a clear role in education due to several characteristic market failures as well as to its impact on the poor. Education can be seen as both a current transfer to the poor as well as a means for increasing future earning capacity. 6.2 The most important market failure is the external benefit associated with basic literacy and numeracy from primary education. While explicit attempts to measure this externality are few, it is widely believed that the benefits of a literate and educated population extend beyond the benefits to the individual. The few attempts to measure the externality in the developing world have focused primarily on agriculture. These have generally found that farmers benefit from the education of their neighbors (Appleton and Balihuta for Uganda) via the ability to copy adoption of farming techniques though some have suggested (Foster and Rosenzweig) that this effect occurs only during periods of rapid technological change. The estimated effects are sometimes quite large, almost as large as the direct effect of education to the individual. This is enough to justify large subsidies. 6.3 Additionally, education, especially female education, has important non-market benefits in the form of lower fertility and better child's health and nutrition. Evidence suggests that girls' education is more effective at improving health than health care itself. 6.4 Further, at primary level at least, a subsidy to education is often very progressive, partly because poorer people tend to have more children than others. This pattern, as described in chapter 2 (poverty) holds for Zambia, particularly if funds are directed toward rural areas. As shown in Table 6.1, school participation rates among children are positively correlated with the economic welfare of the household, increasing from only 52% in the poorest quintile to 85% in the richest quintile.37 However, the poorer quintiles tend to have more children than the richer quintiles. Therefore, even though there are lower participa-tion rates in the lower quintiles, children from these quintiles still account for a substantial share of enrollment. Table 6.1 shows that children from the bottom two quintiles (the bottom 40% of households) account for 44% of enrollnent, even though more than a third of the children in these two bottom quintiles are not in school. One implication is that public expenditure on primary education can reach a large number of children from poor households. 3' The quintiles are defined in terms of household expenditure per adult equivalent. 68 Table 6.1: School Enrollment of Children Aged 7-13 in 1998, by Quintile Percent Percent Distribution Participation of Enrolled Children Richest quintile 85.1% 16.1% Second 78.7% 19.4% Third 73.8% 20.5% Fourth 67.0% 21.6% Poorest quintile 52.4% 22.4% All five quintiles 70.1% 100.0% Source: Analysis of LCMS 1998 6.5 A second market failure associated with education is the general failure of credit markets for human capital investments due to the inability to provide collateral in terns of purchased assets. It is difficult to borrow against future earning ability, a particular problem for more expensive levels of education, preventing an investment that could benefit the individual greatly. While there are few discernable externalities associated with enrollments in higher education, the inadequacy of private credit markets raises a public concern. However, since the problem is with credit and not education per se, the appropriate intervention is not a subsidy to tuition (let alone living allowances) but help with obtaining credit for paying full marginal costs of the education. Tuition subsidies are an extremely indirect and inefficient means of addressing the underlying problem. Recent complaints and protests at UNZA over fee increases seem to be directed more at the requirement of single payments in full at the beginning of the year rather than at the principle of payment in general. Phasing in of full cost charges, flexible payment terms and loan programs (with full repayment expected) should be considered and could well be politically feasible. 6.6 A third market failure is a possible externality associated with basic research. Here again, the more direct intervention would be to pay for research outputs since that is what generates the externality. If researchers need student assistants, they can pass some of these funds on to them. But tuition subsidies, again, are extremely indirect and inefficient means for achieving this objective. 6.7 Subsidies and grants to higher education, on the other hand, are not pro-poor. Proportionate to the number of university students in the population, the LCMS survey had only six households with children attending tertiary education in the stratified sample. All six however lived in Lusaka and were from the richest quintile of the country; all six were also from the richest quintile of people living in Lusaka, the richest region of the country. While the number of cases are too few to generalize from the survey alone, another study, Mwikisa and Lungwangwa (1998) confirrned that public expenditures in Zambia beyond the primary level are not pro-poor. 6.8 Implementation issues appear to be a significant problem. The budget share of education sector has increased since the mid-90's, although the real value of Government 69 expenditure on education remains low by international standards, or in comparison with the Zambian situation of the mid-1980s. Yet enrollments have stagnated, and the efforts undertaken have not reversed long-termn decline - some 77 percent (net enrolment) of children of primary-school age attended school in 1980, but the figure has stayed below 70 percent for several years in recent times. Quality of primary education is falling too - a recent survey found that 75 percent of children now leave primary school illiterate. Partly as a result of perceptions that quality has fallen, and that education no longer guarantees a wage employment, school enrolments have declined even where facilities are within reach. Since increases in enrollments have been recorded in a number of schools assisted by the Social Recovery Program, the stagnation of aggregate enrollment implies that there must have been enrollment declines in a number of other schools. Some of the problem may be attributed to staff losses due to AIDS, or the impact of rising rates of orphanhood and poverty on households' capacity to send children to school. In any case, the data pose questions concerning the effectiveness with which public money can be transformed into real services and more children being taught. 6.9 To address the serious problems of primary education in more concerted way, the Government and donors have recently launched the Basic Education Sub-Sector Investment Program (BESSIP). BESSIPS differs from all previous educational interventions in that it is a coherent and comprehensive program that seeks to address major issues in policy, management, school infrastructure, education materials, school health and nutrition, teacher development and deployment, and equity issues. With the Ministry of Education taking the lead, it involves the government, donors, private sector, local communities and NGOs (see Box 6.1). 70 Box 6.1: Zambia's Basic Education Sub-Sector Investment Program (BESSIP) BESSIP was appraised in September 1998. The objectives of BESSIP are to accelerate enrollment growth to 4% annually, and to improve learning achievement in basic education as measured by periodic National Assessments. BESSIP includes all activities of the Ministry of Education with respect to basic education. As such, it embraces ongoing donor-assisted "projects" in the area of basic education, in addition to new financing commitments. The program covers recurrent costs of primary education (teachers' salaries, education materials, etc.); primary school infrastructure; primary teacher training; curriculum development; and administration of the education system, including capacity building. Developments Thus Far. New financing commitments in support of BESSIP became available from mid-1999 onward. Some external agencies have committed funds to a "Pool" where funds from various partners are mingled in a common bank account, and are available for any BESSIP component, while other agencies have committed funds under more restrictive arrangements. The quality of reporting and the annual work program and budget put forward have improved very substantially, and there is great openness with which issues are discussed. Internal coordination within the Ministry of Education is also very much improved, as compared with pre-BESSIP experience. Disbursements of external aid for BESSIP purposes, which were US$20 million in 1999, reached $28 million in the year 2000, and $8.7 million during the first quarter of 2001 (implying an annual rate of expenditure of $35 million). Performance has been uneven among components and areas: * Good achievement in infrastructure, with 593 classrooms completed in 2000; * the procurement and distribution of education materials is still slow; * the Basic School Curriculum Framework is now developed and finalized; * teacher deployment for rural areas remains a major challenge; problems include late placement of newly qualified teachers and poor conditions of service. * the Ministry's performance in spending external aid is on an improving trend; * the binding constraint on progress is however the Ministry's implementation capacity rather than the magnitude of funds made available by donors - a Medium Term Financial Framework (MTFF) for education still needs to be developed; the Planning Unit of MOE remains understaffed; * decentralization and the creation of education boards have moved slowly; * monitoring is hampered by lack of basic data such as teacher numbers; * coordination between the Ministry of Education and Ministry of Finance and Economic Development could be improved - the budget share targeted in the program and what is actually allocated are inconsistent; data on the GRZ budget for the forthcoming fiscal year usually becomes available only after the consideration by the November/December Semi-Annual Review of the proposed Annual Work Program and Budget (which therefore focuses largely on the external financing). 71 Zambia's Education Indicators 6.10 Educational Attainment of the Adult Population. The Report of the Living Conditions Monitoring Survey (LCMS)1996 found that 38% of Zambian adults had educational attainment of Grade 4 or less (considered insufficient to sustain literacy); 35% had highest educational attainment of Grades 5-7; and 28% had some education beyond primary school. The Demographic and Health Survey 1996 found that the average schooling of Zambians in their twenties (7.8 years for males, 7.2 years for females) was not significantly different from the average schooling of Zambians in their thirties, reflecting stagnation in schooling attainment in recent years. 6.11 Enrollment in Grades 1-7. There are about 1.6 million children attending primary school. LCMS 1998 found the following net enrollment ratios38 among children of primary school age: urban males, 80%; urban females, 80%; rural males, 61%; rural females, 60%. Thus, at the primary level, urban/rural disparity is a much more important problem than gender disparity. School participation rates are positively correlated with the economic welfare of the household, increasing from only 52% in the poorest quintile to 85% in the richest quintile. However, households in the poorer quintiles tend to have more children than the richer quintiles. Therefore, even though there are lower participation rates in the lower quintiles, children from these quintiles still account for a substantial share of enrollment. One implication is that public expenditure on primary education can reach a large number of children from poor households. 6.12 It is important to understand to what extent non-enrollment reflects dropout or, altematively, failure to enter the system at Grade 1. Figure 6.1 illustrates Grade-specific Enrollment Rates (GSERs) for Grades 1 (90%), 4 (80%), and 7 (66%).39 Enrollment decreases by 10 percentage points from Grade 1 to Grade 4, and by 14 percentage points from Grade 4 to Grade 7. The implication of the 66% enrollment ratio for Grade 7 is that 34% of Zambian children never make it as far as Grade 7. About three-tenths of these 34% never entered school at all, while seven-tenths of them dropped out before reaching Grade 7. Provincial student flow profiles reveal that dropout, especially for girls, is greater than the national average in the five provinces which are away from "the line of rail". 38 The most commonly used enrollment indicators are the Gross Enrollment Ratio (GER) and the Net Enrollment Ratio (NER). In both cases, the denominator is the school-age population, which in Zambia is ages 7-13 for Grades 1-7. With the GER, the numerator is the total enrollment in Grades I- 7, while with the NER the numerator is the enrollment of children ages 7 through 13 who are enrolled in school. The advantage of the GER is that it provides a measure of the "capacity" of the system in relation to the school-age population. The advantage of the NER is that I minus the NER gives the percentage of the school-age population who should be in school but are not. 39 The Grade-Specific Enrollment Ratio (GSER) is a number which is calculated for each grade (and not for the primary cycle as a whole). It is the number of non-repeating students enrolled in that grade, divided by the number of children of the age that should be attending that grade, according to a country's enrollment policy. The series of GSERs for a cycle is called the "student flow profile". 72 Figure 6.1: Student Flow Profile in Grade 1-7 Education, 1998 100% 90% 80% 70% 60% 50% Grade 1 Grade 4 Grade 7 6.13 Learning Achievement at the Primary (middle basic) Level. The first-ever National Assessment of Learning Achievement was carried out in 1999, at the Grade 5 level. Average pupil performance (32% correct in English, 33% correct in mathematics) was not much greater than the 25% performnance that would be obtained by answering items randomly. In the three provinces where skills in Zambian languages were tested, performance in these languages was better than in English, but still low. The Assessment authors concluded the system was yielding "pupils who can scarcely read or compute". On a more positive note, teachers were found to be generally well trained, with the potential to be more effective if given the necessary support, supplies, and motivation. Among the steps recommended by the authors to address the situation are the following: (a) replicate nationwide the successful pilot program in initial literacy in Zambian languages; (b)"flood schools with the necessary textbooks and learning resources"; (c) assign only trained teachers teach to the lower grades; and (d) extend the duration of the teaching day in Grades 1-4 beyond the current 3.5 hours. 6.14 Enrollment at Levels Beyond Grade 7. In 1998, there were about 245,000 students in secondary school, distributed as follows: about 80,000 students in each of Grades 8 and 9, and about 30,000 students in each of Grades 10 through 12. As shown in the Figure 6.2, there is a considerable gender gap in secondary education, particularly for Grade 7. About 17,500 students were enrolled in formnal post-secondary education in 1998, distributed roughly as follows: the two universities, 6000; Teacher Training Colleges, 5600; vocational/technical, 5900. 73 Figure 6.2: Profile of Schooling from Grade 1 through Grade 12, 1998 100% 80% 60% - Total 40% - i- Boys 40% - ~~~~~~~~~~~~~Girl s 20% - 0% : 1 3 5 7 9 11 Grade 6.15 In addition, the following observations can be made: - Provision of secondary education cannot meet the demand from the supply at grade 7 and 9. In the late 1990s the progression rate between grades 7 and 8 was in the range 30-35 percent, where as the same for grade 9 and 10 was less than 25 percent. - Secondary education does not attract adequate Government and donor financing; poor students are therefore left out because of the inability of households to supplement. - Boarding expenses are high and yet boarding schools are located primarily in rural areas, which are predominantly poor and households cannot afford the boarding the boarding expenses. As a consequence, most places at boarding schools tend to students from urban areas. * Secondary education is meant to yield high benefits to an individual. However, because secondary curriculum is not relevant to the rural setting, the graduates of secondary education upon graduation migrate to urban areas. Government Expenditures on Education 6.16 Level and Intrasectoral Distribution of Government Expenditure on Education. Table 6.1 places GRZ's expenditure on education, in the range 2.0-2.5% of GDP. The Government has been devoting about 20% of its budget, after debt service, to education. 6.17 In recent years, domestically-financed expenditure on the various subsectors has been roughly as follows: primary schools, 1.2% of GDP; secondary schools, 0.3% of 74 GDP; the two universities, 0.3% of GDP; teacher training, 0.1% of GDP; administration and central functions, 0.5% of GDP. The external aid which is recorded in the Government's accounts is devoted mainly to primary school infrastructure, teacher training, and capacity building. Table 6.2: Government Expenditure on Education (as a percentage of GDP) Post- Primary Primary Education Education Total Zambia, 1999, excl. external aid 1.2 0.8 2.0 Zambia, 1999, incl. external aid 1.6 0.9 2.5 6.18 Expenditure on Grade 1-7 Education (lower and middle basic education). Governnent expenditure on the recurrent costs of delivering primary education is about 1.2% of GDP; when expenditure (mainly donor-financed) on primary school infrastructure is added, total expenditure on primary education becomes 1.6% of GDP. This is a low level; the World Bank's Long-Term Perspectives Study on Sub-Saharan Africa estimated that low-income African countries would need to spend between 3% and 4% of GDP on primary education in order to achieve universal primary enrollment. 6.19 Government expenditure per primary student was about K54,000 on average in 1999, equivalent to 7% of per capita GDP. (For comparison, Government expenditure per primary student is about 10% of GDP per capita found in SSA anglophone countries generally, and 15% of GDP per capita in the francophone countries.) Due to disparities in the distribution of trained teachers between urban and rural areas, and in teaching loads between Grades 1-4 and Grades 5-7, there are undoubtedly substantial variations around the K 54,000 average. 6.20 One consequence of the low level of GRZ expenditure on primary education is a low level of teachers' compensation. Average primary teacher's compensation is below CSO's poverty line for a household of two adults and four children. Average primary teacher's compensation is only 2.7 times GDP per capita in Zambia, while it is 3.6 times GDP per capita, on average, elsewhere in Africa. Another consequence is low public financing for the non-wage costs of teaching children and operating a school. 6.21 Secondary Schools. Typically in Africa, Government expenditure per secondary student is in the range of two to three times greater than the Government expenditure per primary student. This reflects higher compensation of secondary teachers, less efficient use of techers, and sometimes boarding subsidies. In Zambia, the disparity factor between secondary and primary unit Government expenditure is only 1.7, which is relatively low. However, this needs to be further investigated between: Grades 8 and 9 as extensions of primary schools, versus traditional secondary schools; and boarding versus non-boarding. 75 6.22 Universities. Zambia has two universities, the University of Zambia in Lusaka and the Copperbelt University in Kitwe, which together have an enrollment of about 5000 students. There are also several technical and polytechnics schools. Of Zambians aged 20-24 however, only 2 percent are into further or higher education. The emphasis has been on arts and social sciences, with less than 40 percent of graduates going for technical, medical, agricultural or managerial subjects. 6.23 The critical issue is not the importance of higher education, which can develop the human resource and knowledge foundation to the country's programs in economic management, governance, private sector development, health, agriculture, energy, and urban development. What are increasingly being questioned are the role of the state and the continuation of public interventions inherited from the past. The Government supports public university education through two channels: grants to the two universities; and student bursaries. In 1999, the average university student bursary was K 3.8 million, of which K 2.0 million was for tuition, and K 1.8 million for living expenses and books. Meanwhile, the average primary teacher's salary was K 2.0 million. Because of this generous support, Government expenditure per university student is more than 100 times greater than Government expenditure per primary student. Clearly, a disparity exists and the amount of potential savings that can be applied to other priorities are not inconsequential. Including a supplementary appropriation, for 1999 the authorized Government expenditure for university education was 38 billion kwacha (0.5% of GDP), consisting of 22 billion kwacha of grants to the two universities, and 16 billion kwacha of student bursaries. 6.24 The National Policy on Education (1996) recognizes the need to rectify this situation. The Government is gradually (i) replacing state funding of tertiary education with private-sector sponsorship and income generation schemes by the universities and colleges, (ii) reducing the number of Government-sponsored students, and (iii) allowing more self-sponsored students to enter Universities. There is also the perceived need that institutions for higher education could be reorganized and run more efficiently. However, the process has been politically difficult. The budget squeeze and shift to other priorities such as primary education have caused higher education to be chronically in arrears with periodic social unrest from students and teachers. 6.25 The Government plan to gradually eliminate subsidies should be made within a coherent framework with an institutional mechanism for regular open discussions among stakeholders of the issues and strategies for implementation. Possible specific options, counting some lessons from other countries,40 include: raised tuition fees gradually (Armenia, China, Mongolia and Uganda), starting with new incoming students; reduce aid first to non-instructional expenditures (food, dormitories, transportation); protect poor students by applying means testing to student aid (e.g. Kenya); ensuring the growth of the economy is a must to raise the private returns of higher education but allow for private provision and competition (Kenya, Tanzania, Ghana, Uganda, and Mozambique); expand student loan schemes (Mexico, Chile) and consider income-contingent loan systems, sometimes referred to as graduate tax, in which loan repayments are a fixed 40 See, for example, World Bank (2001 c). 