Report No. 26162-ZA Zambia Public Expenditure Management and Financial Accountability Review November 2003 Poverty Reduction and Economic Management 1 South Africa Africa Region Document of the World Bank IMF International Monetary Fund INTOSAI International Organizationo f Supreme Audit Institutions IPSAS International Public Sector Accounting Standards K C M Konkola Copper Mines KDMP Konkola Deep Mining Project LCC Local Cost Compensation LCMS Living Conditions Monitoring Survey LM LineMinistry LPO Local Purchase order MFNP MinistryofFinance andNationalPlanningLPO M o H MinistryofHealth MTBPS MediumTermBudget Policy Statement MTEF MediumTermExpenditure Framework MTPRS MediumTermPayReformStrategy MWS MinistryofWorks and Supply OAG Office o fthe Auditor General OP Operational Policy PAC Policy Analysis and Co-ordination (Division o f Cabinet Office) PE Personal Emoluments PEM Public Expenditure Management PEMFA Public Expenditure Management and Financial Accountability PETS Public Expenditure Tracking Surveys PPP PurchasingPower Parity PRGF Poverty Reduction and GrowthFacility PRSC Poverty Reduction Support Credit PRSP Poverty Reduction Strategy Paper PS Permanent Secretary PSC Public Service Commission PSCAP Public Service Capacity Building Project PSMD Public Service Management Division PSPF Public Service Pensions Fund PSRP Public Sector ReformProgram PSU Procurement Supply Units PTA Parent Teachers Association RDC Recurrent Department Charges RFP Request for Proposals SASE Selective Accelerated Salary Enhancement SOE State OwnedEnterprise SSA Sub-Saharan Africa ST Secretary to the Treasury TA Technical Assistant TAZAMA TamanidZambia Pipeline TAZARA Tanzania -Zambia Railways TNDP Transitional National Development Plan UNCITRAL UnitedNations Commission Intemational Trade Laws USD U.S. Dollars VAT Value-added Tax WSM Works and Supply Ministry WHO World Health Organization ZAMIM Zambia Institute o f Management ZBIC Zambia Insurance Business College Trust ZCCM Zambia Consolidated Copper Mines ZESCO Zambia Electricity Supply Company ZNOC Zambia National Oil Company ... -111- ZNTB Zambia NationalTender Board ZNTBA Zambia National Tender BoardAct ZR Zambia Railways -iv- TABLE OF CONTENTS ACKNOWLEDGEMENTS .......................................................................................... viii ... EXECUTIVE SUMMARY xi PART I INTRODUCTIONAND CONTEXT . ............................................................................................. ............................................................... 1 1 2.. Introduction........................................................................................................... 1 The MacroeconomicContextfor PublicExpenditureReforminZambia ......5 A. 5 B. ProgressinMacroeconomic andPoverty Indicators .................................. Public Expenditure. Poverty and Social Outcomes.................................. 11 14 D. C. KeyAreas for MaximizingImpact ofPublic Expenditure....................... The Scope for Increasing Government Revenue...................................... 16 E . SummaryandConclusions ....................................................................... 18 3. GovernanceandAccountabilityin Zambia ...................................................... 20 20 B. A. UnderlyingFactors for Trends inGovernance......................................... TrendsinGovernanceIndicators.............................................................. 21 C. The Accountability Framework inZambia............................................... 25 PART I1 PUBLIC EXPENDITURE MANAGEMENT AND FINANCIAL . ACCOUNTABILITY ......................................................................................... 36 4. BudgetPreparation ............................................................................................. 36 36 B. A. Comprehensivenessofthe Budget............................................................ Credibility o fthe Budget .......................................................................... 41 Budget Classification ................................................................................ 45 D. C. 49 E. Presentationo fthe Budget........................................................................ Links BetweenBudgeting andPlanning................................................... 55 5. BudgetExecutionandControl .......................................................................... 58 58 B. A. 65 Expenditure Control.................................................................................. Revenue Controls...................................................................................... CashRationing.......................................................................................... 66 D. C. 69 E. Payroll Controls ........................................................................................ Commitment Control andthe Arrears Problem........................................ F. 75 80 G. Capital Expenditure Controls.................................................................... Public Procurement ................................................................................... 83 6. Accounting,Reporting,andAudit .................................................................... 91 A. Legal Basis for Accounting andReporting............................................... 91 B. C. The Current System for Financial Management andReporting............... 92 Moving Towards Integrated Financial Management Information System 100 D. (IFMIS)................................................................................................... Internal Audit .......................................................................................... 104 -V- E. Effectiveness of Extemal Audit.............................................................. 106 PART I11 COMPLEMENTARY POLICY MEASURESAND ISSUES . ................111 7. The HumanResourceChallengesfor EffectiveBudgetManagement .........111 A. B. The Problem of Low Overall Pay........................................................... 115 C. The Fragmented Salary Structure andOveruse ofAllowance................112 116 Recommendations................................................................................... The Linkto Corruption........................................................................... The Impact on Quality of Services ......................................................... D 117 E .. 118 8 . Service Delivery in Practice: Leakages andEquity in the Flow of Fundingto EducationinZambia ........................................................................................ 125 A. 125 B. Expenditure Tracking.............................................................................. Study Methodology................................................................................. 127 C . 129 D. Leakages ................................................................................................. 132 Private Expenditures............................................................................... Equity...................................................................................................... 135 F. E. Discussion and Conclusions ................................................................... 137 PART I V SUMMARY AND A PROGRAMOF PRIORITYACTIONS . 9 . A Programof PriorityActions . . ..............139 ......................................................................... 139 A. MediumTerm Vision For Zambia......................................................... 139 B. 140 Progress Monitonng................................................................................ The Three Year Program ........................................................................ . . 146 D. C. Costing and Support by Development Partners ...................................... 147 REFERENCES .............................................................................................................. 148 ANNEX, Volume I ........................................................................................................ 150 -vi- ListofTables Table 1:Zambia: Structure o f Output and Real GDP Growth. 1970s-2000....................... 6 Table 2: Zambia: Macroeconomic Indicators. 1970s- 1990s (inpercent) .......................... 7 Table 3: Zambia: Poverty and Social Indicators................................................................. 9 Table 4: Zambia: Share o f Social Expenditures. 1996-2001 ............................................ Table 5: Zambia: Central Government Revenue. 1990-2000 .......................................... 12 15 Table 7: Public Perception about Government Performance........................................... Table 6: Zambia Governance Trends............................................................................... 21 22 Table 8: Illustrative Examples o f Use o f Supplementary Appropriations (Year 2000 budget) ...................................................................................................................... 29 Table 9: A Comparison o f Historical Spending with Approved Budget Estimates for Selected Line Items................................................................................................... 38 Table 10: Ministry o f Health (billions o f Kwacha): Budget line items through which personnel costs are paid............................................................................................ 46 Table 11: Cash Releases to Select Ministries for KeyExpenditure Categories during 1998 61 Table 12: Comparison o fBudget Releases and Actual Expenditures ............................. ................................................................................................................................... 