Zambia - Fiscal management report

CONFIDENTAL b a RepOi t No.: 18552 ZA REPUBLIC OF ZAMBIA FISCAL MANAGEMENT REPORT November 19, 1998 Macroeconomics I Africa Region This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. FILE COpy  CURRENCY EQUIVALENTS Currency unit: Zambian Kwacha (K) US$I = 2000 Kwacha (Exchange rate of August 1998) WEIGHT AND MEASURES Metric system FISCAL YEAR OF BORROWER January I-December 31 ABBREVIATIONS AND ACRONYMS BOZ Bank of Zambia BOIMOFED Budget Office, Ministry of Finance and Economic Development CEC Copperbelt Energy Coporation CSPF Civil Service Pension Fund ESAC Economic and Social Adjustment Credit ESAF Enhanced Structural Adjustment Facility GDP Gross Domestic Product GFMIS Government Financial Management Information System GRZ Government of the Republic of Zambia IBRD International Bank for Reconstruction and Development IDA International Development Association IMF International Monetary Fund LASF Local Authorities Superannuation Fund MOE Ministry of Education MOFED Ministry of Finance and Economic Development MOH Ministry of Health MPU Ministerial Procurement Units NEDC National Economic and Development Committee NFPSD Non-Financial Public Sector Deficit PER Public Expenditure Review PFP Policy Framework Paper PIP Public Investment Program PIRC Privatization and Industrial Reform Credits PTA Preferential Trade Agreement VAT Value Added Tax ZCCM Zambia Consolidated Copper Mines ZESCO Zambia Electricity Supply Corporation ZNOC Zambia National Oil Company ZNPF Zambia National ProvidenFund ZNTB Zambia National Tender Board ZIMCO Zambia Industrial and Mining Corporation ZPA Zambia Privatization Agency ZRA Zambia Revenue Authority ZSIC Zambia State Insurance Company  TABLE OF CONTENTS EXECUTIVE SUMMARY ........................................................................................................................... 1 CHAPTER 1 .................................................................................................................................................. 1 FISCAL MANAGEMENT IN ZAMBIA •••....••••..••....•......•..•.....•••.......•.•.•.•.•.•.........•....••••.•....••... 1 A. INTRODUCTION...•...................•...........................•............................................................. 1 B. DEFINING THE FISCAL DEFICIT....•......................................•..•................................... 3 C. THE CENTRAL GOVERNMENT DEFICIT ...............•............•.•............•..............•.......... 4 Central Government Revenues .............................................................................................................. 5 Central Government Expenditures ......................................................................................................... 6 D. THE NON-FINANCIAL PUBLIC SECTOR DEFICIT. ....•................................•......... .•... 7 Local Government ......................................................................................................................... :........ 7 Extrabudgetary Funds ............................................................................................................................ 8 State-Owned Enterprises ........................................................................................................................ 9 Government Guarantees ......................................................................................................................... 9 Estimating the Non-Financial Public Sector Deficit ............................................................................ 10 E. THE QUASI-FISCAL DEFICIT..•.••...........•............•.........•....•......•.....•......•...........•.••...... 12 Profit and Loss on Foreign Exchange .................................................................................................. 13 Bank Failures ....................................................................................................................................... 14 Loan Guarantees .................................................................................................................................. 15 Estimating the Quasi-Fiscal Deficit ..................................................................................................... 15 F. SOLVENCY AND SUSTAINABILITY ANALySIS••....................•.......•.•................••....... 16 G. A PROGRAM OF FISCAL ADJUSTMENT. .•.........••...........•.•................. ........•............... 19 A Program of Fiscal Adjustment. ......................................................................................................... 20 Fiscal Adjustment of State-Owned Enterprises ...................................................................................21 H. SUMMARY OF RECOMMENDATIONS ••.•.•.•.••.•..•.•...•.......•.....•..•..•.•.•.••...•.••.•....•...•..•.. 22 CHAPTER 2 ................................................................................................................................................24 EXTRA-BUDGETARY FUNDS ................................................................................................. 24 A. GRANT-AIDED INSTITUTIONS •.•..••..•.•.•.•••.••.•.......•........•.• .•.••........•.•••.•....••.•..•.••••...••.. 25 B. LOCAL GOVERNMENT ....•••••....•.•.••••••.•.•.•.•.••.•.••.•.••. ...••.•....•.......••••.•••..•.•..•..•.••.•.•..•...... 28 c. SPECIAL FUNDS .............................................................................................................. 31 Rural Electrification Fund .................................................................................................................... 31  -2- The Road Fund .................................................................................................................................... .32 D. PENSION FUNDS ............................................................................................................. 33 h~proving the Financial Management of Extrabudgetary Funds ......................................................... 37 CHAPTER 3 ................................................................................................................................................ 39 ZAMBIA'S BUDGET MANAGEMENT SYSTEM ................................................................. 39 A. THE LEGAL BASIS OF THE BUDGET.......................................................................... 39 B. BUDGET COVERAGE ...................................................................................................... 40 C. BUDGET PREPARATION................................................................................................ 41 D. BUDGET IMPLEMENTATION ....................................................................................... 43 Cash Budgeting .................................................................................................................................... 44 Procurement, Payment, and Banking Arrangements .......................................................................... .46 Cash Management ............................................................................................................................... .48 Rea\1ocation of Spending within Expenditure Categories ...................................................................49 Supplementary Estimates ..................................................................................................................... 49 Accounting and Reporting ...................................................................................................................49 Auditing .............................................................................................................................................. .51 Accountability '" .................................................................................................................................. .51 CHAPTER 4 ................................................................................................................................................ 54 IMPROVIN"G ZAMBIA'S BUDGET MANAGEMENT ......................................................... 54 A. IMPROVING CASH BUDGETING .................................................................................. 55 Enforcing Financial Regulations .......................................................................................................... 55 Improving Accounting Capacity .......................................................................................................... 56 Eliminating Uncertainty over the Legal Status of the Cash Budget.. ................................................... 56 Publishing Data on Cash Releases ....................................................................................................... 57 Adopting a System of Lump-Sum Cash Releases ............................................................................... .57 Linking the Procurement System with Commitment and Expenditure Reporting .............................. .57 Improving Cash Management ............................................................................................................. .58 A Centralized or Decentralized Treasury? .......................................................................................... .59 B. MOVING TOWARD A MEDIUM-TERM EXPENDITURE FRAMEWORK ................ 60 Strengthening Political Support ........................................................................................................... 62 Improving Fiscal Discipline ................................................................................................................. 62 Increasing Planning and Budgeting Capacity ...................................................................................... 63 Improving the A\1ocation Process ........................................................................................................ 64  -3- Increasing Accountability for Results .................................................................................................. 65 Improving the Financial Management of Extrabudgetary Funds and Organizations ........................... 66 C. SUMMARY OF PRIORITY ACTIONS IN THE SHORT TERM.................................... 67 ANNEXES ................................................................................................................................ 69  FOREWORD This report is based on an economic mission that visited Zambia in April-May 1998. The mission was composed of Messrs. Hinh Dinh (Chief), Bert Hofman, Ms. Kendall Schaefer (AFTM1), and Mushiba Nyamazana (AFTZM). A Government Counterpart Team led by Mr. Nonde (Permanent Secretary, Ministry of Finance and Economic Development, MOFED) worked closely with the mission to produce this report. That team included Ms. Florence Sitwala (MOFED) and Messrs. Martin Phillips (MOFED), Clifford Banda (MOFED), Fred Yamba (MOFED), Walter Kasempa (MOFED), Emmanuel Ngulube (MOFED), and Herbert Mhango (MOFED). The mission would like to express its appreciation to Professor Benjamin Mweene and staff of the Ministry of Finance, Dr. F.C. Kani and staff ofthe Bank of Zambia, and other officials of the Ministries of Health and Education, the Auditor General and the Accountant General for their cooperation and support. Peer reviewers for this report have been Messrs. Shahrokh Fardoust (SASPR) and Vinaya Swaroop (DECRG). The Country Director is Ms. Phyllis Pomerantz and the Technical Manager is Mr. Ataman Aksoy. Ms. Ligia Murphy has been in charge of report production. Helpful comments from Ms. Phyllis Pomerantz, Messrs. Ataman Aksoy, Delfin Go, Sudhir Shetty (AFTM1) are gratefully acknowledged.  EXECUTIVE SUMMARY i. The need for fiscal discipline and the setting of spending priorities that reflect development objectives has become acute in Zambia in recent years because of the steep decline in copper revenues and foreign aid, which has necessitated severe fiscal adjustments. While progress has been made in reducing the central government deficit, the country's overall fiscal position, taking into account fiscal and quasi-fiscal activities at all levels of governments, has shown little improvement. In part, this slow progress has been due to the lack of sufficient funds and to delays in privatization. In part, inefficiencies in the budget management process-including arbitrary cuts, ad hoc decisionmaking in response to cash shortfalls, and above all a budget management system that allows for cash allocations and budgetary appropriations to exist side-by-side-have made this adjustment particularly difficult. ii. Despite a number of public sector reforms undertaken since the early 1990s, fiscal management in Zambia remains weak. If Zambia is to enjoy sustainable economic growth, it would need to improve public expenditure management, the topic of this report. But even these measures will not bring about a sustainable fiscal position unless concerted actions are taken by all partners in development, including the government, donors, creditors, NOOs, multilateral institutions etc., to address the debt burden and the underlying structural problems of the non-financial public sector accounts and the quasi-fiscal deficits discussed in this report. 111. Provided political commitment can be obtained, the fiscal management reforms proposed in this report stand a good chance to succeed. The Ministry of Finance and Economic Development (MOFED) has a committed team of budget reform experts who are familiar with good budgeting practices in other countries. Furthermore, many of the necessary changes require only common sense and perseverance. The international community can help by providing the needed funds and by identifying the need for technical assistance, training, and capacity building to strengthen specific areas of budgetary management and control within individual institutions. FISCAL PROBLEMS IN ZAMBIA iv. Fiscal problems in Zambia are deeply rooted in historical experience and in the resulting structure of the economy. Budget revenues are heavily dependent on foreign grants and, until recently, copper-related earnings. Expenditures are dominated by external debt service, the single largest expenditure item. v. The budget deficit of the central government has declined since 1991, despite a sharp reduction in copper revenues and grants. Adjustment has oscillated widely, however, with bouts of sharp fiscal adjustment followed by prolonged periods of  - II - inaction and worsening deficits. This pattern of fluctuating deficits was caused mainly by poor management of expenditures, as revenues were relatively stable throughout this period. In fact, compared with other low- and middle-income countries, domestic revenue performance is good, with revenues representing 19 percent of GDP. Given that fJreign g.·ants are expected to continue to decline and that Zambia's tax revenues are already relatively high for a low- income country, prospects for a significant increase in total revenues are not favorable. Fiscal adjustment will thus have to depend largely on improved control over expenditures, in addition to the restructuring of internal and external debt. VI. Several noteworthy features characterize central government spending in Zambia. First, capital expenditures are financed largely from foreign grants rather than from domestic resources. Second, debt service accounts for a very large proportion of total expenditures. Third, expenditures on operations and maintenance account for a very small and declining part of the budget. Fourth, since 1993 Zambia has used a cash budget system, under which expenditures are authorized only when revenues are received by the Ministry of Finance and Economic Development (MOFED). Use of this system, combined with a lack of fiscal discipline, has resulted in erratic budget performance and rising arrears. vii. It is sometimes difficult for policymakers in Zambia to grasp the full extent of the country's fiscal problems because they have used conventional deficit measures that reflect only the operations of the central government; other components of the public sector account, such as local governments, state-owned enterprises, pension funds, special funds, and quasi-fiscal activities, have been ignored. To determine the sustainability of a country's fiscal stance, all government revenues and expenditures need to be examined. Vlll. The fiscal positions of public institutions that lie outside the central government are weak. Local governments have been adversely affected by declining grants from and rising arrears of the central government. The country's pension funds are suffering because of the enormous level of government arrears and the fact that their investments have been geared toward financing the government budget at subsidized interest rates. State-owned enterprises continue to represent a huge drain on public resources. Although most state-owned enterprises in Zambia have been privatized over the past few years, the few remaining parastatals have continued to experience financial problems. Operating losses of the largest state-owned enterprise, the Zambia Consolidated Copper Mine (ZCCM), accounted for K199 billion in 1996, about 5 percent ofGDP, and about 3.4 percent ofGDP in 1997. These losses will be borne eventually by the public sector, either because ZCCM's debt will be taken over by the government or because public assets will be depleted at the time of privatization. ix. The report estimates the nonfinancial public sector deficit-which includes the central government deficit plus the deficits of local governments, extrabudgetary funds (such as pension funds and the Road Fund), and state-owned  - iii - enterprises-at about 10.4 percent ofGDP in 1997. The central government deficit, which accounted for 2.3 percent ofGDP, was financed by Treasury bills purchased mainly by banks; the remainder of the nonfinancial public sector deficit was financed through arrears and implicit asset depletion. x. To assess the full extent of public sector activities in Zambia, the report also estimates the cost of quasi-fiscal activities undertaken by the central bank. Such activities include the central bank's implicit bearing of foreign exchange risk for domestic residents who borrow externally and the rescue of failed commercial banks and the corresponding issue of liabilities by the central bank to a fragile financial system. The total public sector deficit (including quasi-fiscal activities) amounted to 14 percent of GDP in 1997. These estimates are preliminary and subject to further revision, but they reveal the presence of a considerable quasi-fiscal deficit and indicate that substantial resources will be needed to recapitalize and improve the regulatory capacities of the central bank. xi. The analysis in this report has two implications. First, non-central government public sector activities, particularly the parastatals, have become a major drain on public resources in Zambia. Until 1996, these activities were not as important in part because the favorable world commodity prices made them profitable. However, since 1996, when world commodity prices began to decline sharply, management problems of public enterprises became acute, and problems with the pension system were exacerbated, fiscal measures at the central government level have not been adequate to improve the overall fiscal position of the country. Second, it is also clear from the analysis of solvency and sustainability that even with the best adjustment efforts, Zambia cannot do it alone. A successful fiscal adjustment program would require sustained efforts from the government as well as those of Zambia's development partners, including the creditors, NOOs, donors, and multilateral institutions. Specifically, Zambia's debt burden needs to be reduced to a manageable level, while adequate funding and technical assistance at all government levels need to be provided on a sustained basis. This is particularly important because current fiscal reforms have to be implemented in a difficult macroeconomic environment. The economy is experiencing a recession in 1998 and is likely to have a weak growth recovery in the following year. xii. International experience indicates that while the magnitude of the overall fiscal deficit in Zambia is large, this deficit can be successfully addressed if the government, with the assistance of the international institutions, can identify and implement a realistic medium-term program of fiscal adjustment. Within this program, the government should demonstrate maximum efforts to improve its effectiveness in spending and trim inefficient expenditures along the lines of its medium-term strategy of achieving fiscal balance and a sustainable and significant reduction in poverty. The international community can then help fund the remainder of the program, a large part of which should be accounted for through provision of HI PC debt relief. Details of the program can be worked out in the preparation of the 1991-2001 Policy Framework Paper (PFP), a joint document of the government and World BanklIMF staff.  - IV- Xlll. In the short tenn the government could take a number of actions to help bring the nonfinancial public sector deficit and the quasi-fiscal deficit under control. These actions include the following: • Privatize the remaining parastatals, particularly ZCCM, and use the privatization proceeds to reduce the public sector liabilities, including those of the Central Bank. • Freeze arrears by all public entities and require all public agencies to come up with plans for speedy resolution of these arrears before new funds are released to them in the next budget year. • Collect infonnation on and monitor the developments of the nonfinancial public sector deficit and the quasi-fiscal deficit, and improve estimates of these deficits. • Set out a medium-tenn program to reduce the nonfinancial public sector deficit and the quasi-fiscal deficit to sustainable levels. • Design a plan to restore the solvency of the central bank and other public sector financial institutions. • Improve the management of expenditures, especially the cash budget, to attain the medium-tenn fiscal targets. IMPROVING FISCAL MANAGEMENT xiv. The report identifies three areas in which immediate action is needed to improve the current fiscal management system: tightening control over extrabudgetary funds, improving the cash budget process in the short tenn, and moving toward a medium-tenn expenditure framework in the longer tenn. xv. Tightening Control over Extrabudgetary Funds. Extrabudgetary funds include grant-aided institutions, local authorities, special funds, and pension funds. The report examines these funds in detail because they represent the largest drain on public sector resources after state-owned enterprises. xvi. A variety of problems plague extrabudgetary funds. Grant-aided institutions are insufficiently accountable, and financial regulations are inadequately enforced. Local government budgets are adversely affected by late payment and nonpayment of grants from the central government, confusion over the basis on which grants are detennined, gross mismanagement of funds, and lack of financial discipline. Special funds, such as the Rural Electrification Fund and the Road Fund, are unable to provide adequate services because neither fund has received the monies collected by MOFED on their behalf since 1995. The pension funds have experienced serious financial problems because of government interference, poor management and accountability, and deficiencies in legislation, regulation, and supervision. The government's arrears to the Civil Service Pension Fund are currently estimated at K45-  -v- K50 billion (1 percent of GDP), in addition to debt owed to the fund of about 2 percent ofGDP, and the government continues to borrow from the pension funds. Recently, the government has also shifted responsibility for paying benefits to retrenched workers or early retirees to the fund. xvii. The government should consider improving the financial management of extrabudgetary organizations. It should develop framework legislation and regulation for the financial management of these institutions and refer to this legislation in the laws establishing such organizations. XVlll. The efficiency and equity of grants provided to grant-aided institutions need to be reviewed carefully in light of tight budget constraints. As a first step, the government should require all grant-aided institutions to submit monthly budget reports to the Budget Office ofMOFED,just as line ministries do. Funding for the new budget year should be made conditional on the submission of audited final accounts, as required by law. The local government system requires a major overhaul in terms of how it is funded, how it accounts for its performance, and how it delivers services. The government should help restore the financial health of the special funds and pension funds by ceasing to use these funds to finance its deficit. Doing so will require measures to restore the solvency and sustainability of fiscal policy. xix. The government should gradually expand its reporting on all government activities, including extrabudgetary funds, public sector borrowing requirements, and contingent liabilities, such as guarantees. Reporting must occur in both the budget and in the financial reports of the government. Over time, as budgeting is decentralized, reporting on local government activity and budgets will become necessary to inform Parliament, the general public, and investors about their activities. xx. Improving Budget Management. Given the likely decline in budget revenues, the most prudent way for Zambia to maintain fiscal discipline is through further cuts in expenditures. The need for such cuts in spending makes it imperative that the government increase the efficiency of its expenditures. This report proposes several measures to help Zambia do so by improving its budget management system. xxi. To prevent the excessive spending that led to high inflation in the 1990s, Zambia adopted the cash budget system in 1993. Although the system has helped reduce the deficit, its use has come at a cost, and it did not address what appears to be the root cause of Zambia's fiscal problems -lack of fiscal discipline. Program implementation has suffered because of the limited and unpredictable release of funds, and the effectiveness of public sector spending has been undermined. Moreover, line ministries and agencies have learned how to manipulate the system, further undermining the system's effectiveness in controlling spending. xxii. Although the cash budget has not served Zambia well, the risk associated with abandoning the system at a time of tight resource constraints is high. We therefore recommend a two-step strategy. In the short and medium run, the government should  - vi- maintain the cash budget, but it should take a number of actions to improve budget implementation. At the same time, it should prepare to move toward a medium-term expenditure framework that could eventually link government priorities with the budget within a sustainable spending envelope, highlight the tradeoffs between competing government objectives and link budget~ with tr.e policy choices made, and improve outcomes by increasing transparency, accountability, and predictability of funding. XXlll. Improving the Cash Budget System in the Short Term. The effectiveness of the cash budget system has been hindered by several problems. Enforcement of financial regulations is weak, diminishing motivation to comply with the cash budget; training of government accountants is inadequate; and funds continue to be overcommitted, increasing the government's arrears. xxiv. To address these problems, the government should consider revising the regulations to include enforceable sanctions against offenses, and it should give the highest priority to stepping up ongoing training efforts and providing adequate compensation to trained accountants. If such efforts prove ineffective, it should consider contracting out financial management to the private sector or to an independent organization similar to the Revenue Authority. In order to avoid overcommitments and arrears, MOFED needs to strictly enforce ministry reporting. MOFED should allow ministries to actively manage their own commitments reporting. As a first step, the procurement system should be linked with the commitment and expenditure reporting system, a move that would require only minor adjustments to the recently developed software. xxv. Another problem adversely affecting budget implementation is that especially at lower levels in the administration MOFED staff are often unclear on the amounts and terms of cash releases. This leaves MOFED largely unaccountable for how it releases cash and undermines parliamentary control. To improve information on cash releases, the government could consider publishing the cash releases as an annex in the Yellow Book (Budget Book). To improve cash allocation, the government should move to a system of lump-sum release per subhead rather than per. Over time, when the program budget is further developed, departments should obtain more autonomy over how they allocate their cash. xxvi. Much of the detrimental effects of the cash budget stem from the unpredictability of cash releases. To improve cash management and forecasting, it is recommended that the planning horizon of line ministries be extended to two or three months and that the recently installed cash management unit within the Budget Office be made operational as soon as possible. xxvii. To improve savings on interest payments, it is recommended that the government complete the inventory of government bank accounts immediately. Commercial banks that handle government cash should be required to report daily to MOFED on transactions and balances. If the decentralization of payroll accounts proves  - vii - successful, the government should consider decentralizing all retail payments. It is also recommended that Bank of Zambia provide both MOFED and the line ministries with daily reports on transactions and balances in the control accounts. XXVlll. Moving toward a Medium-Term Expenditure Framework in the Longer Term. Over time the government's goal should be to move away from cash budgeting and toward a medium-term expenditure framework in order to improve expenditure allocation while maintaining a tight fiscal stance. The key to doing so is creating more realistic budgets. xxix. Moving away from cash budgeting also requires revamping of the government's financial management system. Although this system has been improving over the past few years, many deficiencies persist. Before Zambia can successfully implement a medium-term expenditure framework it must establish a better system of Treasury management, create effective management incentives, gamer political support for the new system, improve fiscal discipline, enhance its planning and budgeting capacity, and establish accountability for results. xxx. Zambia's current system of Treasury management is not working well. Expenditure control is centralized by means of the cash budget, but commitments, payments, and accounting are decentralized to line ministries, leaving MOFED inadequately informed of line ministry activity. In the long run Zambia needs to consider which type of Treasury management system will best facilitate budget implementation. Once a system is selected, a good Government Financial Management Information System (GFMIS) can support Zambia's government financial management. xxxi. Appropriate incentives are also critical to effective financial management. Currently, incentives for good budget preparation remain limited in Zambia because budget implementation is driven by cash rationing. To encourage line ministries to identify cost savings and properly manage their finances and accounting, during the transition to the medium-term expenditure framework the cash budget could be abolished in ministries that do so. xxxii. Although budgetary reforms have been ongoing for some time, political support for the revised process is not strong. One reason for the lack of political support is that budgeting is still seen as a technical exercise, driven mainly by MOFED. Another reason is that politicians seem unaware of the benefits of reform and still rely on the old incremental methods to obtain funding. If reform is to be successful, political support must be enhanced. XXXlll. Fiscal discipline is undermined by ad hoc decisions made late in the budget process and by overoptimistic revenue projections. To create incentives and sanctions that will encourage ministries to exercise discipline, the government should obtain Cabinet agreement on spending limits early in the budget cycle, break down sectoral ceilings to ministries, increase the planning reserve, use the contingency reserve only for truly unforeseen spending, and improve macroeconomic and revenue forecasting.  - Vlll - xxxiv. Planning and budgeting capacity in line ministries remains limited. Realistic budgeting is hampered by long delays in reporting, weak accounting, and the lack of timely and reliable data. Although technical assistance can help, the strongest incentive to improving planning and budgeting capacity is to make it matter. Strict enforcement of submission of realistic three-year plans and rewarding of good planning and budgeting with higher budget allocations is the best incentive for line ministries to improve their budgeting capacity. MOFED and some line ministries, such as the Ministry of Education, have strong budget preparation knowledge that could be shared with other ministries. xxxv. Accountability for results is constrained by the lack of published information on policies, programs, and outcomes. Accountability could be improved by revising the budget preparation agenda so that the budget is passed before the budget year starts. Accountability could also be improved by accelerating the production of the final accounts and publishing the program budgets of line ministries, a table with sectoral (or functional) allocation of spending, and a detailed report on budget implementation and outcomes as annexes to the Yellow Book (Budget Book). xxxvi. In summary, for budget preparation, the government should focus on the following actions in the short-run: i) fully involve political decision makers in the budget preparation to increase the credibility and binding nature of the budget; ii) pass realistic budgets with sufficient contingency reserves to absorb adverse revenue developments; iii) accelerate the budget preparation calendar to ensure that the budget is passed before the fiscal year starts. For budget implementation, the government should take the following actions in the short-run: i) finalize the inventory and control of arrears; ii) set up a commitment control system; iii) extend the cash planning horizon from I to 3 months and publish monthly and annual cash release statements; iv) complete the inventory of Government accounts, require daily reporting on those accounts to MOFED and the owner of the account; and instruct the central bank to report daily to line ministries on the position of their subaccount and bounced cheques; and v) consider sanctions against controlling officers who are repeated offenders of regulations. Beyond the short term, the government should take a comprehensive approach to budget reforms. The strategic goal of these reforms should be to improve the results of government spending by increasing transparency and accountability for results. This could be done by introducing a Medium Term Expenditure Framework supported by a government financial management system that no longer relies on cash budgeting.  - 1- CHAPTER! FISCAL MANAGEMENT IN ZAMBIA A. INTRODUCTION 1.1 Zambia's fiscal problems are deeply rooted in its history and in the structure of its economy. For the first quarter century after independence in 1964, economic policy in Zambia was characterized by pervasive state intervention in factor and product markets and by large-scale state ownership of productive assets. State intervention resulted in price distortions and inefficient allocation of resources. 1.2 Because of the wealth generated by the production and export of copper, the effects of these inefficiencies were disguised by rising income and imports. As the world copper price declined, however, these inefficiencies began to manifest themselves, and domestic incomes and output began to fall. The government attempted to offset the decline in copper earnings by borrowing heavily abroad. This strategy failed to halt the economic decline, however, and the level of income fell during the 1980s, while external debt rose rapidly. 1.3 As a result of the development strategy pursued in earlier years, the structure of public finance in Zambia today is dominated by three major factors: foreign grants, copper-related revenues, and debt service payments. Foreign grants accounted for as much as 10.2 percent of GDP in 1992, although this percentage declined to 4.8 percent of GDP in 1997. 1.4 Since 1991 the Zambian government has pursued a program of economic stabilization and liberalization designed to reverse the country's economic decline and put it on the path of sustainable growth. The objectives of the stabilization program are to reduce the fiscal deficit, bring down domestic inflation, and restore external balance. The privatization and liberalization program aims at curtailing public sector ownership of productive assets and eliminating price distortions. Because of the country's heavy debt burden and the uncertain prospects for its major export, copper, the international donor community provided extraordinary balance of payments assistance to ease implementation of these reforms. 1.5 By the end of 1997 the Government's program had proved broadly successful. Prices had been decontrolled and subsidies eliminated. Inflation had been reduced significantly (Table 1.1), and market forces had been allowed to determine the exchange and interest rates. Quantitative restrictions on imports had been eliminated, the tariff structure had been compressed and simplified, and an ambitious privatization program was underway. The country appeared to have one of the most open, market-oriented economies in sub-Saharan Africa.  -2- Table 1.1: Macroeconomic Indicators, 1992-97 Item 1992 1993 1994 1995 1996 1997 Real annual GDP growth (percent) -1.7 6.8 -8.6 -4.3 6.4 3.5 Consumption (percent ofGDP at current prices) Government 16.2 27.9 14.8 14.6 12.6 10.8 Private 82.1 78.1 76.3 77.6 78.5 77.8 Investment (percent ofGDP at current prices) 11.9 15.0 13.4 13.9 14.9 15.0 Gross fixed capital 11.4 14.6 13.0 13.6 14.5 14.7 Public 6.7 4.5 4.2 5.4 5.7 5.2 Private 3.8 7.0 5.6 4.5 8.8 9.5 External debt (millions of U.S. dollars) 6,943 6,788 6,496 6,770 6,349 7,262 Fiscal performance (percent of GDP) Revenues and grants 28.5 23.9 26.9 26.6 26.0 24.0 Expenditures and net lending 31.0 29.4 33.7 31.2 28.5 26.2 Overall balance (accrual) -2.5 -5.6 -6.8 -4.6 -2.5 -2.3 Increase in consumer prices (annual average) 197.4 189 55.6 34.9 43.1 24.4 Current account balance (percent of GDP) -3.65 -2.71 -1.88 -4.51 -3.61 -6.42 Exports (millions of U.S. dollars) 1,120 994 1,066 1,186 993 1,101 Without metals 1,020 870 910 984 754 806 Imports (millions of U.S. dollars) -1,302 -1,019 -1,003 -1,194 -1,055 -1,150 Debt service due (millions of U.S. dollars) 683 522 542 590 453 376 Copper export volume 409 402 360 341 327 301 (thousands of metric tons) Source: Bank of Zambia and the Ministry of Finance and Economic Development. 1.6 Because of the decline in copper production and prices, central government revenues from copper and related products declined from 3.1 percent ofGDP in 1991 to less than 0.5 percent ofGDP in 1997. Although copper and related products now contribute little to the central government budget, their operations constitute a huge drain on the budget, and lack of progress in privatizing the country's largest parastatal, the Zambia Consolidated Copper Mine (ZCCM), is threatening the sustainability of the country's adjustment program.' 1.7 Fiscal expenditures on debt service have been high. Interest payments on external debt reached almost 8.5 percent of GDP in 1995 before declining to about 2 percent in 1997, thanks to the Paris Club debt rescheduling. For the purpose of privatization, ZCCM assets were split into nine packages. Some progress has been made in the sale of four smaller packages, but the largest packages remain unsold.  -3- B. DEFINING THE FISCAL DEFICIT 1.8 No single measure of the fiscal deficit can capture a country's fiscal perfonnance. The conventional measure of the fiscal deficit (the difference between revenues and expenditures) fails to reflect the full impact of fiscal policy on aggregate demand, even within a closed economy. 2 Indeed, an IMF study has concluded that there is no such thing as "the fiscal deficit" but rather a series of alternative measures, the suitability of which depends on the purpose of the analysis. 3 The choice of a particular measure also depends on the particular problems facing the country and the availability of data; in addition, the appropriate size of the fiscal deficit depends on, among other things, a country's targets for growth, inflation, and external and internal debt. Ideally, the deficit considered should be a consolidated public sector deficit that takes into account the entire public sector, including central and local governments, municipal accounts, nonfinancial state-owned enterprises, public sector banks, the social security system, and pension funds. In the case of Zambia, it should take into account also the special funds and grant- aided institutions. 1.9 Past analyses of Zambia's fiscal perfonnance have relied on a number of definitions of the fiscal deficit (Box 1.1). The most commonly used definition is that used by the central government. Several variants of this definition are used. The central government balance including grants represents the difference between total receipts (current and capital revenues plus grants) and total expenditures (current and capital expenditures). To examine domestic fiscal efforts, policymakers often look at the overall central government balance before grants. 4 The primary central government deficit is the overall balance of the central government excluding interest payments on domestic and external debt. It can be measured with or without grants. For a heavily indebted country such as Zambia, the primary deficit excluding grants is a useful measure of domestic fiscal efforts because it isolates the effects of the debt overhang from other fiscal issues (debt dynamics effects). The Zambian authorities also use a measure called the domestic fiscal balance, which is the central government balance on a cash basis, excluding grants, foreign-financed capital expenditures, and interest payments on external debt. This measure is equal to the primary deficit excluding grants plus interest payments on domestic debt. 2 As Tanzi has noted, the conventional measure fails to reflect the differential impacts the deficit composition has on aggregate demand, the endogenous nature of tax revenue, and the impact of alternative financing sources for the fiscal deficit. For developing countries these problems are compounded by inflation, debt arrears, and quasi-fiscal activities as well as the temporary nature of some fiscal measures (such as the sale of public assets). See Mario Blejer and Cheasty, "How to Measure the Fiscal Deficit." Washington, D.C.: IMF, 1993. 4 In Zambia there are very few current grants. All grants are therefore classified as capital grants.  -4- Box 1.1: Alternative definitions of the public sector deficits • Central government balance (before or after grants): Difference between the central government's receipts (before or after grants) and total expenditures, including net lending. • Primary central government deficit (before or after grants): Central government balance (before or after grants) plus interest payments. • Domestic fiscal balance: Central government balance before grants, excluding interest payments and foreign financed capital expenditures. • Nonfinancial public sector deficit: Balance of central government, local government, pension funds, public enterprises, and extrabudgetary funds. • Quasi-fiscal deficit: Nonfinancial public sector deficit plus deficit of central bank not related to financing of the nonfinancial public sector deficit. 1.10 Two new measures are analyzed in this report. The nonfinancial public sector deficit is defined as the central government balance after grants plus the deficits of local authorities, public enterprises, pension funds, and extrabudgetary funds. Although ideally all transactions among public entities should be netted out so that the deficit is presented on a consolidated basis, consolidation of all accounts would require too much time and data; aggregation is therefore done on a gross basis. The quasi-fiscal deficit is obtained by adding central bank losses to the nonfinancial public sector deficit. It represents the broadest measure of public sector activities in Zambia. C. THE CENTRAL GOVERNMENT DEFICIT 1.11 Zambia's overall fiscal deficit after grants, debt rescheduling, and debt relief declined from 7 percent ofGDP in 1991 to 2.3 percent ofGDP in 1997 (Table 1.2). Using a broader definition that excludes grants on the revenue side and includes scheduled interest payments (rather than actual payments) on the expenditure side, the overall fiscal deficit declined from 16.0 percent ofGDP in 1991 to 10.7 percent ofGDP in 1997. These two definitions of the fiscal deficit represent two extreme measurements: the first reflects the net fiscal impact while the second reflects the contractual obligations of the central government. A third measure, the primary deficit indicator, which reflects domestic fiscal efforts, also declined over this period. 1.12 Fiscal progress has been erratic in Zambia; prolonged periods of inaction have followed bouts of sharp fiscal adjustment. After a steep drop in 1992, the deficit excluding grants remained more or less constant until 1995, resuming its declining trend only in 1996--97. The primary deficit declined steadily from 1991 to 1994, worsened slightly in 1995 and 1996, and declined again in 1997. This pattern was caused mainly by the government's inability to effectively control expenditures.  -5- Table 1.2: Central Government Deficit, 1991-97 (percentage of GDP) 1991 1992 1993 1994 1995 1996 1997 Definition of Deficit Overall balance after grants and after -70 -2.5 -56 -6.8 -4.6 -2.5 -2.3 rescheduling Overall balance before grants and before -16.0 -12.6 -13.6 -125 -12.6 -12.7 -10.7 rescheduling Total Revenues, before grants 187 184 159 21.2 21.1 20.0 19 I Expenditures and net lending before rescheduling 34.7 31.0 294 33.7 33.7 32.7 29.8 Expenditures and net lendmg after rescheduling 34.7 31.0 29.4 33.7 31.2 28.5 26.2 Primary balance before grants -7.5 -4.1 -1.3 -0.2 -1.5 -3.9 -3.1 Source: Ministry of Finance and Economic Development and International Monetary Fund Central Government Revenues 1.13 Zambia's domestic revenues represented 19 percent ofGDP between 1991 and 1997, reflecting good revenue performance relative to other low income countries. The country's strong revenues reflected improved collection of income taxes, increased trade tax revenues, and the introduction and increase of fees for vehicle registration and driver's licenses. Out of 55 low- and middle-income countries for which data are available in the 1998 World Development Indicators, Zambia ranked 23rd, in terms of size of revenues (excluding grants). Tax buoyancy with respect to GDP was above 1, and the ratio of direct to indirect taxes was in line with that of other countries at a similar level of development. 1.14 Four sets of recent measures have improved revenues in the past three years. First, on July 1, 1995, a value-added tax was introduced to broaden the tax base. Second, policies to enhance the capacity of the Zambian Revenue Authority for tax collection were implemented. Self-assessment and automation were introduced, customs effectiveness was improved, interest and penalties on late payments by all tax payers were enforced, the tax department was computerized, and the Fines and Fees Act was strictly enforced. Third, in 1996, tariffs on most products were reduced and the differences in rates were narrowed. The elimination of exemptions and the rationalization of the customs duty structure led to an increase in trade taxes collected. Fourth, the income tax base was widened by adjusting personal income tax bands, adjusting fees and charges for inflation, and improving the collection of dividends through the dissolution of the Zambia Industrial and Mining Corporation (ZIMCO). 1.15 Zambia still depends heavily on foreign grants, although the level of foreign aid has fallen since 1992, when massive foreign aid was provided following the launch of the adjustment program. The return to lower levels of aid by bilateral donors has also  -6- reflected tightened budget constraints in donor countries and the suspension of program assistance (but not project assistance) to Zambia in 1996-97 because of donor concerns about governance issues. Table 1.3: Central Government Revenues, 1991-97 (percentage of Gnp) 1991 1992 1993 1994 1995 1996 1997 Total revenues, before grants 18.7 18.4 15.9 21.2 21.1 20.0 19.1 Total tax revenue 18.3 17.5 15.3 19.8 19.3 18.4 18.1 Direct taxes 5.0 6.1 4.5 6.1 6.1 6.0 6.1 Indirect taxes 13.3 11.5 10.8 13.7 13.2 12.4 11.9 On domestic goods & services 7.2 5.4 5.5 7.1 7.4 6.5 6.6 On international trade 6.1 6.0 5.3 6.5 5.8 6.0 5.4 Nontax Receipts 0.4 0.8 0.5 1.4 1.8 1.6 1.0 Memorandum Item External capital grants 9.0 10.2 8.0 5.6 5.5 5.9 4.8 Source: Ministry of Finance and Economic Development and International Monetary Fund Central Government Expenditures 1.16 F our noteworthy features characterize central government expenditures in Zambia. First, capital expenditures are funded by foreign grants, not from domestic resources. Second, debt service accounts for a huge proportion of total expenditures. Interest payments (before rescheduling) reached a peak of76 percent of total domestic revenues in 1993 before declining to 40 percent in 1997. Even after rescheduling, interest payments amounted to 21 percent of revenues and 40 percent of expenditures in 1997. By international standards these figures are high: among the 56 countries for which data are available in the 1998 World Development Indicators, interest payments averaged 13.6 percent of total expenditures. Very few countries in the world face this kind of debt overhang. To a large extent, the high level of total expenditures in Zambia reflects the high level of indebtedness. The share of wages and salaries in total expenditures (excluding foreign-financed expenditures), for example, was only 22 percent in Zambia, compared with about 26 percent in the other similar countries. Third, recurrent departmental charges account for a very small-and declining-proportion of the budget (Table 1.4). Fourth, since 1993 the pattern of expenditures has been shaped by the cash budget system, under which expenditures can be made only when revenues are actually received by MOFED. This system, described fully in chapter 3, served to control the runaway inflation of the early 1990s, but it has proved to be a source of many challenges for fiscal management and it has created distortions at the structural and micro economic levels.  -7- Table 1.4: Current Expenditures, 1991-97 (percentage of GDP) 1991 1992 1993 1994 1995 1996 1997 Total current expenditures (before RESCHEDULING) 29.3 27.2 26.3 29.5 282 22.2 20.6 Interest on external debt (before RESCHEDULING) 6.8 5.8 7.5 74 8.4 5.8 55 Interest on external debt (after RESCHEDULING) 68 58 7.5 74 59 16 18 Interest on domestic debt 1.7 2.7 4.8 4.8 27 3.0 2.1 Other current transfers 2.8 2.4 26 4.2 3.8 24 23 Subsidies 40 0.8 0.2 0.0 00 0.0 0.0 Consumption 140 15.5 11.2 13.0 13.3 11.0 10.8 of which wages and salaries 74 7.9 5.2 5.4 63 54 6I of which: recurrent departmental charges 4.1 3.7 3.7 3.0 4.2 34 2.7 Source: Ministry of Finance and Economic Development and International Monetary Fund D. THE NON-FINANCIAL PUBLIC SECTOR DEFICIT 1.17 The nonfinancial public sector consists of the central government plus local government, extrabudgetary funds, state-owned enterprises, and government guarantees on non-perfonning loans (Table 1.5). Estimating the nonfinancial public sector deficit reveals the effect of fiscal policy and the budget system on the deficit of the public sector as a whole. Because these accounts are being estimated for the first time in this report, we have tried to follow a consistent and conservative approach similar to the definition of the central government deficit. A deficit is defined as the excess of expenditures over revenues, where expenditures are defined as all obligations falling due in a particular year. This is a flow concept, as opposed to a stock concept, e.g., the stock of all contingent claims. Local Government 1.18 Comprehensive information on local government accounts is not available in Zambia. However, a recent Bank study found that local revenues are low and central government arrears to local councils are high. s Local government revenues, which come from property taxes, levies and charges, and government grants, generally comprise less than 1 percent of GDP. In 1997 they amounted to about K.25 billion. Government grants constituted 4 percent of these revenues in 1997, down from 10 percent in 1996. Based on grant information to councils, we estimate local government revenues at K.27 billion in 1995 and K.29 billion in 1996. Other studies have found that council revenues are able to finance about 20 percent of their cost of services. 6 Based on these estimates, we estimate Marianne Fay, "A Review of Local Institutions in Zambia," World Bank, draft, June 1998. 6 Bob Hawkins, LOGOSP, "Note on the Financing of Councils." In Marianne Fay, "A Review of Local Institutions in Zambia." World Bank, draft, June 1998.  - 8- local government deficits at KI00 billion in 1995 and KIlO billion in 1996. For 1997 our estimate of the local government deficit ofK120 billion includes local government arrears for salaries and pensions. Extrabudgetary Funds 1.19 Other sources of the nonfinancial public sector balance are extrabudgetary funds, including pension funds, the Rural Electrification Fund, the Road Fund, and the Privatization Fund. (The effective management of these funds is addressed in chapter 2.) 1.20 The Privatization Fund is the single extrabudgetary fund that has made a positive contribution to the budget. The Zambia Privatization Agency has been very successful in privatizing Zambia's parastatal sector, with 220 firms or units out of a working portfolio of331 privatized as of March 1997. The Privatization Fund brought in K9 billion in 1995, K37 billion in 1996, and K2 billion in 1997. 1.21 The real value of Zambia's pension funds has been seriously eroded over the years by inflation, low-yielding investments in government securities, and declining revenues from unemployment and public sector retrenchment. 7 Poor management and low levels of contribution compliance have caused expenditures to exceed revenues at the largest fund, the Zambia National Provident Fund (ZNPF), which has not been able to provide adequate retirement benefits to its members. The use of ZNPF funds to finance the budget deficit-50 percent of the investible surplus went to the government and another 35 percent went to loss-making parastatals-also hurt ZNPF perfonnance, since the funds were lent at low fixed interest rates, providing an implicit subsidy to the public sector. 1.22 Public sector employees make up about 50 percent of ZNPF membership, and only about 60 percent of employees are estimated to have contributed to the fund in 1994. Assuming that membership contributions comprise about 8 percent of total wages and salaries, the government's liability to ZNPF can be estimated at K2.8 billion for 1995, K3.6 billion in 1996, and K5.2 billion in 1997. These estimates do not include a provision for the implicit interest subsidy to the government. 1.23 The two pension funds specifically for public sector employees are the Civil Service Pension Fund (CSPF) and the Local Authorities Superannuation Fund (LASF). Both of these funds have provisions for early retirement, and the government is responsible for paying early retirement benefits. Responsibility for paying benefits due to retrenchments, which have been growing, is not clear. As indicated in Chapter 2, the government is supposed to pay these benefits but it is not. A study by the Bank and the International Labor Organization reports that as of 1994 the government's outstanding liability for such retirements was KI03 billion to CSPF and K16 billion to LASF, of 7 Monika Queisser, Clive Bailey and John Woodall, Refonning Pensions in Zambia: An Analysis of Existing Schemes and Options for Refonn (World Bank, Washington D.C.: Policy Research Working Paper 1716), January 1997.  -9- which K12 billion to CSPF and K4 billion to LASF was due immediately. For the purpose of our analysis, we included the portion of the liability for early retirement, projected forward to 1997, as part of the nonfinancial public sector deficit. State-0wned Enterprises 1.24 Although most state-owned enterprises in Zambia have been privatized over the past few years, public enterprises continue to contribute to the nonfinancial public sector deficit. The Zambia State Insurance Company (ZSIC) ran a deficit ofK13 billion and Kafue Textiles of Zambia ran a deficit ofK8 billion in FY96/97; ZCCM ran a deficit of K199 billion in 1996 (5.0 percent ofGDP) and K181 billion in 1997 (3.4 percent of GDP). 1.25 At a time when ZCCM's contribution to central government revenues (through corporate income tax and royalty extraction) declined sharply (above), its losses have created a huge implicit liability to the government. In order to finance its operating losses, ZCCM has relied primarily on issuing short-term debt with credit on inventory and future sales of copper. Ultimately, these losses will be born by the public sector, either because the debt will be taken over by the Government or because public assets will be depleted at the time of privatization. By early 1998 ZCCM's massive losses had reached a critical level, and the maturity of its debt had become increasingly short. The delay in the sale of ZCCM has made it difficult for the company to meet its obligations to suppliers, which include many other parastatals, such as the Copperbelt Energy Corporation (CEC), Zambia Electric Supply Corporation (ZESCO), and Zambia National Oil Company (ZNOC). As a result, these parastatals have incurred arrears to other parastatals and other public entities. For example, because it was not paid by ZCCM, CEC could not pay ZESCO. Since CEC accounts for 70 percent of ZESCO sales, ZESCO experienced an immediate cash flow problem and could not pay the Zambia Revenue Authority and other public entities, such as the Zambezi River Authority. Government Guarantees 1.26 The government has extended a number of guarantees to public enterprises that represent extensive contingent liabilities. As of the end of 1997 the government had a total contingent liability ofK26 billion on parastatalloans previously guaranteed by MOFED. Many of these guarantees were extended in the 1980s, although a large percentage of this liability is tied to guarantees issued in the past few years. Contingent liabilities for 1998 rise by K80 million when guarantees extended to ZCCM for financing from UBS and other foreign lenders are included. 1.27 The performance of government-guaranteed loans is extremely poor relative to that of nongovernment guaranteed loans, according to MOFED. Fewer than 1 percent of the loans guaranteed by the government as of December 1997 could be considered as performing, and much of the K26 billion in contingent liabilities reflects large interest charges that accrued after the borrower defaulted on loan repayment. Given that  - 10- repayment was required in 1997, the entire amount was therefore included as part of the nonfinancial public sector deficit for 1997. For 1998 the government anticipates this contingent liability (minus new guarantees to ZCCM) to fall as a result of the privatization of most of these enterprises and the restricted extension of government guarantees. Estimating the Non-Financial Public Sector Deficit 1.28 Under the above assumptions concerning accounts of the local authorities, extra- budgetary funds, and public enterprises (including government guarantees), the non- financial public sector deficit (NFPSD) of Zambia is estimated at about 10.4 percent of GDP in 1997 (Table 1.5). Of this amount, the Central Government deficit accounted for 2.3 percent ofGDP, financed by Treasury bills bought mainly by banks. The remainder of the NFPSD has been financed through arrears and through implicit asset depletion.  ~ 11 ~ Table 1.5 Zambia - Non-Financial Public Sector Deficit (in percent of GDP) 1995 1996 1997 Central Government Defic~t -4.6 -2.5 -2.3 Local Government3 Local Government Revenues 0.9 0.7 0.5 Local Government Expenditures 4.5 3.4 2.2 Local Government Arrears N/A N/A -0.3 Local Government Deficit ~3.5 -2.7 -2.1 Other Government Fundsb Pension Funds ZNPF - Noncompliance -0.1 -0.1 -0.1 CSPF and LASF Outstanding Claims -0.9 -0.8 -0.8 Outstanding Liability to Pension Funds -1 -0.9 -0.9 Privatization Fund Revenues 0.4 0.9 0 Expenditures 0.1 0 0 Privatization Fund Deficit/Surplus 0.3 0.9 0 Rural Electrification Fund (REF) N/A -0.1 0 Road Fund (RF) -0.2 -0.2 N/A Public Enterprises3 ZCCM SurpluslDeficit -1 -4.9 -3.4 Other SurpluseslDeficits 0.7 -1.3 -1.1 of which: energy companies 0.4 -1.2 N/A of which: transport companies 1.6 0.5 N/A Public Enterprise Deficits -OJ -6.2 -4.9 Central Government Guarantees on Parastatal Non- N/A N/A -0.5 performing Loans Nonfinancial public sector deficit -9.3 -11.6 -10.2 a Revenues eStimated based on share commg from grants. Expenditures estimated based on ratio to revenues Arrears includes salary and retiree arrears minus what local governments owed to pensions and parastatals b Governments' noncompliance to ZNPF estimated at 40% and assuming that membership contributions comprise 8% of wages and salaries Outstanding claims based on 1994 estimate for outstanding early retirement claims. c. Public Enterprise data is given by fiscal year. Non-ZCCM public enterpnse deficits for 1997 were estimated at 1% ofGDP, the same share as for 1996 Source: Bank of Zambia, MOFED, and Bank and Fund StafTEstimates  - 12 - E. THE QUASI-FISCAL DEFICIT 1.29 Conventional deficit measures that reflect only the operations of the central government or the non-financial puUic sector are often inadequate for determining the sustainability of a country's fiscal stance. Measures of either deficit may be too narrow because the central bank often performs quasi-fiscal activities that result in macroeconomic effects similar to those created by traditional fiscal activities. The public sector deficit may be significantly undervalued and the relationship among the deficit, inflation, and interest rates misunderstood ifthese quasi-fiscal activities are not considered. 1.30 A quasi-fiscal deficit is a deficit caused by central bank operations that are not directly related to financing the nonfinancial public sector deficit and that are not captured in the normal budgetary functions. Such operations may include provision of subsidized credit to public entities, provision of exchange rate guarantees to domestic residents wanting to borrow externally, transfers to public nonfinancial or financial institutions, and bailing out of failing industries. Because the central bank does not have the authority to tax economic activities to finance its operations, quasi-fiscal deficits are different from regular fiscal deficits. 1.31 A central bank can finance its quasi-fiscal operations by issuing high-powered money or interest-bearing liabilities. The printing of money, beyond seignorage, leads to inflation and erodes public confidence in the domestic currency. Issuing interest-bearing liabilities crowds out private sector activities, raises interest rates, and eventually causes a lack of public confidence in the monetary authority's ability to control inflation. Issuing debt without the counterpart of a performing asset could eventually lead to hyperinflation (which Zambia experienced in the early 1990s) and a collapse of the monetary system. The problem facing the central bank is that while its liabilities must be backed by issuing domestic currency, its assets are not being serviced. As a result, these assets are undervalued and debt cannot be rolled over without disturbing the monetary system. 1.32 As a result of pressure to cut the fiscal deficit and the need to deal with nonviable public enterprises the central bank has undertaken substantial quasi-fiscal activities. These activities have largely been in response to domestic financial and external payments crises, which have created the need to support distressed financial institutions and cover losses on guaranteed public sector liabilities denominated in foreign currency. 