76 proportion of a graduate's annual income (Ghana, South Africa, Sweden, Australia, New Zealand) as well limit loan programs to a fixed share of students to control fiscal exposure and to promote academic competition and performance (Northern Mexico, Costa Rica, Dominican Republic); replace traditional approach of "negotiated" budgets with alternative mechanisms such as incremental or matching resources tied to self- generated resources and linking funding to institutional performance such as the number of graduates (Ethiopia, South Africa); competitive fund to promote quality improvements and innovations (Argentina, Chile, Indonesia, and Romania); consider hiring private companies to manage the universities under well-defined and enabling arrangements etc. 6.26 Institutional Aspects: Planning and Budgeting. A basic problem is that there has not been enough re-examination of what activities should be supported by the Government budget in Zambia, in relation to available revenue. Given the variance between budgeted and actual expenditures, prioritization gets loaded onto budget execution through "cash budgeting", with the Ministry of Finance often releasing less expenditure authority to spending ministries than the ceilings approved by Parliament. Sometimes there are "Supplementary Estimates" for purposes for which inadequate provision was made in the original budget. These factors cause the distribution of actual expenditure to vary considerably from the original budget. A review of GRZ data (all sectors) for 1997 found that actual public expenditures were less development-oriented than the budget enacted at the beginning of the year. There are instances in the education sector of this general problem. 6.27 In order to address the situation, the Government intends to move toward a Medium Term Expenditure Framework (MTEF)-based budget system, as part of the Public Sector Capacity Building Program (PSCAP). In view of the size of the Ministry of Education, and the importance of education in the development process, it would make sense for the Ministry of Education to be one of the three pilot ministries. Household Expenditures on Education 6.28 The Living Conditions Monitoring Survey 1996 found that, on average, 2.7% of Zambian household expenditure was devoted to education expenditure, including school fees, examination fees, PTA contributions, school uniforms, private tuition, books and stationery, and other school expenses. Urban households spent five times more on education than rural households. At the primary level, the school fee of K 500 set by the Ministry of Education is much less important than PTA fees. A survey in 1999 found that the average PTA levy per pupil was K 2100; however, the average levy was as low as K 1000 in some rural districts, and as high as K 10,500 in Lusaka district. In one district, the actual cost to parents of primary education was estimated as over K 100,000 per year, including the costs of uniform, books, and stationery. 6.29 The LCMS findings imply that household expenditures on education were probably about 2.2% of GDP - roughly equivalent in magnitude to Government expenditures on education. Unfortunately, the questionnaire does not ask for which level of the school system such expenditures are being made, or whether Government or 77 private schools are being attended by household members. The questionnaire could usefully be elaborated in these directions. Conclusions and Policy Recommendations. 6.30 Primary education is the foundation for post-primary education and for lifelong learning. In Zambia, results such as the National Assessment indicate that, in general, this foundation is deficient, especially in the rural areas. Primary education in the rural areas is low-cost and low-quality. The remedies - bringing more trained teachers into rural schools; integrating textbooks and other instructional materials into teaching practice; initial literacy in Zambian languages rather than English - are well known, but not always easy to put into practice. Money is likely to be part of the problem, but increased spending in the 1990's has not improved performance. Through the inter- related PRSP, HIPC, and MTEF processes, Zambia can and must find a way to bring higher quality to the rural schools, with supporting measures elsewhere in the education system. 6.31 Evaluate the various alternative ways already being recommended to bringing educational services to remote areas: * Bring more trained teachers into rural schools; integrating textbooks and other instructional materials into teaching practice; e Experiment with alternative ways, such as: improve the living conditions, lighting, sanitation and water facilities of rural schools where rural teachers are bound to live; significantly raise the allowances to teachers for extended stay in rural areas; assists local communities and voluntary organizations that wish to develop their own schools in rural areas, supporting them with necessary resources for education materials and rural teachers; decentralize the monitoring, hiring and wage disbursements of rural teachers to the local communities and consider giving back to church groups, other bodies, and local authorities under enabling guidelines the responsibility they had in the past for the management of schools in selected rural areas etc. * Learn what works from these, and other, experiments. The introduction of any policy measure should be accompanied by a means of careful evaluation in terms of test scores, student attendance and retention. Successful interventions can be replicated and all results should be publicized in order to mobilize parental involvement. 6.32 Government plan to gradually eliminate University subsidies should be made within a coherent framework with an institutional mechanism for regular open discussions among stakeholders of the issues and strategies for implementation. Possible options include: * raised tuition fees gradually, starting with new incoming students; reduce aid first to non-instructional expenditures (food, dormitories, transportation); protect poor students by applying means testing to 78 student aid; eventually, fees should cover marginal cost of education and housing, * ensure the growth of the economy to raise the private returns of higher education but allow for private provision and competition; * institute flexible payment arrangements and aid in the form of loans or student employment; expand student loan schemes and consider income- contingent loan systems, sometimes referred to as graduate tax, in which loan repayments are a fixed proportion of a graduate's annual income as well limit loan programs to a fixed share of students to control fiscal exposure and to promote academic competition and performance; * replace traditional approach of "negotiated" budgets with alternative mechanisms such as incremental resources tied to self-generated resources and linking funding to performance such as the number of graduates; competitive fund to promote quality improvements and innovations; * consider hiring private companies to manage the universities under well- defined and enabling arrangement etc 6.33 To learn quickly what works from these, and other, experiments, the Ministry of Education needs to be more forward looking by improving its planning, management, and monitoring capability. The introduction of any policy measure should be accompanied by a means of careful evaluation in terms of test scores, student attendance and retention. Successful interventions can be replicated and all results should be publicized in order to mobilize parental involvement. 79 7. HEALTH 7.1 The challenges for public spending in health come from a serious deterioration of the health status of the population, primarily due to the AIDS epidemic, as well as formidable problems of allocation and management of health services. A major reform program in health has been underway for most of the decade and a number of goals has been achieved. However, several problems remain relative to expectations of the program (see Box 7.1). 7.2 As far as overall public resources are concerned, the health sector has been relatively protected (Figure 7.1). The large increases in public expenditure in 1994 discussed above have been largely maintained though with some slow decline discemable. This rate of decline is much slower than in other categories of public spending and thus the share of public spending on health has risen somewhat over the period. Zambia spends a relatively high percentage of GDP on health when compared to other African countries of comparable income. The problem is clearly with how the money is allocated to different levels of health care system and with how the money is utilized. Figure 7.1: Ministry of Health - Real Expenditure 1991-99 in 1994 billion Kwacha 60 50 - 40 - 30 - 20 -_ 10 - 0 80 7.3 Efficiency. Health services span a wide range of activities from almost pure public to alnost pure private goods. While most markets associated with the health sector are imperfect to some extent, there are a few characteristically large market failures. The first is the inadequacy of private markets to deal with infectious disease, including HIV/AIDS. Accounting for a large proportion of health problems in Zambia, infectious diseases have efficiency claims on public expenditure for two main reasons. Many of the interventions associated with infectious disease control are public goods such as a mosquito control measure for malaria or establishing a health education campaigns for such basic issues as personal hygiene, the promotion of condom use, STD control, use of insecticide impregnated materials. Even for those activities that are not public goods, many are subject to substantial external effects. Control of diarrhoeal disease is frequently a question of providing safe drinking water or improved sanitation services. The latter, particularly in urban areas, have benefits well beyond the direct recipient of the service. 7.4 Prevention and treatment of certain infectious diseases, such as HIV/AIDS and tuberculosis, also have large external benefits. The tuberculosis problem in particular has been increasing due primarily to its interaction with HIV infection. On the one hand, due to the transmission mechanism, a timely cure of tuberculosis prevents its spread. Also, due to the cost of the drugs and the fact that people often feel much better long before the completion of treatment, people will often choose to end treatment prematurely. This leads both to a resurgence of the disease in the patient (a private cost) which could lead to spread of infection to others (a social cost) as well as an increased risk of developing resistant strains of the disease (a large social cost). Drug resistant tuberculosis increases the risk of death and the cost of treatment for society as a whole. 7.5 Therefore, HIV/AIDS, malaria, diarrheal disease and tuberculosis, all major threats to health in Zambia, are also high priority items for public policy on efficiency grounds. Another area of infectious disease control that is a priority area for public policy is immunization. Whether all immunizations confer external benefits is a subject of some controversy, however, public programs can often be credited with expanding immunization, an effective weapon in the fight against child mortality. 7.6 It should be kept in mind, however, that it is not necessarily in the health sector, per se, that the main public expenditures to attack these problems will be spent. The most important determinants of health status are usually4l income (operating through the channels of nutrition, housing conditions and ability to afford health care), education (operating through knowledge of basic hygienic and nutritional practices as well as the ability to identify health problems which need treatment) and environmental factors such as access to safe water. In Zambia, the effect of safe water and sanitation is significant in urban, but not in rural areas. Mother's education is very important for children's health in both areas. Public spending to improve any of these is not likely to come from the health ministry. 41 This is true in many countries as well as Zambia. For Zambia, analyses of the available Demographic and Health Surveys reported in the PER also identify these factors as important for the reduction of child mortality. 81 7.7 For quite different reasons, appropriate handling of the AIDS problem also extends well beyond the authority of the health ministry. The problem of AIDS is more than a crisis, threatening the overall health and well-being of people well into the future, from those with the disease, to orphans to communities that can no longer cope with the extra dependents of the ill and their families. Current estimates of prevalence are: the national average is 20 percent, urban is at 30 percent, and rural 10 percent. This requires extensive public expenditure on programs that can be shown to be successful and requires an openness and forthrightness concerning the problem the government is increasingly demonstrating. It is unclear whether a single specific public program has had a major impact on the disease but those countries in which the epidemic has proceeded more slowly have invested significantly in a broad mix of interventions including wider publicity and public discussion of the problem, providing people with the information and familiarity with the issues to adopt self-protective behavior, promotion of condom use, STD control, Voluntary Counseling and Testing, and peer programs, while soliciting the active participation of people living with HIV/AIDS. Experimentation with these activities to determine which of these or combination of these is successful in Zambia is a high and urgent priority. 7.8 The other characteristic market failure is the virtual absence of private insurance. For many well-known reasons, insurance markets are likely to be under-supplied by the private sector, often to the point of complete non-existence as is the case in rural Zambia. In these circumstances, people are exposed to catastrophic financial losses in the event of serious illness and to the fear of this loss in anticipation even during times of good health. Even if people would be willing to pay the actuarially fair rate for protection against many expensive-to-treat illnesses (that is, the cost of treatment which is very large times the probability of needing it in any one year, which is very small) they are not able to do so due to the absence of that market. Public intervention is required either to increase insurance use or to avoid needing it with public provision of relatively expensive services, usually in a hospital. 7.9 Equity. Equity in health - both in health status and in access to health services- is a major concern in the health sector and with good reason. Figure 7.2 shows child mortality rates by wealth for 1997, showing substantially higher rates for the relatively poor. While the evidence in Zambia does not exist, evidence from many places around the world confirms that not only do poor people die younger than do the non-poor, they also die of different things. In particular, poor people are dramatically more prone to illness from infectious disease than the non-poor relative to chronic, non-infectious illnesses such as cardiovascular disease and cancers.42 It so happens that these are the very same problems that frequently well-justified on efficiency grounds. Therefore, if one goal of health spending is to redress inequities in health status, the control of infectious disease involves a great deal of complementarity between equity and efficiency goals of government. 42 See World Bank (1998) or Hammer (1997). 82 Figure 7.2: Child Mortality (in percent) by Decile of Wealth, Zambia 1997 30 25 20 1 5 10 5- 0 1 2 3 4 5 6 7 8 9 10 Wealth Source: Zambia Demographic and Health Survey, 1997 (Macro International, Inc., University of Zambia and Central Statistical Office). 7.10 On the other hand, simply providing health services does not necessarily redress these inequities. Figures 3.8 through 3.11 in the section on benefit-incidence above show the usage pattern of public health care facilities. Two points are clear from these pictures. First, use of primary facilities is considerably more equal across income groups than of hospitals. Of course, it is hospitals that absorb the lion's share of health expenditures. Second, even in the case of primary facilities, the public sector doesn't succeed in reaching the very poorest people in the country. 7.11 Related to this point, the relationship between fees in public facilities and poverty reduction is not completely clear. Figure 3.11 in the benefit incidence section shows the use of public facilities by age group. Fee exemptions are given for people younger than 5 and older than 65. In these age groups, the use of public facilities is not well targeted to the poor. All income groups want to take advantage of free care. For people in the intermediate age groups, the use of the private/NGO sector is much higher as income increases and the use of public facilities correspondingly lower. In this case, the better off people self-select out of the public sector. Since fees in the public sector do not cover full costs, there is still a substantial subsidy component and that subsidy is considerably better targeted to the poor under the regime of public fees. 7.12 In the area of coverage of immunization, a high priority for efficiency and equity there should also be concern over the performance of the health system. Figure 7.3 shows immunization rates in 1996 and 1998 by income group. There appears to be a decline across the board in immunization coverage in that period, with the declines somewhat higher among the poor. This decline took place with very little change in the overall funding of the health sector and no increase in fees for immunization, that is, they remained free. 83 Figure 7.3: Immunization Rates(in percent) by Income Quintile 90 80 70 60 50 01 99 6 40 -*19981 30- 20- 1 0 0 1 2 3 4 5 Income Quintle 7.13 The above arguments make it clear that it is not true that simply increasing expenditure in the health sector is the same as helping the poor. There are certainly health programs that would help the poor. But the type of service, the geographic pattern of expenditures and the incentives for use of public facilities can be such that increased spending in health has little impact on the poor. 7.14 Implementability. Ensuring reasonable quality health care, particularly for the poor is further subject to difficulties in implementation. The health sector reform of 1991 had as an expressed goal "...providing equity of access to cost-effective, quality health care as close to the family as possible...". To achieve this laudable goal is easier said than done. A particular difficulty in the attempt to bring services close to the family is that of attracting and managing trained professionals to public service in remote areas where a substantial proportion of the public lives. Also, "cost-effective" services are often associated, in practice, with care at facilities with levels of sophistication lower than those of hospitals. It is difficult, in Zambia and in many countries around the world, to maintain medical care by trained personnel in such facilities and in such locations. Medical personnel are generally among the better educated in society and tend to want urban amenities for themselves and their families as well as urban income-earning opportunities for themselves. Vacant positions in rural areas are common. There has even been experience of people who are offered training at a fairly low level as community health workers leaving rural areas to set up business in more lucrative (urban) areas. 7.15 Even when staffed, it is difficult to assure conscientious or courteous services. In the recent report on "Voices of the Poor" a common complaint in Zambia was discourteous treatment at public health facilities as well as demands for payment of consultation fees when drugs are not available.43 The issue of management and control of 43 "Consultations with the Poor, National Synthesis Report, Zambia" Lusaka, May 1999 84 pharmaceutical stores is long-standing and prominent.44 The issue is not necessarily deliberate misuse (i.e., theft) by professionals but could merely be inadequate incentives for exercising due diligence in supervision of less-well-paid staff. Management problems at a higher level of administration than facilities are also common due to the overlap and unclear division of responsibilities between the Central Boards of Health and the Ministry. This, however, is discussed in the section on management issues. Here we are concerned more with inherent difficulties in providing care at primary level. 7.16 Therefore, even if a highly decentralized network of primary care facilities was the ideal solution for providing health services, it is hampered in its implementation by inherent difficulties in management and administration. The policy questions can be: "how can incentives to public providers be changed so as to obtain better service?" or "are there alternative delivery options that are less vulnerable to these implementation problems?". The answers require some experimentation and research. One possibility is that mobile units may be used for areas where it is hard to find resident staff. Another might be to focus on particular services, such as immunization campaigns, that are easier to monitor and control and do not rely on discretion of medical personnel. It is also possible that if medical staff are more likely to show up for work at hospitals, it might be more important to bring patients to providers with, say, better roads and transport and better referral systems, than it is to bring providers to patients. Recommendations 7.17 Given the history of health sector reform over the past decade, it is important to experiment with policy options and make a concentrated effort at evaluating their outcomes. Monitor health status regularly in order establish the effect of public programs and to replicate successes. * The most pressing threat to health and a very serious development problem is HIV/AIDS, though its claim on public resources has to be carefully evaluated. The government should continue to heavily publicize the problem, its extent, its causes and the means people can take to protect themselves. A wide variety of specific interventions should be tried but their effects should carefully be evaluated so as not to waste potentially large sums of money. * Other infectious diseases are of high priority. Health education activities, particularly for HIV/AIDS but also for basic hygiene, management of water sources and sanitation, malaria protection and nutrition are likely candidates for public intervention and evaluation. * Improvements in the operation of the health care system should be carefully evaluated. Experiments in alternative ways of increasing access of rural people to health care need to be tried and evaluated. Options might include: determining the responsiveness to increases in pay for rural 44 See, as one of many examples, "Hospitals 'dry up' once more" Sunday Times of Zambia, October 22, 2000. 85 service, establishing mobile clinics, "campaign"-style immunization programs or improved communication and referral to urban areas. * Fees for primary care at hospitals should be raised to reflect the marginal cost of such provision both to divert use to cheaper, more efficient places for care as well as to conserve the budget for services that require public expenditure. Exemptions for primary care fees should be based on rural location other than for vaccinations. Payments might differ for urban and rural areas or by region, except for maternal and child care services. The administrative feasibility of specific exemptions for the poor should be explored. 