97 Table 13: Staffing o f the OAG....................................................................................... 108 Table 14: Fundingo f the OAG for the period 1998 2001 ............................................ Table 15 :Scenarios Assumptions Matrix...................................................................... 108 122 Table 16: Enrollment, Urbanization, Wealth, and Distance.......................................... 133 Table 17: First-Year Priority Actions ............................................................................. 141 Table 18: Second-Year Priority Actions......................................................................... 143 Table 19: Third-Year Priority Actions............................................................................ 145 ListofFigures Figure 1:Zambia: Trends inHealthandEducation Indicators. 1990-2000 ..................... 10 Figure2: Index of Zambia's Real Per Capita GDP. 1975-2000..................................... 113 Figure 3: Indices o f Real Minimum Salaries for Various Civil Services Salary Grades. Figure4: Public Service EmploymentTrends. 1978-2001............................................. 1975-2000 ............................................................................................................... 114 120 Figure5: FundingFlows Diagram................................................................................. 127 Figure6: FundingDisbursementinZambianEducation............................................... 129 Figure7: Inequalities inPublic FundinginZambianEducation................................... 134 Figure8: EquityinPublic andPrivate Funding............................................................. 136 Figure9: L o r e n Curves ofPublic andPrivate Funding............................................... 137 Annex. Volume I Annex 1: Comparative Review o fthe Role o fParliament inBudget Passage.............. 150 Annex 2: Government BankingArrangements............................................................... 152 Annex 3: Areas for Improvingthe Cash Release System............................................... 154 Annex 4 Procurement: Short-term Sequenced Actions . ........................................................ 157 Annex 5. Procurement: Mediumto Long-term SequencedActions ........................................ 161 Annex 6.Key Issues inIFMIS Development ................................................................. Annex 7.Description o f a Selected Salary Enhancement Scheme (SASE) ...................163 167 -vii- ACKNOWLEDGEMENTS This report is prepared by the World Bank inclose collaborationwith the Government o f the Republic o f Zambia. The Government team was led by Mr. Chitundu Mwango, the Accountant General, Ministry o f Finance and National Planning (MFNP). Mr. James Mulungushi, Director o f Planning, MFNP, provided crucial leadership at the early stage. The Government team provided two rounds o f detailed comments on the first draft o fthe report, organized a two-day donor workshop in Lusaka, and made presentations o f the draft report at the workshop. Duringthe workshop, usefil comments were received from several donors, which were subsequently integrated into the report. Messrs. Patrick Kalifungwa and Felix Mutati, both Deputy Ministers o f Finance and National Planning, provided guidance and support to the joint Government- Bank team duringthe workshop. Subsequent formal Government comments on the green cover o f the report were received and integratedinOctober 2003. The filllist ofthe Government team appears below. From the World Bank side, the report was prepared by a team led by HinhDinh (Lead Economist) and consisted o f Abebe Adugna (Task Manager and Coordinating Author), Bernard Myers (Principal Author), Mushiba Nyamazana (Government and donor coordinator), Iraj Talai (Country Financial Accountability Assessment Team Coordinator), Subhash Dhingra (Country Procurement Assessment Review Team Coordinator), Harry Garnett (Public Sector Reform Program Coordinator), Jishnu Das (Public Expenditure Tracking Survey Coordinator), Stephano Fassina (IMF), Edward Olowo-Okere, Nicola Smithers (PEFA Secretariat), Ronald Quist (EU Consultant), Jesse Hughes(Consultant), Fenwick Chitalu and Bwalya Mumba. The team visited Zambiain 2001 and 2002. Financial assistance was provided by the Japanese PHRD, the United Kingdom's Department for International Development (DFID) and the EU. The reviewers o f the report were Michael Stevens, Anand Rajaram, and David Shand. Substantive contributions and helpful comments were received from Bruce Jones, Delfin Go, Anthony Hegarty, Richard Allen, and Jos Verbeek. Michael Stevens provided detailed comments and sustained advice to the team on the adequacy o f the legal framework and the institutional structures for accountability. Anand Rajaram also assisted the team with prioritization and sequencing o f the many recommendations o f the report. The IMF team on Zambia led by Robert Sharer and subsequently by David Andrews providedvaluable inputsand comments. Philippe L e Houerou, Yaw Ansu, HartwigSchafer, Laurence Clarke, and Ohene Nyanin provided guidance and support to the team. Mr. L e Houerou guided the team throughout the preparation o f the report, and provided substantive and valuable comments on all aspects o f the report. Dotilda Sidibe provided assistance in document preparation, and Zoe Vantzos assisted in proof-reading. The support o f the Zambia country team, particularly o fthe Lusaka Country Office, i s gratefully acknowledged. The report benefited from active participation o f Zambian Government officials and o f donor representatives. A partial list o f names appears below. -viii- List of GovernmentTeam C.N. Mwango Accountant General (Head o f Government PEMFARCounterpart Team) S. Musokotwane Acting Secretary to the Treasury (formerly PRSP Advisor, MFNP) R.M.KMulenga Deputy Accountant General, MFNP L.Zulu Controller, Internal, MFNP D.Kunda Project Manager, IFMIS, CCSD S. Chungu s.Zulu Director, Budget Office, MFNP Assistant Director, Planning and Economic ManagementDepartment A. Mwakajoka Comm. Finance, ZRA J. KNjolomba Director, Inspection and Standards, National Tender Board M.Chigaga Deputy Director, MOLA R.M.Mwambwa Director, Auditor General M.S. Mwiinga Economist, Bank o f Zambia P.K. Sibitwane Agriculture ChiefAccountant, OAG F.Shandavu Chief Planning, Education dJ. Mulungushi Director, Planning and Economic Management Department G. B.Lintini Director, Investmentand Debt Management Department P.N. Mwangala Director, Budget Office F.M.Siame Former Auditor General M.Philips Advisor, Planning and Economic Department L.Zimba Permanent Secretary, Manpower DevelopmentDivision, Cabinet Office A. Chifungula Auditor General (formerly Acting Secretary to the Treasury, Permanent Secretary ,FMA -MFNPand Controller o f Internal Audit) K.Muleya Director General, Zambia National Tender Board K.S. Chipako Manager, Centralized Computer Services Department (CCSD) E.C. Simukoko Principal Economist, MFNP C.G.Kaluba Permanent Secretary, PSMD, Cabinet Office D.Diangamo Formerly Acting Secretary to the Treasury, MFNP A. Sakala Systems Development Manager, Data Management Center (CCSD) J. Ukwimi ChiefAccountant, Ministryo f Works and Supply D.Lungu Senior Accountant, Ministry o f Works and Supply FrancisMbewe Co-ordinator, World Bank Programs, MFNP M.W. Saviye-Kabaza Secretary, Accountant General, MFNP B.B.Mutemi Secretary, IFMIS Project Listof Donors H.Mealins, K.Wood, M.Mumbwatasai DFID,Zambia P.Rundel1, S. Olander, R.Quist EC, Brussels A. Konard Embassy o f Germany, Zambia Ms.Masako, Mr.Kirao JICA/ JBIC, Zambia C. Kuhnel Embassy o f Sweden, Zambia/ SIDA A. Berglund, E.Von Pistohlkors EU, Delegation, Zambia M.Bergschneider GTZ, Zambia M.Ellyne,L.Mwansa IMF,Resident Representative, Zambia J. Kinsella Ireland Aid, Zambia M.Knudsen Royal DanishEmbassy M.Schuurman Royal Netherlands Embassy, Zambia -ix- A. Olsen, L.Lindback RoyalNorwegianEmbassy,Zambia V. Robinson UnitedNationsDevelopmentProgram(UNDP) L.Mukumbuta USAID,Zambia -X- EXECUTIVESUMMARY I. Overview 1. The challenges faced by Zambia in public expenditure management have been longstanding and will require targeted efforts as well as a strong degree o f political will to address. The recently launched constitutional review which includes issues o f public finance, the anti-corruption campaign o f the new Government, and the renewed interest by Parliament in governance issues and accountability have all been encouraging steps. Nevertheless, for Zambia to assure that public accountability is enduring and not dependent uponthe Government o fthe day, it must take steps to strengthen institutions o f the State that can provide public oversight, and that promote basic checks and balances. Despite a legal framework based on the Westminster tradition o f government, many o f Zambia's public finance laws and regulations are not enforced. This has led to a breakdown o f administrative systems and procedures for the control o f expenditure. Audit systems that are inplace to detect improprieties have beenrendered ineffective due to the lack o f follow-up and enforcement o f the Auditor General's recommendations. Unfortunately, the lack o f sanctions may have only served to embolden those who have been engaged in improper activities. Moreover, a lack o f transparency in budget management - at least partly caused by the weakness in financial reporting systems - have impeded the public from fully understanding the economic trade-offs that are required o f Zambia today. Only through greater disclosure o f information on resources and liabilities o f the State will it be possible for Zambians to develop credible strategies for tackling the economic challenges ahead. 