1.33 The central bank's quasi-fiscal activities need to be estimated and included in the overall fiscal deficit concept. Quantifying and estimating the quasi-fiscal deficit is difficult, however, for two reasons. First, there is no definition of the quasi-fiscal deficit that can be applied across countries, nor is there consensus on the best methodology for computing the quasi-fiscal deficit. Second, quasi-fiscal losses incurred by public financial institutions are often mixed in with operating losses, and it may be difficult to distinguish between normal monetary activities of the central bank and its quasi-fiscal  - 13 - actIvItIes. Different accounting conventions between the government and the central bank compound these complexities. For our purposes, we assume that the central bank should be financially sound in the absence of its quasi-fiscal activities. Central bank operational losses are thus amalgamated with the nonfinancial public sector deficit in order to estimate the quasi-fiscal deficit. 8 Oth~r quasi-fiscal activities of the central bank not included in the profit and loss account are also estimated and added into the adjusted fiscal deficit. Again to be consistent with the definition of the non-financial public sector deficit, we include only obligations falling due during a particular year in the estimate of the quasi-fiscal deficit for that year. 1.34 The data on quasi-fiscal activities in Zambia are limited and inconsistent. Data on net claims on the government and claims and liabilities reported by the Bank of Zambia versus those reported by commercial banks are inconsistent before 1997; data on loan guarantees extended by the government of Zambia as well as the Bank of Zambia are available for only limited time periods; data on the nonfinancial public sector deficit, especially data on local government arrears and government arrears to pension funds, are unavailable for some years. Given these limitations, the figures presented here should be viewed as rough estimates that indicate the presence of a considerable quasi-fiscal deficit rather than precise measures indicating the exact magnitude of all quasi-fiscal activities. 1.35 In order to evaluate the quasi-fiscal deficit in a consistent manner, we have included all potential fiscal and quasi-fiscal liabilities, even though every liability may not be called upon during the year. This methodology allows us to determine the extent of public sector commitments and to identify the corresponding limitations on private sector credit. We examine alternative scenarios and explain the range in which our estimates should be considered. Where current year data are unavailable, our estimates were made conservatively. Profit and Loss on Foreign Exchange 1.36 The Bank of Zambia's implicit bearing of foreign exchange risk for domestic enterprises that borrow externally has contributed to the quasi-fiscal deficit in Zambia. While the central bank no longer covers foreign exchange loss by the private sector, it continues to do so for the public sector. In most cases, the government of Zambia has accepted responsibility for such losses, but it has yet to reimburse the central bank accordingly. 8 Robinson and Stella (1988) and Le Houerou and Sierra (1993) offer a clear and appealing methodology for estimating quasi-fiscal deficits that is based on the practice of adjusting the fiscal deficit ex post through the addition of a transfer from government to the central bank, financed by credit from the central bank. Such an approach builds on the standard assumption that the central bank should be financially sound and should not incur operating losses in the absence of its quasi- fiscal activities. See P. Le Houerou and H. Sierra, "Estimating Quasi-Fiscal Deficits in a Consistency Framework." Washington D.C.: World Bank, 1993; D. Robinson and P. Stella, "Amalgamating Central Bank and Fiscal Deficits." In Mario Blejer and R. Chn, eds, Measurement of Fiscal Impact: Methodological Issues. Washington D.C.: IMF, 1998.  - 14- 1.37 Before December 1996 net losses arising from foreign exchange rate movements on assets and liabilities of the Bank of Zambia were recorded in the revaluation reserve account. This method of accounting was used in order to transfer these losses to the government and maintain the central bank's capitalization. Balances in this revaluation reserve account were covered by the issuance of nonnegotiable and noninterest-bearing government securities. Holdings of such securities, recorded as special holdings of government securities at the Bank of Zambia, amounted to K285 billion by the end of 1995. The new Bank of Zambia Act (No. 43/1996) requires that gains and losses arising on these translations be dealt with in the profit and loss account in order to comply with international accounting standards. 1.38 In addition to the special holdings of government securities, the Bank of Zambia has a portfolio of noninterest-bearing government securities to cover foreign exchange losses in connection with IMF Enhanced Structural Adjustment Facility (ESAF) loans. Although these noninterest-bearing government securities represent an asset of the Bank of Zambia, and have been recorded as such on their balance sheets, they have not been accounted for in net claims on the government or in financing of the central government deficit. This amount represents the exchange rate losses incurred by the central government in connection with external borrowing, and it should therefore be considered in the quasi-fiscal deficit, since foreign exchange resources will be needed to make debt service payments when these obligations fall due. 1.39 The Bank of Zambia has also accrued interest and exchange losses associated with past external debt obligations by state-owned enterprises. In 1992 the government initiated a debt buy-back program through which some of these obligations were settled at a discounted rate, inclusive of interest and exchange losses. A number of accounts have not yet been settled, however, and the government and the BOZ have not yet established who is responsible for the remaining liabilities. The central bank estimates the liability resulting from the exchange losses alone at K21 ,846 billion in 1996. Bank Failures 1.40 Zambia's financial sector has experienced problems in management, lack of confidence and stability, and a series of bank failures. 9 In late 1995 three major commercial banks went into receivership, at a large cost to the financial system. These bank failures complicated monetary policy and put pressures on the central bank to increase liquidity to cover their liabilities. Since 1995 the Bank of Zambia has undertaken a number of actions in order to strengthen its capacity to carry out its regulatory and supervisory obligations. Despite these efforts, three more banks were put 9 Bank failures in 1995/96 included Meridian BIAO Bank of Zambia Limited, which cost the central bank K43 billion; Africa Commercial Bank Limited (KIO.2 billion); and Commerce Bank Limited ( K9.5 billion). Bank failures in 1997 included Manifold Bank (K667 million), Prudence Bank (K10.8 million), and Credit Africa Bank (K15.5 million). Finance Bank failed in 1998 at a cost to the central bank ofK5.3 billion.  - 15 - into receivership in 1997 and one bank went into receivership during the first half of 1998. The central bank is estimated to have lost about US$1 00 million over the past three years in settling the net negative positions of these seven failed banks. Loan Guarantees 1.41 The Bank of Zambia has issued guarantees on loans for two state-owned enterprises, the Food Reserve Agency and ZCCM. These loan guarantees represent K48.6 billion in contingent liabilities. Although the Food Reserve Agency defaulted on its April repayment to Citibank, the central bank's separate loan agreement with the agency means that the guarantee need not yet be considered an actual liability. As of May 1998 loans to ZCCM were either being serviced or rolled over. ZCCM is not likely to continue to be able to service its debt, however, and these contingent liabilities will most likely fall under the quasi-fiscal deficit for 1998. Estimating the Quasi-Fiscal Deficit 1.42 Because losses associated with foreign exchange rate movements and financial instability are mixed in with the central bank's operating losses, we estimated the quasi- fiscal deficit by estimating the size of the loss from these two sources and then adding the cost of these quasi-fiscal activities to our estimate of the nonfinancial public sector deficit. Total quasi-fiscal deficit amounted to about 14 percent of GDP in 1997 (Table 1.6). 1.43 Of special importance to the estimate of the quasi-fiscal deficit is the Bank of Zambia's foreign exchange loss account for settling IMF debt. This account amounted to 12 percent ofGDP in 1997. If this account were included in the quasi-fiscal deficit, the deficit would rise from about 14 percent ofGDP to 25 percent ofGDP (Table 1.6). 1.44 With or without the account for settling IMF debt, the quasi-fiscal deficit in Zambia is very large; its magnitude clearly indicates that the figures for the central government deficit have provided a misleading impression of the viability and sustainability of Zambia's fiscal position. Despite the reduction in the central government's fiscal deficit, the public sector's need for financial resources severely constrains the financial resources available to the private sector. Given that the country's economic development program is relying on a private sector-led economic recovery, the level of the quasi-fiscal deficit needs to be addressed as soon as possible. Although the quasi-fiscal deficit as a percentage ofGDP has fallen over the past three years, its size indicates that government's indebtedness is much larger than previously recognized and partly explains the country's high interest and inflation rates (through the private sector's expectations).  - 16- Table 1.6 Zambia - Quasi-Fiscal Deficit (in percent of GDP) 1995 1996 1997 Central Government Deficit Excluding Grants -4.6 -2.5 -2.3 Nonfinancial Public Sector Deficit -9.3 -11.6 -10.2 Central Bank Foreign Exchange Losses Special Holdings of Govt Securities (prior to 1996) -9.1 -4.6 -2.9 Payments to Failed Banksa NA -1.5 -0.5 Central Bank Deficit -9.1 -6.2 -3.4 Quasi-Fiscal Deficit -18.4 -17.8 -13.6 Memorandum Items: GDP at Current Market Prices 2,826,2724,077,65 5,351,43 6 4 Holdings of Govt Securities for IMF -16.6 -16 -12 Stock of Contingent Liabilities/GDP 0.1 0.1 0.1 a Bank failures occurnng in late 1995 are included In payments made In 1996 Failures occurring In 1997 are included in payments made in 1997. Source: Bank of Zambia, MOFED, and Bank and Fund Staff Estimates- F. SOLVENCY AND SUSTAINABILITY ANALYSIS 1.45 Is Zambia's current fiscal stance sustainable? If not, how much additional fiscal reduction is needed? How will these fiscal adjustments be affected by external shocks? This section addresses these questions by summarizing the analytical results of solvency and sustainability analysis conducted on Zambia's public sector deficit. (See Annex 1.1 for a discussion of how such analysis was performed.) 1.46 The conventional practice of scrutinizing only the budget deficit to assess fiscal efforts and the sustainability of a fiscal stance is often inadequate because fiscal policy is an integral part of the overall policy framework and cannot be assessed independently. A budget deficit of 4 percent ofGDP, for example, may be acceptable in one country but not in another because of different policy conditions and initial endowments. Within the same country, a deficit level may be adequate one year but not the next depending on such factors as economic growth and policies to achieve external and internal balance. A target reduction in the budget deficit of 2 percentage points of GDP may not be adequate in an economy in which the exchange rate is being used as a nominal anchor, but it may suffice in a country in which the money supply serves as the anchor. Thus because of the  - 17 - strong linkages among fiscal, monetary, and other macroeconomic policies, the appropriateness of a particular fiscal deficit target depends on targets for growth, inflation, and external and internal debt. Unless these linkages are brought out explicitly, conventional fiscal measures (such as the deficit to GDP ratio or the government debt service to revenues ratio) shed little light on the appropriate level of the fiscal deficit and consequently on the appropriate speed of fiscal adjustment. 1.47 Solvency and sustainability analysis addresses these shortcomings of conventional analysis by incorporating the debt dynamics and other macroeconomic targets of growth, inflation, and external and internal debt. A public sector is said to be solvent when its discounted revenues are sufficient to cover the public sector debt (including external and domestic debt); it is said to be sustainable if this debt can be serviced without incurring new debt.lo 1.48 Solvency and sustainability analysis should be applied to the broadest definition of the fiscal deficit, in this case, the quasi-fiscal deficit. For illustration, this analysis is also applied to the central government deficit in this report. The common variable for such analysis is the primary deficit, defined as the overall fiscal deficit net of interest payments. In order to evaluate Zambia's fiscal stance, we first calculated the primary deficit that would make the public sector solvent given the stock of public debt at the end of the previous period, the real interest rate, and the real revenue growth rate. (Both the real interest rate and the real revenue growth rate were assumed to be constant throughout the period). This solvent primary deficit may be compared with the actual (or expected) primary deficit to find the adjustment needed to obtain public sector solvencyll. For fiscal sustainability, we first calculated the primary deficit needed to achieve debt sustainability (that is, to maintain the ratio of debt to GDP at the same level). A measure of the required adjustment is calculated by comparing this sustainable primary deficit with the actual (or expected) primary deficit. All variables are expressed as percentages ofGDP. 1.49 As of the end of 1997 , Zambia's fiscal balance remained far from the level required for solvency and sustainability, and at least a 3 percent fiscal adjustment would be needed to bring about solvency and sustainability in the central government deficit (Table 1.7). However, achieving fiscal solvency and sustainability in the public sector as a whole (the quasi-fiscal deficit) would require a fiscal adjustment on the order of 14 percent ofGDP. 1.50 This analysis did not take into account the dire financial situation of ZCCM in early 1998. Incorporating ZCCM's needs in 1998 would increase the level of fiscal adjustment needed. 10 See Buiter (1990), Agenor and Montiel (1996), Frenkel and Razin (1996). II This is equivalent to the so-called myopic one-period gap. see Olivier Blanchard (1990).  - 18 - Table 1.7. Public Sector Solvency and Sustain ability (percentage of Gnp) 1997 Solvency Primary balance 5.3 Real GDP growth 3.5 Domestic debt 4.6 External debt outstanding and disbursed 178.5 Primary balance for solvency 7.7 Central government solvency adjustment 2.4 Quasi-public sector solvency adjustment 14.0 Sustainability Grants 4.8 Sustainable primary balance 7.6 Actual central government primary balance 5.3 Central government sustainability adjustment 2.3 Quasi-public sector sustainability adjustment 14.5 Source: World Bank staff calculation 1.51 The Zambian authorities, as well as donors and creditors, will need to ensure that the social impact of the drastic fiscal adjustment needed is minimized. In order to understand the potential gravity of this issue, we ran additional simulations that examined the impact of slower export growth, lower grants, and a reduced external debt burden (Table 1.8). Reducing the annual rate of export growth from 5.8 percent to 2.9 percent in 1998 would require an adjustment effort of 18.7 percent ofGDP. This result reveals just how difficult the situation is in Zambia and highlights the need for external debt relief. If both the stock of external debt and the international rate of interest were reduced by half, the fiscal adjustment required to restore solvency and sustainability would be reduced to 10.7 percent ofGDP.  - 19 - Table 1.8: Alternative Scenarios for Fiscal Adjustment (percentage of GDP) 1997 Base case Government solvency adjustment 16.0 Government sustainability adjustment 16.5 Reduced export grbwth (2.9 percent instead of 5.8 percent) Government solvency adjustment 18.4 _Government sustainability adjustment 18.7 Reduced external debt burden (50 percent debt burden) Government solvency adjustment 10.1 Government sustainability adjustment 10.7 Source: World Bank staff calculation G. A PROGRAM OF FISCAL ADJUSTMENT 1.52 The above analysis has two important implications. First, non-central government public sector activities, particularly the parastatals, have become a major drain on public resources in Zambia. Until 1996, these activities were not as important in part because the favorable world commodity prices made them profitable. However, since 1996, when world commodity prices began to decline sharplyl2, management problems of public enterprises became acute, and problems with the pension system were exacerbated, fiscal measures at the central government level have not been adequate to improve the overall fiscal position of the country. Fiscal adjustments are now needed at all levels of the government, including the central government, public enterprises, the local government, and the Central Bank. 1.53 Second, it is also clear from the analysis of solvency and sustainability that even with the best efforts, Zambia cannot do it alone. A successful fiscal adjustment program would require sustained efforts from the government as well as those of Zambia's deVelopment partners, including the creditors, NGDs, donors, and multilateral institutions. Specifically, Zambia's debt burden needs to be reduced to a manageable level, while adequate funding and technical assistance at all government levels need to be provided on a sustained basis. This is particularly important because current fiscal reforms have to be implemented in a difficult macroeconomic environment. The 12 The World Bank index of 33 non-oil commodity prices, in current dollars and weighted by 1987- 1989 developing countries exports, shows a rising trend between 1991-1995, a decline of almost 6 percent in 1996, and a further decline of 16% in 1998.  - 20- economy is experiencing a recession in 1998 and is likely to have a weak growth recovery in the following year. This outcome is due in part to adverse terms of trade shocks and in part to failure of the authorities to implement the needed reforms in a timely manner. A Program of Fiscal Adjustment 1.54 While the magnitude of the overall fiscal imbalance in Zambia, at 14 percent of GDP, is large, international experience indicates that this deficit can be successfully addressed if the government, with the assistance of the international institutions, can identify and implement a realistic medium-term program of fiscal adjustment. Within this program, the government has to demonstrate maximum efforts to improve its effectiveness in spending and trim inefficient expenditures along the lines of its medium- term strategy of achieving fiscal balance and a sustainable and significant reduction in poverty. The international community can then help fund the remainder ofthe program, a large part of which should be accounted for through provision of HIPC debt relief. Details of the program can be worked out in the preparation of the 199.1-2001 Policy Framework Paper (PFP), a joint document of the government and World BanklIMF staff. 1.55 Fiscal adjustment at the level of central government needs to be pursued vigorously. The 2 percent central government budget deficit (after grants) in 1997 represented progress toward a healthy fiscal stance. Nevertheless, fiscal efforts need to continue to focus on and enhance the quality of fiscal adjustment, particularly with regard to how fiscal adjustment is achieved and whether or not it contributes to fiscal sustainability. As noted earlier, fiscal progress at the central government level has been erratic and was often achieved at the cost of efficiency and equity of public expenditures. In an environment of low growth and serious fiscal constraint, the effectiveness of spending is ever more important to maintain a rigorous program of poverty reduction. 1.56 A key policy measure to sustain fiscal reforms at the central government level is the revised public sector reform program (PSRP) launched by the government in September 1997. The program aims at improving the quality, efficiency and cost effectiveness of public services through measures to reduce surplus workers, to rectify the incentives needed to attract and retain good workers, to rationalize personnel policies, and to strengthen accountable management with a view to improving service performance. While the government has made some progress toward these goals, the broad deterioration in the overall economy in 1998 combined with internal management problems have caused delays in the phasing of the PSRP and changes in the action plan. It is important that the new, revised action plan and implementation of the PSRP be finalized and implemented as soon as possible. 1.57 Another factor affecting the quality offiscal adjustment is the cash budget system, the problems of which are addressed in detail in Chapters 3 and 4. This report makes recommendations to improve the cash budget system in the short term and to prepare  - 21 - Zambia for moving toward a Medium-Term Expenditure Framework where expenditure policies and priorities can match development objectives. Finally, improving economic governance will also affect the quality of fiscal adjustment. Greater transparency, less government intervention, and better working conditions for civil servants will combine to minimize the desire and opportunity for corruption, thereby contributing to the effectiveness of the public sector. 1.58 At the level of local government, fiscal adjustment needs to focus on two areas. First, a more detailed and better assessment of the accounts of local authorities needs to be made, on the basis of which an action program to reduce sharply the deficit of the local authorities needs to be adopted. Second, the government needs to improve local government performance to ensure that basic services are delivered effectively. Currently, local governments are unable to function effectively as their responsibilities are unclear and their resources are limited. Inefficient council management, bloated payrolls and inappropriate allocation of resources between salaries and other outlays contribute to their inability to deliver basic services. To address these problems, the government has proposed a gradual decentralization of a substantial part of the public service functions over the next ten years. The first step is the adoption of a draft policy paper on decentralization which articulates local government responsibilities in the first phase of decentralization. The paper also contains a fiscal framework that promotes efficient behavior on the part of municipalities, ensures a minimum of funding, and enables municipalities to increase their capacity for self-financing. Fiscal Adjustment of State-Owned Enterprises 1.59 Of the 10.4 percent non-financial public sector deficit in 1997, about half is accounted for by the deficit of state-owned enterprises. Macroeconomic stability can therefore be restored in Zambia only if this deficit can be reduced to a sustainable level. A vigorous privatization program is therefore of top priority not only to rapidly stop this drain but also to improve productivity and boost output of the economy. The current privatization program started in 1993 with a list of about 138 companies to be privatized. The portfolio has now increased to about 272 entities, of which 222 have been privatized 1.60 At the top of the privatization list is the ZCCM. While about 40 percent of ZCCM assets have been sold since 1997, including four small copper mines and ZCCM's Power Division, privatization of the remaining four large assets, the Konkola, Nchanga, Nkana and Mufulira Divisions, has still not been completed. The impact of current production losses on the budget, on supply of foreign exchange, on production and on employment make it imperative that the sale of these four Divisions be completed as soon as possible. 1.61 With regard to the public enterprises in the non-mining sector, the privatization program needs to be pursued vigorously. The top priority at this time is to reach a decision whether to complete the privatization or to liquidate the 50 commercial entities whose preparatory work has been under way. In the petroleum sector, the government  - 22- should pursue a policy to liberalize the retail distribution system for petroleum products and to remove all restrictions on their imports. To cut down on economic losses of ZNOC (para. 1.25) and INDENI refineries, there is a need to adopt a plan of actions based on findings of the fuel options study and the review of the institutional framework for the petroleum sector. 1.62 In the power sector, there is a need to commercialize the operations of ZESCO while encouraging private sector participation in power generation and transmission. The government needs to separate ZESCO's generation, transmission and distribution functions and contract-out ZESCO's distribution system. In the telecommunication sector, completion of ZAMTEL privatization is a priority. In the transport sector, the government has now included Zambia Railways in the tranche of public sector companies to be prepared for privatization. A SIDA-financed management team is in place and is trying to reverse the railways' worsening performance. For government departments, a decision on whether over 30 departments should be privatized, commercialized, closed down or kept as government departments should be made as soon as possible. 1.63 Finally, although the large quasi-fiscal activities of the government pose a serious constraint to the conduct of monetary policy, in the context of Zambia's adjustment program, measures to clean up the portfolio of the central bank and of the public sector banks need to be sequenced appropriately. Thus one should not expect these operations to take place before the above privatization measures in the real sectors have been implemented. However, because it takes a long time for financial operations to take effect, work in this area needs to be initiated as soon as possible. In particular, the Central Bank and the public sector banks should continue to refrain from offering any guarantees, whether on foreign exchange or on interest rates, to public or private entities. The government should discontinue its direct involvement in retail credit operations and to begin the necessary work to privatize the state owned institutions in accordance with the provisions of the Privatization Act. Good candidates in this area include the Zambia National Commercial Bank and Zambia State Insurance Company, the Zambia National Building Society, and the National Savings and Credit Bank. H. SUMMARY OF RECOMMENDATIONS 1.64 Although Zambia has made significant progress in reducing its fiscal deficit in recent years, fiscal and quasi-fiscal activities in the economy remain extensive and further adjustments are needed to attain internal and external balance. In particular, efforts to control central government expenditures need to be complemented by other actions to bring the non-financial public sector deficit and the quasi-fiscal deficit under control. 1.65 While it is beyond the scope of this report to provide detailed fiscal prescriptions, the preliminary analysis of this chapter indicates that:  - 23- • Fiscal and quasi-fiscal activities at all government levels need to be taken into account in the design of an appropriate stabilization cum structural adjustment program; • The magnitude of the fiscal problem in Zambia is so large that it would take a long time, under the most serious adjustment efforts, to restore macrostability. Expectations concerning a rapid return to macrostability therefore would need to be tempered. • The preliminary results of this report should be followed up by further analyses to firm up data and conclusions, to identify the extent to which macro stability has been adversely affected by fiscal and quasi-fiscal activities, and to search for the desired sequencing of appropriate fiscal and quasi-fiscal measures. 1.66 Finally, although the current level of the quasi-fiscal deficit is large, at 14 percent ofGDP, not all of its components call for urgent fiscal measures. Some, such as the deficits of the SOEs, do require immediate measures to stop the drain on public resources. Others, such as the cost of Central Bank recapitalization, could be spread out over a number of years. The timing of these policy measures, and their interaction with one another and with other macroeconomic policies will be critical to the success of the adjustment program. 1.67 In the short-term, the government could take a number of actions: • Accelerate the privatization program, especially the sale ofZCCM's remaining assets, and use the proceeds to reduce the public sector liabilities, including those of the central bank. • Freeze arrears by all public entities and require all public agencies to come up with a plan for a speedy resolution of these arrears before new funds are released to them in the next budget year. • Collect information on and monitor the non-financial public sector deficit and the quasi-fiscal deficit in addition to the central government deficit and improve its estimates of these deficits. • Set out a medium-term program to bring these deficits to a sustainable level. • Design a plan to restore the solvency of the central bank and other public sector financial institutions. • Improve the management of expenditures, especially the cash budget, to attain the medium-Jerm fiscal targets.  - 24- CHAPTER 2 EXTRA-BUDGETARY FUNDS 2.1 As a result of the legacy of state intervention in the Zambian economy, many institutions owned or supported by the central government have come to expect government funding. Because these institutions-local governments, statutory boards, and state-owned enterprises-are not included in the central government budget, their role and contribution to the public sector have remained hidden in the policy debate. 2.2 Until recently, these institutions were not required to charge user fees or to look for new sources of income. If a shortage of funds existed, the government guaranteed commercial loans for these institutions or incurred debt itself on their behalf. 2.3 Extrabudgetary funds include funds for grant-aided institutions, such as universities or special agencies; local governments; special funds, such as the Rural Electrification Fund and the Road Fund; 13 and pension funds. Grants to these institutions in 1997 and 1998 are shown in table 2.1.14 \3 Although revenues from these funds are included in the central budget, they are regarded as extrabudgetary because they are fmanced by surtaxes imposed on users. 14 Because many of the issues addressed have not been documented, the analysis presented here must be considered tentative until subsequent studies confirm its findings.  - 25- Table 2.1. Grants to Extra-Budgetary Funds (millions of Kwacha) 1997 1998 % of Domestic Grant % of Domestic Fund Grant Amount Govt. Revenues Amount Govt. Revenues Local Councils 1,572 0.16% 2,420 0.21% District Health Boards 15,360 1.59% 23,519 2.02% University of Zambia 9,711 1.01% 9,500 0.82% Copperbelt University 3,592 0.37% 3,000 0.26% CSPF Board 24,000 2.49% 20,000 1.72% LASF Board 61 0.01% 100 0.01% Drought Relief Fund 0 5,000 0.43% Food Reserve Agency 0 11,124 0.96% Total Grants 153,629 10.31% 216,385 11.90% Source: Ministry of Finance and Economic Development A. GRANT-AIDED INSTITUTIONS 2.4 Grant-aided institutions, most of which fall under line ministries, supplement grant revenues by charging surtaxes, levies, and user fees. Examples of grant-aided institutions include local governments or councils (under the Ministry of Local Government and Housing), universities (under the Ministry of Education), the Zambia Doctors Flying Services (under the Ministry of Health), Zambia National Broadcasting Services (under the Ministry oflnformation and Broadcasting), the Lusaka Water and Sewerage Company, and the National Housing Authority. 