86 Box 7.1: Zambia's Health Reform Program Economic difficulties since the 1 980s have eroded much of healthcare gains from the 1980s and the health sector is in crisis. In 1994, the Government and donors launched a health sector investment and reform program to coordinate efforts and make the most effective use of limited resources. Shifting the focus away from tertiary-level institutions towards primary-level healthcare, the program aims to improve health status/outcomes by increasing access to and the quality of a national package of essential and cost-effective health services in a decentralized health care delivery system. The long run objective was to provide an essential package of health care services to all Zambians. A key ingredient was the decentralization of the budgetary and decision-making powers to District Health Boards and Hospital Management Boards. The program also aims to tackle health problems at the household and community levels with the health center as the link for the delivery of primary health care services to households. The program was estimated to cost $537 million and the World Bank supported with a Health Sector Support Project (US $56 m). Status. Impact on health care delivery and on health outcomes has been disappointing and slow. Recently, the Government adopted a new 5-year National Health Strategic Plan (2001-2005) to continue the reforms and improve service delivery, but financing support from donors has so far been unenthusiastic because serious problems remain: * At the basic level, health service delivery requires that facilities have trained personnel, the right drugs and equipment. In Zambia, it is a common finding that one of these three essential ingredients for service delivery is frequently missing. As a result, there is a lot of wastage of resources that occurs because facilities are left without personnel, drugs, or adequate equipment. * In particular, a key problem is the frequent shortages of drugs and medical supplies in hospitals and health centers due to erratic distribution methods. * Conditions (low pay, inadequate equipment, poor supplies) in Zambia's state-ran hospitals have deteriorated to alarming levels and were the primary reasons for a long-running doctors strike. * Governance issues particularly in the procurement and distribution of drugs are also perceived to be serious. Although the issue is resolved, ineligible drug expenditures have affected the IDA credit and implementation problems continue to have a negative bearing on project ratings. In the Joint Health (Pre) Appraisal in March 2001, the point was made that present contractual arrangements for management of storage and distribution of drugs and medical supplies through Medical Stores Ltd was found unacceptable and donors were prepared to support a parallel system to ensure accountability and transparency (see Annex 3 of Report). * Other issues include: decentralization lacks authority and flexibility to manage and delever cost-effective health services; inadequate attention to maintenance of physical infrastructure and equipment, which are deteriorating and require repair; wasteful duplication of function between Ministry of Health and the Central Health Board. 87 8. WATER AND SANITATION 8.1 The problem of ensuring the availability of safe water is substantially different in rural and urban areas. Direct provision, which this chapter is primarily concerned with, is an urban phenomenon where government's role is significant on efficiency grounds but ambiguous on equity grounds. In rural areas the problem of unsafe water and inadequate sanitation is more of an issue of education, particularly health education. There are large gains to be had in terms of reduced morbidity from diarrheal disease, gains that disproportionately accrue to the poor, from better and more hygienic management of water and sanitation in rural areas. However, the public infrastructure and expenditure contribution is likely to be modest. 8.2 Reform of public expenditure for expanding and improving water and sanitation services (WSS) faces several constraints inherited from the past. Much of the current infrastructure of the sector was built in the late 1960's and early 1970's when the government budget was not nearly as tight as is the case now. As a result, water and sanitation services were established at a very high level of quality and were free. Since then, government resources have been more limited and the expansion of services has been difficult with actual deterioration in some periods. Table 8.1 shows estimates of the population with access to safe water and sanitation. Table 8.1: Percent of Population with Water and Sanitation Services Access to safe water 1980 1993 1995 Urban 65 76 66 Rural 32 43 37 Total 46 59 53 Access to sanitation Urban 77 76 66 Rural 34 37 Total 55 51 8.3 One legacy from the history of good services is that with the current budget so much tighter, expectations by the consumers are going to be continually disappointed on quality grounds and thus they are likely to be reluctant to pay fees near the cost of delivery. 8.4 Nevertheless, the government has recognized the deterioration in the provision of water and sanitation services and has begun reforms of the sector. The main objectives of the reform are to 1) establish sustainable services and 2) recover costs when possible. These principles have been enacted in the Water and Sanitation Act of 1997. This act transfers authority for the management of water and sanitation services to local authorities and regulatory authority to the National Water and Sanitation Council. The 88 Act provides for several options for local authorities to manage the WSS systems, one of them being the use of a autonomous Commercialized Utilities (CUs) to manage and operate the services. Presently, the various local authorities around the country have jointly formed seven CUs covering the country45 and transferred the WSS assets, operational and management responsibilities to the CUs. Since the reform program is underway with the principle of cost recovery clear and established in law, this review will only try to emphasize a few points of particular importance to assessing any continuing subsidies to the sector. 8.5 Efficiency. The principal market failures that will continue to require public management are the scale economies of urban piped water systems which make the distribution of water, at least, a monopoly that must be regulated. In urban areas, the relatively heavy expense of the initial connections are difficult to recoup up front from consumers among all but quite well-off urban residents due to very limited credit markets. Loans are not forthcoming for consumption purposes such as residential water. Currently, most reticulated areas are located within established town centers, and it is probable that the majority of the urban population who are not served through piped connections are in peri-urban areas. Subsidies aimed at increasing the number of people connected to urban piped systems should mainly be targeted at these areas, through the extension of existing reticulated systems. A rough estimate of required funding for extending services to arrive at a urban coverage ratio of 70% using shared kiosks is about $14 million, but the net amount of the subsidies could be quite low if appropriate fees are charged. 8.6 It should be noted that the preceding issue surrounds the fixed investment costs of the sector and not the marginal costs of delivered water itself. For this, there is strong efficiency justification for prices to reflect marginal costs. Having said this, it should be noted that volumetric production costs are not necessarily high compared to other water utilities in the region. A main issue is the absence of a price mechanism to limit use that has resulted in enormous wastage of water in urban areas. The networks are largely not metered and rely on flat tariff charges. A significant amount of water is wasted through lack of discipline and leakages in house piping which are not repaired, since a flat tariff structure does not provide the incentive to conserve. During a recent supervision mission of the World Bank, it was estimated that up to 60% of taps in low cost housing areas are missing leading to a free flow of water. Flat monthly tariff are in most cases, inadequate to recover the cost when such large quantities of water are consumed (or wasted). It is also probable that most households will not be able to afford the monthly charges if these were to reflect the amounts of water actually used. A rough estimate of required funding for demand management through metering of existing consumer pipe connections is about $8 million. 8.7 Other than the need for intervention to cover capital costs of the sector, a further characteristic market failure in the sector, in urban areas at least, is related to health effects of water and, especially, sanitation services. As described in the health chapter, one important explanatory variable for health status as measured by infant mortality, say, 45 Kafubu, Mulonga, Nkana, Southern, Northwest, Chipata and Lusaka. 89 is access to safe water and sanitation. This is not sufficient to justify government subsidy, however, as this benefit accrues mostly to the family with the service. However, at least in the case of urban sanitation, and arguably in the case, of water supply (often jointly supplied and therefore difficult to address separately), there are possible external benefits from access. 8.8 Most of the CUs also suffer from operating inefficiencies resulting from the deteriorated infrastructure that are the legacy inherited from pre-CU period. At the time of the transfer of responsibility for providing WSS services from the local authorities to the CUs, there were a number of outstanding liabilities which had been accrued for many years while the services were being provided by the local authorities. Some statutory instruments that were gazetted to affect the transfers attempted to allocate specific assets and set limits to the liabilities which were to be assumed by the CUs. In practice, the situation is less clear and there are some uncertainty as to who should be responsible for these accumulated liabilities, which local authorities understandably would want to transfer to the CUs together with assets. It is certain, however, that neither the local authorities nor the CUs have the ability to settle these liabilities. 8.9 CUs have had to accept (in some cases via secondments) all staff previously employed by the local authorities' water and sanitation departments. This include superfluous and under-qualified personnel. An estimate of required finding for the retrenchment of former local authorities' staff no longer required by the CUs is estimated at $500,000. 8.10 Accumulated operating liabilities from prior operations - mainly accrued debts for electricity owed to ZESCO and chemicals owed to suppliers - continue to hinder the CUs from obtaining normal trade credits for their operations. It is probable that ZESCO will increasingly attempt to collect these outstanding arrears, and the CUs' operations increasingly threatened by daily power cuts. Thus, the situation is particularly acute in the case of electricity debt, while a large portion of arrears for chemicals are owed directly with the Ministry of Local Government and Housing. An estimate of required funding to pay off the electricity arrears as of July 2000 amount to $8.4 million. 8.11 The CUs have also inherited WSS utilities which are currently unable to cover their operational costs with their operational revenues. In many cases, physical assets are not operating efficiently owing to a prolonged period of under-investment in the maintenance of these assets. Revenue streams are also low due to non-billing, non- collection of bills and in some cases, low tariff levels which have not been adjusted for some time. Inadequate cash has also contributed to the difficulties in ensuring adequate supplies of chemicals, resulting in incomplete treatment of water. The process of turning around these companies will take time and in the meantime, some support to cover the CUs' cash deficits and working capital requirements are necessary. In practice, this could be in the form of assistance to meet firstly, electricity bills and secondly chemical bills. These two items form a large proportion of a CU's cost. 8.12 It is paramount, however, that any form of subsidies which are provided for this purpose do not defeat the central tenet of the Water Sector Reform policy which states 90 that WSS utilities should be able to operate on a financially sustainable basis and not be eligible for continuing general subsidies. Any such support must therefore be non- permanent, and be provided on a declining basis. CU's who apply for such assistance should need to have an acceptable financial plan in place to ensure that an increasing percentage of its operational costs will be met through operational revenue, and should be monitored on such milestones / basis. One condition could be that the CU are to achieve an annual cash revenue equivalent to at least 60% of operational and maintenance costs (excluding depreciation and debt servicing costs) by the second year of such an assistance scheme being granted. An estimate of required funding for the turnaround time assuming a 4-year period and a linear declining operational cost support scheme are about $9.7 million (Year 1), $7.3 million (Year 2), $4.8 million (Year 3) and $2.5 million (Year 4). Figure 8.1: Percentage of People with Piped Water by Wealth Decile, 1993 and 1997 100 - 80 - 60 - ~~~~~ 60 ~~~~~~~~~~~[01993 ~40 -A1997 20- 1 2 3 4 5 6 7 8 9 10 Wealth Decile 8.13 Equity. Since water and sanitation services are primarily an urban phenomenon, it is unlikely that subsidies to them are pro-poor. This presumption is borne out by Figure 8.1. This shows the source of water by people's decile of "wealth"46. Sanitation services show a similar pattern and are not shown. 8.14 Figure 8.1 illustrates the deterioration of services over a very short period of time, consistent with the aggregate figures of the table above. The share of population with piped water was lower in 1997 than in 1993 in all except the eight decile. More importantly for the equity argument, however, is the fact that it is only at about the middle of the distribution that anyone has piped water at all. This reflects, almost entirely, the fact that it is urban residents with these services and that urban areas are very much better off than rural. This is complicated somewhat by the fact that the deterioration of services has been among the relatively worse off segments of the urban population. 8.15 The pattern of water and sanitation use, and therefore the likely pattern of subsidies for the services, implies two conclusions for policy. First, the imposition of 46 The source of this data is the D emographic and Health Surveys of 1993 and 1997. The definition of income is substantially different from the LCMS surveys and not directly comparable. For an explanation of the definition and use of the concept of wealth, see Bonilla-Chacin and Hammer (1999). 91 charges at current patterns of water connection is not regressive. The beneficiaries are disproportionately urban and well-off within urban areas. Since the efficiency of the system depends on increased charges and the distributional impact is favorable as well, this is a clear area for saving public expenditure. 8.16 Second, the expansion of services is likely to be for the benefit of the middle part of the income distribution, that is, among the less-well-off among the urban population. Certainly if the initial expansion simply undoes the impact of the deterioration of the recent past, this will be the outcome. Therefore, if services are financed from national funds, the distributional impact of expanding any subsidized service is almost certainly regressive. If, however, decentralization proceeds and such services are financed by local taxes, there would be a favorable redistribution within urban areas. Under current financing arrangements, however, there is substantial reason on both efficiency and equity grounds for increased cost recovery. 8.17 Implementation and information requirements. Due to consumers' experience in the sector, there is likely to be resistance to instituting adequate charging mechanisms. As of now, what cost recovery there is amounts to only 30-40 percent of operations and maintenance expenditures alone, let alone any recouping of capital costs. Even this is mostly from commercial users and not households at all. Further, the basis of the charges is for the hookup to piped water, not monitored use of amounts of water as measured by meters. Metering and billing based on volumetric usage is a method of demand management which has proven successful in many instances. A phased metering scheme could be introduced, starting with the category of consumers that are most likely to be able to pay volumetric charges and hence most likely to enable the CU to recoup the cost of metering and meter administration47. The implementation challenge is to establish metering for users in urban areas. This will allow the efficient charging for water as well as the financing needed to extend the system. 8.18 Even with metering, there could well be room for continuing partial subsidies for the external benefits arising from community health. How much this will cost when efficient charges are assessed is guesswork at this stage. However, some of the elements for the calculation of subsidy costs on the basis of the subsidy rate are available. The demand for water as a function of income is sometimes calculated from willingness to pay surveys, though these are often suspect on methodological grounds. A rough and ready calculation of demand might be to assume (as is sometimes done in Zambia) that water costs be about 6% of income. From this, it should be possible to identify the demand curve for water. An illustrative example is shown in the following figure (Figure 8.2). 8.19 Demand can be compared with the marginal cost of providing water to judge the level of use without subsidy. It should then be possible to estimate the degree of coverage of water as a function of the subsidy costs going into the sector. If the marginal cost intersects the demand curve in a region where demand is elastic, costs to the system will 47 Large scale metering is expected to be carried out in the Southern Water and Sewerage Company, using German aid funding. 92 increase rapidly with declining prices but will also yield higher coverage. On the contrary, if the intersection is near an inelastic segment, subsidy costs become simply transfers to people already purchasing water and yield little benefit from overcoming the externalities. It is important to get more accurate information, therefore, on the marginal costs of delivering water in a form that is comparable to its respective demand curve in order to be able to assess the tradeoff between budgetary outlay and actual coverage of safe water. As of now, such cost information is lacking. Figure 8.2: Hypothetical Private and Social Demand for Water (based on demand of 6% of income) 5000 - X 4000 - 3000 >A 2000 ioo 0- 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Proportion of low-cost urban population demanding water - - Private demand - Social demand Recommendations 8.20 Safe water is a high priority for households, but direct provision of water is primarily an urban phenomenon, well justified on efficiency grounds but less on equity. * Outside of digging wells in rural areas, there are large gains to be had from health education in terms of reduced morbidity from diarrheal disease, gains that disproportionately accrue to the poor, and from better and more hygienic management of water and sanitation. * Develop a cheaper method of providing water and sanitation to poor communities in urban and rural areas. * Reversing the deterioration in the provision of water and sanitation services in urban areas require implementing an existing policy, clearing and sorting arrears (among Cus, ZESCO, and Government), metering and collecting fees, and building up the capacity to operate viably. * In particular, aggressively pursue the metering and charging for water on the basis of water used, not hookups. * Eventually prices should reflect marginal costs of water plus a premium to recoup capital expenses over time. Prices can be raised gradually over 3 to 5 years to determine the responsiveness of demand and the appropriate long run subsidy. 93 9. AGRICULTURE The Role of Economic Environment and Policy 9.1 While agriculture is important to the vast majority of the Zambian people, government's role in a market-oriented economy is indirect and is limited to ensuring that farmers' activities reflect real economic opportunities. In contrast to Government's heavy interventions in the 1 980s (establishment of parastatals, excessive market regulations, price controls and subsidies), economic liberalization in the 90s has brought about less public interventions and less share of public expenditure directly allocated to agriculture as the government attempted to promote private sector development in the sector. The mixed results in agricultural performance however indicated that the sector has been largely and indirectly affected by economic policies (outside of weather) that on the surface appear to have been sector neutral or neutral to agriculture, mainly macroeconomic and trade policies as well public expenditure on infrastructure:48 * An important attribute of agriculture is that it is generally more internationally traded than non-agriculture; its competitiveness and growth are therefore highly susceptible to movements of the real exchange rate. In Zambia, the Dutch disease impact of copper revenue was a significant factor, but the size of public expenditure (or the public sector deficit) and the continuing problems of inflation also tended to maintain a relatively higher real exchange rate, which favored non-tradable over tradable goods like agriculture. • While trade taxes and quantitative restrictions on imports and exports have been liberalized, other impediments like high transaction and transportation costs continued to hamper agriculture exports from a substantially higher growth. * Agriculture is also more capital intensive relative to non-agriculture, although much of the capital is external to farmers, like roads to access markets and the availability of water etc. The squeeze on infrastructure spending in the 90's, particularly in the rural areas, was therefore detrimental to agriculture. 9.2 The primary area for public expenditures to improve efficiency is to remove bottlenecks to sales of agricultural production via the improvement of rural infrastructure such as roads. To date, government has tended to intervene in the functioning of commercial markets via the subsidies to fertilizer and maize sales by the Food Reserve Agency. This has hampered the development of private trade as well as distorting farm production. 48 The economic arguments and cross-country evidence are presented, for example, in Mundlak (2000). 