2. This report provides a very detailed analysis o f the public expenditure management and accountability processes inZambia. Yet, many o f the recommendations are not new but have been cited in previous reports o f the World Bank and/or other donors. Effective implementation o f public sector reforms will likely remain a challenge inZambia. The limited capacity of Government suggests the need to target a few major aspects o fpublic finance and to address them persistently: 0 Improvingcompliancewith existingregulations 0 Strengthening the oversight institutions o f the State 0 Promoting public access to information, and 0 Rebuildinginformation management and reporting systems. -xi- Macroeconomic and Governance Context 3. Zambia's challenges faced in public expenditure management and financial accountability should be situated inthe broader context o f macroeconomic performance, poverty, social outcomes, and governance. 4. From a macroeconomic perspective, Zambia's growth record has been one o f protracted decline. The GDP growth rate fell fi-om an average o f 1.5 percent inthe 1970s to 1.4 percent in the 1980s and 0.3 percent in the 1990s. Coupled with the steady increase in inflation and population growth, Zambia's per capita income has dropped by 40 percent in real terms since the 1960s. In addition, almost all indicators o f social and poverty condition deteriorated duringthe 1990s. 5. Public expenditure is key to improving poverty and social outcomes. Zambia's tax revenue, expressed as a percentage o f GDP, has remained stable at around 18-20 percent since the early 1990s. However, due to a relatively highinflation rate, the social sector spending declined inreal terms inthe 1990s. Looking forward, without significant growth, the scope for increasingthe size o f government revenue (and hence expenditures) remains quite limited. Thus improving the efficiency and effectiveness o f existing public expenditure remains critical to making a significant dent on poverty. Indeed, without better budget management, accountability and greater value-for-money, the impact on poverty will remain limitedat best. 6. From a governance perspective, the quality o f governance has declined in Zambia on many fronts during the late 1990's. Looking across the dimensions o f governance, Zambia's worse scores have been in "Control o f Corruption" and "Government Effectiveness" respectively (World Bank 2000/01: Governance Research Indicator Country Snapshot, 11). On a scale from 0 to 100 (with 100 being the best), Zambia's percentile rank in "Control o f Corruption" was only 19.9. In "Govemment Effectiveness", it was only 26.9. These statistics serve to highlight the importance o f improvedpublic expenditure management and financial accountability. 7. In its Poverty Reduction Strategy Paper (PRSP), the Government of the Republic o f Zambia (GRZ) recognizes that good governance goes hand in hand with effective poverty reduction. The PRSP spells out three objectives in the area o f governance: (i)to enhance the normal checks and balances among branches of government, especially by empowering the legislative and the judiciary in preventing corruption, and to empower local and traditional authorities and strengthen institutional linkages between the local authority and the community; (ii)to ensure efficient, equitable, and transparent management o f public resources; and (iii) and guarantee justice for all by advancing human rights, improving the efficiency o f the legal system, and implementingconstitutional andlegalreforms. 8. This report deals with the second objective of the PRSP, i.e., with ways and methods by which the Government can ensure efficient, equitable and transparent management o f public resources. It also focuses on the dimension o f governance often referred to as "administrative" governance, i.e., the effectiveness o f government to be -xii- able to provide public services. Proper implementation o f the key findings and recommendations o f this report can thus help improve "administrative" govemance, a dimension o f governance in which Zambia's performance has further deteriorated in recent years. The specific objectives o f the report are to: (a) provide comprehensive and integrated assessment o f Zambia's overall fiduciary risk, Le., budget management, financial systems and auditing, and public procurement; (b) document PEM reforms progress to-date and challenges facing Zambia; and (c) develop a realistic action plan outlining short and medium term remedial measures which the Government should implement with donor support. 11. MainIssues The LegalandInstitutionalFrameworkfor Accountability 9. One of the most important issues facing Zambia is the inconsistency between the legal framework for public finance and the actual structures and incentives in place to assure accountability. First, this i s reflected in the fact that the Constitution seems to grant an excessive amount o f discretion or authority to the Ministry o f Finance and National Planning (MFNP). Typically, in a system with democratic control of public expenditure, Parliament approves the budget presented by the Executive as well as any major changes that are needed during the budget year. In contrast, under the current financial provisions in the Zambian Constitution, the Ministry o f Finance i s granted the power to make sweeping changes in the budget allocations without any prior consent o f Parliament. These flexibilities given to the Minister o f Finance could easily be misused inthe absence o f a proper Parliamentary oversight. A review o f other Constitutions and Finance Acts in Anglophone Africa provides important insights on how Zambia might achieve a better balance. This report gives a summary o f some o f the legal provisions that could be improved. 10. Secondly, even where the legal provisions are appropriate, they hinge upon a system of enforcement and accountability that has not functioned in practice. The current public finance rules are based on an assumption that Ministry o f Finance plays a primary role o f stewardship or guardianship over the public purse and that it enforces an adherence to the rules by line ministries. However, Ministry o f Finance has not held Controlling Officers (CO) accountable for their management o f public resources, nor has the Parliament generally been in a position to hold the Executive accountable. As this report will show, without these checks and balances, the system simply breaks down. A number o f steps are described that could enhance Parliament's capacity to provide oversight. BudgetPreparation 11. First and foremost Zambia must improve the credibility o f the budget. Lack o f effective and realistic budget preparation has follow-on implications for budget execution, procurement, and ultimately for service delivery. For years now, budgets have been approved that are structurally over-committed and that lack the resources to meet the on-going obligations of the government. The most obvious result has been that the -x111- ... structure of actual expenditures at the end of the year bears little relationship with the original budget estimates. Consequently, most line ministries realize that the budget i s not a reliable basis on which to plan activities since actual releases are unpredictable and vary substantially from the amounts indicated inthe budget. 12. Government budgets have largely reflected only incremental changes from the prior year, without tackling the kindo f substantive policy decisions needed to bringcosts inline with resources. Evenwhen the budget appears to be inbalance on an aggregate basis - with sufficient revenues and financing to cover expenses - there are significant imbalances that exist on a ministry-by-ministrylevel, i.e., budgets for some ministry activities are insufficient to cover the actual costs. As MFNP acknowledges, some activities have essentially fixed costs associated with them, which have to be met, whether the activities are fully funded inthe budget document or not. Ultimately, these activities receive a supplemental budget, which squeezes out funding for other activities that had been previously budgeted. Going forward, Zambia should (i) reconcile the large differences between historical cost and proposed budget estimates, and (ii) strictly limit the use o f supplementaryappropriations. 13. In addition to the need to improve credibility, Zambia's budget preparation still requires substantial improvement in several additional areas: (i) comprehensiveness, (ii) budget classification, (iii) presentation o f the budget document itself, (iv) the link the betweenbudgeting, policy, and planning, and (v) timingo fthe budgetcycle itself. 14. First, comprehensiveness should be addressed because Zambia's budget is confined to central government and lacks information on state enterprises, pension funds, special funds, quasi-governmental activities, and local government expenditures. This presents a misleading picture o f the financial position o f the State, especially due to the heavy losses and contingent liabilities coming from state enterprises and public sector pension funds. 