2.5 The legal basis of the extrabudgetary funds and levies is defined by the acts of Parliament that established the grant-aided institutions. The act creating each grant-aided institution defines its role and functions, management and administration, and mechanism for funding. The surtaxes, levies, and user fees that these institutions charge are usually subject to ministerial or Cabinet approval. 2.6 Although most grant-aided institutions are specialized government agencies or statutory bodies created by acts of Parliament (which either define how they are to be funded or raise revenues to sustain their operations), the Budget Office of MOFED lacks detailed information on the revenues generated by the institutions, which submit their budget proposals through their respective line ministries. IS In some cases even line ministries are not provided with full information on the revenues earned by the grant- aided institutions. The University of Zambia, for example, has not fully disclosed its revenues in recent years. Although the Budget Office of MOFED is not provided with detailed revenue and expenditure information, it is nevertheless expected to fund the 15 By law such submissions are supposed to go directly to MOFED.  - 26- short-fall between projected revenue and projected expenditure for grant-aided institutions. 2.7 Lack of adequate information on the levels of revenues earned by the grant-aided institutions means that there is no way of knowing if government grants are being applied optimally. Some MOFED officials believe that large city councils, such as those in Lusaka, Ndola, and Kitwe, have adequate sources of income to sustain their operations without government grants. With improved transparency and accountability, they argue, such institutions could be weaned from the government budget and resources could be directed to other areas. 2.8 The case of the University of Zambia illustrates how a grant-aided institution draws resources from the central budget. The University was established by the University Act of 1965, which defined its main functions as encouraging the advancement of learning and research throughout Zambia and providing all Zambians with the opportunity to acquire higher education. Since 1965 the University Act has been revised four times. 16 Under the 1992 University Act, which provides for the registration and regulation of public and private universities, the government may award grants to any public university out of monies appropriated by Parliament for the purpose of enhancing learning and research in the country. The University Council can charge fees and receive money by way of grant or donation or in any other way from any source. It must prepare and submit to the Auditor General statements of income and expenditure, assets and liabilities, and progress reports. The Act also requires the Auditor General, or an auditor appointed by him or her, to audit the accounts and submit the report and financial statements to the Minister of Education for tabling in Parliament. 2.9 During the 1995 and 1996 period, the government provided the University of Zambia with K18.5 billion in grants and K3.4 billion in scholarships. The University earned K3.8 billion from other sources, such as consultancy, tuition, interest on investments, and rent. In 1997 government funding to the University was K9.7 billion in grants and about K2.5 billion in scholarships; for 1998 K9.5 billion in grants was budgeted, of which K3.9 billion had been paid as of June 23, 1998. 2.10 The University has experienced recurrent financial crises. An audit by the Auditor General's Office of the University's accounts for the 1995 and 1996 fiscal years highlighted the lack of accountability and inadequate enforcement of financial regulations. 17 Major findings of the report include the following: 16 The 1987 Act that repealed the 1979 Act established two universities, the University of Zambia and the Copperbelt University. 17 Republic of Zambia. 1997. Report of the Auditor General on the Review of the Operations of the University of Zambia for the Financial Years Ended 31 st December 1995 and 1996. Lusaka: Government Printers.  - 27- • The internal control and internal audit of the University are very weak. Management does not act on internal audit reports. Payments are almost never referred to internal audit for preaudit and are sometimes effected through letters to banks unaccompanied by payment vouchers. Accounting records such as cash books are poorly maintained, and bank reconciliation is not up to date. • Although it incurred auditing and accountancy fees of K81 million between 1988 and 1996, the University did not prepare final accounts for presentation to Parliament, as required by the University Act. Without such final accounts, it is not possible to ascertain whether funds were used for their intended purposes. • Financial management ofthe University is weak. The University operates many bank accounts, which were not reconciled, as required by law. Failure to reconcile these accounts means that it is not possible to monitor the movement of funds; as of April 1997, it was difficult to determine the University's cash position. University funds were deposited in investment accounts without prior authority of the Minister of Education, as required by the University Act; poor recordkeeping made it impossible to determine how much interest was earned and received on such investments. Shares in the ZAMNET Communications System were issued to senior managers and not placed in trust by the University. • No inventory of the University's assets has been conducted since 1983. 2.11 The Auditor General's report demonstrates the need to improve transparency, accountability, and enforcement of financial regulations to ensure that government and public funds are managed properly. 2.12 Other problems are also troublesome: • Budget estimates submitted by the universities do not identify how they were made, and they tend to overstate expenditures and understate revenues in order to maximize government grants. • The universities often overlook the Budget Officer of the Ministry of Education and deal directly with the Budget Officer of MOFED. • The pattern of university enrollment does not reflect Zambia's human resource needs. Humanities, social sciences, and natural sciences continue to dominate enrollment, despite the fact that graduates in these disciplines remain unemployed for long periods of time. • Allowances paid to university students remain too high. In some cases, university students receive more in scholarships than entry-level civil servants  - 28- earn in salary. In a country in which more than 60 percent of the population live in abject poverty, a review of student allowances is imperative. 2.13 The Budget Office of MOFED should not confine its monthly budget to line ministries but extend it also to grant-aided institutions. Additional funding should be made conditional on submission of acceptable itemizt!d returns that detail how grants are spent and identify other revenues the grant-aided institutions generate. Annual funding should be conditional on production of audited final accounts for the previous year, as required by law. B. LOCAL GOVERNMENT 2.14 Local government revenues are extrabudgetary because they are not included in the central budget; council expenditures are channeled through the Ministry of Local Government and Housing for inclusion in the central budget. 2.15 Under the 1995 Local Government Act, councils may impose levies, fees, and charges to raise revenues. They are supposed to keep accurate accounts of their revenues and expenditures and to prepare and adopt annual estimates of revenues and expenditures, which should be submitted to the Minister of the Ministry Local Government and Housing (MLGH) for approval at least 60 days before the start of the financial year. For its part, the government is supposed to provide councils with grants for water and sanitation, health services, fire and road services, primary education, and agricultural services. MLGH may also decide to make constituency development grants or loans to councils. 2.16 Regulations for the control and management of council finances are to be established by the Minister of Local Government and Housing by statutory instrument, and an auditor appointed by the Minister is to audit each council's financial statements every year. The Minister of MLGH may suspend or withhold grants to a council if it fails to transmit a copy of the auditor's report to the Minister within 60 days of receipt of the report. 2.17 Although the Department of Local Government Administration (Finance) normally issues budget preparation guidelines to councils on time, submission of estimates of revenues and expenditure is usually late. Other problems include the following: • Resource constraints, poor controls, lack of accountability, and political interference adversely affect the capacity of local governments to deliver services. While residents bear the burden of personal levies; fees, and other charges imposed by the councils, the quality of services provided is inadequate or nonexistent. • Payment of grants by the central government is late or nonexistent.  - 29- • The basis upon which the Budget Office calculates the council grants is unclear. The final figures differ substantially from those submitted by the Department of Local Government Administration. • Council audits reveal gross financial mismanagement and lack of discipline. Although the Parliamentary Committee on Local Administration summons local authority officials to respond to queries raised in the audit reports, few corrective measures have been implemented to improve accountability and financial management. • The central government's arrears for rent, water and sewerage charges, and other services rendered by councils continue to adversely affect their operations. As of July 31, 1997, the government's arrears to councils had reached almost Kl 0 billion (table 2.2). City councils are most severely affected by the government's arrears. Since 1997 the government's budgetary problems have worsened; as a result, many councils are having difficulty paying salaries and providing other essential civic services, such as water and public health facilities. • Local taxes and charges account for too large a share of council revenues. The 1996-97 estimates (table 2.3) show that the share of local taxes and charges in total local authorities revenues was expected to increase from 49 percent in 1996 to 63 percent in 1997. Over the same period Central Government Grants (national support) was estimated to decline from 8 percent in 1996 to 3.5 in 1997. • Council expenditures are dominated by personal emoluments and other expenses, with very little spent on their provision of services. For example, over 1996-97 period, personnel emoluments and other expenses (such as running costs, fuel and repair expenses, and councilor allowances) accounted for at least 88 percent of estimated total expenditures. This contrast with their provision of services (refuse collection, markets, water and sewerage services) which accounted for only 12 percent of estimated total spending.  - 30- Table 2.2: Ministry of Local Government and Housing Central Government's Indebtedness to Councils as at 31 July 1997 (millions ofKwachas) Central Government's indebtedness by province House rent Water Rates Ot"er Total Central Province 301.6 178.2 15.8 0 495.7 Copperbelt Province 597.0 4,478.0 223.7 1,037.4 6,336.0 Eastern Province 218.5 4.1 0 0 222.5 Luapula Province 98.5 147.4 2.3 1.6 249.8 Lusaka province 361.8 52.5 16.8 0 431.0 NlWestern Province 135.0 45.8 1.1 7.6 189.5 Northern Province 456.2 237.2 93.5 192.7 976.6 Southern Province 401.7 268.4 66.7 736.8 Western Province 285.0 0 0 5.7 290.6 Totals 2,855.2 5,411.4 420.0 1,245.0 9,931.6 Central Government's indebtedness by council type City 752.4 3,930.1 0 826.2 5,508.8 Municipal 857.7 1,144.5 226.1 218.9 2,447.1 District 1,245.1 336.8 194.0 200.0 1,975.8 Totals 2,855.1 5,411.4 420.0 1,245.0 9,931.6 Source: Ministry of Finance and Economic Development  - 31 - Table 2.3: Council Revenues and Expenditures 1996-97 (percent) Revenues 1996 1997 Expenditures 1996 1997 General Rate and Develo~ment Funds Accounts Local Taxes 23.4 41.3 Personal 33.9 39.4 Emoluments Charges 19.7 21.3 Provision of 8.0 10.0 Services Other Receipts 27.6 25.8 Other Expenses 39.3 43.9 National Support 8.2 3.5 Liquor Undertakings 0.6 0.6 Rest HouseslMotels 0.5 0.5 ShopslMarkets 0.1 0.1 Other Undertakings 0.2 0.3 Water Fund Account Sale of Water 11.7 6.2 Personal 5.0 1.7 Emoluments Sewerage Charges 5.5 0.1 Provision of 4.3 1.8 Services Other Income 2.5 0.3 Other Expenses 9.4 3.2 TOTAL 100 100 100 100 Source: Ministry of Local Government and Housing, Summary of Local Councils Revenue and Capital Estimates, 1996 and 1997 2.18 Given the poor performance of local government and the low income levels in Zambia, an increase in service charges could be counterproductive. Instead, a major overhaul of the local government system is needed to improve funding, accountability, and service delivery. The proposed decentralization policy should focus on how the councils can increase their efficiency by becoming more cost-effective and accountable. Local authorities are, at the moment, not viable since the system for local administration is not clearly defined and open to different interpretations. C. SPECIAL FUNDS 2.19 The largest special funds in Zambia are the Rural Electrification Fund and the Road Fund. The central government's arrears to both funds are significant and are adversely affecting the funds' ability to deliver services.  - 32- Rural Electrification Fund 2.20 The Rural Electrification Fund was created in January 1994. It initially received 3.45 percentage points of the 23 percent sales tax on electricity bills, to finance the expansion of the electricity network into rural areas. 2.21 The Rural Electrification Fund finances projects that have strong potential for increasing agricultural or industrial/commercial growth; addressing the need for more equitable regional distribution of income while balancing cost considerations; and improving rural living conditions by providing power to social amenities, such as health, schools, and community centers. 2.22 The Rural Electrification Fund was not originally an extrabudgetary fund, since it was financed from a portion of the excise duty due to the government. It became an extrabudgetary fund on July 1, 1995, when, following the introduction of a 20 percent value-added tax (V AT), 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. The Ministry of Energy maintains that the entire 10 percent is supposed to be channeled to the Rural Electrification Fund. The issue is currently being debated and appears far from resolution. 2.23 In theory, the Zambia Revenue Authority collects the 10 percent electricity levy from the Zambia Electricity Supply Corporation (ZESCO) and deposits the proceeds in the government's revenue account at the Bank of Zambia. The Budget Office then writes a check to the Ministry of Energy, which deposits the check in the Rural Electrification Fund account at Zambia National Commercial Bank. 2.24 In practice, MOFED has transferred no funds to the Ministry of Energy since 1995. During 1996 and 1997 MOFED collected K27.3 billion in electricity levies, of which K5.8 billion was due to the Rural Electrification Fund (table 2.4). On May 4, 1998, the Treasury credited the Rural Electrification Fund account with K600 million, and regular monthly payments were promised. No payment was made in June, however. MOFED's arrears to the Rural Electrification Fund, up to the end of April 1998, amounted to more than K5.2 billion.  - 33 - Table 2.4: Funding ofthe Rural Electrification Fund, 1994-98 (millions of Kwachas) Total Amount Amount due to REF Actual Amount Year Collected (100%) (30%) Remitted to REF 1994 9,449.8 1,417.5 937.6 1995 2,983.1 502.3 2,480.0 1996 17,186.1 4,457.1 no payment 1997 10,193.3 1,335.0 no payment 1998 5,120.0 1,536.0 600.0 Source: Ministry of Energy and Water Development 2.25 The Rural Electrification Fund is administered by a committee chaired by the Ministry of Energy. Members include: the Ministry of Local Government and Housing, the Engineering Institute of Zambia, and ZESCO (the implementing agency). Applications for funds are channeled through provincial permanent secretaries and submitted to the committee, which decides which applications to fund. Once funding is approved, ZESCO is hired to provide electrification. 2.26 The committee has not met since 1996, severely constraining the Rural Electrification Funds' ability to electrify more rural areas. The loss by the Fund of almost Kl billion deposit as a result of the failure of Meridian BIAO Bank has also reduced its ability to provide services. The Road Fund 2.27 The National Roads Board was constituted by Statutory Instrument No. 4211994 to administer and manage the Road Fund. Funded by the fuel levy and other resources allocated by Parliament and donors, the fund maintains roads in Zambia. The National Roads Board reports to a committee made up of the Ministers of Energy, Local Government, and Communications and Transport that is chaired by the Minister of Communications and Transport. 2.28 The Zambia National Oil Company collects the 10 percent surcharge on gasoline, which it hands over to the Zambian Revenue Authority, which sends the funds to MOFED. 18 The National Roads Board receives Road Fund proceeds through the Ministry of Communications and Transport. 2.29 MOFED's arrears to the Road Fund exceed K20 billion, not including road user charges (international transit tolls, weigh-bridge fines, and vehicle license fees). These arrears include K5.2 billion lent by the National Roads Board to the Ministry of Works and Supply in 1995 to finance budgeted road works that were not financed by the Road 18 At one time, the Zambia National Oil Company deposited the proceeds of the fuel levy directly into the National Roads Board account. This practice was stopped after the Zambia Revenue Authority claimed that it was the only agency authorized to collect revenue on hehalf of the government.  - 34- Fund, K9.8 billion in unpaid fuel levy proceeds from 1997, and K5 billion of the KI0 billion fuel levy proceeds that had been collected as of March 31, 1998. A draft Cabinet Memorandum (which is one of the conditions for accessing the ROADSIP IDA Credit) on this has been prepared but not yet presented to Cabinet. 2.30 The National Roads Board has another K6 billion in arrears with MOFED that represent counterpart funding for ROADSIP. Without counterpart funding, IDA resources cannot be accessed to implement ROADSIP. 2.31 Because of these arrears, the Road Fund owes road contractors more than K4 billion and is incurring interest and exchange loss charges. In addition, it has about K14 billion of work-in-progress and fee notes. Faced with such a situation, the National Roads Board suspended commissioning new contracts for road repairs until the fuel levy arrears have been cleared and all contractors have been paid. D. PENSION FUNDS 2.32 Pensions funds in Zambia are under severe financial stress as a result of government interference in their operations, poor management, lack of accountability, low interest rates on member contributions, and high rates of inflation. A recent study concluded that all of the Zambian pension schemes have major weaknesses due to deficiencies in design, financing, and administration. 19 The retirement benefit rules of the schemes are too costly, and the funds are mismanaged. Mismanagement is compounded by a lack of appropriate legislation, regulation, and supervision; the absence of secure and profitable investment opportunities, which confines the funds to Treasury bills and real estate; and the dearth of qualified management personnel. None of the Zambian pension schemes provides adequate 2.33 The study presents various alternatives for improving pension and social security system. It also discusses the need to revise the legal framework to ensure portable pension schemes and transparent management systems. 2.34 The 1996 Pension Scheme Regulation Act that established the Insurance and Pensions Authority addressed many of the concerns raised by the study. However, the Act still leaves pension schemes vulnerable to political considerations. For example, under the Act pension fund managers are allowed to invest only in investments approved by the Registrar, and the Minister of Finance regulates the minimum interest rate to be applied for calculating portable benefits. The Act also authorizes the Minister to issue investment guidelines that establish the upper limits for investment categories, and it bans pension funds from making unsecured loans to affiliated employers or investing its assets abroad. 19 Monika Queisser, Clive Bailey, and John Woodall. 1997. "Reforming Pensions in Zambia: An Analysis of Existing Schemes and Options for Reform." Policy Research Working Paper 1716. World Bank, Washington, D.C.  - 35 - 2.35 The Civil Service Pension Fund was established under the Pensions Act of 1968 as a benefit scheme to provide retirement pensions to permanently employed civil servants. The fund operated as a government department until 1986, when it was transfonned into a corporate body. It opened its own account with the Bank of Zambia in December 1990. Since 1991 all members have made their 7.25 percent contribution; the government, however, has not made its 7.25 percent employer contributions to the fund since 1991, and its arrears are estimated at K45-K50 billion. 20 2.36 The 1968 Pensions Act was repealed by the 1996 Public Service Pensions Act, which established the Public Service Pensions Fund Board. The board is required to submit to Parliament an annual comprehensive report covering the operations of the fund, which serves civil servants, police, prison staff, teachers, members of armed forces, and jUdiciary staff. 2.37 The chairman of the Public Service Pensions Fund Board is appointed by the President of Zambia. Directors include the pennanent secretaries for the Ministry of Labour and Public Service Management Division at the Cabinet Office, the Budget Director of MOFED, the Attorney General, the National Secretary of the Pensioners Association, the General Secretaries of the Civil Servants and Teachers Unions, representatives of the defense and security forces, representatives of the Chamber of Commerce and the Lusaka Stock Exchange, and two persons appointed by the President. 2.38 The board continues to respond to political influence. In June 1998, for example, the President appointed a new board chairman, whose first action was to fire the chief executive of the Fund. 2.39 Based on the principle of pooling resources and sharing risks, the fund is expected to pursue the following objectives: . • Undertake only business activities that are consistent with the basic objectives of the fund and the interests of fund members. • Distinguish between employment conditions of service (such as those relating to retrenchment and early retirement) and eligibility for pension benefits. • Maintain the real value of benefits. • Make adequate arrangements for the transfer or preservation of pension rights. • Maintain a realistic contribution structure. • Make adequate arrangements to avoid noncompliance. 20 The mission had access to the 1985 reprint of the 1976 regulations, which seems to be the most recent  - 36- • Avoid delays and other constraints to the effective provision of adequate benefits. • Maintain current records of contributions. I • Curb excessive administrative costs and ensure prudential management of fund assets. 2.40 In the context of the current Public Service Reform Program, which aims, among other things, at reducing the size of the public service, the constitutionally guaranteed pension benefits are proving to be too expensive for the central budget to fund. The retirement age for civil servants is 55 years; defense, police, and prison service employees retire at age 45 or after 20 years of continuous service, whichever comes first. Employees who retire at the normal retirement age, survivors of employees who die in service before reaching retirement age, and employees who retire on medical grounds before reaching the official retirement age are entitled to a lump-sum retirement package plus monthly stipends payable by the fund. Public service employees who are retrenched or forced to retire prematurely because their jobs are abolished are also entitled to full retirement benefits. For these employees the government is supposed to pay their monthly benefits until they reach the normal retirement age, at which time benefits are paid by the fund. Currently, all benefits are being paid by the fund, however, whose monthly benefit bill currently stands at K800 million, 10 percent of which goes to individuals who have not attained retirement age. 2.41 As part of its voluntary separation package, the government has promised to refund individual members' contributions plus interest. Doing so has obvious welfare implications, since it disqualifies such individuals from pension benefits when they become eligible, since they cease to belong to the fund. 2.42 Growing pressure on fund resources has made it imperative for the government to clear its outstanding employer contributions arrears. Recognizing the inflationary implications of the government settling its entire K45-K50 billion arrears at once, the fund has proposed that the government issue a IO-year KIO billion bond on which interest would be paid immediately, a 3-year Treasury bill on which interest would be paid at 6-month intervals, and KI 0 billion worth of I 82-day Treasury bills and K5 billion worth of91-day Treasury bills on which interest would be paid at 6-months intervals. The government has not announced whether it accepts the fund's proposal. 2.43 The fund believes that settlement of the government's arrears would increase its liquid assets from about Kl2 billion to K60 billion. With that kind of liquidity and current interest rates of 20 percent, the fund believes it could generate sufficient interest income (in addition to its K3 billion in annual dividend income) to cover its KIO billion pensions benefits bill. 2.44 The effect on the fund of the reduction in the number of public service employees remains unclear. The funds' chief executive believes that the drop in pension  - 37- contributions as a result of a reduction in the number of employees in numbers will be more than offset by the rise in contributions associated with the decompression of the pay structure. The fund will be adversely affected if the pay structure is not decompressed. 1:45 Given current pension provisions, the government's employer contributions . arrears, increased retrenchment payments, and political interference, the fund's viability cannot be guaranteed. Although the 1996 Act requires the Board to "make adequate arrangements to avoid noncompliance" and to "avoid delays and other constraints to the effective provision of adequate benefits under the Fund," there appears to be little it can do to ensure the fund's solvency. A change in pension legislation could address the problems in the system, but such a change may not be politically feasible. 2.46 The government continues to borrow from pensions funds at low interest rates to finance its recurrent and capital expenditures (see table 2.5). Relying on loans from the Zambia National Provident Fund to finance certain aspects of local government adversely affects the operations of the pension schemes and undermines their ability to safeguard the interests of their members. The new pension legislation indicates that the government is not prepared to loosen its control over the operations of the fund-a position that is in stark contrast to the government's institutional reforms and market-driven economic policies.  - 38- Table 2.5. Internal Government Guaranteed Loans by Councils (From ZNPF) Councils Original Product Issue Date Amount Outstanding Outstanding 31112/96 31112/97 Chadlza District Council ZNPF 10/26/84 2012 200,000.00 212,180 194,417.07 Lusaka Cit) Council ZNPF 12/20/83 800,000.00 607,187 Kabompo District Council ZNPF 10/15/84 150,000.00 201,849 145,229.26 Kabompo District Council ZNPF 8/19/88 152,000.00 205,720 148,272.51 Kabwe District Council ZNPF 3/31183 881,000.00 836,956 Kalomo DIstrict Council ZNPF 6/25/86 50,000.00 16,214 Mufulira Municipal Council ZNPF 9/2/69 1,500,000 1,420,926 Livingstone Municipal ZNPF 4/8170 450,000 535,942 Council Livingstone Municipal ZNPF 6/30/81 14,400,000 14,400,000 Council Luanshya Muntcipal CouncIl ZNPF 2/23/84 230,000 97,536 Lusaka City council ZNPF 6/26/87 3,000,000 3,107,111 Lusaka City Council ZNPF 6/26/87 1,000,000 979,205 Lusaka City Council ZNPF 119/88 9,800,000 10,386,231 Solwezi District Council ZNPF 217191 1,200,000 900,000 Mwmilunga District Council ZNPF 10/15/84 100,000 133,032 94,132.00 Senanga District Council ZNPF 7/1/83 1,533,000 2,153,962 1,549,617.00 Senanga District Council ZNPF 4/6/90 722,430 1,003,009 722,430.00 Solwezi District Council ZNPF 3/15/84 200,000 1,651,148 Solwezi District Council ZNPF 8/22/86 170,000 154,441 Lusaka Urban District ZNPF 9/14/89 Jun-96 16,000,000 6,000,000 Council Livingstone Municipal ZNPF 10/26170 1994 500,000 680,415 Council Livingstone MuniCipal ZNPF 12/4/89 1996 600,000 2,150,619 Council Katete District Council ZNPF 6/11186 1996 2,894,000 20,448,877 Luanshya District Council ZNPF 11122171 1996 350,000 2,031,879 Ndola City Council ZNPF 11122171 1996 550,000 2,840,046 Kawambwa District Council ZNPF 4/11183 1996 150,000 3,096,402 Zambia Airways ZNPF 8/15/84 4.000,000 1,785,714 1,785,71400 Sub-Total 61,582,430.00 78,036,601.00 4,639,811.84 Miscellaneous National Housing Authority ZNPF 3/8/84 1994 7,500,000.00 2,740,297.00 0 Source: Mmistry of Finance and Economic Development Improving the Financial Management of Extrabudgetary Funds 2.48 The government should consider improving the financial management of extrabudgetary organizations. Rather than doing so by applying the revised financial regulations to these organizations, it should develop framework legislation and regulation for the financial management of these institutions and refer to this legislation in the laws establishing such organizations (see Chapter 4).  - 39- 2~49 The efficiency and equity of grants provided to grant-aided institutions need to be reviewed carefully in light of tight budget constraints. As a first step, the government should require all grant-aided institutions to submit monthly budget to the Budget Office of MOFED, just as line ministries do. Funding for the new budget year should be made conditional on the st:.bmission of audited final accounts, as required by law. 2.50 The local government system requires a major overhaul in terms of how it is funded, how it accounts for its performance, and how it delivers services. 2.51 The government should help restore the financial health of the special funds and pension funds by ceasing to use these funds to finance its deficit. Doing so will require measures to restore the solvency and sustainability of fiscal policy. 2.52 The government should gradually expand its reporting on all government activities, including extrabudgetary funds, public sector borrowing requirements, and contingent liabilities, such as guarantees. Reporting must occur in both the budget and in the financial reports of the government. Over time, as budgeting is decentralized, reporting on local government activity and budgets will become necessary to inform Parliament, the general public, and investors about their activities.  - 40- CHAPTER 3 ZAMBIA'S BUDGET MANAGEMENT SYSTEM 3.1 Although improvements have been introduced in recent years, Zambia's budget management system continues to suffer from overgenerous budgeting, poor estimation of revenues, weak revenue collection, inadequate cash management, poorly trained staff, and lack of accountability. Its system of cash budgeting, introduced in 1993, has hindered budget implementation, and inadequate information systems have caused the government to bounce checks and accumulate significant arrears. A. THE LEGAL BASIS OF THE BUDGET 3.2 Part X of the 1996 Constitution provides the legal basis for Zambia's government expenditure management. It requires each expenditure item to be authorized by an appropriation act or supplementary estimate issued by the President. The Constitution allows the President to issue a warrant without an appropriation act or supplementary estimate if he considers the expenditure "urgent." Such expenditures must be included in a supplementary estimate to be approved by Parliament within four months. Table 3.1: Budgeted and Actual Domestic Expenditures, 1994-97 (billions of Kwacha) 1994 1995 1996 1997 Budget Outturn Budget Outturn Budget Outturn Budget Outturn Total Expenditures 397 503 527 557 603 721 1127" 1293" Wages and salanes 77 115 156 178 216 220 259 324 Recurrent Departmental 68 63 109 101 142 120 145 137 charges Interest na 103 45 77 77 121 198 213 Memorandum Items Difference budget-outtum 106 30 118 167 Supplementary estimates 43 111 104 136 " Different defimtlon Source: Expenditure Estimates, Budget Speech 1998, IMF. The above table shows that constitutional rules are not always applied. The Auditor- General's report of 1996 (issued in February 1998) mentions several instances of unconstitutional expenditures that add up to K30 billion. 3.3 Because of these urgent expenditures and because some donor funding becomes available only during the fiscal year, supplementary estimates are a regular feature of Zambia's budget system; actual expenditures have exceeded budgeted expenditures every  - 41 - year since 1994 ( Table 3.1). The Auditor General's report of 1996 (issued in February 1998) estimates that K30 billion in unbudgeted expenditures were made in 1996. 3.4 The Finance Control and Management Act (Chapter 600, Laws of Zambia), amended in 1980 (Chapter 347), vests the management, supervision, control, and direction of all matters relating to the finances of the Republic in the Minister of Finance. The Minister designates senior officers in the ministries as controlling officers for each category of expenditures. 3.5 The Finance Act is supported by a set of financial regulations that describe the budget process in detail and that specify various agents' duties and responsibilities. 21 The regulations are statutory instruments under the Finance Act and carry the force of law. 3.6 The budget year runs from January 1 to December 31. Traditionally, the budget is not submitted to Parliament until late January, and the appropriation act is usually passed in early February. B.BUDGETCOVERAGE 3.7 The budget covers current and capital expenditures of central and provincial governments, including statutory expenditures, such as interest and amortization on debt. In principle, it includes foreign aid, which is handled by the Finance Minister. Some technical assistance and funding for local governments and NGOs remain outside the purview of MOFED, however, and is thus outside the budget process. Moreover, because the level of aid is not known at the time the budget is prepared, a significant portion of such aid is allocated through supplementary estimates. 3.8 Appropriations-in-aid are identified separately in the budget, since some institutions receiving such appropriations (such as the Zambian police) recover some of their operating costs through fees or fines. These fees and fines are included in the budget estimates, but they are listed separately because the collecting institution is allowed to retain fees up to an amount authorized in the budget. In 199x some K86 billion, or about 10 percent of government spending, was spent by these institutions. 3.9 Institutions receiving government appropriations differ from independent institutions such as statutory boards and universities, which receive government grants. These independent institutions received K150 billion in grants in 19xx. Their expenditures are not reported in the budget, even though most of them are government institutions. 21 The mission had access to the 1985 reprint of the 1976 regulations, which seems to be the most recent.  - 42- 3.10 With the exception of central government grants to local governments, local government revenues and expenditures are not included in the budget, and data on these budgets are incomplete. The Central Statistics Office collects some local government statistics, but the data are made available only after a long delay. Data obtained by the mission for 1993 and 1994, which cover about thr<;:e-fourths of all local governments, show an aggregate deficit of about K 10 billion. 3.11 No consolidated accounts for the public sector as a whole are prepared. Guarantees and other contingent claims on the government are not included in the budget, parastatals are included only on a net basis, and no public sector accounts are presented. The Central Statistics Office is currently working on assembling public sector accounts, but the effort is severely hampered by incomplete reporting by parastatals and local governments. c. BUDGET PREPARATION 3.12 Major changes in the budget preparation process have been made in the past two years (Table 3.2). Before the 1996 budget, little planning went into budget preparation and ad hoc changes were routinely introduced during budget preparation and implementation. The budget failed to reflect the government's priorities, and it regularly overcommitted resources. 3.13 In preparing the 1997 budget, the government began to adopt wide-ranging reforms. Activity-based program budgeting was introduced to meet strategic goals defined within the context of a medium-term expenditure framework. These reforms are ongomg. Table 3.2: Budget 1999 Preparation Calendar Update of General Policy Framework May-July Update Medium Term Financial Framework May-July 3.14 The budget is prepared Prepare Sector Papers May-July Issue Budget Circular July within the context of the "Insaka" diSCUSSIOns on GPF, Sector Papers government's general policy July/August Permanent Secretary Workshop August framework. This annually Submission of Activity Based Budgets October Discussion of ActiVity Based Budgets with MOFED updated framework, which is OctoberlNove mber not made public, sets out the Finalize GPFIMTFF Sector Chapters November primary government objectives Cabinet DiSCUSSIOns of Budget December Budget Day Januaryfor the following three years. It is drafted by the Budget Office and the Directorate of Macroeconomic Policy Analysis in MOFED. The general policy framework for the 1998 budget became available only in October, leaving little time for adjusting the draft budgets. For the 1999 budget, the general policy framework should be ready by July, when a series of consultation meetings (insakas) with members of Parliament is planned to build consensus on government policy.  - 43 - 3.15 The medium-term financial framework, drafted by the Directorate of Macroeconomic Policy Department and the Budget Office, sets forth the macroeconomic and fiscal framework for the coming three years. It includes projections for key economic and fiscal variables, estimates for donor funding, and expenditure limits within which the sectoral plans and ministerial budgets need to be rrepared. For the 1997 budget the medium-term financial framework did not include a planning reserve for new policy. 3.16 IMF and World Bank staffwork closely with Zambian policymakers in preparing the medium-term financial framework, which is based on assumptions similar to those underlying the Policy Framework Paper. The medium-term financial framework is not a public document, and it is not explicitly endorsed by the Cabinet or the Parliament. 3.17 MOFED desk officers for the various ministries meet with the permanent secretary of MOFED and the line ministry to discuss each ministry's budget. For each department or province, expenditures are identified in five categories: personal emoluments (salaries, wages, and other emoluments); recurrent departmental charges (allowances, purchase of goods, purchase of services, and training expenses); grants and other payments; and capital expenditures (movable assets, projects, technical assistance, long-term training, civil works, and private sector support). These economic classifications deviate somewhat from the international norm defined in the IMF's GFS Manual in that training expenses are listed separately. The budget thus contains no functional classification (in Zambia usually referred to as "sectoral classification") or classification by program. 3.18 In preparation for the 1997 budget, personnel costs were assigned to each activity and program. This proved to be unmanageable, and the 1998 and 1999 budgets include separate budgets for personnel costs. 3.19 Until this year, the investment budget was prepared separately. The PIP served mainly as a document for donors, who finance most, if not all, public investments. Beginning in 1999, the ministries are supposed to prepare their investment budgets as an integral part of their program budgets. 3.20 Until recently, sectoral plans were not broken down by ministry and agency, creating considerable confusion over who was responsible for what program and causing ministries to duplicate efforts. Since 1997 sectoral planning teams that include ministry staff prepare the sectoral chapters of the medium-term financial framework. For the 1999 budget the sectoral chapters will contain ministerial spending limits within which the activity budgets are to be prepared. 3.21 In preparing the 1999 budget, policymakers will also be able to draw on strategic planning papers submitted by the ministries. These strategic plans, which will include a mission statement, a description of recent sectoral performance, a breakdown of expenditures by major program for the coming three years, and a list of program priorities, will serve as the basis for the sectoral chapters of the medium-term financial  - 44- framework. To coordinate the plans and eliminate overlap among the ministries' work plans, permanent secretaries from the line ministries will meet to discuss the papers. 3.22 Within the context of the sectoral plans and the strategic planning papers, individual ministries prepare their budgets for the coming year. The quality ofthese plans and budgets varies widely from ministry to ministry. Some ministries, such as the Ministry of Education, describe the goal of each activity and expenditure based on the general policy framework and the sector plans. Other ministries simply request an increase over the previous year's budget. 3.23 All budget proposals request significantly more funding than is available. In 1997, for instance, only about a third of nonpersonnel ministerial requests were included in MOFED's submission to the Cabinet. For 1999, the ministries are supposed to prepare budgets that keep expenditures within the limits agreed upon by the Permanent Secretaries. 3.24 The budget submitted by MOFED is discussed by the Cabinet in December. Until 1997 the Cabinet made significant changes in the budget proposal, increasing expenditures beyond the medium-term financial framework ceilings. These increases reflected the fact that the Cabinet was not involved earlier in the budget process. The lack of a planning reserve during budget preparation meant that these last-minute changes could not be absorbed without exceeding the spending limits. 3.25 The Finance Minister submits the budget to Parliament at the beginning of January and gives the budget speech at the end of January. The budget speech discusses global and national economic developments, briefly reviews budget implementation over the past year, presents a medium-term perspective on the budget based on the general policy framework and medium-term financial framework, and presents the main objectives for the coming fiscal year. The speech presents the economic classification of expenditures and revenues on a highly aggregated basis. 3.26 Although the Constitution gives Parliament the right to change the budget, Parliament rarely exercises this privilege. Once parliamentary approval is obtained, MOFED informs the ministries of their approved budgets. The permanent secretary of MOFED authorizes the controlling officers in the line ministries to incur and commit expenditures within the amount of the annual estimates by issuing a Treasury Authority. D. BUDGET IMPLEMENTATION 3.27 Budget implementation has been weak in Zambia. Accounting personnel are scarce and inadequately educated, and program implementation is frustrated by cash rationing. Accountability is undermined by lack of enforcement of the financial regulations-in part because controlling officers are appointed for political reasons. Many accounting and financial officers in line ministries appear unfamiliar with all of the  - 45 - regulations and procedures governing budget execution, and they often lack access to the financial regulations and MOFED circulars. 3.28 Budget implementation is further hindered by limited information technology support. Many agencies still use a manual accounting system and a computerized system that is not always compatible with those of other ministries. Delays in reporting and insufficient information on commitments and bank accounts cause large arrears and repeated bouncing of government checks. Cash Budgeting 3.29 Budget implementation is based on a cash budget. Under this system no cash is released before revenues are deposited in the central Treasury account. The cash budget was adopted to stop the once common practice of requiring the Bank of Zambia to finance expenditures-a practice that contributed to the runaway inflation of the early 1990s. 3.30 MOFED derives its authority for cash budgeting from Section 19 of the financial regulations, which states that "the Permanent Secretary [of Finance] may impose a restriction on expenditures under any subhead or item appearing in the estimates." The procedures for cash budgeting are spelled out in Circular No. 111994, but line ministries are not always aware of these regulations. 3.31 Originally, a Joint Data Monitoring Committee comprising MOFED and Bank of Zambia staff was in charge of budget implementation. The committee met daily to monitor financial and monetary conditions, the government's cash position, and expected disbursements and revenues. Key departments had to report their spending and spending plans on a daily basis, and the Bank of Zambia reported daily on the Treasury position and the subaccounts. The committee reported weekly to the Finance Minister and the governor of the Bank of Zambia to bring possible issues to the attention of the political leadership. 3.32 When the cash budget was introduced in 1993, cash releases were made on a daily basis and no Bank of Zambia financing was allowed; all cash shortfalls were financed with Treasury bills issued to the private sector 22 This system was changed in 1997. Cash releases are now made monthly, and the Bank of Zambia provides for some within-month financing. If insufficient funds are available to cover a check that is considered important, the Bank of Zambia contacts MOFED, which can arrange for funds to cover the check. If provision is not made by the following morning, the check bounces. 3.33 Currently, the government has negative domestic borrowing and has sufficient liquidity over time to repay the Bank of Zambia. However, more revenues tend to come 22 The Bank of Zambia did provide some liquidity in January 1993, which was repaid before the end of the month (Bolnick 1995).  - 46- in to the Treasury toward the end of the year; some issuance of short-time Treasury bills thus takes place earlier in the year. 3.34 MOFED decides how much cash to release and for what purpose. Every month the secretary to the Treasury, the permanent secretaries, the Budget Director, and other high-level staff from the Budget Office meet to decide how much cash is available for release. 3.35 Although cash release statements specify an economic category, cash is released for particular activities. 23 For instance, MOFED determines what part of the release to the Ministry of Education should be spent on wages in Luapula province. The cash budget gives priority to statutory expenditures, such as domestic and foreign debt service. Next in line are personal emoluments. If there is cash left, grants and then recurrent departmental charges are funded; funds for investment are released last. Line ministries, the Accountant General, and the Bank of Zambia are informed of the planned cash release by desk officers in the Budget Office. 3.36 Departments can also make requests for special cash allocations (to pay for a large delivery of goods, for example). Although future allocations are reduced by the amount released for the special purpose, special allocations nevertheless provide an incentive for ministries to request more funds than they need, and they render the cash budget less transparent. 3.37 MOFED does not check whether the monthly releases are actually spent as prescribed, and it has no legal recourse if a ministry does not spend the cash according to the release, since the ministry is legally bound by the appropriations law, not by the cash release. The ministries report their overall spending to the Budget Office and their detailed spending per category to the Accounting Department. The detailed data are submitted only months aner the expenditures are made and are usually not verified. 3.38 In the 1997 budget, allocations for personal emoluments far exceeded the original expenditure estimate because of the 40 percent wage increase. As a consequence, cash releases for all other categories were reducea. In one ministry, for example, 88 percent of grants, 67 percent of recurrent departmental charges, and only 35 percent of budgeted capital expenditures were funded (figure 3.1). Worrisome as these figures are, they nevertheless represent an improvement over 1996, when only 50-60 percent of recurrent departmental charges and 25 percent of capital appropriations were released. 23 Although cash releases are supposed to be made by category, MOFED releases the funds lump sum, leaving the division of funds to the discretion of the line ministry, according to the budget director.  - 47- 3.39 Perfonnance during the first quarter Figure 3.1 a: Budget and Outcome of of 1998 shows marked improvement over Expenditures 1997, with line ministries obtaining some ,ao 90-95 percent of budgeted cash. The .. eo I.::::=::.nm._l :Otutal aprwad I improvement was due in part to the larger 0_'" than usual contingency reserve of K 112 '" 110 billion. Some K29 billion was used to 50 compensate for revenue shortfalls and expenditure overruns on certain budget items. '0 3.40 The cash budget severely undermines the budget process by making cash releases ROC. - Figure 3.1 b: Share of Expenditure per Coplal unpredictable and undennining fiscal Subhead, Budget and Outcome discipline. Ministries and departments are ,""" I----,~--r---T""--,--I unsure about the release of funds, and they "'" usually receive significantly less than "'" allocated in the expenditure estimates. Plans "'" to move to a three-month horizon for cash "'" ~ releases have not yet been implemented. Although inclusion of a large contingency reserve in the 1998 budget has improved the ..... - !OGnlnts I ~1 I IORDCa situation, line ministries still maneuver to ,... ... avoid the negative impact of cash budgeting by budgeting for lower priority expenditures - ....... - Note Total approved equals Original plus supplementary and then requesting supplementary estimates estimates Source Data provided by a Ime mmlstry for high-priority spending; paying for lower- priority spending with regularly allocated cash and then making special requests for high- priority payments; cutting disbursements to highly vocal institutions, such as the university; overcommitting; and going into arrears-sometimes at the cost of heavy penalties-because expenditure control is exercised at the cash level and not at the commitment level. 3.41 Improving the underlying factors necessitating cash budgeting-overgenerous budgeting, poor estimation of revenues, weak revenue collection, inadequate cash management-would obviate the need to use the system and help to improve fiscal management in Zambia. Procurement, Payment, and Banking Arrangements 3.42 Under IDA Second Economic and Social Adjustment Credit (ESACII), the Government initiated a revision of the public procurement system in 1996 to improve its timeliness, transparency, and accountability. A Procurement Action Plan establishing a schedule for implementing specific measures to improve the quality and speed of public procurement was developed with SIDA and World Bank support. The plan called for  - 48- decentralizing the implementation of public procurement to individual ministries and local governments, training and institutional strengthening of Zambia National Tender Board (ZNTB) and the Ministerial Procurement Units (MPUs), and providing stronger and more effective oversight of major procurement decisions. 3.43 Implementation of the program is progressing well. Procurement units have already been established in six Ministries and several other agencies, standardized bidding documents have been prepared, and procedures have been simplified. The National Tender Board is being transformed into an oversight agency. The ZNTB will review awards on contracts above certain limits. It will also review procurement policies and procedures and propose procurement policy changes as appropriate. A World Bank Country Procurement Assessment Review issued in June 1997 found that implementation progressed well and that, when the reform is completed, Zambia will have a sound public procurement system. 3.44 The procurement system is separate from the accounting system. As a result, awarded tenders are not automatically entered as commitments in the accounting system. This leads to delays in payments and ultimately to arrears and delays in program implementation. It has also caused suppliers to lose confidence in the government as a reliable client and to charge higher prices to compensate for the anticipated delays in payments. 3.45 Zambia has a decentralized payment system, with each department responsible for its own payments. Departments are issued checkbooks by MOFED. When making a payment, one accounting officer prepares a payment voucher, one confirms the voucher, and one approves the payment. The checks include the appropriation and warrant number for the payment. Checks are written whether or not the goods have been delivered, an invoice has been received, or the proper tendering procedures have been followed. In some cases, the internal auditor verifies whether all procedures have been followed and payment is authorized. The internal auditor cannot block a payment, however. After the payment voucher is approved, two accounting officers must sign the check. 3.46 Payments on current expenditures made in the past fiscal year can be made until six months into the new fiscal year. The Bank of Zambia maintains ex-accounts to which remaining funds at the end of a fiscal year are transferred. 3.47 Government bank accounts are controlled by the Accountant General ofMOFED. The government has some 204 bank accounts, 99 of which are at the Bank of Zambia. MOFED believes it has sufficient control over these accounts and that it is aware of all government bank accounts in the system. However, an IMF-recommended inventory has not yet been completed. 3.48 Except for payments for personal emoluments, line ministries make payments from their subaccounts at the Bank of Zambia. Funds are disbursed to ministerial subaccounts ("control accounts" 1-99) from the general revenue account (account 99). Ministries hold separate accounts for recurrent and capital expenditures. In addition,  - 49- subaccounts for donor projects are kept at the Bank of Zambia. Personal emoluments are paid from a central account controlled by MOFED. 3.49 Ministries hold a variety of accounts at commercial banks. These accounts are used to facilitate payments in remote areas. They are also used for revenues line ministries consider nongovernment revenues. This practice seems to rest on the mistaken assumption that extrabudgetary funds are not public funds. These commercial bank accounts can contain substantial amounts of funds, forcing MOFED to cut cash allocations, issue Treasury bonds, or both. Donors contribute to the inefficient management of cash when they require deposit of aid in a commercial bank account, since this cash often remains unused during the year because of lack of counterpart funds. 3.50 The GRZ is in the process of decentralizing payments for personal emoluments. Already five departments make payments for their personnel from their own accounts in commercial banks. The GRZ plans to have all ministries and departments using such a system before the end of July 1998. The new procedures aim to increase line ministries' responsibilities for their payrolls and provide incentives to save on personnel by removing ghost workers and no-shows from the payroll. However, decentralizing in an environment in which financial management is weak and accountability low is risky, and decentralizing the payroll could increase cash management problems. Cash Management 3.51 Except for its cash budgeting system, the government engages in little cash management. MOFED has recently set up a cash management unit in the Budget Office, but this unit does not yet play an active role. Line ministries do not seem to manage their accounts actively, and they have no cash management staff. 2s Cash management is hampered by poor reponing on commitments and lack of information on the cash position in the line ministries' subaccounts at the central bank. MOFED receives a report on all government accounts every two weeks, but line ministries are not automatically informed of their cash position. MOFED can forfeit money from line ministries' accounts, although this has not happened since 1992. 3.52 The lack of cash management causes repeated payment problems, and government checks bounce regularly. On average, 100 of3,500 checks bounce every day; another 150 checks a day are improperly prepared (wrong dates, lack of signature). Bounced checks are returned to MOFED rather than the issuing line ministry. As a result, the line ministry may remain unaware of the lack of funds and continue to write uncovered 2S The government has recently decided to set up networks of staff engaged in similar functions across government departments. Networks for budget preparation, accounting, and cash management are planned.  - 50- checks. Suppliers whose checks bounce have a hard time proving that they have not been paid, as ministries are often unaware that their checks did not clear. 3.53 Poor cash management also increases the government's borrowing needs. Zambia's revenues show significant month-to-month variation, with revenues increasing. at the end of the year. Better cash management would result in lower borrowing and lower interest payments. Reallocation of Spending within Expenditure Categories 3.54 Ministries and departments may transfer savings within categories, provided that expenditures for personal emoluments and capital expenditures are not increased. According to the financial regulations, transfers within categories require the approval of the permanent secretary of MOFED. In practice, however, such reallocation can be done by a controlling officer. 3.55 The transfer of funds for personal emoluments or capital expenditures requires parliamentary approval by means of a supplementary estimate. Pay can be increased without parliamentary approval, however, by increasing expenditures on recurrent departmental charges, which include some personal emoluments (such as various daily allowances). Supplementary Estimates 3.56 Supplementary estimates are a regular feature of Zambia's budget process. Although declining inflation in recent years has reduced the level of supplements, about 10 percent of expenditures were still made under supplementary estimates in 1997 (down from almost 20 percent in 1995). 26 Supplementary estimates are often used to confirm expenditures already made under Presidential warrants issued according to Article 115 (2) of the Constitution; that is, Parliament approves expenditures after the fact. Supplementary estimates are not subject to a rigorous preparation process, and they are prepared on an ad hoc basis very late in the year rather than as part of a midterm budget review. In principle, supplementary expenditures should be met from savings on other items within the same category. Accounting and Reporting 3.57 The 1992 Manual of Accounts serves as a guideline for preparation of financial statements. The Accountant General in MOFED is responsible for preparing the government's accounts, which are prepared annually, long after the close of the fiscal year. The 1996 accounts, for example, were finalized and audited only in May 1998, and the Parliamentary Accounts Committee was only then meeting to discuss the audited accounts of 1995. MOFED's Chief Accountant believes that the statutory limit of nine months will be met for the 1997 budget. 26 See IMF report, ibid.  - 51 - 3.58 Line ministries report their expenditures to MOFED monthly, but the reports are not reviewed or processed until after the end of the fiscal year. Until recently, the reports were of poor quality and were submitted several months late, particularly in ministries with large provincial operations, such as Agriculture and Education. The timeliness of reporting has improved, however, and the chief accountant estimates that some 90 percent of the reports are now accurate and submitted on time. The key reason for the improvement is that cash release is now tied to the delivery of the expenditure report, which is prepared on diskette and sent to MOFED, where the data are loaded onto the mainframe computer financial management system. 3.59 In addition to the full report on expenditures to MOFED's accounting unit, line ministries report their aggregate spending, commitments, and arrears to the Budget Office. This practice was announced in Circular No. 111994 and is just now being implemented. Commitment reporting was made obligatory under MOFED Circular No.1I1998, which states that no funds will be released until the commitment report is delivered. 3.60 Despite these changes, however, difficulties with commitment reporting persist. Line ministries are not always clear on what constitutes a commitment, and data on arrears remain incomplete, despite repeated efforts to take stock of arrears. In November 1997 the level of arrears was estimated at K37 billion (excluding significant arrears to extrabudgetary institutions [see chapter 2]), and funds were budgeted to clear those arrears. This estimate is probably low, however, since line ministries are not required to keep commitment ledgers. 3.61 The high level of arrears partly reflects MOFED's tougher stance on unbudgeted expenditures. Unauthorized spending was once routinely legitimized after the fact by supplementary estimates. Recently MOFED has been much less willing to grant supplements. Since line ministries have not yet changed their behavior, unauthorized spending shows up in arrears. The key problem-unauthorized spending-remains unsanctioned. 3.62 The Bank of Zambia provides MOFED with daily reports on revenues received; it reports on the balance of each control account weekly. Within five days after the end of the month, the Bank of Zambia reports on all transactions on all accounts. MOFED wants to move to a daily reporting on the control accounts in 1998. 3.63 Reporting on the implementation of the budget is minimal. The Finance Minister's budget speech contains only one paragraph on implementation of the previous year's budget, and ministries are not required to submit reports on their programs or policies. Expenditure reports are not reconciled with bank account statements. 3.64 Despite recent improvements in reporting, the accounting system remains weak. The Auditor General's report for 1996 identified numerous deficiencies, including weaknesses in the control system, failure to reconcile the final accounts with the Bank of Zambia statements, numerous unconstitutional expenditures, unauthorized expenditures,  - 52- unvouched and inadequately vouched expenditures, irregular accounting, duplicate payments, questionable payments, and nondelivery of paid goods. Weak. accounting results in considerable loss of funds and reduces the effectiveness of spending. 