94 Figure 9.1: Source of All Maize Sold Poorest 5% Quintile 2 4/~~~~~Q ulnt~~Qunile 3 14% 52% | 520/o \ \ \Qulnti~~~~~unle 4 19% 9.3 The equity effect of subsidized grain purchases and fertilizer sales is not likely to be oriented towards the poor. First, the majority of fertilizer users are commercial farmers who consume about 70 percent of fertilizers in the country; 30 percent are consumed by medium-scale and small-scale farmers mostly along the line of rail. Moreover, both private distributors and FRA are known to have problems reaching poor farmers in the remote areas. Second, source of maize sale as depicted by Figure 9.1 confirms that the beneficiaries of higher maize prices or subsidized inputs are not likely the poor. Very little of the marketed surplus of maize comes from poor farmers and over 50% of marketed maize comes from the few people in rural areas who are in the richest 20% of people in the country. Third, the acquisition and distribution of fertilizers by FRA, which is beset by non-transparency, also indicate that small and poor farmers were not the beneficiaries. The following observations can be drawn from the public report of the Auditor General on the account of FRA:49 (i) fertilizers distributed were loaned to companies and enterprises as well as individuals including politicians; (ii) loan applications were not subject to credit scrutiny and lack of records did not allow analysis of the borrowers; (iii) only 2.4 percent of the total loans had been repaid (as of March 1999); and (iv) there is no proper system for monitoring loan repayment. The value of the fertilizer subsidy averaged US$20-25 million a year, about 0.7 percent of GDP (see also public expenditure in agriculture below). 9.4 The equity effect of investments in rural infrastructure, particularly roads, is hard to determine, primarily because the ways in which integration with the rest of the economy is complicated. Not only would this increase output prices (which would benefit the larger farmers more) but would also decrease input prices (with likely better 49 Republic of Zambia (1999). 95 distributional effects). Of course, roads would also improve access to urban areas for reasons beyond farm production such as for educational or health services. In this way it is potentially quite progressive. Comparing the map of physical infrastructure in Figure 10.1 with the poverty map of Figure 3.1 shows that much of the nation's improved road system links the poorest areas of the country with the urban center. This gives some indication of the progressive nature of these investments. 9.5 Implementation. In order to undertake critical reforms and improve the performance of agriculture, the Zambian Government initiated the Agricultural Sector Investment Program (ASIP) with the assistance of donors. ASIP is best understood as a key ingredient to the economic liberalization undertaken in the 1990s. Its achievements in helping reverse three decades of Government dominance particularly in the agricultural sector are substantial. However, problems have plagued the delivery of services to farmers and the impact of the reforms in terms of the performance of the sector is much less than expected. Moreover, there are persistent complaints about late delivery of fertilizer. The visibility of the problems surrounding the fertilizer subsidy and the wrong signals sent to private suppliers by Government intervention in the input supply continue to raise concerns to various stakeholders (see Box 9.1). Zambia's Agriculture 9.6 It is estimated that about 60% of the population depend on the production of crops, livestock and fisheries for income and about 70% of the labor force for employment. Zambia is one of the few countries experiencing urban to rural migration; more people are now living in rural areas than a decade ago. In the period 1994 to 1998, the contribution of agriculture to GDP was in the range of 15 to 19 % and increased to about 21% in the last two years. Agricultural export also increased dramatically in the last ten years. In 1990, agricultural export accounted for less than 2% of total export value and in 1996 and 1999 increased to 10% and 20%, respectively. Of the total non- traditional exports, the share of agriculture export was 37% in 1996 and increased to 55% in 1999. Most of the employment and manufacturing output, which accounts for 10- 12% of the GDP, comes from agro-based operations. About 75% of the manufacturing sector output and almost the same percentage of employment comes from agro-based industries (food, beverages, tobacco, textiles, leather, etc.). 9.7 There have been severe year-to-year fluctuations in agricultural production and the trend has been that of stagnation, with occasional declines. Since 1995, which was one of the best years in the 90s, production has been on the decline until 1999 due to shortage of rainfall, decline in area cultivated, change in crop-mix and lower use of agricultural inputs. Production growth rates were negative in 1996, 1997 and 1998 at -.6, -5.3 and -6.1 percent, respectively. This trend was reversed in 1999 with 4.3% growth rate. Although figures are not available, year 2000 production is expected to be higher than the previous year. The good harvest for two consecutive years has made Zambia a net food surplus country. 9.8 Despite the abundant land and water resources, area cultivated per household has, however, declined during the same period. This is partly explained by the significant 96 losses in draft animals caused by persistent problems of disease outbreaks, and decline in area under maize. The dominance of maize in the Zambian agriculture has virtually ceased following its drop from 57% of total hectarage cultivated by small and medium scale farmers in 1996 season to about 32% in the 1999 season. Combined with a drying up of farm credit, following the collapse of the agricultural banking institutions, the risks in maize cultivation increased and incomes of farmers were reduced. It has affected farmers of all sizes, especially those in outlying areas, and required them to adjust their production systems; reducing their dependence on maize, diversifying into new farming systems and making better use of locally available means of maintaining soil productivity. Challenges and Prospects for Growth 9.9 The main challenge facing Zambia's agriculture is low productivity. In the 80s when high agricultural growth rates were experienced (due to improvements in prices, good weather, input subsidy) increased production was coming from area expansion. In spite of the input/output subsidy, growth in yields has been very low. Since the implementation of the economic reform program in the early 90s, there is notable structural change (diversification) in Zambian agriculture but there is no improvement in productivity. Technologies that increase yield (fertilizer and agro-chemicals) and improve cropping intensity (irrigation, reduction in fallow, etc.) are limited and reach only few farmers. The other challenge is the sustainable and extensive use of the abundant land resource the country has. The potential impact of an abundant cultivable land on production is not realized, as it is countered by inelastic supply of farm labor. In recent years, the situation has been exacerbated by the rapid spread of HIV/AIDS and malaria. This calls for an encouragement of urban to rural migration or the introduction of labor-saving technologies to expand cultivatable land and for faster agricultural growth. Another major challenge is completing the sector reform agenda that is already underway. In the last seven years, substantial progress has been made in implementing sector policies that create an enabling environment for agricultural growth. However, the policy and institutional reform agenda is incomplete and the task of improving the incentive framework comprising price and non-price factors that allow farmers to respond and entrepreneurs to invest has to be improved. Public Expenditure in Agriculture 9.10 There is no complete and consolidated data on agricultural expenditure, mainly due to absence of records of donor contributions to the sector.. Pooling data from various sources, some lessons could be gleaned on the patterns of total expenditures, the intra- sectoral allocation and the relative importance of different sources of funding. 9.11 The share of the agriculture sector in the aggregate public expenditure has been erratic since 1994. Its percentage share in the overall public expenditure came down to 2.5 percent in 1996 as against the 10.6 percent in 1994. The upward trend began in 1997, by attaining 6.0 percent, and again declined to 3.1 percent in 1998. The 1999 allotment fell further down to 3.0 percent. 97 9.12 Moreover, the appropriated budget for the sector has never been fully released. Disbursement is guided by cash availability and subjective judgments, not budget, thus reducing the credibility of the budgetary system. For instance, in 1997 disbursement to the sector was 55 percent of the allotted budget (Table 9.1). As discussed in chapter 3, agriculture has been a consistent loser in the release of funds relative to other sectors due to the investment share in its overall budget. Table 9.1. Agriculture, Public Budget Allocation and Actual Expenditures (K'm) Year 1996 1997 1998 1999 2000 (Estimate) A. Estimated Budget Allocation 1. Government a) Recurrent expenditure 29 146 28362 34 189 38 197 37 176 b) Capital expenditure 27 632 6 980 4 636 3 628 Subtotal 29 146 55 994 41 169 42 833 40 804 2. Donor a) Recurrent expenditure 16435 24 058 8 507 59 864 51 500 b) Capital expenditure 41 810 30 635 24 412 20 968 24 886 Subtotal 58 245 54 693 32 919 80 832 76 386 Grand total 87 391 110 687 74 088 123 665 117 190 B. Actual expenditure 1. Government a) Recurrentexpenditure 19590 31 118 29503 35753 20388 b) Capital expenditure NA NA 614 N/A N/A Subtotal 19590 31118 30117 35753 20388 2. Donor Recurrent and Capital 11 800* 32 144 37 274 37 999 N/A Subtotal 11800 32 144 37 274 37 999 N/A Grandtotal 31390 63 262 67 391 73 752 20 388 Source: FMU, MAFF * Breakdown ofDonors actual expenditure by economic classification is not available 9.13 Intra-sectoral allocations are highly skewed towards headquarters, which includes procurement of maize and fertilizer through the Food Reserve Agency (FRA). The historical pattern indicates that with a share of 66.9%, on average, Headquarters received the highest share of the allocation to agriculture, followed by Agriculture Extension, which averaged 5.5%. Other relatively important recipients were Animal Production and Health (5.4 percent); Soils and Crops Research (5.3 percent); and Agricultural Training (4.3 percent). At times, payments to Headquarters are higher than the budget, pointing that the budget management is discretionary rather than planned. Technical departments and field level operations, therefore, do not have allocations that will allow them to carry out their core functions (Table 9.2). 98 Table 9.2: Historical and Projected intra-sectoral Shares of Expenditures50 Description Historical Allocation, Projected Avg. % Allocation, Avg.% GRZ Donor GRZ Donors s Headquarters 66.9 0.0 33.3 10.7 Soils and crops Res. 5.3 10.74 11.6 9.6 Seed control and cert. 0.5 6.0 0.94 1.1 Policy and planning 3.5 8.4 4.3 4.7 Animal P&H 5.4 16.8 8.7 19.5 Agriculture training 4.3 0.0 7.0 8.2 Farm P&M 1.3 4.2 2.1 3.2 Irrigation and land use 2.2 20.2 3.5 5.3 Marketing and trade 1.7 6.9 7.2 16.1 Agriculture extension 5.5 0.0 14.6 5.1 Fisheries research 0.4 6.5 1.0 1.1 Fisheries extension 1.2 0.0 3.1 2.3 Agriculture inf 1.8 10.3 2.7 3.2 Total 100 100 100 100 Source: FMU, MAFF 9.14 Donor contributions, by and large, followed the opposite pattern. When compared with the budget for the years 1996-99, it is evident that the priorities and the corresponding resource allocations are different. The pattern reflected donors' disinclination to pay for the cost of administration and staff. While the disbursement for Irrigation and Land Use, Marketing and Trade, Fisheries Research, Soils and Crops Research and Agriculture Information increased, there was no allocation for headquarters from donors. 9.15 The allocation of budget between recurrent and capital expenditure, and recurrent departmental charges (RDCs) and personal emoluments (PEs), are good indicators of the extent to which the stock of capital is being built and utilized, the efficiency of use and the degree of service delivery in the sector. Table 9.3 provides the relevant ratios for the period 1996-2000, based on actual data for 1996-99 and budget data for 2000. Since the last year does not reflect actual performance, the analysis will concentrate on the first four years. Table 9.3 Agriculture - Expenditure Distribution (%) Year PE RDC Grants Total RC Capital RDC/PE RDC/CAP 1996 20 31 10 61 38 1.5 1.6 1997 16 28 5 94 51 1.7 0.5 1998 21 19 4 44 56 0.9 0.5 1999 20 7 13 40 60 0.3 0.7 2000' 10 13 0.3 24 76 1.2 0.3 Source: Computed from FMU, MAFF raw data Budgetary allocation 50 The Historical Allocation covers 1996-1999 expenditure; the projected allocation is for ASIP, 2000 budget. 99 9.16 During the four years, 1996-99, capital expenditure accounted for an average of about 51% of the sectoral expenditure, while 49% was devoted to recurrent expenditure. The ratio between recurrent and capital expenditure decreased steadily from 1.6 in 1996 to 0.67 in 1999. The increase in capital expenditure is a welcome development but such ratio has to be interpreted with caution, as the expenditures reported as capital, particularly, donor contributions tend to be recurrent and overstate the capital. The mix between RDCs (the materials and supplies component) and PEs (the wage and salaries component) within the recurrent expenditure has varied widely during 1996-99. The ratio of RDCs/PEs stood at 1.5 in 1996, increased to 1.7 in 1997, declined to 0.91 in 1998, and further plummeted to 0.39 in 1999. Continued salary increases, without a corresponding increase in RDCs, have contributed to the declining ratio. 9.17 Though major agriculture sector subsidies have been removed, the government is still subsidizing certain activities indirectly. The FRA was established under the Food Reserve Act of 1995 to administer the national food reserve; provide market information, and lease or sell of government storage facilities. Outside of these mandates, FRA imports fertilizer. In order to maintain the strategic grain reserve, FRA also either imports maize from countries in the region, or buys grain from the domestic market by fixing prices. Since its inception, FRA has been funded by Government through budgetary allocations and through external borrowing guaranteed by Government. Budgetary allocation for the FRA has reached 30% of the total budget allocation for the agriculture sector in some years. Although the Government guarantees are not provided for in the budget, due to poor recovery of credit given to farmers, Government eventually has to pay its contingent liability and in most cases this expenditure crowds out the MAFF budget. 9.18 In adddition to the support provided to FRA, Government has rendered direct financial support to other organizations notably statutory boards like Coffee Board of Zambia, Tobacco Board of Zambia; trusts i.e. Golden Valley Agricultural Research Trust, Cotton Development Trust; NGOs and farmer associations like PAM, Zambia Export Growers Association, Tobacco Association of Zambia, Zambia Coffee Growers Association and Zambia National Farmers Union. Over the period 1996-2000 a total of about K178 billion and US $5.5 million has been disbursed to the statutory bodies. Priorities for Public Expenditure in Agriculture 9.19 Agriculture can move to a higher growth path provided there is a rapid transfornation of the sector to encourage dynamic private enterprises in both production and marketing. Zambia has abundant natural resources (land, water, forestry, fisheries and wildlife) that permit a variety of farTning. Considerable potential lies in expansion of area under cultivation, in diversification of production, and development of water resources for irrigation. 9.20 From the technical perspective, the immediate growth prospect would come from taking Zambian agriculture on growth path based on transforming the performance of existing individual farming units through the application of available technologies, developing new varieties and introducing better farming and resource conservation 100 practices. The second source of growth would be to encourage expansion of cultivated area (less than one-fifth of the estimated 10 million hectares of cultivable land is currently utilized). The, third source of growth would come from transforming smallholder farms from subsistence orientation to farm enterprise orientation, by accelerating the change in the mix of farming activities to those which add more value per unit land and labor. A shift from low return farming system (food crops such as maize, sorghum, millet, cassava) towards enterprises with high value added (dairying, coffee, horticulture, etc.) is critical for Zambia, which is landlocked and a big country with underdeveloped infrastructure. Another option is to increase vertical integration and increase processing of agricultural products and their linkages to other sectors. 9.21 From the perspective of priorities for public expenditure, it is important to determine the bottlenecks for such developments (e.g. transport costs, extension services etc.) with farmers and traders. Public expenditure in the agriculture sector should focus on supporting the main sources of growth and help achieve the sector objectives without displacing the private sector - policy (sector specific as well as macro and trade), extension, research, infrastructure spending etc. should aim at increasing agricultural production and diversification to raise rural incomes, reduce poverty and improve national food security; raising the output of marketed commodities and livestock products to increase employment opportunities and incomes in production, commerce and agro- processing, and to expand export earnings; and promoting farming systems, practices and institutions that will ensure sustainable use of the natural resources. 9.22 The government should introduce strict budgetary discipline, in which budget should be driven by activities, and be implemented according to plan. The present budget orientation should also change by balancing the weights given to personnel emoluments (PEs) and recurrent departmental charges (RDCs). Allocation to budgetary units of government should be released according to program; units should report on their budgetary utilization regularly and the budget should be reviewed every quarter incorporating the changes in the donors funding so that realistic budgets are prepared and compared with performance. 9.23 The existing inconsistencies between budget and its implementation and among departments should be redressed to ensure correct budget management and balanced resource allocation for growth and development of the agriculture sector. 9.24 It is very urgent that the financial operation system of different agencies is regularized and monitoring and reporting mechanisms are improved/established. 9.25 In addition to support from the public sector, the sector objectives have to be met through further expansion of private sector involvement in the sector, particularly, private sector provision of services for farmers; further encouragement of private sector involvement of agricultural input and output markets; and improvement in the effectiveness of public sector agricultural services and increasing their impact on national and household food security and sustainable use of agricultural lands and water resources. 101 9.26 Commercialization of some of the publicly provided services, cost-sharing with users of the service and partnership, on a joint-venture basis, with the private sector where it is feasible be vigorously pursued. The government should limit its intervention and should facilitate private sector participation in input supply and output marketing by providing appropriate regulatory framework, infrastructure and market information. Any government support to targeted groups such as farmers in outlying areas should be made in a transparent manner. Conclusions and Recommendations 9.27 Growth in the vital agriculture sector can be enhanced: (i) by improving the performance of existing individual farming units, adding more value to land and labor through the application of available technologies, developing new varieties, and introducing better farming and resource conservation practices; (ii) by encouraging the expansion of cultivated area (less than one-fifth of the estimated 10 million hectares of cultivable land is currently utilized); and (iii) by integrating smallholder farms to the markets and transforming their subsistence orientation to more commercial operations. 9.28 Public expenditure for agriculture should broadly focus on: * removing the bottlenecks for agriculture developments (e.g. transport costs, extension services etc.) with farmers and traders; X supporting growth without displacing the private sector through various measures - policy (sector specific and macro and trade), extension, research, infrastructure spending, promotion of farming practices and institutions etc. that all aim at increasing agricultural production and diversification and raising the output of marketed commodities and livestock products. 9.29 Policy reforms should include: * FRA should divest from fertilizer marketing and refocus its mandate to maintenance of strategic food reserves. * Increase the delivery of public agricultural services to resource-poor smallholders by expanding local participation and ownership of agricultural programs. * Improve the rural road network by increased maintenance. The availability of cheap and reliable transportation is also an important factor. * Encourage small scale irrigation schemes. * Maintain good policies in the areas of macro, trade, and real exchange rate, which have great indirect effects on agriculture outputs and exports. The availability of cheap and reliable transportation is also important factor. 102 Box 9.1: Zambia's Agricultural Sector Investment Program (ASIP) A key feature of ASIP is a clear strategy and policy framework aimed at increasing private sector participation and attaining Government's main agricultural objectives: improve household food security, promote better use of natural resources, generate employment and raise incomes through greater agricultural production and productivity, provide locally produced agro-based materials for further processing and increase export earnings and contribute to the balance of payments. Started in 1995, it was originally part of the liberalization and deregutation efforts that removed substantial state interventions, controls and subsidies. Although slower than expected, the emergence of private sector players in various aspects of agriculture and the recent diversification of agricultural products and exports are some of the main results. Lack of access to agricultural credit by farmers remains an unresolved constraint with the collapse of traditional public credit institutions and the lack of vitality and reforms in the banking sector and rural finance. The agricultural reform program was complemented later by the Rural Investment Fund (RIF) to tackle feeder roads and transport problems in rural areas and to improve the participation of local communities. ASIP is guided by sector investment program (SIP) concepts and principles outlined in Box 4.3. However, application of several of these principles proved more difficult than expected. It took much longer for the Ministry to establish its new institutional framework, and donors were hesitant in moving towards common implementation arrangement in the absence of proven institutional capacity. The national process of public sector reform also moved slowly and dampened the momentum for a common implementation arrangement. In designing a program in the future, the vision and the policy framework for the sector has to be clear and, the planning and coordination mechanisms necessary to guide the rational allocation of resources defined. Effort has to be made to minimize the dangers of counterproductive donor competition, duplication and the creation of parallel structures, by setting guidelines aimed at harmonizing activities, within sub-sectors and geographical areas. Deepening the improvements in financial management, procurement and management information systems is necessary to raise the efficiency with which internal and external resources would be managed and program performance would be monitored. Donors support in the sector should be consistent with the government priority and strengthen the institutional structures and management systems established for program implementation. Further strengthening the public-private sector partnerships, decentralization and empowering communities to plan and implement development initiatives. While effort can be made to have a country-wide program, the growing duality of the agricultural sector requires that, on food security and social equity grounds, publicly managed and funded agricultural services would be increasingly focused on resource-poor smallholders. 