15. Secondly, the budget classification should be improved to make it easier to discern accurately the policy priorities o f the government. Lacking any functional or programmatic classification, it is difficult to identify poverty-reducing expenditures inthe budget, except in very broad terms. In addition, the economic classification often hides the true nature of expenditures. For example, personnel-related costs are spread across several categories. The Government's attempts to pilot Activity Based Budgeting (ABB) are promising, but have born limited fruit to date. 16. Third, budget presentation should be enhanced because the poor format and quality o fthe Government's budget document has a direct bearing on Parliament's ability to provide effective oversight, and for other stakeholders to understand the budget. An immediate objective should be to include expenditure data from the prior year, alongside with an estimate of the current year to enable comparison with the proposed budget estimates. Other needs to consider are: (i) information on the status o f capital projects; (ii) summarytables; and(iii) narrativeonthemajorchangesinthebudget more greater and the reasons for them. -xiv- 17. Finally, the Government should ensure that new government activities are not initiated without a realistic and comprehensive multi-year costing, and more importantly, without an explicit indication o f where the additional resources or off-setting savings would come from. The link between policy, planning, and budgeting has been very weak to date. Poor communication from donors on possible resource inflows compounds the Government's budget andpolicy planning. The problems o f credibility and structural over-commitment o f the budget underline the importance o f integrating some type o f medium term planning horizon and decision-making into the annual budget process. While a fully functioning Medium Term Expenditure Framework (MTEF) may realistically be more a medium term objective in Zambia, fundamental steps would need to be taken immediately to establish a foundation for multi-yearbudgetplanning. Budget Execution and Control 18. Improvements in budget execution must address multiple dimensions o f the problem: (i) abolishing cash rationing, (ii)enforcing expenditure controls, (iii) properly recording commitments, especially for capital projects, (iv) updating payroll lists systematically, and (v) reforming procurementpractices andregulations. 19. Improvingbudget execution fundamentally means ending Zambia's reliance upon "cash rationing'' to determine resource allocation and restoring the proper role o f the budget. Although an understandable response to macroeconomic crisis at the time o f its creation in 1993, the cash budgeting process has persisted until today in ways that are highly destabilizing to effective public expenditure management. One o f the major consequences o f the cash budget i s that actual expenditures by line ministries bear little relationship to the budget initially approved by Parliament. Moreover, because o f the unpredictable nature o f the cash releases from month to month, line ministries have had little means by which to plan basic operational and management needs. This has an inevitable impact on procurement practices and encourages an accumulation o f supplier credits and higher prices charged to Government. 20. Enforcement o f existing financial regulations is essential to addressing the various weaknesses in expenditure control. First, many spending departments in Zambia operate as if standing instructions on the cash control mechanism did not exist. Inmany cases, orders are placed and payment vouchers are prepared and kept (often unrecorded) until availability o f cash. Second, in some cases, checks are issued before delivery o f goods, and presented to the procurement staff (or stores) for exchange against goods. Third, overdrafts are used by the spending agencies to pay for expenditures over and above the budget. Fourth, the present control mechanisms are largely designed for a manualsystem based on segregation o f duty, and discipline to apply the rules. However, changes have occurred and made the controls ineffective, among which is the increased volume o ftransactions. Furthermore, although manuals and guidelines exist, staff are not well trained, and do not have copies for reference. 21. Ineffective control over commitments has led to a chronic build-up of arrears, which in turn helps to undermine the basic integrity and value o f the annual budget. Although Financial Regulations prohibit line ministries from committing more than they -xv- have resources to pay, the line ministries continue to do so. Though quarterly commitment reports prepared by the Accountant General are supposed to be used to identify over-committing ministries, sanctions have been rarely forthcoming. The worst problems with arrears lie inthe area o f capital projects. 22. Ineffective payroll controls create a significant area o f potential financial risk because o f the high share o f payrolls in the budget and the lack o f effective controls for updating the payroll list. In 2001, 37.5 percent o f budget releases were for personal emoluments (K907 billion). Some degree o f loss may be introduced by the use o f cash payments in districts that are outside the banking network. Yet, o f bigger concern in Zambia are the weak procedures for managing and controlling the payroll database. One o f the ways "ghosts" may be introduced i s through inefficient recording o f personnel separations so that payments continue long after the person has ended hisher services with the Government. Currently, the payroll management system may be described as passive, in that the management system awaits notification in the case of employee separation or transfer. Reports that can be generated on who has been paid each month do not reachthose who can most effectively exercise control. 23. Arrears on capital projects have been amongst the most pervasive and costly, and will require a much deeper linkage between commitment monitoring and cash management. Although only a small portion o f the total budget, losses on domestically- financed capital projects can be quite significant. Some have estimated that capital projects in Zambia may be costing twice as much as they should. Two central problems faced with capital projects are: (i)insufficient budget releases for the Ministry o f Works and Supply (MWS) to cover the obligations the Government has incurred; and (ii) lack o f comprehensive and transparent accounting, which inhibits a carefil prioritization o f cash needs. Moreover, the structure of the contracts i s such that if not managed aggressively the overall costs can skyrocket. 24. Another area where Zambia needs to do better is in public procurement. The total value o f public procurement in 2001 was estimated at USD 485 million, equivalent to 15 percent o f Zambia's GDP. Even at a conservative estimate o f 10 percent saving through better procurement policies, practices and institutional arrangements, this would mean a yearly gain to the Treasury of the order o f USD5O million. Development o f a well-functioning procurement system, based on transparency, competition, economy and efficiency, fairness, and accountability, thus becomes critical. 25. Several factors limit effective public procurement in Zambia. First, a gap between intentions and practice pervades various aspects o f Zambia's public procurement. Despite a legal and institutional framework, weaknesses in its structure and content allow undesirable practices and procedures. While the Zambia National Tender Board (ZNTB) i s expected to enforce procurement rules, in practice it i s liberal in permitting exceptions and allowing negotiations to replace clear procurement guidelines. Current practices that foster corruption and higherprices include usingnegotiations as an -xvi- accepted procurement method and misusing the registration system for purchases from short-listed firms. Secondly, the legal framework' lacks robustness and features structural, and content, inadequacies. Basic principles o f procurement are not captured comprehensively. Further, aspects that should be in the Zambia National Tender Board Act are in the regulations or guidelines and vice versa. Third, procurement management is weak. In the few departments where they do exist, procurement files are often incomplete, and procurement planning as a tool for conducting efficient and economic procurement i s largely non-existent. Fourth, there has been no progress in setting up a procurement cadre, despite the fact that over 600 people have been trained at different levels. Lastly, activities o f the Anti-Conuption Commission and the Office o f the Auditor General (OAG) are critical ifZambia i s to achieve a well-functioning, fair, and transparentprocurement framework. Accounting,ReportingandAudit 26. The current systems used by MFNP for financial management are either manual or rely on outdated technology with limited applications. Budget preparation, budget execution, and payroll management are all separate and independent applications within MFNP that do not interact with each other. Inline ministries, the situation is generally worse, with reliance upon basic spreadsheet applications where available. There are few personal computers available within MFNP and many o f those are also outdated. The accounting hnction is heavily reliant upon manual record-keeping, especially in the line ministries. Line ministries and provinces submit their monthly expenditures reports on diskette to MFNP with the help o f a computer-based Financial Management System (FMS). The FMS records the details o f transactions after the payments have beenmade. 