3.65 Human resources remain a major constraint for accounting. Few of the 875 government accounting personnel are fully qualified, and job performance is poor. Since adoption of the Public Sector Reform Program, productivity has declined, and internal audit has observed an increase in fraud by people expecting to lose their jobs. 3.66 A further constraint on accurate accounting and reporting is the lack of a government financial management information system. Although a variety of individual systems (for budget preparation, procurement, accounting, expenditure, commitment, and arrears reporting) are in place in the ministries, the systems are not integrated. As a result, expenditures and commitments are inadequately controlled, and overspending and overcommitment cannot be automatically prevented. Auditing 3.67 Auditors from MOFED's internal audit unit work within the main spending units and report regularly to the ChiefInternal Auditor. Internal auditors send monthly reports to the Chief Internal Auditor, with copies sent to the appropriate controlling officer. The Chief Auditor sends a summary report to the permanent secretary of MOFED. 3.68 As prescribed by the 1980 Public Audit Act, the Auditor General's Office performs external audits on the basis of the government's final accounts. Lack of funding and personnel have meant that only partial audits have been conducted. The Auditor General produces separate reports for government expenditures and parastatal bodies. The reports are submitted to Parliament, where they are discussed in the Public Accounts Commission. The audit reports are usually late because the government's final accounts are submitted very late. The audit report for 1995 expenditures, for example, was delivered in February 1998 and discussed by the Public Accounts Committee in May 1998. 3.69 The Auditor General also has a mandate to conduct value-for-money audits. These have been very limited, however, because of lack of personnel and lack of information on what government spending seeks to achieve. 3.70 The effectiveness of auditing is limited by the lack of well qualified personnel and lack of training of the government's 170 auditors, only one of these is a certified accountant. Accountability 3.71 MOFED is responsible for managing the preparation and implementation of the budget (table 3.3). The controlling officer is the key person responsible for financial management within line ministries and departments. In principle, the officer, usually the  - 53 - permanent secretary, is personally responsible for any payments made without proper authority (Financial Regulation No. 27); in practice, controlling officers are rarely fired or forced to reimburse ~e government for improper payments, and no criminal cases have been brought to court since 1993. Instead, the government often legitimizes improper payments after th~ fact by authorizing supplementary spending. Table 3.3: Fiscal Management in Zambia Functions Responsibility Macroeconomic and Fiscal MOFED Budget Office and Directorate of Policy Formulation Macroeconomic Analysis; Bank of Zambia Budgetary Policy MOFED Budget Office Social Security Policy Ministry of Labor and Social Security Tax policy MOFED Budget Office and Directorate of Macroeconomic Analysis Tax Administration Zambia Revenue Authority Domestic Debt Policy Budget Office, External Resource Mobilization Department Foreign Debt policy External Resource Mobilization Department Aid Coordination External Resource Mobilization Revenue Forecasts Budget Office/Zambia Revenue Authority Expenditure Forecasts Budget Office; line ministries Treasury Management Expenditure Budget Office Authorization Financial Planning Budget Office; BOZ Cash Management Budget Office, Internal audit Internal Audit Department Domestic Debt Budget OfficelExternal Resource Mobilization Management Foreign Debt External Resource Mobilization Government Accountant General (in MOFED); controlJing Accounting officers; BOZ Control over Accountant General (in MOFED) Government Bank Accounts Asset Management Source: MOFED 3.72 In the case of suspected improprieties, the Auditor General reports to the Public Accounts Committee, which decides whether action should be taken. Since the audit report on the final accounts is submitted more than two years after the close of the fiscal year, however, the possibility that action may be taken does not constitute a serious threat. The Finance Minister does not seem to use his right to appoint, suspend, or dismiss controlling officers from their duties or to turn over apparent criminal cases to the police.  - 54- 3.73 Some confusion appears to exist over responsibility for accounting officers. According to the financial regulations, accounting units in line ministries are responsible to the controlling officers. However, the Accountant General's office seems to be under the impression that it is responsible for accounting officers. Except for those internal auditors detached frvm MOFED, internal auditors are responsible to the line ministry in which they are stationed. 3.74 Currently, a revised draft of the financial regulations is being circulated. This draft, which was not made available to the mission, apparently contains specific sanctions for a variety of financial improprieties. The revised regulations have been reviewed by the Cabinet on various occasions, but no decision about their introduction has been made.  - 55- CHAPTER 4 IMPROVING ZAMBIA'S BUDGET MANAGE~ENT 4.1 Zambia could achieve better results from its public expenditures by improving its budget management system. The need to do so is imperative given the country's tight resource constraints, which are likely to dictate cuts in total expenditures in the coming years. 4.2 Zambia's budget process has undergone significant reforms over the past several years. Both budget preparation and budget implementation have improved, albeit from a very low base. Budget reforms have been strongly endorsed and supported by the donor community, and both the European Union and the IMF have budget advisors in Zambia to help the government strengthen its budget preparation and budget implementation. 4.3 Despite progress in reforms, however, much remains to be done. The budget still does not provide an effective constraint on the level of government spending, and it fails to determine the allocation of government resources. Cash rationing during budget implementation rather than the budget or the supplementary estimates determines government spending. 4.4 While the cash budget has effectively reduced inflationary financing of fiscal spending, it has come at a cost. As chapter 3 has shown, the government has no control over commitments, cash management is weak, reporting is slow and inaccurate, and accountability is lacking (table 4.1).  - 56- Table 4.1 Main Deficiencies in the Government's Financial Management System Function Current practice Result Expenditure • Cash rationmg • ElimmatlOn of cash deficit but no control control over commitments, mcrease in arrears, ad hoc allocations • Unclear cash budgeting procedures • Diluted accountability and uncertainty in program implementation Commitment • No commitment control, initial • Continued overcommitment, arrears, control commitment monitoring has penalties for late payment, and higher prices started, but concepts are unclear for goods purchased by government and not hnked to procurement system Cash • No cash management function in • Bounced checks, delays in payments, late management MOFED or mmistnes payment fines, uncertainty 10 program implementation • No reconCIliation expenditures and bank account statement • Weak accountability and expenditure control • Multitude of control accounts; • Idle balances, higher interest, and debt unregistered bank accounts service costs Reporting • Delays in reportmg; maccuracles • Inaccurate basis for budget preparation and in reports implementation Accounting • Inaccurate accounts; long delays • Inappropriate information in budget in final accounts preparatIOn, undermming of accountabihty Accountability • Weak enforcement of financial • Weak financial management framework regulations Source Zambian authorities and Bank staff assessment. A. IMPROVING CASH BUDGETING 4.5 The numerous problems associated with the cash budget suggest that in the long run the system should be abandoned. In the short and medium term the government can take numerous actions to improve budget implementation. Enforcing Financial Regulations 4.6 The Finance (Control and Management) Act and the financial regulations provide a solid framework for government financial management and clearly define responsibilities for financial management. The regulations describe procedures to ensure proper management control and vest responsibility for financial management in a department's controlling officer, who is personally liable for spending made without proper authorization. 4.7 Weak enforcement of financial regulations lies at the core of Zambia's budget implementation problems. Despite repeated violations of the regulations, no controlling  - 57- officer has been fired or suspended in the past five years. Controlling officers have no incentive to comply with the financial regulations, let alone with MOFED's administrative guidelines, which do not have the force of law. The Government should consider revising the regulations to include enforceable sanctions against offenses. Such sanctions should be predominantly administrative t(l ensure swift implementation. Improving Accounting Capacity 4.8 Lack of accounting capacity is a serious constraint to better financial management in Zambia. Few of the government's 875 accounting personnel are adequately trained, and the government seems unable to attract competent accountants at current wages, which are much lower than in the private sector. Uncertainty caused by the Public Sector Reform Program has adversely affected performance by all government workers in Zambia, including accounting staff. The government should give the highest priority to stepping up ongoing training of accounting staff, and it should improve compensation in order to retain trained staff. However, the government should realize that turnover is likely to remain high, because well-trained accountants remain scarce in Zambia. One way to accommodate this is to keep the complexity of the accounting system within reason and to seek computer-supported solutions that can be easily learned by new staff. 4.9 If the government fails to improve its accounting capacity, it could consider contracting out financial management, either to the private sector or to an independent organization similar to the Zambia Revenue Authority. In some countries the Treasury is kept independent of the finance ministry so that the finance ministry can focus on fiscal policy and budget preparation. Although running the accounting function of a Treasury outside the civil service is rare, it could provide an opportunity for Zambia to improve financial management, attract competent staff, and remove political interference from day-to-day financial management. Eliminating Uncertainty over the Legal Status of the Cash Budget 4.10 Currently, the cash budget is operated under administrative regulation of MOFED. Although these regulations seem to be binding under Zambian law, some line ministries argue that the cash budget is only an administrative measure, which is superseded by the appropriations law. As a consequence, they commit resources and accrue arrears when cash allocations fall short. Integrating the cash budget procedures in the financial regulations or the Finance Act would reinforce the binding nature of the cash budget. Even if the goal is eventually to phase out cash budgeting altogether, proper regulation of the procedures is desirable, if only to maintain the possibility of imposing the procedure again in times of emergencies. If this proves impossible, the government could consider limiting the release of warrants to the amount of cash available. To do so would require quarterly or even monthly release of warrants.  - 58- Publishing Data on Cash Releases 4.11 The procedures used to make cash releases are not clear to parliamentarians, the general public, and even line ministries, especially at lower levels in the administration. This leads to frustration and disgruntlement with government performance. It also leaves MOFED largely unaccountable for how it releases cash and undermines parliamentary control. The Government could consider publishing the cash releases for the previous year as an annex in the Yellow Book. Cash releases provide a fairly clear idea on how the government has spent its money and reveals to what extent scarcity of funds or mismanagement underlies the budgetary outcomes Adopting a System of Lump-Sum Cash Releases 4.12 Currently, MOFED releases cash for very specific purposes. This means, for instance, that it determines how much cash it releases for capital expenditures on education in the Western Province. This kind of detailed prescription of spending patterns by MOFED seems inefficient, since line ministries are probably in a better position to make such decisions. Detailed cash releases are not monitored, because detailed spending statements come in after the next cash release and are checked only at the end of the year. If the government believes that certain spending needs to be protected, it could do so by general rule. The government should, therefore, move to a system of lump-sum release for each ministry. Over time, when the program budget is further developed, departments should be granted more autonomy over how they allocate their cash. At that point, the rules for reallocation of funds may need adjustment. Linking the Procurement System with Commitment and Expenditure Reporting 4.13 MOFED has recently ordered line ministries to report on their commitments and arrears, and it has ruled that no line ministry may commit more than 20 percent of its budget. Much confusion remains over what constitutes a commitment, however. MOFED should intensify its training efforts and actively manage commitments. As a first step, the procurement system should be linked to the commitment and expenditure reporting system, a move that would require only minor adjustments to the recently developed software. Under a linked system future commitments would be entered into the system as soon as the procurement process began; they would be transformed into actual commitments after the contract was awarded. MOFED could also automatically cut off commitments once the 20 percent threshold is reached. To create an effective commitment management system, the government should change the financial regulations to require departments to maintain such systems.  - 59- Improving Cash Management 4.14 Much of the detrimental effects of the cash budget come frot:J1 the unpredictability of cash releases. Extending the planning horizon of line ministries to two or three months could bring significant gains. 4.15 Three-month forward projections of cash disbursements should becolTIre feasible shortly. Statistics on previous years' monthly revenues (closely monitored since the start ofthe cash budget) are available for several years and enable the seasonal patterns of cash inflow to be determined. Combined with data on future debt service from the External Resource Mobilization Department, these data could be used to make simple cash release projections for the next three months. Although these projections could not form the basis for a MOFED commitment to disburse, they could improve operational planning in the line ministries. Once commitment monitoring and reporting have improved, cash flow projections can be used to better plan the timing and the amount of Treasury bill issuance, thus reducing the Government's interest costs. 4.16 Debt service can be further reduced by making better use of idle cash balances in the control accounts. The government's many subaccounts sometimes hold abundant cash while the Treasury account is empty, forcing the government to issue Treasury bills needlessly. There seems to be little reason for line ministries to maintain subaccounts at the Bank of Zambia. Instead, the government could run a consolidated account with a ledger to keep track of cash allocated to individual line ministries. The Bank of Zambia should be informed about cash allocation in order not to disburse more cash than allocated to an individual line ministry. 4.17 Further savings on interest could be generated if the government had a better handle on its accounts outside the Bank of Zambia. Despite repeated attempts to obtain a comprehensive inventory, some bank accounts remain outside MOFED's supervision. Reporting on balances of known accounts is irregular, and MOFED seems unable to access funds from these accounts for cash management purposes. The current plan to decentralize the payroll to ministerial bank accounts outside the central bank may actually worsen the problem of idle balances. The government should complete the inventory of government bank accounts immediately. Commercial banks that handle government cash should be required to report daily transactions and balances to MOFED and the owner of the account. 4.18 A major benefit of decentralized payments would be a better separation of fiscal and monetary policy. Under a decentralized system, the central bank would no longer be forced to keep unpaid checks overnight and would no longer provide a liquidity injection through the clearing system. Such a system would shift responsibility for expenditure control to the line ministries. If the decentralization ofpayroll accounts proves successful, the government should consider decentralizing all retail payments. 4.19 The Bank of Zambia is eager to relinquish the retail banking function it currently performs for the government. However, only banks meeting all prudential regulations  - 60- should be considered for handling government funds. Service agreements with commercial banks must be carefully drafted and financial regulations should allow MOFED to access the balances for cash management purposes only. 4.20 Reducing the number of bounced checks and the associated delays in payment would save penalties for late payment and bank processing fees. It would also make the government a more reliable client and lower the prices for goods delivered. The Bank of Zambia should, therefore, report daily on the transactions and balances in the control accounts Reports should go both to MOFED and to the line ministry that owns the account. Copies of bounced checks should be sent both to MOFED and to the originating line ministry. A Centralized or Decentralized Treasury? 4.21 Zambia's current Treasury management system is not working well and should be replaced by a system that facilitates better budget implementation. Expenditure control is centralized by means of the cash budget, but commitments, payments and accounting are decentralized to line ministries. There is insufficient reporting to MOFED, hindering it from enforcing spending limits and tracking actual spending. Banking is centralized in the central bank, but cash management is largely decentralized to line ministries, which receive insufficient information to effectively carry out this responsibility. Accounting officers are in charge of checking the controlling officers' behavior, but the controlling officers determine their career prospects. The change to program budgeting increasingly keeps line ministries responsible for program implementation and their results, but MOFED continues to prescribe what ministries should spend released cash on, thereby frustrating initiative and cost saving in implementation. 4.22 The medium-term challenge for Zambia is to move toward a system of Treasury management that matches responsibility with authority and the capability to perform the function assigned. Examples of good Treasury management can be found in both highly decentralized systems (as in New Zealand) and in highly centralized systems (as in France). Even mixes of both (as in Australia) can operate well as long as clarity over functions prevails. Annexes 2 provides details on these systems. 4.23 Once a Treasury system is selected, a good government Financial Management Information System (GFMIS) can support Zambia's government financial management. Designing, developing, and implementing a full-blown GFMIS takes considerable time and effort; managing the process requires skilled and dedicated staff. Developing a GFMIS also requires considerable expertise from outside the government. A GFMIS is thus not a short-term solution for Zambia's woes, and the country will have to rely on the various stand-alone systems in the ministries and the central mainframe system on MOFED for some time. However, the government should start studying a GFMIS, with the goal of implementing such a system over the longer Tun.  - 61 - B. MOVING TOWARD A MEDIUM-TERM EXPENDITURE FRAMEWORK 4.24 Zambia is gradually moving toward a program budget within a multiyear expenditure framework. This move promises to improve expenditure allocation while maintaining tight fiscal control. 4.25 Zambia's budget preparation is starting to take the shape of a medium-term expenditure framework, a framework that has been implemented in a variety of other countries. Cross-country experience shows that a well-functioning medium-term expenditure framework should link the government's priorities with the budget within a sustainable spending envelope, highlight the tradeoffs between competing objectives and link budgets with the policy choices made, and improve outcomes by increasing transparency, accountability, and the predictability of funding. 4.26 Zambia faces numerous challenges in implementing a medium-term expenditure framework. The most important challenges are summarized in table 4.2.  - 62- Table 4.2: Current Practices in Zambia and Practices under a Medium-Term Expenditure Framework Practice under an MTEF Current Practice Goal Purpose Processes Define aggregate Provide a realistic Macroeconomic Limited macroeconomic analysis, optimistic resources estimate of total analysIs, revenue revenue forecasts: lack of planning, contingency resources available forecasting, and reserves; volatility of donor funds and copper to allocate to definition of revenues; no Cabinet endorsement of spending spending programs sustainable fiscal limit in the medium term policy Set sectoral Commumcate to DecislOnmakers No sectoral! mimstenallimlts early In the budget allocations ministries a sectoral allocate aggregate process; no Cabinet endorsement of spending expenditure policy resources to limits constrained by sectors based on aggregate resources relevant data Formulate and Show sectoral Spending Widely varying planning and budgeting capacity In determine costs of objectives, ministries line ministries; activities increasingly represent sectoral spending programs, and formulate sectoral basis for budget; weak link between goals and plans actiVIties and their expenditure actiVities costs programs Reconcile available Reach agreement on PolitiCians and Elaborate negotiation process With MOFED. resources with medium-term other sectoral spending expenditure decislOnmakers plans programs reconcile top-down constraints With bottom-up spending demands Announce sectoral Ensure that budgets Formulation of Done late In the budget process; constraints are soft expenditure limits prepared by annual budget because of supplements for year one of the mimstries reflect MTEF agreed sectoral expenditure programs Ensure that budget Prevent excessive Accounting, Weak expenditure control, focusing on cash execution is in line deviation from the reporting, and distribution; weak cash management; little with budget annual budget and expenditure autonomy of line ministries; little incentives for intentions MTEF controls are used savings; ad hoc supplementary estimates; large during execution delays in reporting of annual budget Ensure that desired Align Civil servants' Incentives for civil Lagging Civil service reforms, focused on staff results are achieved and politicians' servants to cuts; weak financial accountability; limited and incentives with perform; expost delayed audit, little evaluation capacity public goals audit and evaluation Source: Adapted for Peter Dean, " The Relevance ofa MedIUm-Term Expenditure Framework to the Needs ofa Developing Country." World Bank. Government of Zambia.  - 63 - Strengthening Political Support 4.27 A key challenge for Zambia's budgetary reforms is political commitment. Although the reforms have been ongoing for some time, political support for the revised process remains lackluster at best. Few, if any, policy speeches mention budgetary reforms; the 1998 budget speech failed to cite the new budget methods or the government's goals in adopting them. No presentation to the Cabinet has been made on the ongoing reforms, and the Cabinet has not signed off on the evolving budget process. Line ministries remain generally unaware of the ultimate goals of ongoing reforms. 4.28 One reason for the lack of political support is that budgeting is still seen as a technical exercise driven mainly by MOFED. Another reason is that politicians seem unaware of the benefits of the new process and still rely to a large extent on the old incremental methods to obtain funding. Incentives for good budget preparation remain limited because implementation is still driven by cash rationing. 4.29 Experience in other countries has shown that political support is key to successful budget reform. 27 The government, especially the Minister of Finance, needs to build this commitment by: • Clearly expressing a vision for Zambia's budget management; disseminating that vision through budget circulars, budget management handbooks, and training seminars; and seeking Cabinet endorsement of the vision • Developing an explicit implementation strategy for budgetary reforms and creating a budget reform management structure, including a high-level steering committee and a working level implementation team • Enforcing the budget rules for each ministry and agency • Ensuring that realistic budgets are implemented rather than derailed by cash rationing. 4.30 In implementing the reforms, closer cooperation among those involved in budget preparation reform and those involved in financial management is desirable. Although the two teams work together infonnally, much could be gained if budget management issues were address by a team consisting of experts from both areas. Improving Fiscal Discipline 4.31 Fiscal discipline is undermined by ad hoc decisions made late in the budget process and by overoptimistic revenue projections. As a result, spending continues to be determined by cash budgeting in implementation, and decisions on spending and budget 27 The best examples are New Zealand and the United Kingdom. Political buy-in was also identified as a key detenninant for the successful introduction of a MTEF in South Africa (see Owen Barder, "A Medium-Tenn Expenditure Framework: Lessons from South Africa." Presentation at the World Bank, June 1998).  - 64- outcomes have little to do with the budget approved by Parliament. Fiscal sustainability in Zambia could be improved in several ways: • Obtain Cabinet agreement on overall spending limits and sectoral and ministerial limits early on in the budget cycle. Currently, these limits are established as an outcome of the budget process rather than as its starting point. • Increase the planning reserve. Last-minute changes by the Cabinet often exceed fiscally prudent spending limits. The planning reserve for the coming budget year and for subsequent years is currently no more than 2.5 percent of planned aggregate spending. Increasing the planning reserve could help accommodate last-minute changes and leave room for future policy changes. • Use the contingency reserve only for truly unforeseen spending. Zambia has wisely decided to increase its contingency reserve to almost 10 percent of the budget. Almost half of the reserve has been allocated for reducing arrears, however. Moreover, budgets are often not realistic, and the rule requiring departments to offset costs overruns with savings is not enforced. The contingency reserve is thus rapidly depleted by spending on foreseeable expenditures. All of the K29 billion reserved for the first quarter was already spent in mid-March-despite the fact that regular appropriations had not been fully disbursed; by the end of May, the K 115 billion appropriated for 1998 seemed already allocated for the rest of the year. 28 The contingency reserve thus seems to work as a second budget rather than as a safety cushion. More realistic budgeting and strict enforcement of rules governing savings and reallocation of funding should be established and enforced so that the contingency fund is reserved for true contingencies. • Improve macroeconomic and revenue forecasting and obtain Cabinet agreement on the common assumptionsfor the budget. Because Zambia's economy and its revenue base is highly dependent on copper and aid, revenues are volatile. Realistic budgeting could nevertheless take place. The government has started to improve its macroeconomic forecasting with the support of technical assistance. This effort should lead to better revenue forecasts. However, experience in other countries shows that forecasters within the government are often under pressure to be overoptimistic. Increasing Planning and Budgeting Capacity 4.32 Planning and budgeting capacity in line ministries remains limited. Realistic budgeting and costing of programs is further hampered by weak accounting and long delays in reporting and by forcing budgets to be prepared based on dated and unreliable data. Although technical assistance can help, the strongest incentive to improve planning and budget capacity is to make it matter; strict enforcement of delivery of realistic three- 28 The government seemed to have no more cash for a contingency that would require extra cash in the second and third quarter.  - 65- year plans and rewarding of good planning and budgeting with higher allocation is the best incentive for line ministries to create capacity. 4.33 MOFED and some line ministries, such as the Ministry of Education, have strong budget preparation capacity. Their knowledge could be disseminated throughout the government. The recent initiative to set up networks among government officials therefore deserves strong support. Improving the Allocation Process 4.34 Allocation of budgetary resources continues to be undermined by ad hoc decisions in preparation of the budget and supplementary estimates and by faulty implementation of the cash budget. Allocation could be improved in several ways: • Set sectoral ceilings earlier in the budget process, break down sectoral ceilings to ministries, and gain Cabinet endorsement of the sectoral and ministerial allocations. • Require output and outcome indicators for ongoing and new policy (box 4.1). • ReqUire three-year cost estimates of ongoing and new policies. • Leave subsectoral allocation decisions to line ministries or departments. MOFED should resist engaging in detailed budget negotiations with line ministries. Instead it should focus on assumptions and methodology for cost calculations, checking whether a line ministry can realistically remain within its budgetary calculation. Line ministries should be put firmly in charge of subsectoral policy decisions and allocation, both to increase ownership and to leave policy choices to those in the best position to make them. • Require ministries to identify cost savings. Currently, ongoing policy is rarely challenged, and the budget process effectively focuses on the distribution of additional resources. Requiring ministries to identify cost savings and policies they want to phase out would increase budgetary headroom. Zambia could consider adopting a mechanism for reviewing ongoing spending similar to that used in the Netherlands in the early 1980s (box 4.2) and making spending reviews a regular feature of the budget process • Accelerate the production offinal accounts. Although by law, the final accounts in Zambia are to be published within nine months of the close of the fiscal year, the current delay is two years, making it all but impossible to base the budget on reliable information. Even a nine-month delay may make it difficult for line ministries to prepare their budgets on the basis of realistic costs. Submission of preliminary figures at the beginning of budget preparation should be required. • Provide incentives for gathering policy data. Many of the data necessary for good budgeting and policymaking are lacking in Zambia. Rather than waiting for the Statistical Office to produce them, line ministries should be encouraged to gather and report policy relevant information themselves.  - 66- • Introduce a midterm budget review. To avoid granting ad hoc supplementary estimates throughout the year, the government could consider holding a midterm review, during which all budget proposals are considered concurrently. Such a review would force proposals for extra spending to compete with each other. Additional spending outside the review would have to be met by savings within budgets for particular economic categories, as prescribed by the financial regulations. Such a midterm (or even quarterly) review could even become a substitute for the cash budget. Box 4.1 Outcomes, Outputs, and Inputs Perfonnance can be measured in various ways For policies-and therefore for policymaking budget units-outcomes are the most relevant indicators; for service delivery units, outputs are usually more appropriate. Outcomes indicate the ultimate goals of government strategies and policies. For example, education policies could aim to increase per capita productivity or income; health policies could aim to increase life expectancy. But productivity and life expectancy depend on much more than education and health policies alone. Policy research can try to isolate the effect of policies on these outcomes, but it cannot eliminate uncertainty over the policy outcomes themselves. Outputs specify the operational results of departments or other agents of government. For example, education could target nine-year compulsory education; health policies could aim for full immunization of 5-year-olds. Output indicators are more specific than outcomes and often require quality standards to link them with outcomes. A quality standard for education could be the literacy rate of 12-year-olds. Significant policy outcomes may be the product of several outputs together with direct budgetary payments and other policy measures, such as regulation. Input indicators show the resources used to achieve outputs and outcomes. Traditionally, the government has tracked only registered cash outlays, but increasingly full costs are measured by including accrual elements in the government's accounting system. Other input indicators relate to actions taken to achieve certain goals, such as the purchase of textbooks, the construction of hospitals, and the hiring of teachers and doctors. Box 4.2 The Netherlands Reconsideration Procedure The Reconsideration Procedure has been the main instrument of policy review in the Netherlands since 1981. The procedure reviews existing policies in the public sector with the aim developing more cost-effective alternatives or abolishing policies. Policy areas for reconsideration are determined by the Cabinet. The departments responsible for the areas of review come up with savings options, including one that would reduce overall funding by 20 percent. Sectoral departments can use outside reviewers to develop policy alternatives. Policy are made public by means of submission to Parliament. Savings proposals are reviewed in the normal budget cycle, but reconsideration can be initiated at any time. The reconsideration procedure is guided by a small ministerial commission led by the Prime Minister, the Vice Premiers, and the Minister of Finance. The Ministry of Finance provides advice and administrative support for the reconsideration. Increasing Accountability for Results 4.35 Accountability for results is constrained by the lack of public information on policies, programs, and outcomes. Recent civil service reforms have proved counterproductive because they have paralyzed the civil service with uncertainty and are  - 67- seen as a staff cutting exercise that will not lead to the achievement of government objectives. 4.36 The government could increase accountability by taking the following actions: • Publish the program budgets of line ministries (as an annex to the Yellow Book, for example). • Include a table with sectoral (or functional) allocation of spending in the Yellow Book by creating a template that translates the current organizational budget into a sectoral one. (Given that the current budget preparation system is already automated, creating such a template would be straightforward.) • Include a detailed report on budget implementation and outcomes as an annex in the Yellow Book. The budget speech would be the best vehicle for summarizing such an annex. • Revise the budget preparation agenda so that the budget is passed before the budget year starts. Increasing the certainty of funding would also increase the accountability for line ministries to deliver results. 4.3 7 Provided political commitment can be obtained, these reforms stand a good chance of succeeding. Many of the necessary changes require common sense and perseverance rat~er than high-tech solutions, and MOFED has an enthusiastic team of experts working hard on budget reforms who are familiar with good budgeting practices in other countries. 4.38 Including the new budget framework in a revised Finance Act would add credibility to the process over time. Countries that have gone through major budgetary reforms, such as the United Kingdom and New Zealand, have revised their budget laws to reflect the reforms. Embedding the new budget framework in the law would enhance its credibility and, consequently, the chances for achieving better budgetary outcomes. Improving the Financial Management of Extrabudgetary Funds and Organizations 4.39 Extrabudgetary funds and organizations are a growing threat to government financial discipline. If not well managed, these entities can undermine accountability for delivering results. As part of the Public Sector Reform Program, setting up independent agencies outside the budget has become a popular means to deliver services. For a supervisory ministry these agencies can serve to meet staff cutback targets without affecting service delivery. When well managed, such independent agencies can boost government productivity and improve the delivery of government services. However, the financial management framework for these organizations and funds is often inadequately defined and could create contingent liabilities for the government. Moreover, because civil service salary rules do not apply, government services may end up costing more if produced by independent agencies.  - 68- 4.40 The government should consider improving the financial management of extrabudgetary organizations. Although management could be tightened by applying the (revised) financial regulations, the government would be better served by developing special framework legislation and regulation for the financial management of these institutimis. 4.41 The government should gradually expand its reporting on all government activity, including extrabudgetary funds, public sector borrowing requirements, and contingent liabilities. Reporting must be done both in the budget and in the financial reports of the government. Over time reporting on local government activity and budgets will also become necessary to inform Parliament, the general public, and investors about the government's activities. C. SUMMARY OF PRIORITY ACTIONS IN THE SHORT TERM 4.42 Zambia has made important strides in improving the budget system over the last several years. However, the task ahead remains large, and the government needs to set clear priorities for reforms in budget preparation and implementation. 4.43 In the short run, for budget preparation, the government should focus on the following actions: • Fully involve political decision makers in budget preparation to increase the credibility and binding nature of the budget; • Pass realistic budgets with sufficient contingency reserves to absorb adverse revenue developments • Accelerate the budget preparation calendar to ensure that the budget is passed before the fiscal year starts For budget implementation, the government should: • Finalize the inventory and control of arrears; • Set up a commitment control system; • Extend the cash planning horizon from 1 to 3 months and publish monthly and annual cash release statements; • Complete the inventory of Government accounts, require daily reporting on those accounts to MOFED and the owner of the account, and instruct the central bank to report daily to line ministries on the position of their subaccount and bounced cheques; • Consider sanctions against controlling officers who are repeated offenders of regulations. 4.44 Beyond the short term, the government should take a comprehensive approach to budget reforms. The strategic goal of these reforms should be to improve the results of  - 69- government spending by increasing transparency and accountability for results. This could be done by introducing a Medium Term Expenditure Framework supported by a government financial management system that no longer relies on cash budgeting. This chapter has discussed many steps involved in such a comprehensive reform, but it is also important for the government to define a vision for these reforms, a strategy to achieve such a vision, and devote the resources and manpower to realize such a vision.  Annex 1 Page 1 of7 FISCAL SOLVENCY AND SUSTAINABILITY 1. Just like any other entities in the economy, the public sector has to satisfy the intertemporal budget constraint: a fS e - 1 rl dt = Bt . (1) o where S, is the primary balance (budget balance, excluding interest payments) in period t, r is the discount rate, and Bt is the initial level of public debt. Ideally, this public debt should be net - that is, include other public-sector assets and liabilities29 • It can be shown that equation (l) is true if, and only if, the transversality condition is satisfied30 : LimBP * e- rl , .... «) , =0 (2) 2. The above condition says that the present discounted value of a country public- sector debt falls to zero as time progresses. This does not mean that debt should go to zero or even stay constant. Debt can grow at a positive rate in the long run. Of course, a permanent fiscal deficit is inconsistent with the above condition. A deficit at any point in time (or over a period of time) has to be offset by a surplus at another point in time. 3. Define s* as the proportion of output that holds equation (1) above and use the relatlOn • y, = y 0 g. e ,we have a B P s *e (g - r )1 d t - f o - -0 - Yo (3) Solving for s* yields 29 For the above integral to be bounded, r has to be positive. Ifr is negative, the debt would explode - that is, it would be beneficial to borrow forever. As shown by Diamond (1965), in these economies, current debt increase has no impact on future surpluses. 30 See, for example, Cohen (1991) for the external debt for an infinitely lived economy with finite wealth.  Annex 1 Page 2 of7 s*= y: (r-g)PI BP (4) 4. Where Y is output (GDP), g is output (GDP) growth rate, and Bt is the initial government debt. Note that one could define s'" accOlding to the public sector's revenue rather than output. 5. As the above equations show, s* is the proportion of output that would keep the public-sector solvent. The higher s* is, the greater is the proportion of output that must be devoted to debt service to keep the public-sector solvent. Thus, for a public sector debt- to-GDP ratio of 0.65, a differential in interest rate and GDP growth of 0.05, the public sector solvency index, s*, is 0.03 - that is, about 3 percent of GDP would have to be devoted to debt servicing for the government to be solvent. 6. s* could also be seen as the primary balance. s*>O implies that a primary budget surplus is required for fiscal solvency if the country started out in a deficit position and if the real rate of interest exceeds output growth, i.e. ,(r-g»O. Thus, if there is initial public debt and r is greater than g, the public sector has to make debt service payment at least equal to s*, or equivalently, it should have a primary surplus equal to s*. A primary fiscal surplus less than that amount (or a primary fiscal deficit) in that case implies perpetual public sector borrowing and infinite debt. For a country whose rate of output growth exceeds the real rate of interest, (r-g)<O, incurring a primary deficit is still consistent with solvency. However, a deficit higher than s* implies that the country's fiscal position is moving away from a solvency position. As will be shown below, many developing countries are facing either a GDP growth rate lower than the real rate of interest, or a primary deficit much larger than one consistent with solvency defined in (4) above. Note that for any given actual primary surplus, the higher output growth, the smaller is the required primary deficit needed for solvency and hence the smaller is the needed adjustment. The real discount rate is usually approximated by the long-run cost of capital. 7. In equation (4), the solvency index depends on a static component (the traditional initial value of debt to output ratio) and on a dynamic component, (the expected real interest rate and real output growth). It is this dynamic component that the solvency concept introduces to conventional measures of the public-sector debt overhang. It could be argued that it makes little sense to use conventional ratios, such as debt-to-GDP, to measure the domestic debt overhang because a debt-to-GDP ratio of, say, 65 percent may be low for a country whose growth prospects are considerable but high for a country where growth is declining. Similarly, a given amount of fiscal adjustment may be adequate in a country where growth can be readily restored (for example, where structural 31 The above equation was constructed on continuous tenos. In discrete tenos, it could be shown that (r- r-g y) becomes - - . 1+ g  Annex 1 Page 3 of7 refonns are already undertaken) but inadequate in one where economic growth is low due to policy distortions32 • 8. No matter what the initial public debt stock, if g > r, and if the public-sector's primary surplus is equal to interest payments, (that is, net debt is rolled over), the country's debt remains constant and its discounted value goes to zero when the real interest rate is positive. The public-sector will remain solvent as long as the real interest rate is less than the real output growth. This is because the mechanics of compound interest rates are such that as long as the real rate of interest is positive, the discounted value of debt will vanish as it approaches infinity. However, a situation where the long run growth rate of output is pennanently higher than the real interest rate is not possible. Discount Rates and Solvency Adjustment 9. The solvency of the public-sector is partially detennined by the difference in the long-run real interest rate and the long-run, real growth of output. The real interest rate should be detennined free of restrictions and distortions in the economy. For example, domestic interest rates (on Treasury bonds or money-market funds) in developing countries often do not reflect the scarcity of funds because public-sector banks act like agents of the Treasury department. Care should be exercised in choosing a domestic interest rate which would reflect the true cost of capital. 10. In the above fonnula, Bt is the public-sector debt, and includes both domestic and external debt. The discount rate applied to this debt is a weighted sum of the discount rates used for each component, the weights being their respective shares. In the long run, these rates reflect the opportunity cost of capital at home and abroad and are closely approximated by the real interest rates that would prevail in a market free of distortions. In industrial economies, the long-run real interest rate is often equated with the long-run rate of return to capital and estimated between 7 and 12 percent33 • For developing countries, the scarcity of capital would imply a higher real rate of interest. A 10-15 percent real rate of interest is usually considered appropriate. 11. While the concept of solvency is important for new creditors, it is a moot point for existing creditors. In the late 1970s, for instance, New York city was broke - that is, both illiquid and insolvent - but this did not prevent creditors from getting together to bailout the city. The issues facing existing creditors are different from those facing new creditors. Existing creditors are looking at alternative actions to recover assets, while new creditors are faced with choosing the best rate of return for their money among a 32 This statement assumes that the fiscal adjustment undertaken is growth neutral. 33 See, for example, M. Boskin , "Taxation, Saving, and the Rate of Interest," Journal of Political Economy, Vol. 86, April 1978.  Annex 1 Page 4 of7 range of investments. Thus knowing a country is insolvent may help new creditors but does little for existing creditors. 12. What matters for existing creditors is not whether the public-sector is insolvent, but whether the current fiscal stance would move the country away from or closer to solvency. The direction of policy change is an important factor in determining whether new money should be given under existing conditions, or whether tougher conditions should be imposed before new money is forthcoming. 34 In this paper, fiscal solvency adjustment is defined as the difference between the primary balance required for solvency and the actual primary balance. A positive number, therefore, indicates that the country in question needs to make fiscal adjustments to restore solvency. A negative number indicates that no adjustment is required. FISCAL SUSTAINABILITY AND LIQUIDITY 13. The above concept of fiscal solvency requires knowledge of several long run variables: real interest rate and real growth rate, both are not directly observable and have to be approximated. As discussed above, one could use a real rate of interest in the range of 10-15 percent as a proxy for the long run real interest rate and real GDP growth in the past as a proxy for long-run output growth. These assumptions will be used in the empirical section below. 14. Alternatively, one could observe a one-period budget constraint to infer about solvency condition in the future. We define fiscal sustainability this one-period condition such that solvency can be assured in the future. Thus, fiscal sustainability can be derived from the instantaneous view of the budget constraint: D + iB + Ei" (I-Jl)B* = B + EB*(I-Jl) +M (5) where D is the primary fiscal deficit, i is the nominal interest rate paid on domestic debt, E is the nominal exchange rate (domestic currency per unit of foreign exchange); B is the public-sector's domestic debt, B* is the public-sector's foreign debt, M is the monetary base, and Jl is the grant or "soft money" component of the budget. A dot above a variable indicates its rate of change. Divide (5) by P and arrange to express in real terms, defining b=BIP and b*=EB*IP: and note that : 1\ ( B / p ) :::: b + b p 34 This was the situation for Washington, D.C. in 1995 when Congress imposed a control board on the district.  Annex 1 Page 5 of7 1\ 1\ and ( E B * / P ) b * - b * (e p *) Equation (4) can be expressed: • 1\ • 1\ 1\ 1\ d + ib + i* (l-I1)b* = b + b p+ (1-I1)(b * -b * (e- p*) + m(p+ g) 1\ 1\ 1\ • • 1\ d + ib - b p+ (i* - p* +e)(I-I1)b* = b + (1-I1)b* +m(p+ g) or, . . " d + rb + (1- l1)b * (r * +e) = b + (1- l1)b * +m(p + g) where r = i - P Divide the above by y d I y + rb I y + (b * I y)(1-I1)(r * +e) = bl y+ (1- p)b * I y + (m I y)(p+ g) (6) A b ,. .., A· b· D e fime..,= y x • • • • 1\ Hence bl y = /3+ ~ g; and b* Ix = f3*+ ~* X or, in terms ofy, . b* = x (ft * + P* ~) (7) y y Hence ~ + _rb + ..;..(l_-..;...J.l.:.....)b_*-,(r_*_+_e..;..) = _b * + (1- J.l )_b *+ m (p + g) (8) y y y y y y or, d "'.. 1\ m - +rp + (l-J.l)(x/y)p*(r*+e)= p+f3g +(1-p)p+p* x) + - (p + g) y y  Annex 1 Page 6 of7 Also from the quantity theory of money: Mv=Py (9) In the short term, assuming a fixed velocity of money demand v, m/y= lIv Hence d • x · " f3 +(I-fl)- f3*+p(g-r)+ (I-fl)(x/y)f3*(x -r*-e) + lIv (p + g) (10) y y 15. Equation 10 shows a snapshot of the government budget constraint. This constraint depends on a number of factors, including the existing stock of domestic and foreign debt p and p * ; real interest and growth rates, r and g; the proportion of exports in national output ~; export growth rate i; real international interest rate r*; real y exchange rate change e; the inverse of the velocity of money demand m, and inflation y rate jJ. 16. . . We now define the condition for public sector sustainability as one with p = p* = 0 ,or d p(g-r)+ (1-Il)(x/y)f3*(x -r*-e) + lIv (p + g) (11) y 17. The condition states that there are three possible sources of financing the primary fiscal deficit in a sustainable way : domestic borrowing if output growth is greater than the interest rate on domestic debt; by external borrowing when export growth is higher than international interest rates, plus currency depreciation; and by money financing when it is consistent with seignorage. Because it relates to a one-period budget constraint, it also shows the liquidity constraint of the public sector. 18. Define s2 = - d as the primary surplus (expressed as a percentage of output) y needed to achieve debt sustainability for the public sector, s**= p(r-g)+ (l-fl)(x/y)f3*(r*+ e -x) - lIv (p + g) (12) 19. Equation 12 constitutes the necessary and sufficient condition for debt sustainability of the public-sector. 20. To assess progress of fiscal policy, the above formulation of fiscal sustainability condition needs to be compared to the actual fiscal deficit. What is important from a  Annex 1 Page 70f7 policy standpoint is whether the existing fiscal stance would drive the country away or towards sustainability. To measure the direction of policy change, we define fiscal sustainability adjustment the difference between the sustainable primary balance defined in equation 12 above and the actual primary balance. A positive number indicates the need for fiscal adjustment and a negative number indicates no adjustment is required as far as liquidity and sustainability are concerned. While the above formulation seems to be mechanical, it should be noted that in practice, the calculation requires a great deal of data interpretation. What is more important, the exercise is expected to help the analyst be aware of the interrelationship among critical variables. 21. The sustainability condition of the fiscal deficit, equation 12, differs from the solvency condition, equation 4, in several aspects. First, it is a one-period-budget constraint, unlike the intertemporal budget constraint of equation 4. Second, data for equation 12 are observable and are readily available. Third, for j.l = p * = 0, s** is smaller than s* by the amount of inflation tax (including seignorage) which one can extract from the public in any period but presumably not in the long run. 22. The two indicators proposed above, s* and s* *, can address the shortcomings of the conventional fiscal indicators in a number of ways. They take into account the existing stock of internal and external debt, as well as other macroeconomic variables such as export growth, real interest rates, real exchange rates. Fiscal performance is assessed against long run solvency perspective, as well as short run liquidity. The adequacy of fiscal efforts therefore can be evaluated both within a country and across countries over time. There is less dependence on the coverage of fiscal data than on movements away from or towards a policy objective. However, given the linkages between fiscal policy and other macroeconomic policies, these indicators need to be supplemented by a qualitative analysis of fiscal issues. 23. The short-term sustainability condition in equation (12) applies to the fiscal deficit at any point in time. Summing up equation (12) for all time periods will lead to the solvency condition (4), although the converse is not true. Equation (12) differs from equation (4) in several aspects. First, (12) is a snapshot of the fiscal situation at a particular time, whereas equation (4) covers all time periods. Second, because of its short-term nature, equation (12) deals with actual economic conditions at a particular time. In discussing empirical results of the model, two other concepts are also used to compare policy performance time and across countries - external solvency and external sustainability. These concepts are developed in another paper and refer to the same approach as above, except that the model is applied to external instead of internal debt.  Annex 2 Page 1 of2 CENTRALIZED VERSUS DECENTRALIZED PAYMENTS PROCESSING I. Government payments processing can be done in a variety of ways, ranging from centralized processing to decentralized processing (Annex Figure). Each of the four principal modes of payment system can provide adequate expenditure control, provided that the flow of infonnation is adequate and authority is matched with responsibility. In Zambia, this is currently not the case. Zambia's payment system comes closest to case 2a, but neither the infonnation flow from central bank to ministries, or from ministry level ledger system to Ministry of Finance Ledger is adequate. 2. In further refonning its treasury system, Zambia has to choose the mode it deems adequate, and develop the appropriate system accordingly, whether centralized, or decentralized. In Australia, for example, all commonwealth payments are processed centrally, there is a Commonwealth account in the central bank, and retail payments are handled by a fiscal agent bank (case la). In New Zealand, payments are made by ministries themselves, drawn on a bank account in a commercial bank acting as fiscal agent, which also handles the government's revenue account (case 2b). At the end of the day, for cash management purposes, all accounts in the commercial bank are emptied, and the balance is put in the Government's account at the central bank.  Figure. Alternative Models for Expenditure Processing Page 2 of 2 Case la: Treasury is responsible for making payments. Central Bank directl~ responsible for retail banking operations. B"'I~"P_Ul~ BIICIgOI AoProPn'-s • InformaUon Flows Mlnlatry of Rnance ~ ....... ............... ........• C.... lOIlcla .... FI ..., R.pons FromS,.lIm y .- apen«*tufe TranlKCton. from _ . : Spending Mlnl.lI1n & SUa ~ Sprendang MI""tne" Ul'hta To T,..••..,., finanCial Reporu .. Tre. . ury (c..ml y. T,.•• ury LodtllrS,... m (C.ntral .nd ~: Eapolldo...,. Trans....,.,s R.ports .. T....u.... • Central Bank TSA P..,_n"IOSu_ Accounl\. Commercial Gowemmon,R_pts Banks :4 ~ """'T........, & RogIonII Reglonel PoymentIIR_s 0IIIce.) I- - ,,.,.m.) Case Ib: Treasury is responsible for making payments. Retail banking ope~ions ca,!!ed out through a fiscal agent. BudQe' <:;_".., Proposols .. Central Bank Ministry of Finance Con.ollclated Fllc.1 R.porta From S,...m TSA ................................. ,tl DaotyR_ y Eapor1dllUtll Tron..dlons of F'.... Agent IIattI< Accaunl - 0ItIy DePOSI"1D lSA ""'" FIscal Avon'_ SpendinG MINltnea I Units l ....ury _.~~n~.lut1lT~_. Spending _._ !I!.!".!i''Y._ .• Treasury Lodtllr S,...m ~".'" SupplI.r Accoun!\,. Commercial Mlnl.lI1e. & SUs . Fui.".. .. ifepOrts-... . (Conftl (Central.nd ~.;t- R.po,," ..-i.......... -- _~..m!!!!"lB.C!'l!ts__ Banks from Treasury & Regional Rag'....1 PoymonlJlR""pto 0IIIc0.) 1--' S,.lImo) Case 2a: Ministries and spending units are responsible for making payments. Central Bank directly responsible for retail banking operations. BudQel G_Io••., ProPOSl'S BudQel AWOplllbons -----~ Ministry of Rnanc. •............................... fIs,... Conaohdatecl F ••cal Reports m ~. Treasury Centrol (central 0IIIca) Tre••ury ~ Eapendltunt Transac:bons Lodg.rS,...m from Spending M""s"'" I Urn" 1-- R.pon. ID T....ury .. Posted m MiniStry SYllem y 1 1 ------ _ _---._-_._-. - . ~.pons ..I P.nodoc .. from MI.lStry SyJlemJ on Poymon\sIRlCIIPIJ ° .---_. . Spending Monistry ContrOl T....ury lldQer Sysl.m ClntralBank Poymo." ID Supobor Aa:autlts Commercial Mlnlall1e. & SUs Low' ExpendoIUt'll T,....c:\JOIIl F""" Mlnlltry IyJloms . ----- ------ .. - -_. -- TSA .. _~,!,!!!!~I RK!!P.!!.1 Banks L.dgerS,....... ~------.- ..- RoportIlD Mmostnos .. on poymon\sIRlCIIPIJ Case 2b: Ministries and spending units are responsible for making payments. Retail banking operations carried out through a fiscal agent. BudQel ~cIo~" Propoul~ ~ .I!.UdQe~'~""', , Mlnlatry of Rnance C ......IIcIa .... F....I R.pone ................................• FromS,.tem _I Tre •• ury Central : Central bank , (centralomc.) : TSA I I ~Tron_s Recoupment or; ""'" Sponclong M..s",. . I Uno" 0/ F,.... AQonIIlattl< --...,r---oI"lIIy DepoSIts ID TSA ""'" FlScot ~nl _ _ POlted 1ft MIJ1t.try 5y.tom 1! R.pons ID Tr..Jury P.nocIoc Ropons from MU1Istry Syst~ms • J.pon~ '! ~~."!! on~~~I~'~'pto: Spending ID 6ontrol T....ury..,.r System Mlnlall1n & SUa ..  Annex 3 Page 1 of 11 1 Case studies: MTEFs in selected African countries 1. Malawi 1. In 1994 the World Bank and Government of Malawi collaborated in the Malawi Budget Management Review. As an outcome, in 1995 GOM decided to adopt an MTEF (Medium tenn expenditure framework) as suggested by the World Bank. 2. The intention was to carry out the MTEF within a macroeconomic framework with three alternative scenarios (high, medium, low), the central case being the existing scenario. Resources were to be split between capital and recurrent expenditure, and within recurrent expenditure between wage-related and other recurrent expenditure (ORE). Recurrent expenditure was to be allocated to economic sectors using the three scenarios. Sector expenditure allocations would create the conditions for ministries to plan for 3-5 years ahead with reasonable assurance of availability of resources; to assess the need for repair and maintenance expenditure to maintain existing capital stock; to address the imbalance between wages and ORE; and to confront choices and constraints to support rational resource allocation decisions. 3. Beginning in 1995, this approach was applied to the four largest spending ministries (education, health, agriculture and public works) and to the police. The exercise was extended to 12 ministries in 1997-98 and will apply to all ministries in the following year. At the center, the approach involves the development of a macroeconomic framework, the derivation of aggregate and sectoral ceilings and reprioritization of expenditure. The intention is that in due course there will be one budget for both recurrent and development expenditures, and that the MTEF will be prepared in constant prices for the budget year plus three further years. 4. The major achievement to date has been in the field of sectoral programs. Here a strategic review of the whole of government helped to define ministries' objectives and functions, besides indicating opportunities for modification, reduction and elimination of activities (duplication of function, transfer to the private sector, contracting out, cost recovery). Initially the emphasis was on the recurrent budget. As a result the quality of budget requests has improved and managers have been involved actively in budgeting, some for the first time. Some problems remain with regard to defining activity costs and bringing together both capital and recurrent costs. 5. In practice, several difficulties have emerged. Government had not developed its own macroeconomic framework and had been using that of the IMF at the time that the MTEF was introduced. A working group to create a macroeconomic model for the I This annex was prepared by Peter Dean  Annex 3 Page 2 of 11 economy has only recently been set up. As a result, the macroeconomic framework has not provided an adequate basis for fiscal planning. Consequently the process for communicating expenditure ceilings to ministries was also imperfect. Government, the IMF and the World Bank agreed the policy framework paper on which ministry ceilings were based. But late announcement of the ceilings for the 1997-98 budget cycle severed the essential link between the MTEF and the annual budget. Besides, the announced ceilings were far below expectations, giving rise to frustration in ministries and confusion about the nature of the MTEF process. On top of this, the failure to obtain overt and sustained political commitment to the MTEF process, even at the basic level of parliamentary participation, means that its influence remains superficial. 6. The reported difficulties are being played out against a fundamentally discouraging background for the introduction of an MTEF. Malawi is one of the poorest countries in the world. The World Bank noted in 1996 that it had slipped into its worst economic crisis in a decade, including the collapse of its fiscal system. A year earlier the Malawi Budget Management Review noted " ... the budget is not working well. Line ministries have little confidence in the capacity of the budget to provide realistic budgets. This has weakened expenditure controL .. the links between the recurrent budget and the development budget have progressively weakened. Government finances have become seriously over-extended. The Government has found it difficult to maintain competitive wages for technical and professional cadres, project implementation is delayed by unreliable funding of counterpart costs and departments lack funds to maintain public sector investments when completed. Donors have increasingly sought to compensate for the weakness of the domestic budget through projects and by other means". 