103 1 1. INFRASTRUCTURE Introduction 11.1 Zambia's challenge of attaining broad based economic growth and sustainable poverty reduction will depend, to a large extent, on the state and spatial distribution of the economic and social infrastructure in the country. Moreover, it is shown in the macro section that the efficacy of infrastructure spending will be enhanced with significant policy and institutional reformns. At the micro level, it is clear that the geographical distribution of population and economic activities has followed the physical location of roads, railways and other infrastructure and that the rural poor tended to live away from these facilities (compare Map 1 and 2 at the end of the document). Both private sector development and the delivery of quality government services are facilitated by functional basic infrastructure such as roads, railways, air transport, national electricity grid, telecommunications, etc. In particular, given Zambia's desire to pursue an export-led growth strategy, the level and quality of its infrastructure development will have a direct bearing on its global competitiveness as it has a direct impact on the cost of doing business. In this context, the effectiveness of public expenditure plays an important and key role in infrastructure (transport and energy) development. The State of Zambia's Infrastructure 11.2 Zambia's current infrastructure development indicators are quite poor. For example, while the country has an installed capacity of 1,670 MW, which is in excess of the country's demand, and an estimated hydro-power potential of 6,000 MW, only about 20% of the population have access to electricity, the rest depends largely on wood fuel. Figure 10.1 shows that the structure of electricity consumption has not changed since the mid-1980s as the mining sector continues to account for 70% of total electricity consumption. The combined electricity consumption of the non-mining productive sector (agriculture, industry/commerce and transport) in 1999 was much lower than that consumed by households. This demonstrates a further decline in Zambia's industrial development and hence the major challenge of turning around the economy to a sustainable growth path. To tackle the development and maintenance of roads as well provide a coherent framework, the Govemment and the donors have initiated a Road Sector Investment Program (see Box 10.1). 104 Figure 10.1 Electricity Consumption by Sector 1986 and 1999 80 70 -_ _ _ _ _ _ _ _ _ 60 - OAgriculture 50 - * Industry/Commerce 40 - 1 mMining 30 3 OTransport 20 - M Households 10- vl Govemment 0 1986 1999 Source. Republic of Zambia, Energy Statistics Bulletin 1980-1999, October 2000. 11.3 Zambia's energy infrastructure development challenges are further demonstrated in Figure 10.2 by the final energy consumption (in thousand tons of oil equivalent) which show the predominance of wood fuel (charcoal and firewood) as the dominant sources of energy. This energy consumption pattern has major adverse environmental implications. Figure 10.2 Final Energy Consumption 5 5000.0 X 4500.0 5 4000.0- .4 3500.0 - U Petroleum 9 2000.0 - Electricity 0 .O 2000.0 . 00* 1500.0 - Woodfuel tA 1000.0 fTotal .° 500.0 c 0.0 1986 1990 1995 1999 Source: Republic of Zambia, Energy Statistics Bulletin 1980-1999, October 2000. 11.4 The major issues in the energy sub-sector which have direct implications on public expenditure include the following: * Quasi-fiscal deficit. The procurement, transportation and processing of the oil feedstock is undertaken by state-owned enterprises (Zambia National Oil Company (ZNOC), Tanzania Zambia Pipeline and Indeni Refinery) whose commercial credit rating is poor and can only borrow commercially 105 through Government guarantees. Being a state-owned enterprise, ZNOC does not operate under the profit maximization principle and, consequently, incurs huge losses which it cannot sustain and hence the calling of Government guarantees by its creditors. To the extent no formal provision was made for such guarantees in the Government budget in the past, it formed part of the huge quasi-fiscal deficit which the budget had to bear. * High cost operations. Despite having the cheapest source of electricity (hydro), the cost of electricity in Zambia is unnecessarily high on account of inefficiency arising from, among other things, high overheads, low re- investment and poor maintenance, weak management regimes, etc. which translate into high consumer prices. * Limited access to energy. As stated above, only about 20% of the population have access to electricity and that 70% of domestic consumed electricity power is accounted for by the mining industry. Given such a dominance of the mining industry in the energy sector, it has monopsony power which has adversely affected the expansion of the electricity infrastructure to other productive sectors such as agriculture. * Uneven spatial distribution of energy infrastructure. Both the national grid and the petroleum products distribution network are biased towards the urban areas. Rural districts rely on diesel generators whose cost has increased substantially. Since there is a limit to which electricity tariffs can be increased at any given time, ZESCO's capacity to raise necessary resources to ensure continuous electricity supply in rural areas is limited. 11.5 In the transport sub-sector, the current road network is still very small relative to the size of the country and the needs of the population. For example, out of the 37,000 kilometers of gazetted road network, only 6,316 km (17%) are bituminous and surfaced to class 1 standard; 8,478 km (22%) are gravel and 21,679 km (58%) are earth roads. In addition, there are about 30,000km of ungazetted comnmunity road network comprising tracks, trails and footpaths. Given Zambia's sparse population distribution, the challenge for increasing the gravel and asphalt road network necessary for enhancing broad based growth is quite significant. However, the Government's current focus is on existing road network rehabilitation and not necessarily the construction of new roads. 11.6 The transport sub-sector also has a number of challenges which have direct bearing on public expenditure and for which urgent remedies should be sought:5' * High cost of transport. Transport costs in Zambia are excessive with road tariffs ranging from US$0.07 to US$0.10 per net ton kilometer. Zambia Railways users, for example, face further costs caused by unpredictability of the schedules, delays, improper handling and storage, derailments and accidents. For road haulage, high rates of accidents, lack of investment in 51 For further discussion of railways operational problems in Zambia, see World Bank, Project Appraisal Document for Railways Restructuring Project, Report No. 21073-ZM, October 18, 2000. 106 safety equipment and inadequate enforcement of safety regulations also contribute to high costs of road transport. Such high transport costs adversely affect the competitiveness of Zambian exports. Non-traditional exports are further adversely affected by lack of a national wide-bodied and long haul airline service to transport perishable cargo expeditiously. * Uneven distribution of public expenditure among different modes of transport. Railway companies are responsible for the maintenance of both the rail tracks and their rolling stocks. On the other hand, road hauliers who compete directly with railway system do not directly pay for road maintenance. The government road users charges for large capacity vehicles have little relationship to the high axle loads and the resulting road degradation. As a consequence, and given the current poor state of the rail system in the country, more cargo is transported on the roads but this reduces the life of the road network and hence the high fiscal cost of frequent repairs and rehabilitation. * High dependence on Government financial support. In the road sub-sector, despite the Government's cost recovery efforts, revenues continue to lag behind the road network's capital and maintenance requirements. In the railways sector the Government is no longer providing a direct operating subsidy but has been forced to write off its loans to Zambia Railways Limited as the latter has failed to service them. Continuous neglect of maintenance of the railways assets by the two railway companies has resulted into serious deterioration of the assets to the extent that it would require a substantial injection of private or public capital to prevent the system from total collapse. * Deteriorating condition of railways infrastructure. This arises from three factors: (a) poor quality of maintenance, (b) concentration of huge resources to certain sections of the infrastructure and the development of these sections to unnecessary high standards while starving others of even the minimum of basic maintenance, and (c) adoption of highly sophisticated track and signaling technology which resulted into high dependence on imported parts and skills. The road network has also deteriorated on account of inadequate cost recovery from road users, lack of enforcement of axle loading regulations, and inefficient maintenance (especially for the feeder and non-trunk roads). However, in the wake of the Road Sector Investment Program (ROADSIP), the proportion of the condition of the paved road network which was rated good has increased from 20% in 1995 to 43% in 2000 * Poor service of ruralfeeder roads network. The feeder (gravel) roads for which the poor rely are poorly serviced in terms of regular repair and maintenance (grading). In years where rainfall is above normal (like the 2000/2001 rain season), a lot of feeder roads bridges are washed away and thus, make certain areas inaccessible. To the extent over 80% of the rural 107 population are very poor, the lack of a well maintained feeder roads network further disadvantages them both socially and economically. * Reliance on contractors to maintain the road network. Under ROADSIP, the role of Roads Department to undertake periodic maintenance of trunk roads and local councils to undertake repairs of township and feeder roads have been superceded by contractors. The rationale for this was to achieve cost-effectiveness and quick outturn as the private sector contractors are assumed to be more efficient. However, given the problems of rent seeking and erratic funding releases by the Ministry of Finance to pay such contractors, the outcome seem to have increased the cost of maintaining the road network. The Government continues to incur huge interest payments on outstanding balances and, in some cases, private contractors suspend work if they are not paid on time. Atwell funded and functional Roads Department is better placed to carry out periodic maintenance without the Government incurring unnecessary costs associated with contractors. * Proliferation of institutions involved in road infrastructure. There are a number of institutions responsible for different aspects of the road network. The Ministry of Communications and Transport is responsible for policy formulation and monitoring of the road sector; the Roads Department in the Ministry of Works and Supply superintends the construction of maintenance of the trunk, main and district road network; Department of Infrastructure and Support Services in the Ministry of Local Government and Housing is responsible for over 90% of feeder roads and all council roads; the Ministry of Tourism and, sometimes, the Ministry of Agriculture, Food and Fisheries are also responsible for some feeder roads in their respective sectors. The Social Recovery Project (now Zambia Social Investment Fund) under the Ministry of Finance and Economic Development has been mandated to undertake rehabilitation of community roads. The National Roads Board, a very recent creation, is responsible for the maintenance and rehabilitation of trunk roads. This proliferation of responsibilities for the road network has resulted in an inefficient management of the sub-sector. 11.7 The importance of infrastructure in poverty reduction strategies is demonstrated by the 1998 LCMS results (see the benefit incidence analysis in chapter 2) which show that the poorest income groups do not perceive public expenditure on health (in term of building or rehabilitation of a health facility) as being important in improving the quality of lives. The projects which they rate highly in terms of improving their lives a great deal include: construction of a new road where none existed before; sinking or rehabilitation of a well and/or bore hole; provision of hammer mill; buyer of agricultural produce available; and improving the supply of agricultural inputs in the locality. This demonstrates the fact that the poor value improvements in public health, food security, and infrastructure as being important elements which directly improve their livelihoods. 108 11.8 The level and quality of infrastructure development is one of the key elements in determnining the size of foreign direct investment (FDI) as it has direct impact on investor confidence. It is estimated, for example, that road transport charges increase the cost of intermediate products shipped through Durban (South Africa) by up to 40%. Such high costs are further compounded by long delivery periods which make the Zambian productive sector un-competitive. Given such a situation, a case for increased public expenditure on infrastructure is compelling. However, since certain types of infrastructure development can be appealing to the private sector, the issue is to determine what incentives (fiscal and others) could be put in place to make it attractive for private sector participation. Special Infrastructure Extra-Budgetary Funds 11.9 Given the dilapidated condition of the road network and its consequent adverse impact on the economy a stakeholders' workshop in 1993 recommended for an introduction of fuel levy to be used for timely and pro-active maintenance of the road network. The road fund was to be managed by the National Roads Board through private and public sector partnership. In the same vein, the high concentration of the electricity infrastructure in the urban areas gave rise to creation of the Rural Electrification Fund (REF) in 1994 with the aim of extending the national grid to rural areas that have strong potential for increasing agricultural or industrial/commercial growth; promoting a more equitable regional distribution of income; and improving rural living conditions by providing power to social amenities, such as health facilities, schools, and community centers. 11.10 Initially, the levy for the REF was based on the 3.45 percentage points of the 23 percent sales tax which was charged on all electricity bills. When 20% VAT was introduced in July 1995 and replaced sales tax, the rural electrification levy was reduced to 3 percent of all electricity bills after VAT. Later in 1995 MOFED decided to raise the levy to 10 percent, with 3 percent going to the Rural Electrification Fund and 7 percent being retained by MOFED for other uses (i.e., excise tax). The Zambia Revenue Authority was mandated to collect the 10 percent electricity levy from the Zambia Electricity Supply Corporation (ZESCO) and then deposit the proceeds in the government's revenue account at the Bank of Zambia. The Ministry of Finance was then expected to remit the REF proceeds to the Ministry of Energy and Water Development. More recently, ZESCO remits directly to MOFED. 11.11 In the case of the Road Fund levy, it was initially collected by the Zambia National Oil Company through a 10 percent surcharge on gasoline, now at 13.5 percent, which it then handed over to the Zambian Revenue Authority to deposit the proceeds in the Government revenue accounts at Bank of Zambia (for MOFED). 11.12 Both funds suffer from persistent delays by MOFED to remit the proceeds on time. For example, Tables 10.1 and 10.2 show an accumulation of arrears to the two funds by MOFED from the time of their inception in 1994. 109 Table 10.1. Rural Electrification Releases to Ministry of Energy and Water Development (MEWD) against Remittances from ZESCO to Ministry of Finance. Year ZESCO Remittance to MOFED MEWD/REF Receipts from MOFED 1996 KI,309,646,260 Nil 1997 K2,700,000,000 Nil 1998 K4,360,000,000 K600,000,000 1999 K9,700,000,000 K1,000,000,000 2000 K39,351,784,0231 K800,000,000 2001 K2,600,000,0002 K400,000,000 1K37,924,987,941.78 was paid through debt swap arrangement between MOFED and ZESCO. Therefore, the only amount available for REF allocation was K 1,426,796,084 2includes only January and February 2001 remittances Source. Department of Energy, MEWD. Table 10.2. Fuel Levy Collections and Arrears to the Road Fund, 1994-2001. ZNOC ZNOC ZRA ZRA ZRA ZRA 1994/95 1996 1997 1998 1999 2000 2001 Total January 1,005,498,000 1,289,319,200 2,120,614,770 1,600,273,719 1,513,637,850 3,201,520,428 10,730,863,967 February 926,867,000 1,104,578,000 1,504,785,354 1,565,764,413 2,136,045,539 4,879,731,562 12,117,771,868 March 1,144,384,000 1,134,443,600 1,922,875,681 1,226,324,198 2,071,602,674 4,333,312,127 11,832,942,280 April 1,253,465,000 1,396,725,600 1,818,491,462 1,543,486,094 2,168,164,648 8,180,332,804 May 1,306,995,000 1,288,676,800 1,853,155,791 2,194,053,687 1,392,329,028 8,035,210,306 June 1,357,900,000 1,351,651,029 1,962,963,782 1,938,933,274 1,745,773,268 8,357,221,353 July 1,234,554,000 1,973,473,376 2,359,995,143 2,094,452,348 1,309,498,838 8,971,973,705 August 1,163,046,000 1,991,377,001 1,952,260,220 2,850,825,677 1,675,908,613 9,633,417,511 September 1,113,862,000 2,052,993,183 2,157,409,061 2,458,574,188 1,656,228,095 9,439,066,527 October 1,199,979,000 2,129,302,062 2,129,508,814 1,096,939,353 3,862,851,755 10,418,580,984 November 1,050,486,000 1,877,575,271 1,902,401,276 2,170,842,571 12,136,528,297 19,137,833,415 December 1,106,230,000 1,846,631,562 1,429,116,468 2,718,654,840 3,782,060,719 10,882,693,589 11,913,000,000 13,863,266,000 19,436,746,684 23,113,577,822 23,459,124,362 35,450,629,324 12,414,564,117 139,650,908,309 Total Remittances 11,312,000,000 10,600,000,000 13,500,000,000 23,000,000,000 20,200,000,000 22,600,000,000 6,000,000,000 107,212,000,000 Arrears 601,000,000 3,263,266,000 5,936,746,684 113,577,822 3,259,124,362 12,850,629,324 6,414,564,117 32,438,908,309 Source. National Roads Board. 11.13 According to the National Roads Board (NRB) data, the K32.5 billion arrears have serious cost implications on the repair of the road infrastructure. For example, NRB is unable to commission road projects for its routine repair of roads. Such delays mean that instead of spending an average of K50 million per kilometer for periodic maintenance of a road, a cost of K500 million per kilometer for rehabilitation or reconstruction of the same road could be incurred if the funds are not released timely by MOFED. Further, the following projects which were approved by the Zambia National Tender Board and for which contract documents have already been signed by the Ministry of Local Governmuent and Housing and the Ministry of Works and Supply but are on hold due to non remittance of fuel levy by MOFED: 110 Turnkey projects to rehabilitate main roads in the five mining towns on the Copperbelt Resurfacing of Mpika-Kasama road Rehabilitation and maintenance of feeder roads throughout the country Improvement of urban roads throughout the country Periodic maintenance of the Nyimba-Katete stretch of the Great East Road Resurfacing of Chingola-Soiwezi road. The Rural Electrification Fund projects have also been on hold on account of the non-remittance by MOFED of the electricity levy. 11.14 The Rural Electrification Fund projects have also been on hold on account of the non-remittance by MOFED of the electricity levy. Levels and Composition of Infrastructure Public Spending 11.15 Given her desire to accelerate the development process, Zambia's infrastructure development bill is well beyond what can be afforded by Government revenue. In which case, infrastructure capital development costs have increasingly been financed externally either through loans or development assistance grants. As a result, a significant portion of domestic revenue is directed at recurrent expenditure while external finance is directed at capital expenditure. For example, Table 10.3 shows that for the period 1995-2000, external financing for road works ranged from 61% in 1995 to 93% in the year 2000. The lower figure of 28% for 1998 was on account of donor freeze on aid on account of the delay of the privatization of ZCCM. Table 10.3 Structure of Funding to the Roads Department, 1995-2000 (K million) Year GRZ Actual GRZ Actual Total GRZ Actual as % % Extemal Budget Funding Extemal Funding of Budget Funding Provision Funding Provision 1995 33,940 17,913 28,076 45,989 46 61 1996 57,714 11,338 46,246 57,584 20 80 1997 70,461 22,791 43,772 66,563 32 66 1998 112,919 224,889 85,577 310,466 199 28 1999 221,407 34,727 176,862 211,589 16 84 2000 185,598 12,102 156,000 168,102 7 93 Source. Department of Roads, Ministry of Works and Supply. 11.16 Table 10.3 also shows that GRZ actual funding when compared to the budget provision ranges, in a normal year, from as low as 7% to 46%. It is interesting, however, that when donor resources were not forthcoming in 1998, actual GRZ funding for roads doubled that year. This might suggest a moral hazard or fungibility problem, that is, when GRZ knows that external resources are forthcoming, it diverts its own domestic resources to other uses. However, to the extent the level of external resources cannot be predicted with a great degree of accuracy, excess dependency on such resources has proved to be risky and made strategic planning difficult. 11.17 Being complex and costly capital projects, road works cost estimates cover a longer time horizon in excess of the 12 months Government fiscal year. The budget 111 estimates for road works given by the Roads Department are for the whole contract period which could be eighteen months, while those indicated in the Yellow Book are only for 12 months. However since the actual funding releases by MOFED are very minimal, contractors are usually forced to use such funds for running costs (wages and maintenance of equipment) with no actual work on the projects. 