27. The current systems produce a number o f serious problems that must be addressed. First, expenditure reports are prone to errors and inconsistencies that are difficult to find and to reconcile. Second, reporting tends to be significantly delayed because o f the various levels o f compilation required. Third, the focus on bookkeeping tasks distracts accounting staff from more meaningful analysis o f the data and trends. Fourth, the delays and inconsistencies greatly diminish the ability o f MFNP to provide sufficient oversight and control over the nature o f the expenditures. Because o f these inadequacies, the MFNP has found it difficult to provide an accurate, complete and transparent account of its financial position to Parliament and other stakeholders, includingthe general public and donors. 28. Compounding matters, a variety of shortcuts and deviations from accountingand reportingregulations have developed over time that harmthe accuracy and reliability o f ledgers for recording commitments. The shortcuts may originate from the shortages in staffing, the absence of training, office reorganizations, and the increase in the sheer 'The legal framework i s mainly based o n the 1982 Zambia National Tender Board Act (ZTBA), which has limited scope and unclear enforcement mechanisms. It provides for the establishment o f the Zambia National Tender Board (ZNTB). -xvii- volume o f (largely manual) transactions. Quality o f monthly outturn reporting by line ministries remains uneven although it has improved somewhat recently. The audited Financial Report for FY 2000 was the first delivered to Parliament on time in about a decade, but significant problems remain that could affect the quality o f the information contained in the Financial Report. For example, key information appears to be missing from bank reconciliation statements; ledgers are improperly kept; intemal control mechanisms are sometimes missing form reorganized agencies; quarterly consolidated reports are subject to double counting; and the role o f the Accountant General i s not adequately defined. 29. Though not a panacea, an Integrated Financial Management Information System (IFMIS) could help improve reliability and timeliness o f financial data, strengthen financial controls, improve data comprehensiveness, and generally enhance transparency o f budget execution. Development o f an IFMIS has been progressing steadily, but IFMIS i s only a tool and not a substituteto implementingthe basic measures neededfor improving budget accounting and financial reporting. Indeed, IFMIS should bepursuedas complementary to, and inparallelwith, such basic measures. 30. The Internal Audit department o f the MFNP has a critical role to play in maintaining the pre-audit controls on expenditure, as well as in assessing the overall adequacy o f the financial management systems and procedures inplace. InZambia, the Internal Auditors have been effective in diagnosing problems in the financial management procedures, but their potential impact is still severely limited by the lack o f adequate humanand financial resources to carry out their work and insufficient follow-up on their recommendations by Controlling Officers. 3 1. Although External Audit has benefited from various training opportunities in recent years, it still remains weak in many respects. While the Auditor General (AG) himself enjoys a constitutional backing and adequate independence, in practice his independence is hampered by limitations in financing imposed by MFNP and staffing imposed by civil service regulations. Audit planning and methods are, inprinciple, those recommended by the Intemational Organization o f Supreme Audit Institutions (INTOSAI) but the degree o f compliance is low. The audits consist mainly o f financial and regularityaudit, which are often limitedinscope. The audit reports are inconclusive and untimely. The Office o f the Auditor General (OAG) i s operating with severe staffing problems, and has one professionally qualified auditor, although many are at different levels for passing their qualifications. 111. ComplementaryPolicyIssues 32. Human resources and pay reform Even with excellent policies and procedures, sound financial management would be difficult to achieve without properly motivated and skilled staff. Low remuneration in the public sector is a major factor contributing to the problem o f poor productivity, motivation and recruitment and retention. At a time when Government is seeking efficiency improvements, in part by reducingthe size o fthe civil service, there exist significant staff shortages inawide range -xviii- o f professional and technical jobs owing to poor pay compared to that available in the private sector andwithin regional labour markets. 33. The real wages for civil servants has declined since the mid-1970s. While this decline to some degree mirrors the overall macroeconomic decline o f the country, real public service pay declined much faster than did real GDP per capita from 1975 to 2000. Adding to this problem o f low and declining pay is a poorly constructed internal grading structure and an increments system that i s based on the length o f service rather than performance, which stifles initiative and motivation. Furthermore, the fragmentation in job grade scales and the growing dependence upon allowances to enhance compensation has severely weakened horizontal andvertical equity. As a result o f implicit Government pay policy, fringe benefit and monetary allowances have progressively been used as a major vehicle for increasingcompensation, particularly for upper-middle level and senior civil servants. 34. The problem o f low pay and fragmentation notwithstanding, the wage bill in Zambia has remained large relative to overall government recurrent expenditures, thereby crowding-out operational expenditures. The challenge for Zambia is therefore how to design and implement a pay reform strategy that is consistent with the macroeconomic goal o f containing the size o f wage bill (as a proportion o f GDP). The characteristics of such a pay reform strategy would include the following: (i) consistent with an being affordable resource envelope; (ii)controlling the size and growth o f government employment; (iii)eliminating the distortions and anomalies, and paying salaries commensurate with skills, experience and responsibilities; (iv) progressively consolidating all allowances into the basic pay; (v) progressively decompressing the pay o f senior and middle ranking staff; and (vi) designing and implementing a new performancemanagement system and employment policy. 35. Pressure to increase salaries in2003 may have led to an overshooting o f the wage bill target agreed with the International Monetary Fund (IMF). This will make it necessary for Government to find measures to reduce the wage bill inthe near future, and to develop strategies to contain it adequately in the medium term. The short-term measures will include completing the payroll data cleaning exercise and retrenchment packages to off-set the cost o f the 2003 salary increases. The medium term actions may include restructuring and rightsizing the public sector to control the size o f the wage bill, and its structure and composition. However, this will need to be complemented by further decompression and consolidation o f allowances within the feasible budget envelope to be able to attract as well as keep skilled and qualified personnel in the civil service. 36. Finally, improvements in compensation levels can be only part o f a package to change civil servant behavior. Other complementary measures are essential to reduce deviant or corrupt practices. To the extent that corrupt behavior i s a response to opportunity, increasing the official pay o f civil servants raises the cost o f getting and possibly losing one's job. But unless actions are taken to punish corrupt behavior, pay increases alone are likely to be insufficient to change the costhenefit calculation and/or to bring an end to corrupt practices inthe public sector. -xix- 37. Leakages and equity of public expenditure in education. An Education Service Delivery Survey (ESDS) was recently completed in Zambia to track the flow o f funding in the administrative hierarchy, assess the extent o f leakages in the system, and evaluate the distributional and equity consequences o f such public expenditure in the education sector. The findings o f the survey have important implications inthat ifbudget resources cannot reach their intended beneficiaries, then higher spending will have little effect on social or economic outcomes. Furthermore, the findings could be o f broader relevance in informing the kinds o f institutional reforms that the Government needs to carry out at lower levels of government in addition to addressing the central government problems discussed throughout this report. 38. The flow o f public funding to schools in Zambia is organized through a three- tiered administrative hierarchy involving provincial offices, district offices and schook2 For the purposes o f this analysis, funds were classified into four different categories: rule-based allocations to schools, discretionary allocations funds to schools, rule-based allocations to teachers, and discretionary allocations to teachers. 39. The tracking exercise focused on all non-salary funding flows through the administrative hierarchy. It found that on average about K28,OOO per pupil enters the educational system for the four provinces o f the ESDS (Eastern, Northern, Lusaka and Copperbelt provinces), although there i s a significant degree o f variation within these provinces. O f this amount, discretionary funds at the level o f the province and district account for 70 percent o f all fundingwhile rule-based funds account for the remaining 30 percent. 