7. Given this background, and the fact that inadequate skill levels throughout the public service in budgeting and financial management are seen as a serious barrier to the cost effective implementation of the government's development priorities, the decision to attempt an MTEF seems heroic. A less ambitious budget reform policy might have been more appropriate. For instance, given the serious fiscal problems faced by Malawi, a realistic objective might have been a macro framework and an MTEF-type approach to budgeting for the budget year only, rather than an MTEF primarily addressing sector expenditure programs but without an adequate macro context, and requiring four year forecasts expressed in constant terms. A less ambitious technical approach targeting aggregate fiscal discipline (as in Uganda), might also have activated greater political support by addressing issues of greater political relevance. In particular, claiming that MTEF would provide greater funding predictability, in a fundamentally unpredictable resource environment, resulted in inevitable contradictions. 8. Also the budget reform policy might have targeted the improvement of basic systems on a time line in which a full MTEF would have been placed as a more distant objective. Inadequate basic systems pose a serious obstacle to the realization of the MTEF concept: separation of decisions and information regarding recurrent and development expenditure; the presence of significant unrecorded and uncontrolled donor flows; the weak control of within-year expenditure; and the absence of up to date,  Annex 3 Page 3 of 11 accurate and relevant accounting information. Substantial improvement of these processes will require both resources and time. 9. In conclusion, the MTEF continues to be relevant as a concept for Malawi. It may yet be real~zed in practice, if it can be properly adapted to the country's capacities and priorities. In particular the lack of a macroeconomic framework and explicit fiscal. policies derived from the framework, deprive the MTEF of its starting point: a definition of the aggregate resources available to fund expenditure over the medium term. Thus the annual ceilings announced to ministries as part of the annual budget cycle have proved to be largely irrelevant. To achieve relevance, the MTEF process requires better fiscal underpinning, sustained and explicit political support and stronger means of enforcing budget discipline. 2. Mozambique 10. In 1996, the World Bank, Government of Mozambique and several donor countries undertook a review of the country's fiscal management, the outcome of which was a report by the World Bank (1996). During the review, an MTEF was suggested as a method for improving inter-sectoral expenditure allocations. In the following year, Government prepared its own reform plan (Republica de Mocambique, 1997) which sets out a comprehensive plan for expenditure management reform, including an MTEF. A consultant was hired to advise government on the adoption of an MTEF and reported on its advantages and how to go about introducing it (Muggeridge, 1997). 11. The expenditure reform strategy consists of improvements in several key processes: the budget law, budget classification and coverage, accounting and monitoring, cash management, debt management and auditing. Recently, as part of this strategy a budget framework law has been passed and a new system of budget classification introduced. However, much remains to be done if public finance systems are to perform adequately. This is because the starting point for reform is quite basic: single entry accounts (often hand-written); a proliferation of bank accounts for conducting government business; severe problems of cash flow; disregard of the budget calendar; extreme reliance on donor funding (most of which escapes budget disciplines); strictly segregated procedures (including budgeting and accounting) for recurrent and investment expenditures; and weak controls. 12. The country has recently emerged from a long civil war and is trying to move from a communist system to one which is more democratic and market-orientated. Moreover considerable fiscal adjustment is required. Even assuming very considerable debt relief, the budget deficit needs to be further reduced at a time when donor funding is decreasing. This has serious implications for the sustainability of spending programs. The need for fiscal adjustment and for consequent re-examination of spending programs explains why an MTEF is in principle an attractive proposition. However, in practice it is beset by potential problems. Besides the lack of supporting systems already referred to, a particular problem is evident in the field of human resources. There is a severe lack of  Annex 3 Page 4 of 11 skilled staff needed for budgetary reforms. Problems include lack of basic skills, lack of staff with higher qualifications, loss of staff to the private sector, immense salary differentials between the public and private sectors and moonlighting to supplement inadequate wages. The Ministry of Finance and Planning is said to be one of the weaker agencies in terms of skills. 13. Nevertheless Government is said to be introducing a rolling, five year MTEF starting with the preparation of the 1999 budget. Some ground work has been done at the level of sector programs. Now there are two immediate challenges: to link the sectoral work with macroeconomic realities (defining the aggregate resources available) and to ensure that the 1999 budget is prepared within the MTEF framework (defining sectoral allocations and seeing that they are observed). 14. It is by no means clear that the basic building blocks for an MTEF are yet in place. The sentiment that "Mozambique has the advantage of not yet starting some of the refonns that other countries have implemented and can benefit from their experience." 2 needs to be balanced against the difficulty of achieving something as exacting as an MTEF in Mozambique's present situation. Certainly the timetable for effective adoption will stretch to several years. The World Bank writes with caution on the timing issue: "Once the basic elements of a new budget and accounting system and its associated planning, legal, accounting, training, staffing and governance structures are in place, or are being put in place, it will be possible to move on to other elements of fiscal refonn ....." (1996, p.55). 15. Given the weaknesses still experienced in budget and related systems, it might be more realistic to plan for an MTEF of a less demanding nature or to postpone its introduction. Following the first option, it would be possible to start to apply the concept just to the budget year. Then when experienced had been gained with linking the macro framework to expenditure plans, and in coordinating expenditure plans with the requirements of the annual budget, the MTEF could progressively assume a medium tenn orientation. To aim at the outset at a five year MTEF is over-ambitious, especially in view of the lack of real meaning attached to two and three year expenditure figures in far more predictable environments. 3. South Africa 16. The Budget Speech for 1998 announced the adoption of a medium tenn expenditure framework. It is described as" the operational plan by which we give substance to our reconstruction and development efforts". Its goals are given as: 2 Muggeridge 1997, p 26.  Annex 3 Page 5 of 11 to strengthen political decision-making in the budget process ... to enable Cabinet to make the link between the budget allocations and the services which we intend to deliver. to strengthen co-operative governance and Jecision-making. to make the Rand go further. .. to deliver better services. to create a medium term environment for public sector planning. to relate what we spend to what we buy. to achieve greater transparency and openness in making budget policy. Strong, public, political commitment to MTEF is an excellent basis for its successful adoption. 17. The National Budget Review provides details of the MTEF process. It is a three year rolling process (the budget year plus two further years) in which expenditure is expressed at estimated current prices. It applies to both national and provincial government and reflects Government's social and economic priorities and the country's reconstruction and development commitments. 18. The benefits of an MTEF are explained as: allocation of resources to priority services. more efficient planning and management. a framework within which policy proposals can be assessed. more transparency in government. a reduction in rollovers (expenditure commitments from previous years). a clear demonstration of how fiscal targets will be met. 19. South Africa's approach to MTEF is comprehensive. The model, which is in course of implementation, contains all the necessary components for successful adoption, (a macroeconomic framework, a fiscal policy, medium term expenditure forecasts for three years, an annual budget process linked to the MTEF, and draft sectoral programs). The components are linked to one another and there is strong political support for the entire process. South Africa's MTEF is at an early stage of development. Nevertheless, already much has been achieved. The following tables from the National Budget Review illustrate in aggregate terms the macroeconomic link and the breakdown by sector.  Annex 3 Page 6 of 11 Consolidated National and Provincial Spending: 199~6 to 2000/01 (In billion rand) Outcome Preliminary MTEF Projections 1995196 1996/97 1997/98 1998/99 1999/00 2000/01 Total spending 154.4 182.7 196.3 205.2 219.8 239.6 as % of GOP 31.1% 32.9% 32.0% 30.7% 29.9% 296% % growth 183% 7.5% 45% 7.1% 90% less debt service costs 31.3 34.3 39.6 43.0 45.3 48.1 Total non-Interest spending 123.1 148.4 156.7 162.2 174.5 191.5 as % of GOP 24.8% 26.7% 25.6% 24.2% 238% 23.7% % growth 20.5% 5.6% 35% 7.6% 9.7% GOP Inflation rate 82% 8.5% 8.5% 60% 55% 5.0% GOP (R billion) 497.3 556.0 613.0 6693 734.3 809.6 (Source: Chapter Sof the National Budget Review, 199B)  Annex 3 Page 7 of 11 Consolidated National and Provincial Expenditure: Functional Breakdown: 1995/96 to 2000101 (in billion rand) Budget MTEF projections Average real growth forMTEF 1995/96 1996/97 1997/98 1998/99 1999/00 2000/01 Protection services: 25.1 28.1 29.6 33 35.2 37.8 2.8% Defense 11 6 11 8 10.7 11.0 11.6 123 -0.8% Justice '1.6 1.7 2.0 25 2.7 30 9.5% Police 9.3 11.4 13.1 141 15.1 16 15% Prisons 27 31 3.9 5.4 5.8 64 12.3% Social services: 73.9 81.4 88.6 102.1 107.1 114.4 3.2% Education 34.6 39.2 40.3 468 49.2 52.2 3.4% Health 16.1 185 20.2 25.1 264 281 5.8% Social security and welfare 15.2 164 184 19.8 207 222 0.9% HOUSing 30 1.6 42 39 4.3 4.3 -39% Other 51 5.7 5.5 6.5 65 7.5 5.0% Economic services: 17.5 19.5 18.9 17.1 18.6 20.6 -2.5% Water schemes, related services 1.4 2.2 1.9 1.9 2.0 2.3 1.2% Fuel and energy 0.2 0.2 0.2 0.1 0.1 01 -9.9% Transport, communication 7.3 77 7.3 6.5 70 7.7 -35% Agriculture, forestry, fishing 35 44 45 4.3 45 4.9 -23% Mining 0.2 0.3 0.3 0.3 0.3 0.3 -1.4% Manufacturing 0.8 1.0 1.1 0.6 06 0.6 -20.6% Regional development 0.9 07 0.9 07 0.7 0.7 -10.2% Other 3.3 3.2 2.8 26 3.2 3.8 4.7% General Admin., other 10.8 13.5 13.2 9.0 10.7 11.7 -9.0% Interest 29.5 34.6 38.6 43.0 45.3 48.1 2.0% Reserve 1.3 05 1.3 1.0 30 70 Total Budget expenditure 158.1 1n.6 190.2 205.2 219.8 239.6 2.4% (Source: Chapter 5 of the National Budget Review, 1998) 20. Also significant work has been done on sectoral programs. Sectoral teams were established in August 1997. They were assigned to six expenditure areas: education, health, welfare, criminal justice and personnel management. Their aims have been to analyze budget submissions so that understanding of the service implications of the proposed budgets is improved; to develop simple spending models linking expenditure, unit costs and outputs; to produce a menu of policy choices linking budget allocations to service delivery; to make recommendations on the structure of the votes in question; and in the cases of education, health and welfare to reach agreement on the amount and structure of unconditional grants.  Annex 3 Page 8 of 11 21. Besides the obvious links needed for any MTEF, the South African version has additional links designed to increase its relevance and chances of success. First, it is located within a larger program for budget reform which focuses on fiscal decentralization, translation of policy goals into service delivery, increased efficiency and transparency, and promoting partnership between the public and private sectors. Second, budget reform is seen as dependent on improvements in financial management which receives specific incremental funding under the MTEF program. Third the extent, cost and conditions of public sector employment are seen as significant issues within the MTEF program. The program specifically addresses personnel management, public service regulations, budgeting for personnel costs and decentralization of personnel decision~making. This establishes a strong link between expenditure management and personnel issues which is quite rare in programs for public sector reform. 22. The MTEF approach deserves to succeed in South Africa because it has been well~conceived and planned and because it addresses the right questions. Its chances of success are enhanced by strong government commitment and a realistic approach to the magnitude of the task which is seen as requiring several years. Whether it can provide the right answers is as yet unknown. Certainly, innovations rarely match the rhetoric used at their introduction. Consequently some redrawing of the boundaries of expectation may occur. Nevertheless, MTEF in South Africa has a good chance of achieving its central aims if it is pursued with energy and detennination. 4. Tanzania 23. Tanzania has had a rolling plan and forward budget (RPFB) for four years but it has been a disappointment. Its status in late 1997 was unclear: the analytical work was done for the three year period to 1998/99 but not published, and it is not known whether RPFB will continue. Budget administration has been dominated by day to day issues of cash management which has reinforced skepticism concerning the feasibility of medium term expenditure planning. 24. The World Bank (l997b) refers to the following problems in Tanzania: "Expenditures under the RPBF have always departed from the projected path and projections for the forward years have not been taken seriously. Although line ministries have prepared and presented combined submissions for their recurrent and development estimates, recurrent and development submissions have in practice been reviewed separately by Finance and Planning. A large proportion of aid flows still fails to be reported in the government budget and accounts ...... There are several reasons why the RPFB has so far failed to make the concept of a medium~term expenditure framework fully operational. Budget projections were over optimistic, thus evading the problem of attaining a realistic initial budget. The Government's political commitment to the budget figures it announced was weak; thus there were frequently large reallocations early in the budget year. General expenditure discipline was poor, and budgets were also thrown off by expenditure arrears carried forward. And weaknesses in estimating, monitoring and accounting were another source of uncertainties" (page 71). Elsewhere we find that "the  Annex 3 Page 9 of 11 RPFB exercises have been diluted by rapid departure from the main macroeconomic assumptions in the plan; it is also not evident that the proposed macroeconomic framework was in fact consistent with maintenance of macroeconomic stability" (page 74); "the development budget accounts for a small fraction (10 per cent) of the recorded budget and an even smaller proportion of actual disbursements ... " (pabe 75); the number of projects is unmanageably large; projects are severely underfunded ... " (page 76); "a comprehensive sector-wide view of requirements is inhibited by the fragmentation of expenditure planning between recurrent and development budgets ... " (page 79); "RPFB has tended to be crowded out by urgency of annual budget preparation and even annual expenditure projections have been quickly derailed by expenditure indiscipline and failure to anticipate and control commitments effectively .... " (page 81). 25. In addition, the following are noted in the same report: as much as 80% of donor funding not recorded in the budget; ministries incurring significant expenditure comrriitments outside budget limits and consequently carrying forward large payment arrears; a high level of within-year deviation from budgets; only a fraction of the development budget released during the budget year; weak supervisory controls at spending agencies and no disciplinary action for over-spending; a two year backlog in Exchequer Account reconciliations and consequent absence of final accounts. 26. Public expenditure reviews carried out by the World Bank in 1989 and 1994 recommended the RPFB approach. The World Bank's latest review concludes " .... the principles and objectives of the RPFB approach remain appropriate. Given the present Government's seriousness of purpose the RPFB, if suitably modified, can still be an effective instrument" (page 71). This is despite the fundamental problems reported above. Comparing this situation, with the conditions which would favor successful introduction of an MTEF as listed in annex 3 below, the likelihood of a successful MTEF for Tanzania is very low indeed. 5. Uganda 3 27. Since 1992, budget strategy has been guided by an exercise known as the annual budget framework. This is an adaptation to Ugandan conditions of the MTEF approach. Its objectives are the same as those cited in the case of Ethiopia in the main text. Before the change, the budgeting horizon was very close: the end of the budget year. This had two major results: budget allocations were made routinely by civil servants rather than by decision makers, and only a weak link was established between the budget and the nation's strategic priorities. The budget framework clarifies strategic choices and is adapted to the needs of Cabinet decision making. Annually, the budget framework paper (BFP) abstracts from the enormous volume of detail contained in annual budgets to focus on a strategic view of government's policies and the resources available to pursue them. 3 Based extensively on Moon, 1997.  Annex 3 Page 10 of 11 28. The BFP was introduced in 1992 to address a serious fiscal crisis. Its early emphasis was therefore on macroeconomic factors. In the next two years it was expanded to cover allocatory issues including planning for civil service changes (cost, numbers and remuneration). The BFP for 1995/96 contained comprehensive three year projections by sector and by ministry, and priority expenditure areas were identified. This provided ~he first occasion for Cabinet to consider ministries' expenditure plans over an extended time horizon, engendering intense interest and debate. However in 1996/97 a large revenue shortfall coincided with the demerger of the Ministries of Finance and Economic Planning. The consequent uncertainty meant that the full three year BFP was not prepared, and the budget exercise concentrated on issues within the budget year. 29. Moon (1997) sees several major constraints preventing the achievement of a more successful budget framework. First, the within-year budget process lacks discipline in the sense that spending ministries do not respect hard budget constraints. He analyses budget and actual data by ministry over a three year period and concludes that a fairly stable pattern of winners and losers emerges, with the "losers" representing sectors which are identified among Government's apparent "priorities". Second, the budget framework is vulnerable to abrupt revenue shortfalls. In this context he notes that domestic revenue instability in sub-Saharan Africa is roughly three times as high as in other developed countries. Third, in the case of Uganda, donor dependence may have been an additional cause of resource instability. Fourth, Uganda's financial markets lack depth and this implies that demand has to be managed by fiscal rather than monetary means. Again, the result is a more volatile environment for planning public expenditures. 30. Moon also indicates the steps now needed to achieve a more realistic budget framework in Uganda. First, care is needed to ensure that the BFP does indeed reflect national priorities. Second, broader ownership is needed to ensure a more disciplined budget implementation. Third, broader ownership implies the need for greater involvement of line ministries and Parliament. He argues that the BFP be distributed to Parliament in good time, that Parliament monitor the BFP against actual budget implementation and that the BFP be published. Finally he sees the BFP as a way of achieving better coordination between donors. 31. The BFP has achieved the following: it has contributed strongly (along with better cash flow management) to macroeconomic stability by promoting aggregate fiscal discipline; it has introduced a medium term focus in cabinet's budget decisions; and has provided a framework in which key policy changes can be more effectively evaluated. At the same time it has encountered problems: it has been a weak aUocatory device because of significant variation between budgeted and actual expenditure; and the sector analysis underpinning the BFP has lacked clear objectives, links between resources and results, integration of recurrent and development expenditure and adequate coverage of donor funded expenditure. 32. Despite reverses, BFP may still have a future in Uganda. High on the list of preconditions for a successful future are the changes listed by Moon above, including  Annex 3 Page 11 of 11 better mechanisms for ensuring consistency between budgeted and actual expenditure which he sees as critical to the future ofBFP. Thus realistic budgets, effective expenditure controls, good reporting systems and strong political oversight are key to future deVelopment of the MTEF in Uganda.  Annex 4 Page 1 of7 Table 1: Central Government 1991-97 (millions of Kwacha) 1991 1992 1993 1994 1995 1996 1997 Revenues Total Revenues, before Grants 40,814 104,521 235,255 449,618 595,911 817.000 1,023.068 Total Tax Revenue 39.954 99,819 227,247 418,879 545.911 752,000 967,068 DIrect Taxes 11,000 34,515 67,293 129,213 I 72,S08 24S,OOO 327,823 IndIrect Ta.xes 28.954 6S,304 IS9.9S4 289,666 373.403 507,000 639,245 On domestic goods & servIces 15,738 30,891 81.513 151,130 2\0,248 264,000 351,000 On Intematlonal trade 13,216 34,413 78,441 138,536 163,155 243,000 288,245 Nontax ReceIpts 860 4,702 8,008 30.739 50,000 65,000 56.000 Memorandum Item External capItal grants 19,656 57,913 1\8,250 119,671 155,689 242,000 259,000 Expenditures Total Current Expenditures before rescheduling 63,896 155,063 389,558 624,735 797,965 906,200 1,105,041 Interest on External Debt before rescheduling 14,783 33,014 1\ 1,155 IS7,866 238,051 237,600 291,732 Interest on External Debt after rescheduling 14,783 33,014 1\ I,ISS IS7,866 167,000 66,000 98,400 Interest on Domestic Debt 3,662 IS,577 71,463 102,627 76,971 122,300 115,000 Other Current Transfers 6,162 13,954 38,498 89,663 106,997 96,000 121,809 SubSIdIes 8,744 4,437 2,500 0 0 0 0 Consumption 30,545 88,081 165,942 274,579 375,946 450,300 576,500 o/w Wages and Salanes 16,100 44,800 77,300 115,051 177,900 221,000 324,000 o/w Recurrent Departmental Charges 8,846 20,907 54,425 63,181 119,636 137,000 147,000 Deficit - Different Definitions Overall balance before grants and before rescheduling -34,837 -71,966 -201,192 -264,289 -356,310 -517,200 -569,973 Overall balance after grants and after rescheduling -15,181 -14,053 -82,942 -144,618 -129,S70 -103,600 -122,479 Total Revenues, before Grants 40,814 104,521 235,2S5 449,618 59S,911 817,000 1,023,068 Expenditures and net lending before rescheduhng 75,651 176,487 436,447 713,907 952,221 1,251,200 1,495,041 Expenditures and net lending after rescheduling 75,651 176,487 436,447 713,907 881,170 1,162,600 1,404,547 Pnmary balance before grants -16,392 -23,375 -18,574 -3,796 -41,288 -157,300 -169,432 Source: MinIStry of Finance and EconomIc Development (MOFED) and IMF  Annex 4 Page 2 of7 Table 2: Nonfinancial Public Sector Deficit (millions of Kwacha) 1995 1996 1997 Central Government Deficit -129,570 -103,600 -122,479 Local Government" Local Government Revenues 26,550 29,370 24,686 Local Government Expenditures 126,378 139,801 117,505 Local Government Arrears N/A N/A -18,167 Local Government Deficit -99,828 -110,431 -110,986 Other Government Fundsb Pension Funds ZNPF - Noncompliance -2,800 -3,600 -5,200 CSPF and LASF Outstanding Claims -25,000 -31,000 -45,000 Outstanding Liability to Pension Funds -27,800 -34,600 -50,200 Privatization Fund Revenues 12,468 38,733 1,775 Expenditures 2,582 1,000 o Privatization Fund Deficit/Surplus 9,886 37,733 1,775 Rural Electrification Fund (REF) N/A -4,457 -1,335 Road Fund (RF) -5,200 -9,800 N/A Public Entetprises J ZCCM Surplus/Deficit -29,281 -198,722 -181,406 Other SurpluseslDeficits 20,944 -52,092 -53,514 of which: energy companies 11,519 -50,624 N/A of which: water companies -777 -340 N/A of which: transport companies 44,567 19,147 N/A Public Enterprise Deficits -8,337 -250,814 -234,920 Central Government Guarantees on Parastatal Non-peforming Loans N/A N/A -26,170 Nonfinancial Public Sector Deficit -260,849 -475,969 -544,315 a Revenues estimated based on share coming from grants Expenditures estimated based on ratio to revenues. Arrears includes salary and retiree arrears minus what local governments owed to pensions and parastatals. b. Governments' noncompliance to ZNPF estimated at 40% and assuming that membership contributions comprise 8% of wages and salaries Outstanding claims based on 1994 estimate for outstanding early retirement claims c Public Enterprise data is given by fiscal year Non-ZCCM public enterprise deficits for 1997 were estimated at 1% of GOP, the same share as for 1996 Source: Bank of Zambia, MOFEO, and Bank and Fund Staff Estimates  Annex 4 Page 3 of7 Table 3: Quasi-Fiscal Deficit (millions of Kwacha) 1995 1996 1997 Central Government Deficit Excluding Grants -129,570 -103,600 -122,479 Nonfinancial Public Sector Deficit -260,849 -475,969 -544,315 Central Bank Unsettled Losses on Parastatal Debts· N/A N/A -21,846 Payment made to BOZ by GRZ N/A N/A 6,720 Foreign Exchange Losses Special Holdings of Govt Securities (prior to 1996) -256,036 -187,562 -152,572 Trading Losses on GRZ Transactions (1996 and after) -148 -2,176 Trading Losses on ZCCM Transactions (1996 and after) -1,114 -75 Payments to Failed Banksb -62,700 -27,031 Central Bank Deficit -256,036 -251,524 -196,980 Quasi-Fiscal Deficit -516,885 -727,493 -741,295 Memorandum Items: GDP at Current Market Prices 2,826,272 4,077,656 5,351,434 Holdings ofGovt Securities for IMF -469,802 -651,962 -640,178 Stock of Contingent Liabilities/GOP 298,080 433,371 639,954 a Unsettled losses on parastatal debts covers foreign exchange losses on parastatalloans which have not yet been settled through the Government's debt buy-back program These losses are all considered as a flow for 1997 given that payment is required as soon as responsibility is determined. b Bank failures occurring in late 1995 are included in" payments made in 1996. Failures occuring In 1997 are included in payments made in 1997. Source: Bank ofZarnbla, MOFED, and Bank and Fund Staff Estimates  Annex 4 Page 4 of7 Table 4: Contingent Liabilities of Various Public Entities (millions of Kwacha) Realized Stocks Losses· As of 12131196 As ofl2131197 98 Estimate 1997 GRZ Loan Guarantees to Parastatals Dairy Produce Board 129.15 129.15 129.15 129.15 Development Bank of Zambia 101.97 101.97 101.97 101.97 Medtcal Stores Limtted 254.91 171.08 171.08 171.08 Nitrogen Chemicals of Zambia 855.17 1,595.01 1,595.01 1,595.01 Untted Bus Co of Zambia 828.50 2,200.04 2.200.04 2,200.04 Zambia Airways 3,909.14 3,905.49 3,905.49 3,905.49 Other Parastatals 1,121.72 11.88 11.88 11.88 Guarantees to Cooperatives 4,483.86 5,511.54 5,511.54 5,511.54 Guarantees to Councils 76.25 2.85 2.85 2.85 Guarantees to Corporations 3,780.59 7,729.75 7,729.75 7,729.75 Zamibia Coop. Federationbt 4,046.46 4,811.86 6,121.09 4,811.86 ZCCM - Swiss Bankb2 180,000.00 Total Guarantees 19,587.73 26,170.62 207,479.86 26,170.62 Local Government Arrearsc 27,890.00 36,602.00 42,014.00 18,167.00 Pension Fundsd 228,756.00 300,215.00 344,599.00 45,000.00 BOZ Loan Guarantees Food Reserve Agencyb3 3,537.50 2,700.00 ZCCM - Standard Chartered 20,000.00 10,000.00 ZCCM - ZANACO 10,000.00 11.700.00 ZCCM - Citibank 2,500.00 ZCCM - Finance Bank 6,000.00 ZCCM - Stanbic 6,500.00 Unsettled Losses on Parastatal Debts· 21,846.00 21,846.00 21,846.00 21,846.00 Total Guarantees 21,846.00 70,383.50 46,24600 21,846.00 Liabilities Accounted for in Budget GRZ Allowance for Guarantees 0.00 385.00 0.00 Total Liabiltties 298,080 433,371 639,954 111,184 Asa%ofGDP 7.31% 8.10% 10.42% a Guarantees offered by GRZ are assumed as part of the quasl.fiscal deficit m 1997 given that less than I% were perfonnmg b Assumes 1997 end ofpenod DOZ exchange rate (2198 SJaltstics Fortmghtly) and 1998 exchange rate ofKI8001S. c Local Government Arrears as of 1211997 (plus salary arrears up to 3198) amounted to 7% of GOP. This share used to estimate outstanding arrears 1996-1998 Amount of liability Included In 1997 quasl.fiscal defiCit IS for salary arrears only d Government's outstandmg liabilities to CSPFILASF. 1994 liablhty estimated at K119 billion (5.6% of GOP) This share used to eSllmate 1995·1997 amounts. Outstandtng claims used In quasl·fiscal defiCit e Covers foreign exchange losses on parastatalloans not yet sellied through the Government's debt buy.back program Losses are conSidered as liability for 1997 given that payment IS reqUired as soon as responSibility is detennmed. Souru: Bank of Zambia Accounts. MOFED, and Dank staff eSllmates  Annex 4 Page 5 of7 Table 5: Holdings of T -Bills and Banking System Claims on Government (millions of Kwacha) Dec-93 Dec-94 Dec-95 Dec-96 Dec-97 Nonbank Holdmg ofT·Bills 23,466 23,637 44,865 58,410 45,827 Commercial Bank Holding ofT-Bills 71,244 79,952 141,876 164,057 195,343 Central Bank Holding ofT-Bills 255 21,781 24,662 12,850 6,863 Total T- Bills 94,965 125,370 211,403 235,317 248,032 Bank of Zambia Net Claims -14,305 5,018 -52,344 -54,041 -60,603 Commercial Banks Net Claims 66,115 71,815 147,639 159,310 181,339 Total Net Claims 51,810 76,833 95,295 105,270 120,736 Source: Bank of Zambia and MOFED  Annex 4 Page 6 of7 Table 6: Local Government's Arrears As of 12/31/97 (Thousands of Kwacha) COUNCIL PAYE LASFI\.'CF SALARIES RETIREES ZAMTEL ZESCO ZNPF OTHERS TOTAL Central Provmce 311 941 387 459 21 87 10 230 2.445 Copperbelt Province 4,364 4,227 958 1,649 126 4,706 48 615 16,692 Eastern Province 224 914 561 109 32 18 12 117 1,988 Luapula Province 137 316 399 28 71 2 53 379 1,384 Lusaka Province 107 1,984 245 161 389 5 134 2,586 5,610 N Western Province 56 426 516 24 43 18 216 154 1,454 Nonhern PrOVince 263 778 1,308 127 199 6 176 310 3,167 Southern Province 725 1,628 703 237 91 154 7 808 4,353 Western Provmce 217 354 522 543 5 21 5 223 1,891 Total 6,404 11,567 5,601 3,336 977 5,017 660 5,421 38,985 Source MiniS!!), of Local Government and HOUSing  Annex 4 Page 7 of7 Table 7: Local Government Councils' Revenues and Expenditures Approved Estimates for 1996-1997 (Percent) Revenues 1996 1997 Expenditures 1996 1997 General Rate and DeveloEment Funds Accounts Local Taxes 23.4 41.3 Personal Emoluments 33.9 39.4 Charges 19.7 21.3 Provision of Services 8.0 10.0 Other Receipts 27.6 25.8 Other Expenses 39.3 43.9 National Support 8.2 3.5 Liquor Undertakings 0.6 0.6 Rest HousesIMotels 0.5 0.5 ShopslMarkets 0.1 0.1 Other Undertakings 0.2 0.3 Water Fund Account Sale of Water 11.7 6.2 Personal Emoluments 5.0 1.7 Sewerage Charges 5.5 0.1 Provision of Services 4.3 1.8 Other Income 2.5 0.3 Other Expenses 9.4 3.2 TOTAL 100 100 100 100 Source: MLGH, Summary of Local Councils Revenue and Capital Estimates, 1996 and 1997 r---- - - -- --- - ---- CATALOGUERS/FilE CONFIDENTIAL Report No:. 18552 ZA Type: ER '- - -- - ----