11.18 Since a lot of political pressure is applied on the Ministry of Works and Supply to undertake additional road works, more project estimates are included each year in the budget even when the ongoing projects are not finished. This results into a thin spread of the limited resources and hence tend to minimize the impact on the ground. Consequently, contractors are forced to charge interest on certificates due for payments and make the Government pay for works which have not yet actually been carried out. Table 10.4 Public Spending on Transport by Type, 1995-99. WATER TRANSPORT CIVIL AVIATION RAIL TRANSPORT YEAR BUDGET ACTUAL BUDGET ACTUAL BUDGET ACTUAL 1995 PES 0 0 96,037,001 68,118,325 0 0 RDCs 0 0 200,998,000 106,662,441 0 0 GRANTS 186,512,000 189,230,250 57,407,000 6,833,100 47,657,000 20,635,751 CAPEX 50,636,000 18,697,648 256,941,000 94,943,000 86,793,000 17,000,000 TOTAL 237,148,000 207,927,898 611,383,001 276,556,866 134,450,000 37,635,751 1996 PES 41,222,001 17,261,013 109,915,001 75,065,575 0 0 RDCs 181,460,000 80,622,369 226,094,000 167,911,582 0 0 GRANTS 215,000,002 116,100,000 51,424,000 0 143,000,000 71,500,665 CAPEX 331,100,000 0 140,735,002 51,098,953 0 10,000,000 TOTAL 768,782,003 213,983,382 528,168,003 294,076,110 143,000,000 81,500,665 1997 PES 53,000,001 32,156,906 177,363,001 120,792,549 0 0 RDCs 87,201,000 61,643,052 163,669,000 189,068,070 0 0 GRANTS 119,794,000 119,613,345 0 0 99,490,000 99,342,571 CAPEX 78,031,000 0 320,704,000 515,363,211 36,750,000 27,000,000 TOTAL 338,026,001 213,413,303 661,736,001 825,223,830 136,240,000 126,342,571 1998 PES 115,243,001 23,478,470 176,041,001 61,096,662 0 RDCs 190,717,000 101,961,485 178,296,000 679,878,377 0 GRANTS 341,000,000 181.334,686 70,000,000 0 120,000,000 CAPEX 900,000,000 0 690,000,000 650,793,413 0 TOTAL 1,546,960,001 306,774,641 1,114,337,001 1,391,768,452 120,000,000 0 1999 PES 121,982,001 12,621,483 161,988,001 29,325,616 0 RDCs 168,241,000 68,053,449 171,259,000 123,390,520 0 GRANTS 281,000,000 24,918,972 70,000,000 0 100,000,000 CAPEX 820,000,000 0 500,000,000 561,910,532 0 TOTAL 1,391,223,001 105,593,904 903,247,001 714,626,668 100,000,000 0 Source. Ministry of Finance and Economic Development. 112 1 1.19 With regard to strategic application of limited fiscal resources, Table 10.4 shows that GRZ continues to undertake activities which can easily be done by the private sector. For example, its funding of the Mweru and Bangweulu Water Transport companies in the Luapula province can neither be justified on equity nor efficiency reasons as these are activities which are currently undertaken by the private sector on the same lakes and elsewhere in the country. The Department for Civil Aviation was scheduled to be commercialized and hived-off from the Government but it has yet to be approved and implemented. Table 10.4 also demonstrates efficiency problems associated with Government funding of activities which can easily be taken over by the private sector, that is, paying only for personnel emoluments and other recurrent expenditure at the expense of capital expenditure is not the most efficient way of managing any undertaking. Policy Options for Improving Impact of Public Expenditure on Infrastructure 11.20 The discussion above shows that current levels of public spending on infrastructure are still much lower to have a significant impact on poverty outcomes. The positive impact on poverty outcomes is also adversely affected by weak budget design and execution regimes; lack of prioritization, strategic policy formulation and execution capacity; urban bias; and a weak institutional framework. Given such a situation, there are a number of policy options which the Government could consider in order to improve the poverty outcomes: - Focus on development and implementation of least cost alternatives for access provision. This means not only looking at pavement surfacing alternatives to bitumen and to gravel, but also at standards with regard to road width and other engineering specifications which could be modified to reduce cost, i.e. a given level of access goal could be attained for a far lower cost than is currently the case. v A similar argument can be used for the electricity infrastructure. More poor people can be reached through a solar power program and in mini hydro and alternative sources of energy than through the extension of the national grid. * Privatize utilities as soon as possible and minimize direct Government participation to policy and regulation of utilities. In addition, provide a conducive environment for private sector infrastructure development in areas such as hydro power generation. * Develop a balanced infrastructure network. This requires a revised public investment plan which would follow on from the adoption of a coherent transport policy framework which is expected imminently. The transport policy framework - and associated regulatory reforms to be undertaken by the government especially in the rail and air sectors - are expected to encourage greater private sector participation in management and financing. Further reforms to road user charges, in particular the 113 introduction of economic tariffs for heavy vehicles, are expected to contribute to a more optimal balancing of freight traffic using the full transport infrastructure. * Use special funds proceeds for intended purposes. There is no point in establishing special funds for road maintenance and rural electrification and then starve them of the proceeds from such levies. Immediate attention thus needs to be given: in the short term to clearance of all arrears and to provision of more direct and transparent methods of transferring the agreed resources into these funds; in the medium term to setting in place the required legal and institutional reforms to secure the proper, independent management of these funds and the methods by which future resource requirements of these funds can be met. * The road sector institutional framework has still to be rationalized in order to: address overlapping and unclear mandates; give impetus to the commercialization of road management; provide opportunities for decentralization where appropriate. The revised framework should follow on from the adoption of the transport sector policy and would involve consolidation and further restructuring of the agencies responsible for road management to improve efficiency and build capacity. * In the electricity sub-sector, however, ZESCO also needs to be unbundled, i.e., generation, transmission and distribution operations, respectively, should be undertaken by different companies in order to remove costs associated with such a large natural monopoly. However, this would also require a restructured energy sector regulatory framework. 114 Box 10.1: Zambia's Road Sector Investment Program (ROADSIP) The ROADSIP is a 10-year $1 billion investment program for the road sector. The first phase started in mid-1997 and will end in 2002. IDA money (about $70 million) became effective in March 1998. The implementing agencies include: Ministries of Finance and Economic Development, Communications and Transport, Local Government and Housing, and Works and Supply. Program objectives are: reform of road sector policy and institutional framework (enhance capacity, increase efficiency of use of resources and improve planning and implementation performance); strengthening of road sector financing (Road Fund to cover minimum 70% of maintenance requirements by 2002); development of the local construction and consulting industry (create 14,000 new jobs by 2002); addressing the road maintenance and rehabilitation backlog (increase roads in good condition to 45% for paved and 15% for unpaved by 2002); and establishment of a pilot program for community road management. Status: There are three important concerns. Firstly - Progress has been slower than expected with regard to finalizing the transport sector policy document. Action has been taken on drafting a revision of road sector legislation, but this cannot be finalized until a decision has been taken on the medium term institutional framework. Secondly - Development of the local construction industry is not progressing as well as planned. NCC is now established but has made slow headway on addressing constraints to increase local participation in ROADSIP. Thirdly - Adequacy of Road Fund resources needs close attention and is considered unsatisfactory in relation to the objectives of assuring funding of routine maintenance on the network. Less than 50% of requirements are currently being met and there is no firm strategy in place to address the gap. The mid term review in December 2000 was intended to address these concerns but was not decisive in so doing. Implementation management and monitoring remains satisfactory. An Annual Work Program process has been set in place and is working quite well. Steps have also been taken to improve reporting and the flow of information on ROADSIP implementation to all stakeholders. In addition a locally organized Donor Forum has been established since April 2000 and this is improving coordination of external support. Good progress is being obtained particularly in two areas; first on the environmental management and the second on community road development. Most of the credit is now committed (over US$56 million) and disbursement has accelerated as a result of a number of major contracts coming on stream - at US$42.5 million at end June 2001 which is ahead of forecast. The undisbursed balance as of June 30, 2001, was SDR 19.4 million. 115 REFERENCES Ablo, Emmanuel and Ritva Reinikka (1998). "Do budgets really matter? Evidence from public spending on education and health in Uganda." World Bank Policy Research Working Paper 1926. Adams, Christopher S. and David L. Bevan (2000). " The cash budget as restraint: the experience of Zambia" in Paul Collier and Catherine Patillo, eds., Investment and Risk in Africa. NY: st. Martin's Press, Inc. Agenor, Pierre-Richard, Alejandro Izquierdo, and Hippolyte Fofack, "A quantitative macroeconomic framework for the analysis of poverty-reduction strategies." World Bank mimeo (2001). Arndt, Channing and Jeffrey D. Lewis (2000), 'The macroeconomic implications of HIV/AIDS in South Africa: a preliminary assessment," World Bank Africa Region Working Paper Series 9 (November). Aron, Janine (1999). "The Zambia copper boom and crash, 1964-1980," in Paul Collier and Jan willem Gunning and associates (eds), Trade Shocks in Developing Countries, Vol.1- Africa. Oxford: Oxford University Press. Baldwin, Robert E. (1966). Economic Development and Export Growth: A Study of Northern Rhodesia, 1920-1960. Berkeley: University of California Press. Barro, Robert J. (1997).Determinants of Economic Growth, MIT Press, Mass. Barro, Robert J. (1995). 'Inflation and Economic Growth', Bank of England Quarterly Bulletin, May 1995: 166-76 Bonnel, Rene (2000). 'HIV/AIDS: does it increase or decrease growth in Africa?" Mimeo, World Bank; Bruno, Michael and William Easterly (1998). 'Inflation Crises and Long-Run Growth'. Journal of Monetary Economics 41/1:: 3-26 Castro-Leal, Florencia, Julia Dayton, Lionel Demery, and Kalpana Mehra (1999). "Public social spending in Africa: do the poor benefit?" The World Bank Research Observer, vol. 14, no. 1 (February), pp. 49-72. Chomitz, Kenneth, Gunanwan Setiadi, Azrul Azwar, Nusye Ismail, and Widiyarti (1998). "What do doctors want? Developing incentives for doctors to serve in Indonesia's rural and remote areas." World Bank Policy Research Working Paper 1888. Devaraj an, Shantayanan and Delfin Go (2001), with contributions from Florence Charlier, Andrew Dabalen, William R. Easterly, J. Alejandro Izquierdo, and Leonid Koryukin. "The 1 23PRSP Macroeconomic Framework for Poverty Reduction Strategy Papers - Illustrations using Zambia data," World Bank. Devarajan, Shantayanan, David Dollar, and Torgny Holmgren, eds. (2001). Aid and Reform in Africa -Lessons from Ten Case Studies. World Bank. 116 Devaraj an, Shantayanan ,William R. Easterly, and Howard Pack (2000). "Is investment in Africa too low or too high? Macro and Micro Evidence." World Bank Policy Research Working Paper 2519. Devarajan, Shantayanan, Vinaya Swaroop, and Heng-fu Zou (1996). "The composition of public expenditure and economic growth." Journal of Monetary Economics 37(1996) 313-344. Devarajan, Shantayanan, Karen E. Thierfelder, and Sethaput Suthiwart-Narueput (1999). "The marginal cost of public funds in developing countries." Draft, World Bank. Forthcoming in Amedeo Fasati and Wolfgang Wiegard, eds. Policy Evaluations with Computable General Equilibrium Models. Routledge Press. Dinh, Hinh (1999). "Fiscal solvency and sustainability in economic management," World Bank Policy Research Working Paper No. 2213 (October). Dollar, David and Aart Kraay (2000). "Growth is good for the poor," World Bank, Washington D.C. See also The Economist, May 27, 2000, p.82. Dollar, David and William Easterly (1999). "The search for the key: aid, investment, and policies in Africa." World Bank Policy Research Working Paper 2070. Dollar, David and Lant Pritchett (1998). Assessing aid: what works, what doesn 't, and why. A World Bank Policy Research Report. Easterly, William (2000). "The lost decades: developing countries' stagnation in spite of policy reform." World Bank. Journal of Economic Growth, forthcoming. Easterly, William (1997), "The ghost of financing gap: how the Harrod-Domar growth model haunts development economics." World Bank Working Paper No. 1807. Easterly, William and Stanley Fischer (2000). "Inflation and the poor." World Bank Policy Research Working Paper 2335. Also forthcoming in Journal of Money, Credit, and Banking. Filmer, Deon and Lant Pritchett (1999). "The impact of public spending on health: does money matter?" Social Science & Medicine 49: 1309-23. Filmer, Deon, Jeffrey S. Hammer, and Lant Pritchett (2000). "Weak links in the chain: a diagnosis of health policy in poor countries," The World Bank Research Observer, vol. 15, no. 2 (August), pp. 199-224. See also "Health Policy in Poor Countries: Weak Links in the Chain?," World Bank Policy Research Working Paper No. 1874, (1998.) Filmer, Deon, Jeffrey S. Hammer, and Lant Pritchett (2001), "Weak links in the chain II: a prescription for health policy in poor countries," revised version World Bank Policy Research Working Paper 1874. Fisher, Stanley (1993). "The Role of Macroeconomic Factors in Growth, " Journal of Monetary Economics 32/4; 485-512 Fisher, Stanley (1991). "Growth, Macroeconomics, and Development', NBER Macroeconomics Annual. 329-64. 117 Gelb, Alan, Ali A.G. Ali, Tesfaye Dinka, Ibrahim Elbadawi, Charles Soludo, and Gene Tidrick (2000). Can Africa Claim the 215' Century? The World Bank. Ghosh, Atish R (2000) "Inflation and Growth," IMF Research Bulletin Vol. 1, June Gylfason, Thorvaldur (1999). "Exports, Inflation, and Growth," World Development 27. Gylfason, Thorvadur, Tryggvi Thor Herbertsson, and Gylfi Zoega (1998). "Ownership and growth," CEPR Discussion Paper 1900. (June). Gylffason, Thorvaldur and Tryggvi Thor Herbertsson (1996). "Does Inflation Matter for Growth?", CEPR Discussion Paper 1503. ILO. International Labor Organization (1976). Meeting Basic Needs. Geneva: ILO. International Monetary Fund (2001). "Zambia: Tracking poverty-reducing expenditure --- - assessment and action plan." Fiscal Affairs Department (forthcoming June). International Monetary Fund (2000). "Zambia: Improving public expenditure management." Fiscal Affairs Department (June). International Monetary Fund and International Development Association (2000). Zambia: Decision Point Document for the Enhanced Heavily Indebted Poor Countries (HIPC) Initiative (November). Kambou, Gerald, Shantayanan Devarajan, and Mead Over (1991), "The economic impact of the AIDS crisis in Sub-Saharan Africa: simulations with a computable general equilibrium model," Revue D 'Economie du Developpement (France): 1, No. 93/1:37-62; Levine, Ross and Sara J. Zervos (1993). "What we have learned about policy and growth from cross-country growth regressions?" American Economic Review, v83 n2 (May): 426-30. Mwikisa, C.N. and Lungwangwa, G. (1998). Education Indicators, Costs and Factors Associated with Primary School Effectiveness in Zambia. Report to Ministry of Education, Lusaka, Zambia (April). Mundlak, Yair (2000). Agriculture and Economic Development: Theory and Measurement. Harvard University Press. Oxfam (2000). Break the Cycle of Poverty. A report of Oxfam International's Education Now Campaign. Pritchett, Lant (1977). "Where has all the schooling gone?" World Bank Policy Research Working Paper 1581. Ravallion, Martin (2000). "Growth, inequality and poverty: looking beyond averages." World Bank, Washington D.C. Republic of Zambia (1999). Report of the Auditor Generalfor the 1996 and 1997 on the Accounts of the Food Reserve Agency. Roubini, Nouriel and Xavier Sala-i-Martin (1992). "Financial Repression and Economic Growth," Journal of Development Economics 39/1; 5-30. 118 Sackey, James and Tejaswi Raparla (2000), "Lesotho: the development impact of HIV/AIDS - selected issues and options." World Bank Report No. 21103-LSO plus forthcoming reports on Botswana, Swaziland, and Namibia. Swaroop, Vinaya and Sunil Rajkumar (2001). "Does governance matter? The link between public spending and outcome." World Bank, forthcoming. United Nations (1990). Human Development Report (annual). World Bank (2001). Implementation Completion Report on the Public Sector Reform and Export Promotion Credit. World Bank (2001b). "Action plan for improving fiduciary accountability in Zambia." Draft. World Bank (200 1c). "Constructing knowledge societies: new Challenges for tertiary education." Draft. World Bank (2000). The President Report on the Fiscal Sustainability Credit, Report No. P-7379-ZA. World Bank (2000b). " The cash rationing system in Zambia." Draft, Macroeconomics, Southern Africa Region. World Bank (1998). Zambia: Fiscal Management Report. Bank Report 18552-ZA. World Bank (1997). Zambia Country Assistance Review: Turning an Economy Around. An OED Report by Gladstone G. Bonnick. World Bank (1996). Prospects for Sustainable Growth. Report No. 75477-ZA. World Bank (1995). Republic of Zambia: Public Expenditure Review. Report No. 13854- ZA. World Bank (1993). Prospects for Sustainable and Equitable Growth. Report No. 11570- ZA. World Bank (1987). Republic of Zambia: Public Expenditure Review. Report No. 6438- ZA. 119 Annex 1: Table Al.l: Central Goverrnment Budget, 1990-2000 In K, million 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Revenue and grants 27,927 71,906 188,113 390,342 676,344 870,608 1,058,415 1,282,895 1,529,324 1,921,000 2,527,000 Revenue 23,001 40,814 104,521 235,256 449,616 595,711 816,579 1,022,656 1,131,675 1,324,000 1,952,000 Tax revenue 22,577 39,954 99,819 227,248 418,878 545,711 751,446 966,923 1,093,819 1,290,000 1,930,000 Income taxes 8,821 14,334 37,115 81,346 134,034 190,262 245,069 327,064 398,000 483,000 634,000 Excise taxes 3,266 7,804 15,526 33,067 70,204 84,616 126,623 168,189 211,000 222,000 278,000 Value-added taxes (VAT) 2,182 4,600 12,765 34,394 76,104 107,678 136,854 183,688 200,000 248,000 230,000 Trade taxes 8,308 13,216 34,413 78,441 138,536 163,155 242,900 287,982 284,819 337,000 597,000 Clearance of tax arrears 0 0 0 0 0 0 0 0 0 0 191,000 Non-tax revenue 424 860 4,702 8,008 30,738 50,000 65,133 55,733 37,856 34,000 22,000 Grants 4,926 31,092 83,592 155,086 226,728 274,897 241,836 260,239 397,649 597,000 575,000 Total Current Expenditures (after rescheduling) 30,282 63,997 155,063 389,563 624,735 712,210 734,436 906,782 1,044,749 1,133,169 1,608,000 Interest on External Debt (after rescheduling) 7,275 14,783 33,014 111,155 157,866 167,000 66,000 97,400 123,000 107,000 167,000 Interest on Domestic Debt 1,584 3,662 15,577 71,463 102,627 77,029 121,846 115,228 80,139 105,000 140,000 Transfers, pensions, and subsidies 6,969 14,906 18,391 40,998 89,663 106,997 96,165 130,008 149,000 181,000 219,000 Consumption 11,117 22,218 56,655 129,050 10,212 279,699 340,637 462,995 565,627 645,169 912,000 Wages and Salaries 5,960 13,372 34,045 70,524 115,051 177,900 221,000 324,226 327,273 402,169 538,000 Retrenchments 0 0 1,703 4,101 1,980 1,127 0 2,000 76,957 51,000 74,000 RDCs 5,157 8,846 20,907 54,425 63,181 100,672 119,637 136,769 161,397 192,000 300,000 Other current expenditures 3,337 8,428 31,426 36,897 94,367 81,485 109,788 101,151 126,983 95,000 170,000 Total Capital Expenditures and Net Lending 7,059 23,831 57,000 141,232 229,700 273,256 344,847 389,571 680,430 942,000 1,009,000 Budgetary Investment 7,059 23,831 57,000 141,232 229,700 273,256 344,847 389,571 680,430 790,000 1,009,000 foreign financed 3,059 17,573 47,330 123,000 193,001 215,000 304,251 319,545 567,067 666,000 781,000 domestically financed 4,000 6,258 9,670 18,232 36,699 58,256 40,596 70,026 113,363 124,000 228,000 Net lending 0 0 0 0 0 0 0 0 0 152,000 413,000 foreign financed 0 0 0 0 0 0 0 0 0 152,000 0 domestically financed 0 0 0 0 0 0 0 0 0 0 413,000 Overall balance (cash basis)* -9,155 -19,064 - 19,641 -140,586 -153,384 -113,863 -212,409 -213,381 -266,855 -297,000 -708,000 Domestic balance (cash basis)' -3,747 -17,800 -22,889 -61,517 -29,157 7,940 49,180 64,434 25,185 31,000 -335,000 * Payments of domestic arrears and changes in line ministries' balances are not shown in the table Sources: IMF, World Bank - various statistical annexes 120 Annex I Table Al.2 Central Government Budget, 1990-2000, % of GDP 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Revenue and grants 24.6 32.9 33.0 26.3 30.2 29.0 26.8 25.0 25.4 25.9 27.9 Revenue 20.3 18.7 18.4 15.9 20.1 19.9 20.7 19.9 18.8 17.9 21.6 Tax revenue 19.9 18.3 17.5 15.3 18.7 18.2 19.0 18.8 18.1 17.4 21.3 Income taxes 7.8 6.6 6.5 5.5 6.0 6.3 6.2 6.4 6.6 6.5 7.0 Excise taxes 2.9 3.6 2.7 2.2 3.1 2.8 3.2 3.3 3.5 3.0 3.1 Value-added taxes (VAT) 1.9 2.1 2.2 2.3 3.4 3.6 3.5 3.6 3.3 3.3 2.5 Trade taxes 7.3 6.1 6.0 5.3 6.2 5.4 6.1 5.6 4.7 4.5 6.6 Clearance of tax arrears 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.1 Non-tax revenue 0.4 0.4 0.8 0.5 1.4 1.7 1.6 1.1 0.6 0.5 0.2 Grants 4.3 14.2 14.7 10.5 10.1 9.2 6.1 5.1 6.6 8.1 6.4 Total Current Expenditures (after rescheduling) 26.7 29.3 27.2 26.3 27.9 23.7 18.6 17.6 17.3 15.3 17.8 Interest on External Debt (after rescheduling) 6.4 6.8 5.8 7.5 7.0 5.6 1.7 1.9 2.0 1.4 1.8 Interest on Domestic Debt 1.4 1.7 2.7 4.8 4.6 2.6 3.1 2.2 1.3 1.4 1.5 Transfers, pensions, and subsidies 6.1 6.8 3.2 2.8 4.0 3.6 2.4 2.5 2.5 2.4 2.4 Consumption 9.8 10.2 9.9 8.7 8.0 9.3 8.6 9.0 9.4 8.7 10.1 Wages and Salaries 5.3 6.1 6.0 4.8 5.1 5.9 5.6 6.3 5.4 5.4 5.9 Retrenchments 0.0 0.0 0.3 0.3 0.1 0.0 0.0 0.0 1.3 0.7 0.8 RDCs 4.6 4.1 3.7 3.7 2.8 3.4 3.0 2.7 2.7 2.6 3.3 Other current expenditures 2.9 3.9 5.5 2.5 4.2 2.7 2.8 2.0 2.1 1.3 1.9 Total Capital Expenditures and Net Lending 6.2 10.9 10.0 9.5 10.3 9.1 8.7 7.6 11.3 12.7 11.1 Budgetary investment 6.2 10.9 10.0 9.5 10.3 9.1 8.7 7.6 11.3 10.7 11.1 Foreign financed 2.7 8.1 8.3 8.3 8.6 7.2 7.7 6.2 9.4 9.0 8.6 domestically financed 3.5 2.9 1.7 1.2 1.6 1.9 1.0 1.4 1.9 1.7 2.5 Net lending 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.1 4.6 Foreign financed 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.1 0.0 domestically financed 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 4.6 Overall balance (cash basis) -8.1 -8.7 -3.4 -9.5 -6.8 -3.8 -5.4 -4.2 -4.4 -4.0 -7.8 Domestic balance (cash basis) -3.3 -8.2 -4.0 -4.2 -1.3 0.3 1.2 1.3 0.4 0.4 -3.7 GDP,K billion 113 218 570 1,482 2,241 2,999 3,951 5,141 6,029 7,412 9,055 121 Annex I Table A1.3 Structure of the Central Government Budget, 1990-2000 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Revenue and grants, % of Total Expenditures 74.8 81.9 88.7 73.5 79.2 88.3 98.1 99.0 88.6 92.6 96.6 Revenue, % of Revenue and Grants 82.4 56.8 55.6 60.3 66.5 68.4 77.2 79.7 74.0 68.9 77.2 Tax revenue, % of Revenue 98.2 97.9 95.5 96.6 93.2 91.6 92.0 94.6 96.7 97.4 98.9 Income taxes, % of Tax Revenue 39.1 35.9 37.2 35.8 32.0 34.9 32.6 33.8 36.4 37.4 32.8 Excise taxes, % of Tax Revenue 14.5 19.5 15.6 14.6 16.8 15.5 16.9 17.4 19.3 17.2 14.4 Value-added taxes (VAT), % of Tax Revenue 9.7 