40. The main conclusions o f the education study are: 0 Only between one-sixth and one-third o f total funding in the system eventually reaches schools: O f this total amount, between 14 percent (Lusaka) and 34 percent (Copperbelt) eventually reaches the schools as a combination o f rule-based and discretionary spending while the rest i s spent at the provincial anddistrict levels. 0 "Decentralization" has shifted spending from the provincial to the district level, but has not resulted in greater disbursements to schools. The extra funding that reaches the districts in decentralized provinces just results in higher spending at the district level and is not associated with greater funding to schools. 0 The specific rule usedinthe case of rule-based allocations (a fixed amount per school) has led to greater per-pupilhnding for poorer andmore rural schools. However, these allocations are the only progressive disbursements in the 'Two o f the provinces in the survey are considered "decentralized" because the contain district education boards, while the other two are referred to as "centralized" because they do not contain district education boards. -xx- ESDS. Once all sources o f public funding are factored in, public school funding inZambia is regressive with almost 30 percent higher allocations to richer schools. IV. A Programof PriorityActions 41. The analysis in this report clearly shows that Zambia must significantly improve its public expenditure management and financial accountability in order for public expenditures to have any significant impact on poverty. The Highly Indebted Poor Countries (HIPC) expenditure tracking assessment and action plan o f 2001 showed that Zambia met only 3 out o f 15 public expenditure management benchmarks3 - one o f the lowest o f any HIPC country. The country must therefore find strategies to restore a public expenditure system that has lost credibility among both internal and external stakeholders. Because many o f the problems are inter-related, the problem calls for a comprehensive approach to reform o f the public finance system, including the laws and institutions that underpin public accountability. Yet, in view o f the country's limited institutional capacity, scarce resources, and the overall socio-political environment, the implementation o f these reforms must be sequenced carefully. To that end, the table at the end o fthis section summarizes a three-year program o fpriority actions for Zambia. 42. The table contains numerous measures o f both policy and technical nature. However, Zambia's history has shown that public sector reforms that rest solely at a technical level have only a short-lived impact, if any. Therefore, implementation of an effective reform agenda will require strong and sustained political support from the highest levels o f Government. Consequently, the most important recommendations in this report are not the technical ones, but rather those that require a policy decision to be implemented. Below are six recommendations which stand out from the others in the three-year action matrix. e Revise the legal framework for jinancial management. Supplemental appropriations should not be legally permitted without the prior approval o f Parliament. Emergency expenditures and excess expenditures should also be much more strictly limited. The Financial statements o f the government should be submitted to Parliament within 6 months o fthe end o fthe fiscal year. e Strengthen Parliament's oversight over the budget. Submit the macroeconomic framework and functional priorities to Parliament earlier in the budget calendar, submit regular reports on budget execution for information (e.g., semi-annually), and ensure appropriate follow-up on the Auditor General's reports. The benchmarks met by Zambia were the following: (i) includes capital and current expenditure "Budget finance by donors"; (ii) and banking reconciliation undertaken routinely"; and, (iii) "Fiscal "Closure o f the accounts occurs within two months after the end of the fiscal year". -xxi- Improve the presentation of the budget. Present to Parliament a budget that clearly compares the proposed budget with the actual expenditures (not budget) from the preceding two years. This should facilitate an informed debate on the realism o fthe budget and the Government's policypriorities. Strictly limit the use of supplemental appropriations. Prevent ministries fi-om relying upon supplemental appropriations, e.g., limiting the total amount o f supplemental spendingto about 5-10% o fbudget. Establish quarterly expenditure ceilings for line ministries, back ceilings with actual cash releases, and sanction controlling officers who over-commit. MFNP must first develop and enforce credible expenditure ceilings. Controlling officers who persistentlyover commit should be sanctioned. Ensure independence of the Office of the Auditor General by creating an Audit Board (oversight body) that reports to Parliament. The OAG's budget and staffing shouldbe determinedbythe Audit Boardrather thanbythe Executive. 43. Implementation o f the measures above forms the foundation for the larger reform agenda. Moreover, because they generally require only a policy decision by Government to implement (rather than extensive technical assistance), Government can address them immediately. They are essential to the success o f the public expenditure management system, and therefore, should take precedence over technical recommendations contained inthe report. Conversely, without these hndamental policy measures, the longer term impact o f any technical interventions is likely to be limited at best. 44. Implementation o f some o f the technical recommendations i s already underway (e.g. ABB, MTEF, and IFMIS). However, others may require technical and financial assistance from development partners. Donor support for the first year program could be earmarked for specific actions. Provided measures in the first year have been implemented successfully, support for the second and third years could be inthe form o f a common donor hndto be channeled directly to the budget. 45. This program o f priority reforms represents the bulk of Public Expenditure Management and Financial Accountability (PEMFA) reform agenda for the next three years. However, hrther actions going beyond this three-year horizon may be needed, especially inthe areas o f assuring accountability inthe use o f public funds, procurement, andimplementationo fthe MTEF andIFMIS. -xxii- A Three-Year Program of Priority Actions RecommendedAction Year Institutional 3 Responsibility Revise the legal framework for financial management to require that: a) the budget be submitted to Parliament with sufficient time [e.g., 60 days] to review and enact itbefore the start o fthe fiscal year, b) Parliament approve any supplementary estimates before the spending occurs, c) emergency expenditures be made through a voted Contingency Fundonly, TBD d) the amount o f time allowed for submitting an excess (unauthorized) expenditure bill be reduced, and e) the financial statements be submitted to Parliament within 6 months o f the end o f the fiscal year. Strengthen Parliamentary Oversight over spending by: (a) improving the budget presentation (as described MFNP, below under budget preparation), Parliament (b) presenting to Parliament the macro framework and functional priorities o f the budget earlier in the budget cycle, (c) assuring that Parliament provides adequate funding for the Budget Estimates and Public Accounts committees, (d) submittingregular (e.g. semi-annual) reports on the execution o f the budget accompaniedby any X requests for supplementary appropriations, (e) developing arrangements to ensure government response to P A C reports X Budget Preparation Reconcile the large variances between budget estimates ~ and actual expenditures for each budget head and subhead, and assure that the proposed budget reflects MFNPIBudget either the higher cost or specific policy action taken to Office reduce future services. Develop a multi-year planning framework (or MTEF) by building on the recent budget framework paper to (a) link national development priorities (TNDPFRSP) to annual MFNP, budgets, (b) to reconcile the differences between budget x Cabinet, estimates and actual program costs., and (c) to engage Parliament Cabinet and Parliament earlier inthe budget calendar. Reduce the amount o f supplementary appropriations, so that they are used only in exceptional circumstances [e.g., MFNP 5% o f the original budget]. -xxiii- RecommendedAction Institutional Responsibility revenues and expenditures o f BOZ, state enterprises, pension MFNP MFNP MFNP MFNP, Parliament M F N P MFNP MFNP an assuran MFNPDudget Office MFNPiAccG sanctions incases o f persistentnon-compliance. Complete and maintain the domestic arrears database, and 1 develop a plan for liquidating past arrears. A MFNP Update payrolls more effectively by having the senior pay point officer check and sign o f f on the payroll lists each month and return them to the provincial headquarters. MFNP These verified payroll lists should serve as the basis for updating the payrollmaster file. Develop unique codes for each post in the establishment registry and use them to control unauthorized additions to PSMD the payroll database. -xxiv- RecommendedAction Year Year Year Institutional 1 2 3 Responsibility Institute a reporting mechanism between MFNP and M W S that provides detailed financial information on each capital project including the financial terms o f the contract, work variations, price fluctuations, interest X M F " , M W S penalties, etc., inorder to inform cash release decisions by MFNP. Assure that recurrent costs o f capital projects are adequately planned for inthe MTEF, as a pre-condition to X MFNP MFNP ZNTB MFNP, ZNTB X