11.5 12.8 15.1 18.2 19.7 18.2 19.0 18.3 19.2 11.9 Trade taxes, % of Tax Revenue 36.8 33.1 34.5 34.5 33.1 29.9 32.3 29.8 26.0 26.1 30.9 Clearance of tax arrears, % of Tax Revenue 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 9.9 Non-tax revenue, % of Revenue 1.8 2.1 4.5 3.4 6.8 8.4 8.0 5.4 3.3 2.6 1.1 Grants, % of Revenue and Grants 17.6 43.2 44.4 39.7 33.5 31.6 22.8 20.3 26.0 31.1 22.8 Total Current Expenditures (after rescheduling), % of Total Expenditures 81.1 72.9 73.1 73.4 73.1 72.3 68.0 69.9 60.6 54.6 61.4 Interest on External Debt (after rescheduling), % of Total Current Expenditures 24.0 23.1 21.3 28.5 25.3 23.4 9.0 10.7 11.8 9.4 10.4 Interest on Domestic Debt, % of Total Current Expenditures 5.2 5.7 10.0 18.3 16.4 10.8 16.6 12.7 7.7 9.3 8.7 Transfers, pensions, and subsidies, % of Total Current Expenditures 23.0 23.3 11.9 10.5 14.4 15.0 13.1 14.3 14.3 16.0 13.6 Consumption, % of Total Current Expenditures 36.7 34.7 36.5 33.1 28.8 39.3 46.4 51.1 54.1 56.9 56.7 Wages and Salaries, % of Consumption 53.6 60.2 60.1 54.6 63.8 63.6 64.9 70.0 57.9 62.3 59.0 Retrenchments, % of Consumption 0.0 0.0 3.0 3.2 1.1 0.4 0.0 0.4 13.6 7.9 8.1 RDCs, % of Consumption 46.4 39.8 36.9 42.2 35.1 36.0 35.1 29.5 28.5 29.8 32.9 Other current expenditures, % of Total Current Expenditures 11.0 13.2 20.3 9.5 15.1 11.4 14.9 11.2 12.2 8.4 10.6 Total Capital Expenditures and Net Lending, % of Total Expenditures 18.9 27.1 26.9 26.6 26.9 27.7 32.0 30.1 39.4 45.4 38.6 Budgetary Investment, % of Total Capital Expenditures and Net Lending 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 83.9 100.0 foreign financed, % of Budgetary Investment 43.3 73.7 83.0 87.1 84.0 78.7 88.2 82.0 83.3 84.3 77.4 domestically financed, % of Budgetary Investment 56.7 26.3 17.0 12.9 16.0 21.3 11.8 18.0 16.7 15.7 22.6 Net Lending, % of Total Capital Expenditures and Net Lending 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 16.1 0.0 foreign financed,% of Net Lending 100.0 0.0 domestically financed, % of Net Lending 0.0 100.0 Overall balance (cash basis), % of Revenues and Grants -32.8 -26.5 -10.4 -36.0 -22.7 -13.1 -20.1 -16.6 -17.4 -15.5 -28.0 Domestic balance (cash basis), % of Revenues -16.3 -43.6 -21.9 -26.1 -6.5 1.3 6.0 6.3 2.2 2.3 -17.2 122 Annex I Table A1.4. Zambia: Allocated and Actual Expenditures for Key Sectors, 1990-2000 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 EXPENDITURES (in million Kwacha) I. Education Total Allocation 231.4 1916.8 5348.5 13440.7 55754.3 78037.4 98950.8 130828.3 192900.2 235049.4 382826.0 Personnel Emoluments 2.9 204.0 358.1 690.0 17200.1 28538.9 50360.0 57424.4 91898.3 100302.9 136768.9 Recurrent Departmental Charges 1.3 441.0 1542.9 3585.7 12115.3 14408.1 14197.4 24647.4 25474.3 39099.1 36497.6 Grants and Other Payments 52.1 73.0 1527.7 4610.3 8812.3 13567.2 15651.3 20559.7 25185.2 32132.1 37632.5 Capital Expenditure 175.1 1198.7 1919.8 4554.8 17626.6 21523.2 18742.1 28196.9 50113.8 63514.9 171927.0 Total Expenditure 524.2 1081.9 5321.5 13534.6 47394.7 126865.0 102100.7 208964.4 139539.0 156505.8 231503.7 Personnel Emoluments 28.5 108.7 199.1 630.7 27311.1 83093.9 57718.6 162372.6 94762.9 109623.3 135121.0 Recurrent Departmental Charges 172.4 185.9 1637.5 3103.4 9693.0 13997.2 19139.2 22439.5 21084.3 22724.1 37335.4 Grants and Other Payments 158.8 422.0 2673.4 8223.4 9095.1 12637.1 17646.7 17108.5 22879.3 23991.2 33109.8 Capital Expenditure 164.4 365.3 811.5 1577.2 1295.5 17136.8 7596.2 7043.8 825.1 166.7 25937.5 2. Health Total Allocation 535.1 2191.7 5135.6 17480.0 56506.7 84302.3 122273.4 170442.0 194186.3 254469.8 282227.4 Personnel Emoluments 34.4 111.6 216.0 353.1 6684.9 9057.1 15528.0 19471.0 24440.6 34044.2 75204.9 Recurrent Departmental Charges 44.1 686.6 1389.2 220.3 5479.8 7475.8 8197.7 13915.3 14982.0 24337.8 28055.6 Grants and Other Payments 270.9 908.1 2188.2 10241.1 31543.5 39124.7 43690.3 49638.4 71852.5 77834.3 50640.0 Capital Expenditure 185.8 485.3 1342.3 6665.6 12798.5 28644.7 54857.4 87417.3 82911.1 118253.6 128326.9 Total Expenditure 1320.4 3154.1 5856.3 14723.0 49658.6 62768.7 82182.2 100194.0 118303.0 116185.6 140950.2 Personnel Emoluments 45.7 199.4 223.8 1334.3 10482.1 16155.1 28100.1 36443.2 51368.7 63315.5 32431.0 Recurrent Departmental Charges 569.3 1304.3 2365.7 4822.5 5983.4 22694.4 20500.1 14580.2 13691.8 10018.5 32340.5 Grants and Other Payments 563.0 1396.5 2763.8 7223.5 32790.7 23054.3 32500.0 47108.6 85969.1 31551.7 41104.7 Capital Expenditure 142.4 253.9 502.8 1342.7 402.5 864.9 1082.0 2062.1 2194.9 11300.0 35074.0 3. Agric., Food & Fisheries Total Allocation 74.8 446.4 3128.5 10921.0 53171.4 53045.9 87389.8 110687.4 74088.2 80832.0 105178.1 Personnel Emoluments 31.8 151.5 339.5 549.4 2723.9 1485.1 11504.3 22706.1 10732.1 10341.0 11435.1 Recurrent Departmental Charges 26.1 136.1 400.5 2142.7 6074.0 9832.7 43214.3 46153.3 17694.5 28604.0 64318.8 GrantsandOtherPayments 2.0 36.6 127.8 402.7 21232.0 321.7 1911.2 11446.0 14925.2 5800.0 910.1 Capital Expenditure 14.9 122.1 2260.7 7826.2 23141.4 41406.4 30760.0 30382.0 30735.0 24959.5 105178.1 Total Expenditure 681.2 885.8 1852.3 4904.0 8473.0 5801.3 17660.0 52940.8 39188.5 31478.4 30521.0 Personnel Emoluments 167.0 131.5 390.3 775.1 1455.4 1634.7 6737.6 11045.9 14692.0 6798.2 8624.8 Recurrent Departmental Charges 147.4 132.3 378.5 1785.6 2605.1 1339.0 7887.6 20648.4 17579.4 12530.0 21542.5 Grants and Other Payments 16.9 29.0 166.3 361.3 2457.6 360.1 1002.5 1604.8 2134.1 199.0 130.1 Capital Expenditure 350.0 593.0 917.1 1982.0 1954.9 2467.5 2032.4 19641.6 4782.7 1773.8 223.6 123 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 EXPENDITURES 4. Comm. & Transport Total Allocation 328.3 278.7 599.8 1342.2 3253.5 3314.9 3324.4 15700.7 20957.6 26316.9 30299.8 Personnel Emoluments 36.1 70.4 153.3 187.1 384.2 537.4 539.8 945.8 926.1 865.4 1720.9 Recurrent Departmental Charges 45.8 79.4 132.5 338.1 888.5 1224.9 1359.1 718.2 1104.2 1392.3 560.6 Grants and Other Payments 36.6 8.0 17.6 408.4 613.9 537.8 654.1 471.0 711.4 281.0 958.3 Capital Expenditure 209.8 121.0 296.4 408.7 1366.9 1014.8 771.4 13565.8 18215.8 23239.0 27060.0 Total Expenditure 128.7 219.4 432.9 982.9 983.0 1929.8 1482.5 8234.1 6950.6 11839.9 33168.0 Personnel Emoluments 23.1 54.6 63.3 168.3 168.4 531.1 37.9 721.0 874.6 913.2 1548.7 Recurrent Departmental Charges 52.3 77.8 131.1 301.7 301.7 642.4 865.9 1131.9 1839.7 1901.4 3089.5 Grants and Other Payments 5.7 4.5 64.3 380.2 380.3 405.3 310.3 429.9 416.3 24.9 651.5 Capital Expenditure 47.6 82.5 174.2 132.7 132.7 351.1 268.4 5951.4 3819.6 8659.9 27878.3 5. Works & Supply Total Allocation 687.7 1896.8 3062.5 11976.2 32052.9 46339.6 64114.2 77942.4 119844.7 241137.0 206178.4 Personnel Emoluments 48.0 204.0 307.3 805.6 1322.0 1557.9 1655.2 1890.0 1822.8 2346.1 3095.7 Recurrent Departmental Charges 144.5 441.0 1144.2 2335.2 2604.5 3100.2 2647.2 1346.7 2313.5 4230.5 8713.1 Grants and Other Payments 0.0 73.0 57.0 249.0 446.8 466.1 340.0 382.4 185.0 1250.0 1320.0 Capital Expenditure 495.3 1178.7 1554.0 8586.5 27679.6 41215.4 59471.7 74323.3 115523.3 283310.2 193049.6 Total Expenditure 735.2 1809.5 11545.7 14541.0 8039.2 13959.0 15358.2 31605.3 34484.1 64751.9 67625.7 Personnel Emoluments 66.5 179.0 238.3 608.4 1033.5 1389.0 1518.2 2112.7 2738.8 1938.4 2708.5 Recurrent Departmental Charges 137.7 387.7 1785.2 1768.7 1998.7 2059.5 1617.0 944.6 3710.9 2666.9 6167.8 Grants and Other Payments 62.8 85.0 100.8 181.4 364.8 307.3 217.0 176.1 216.5 645.3 611.5 Capital Expenditure 468.2 1157.9 9421.5 11982.4 4642.1 10203.3 12006.0 28371.9 27817.9 59501.3 58137.9 6. Energy & Water Total Allocation 328.2 424.3 1206.1 2575.4 15437.4 31813.6 35402.3 28728.8 35311.6 34347.5 20888.4 Personnel Emoluments 22.4 30.2 73.8 148.2 237.5 331.7 517.5 649.4 757.1 710.7 815.8 Recurrent Departmental Charges 26.7 27.5 151.6 450.5 72.5.3 761.2 993.3 550.8 1259.4 1594.9 3275.0 Grants and Other Payments 0.0 0.0 3.5 25.5 45.2 67.1 176.3 136.9 617.0 621.0 1340.0 Capital Expenditure 279.1 366.5 977.2 1951.2 14429.4 30653.6 33715.3 27391.8 32678.2 31420.8 15457.6 Total Expenditure 87.7 117.9 470.4 1069.0 3108.2 1917.1 2120.5 1750.0 4226.5 2381.0 4455.4 Personnel Emoluments 16.1 33.7 112.1 211.6 224.1 248.2 328.4 479.1 548.4 313.7 572.0 Recurrent Departmental Charges 26.2 19.3 105.0 376.8 487.0 482.0 478.3 467.7 1033.1 404.9 1582.1 Grants and Other Payments 0.0 0.0 3.5 22.0 32.3 8.2 153.9 189.0 594.3 249.6 736.6 Capital Expenditure 45.4 64.8 249.9 458.7 2364.8 1178.7 1159.9 614.2 2050.8 1412.8 1564.7 Source: GRZ, Financial Reports, various issues. 124 Annex I Table A1.5. Zambia: Selected Macroeconomic Indicators, 1990-2000 DESCRIPTION SCALE 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 1. National Accounts 1.01 Gross National Product (GNP), real, $1995 prices Mil 3,376 3,303 3,326 3,657 3,321 3,221 3,480 3,638 3,535 3,675 3,802 1.02 Gross Domestic Product (GDP), real, $,1995 prices Miu 3,705 3,704 3,640 3,887 3,552 3,463 3,691 3,813 3,742 3,818 3,950 1.03 VA in agriculture, real, $, 1995 prices Mil 439 462 309 520 422 562 559 530 537 574 584 1.04 VAin industry, real, $,1995 prices Mil 1,415 1,398 1,532 1,410 1,271 1,093 1,100 1,178 1,051 998 1,044 1.05 VA in services, real, $, 1995 prices Mil 1,466 1,556 1,527 1,576 1,425 1,393 1,602 1,666 1,738 1,853 1,963 1.06 GDP, nominal, $ Mil 3,742 3,377 3,307 3,250 3,347 3,463 3,273 3,910 3,238 3,104 3,443 1.07 Total consumption, % GDP % 83.4 91.6 99.7 91.8 92.6 87.8 94.7 90.6 96.1 100.9 96.9 1.08 General GovemmentConsumption, % GDP % 19.0 31.8 15.0 18.4 13.1 12.5 11.4 10.9 11.2 9.7 10.6 1.09 Gross Domestic Investment, % GDP % 17.3 11.0 11.9 15.0 8.2 15.9 12.8 14.6 16.4 17.9 18.3 1.10 Gross Public Investment, %GDP % 6.2 7.8 6.7 4.5 4.0 5.1 6.0 5.4 9.6 10.7 11.0 1.11 Gross Private Investment, % GDP % 7.2 3.5 3.8 7.0 7.3 7.3 5.2 7.7 5.2 5.7 5.7 1.12 Gross Domestic Savings, % GDP % 16.6 8.4 0.3 8.2 7.4 12.2 5.3 9.4 3.9 -0.9 3.1 1.13 Gross National Savings, % GDP % 6.9 -4.2 -10.6 0.7 -0.2 4.6 -1.5 3.7 -3.6 -6.5 -2.3 1.14 Resource Balance, % GDP % -0.7 -2.6 -11.6 -6.8 -0.8 -3.8 -7.5 -5.2 -12.5 -18.9 -15.2 1.15 Exports of Goods and Non-factor Services, nominal $ Mil 1,343 1,169 1,204 1,091 1,205 1,250 1,025 1,177 865 701 890 1.16 Imports of Goods and Non-factor Services, nominal $ Mil 1,369 1,258 1,586 1,313 1,232 1,380 1,272 1,381 1,268 1,287 1,333 Exports of Goods and Non-factor Services, real, $, 1.17 1995 prices Mil 1,228 1,284 1,330 1,376 1,497 1,250 1,333 1,589 1,668 1,750 1,837 Imports of Goods and Non-factor Services, real, $, 1.18 1995 prices Mil 1,361 1,106 1,164 1,078 1,214 1,380 1,319 1,382 1,446 1,469 1,575 1.19 GDP growth, real, % p.a. % -0.5 0.0 -1.7 6.8 -8.6 -2.5 6.6 3.3 -1.9 2.0 3.5 1.20 GNP per capita, Atlas $ Units 440 400 370 380 360 350 370 370 330 320 .. 1.21 Total consumption per capita, current $ Units 401 386 399 351 355 339 336 375 322 317 280 2. Prices 2.01 GDP deflator, 1995=100, LC Units 3.5 6.8 18.1 44.0 72.9 100.0 123.6 155.7 186.0 224.2 264.7 2.02 GDP deflator, 1995=100, $ Units 101.0 91.2 90.9 83.6 94.3 100.0 88.7 102.5 86.5 81.3 87.2 2.03 CPI, 1995=100 Units 3.2 6.2 16.8 48.3 74.1 100.0 143.1 178.1 221.6 281.0 365.5 2.04 Official Exchange rate, LC/$ Units 30.3 64.6 172.2 456.0 669.4 866.0 1207.0 1315.0 1862.0 2388.0 3110.8 2.04 Real Effective Exchange Rate index, 1990=100 Units 100.0 93.2 89.3 101.4 97.5 93.4 97.7 117.0 127.4 134.4 137.8 3. Balance of Payments 3.01 Merchandise exports, FOB, $ Mil 1,264 1,085 1,120 994 1,067 1,186 993 1,119 816 756 789 3.02 Merchandise imports, FOB, $ Mil 889 952 1,302 1,019 1,003 1,194 1,056 1,218 971 871 1,008 3.03 Exports of Total Services, $ Mil 79 87 94 107 123 152 143 228 142 130 137 3.04 Imports of Total Services, $ Mil 799 693 623 530 566 588 490 622 536 480 488 3.05 Net Private Transfers, $ Mil -43 -40 -33 -19 -19 -20 -17 -16 -27 -16 -18 3.06 Net Official Current Transfers, $ Mil 00 00 00 00 00 00 00 00 00 00 00 Curent Account Balance, excluding net capital grants, 3.07 $ Mil -388 -513 -743 -467 -398 -464 -427 -509 -576 -481 -588 Current Account Balance, excluding net capital grants, 3.08 % GDP % -10.4 -15.2 -22.5 -14.4 -11.9 -13.4 -13.0 -13.0 -17.8 -15.5 -17.1 3.09 Net Capital Grants, $ Mil 296 521 635 340 338 318 304 198 221 214 185 3.10 Net Foreign Direct Investment, $ Mil 00 00 00 03 40 97 117 207 198 163 126 3.11 Net Long-term borrowing, $ Mil -149 -20 109 154 -97 -104 -49 09 -45 121 14 3.12 Other Capital Flows Mil -109 -79 -260 -212 43 -62 -27 -31 -216 -190 106 125 3.13 Use of Reserves Mil -1.16 -31 -76 -17 -77 1,210 -52 -33 211 -19 -51 3.14 Export unit values, index, 1995=100 Units 109.4 91.0 90.5 79.3 80.5 100.0 76.9 74.1 51.8 40.1 48.5 3.15 Import unitvalues, index, 1995=100 Units 100.6 113.8 136.3 121.9 101.5 100.0 96.5 99.9 87.7 87.6 84.6 3.16 Terms of Trade, index, 1995=100 Units 108.7 80.0 66.4 65.1 79.3 100.0 79.7 74.2 59.1 45.7 57.3 3.17 Export unit values, index, 1995=100 (TRADE) Units 93.9 82.7 80.8 70.2 80.3 100.0 79.2 79.3 66.1 64.0 70.1 3.18 Import unit values, index, 1995=100 (TRADE) Units 85.6 87.5 91.2.. 91.5 100.0 97.4 92.1 88.8 87.7 92.9 3.19 Terms of Trade, index, 1995=100 (TRADE) Units 109.7 94.5 88.5.. 87.8 100.0 81.3 86.1 74.5 73.0 75.5 3.20 Manufactured goods exports, $ Mil 67 46 62 103 129 155 168 180 194 189 175 3.21 Manufactured goods exports, growth, % p.a. % 74.4 -32.4 28.9 67.3 19.9 11.3 13.5 12.5 12.2 -2.8 -9.5 3.22 Food Imports, $ Mil 09.. 258 50 55 98 25 12 108 00 09 3.23 Food Imports, growth, % p.a. % -33.1.. .. -80.3 0.7 64.0 -75.9 -47.3 864.4 -100.0.. 3.24 Non-food Consumer Goods Imports, $ Mil . .. .. 167 199 200 125 121 128 408 3.25 Non-food Consumer Goods Imports, growth, % p.a. % .. .. .. .. .. 10.2 4.9 -33.9 0.0 6.0 226.9 3.26 Fuel Imports, $ Miu 119 83 53 47 56 41 51 87 42 115 177 3.27 Fuel Imports, growth, % p.a. % -9.9 -17.5 -35.0 0.1 25.2 -31.7 4.6 82.1 -29.5 98.0 -0.4 3.28 Primary intermediate goods imports, $ Mil 49 34 48 20 72 80 80 50 54 02 02 3.29 Primary intermediate goods imports, growth, % p.a. % -27.7 -32.5 51.0 -52.3 222.6 0.1 -13.5 -37.5 5.1 -95.8 23.2 3.30 Manufactured goods imports, $ Mil .. .. .. .. 40 54 55 116 108 103 119 3.31 Manufactured goods imports, growth, % p.a. % 47.1 .. ... 54.3 55.7 130.3 -0.1 -9.4 12.8 3.32 Capital goods imports, $ Mil 346 285 349 328 296 396 361 388 284 253 292 3.33 Capital goods imports, growth, % p.a. % -11.9 -19.4 17.4 -5.8 -13.0 23.7 -4.8 13.1 -23.9 -11.3 18.9 126 Annex II Public Expenditure, Growth, and Poverty: Alternative Scenarios from the 123PRSP Model 1. Public expenditure, economic growth, and poverty are clearly interrelated - the level and composition of public expenditure are important factors to the government objectives of growth and poverty reduction; vice-versa, the objectives of growth and poverty reduction will also affect the level and sectoral allocation of public expenditure. While it is difficult or not always possible to disentangle their interactions, this section will try to examine the prospects and risks for public expenditure, growth, and poverty in Zambian a practical way. The starting point is the consensus forecast agreed by the Government of Zambia, the International Monetary Fund, and the World Bank; this consensus forecast is contained in the current medium-term macroeconomic framework supported by the IMF's Poverty Reduction and Growth Facility (PRGF) and the Bank's adjustment credit, Fiscal Sustainability Credit (FSC). Possible expenditure pressures and other risk factors may however cause significant deviations from the consensus forecast in the short and long term and the next section investigates what the likely impact might be on the economy, the budget, and various household groups. These steps are combined by utilizing an operationally oriented macroeconomic framework for poverty reduction strategy papers, the 123PRSP model (see below). It is important though to look beyond numbers in the growth and expenditure scenarios and the last section of the macro chapter examines the policy lessons and caveats on the composition of public expenditure. Consensus Forecast - The Medium-Term Macroeconomic Framework 2. The privatization of the ZCCM (copper mines, Zambia's biggest state-owned enterprise) in March 2000, has brightened economic outlook and has led to an increase in donor support and funding, including the approval interim of HIPC debt relief in December 2000. Economic outlook is expected to improve with higher investments in the mines and rising copper prices. Inflation, which reached close to 30 percent in end-2000 partly because of the recent oil price shock and monetary expansion, need to be contained. Expected to be temporary, the oil shock in 2000 will likely require additional external financing of around $60 million in 2000-02, or some 1.7 percent of annual GDP.52 Moreover, much remains to be done to ensure that the government continues with economic reforms, including governance. Key aspects of the consensus forecast include the following points (Table A2. 1): GDP is projected to grow by 5 percent in 2001 on the strength of the expected recovery in mining production and will rise to a modest 5.5 percent in 2003. Inflation is expected to go down to less than 20 percent in 2001, falling further to 10 percent in 2003. The overall fiscal deficit (cash basis), about 7.3 percent of GDP in 2000, will fall to about 5 percent in 2001, and eventually to 4.5 percent by 2003. 52 IDA provided about US $30 million at end-2000 for the oil shock. PER Macro... 127 12/19/01 * Public revenue and expenditure as percent of GDP will remain relatively stable at about 18 and 30 percent, respectively. * Extra-budgetary headroom is however created with HIPC debt relief, amounting to about 2.7 percent of GDP in 2001 and rising to 3.2 percent in 2003. The government has identified about US$86 million (about 2.7 percent of GDP) in poverty-related activities that will be funded by HIPC resources (See Box 2.1). 3. The consensus forecast is contingent on the realization of key reform measures identified in the credits and should therefore be interpreted as policy target: * The main priorities continue to be macroeconomic discipline and reduction of the public sector deficit, implementing key structural and public sector reforms, strengthening budget management and governance measures, and laying the basis for poverty reduction by increasing social and poverty related spending. * Overall fiscal deficit (cash basis), currently 7.3 percent of GDP, is expected to fall and remain at about 5 percent of GDP. * Inflation is targeted to fall to about 10 percent by 2003. * While the sale of ZCCM marked a big step towards reducing the public sector deficit, further reforms and privatization of other key parastatals are needed. Without significant reforms, a stable macroeconomic and fiscal environment will be difficult and Zambia's growth record will remain uneven. Hence, the focus of key structural reforms continue to be state- owned enterprises, including the deregulation and privatization of the petroleum sector and privatizing the controlling interest of the largest domestic bank, the Zambia National Commercial Bank (ZNCB). Pension reform is another area requiring attention. X The efficient and transparent delivery of public services requires improving fiduciary accountability, including the implementation of a medium term expenditure framework, strengthening of the budget and financial management, and the effective control of expenditure commitments and arrears. 4. Two risks factors to the concensus forecast in the short and medium-term are: * weakening of copper price because of a possible slowdown in world demand; and * expenditure pressures from the presidential election in late 2001. 5. Over the long run, failure to implement the structural and policy reform measures are key risks. Additionally, the benefits of reforms may not be realized quickly or wihtin the time horizon of the medium-term framework. We will explore what may elevate the long- term growth of Zambia, by considering the lessons and determinants of growth compiled PER Macro... 128 12/19/01 from many countries and their implications to the composition of public expenditure and policy reform. We also look at the possible consequences of HIV/Aids. Table A2.1: Consensus Forecast: Medium-term Macroeconomic Framework, Zambia"3 (IMF Financial Programming/Bank RSMS-X; in percent change, unless otherwise specified) 1999 2000 2001 2002 2003 Real GDP 2.4 3.6 5.0 5.0 5.5 Inflation- end year CPI 20.6 30.1 17.5 12.0 10.0 Inflation - average annual CPI 26.8 26.1 22.5 13.7 10.8 GDP deflator 21.7 27.9 24.9 17.2 13.2 Investment (percent of GDP) Term oftrade - 5.7 1.9 3.1 3.5 2.4 Copper price (US$/lb) 0.70 0.83 0.87 0.91 0.92 End-period Exchange rate (kwacha/US$) 2630 4160 Average Exchange rate (kwacha/USl 1 $) 2380 3110 Current account balance (percent of GDP) -15.2 -18.5 -19.9 -21.1 -19.3 Broad money 29.2 67.5 17.1 Net foreign assets -14.7 63.9 16.3 Net domestic assets 43.9 1.6 0.8 Net claims on government 10.2 12.4 -4.2 Claims on public enterprises 12.0 3.6 -4.4 Revenue and grants (percent of GDP) 25.5 25.4 26.7 23.5 23.4 Revenue (percent of GDP) 17.6 19.6 17.3 17.7 18.1 Expenditure (percent of GDP) 29.2 31.2 31.3 29.7 28.2 Overall balance (cash basis - percent of GDP) -4.0 -7.1 -5.0 -6.6 -5.2 0. balance w/o grants(cash basis-percent of GDP) -11.9 -12.9 -14.5 -12.4 -10.5 Domestic balance (cash basis- percent of GDP) 0.4 -3.4 -3.3 -2.2 -1.2 The 123PRSP Macroeconomic Framework 6. Alternative scenarios from risk factors to the consensus forecast are developed by using the 123PRSP macroeconomic framework (see Figure A2.1). The framework combines the flexibility and best features of four existing modules to deal with key determinants of short-term and long-term growth, economy-wide impact on key relative prices such as the real exchange rate and wages, and, along with growth elasticities of household expenditures of different household groups, to estimate the likely implications on various household expenditures and incomes. It should be emphasized that these steps are not combined by building a full-blown, integrated and sophisticated analytical framework - which would be beyond the scope the present PER.54 However, the 123PRSP model is a simple, quantitative 53 The figures and ratios to GDP may differ from PRGFIFSC numbers for a number of reasons - the assumptions of the consensus forecast are run through to the 123PRSP model to derive the reference run numbers; the tax rates are smoothed out if the implied taxes from the revenue to GDP ratios suggest wide variations in the tax rates. 54 Such an integrated framework would likely entail a multi-sector, multi-household, dynamic general- equilibrium model such as the one proposed in Agenor, Izquierdo, and Fofack (2001). PER Macro... 129 12/19/01 framework developed at the World Bank for immediate and easy operational use and specifically to examine critical tradeoffs in poverty-reducing macroeconomic policies, such as, how do the costs (in terms of poverty) of higher spending (and higher fiscal deficits) compare with the benefits of targeting that spending on the poor? While not a substitute for a more sophisticated framework in terms of scope and in-depth analysis, it is meant to provide first approximations and to anticipate the results of more complex models, particularly if data, budget, time, and capacity problems exist. Macroeconomic Policies (Consensus Forecast- IMF Financial Programming Model Short-run Growth Long-run Growth (Trivariate VAR Model) ("Get Real" Model) Relative prices, wages etc Relative prices, wages etc. (1-2-3 Model) ] (1-2-3 Model) l Household data Household data (poverty) (poverty) (t=1,2, ...,5) (t=6,...,10) Figure A2.1: Schematic View of the 123PRSP Model 7. The first module begins with a consensus forecast - an agreed set of macroeconomic projections normally derived from aggregate accounting framework, such as the IMF Financial Programming Model (FPM) or the Bank's RMSM-X Model. They have the advantage of having a consistent set of national accounts, linked with fiscal, balance of payments and monetary accounts. Most of the macroeconomic policies, such as the level of government spending, taxation, and the composition of deficit financing, will be contained (as exogenous variables) in this module. However, unlike the standard practice with RMSM-X or FPM, the economy's growth rate and its real exchange rate are not taken as given (or as targets) but rather they are explicitly derived in the models that follow; in addition, the last module calculates the possible impact on household expenditures and incomes. The steps are as follows: (i) possible deviations of GDP growth from the consensus forecast is calculated from one of two growth models, either the "Get Real" model, which is a long-run growth model, or a Trivariate VAR model that captures short- run growth effects; (ii) the GDP growth and macro information in the accounting and growth modules are read into the "1-2-3 Model," a static, multi-sector, general-equilibrium model to obtain economy-wide implications; and finally (iii) key prices, wages and growth elasticities are mapped into household data to derive poverty implications.55 55 A more detailed description of the framework is found in Devarajan and Go (2001). PER Macro... 130 12/19/01 Deviations to the Medium-term Projections 8. The key two risk factors in the short to medium term are expenditure pressures and a possible terms of trade shock.56 The short-run growth elasticities of these shocks in the first two years from one of the regressions for Zambia are given in Table A2.2. These coefficients are derived from time-series analysis, trivariate vector autoregression (VAR), to captures the dynamic interrelationships between such variables as the world copper price (terms of trade shocks), public expenditure, and GDP growth: Table A2.2: Growth Elasticities in the First Two Years Year 1 Year 2 Terms of trade shock (price of copper) 0.053 0.024 Government spending 0.038 -0.033 *Note: Impulse response from the VAR for the first two periods. 1) Shocks in public expenditure: The growth elasticities indicate that increases in government consumption do not have much short-term beneficial impact on growth (see discussion in the previous section) - the slight positive effect in the first year is offset in the second year. In fact, the cyclical behavior between GDP and shocks to government consumption was broken in 1987, when macro balances deteriorated rapidly. Prior to 1987, the connection was more positive and significant; after 1987, it began to move in opposite directions when large swings in government expenditure occurred particularly between 1987 and 1992. 2) Shocks in copper price: Changes in Zambia's GDP can be tracked almost completely by changes in world price of copper (except for a brief period between 1970-74). The response of GDP to changes in copper prices persists for 2 to 3 years, but the response has been tighter since the balance-of-payments situation deteriorated in the late eighties. 9. Alternative Scenarios in the Medium Term: the results of two simulations are shown in Table 1.10. The first simulation is the expenditure shock. * Using the consensus forecast as a reference run, election pressures raise expenditure (government consumption) by 15 percent in 2001 and 10 percent in 2002. The mode of adjustment to the shocks assumed is to keep foreign borrowing as percent of GDP constant relative to the reference run (i.e. a constraint); investment adjusts to available savings. 56 A third possible risk is shock in the agriculture sector. Random weather changes such as a drought continue to be important factors in agricultural production and its impact to GDP. However, we generally did not find that agricultural value added to be a good predictor of the underlying shocks to agricultural production and GDP. Nevertheless, there has been a stronger responsiveness of GDP growth to changes to the value added in agriculture since the late eighties. PER Macro... 131 12/19/01 * Relative to the consensus forecast, the small Keynesian effect of additional expenditure on GDP, a percentage point deviation of 0.6 in the first year is largely dissipated by the second year. * Even without the availability of foreign financing, the overall fiscal deficit of the central government deteriorates to 6.0 percent of GDP in 2001, 7.8 percent in 2002, and 6.3 percent in 2002, raising issues of sustainability. * With no room to expand the current account deficit (by assumption), macro adjustment falls heavily on investment, crowding out private investment in particular. Investment falls from the reference run by about 4 to 5 percent in general. * There are very little effects in the external sector in terms of the real exchange rate or relative price of exports and imports and the volume of exports and imports. * Consumption and household incomes by decile increased slightly relative to the reference run, but nowhere near the magnitude of the expenditure increase. In fact, if fiscal or tax adjustment is needed to reduce the fiscal deficit, the small gains would be reversed.57 * The findings confirm the arguments about expenditure policy in Zambia - the benefits are few but the risks of fiscal deficit and the crowding of investment are high. In the next and last section, we return to the issues about the links between public expenditure, growth and poverty. 10. In the second simulation, a terms-of-trade shock is added to the above: - Over the first simulation, copper price also deviates from the medium term framework by -20 percent in 2001; -15 percent in 2002 and -10 percent in 2003. For comparison, the average volatility of the spot price of copper is about 23 percent; during the period 1997-99, copper price fell by more than 45 percent from peak to trough. * The adverse effects of such a copper price shock will be devastating in Zambia. GDP growth will lose by -0.5 percentage point in 2001, -1.4 percentage point in 2002, and -1.3 percentage point in 2003. * The loss of tax base means that fiscal deficit will rise to 7.3 percent of GDP in 2001, about 10 percent in 2002. * The depreciation of the real exchange rates or relative prices of foreign goods will be high particularly for imports. Imports will be much more expensive relative to domestic goods by 22 percent in 2001 and over 35 percent in 2002 57 A fiscal adjustment simulation (not shown) confirms the results. PER Macro... 132 12/19/01 and 2003. The relative price of exports over nontraded goods also rises by about 8 percent in 2001 and over 13 percent in 2002 and 2003. * As a consequence, expenditure and production switching will take place. Imports fall strongly relative to the reference run, by about 13 percent in 2001 and over 20 percent in 2002 and 2003. As the economy contracts but the real exchange rates of exports rising, exports will rise slightly in real terms, by about 2 to 3 percent in spite of the general fall in their dollar prices. * The contraction in aggregate demand is strongest in investment, which deviates from the reference run from 15 to 25 percent. Aggregate consumption falls by 8 to 13 percent. * As a consequence, household welfare generally declines. Relative to the reference run, household incomes in general will fall by 2 percent in 2001, cumulating to about 6 percent in 2003. The impact on each income decile is shown in Table A2.3. The results generally confirm that such a terms-of trade shock will be hard on all household groups, particularly the poorest. Table A2.3: Impact of Shocks in Government Expenditures and Copper price (Percent Deviation from the Reference Run; unless otherwise stated) 2001 2002 2003 Real GDP (Dercentaze Doint +1-) Expenditure shock 0.6 -0.1 0.1 Expenditure + TOT shock -0.5 -1.4 -1.3 Overall fiscal balance (cash basis-% of GDP) Reference run -5.0 -6.6 -5.2 Expenditure shock -6.0 -7.8 -6.3 Expenditure + TOT shock -7.3 -9.8. -8.6 Real exchange rate of imports (depreciation >0) Expenditure shock 0.0 0.0 0.0 Expenditure + TOT shock 21.8 35.9 35.8 Real exchange rate of exports (depreciation>0) Expenditure shock 0.0 0.0 0.0 Expenditure + TOT shock 7.9 12.7 13.6 Imports (real) Expenditure shock 0.5 0.4 0.6 Expenditure + TOT shock -12.8 -20.1 -21.2 Exports (real) Expenditure shock 0.6 0.4 0.6 Expenditure + TOT shock 2.7 3.2 2.2 Investment (real) Expenditure shock -4.3 -4.9 . -4.5 Expenditure + TOT shock -14.6 -25.3 -25.1 PER Macro... 133 12/19/01 Cnn-iimntinn (ren l Expenditure shock 0.6 0.4 0.6 Expenditure + TOT shock -7.6 -11.6 -13.0 Household income in real terms Expenditure shock Decile 1 (poorest) 0.5 0.4 0.6 Decile 2 0.6 0.6 0.7 Decile 3 0.6 0.6 0.7 Decile 4 0.6 0.5 0.7 Decile 5 0.6 0.5 0.7 Decile 6 0.6 0.5 0.7 Decile 7 0.6 0.5 0.7 Decile 8 0.6 0.5 0.7 Decile 9 0.6 0.5 0.7 Decile 10 0.6 0.5 0.7 Total 0.6 0.5 0.7 Expenditure + TOT shock Decile 1 (poorest) -4.6 -8.0 -9.8 Decile 2 -4.2 -7.4 -9.2 Decile 3 -4.0 -7.0 -8.8 Decile 4 -3.6 -6.4 -8.2 Decile 5 -3.4 -6.1 -7.8 Decile 6 -2.6 -5.0 -6.6 Decile 7 -2.4 -4.7 -6.3 Decile 8 -2.7 -5.1 -6.7 Decile 9 -2.5 -4.8 -6.4 Decile 10 -1.7 -3.7 -5.3 Total -2.1 -4.4 -5.8 Deviations to the Long-term Growth Path 11. Is it possible for the Government of Zambia to undertake further reforms and reallocate public expenditure in order to raise long-term aggregate growth? We turn to that important question next and examine possible deviations to the long-term growth or historical trends. 12. Get Real Model: To derive alternative long-term growth in Zambia, we employ the Get Real model (Table A2.4), which presents a parsimonious set of cross-country growth regressions and captures the long-run growth impact of key factors and policies. Determinants generally fit in four broad groups - physical capital, human capital, policy, and shocks. Because it is difficult to come up with comprehensive and non-arbitrary measurements of factors like human capital, physical capital, and policy, specific proxy variables are customarily employed. The coefficients of one of the more robust Get Real regressions are given in the table. It is a reduced-form model, and since it is based on cross- PER Macro... 134 12/19101 country regressions, the coefficients are the same for all countries.58 An alternative would be to estimate a country-specific model of Zambia. The problem here is that there is not enough intertemporal variation in policies to obtain significant coefficients. Other long-run growth determination may also be used in place of the Get Real Model at this stage in the framework. Note, for example, that the long-run growth effects of increases in secondary- school enrollment and in infrastructure stocks (telephone lines) are captured by this model. The version presented in Table A2.4 is the expanded Get Real model that also found favorable external shocks (or the lack of) to be important factors to growth (contraction) of developing countries in the last two decades. TABLE A2.4: GROWTH COEFFICIENTS OF THE GET REAL MODEL Policy and determinants Black market premium -0.0153 (-4.02) M2/GDP 0.0004 ( 3.31) Inflation -0.0014 (-0.21) Real exchange rate -0.0087 (-2.36) Secondary enrollment 0.0003 ( 2.40) Telephone lines/1000 0.0054 ( 2.13) Shocks Terms of trade as % of GDP 0.2125 ( 2.45) Interest on external debt as % of GDP -0.0029 (-3.28) OECD trading partner growth 0.0210 (3.56) Initial conditions Initial income -0.0105 (-2.33) Intercept 0.0236 ( 0.71) Shifts 80s -0.0021 (-0.41) Shifts 90s 0.0046 (0.60) Zambia's economic growth Actual, 90s -1.5 % Estimated -2.1% Source: William Easterly (2000). Figures in parenthesis are t-statistics. 13. Impact of HIV/AIDS: With one of the highest prevalence rate, the impact of HIV/AIDS in Zambia is a very significant factor to the budget and long-termn growth. Below is a practical approach to account for HIV/AIDs, cognizant of the fact that there may be no single way of handling this factor. Regarding the budgetary implications, the cost of prevention and treatment, currently estimated to be near US $800 per person per year, is certainly 5S Although the coefficients are the same for all countries, the long-run growth rates will be different since the levels of the explanatory variables will be different. PER Macro... 135 12/19/01 beyond the means of governnent resources without additional help. Government strategy is to seek foreign assistance and to coordinate the efforts of government, donors, and NGOs through a National HIV/AIDS Council (see the health chapter). Part of the HIPC savings are also projected to defray the start-up cost of and to fund various information campaign and preventive measures. Consequently, the net effect on the budget necessarily has to be negligible because of the resource constraint and the expectation of foreign assistance (assumed to be the case.) Beyond the budget, the impact on the economy is going to be significant. Based on studies on nearby countries in Africa, the impact on GDP is estimated to range from -0.5 to as much as -1.5 percentage points, depending generally on which segment of the labor market is affected the most (see Table A2.5).59 In the case of Zambia, the mid-point -1.0 percentage point is assumed and applied as a straightforward adjustment to the results of the Get Real model. Table A2.5: Impact of HIVIAIDS on GDP Growth Selected World Bank Studies on Africa No. of Data/method Approx. Effect on Countries GDP growth (%) Arndt and Lewis (2000) South Africa CGE simulations -0.8 to -1.0 Bonnel (2000) 47-86 Cross-section -0.7 Kambou, Devarajan, and Over (1991) Cameroon CGE simulations -0.5 to -1.2 Over (1992) 30 Demographic/econ -0.3 to -0.6 omic modeling Sackey and Raperla (2000) - various Botswana, Demographic/econ -1.0 to -1.5 Reports Lesotho, omic estimates Namibia, __________________________________ Swaziland 14. Alternative Long-Term Growth Rate: Strictly speaking, there is no long-term growth rate in the consensus forecast, a medium-term framework covering the next three years from 2001 to 2003. The GDP growth rate of 5.5 percent in 2003 is therefore assumed to continue and is taken as the implied long-term growth rate in the consensus forecast. It should be noted further that initial conditions and inertia are significant factors in the get real model and substantial reforms are necessary to attain such a growth rate. Starting from the realities of the 1990s, a 5.5 percent GDP growth would require, for example, the following: (i) no economic distortions (zero black market premium and zero overvaluation of the real exchange rate); (ii) a HIPC debt relief that reduce the interest payment as ratio to GDP by more than half; (iii) an inflation rate of less than 10 percent;60 (iv) a very favorable export environment; and (v) substantial increases (over 30 percent) in school enrollment rate and infrastructure like telephone. 59 If the skilled labor market is affected most as in smaller countries like Botswana, Lesotho, and Namibia, the impact is projected to primarily fall on population rather than GDP per capita. 6 Note the coefficient for inflation is not significant. The variable maybe multicolinear with other determinants. The growth effects of inflation may also cancel out over the cycle of inflation crises. See caveats in the discussion of Table 1.7. PER Macro... 136 12/19/01 Table A2.6: Alternative Long-Term Growth (Percent Deviation from the Reference Run) Sim 3 Sim 4 Sim 5 Real GDP (DercentaLe Doint +1-) 0.9 -4.6 -1.0 Household income in real terms Decile 1 (poorest) 1.1 -4.7 -0.9 Decile 2 1.0 -4.7 -0.9 Decile 3 1.0 -4.6 -0.9 Decile 4 1.0 -4.6 -0.9 Decile 5 1.0 -4.6 -0.9 Decile 6 1.0 -4.6 -0.9 Decile 7 1.0 -4.5 -0.9 Decile 8 1.0 -4.5 -0.9 Decile 9 1.0 -4.5 -0.9 Decile 10 0.9 -4.5 -0.9 Total 0.9 -4.5 -0.9 15. Three alternative long-term growth paths are examined (counting the two VAR simulations, Simulation 3 is the starting point): * Simulation 3: A 10 percent improvement across the board on all the underlying variables in the Get-Real Model (except initial income). The net effect is to add 0.9 percentage point to the economic growth of Zambia. Household income in all deciles increase by about I percent relative to the reference run. * Simulation 4: As a policy target, the consensus forecast is generally higher and more optimistic than the historical trends or what the get real model would suggest.6' In Zambia's case, the latter would imply a lower GDP growth of less than 2 percent per year and a negative growth for GDP per capita. If progress is not made and all growth determinants remain the same as in the nineties, a -4.6 percentage point reduction in GDP growth from the reference long-term growth is projected (i.e., 0.9 percent GDP growth). The net effect of about the same magnitude on all households (Table 1.13) measures the income foregone from a reform program that is seriously off track. * Simulation 5 shows the impact of a one percentage point reduction in GDP growth that may be due to HIV/AIDs for example. 61In many countries, there is a tendency for overestimation with regard to the execution of reforms and the impact of future investments on growth. See, for example, William Easterly (1997). PER Macro... 137 12/19/01 ANNEX III: GOVERNMENT INTERIM SYSTEM FOR TRACKING HIPC-RELATED EXPENDITURES* Recognizing the importance of public accountability in the use of its HIPC resources and, while the PRSP, MTEF and a better budget management system is being developed, the Government is setting up an interim system to track, monitor and report on HIPC-related expenditures. The key features of the interim system are: I. The HIPC-funded activities are integrated in the Budget like all other budget items, (line items, code and amount). HIPC-related expenditures will be identifiable in all budgetary procedures (revenue, disbursement, expenditure, and reporting), supported with proper documentation. 2. The programs to be supported by HIPC have been determined in consultation with the line ministries. A summary table has been compiled that lists the HIPC-funded activities for 2001, with the current budget codes and the amount allocated for each item, together with the total allocated to these items in the 2000 and 2001 Budget. For 2002, the programs will be determined according to the priorities set out in the full PRSP. 3. In the recently published 2001 Budget (Yellow Book), HIPC-funded or co/funded programs have been identified with a footnote within the existing administrative subhead, and added to the list of donor-funded grants. However, as from March 2001, a separate HIPC subcode is to be introduced at the subhead level, which will enable the resources to be tracked through the disbursement, expenditure, reporting, and auditing processes-the new code will be made retroactive, and past funding sheets revised, so as to allow proper categorization of all HIPC Initiative-expenditures throughout the year. When the Integrated Financial Management Information System (IFMIS) is installed and IT capabilities are improved, the HIPC coding system will be further modified. 4. In order to identify the HIPC revenues, a separate Ministry of Finance (MOFED) account has been created in the Bank of Zambia, as an integral part of the government's cash position. HIPC resources will be channeled into this account through the year, reflecting reductions in debt service due and inflows of cash refunds that materialize in the period. 5. The line ministries will provide quarterly projections of expenditures to the Budget and Economic Affairs Division, who will use this information as the basis for allocating and releasing HIPC resources, as part of the normal execution of the Budget. 6. From the HIPC account, the Budget Office will, on a monthly basis, release resources to the line ministries, with a funding sheet indicating the HIPC programs to which the PER Macro... 138 12/19/01 funds should be applied. As the HIPC code is at the subhead level, controlling officers (COs) will have no authority to transfer HIPC resources to non-HIPC activities. 7. Budget Office will prepare a monthly report for the Accountant General, detailing sources, receipts and disbursement of HIPC resources. 8. COs in the line ministries will continue to report monthly to the -*ntant General on all expenditures, but will also report separately on HIPC-related expenditures, using the HIPC codes. The COs will ensure that the actual expenditures reported for HIPC programs are consistent with the amounts approved. 9. In his monthly consolidated report to the Permanent Secretary (FMA), the Accountant General at MOFED will include a summary report on HIPC receipts, disbursement and expenditures, following the accounting procedures and reporting currently in use. The summary of HIPC sources and utilization of funds will be published on a quarterly basis in at least three newspapers, beginning August 2001. 10. HIPC transactions will be subject to internal and external audit. They will be pre- audited internally by the Head of Internal Audit, to ensure accuracy, validity and authenticity of the transaction; conformity with the regulatory framework; and reporting of any misappropriation and wastage. The Auditor General will also audit HIPC transactions as part of the Financial Report (Blue Book), which is presented to Parliament. 11. Semi-annual monitoring and inspections of HIPC expenditures will be carried out by a committee comprising officials from the Zambia Institute of Certified Accountants, the Central Statistical Office, the Zambia Association of Public Finance and Auditing, the Institute of Purchasing and Supplies and representatives of civil society. Their reports to the Minister of Finance and the Secretary to the Treasury will be available to the public. Additionally, with donor support, MOFED will be carrying out some public expenditure surveys (PETS) at selected facilities (e.g. health clinics, schools) in order to track the flow of public resources to these facilities and to assess the delivery and quality of public services to the intended consumers. 12. To promote transparency, public discussion and monitoring of the amounts and use of HIPC resources, and to improve the HIPC expenditure system, the Ministry of Finance will hold semi-annual public meetings for donors and civil society organizations. The monitoring and inspection reports, the PET reports, audit reports and other HIPC related-reports will be distributed prior to these meetings. A percentage of the HIPC funds will be retained to finance the semi-annual monitoring and inspections and public meetings. *Source: MoFED, IMF PER Macro... 139 12119/01 22' 23' 24' 25' 26' 27' 28' 29' 30" 31" 32" 33' 34 35" 36- a .PREPUBLIC OF 2AMBIA a PERCENTAGE OF PEOPLE LIVING BELOW THE POVERTY DATUM LINE BY DISTRICT, 1998 lun_u 9- O elong~~~~~~~~Nao .n 0 Klomete; 40 o 40 80 120 160 200 240 280 320 388 400 440 480 S20 po oso O t_7 } Mansama a~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sl a~~~~~~~~~~~~~~~~~~~~~~as O 12- Mwinilunga Mpika r_12° O 0tD gISI _ _ \ ; 0 6uhoezi ulira Lundazi a~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~l a _K JKabompo ' | -'; -- ab otsaitWz UMambompo { iSD - Zamb0i | i ~~~Mufurnbwe \0,r,(hp< 01 14' - 14''- KY, a , . m A V M 1391-99%~~~~~~~~Mufdya 0~~~~Kao C46 - -h65a M 17" Kalomo - 757 -~~~~~~~~~mk 75- 22" 23 24" 2 26" 27 28" 29" 30 33 s st j .,2' 2'5 26 272' 2' W31" 32" 33" 34 35" 36" MAP 2: ZAMBIA'S MAIN INFRASTRUCTURE _ ~oe 0 -I N, 0p e 0'pi 5 Z~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Z 0 en C') S QI _ Cy Source: Co Source: Central Statistical Office (CSO); Govenment Republic of Zambia (GRZ,) 141