ZNTB. MFNP,ZNTB Revise the legal framework. X MFNP,ZNTB Introduce an improved registration list system. Improve X procurement practices, through greater restriction o f ZNTB I negotiation and informing tenders. Improve business practice through development o f a code o f ethics, training, and development o f competition and MFNP,ZNTB fair trade regulation Improve trade practices through increased capacity o f Modify the accounting procedures in order to capture commitments and arrears, and to highlightinterest charges X I MFNP and penalties o n capitalprojects. Clarify the role and responsibilities o f the Accountant General in legislation, and have the office report directly X to the Secretary to the Treasury. I MFNP =E- Ensure that LPOs and contracts are entered into the F M S at the time the commitment i s entered into and not at X MFNP ~ payment stage. Reconcile expenditure retums with budget releases and banking formation for each ministryto verify compliance X with financial regulations. Submit audited financial statements to Parliament within 6 months o f the end o f the fiscal year. X Procure hardware and software for the IFMIS, pilot test, evaluate. Roll out IFMIS to all ministries andtrain users. x X I MFNP Increase the capacity o f Intemal Audit to execute its annual workplans through improved staffing, greater operating resources, and better coordination with the X Office o f the Auditor General (OAG). Workplans should include materiality audits o f capital projects. -xxv- RecommendedAction IYear Year Year Institutional 1 1 2 1 3 IResponsibility Government has most difficulty in attracting and -xxvi- PARTI.INTRODUCTION AND CONTEXT 1. INTRODUCTION 1.1 Over the last few years, Zambia has undertaken a number of fiscal reforms aimed at improving public expenditure and financial management. The international community, including the World Bank, IMF, AfDB, EU, and bilateral donors, has supported these reforms. While sizable progress has been made recently, considerable weaknesses remain in the areas o f budget management, financial reporting and audit, public procurement, and financial accountability. 1.2 The Bank's Public Service Capacity Building Project (PSCAP) and Fiscal Sustainability Credit 1 (FSC1) were among the key instruments focused on helping the Government o f Zambia in improving public expenditure management and financial accountability over the last few years. 1.3 PSCAP became effective in October 2000. The project's objective i s to make public service delivery processes more effective and efficient inorder to facilitate growth and poverty reduction. During its Phase Iimplementation (October 2000 - June 2002), PSCAP helped achieve numerous positive developments in terms o f improving service delivery. Considerable training and capacity building as well as significant purchases o f computers, vehicles, and office equipment took place under the assistance o f the Project. However, more needs to be done on some key policy reforms such as right-sizing the public sector, procurement, decentralization, and ensuring the independence o f the Auditor General. 1.4 Onekey component of PSCAP was to improve financial management, accountability, and transparency by helping the Government to implement MTEF and IFMIS, training the Auditor General's staff, increasing the frequency and regularity of audit reports, improving the quality and timeliness o f Office of Auditor General (OAG) reports, and trainingthe Zambia National Tender Board's (ZNTB) staff as well as other procurement support units at line ministries. Some progress was achieved with respect to many of these planned interventions. For example, the Office o f the Auditor General undertook site visits for activities that had not been audited for many years; the Zambia National Tender Board enhanced its inspection and oversight function; computers and other equipment allowed the Office o f the Auditor General to improve the quality and timeliness o f reports; the Ministry o f Finance and National Planning (MFNP) implemented improvements to Financial Management System (FMS) utilizing PSCAP financed equipment; computers were provided for urgent commitment control needs in -1- the Treasury; and, action plans have been developed for implementation o f IFMIS and MTEF.4 Yet, what has been achieved so far is just the beginningo f a long process o f reform that will be required to bring about system-wide gains and improvements in effectiveness and efficiency o f public expenditures in Zambia. This process needs to be sustained and intensified inthe future. 1.5 Fiscal Sustainability Credit 1 (FSC1) became effective in August 2000. Public expenditure reforms under FSCl aimed at introducing spending rules that would help achieve greater predictability, accountability, and transparency at the Central Government level. Interms o f ensuringgreater predictability, FSCl sought to protect the share o f expenditure allocated to line Ministries in the budget by requiring the MFNP to allocate to these Ministries at least 80 percent o f what they are entitled to receive from the voted budget. In terms o f accountability, FSCl primarily helped the Government introduce measures to strengthen control over commitments o f line Ministries to suppliers when no resources have been approved for such spending. This measure is related to the predictability measure inthat lack o f commitment control i s to some extent a consequence o f the unpredictability o f cash flow, and predictability would help reduce commitments for which there is no budget approval. Interms o f transparency, a measure was introduced to publish a monthly report on Government expenditures. In general, while these measures have incrementally helped Zambia move towards better budget management, their overall impact was very limited.5 1.6 Clearly, a longer and sustained reform i s needed to reverse institutionai and behavioral problems that have accumulated over the past several years in Zambia. Looking ahead, much remains to be done in terms o f improving fiscal management and judicious use o f the limited resources in order to further reduce the public sector deficit, improve government transparency and accountability, and increase the impact o f public services on growth and poverty reduction. Institutional and complementary reforms need to move ahead to ensure the effectiveness and efficiency o f public resources and HIPC debt relief, provide adequate public infrastructure and social services, and help in the diversification o f the economy away from copper. 1.7 Fromthe Bank's point o f view, issues related to public expenditure management and financial accountability will form the core o f the Bank's country assistance strategy (CAS) and dialogue for a number o f reasons. First, the PRSP emphasizes the need to monitor poverty and to improve public expenditure effectiveness and service delivery as key components o f poverty reduction in Zambia. Second, the Bank's future support o f the PRSP will be based on the Government's ability to implement poverty reducing programs in an efficient, equitable, and transparent manner. Intum, this ability depends on, among other factors, the quality and quantity o f spending for the poor. Third, donors are increasingly focusing on public sector accountability as a key indicator for aid For more details o n PSCAP, see World Bank Aide Memoire: Zambia Public Sector Capacity Building Project (PSCAP) Mid-term Review; October 14-25, 2002. For more details o n FSC 1, see the credit's ImplementationCompletion Report; December 28,2002. -2- effectiveness and, inthe context o f HIPC debt relief, they have placed these issues at the top o f their country assistance strategies. Recognizing these factors, the Bank's Interim Guidelines for preparation o f PRSCs call for an ex-ante assessment o f a country's public financial accountability, including an assessment o f public expenditure, procurement, and financial management systems before the development o f a CAS and/or PRSC program inthe country.6 1.8 Interms ofdiagnostic work (Economic andSector Work), over the last few years the Bank has prepared two major reports on public expenditure management in Zambia, which have addressed different but overlapping dimensions o f public expenditure management. The 1998 Fiscal Management Report focused on fiscal sustainability issues, and underscored the importance o f reducing quasi-fiscal deficits - i.e., deficits o f parastatals, Bank o f Zambia, pension funds, extra-budgetary funds - in achieving full sustainable fiscal position and full macroeconomic stability. The 2001 Public Expenditure Review focused on budget allocation and execution problems in Zambia, and provided sector-specific recommendations for better public expenditure management and effectiveness. This current report is the thirdmajor analytical report inthis field. 1.9 This report builds on the existing Bank ESW and IMF diagnostics to provide a comprehensive and integrated assessment o f Zambia's overall fiduciary risk, as seen in the areas o f budget management, financial systems and auditing, andpublic procurement. Itprovides not only an assessmento fcurrent practices andinstitutional arrangements, but also a guide to the substantive progress made inrecent years and the key priorities for the Government to focus on in the years ahead. The report i s intended to provide both donors and the Government o f the Republic o f Zambia (GRZ) with a candid review o f public sector problems and a roadmap of short and medium term actions inwhich donors can provide as~istance.~Such an action plan will take into consideration the limited financial and human resource capacity o f the Government and the importance o f closer donor collaboration. This report, in conjunction with the previous two other reports, is hoped to provide a solid analytical basis from which the Bank can engage the Government in the area o f public expenditure management and accountability over the coming years. 1.10 The rest o fthe report is organized as follows. Chapter 2 will assess Zambia's past performance interms o f the macro-economy and poverty reduction. It will then focus on two key areas for refocusing public expenditure towards poverty reduction-reducing the fiscal drain o f parastatals and improving the efficiency and effectiveness o f existing public expenditure. Chapter 3 gives an overview o f the trends in governance in Zambia, the governance and accountability framework, the constitutional provisions for financial accountability, and the role o f parliamentary oversight for public expenditure management and financial accountability. Chapter 4 explores budget preparation issues, %ee OP 8.6, Memo from Ms. Salop to Staff Recipients of the Operational Manual on Interim Guidelines for Poverty Reduction Support Credits, dated M a y 31,200 1, Attachment I. 'In addition to the main report, the Country Financial Accountability Assessment (CFAA) and the Country Procurement Assessment Review (CPAR) constitute annexes to the main report. -3- including the budget's credibility, comprehensiveness, classification, and presentation to Parliament as well as the development o f a medium term expenditure framework (MTEF). Chapter 5 focuses on analysis o f budget execution and expenditure control issues. It covers expenditure control, commitment control and arrears management, payroll controls, capital expenditure controls and asset management, and public procurement. Chapter 6 deals with accounting, financial reporting, and budget audit. It covers a description of the current accounting and reporting systems, the role o f IFMIS as a medium- to long-term tool for improving financial management, accountability, and reporting, and the role and status o f internal and external audits in Zambia. Chapter 7 focuses on humanresources andpayreform issues as they relate to capacity enhancement for effective implementation o f public sector reforms in general and effective budget management in particular. Chapter 8 uses the Education sector as a case study to examine the actual flow o f funds through the administrative network, and assesses leakages and inequity in distribution o f public resources in the education sector. Chapter 9 discusses the public expenditure and financial accountability vision that the Government aspires to in the medium term and, in that context, a program o f priority actions that would enhance Zambia's accountability, transparency, and predictability in public spending. 1.11 This report has three main annexes. Annex I, is provided at the end o f the which report, covers a number o f key issues corresponding to the chapters o f the report. Annex I1is the Country Financial Accountability Assessment (CFAA), which provides detailed assessment o f financial accountability issues. Annex I11i s the Country Procurement Assessment Review (CPAR), which covers public procurement issues in detail. The CFAA and CPAR annexes are issued separately. -4- 2. THE MACROECONOMIC CONTEXT FOR PUBLIC EXPENDITURE REFORMINZAMBIA 2.1 Inthis chapter, we will address three main issues. First, we will review the progress in macroeconomic and poverty indicators in Zambia, particularly during the 1990s. This section shows that while Zambia's macroeconomic stability has gradually improved, full macroeconomic stability i s yet to be achieved. Furthermore, the country saw an increase in poverty and a deterioration in the key social indicators in the 1990s. Second, we will highlight the connection between public expenditure, macroeconomic performance, and improvements in social indicators and poverty. Although Zambia has made considerable efforts since the mid-1990s in increasing and subsequently maintaining the nominal expenditure shares on social sector, expenditures in real terms have fallen. While the decline in real expenditures was underpinned by lack o f growth and high inflation, the deterioration in the social indicators was because of low real spending, poor public expenditure management and financial accountability, low public sector efficiency and poor quality o f services, and the HIV/AIDS pandemic. Third, the scope for increasing domestic revenuethrough taxation i s limitedinthe short term. Given this limited scope, the GRZwould needto focus on maximizing the impact onpoverty o f existing public expenditures. Inthis respect, reducing the fiscal drain o f parastatals, and promoting effective and transparent utilization o f existing resources through better public expenditure management, financial accountability, and procurement systems is critical. A. PROGRESSINMACROECONOMIC AND POVERTY INDICATORS Progress in Macroeconomic Performance 2.2 With an estimated per capita income in2002 of about USD320, Zambia is one o f the poorest countries in sub-Saharan Africa. Zambia's economy exhibits heavy dependence on mineral resources and exports, inparticular copper, which generates over 50 percent o f the foreign exchange earnings o f the country. Over the last three decades (first three columns of Table l), structure o f the economy and composition o f output the changed perceptibly: the share o f agriculture in the economy increased from around 15 percent in the 1970s to about 21 percent inthe 1990s; manufacturing from 16 percent to 21 percent; services from 35 percent to 40 percent; and energy from 2.4 percent to 2.8 percent. On the other hand, the share o f mining declined from 24 percent to about 11 percent and that of construction from 7.3 percent to about 4.5 percent. As o f 2000, nearly half o f the Gross Domestic Product (GDP) came from services, about 27 percent from agriculture, and about 13 percent from manufacturing. Mining contributed only about 3 percent o f GDP (last column o f Table 1). -5- Table 1: Zambia: Structureof Output andRealGDP Growth, 1970s-2000 1970s 1980s 1990s Avg. Avg. Avg. YR95 YR96 YR97 YR98 YR99 YROO Y R O l YR02 PercentShareinGDP Agriculture 14.6 15.8 21.1 18.4 17.6 18.7 21.1 24.1 27.3 22.1 22.0 Manufacturing 16.2 25.0 21.2 11.3 13.4 13.2 13.0 12.2 12.7 11.1 11.6 Mining 24.4 15.5 10.7 16.4 13.7 11.3 7.1 4.2 3.2 4.4 3.9 Construction 7.3 3.0 4.5 4.6 4.0 5.0 5.0 5.2 5.0 6.2 7.4 Energy 2.4 1.7 2.8 3.6 3.7 4.7 4.1 3.7 3.2 3.8 3.4 Services 35.1 38.9 39.6 45.7 47.7 47.1 49.7 50.6 48.6 52.3 51.7 .Growth Rates, 1994 prices GDP 1.6 1.4 0.3 -2.5 6.6 3.3 -1.9 2.0 3.5 4.9 3.0 Agriculture 2.2 3.5 4.8 33.4 -0.6 -5.1 1.2 6.9 1.8 -2.4 -4.1 Manufacturing 4.5 3.6 1.8 -0.4 5.5 5.1 1.8 2.8 13.5 4.2 5.8 Miningandquarrying -2.2 -0.8 -10.2 -27.5 2.8 2.2 -25.1 -24.8 -5.1 14.0 16.4 Construction -1.9 -2.7 -3.2 -3.3 -11.0 29.0 -9.1 10.2 1.2 11.5 17.4 Energy 2.5 -2.0 2.8 -1.5 -5.6 4.2 0.6 2.6 1.1 12.6 -3.2 Services -0.2 1.1 1.8 -2.3 15.0 4.0 4.3 6.6 6.0 4.9 3.2 2.3 From a long-term perspective, Zambia's growth record has been one o f protracted decline (Table 2). The GDP growth rate fell from an average o f 1.5 percent inthe 1970s to 1.4 percent in the 1980s and 0.3 percent in the 1990s. Inflation increased steadily, on average, from around 10 percent in the 1970s to about 70 percent in the 1990s. The steady increase in inflation, coupled with population growth which was above the GDP growth rate, resulted in a decline inreal per capita income o f 1.6 percent per annum since independence. In2002, Zambia's per capita income at constant prices was only about 60 percent o f its per capita income in the 1960s. The country has therefore gotten much poorer. 2.4 Among the reasons for Zambia's poor growth record are: (i) poor performance the of the copper sector and adverse terms o f trade shocks; (ii) macroeconomic instability,in particular high inflation and high interest rates, which deterred private investments; (iii) the lack o f timely structural reforms aimed at reducing the cost o f inefficient public enterprises; and (iv) failure to realize anticipatedbenefits from privatization. 2.5 Duringthe 1990s, Zambia undertook ftndamental changes inits economic policy. Exchange and interest rates were liberalized, the trade reforms in 1995-98 simplified the tariff structure, removed quantitative restrictions, and transformed the Zambian trade regime into one o f the most open in the sub-region; and, the Government successfully concluded debt reduction and rescheduling agreements with the Paris Club and reached the HIPC Decisionpoint inDecember 2000. 2.6 Despite significant structural reforms, however, the economy continued to grow at a very low rate. Over the period 1995-2000, the GDP grew by an average o f only 2.1 percent per annum, compared with an average population growth o f 3.1 percent per year. Inflation averaged about 25 percent, reaching nearly 30 percent in 2000; in addition, -6- domestic savings remained low, interest rates remained high, the current account deficit increased, and the exchange rate depreciated precipitously. 2.7 In2000, Zambia's nominal extemal public and publicly guaranteed debt stood at about USD6.5 billion, more than twice the level o f GDP (Table 2). As o f 2000, debt service stood at 5.5 percent o f GDP, or about 21 percent o f the value o f exports. On the domestic front, the fiscal deficit o f the government averaged about 2.5 percent o f GDP over 1995-2000, and stood at about 5.4 percent in2000. The overall public sector deficit (quasi-fiscal deficit) in Zambia - which includes the deficits o f the central government, the local governments, extra-budgetary accounts, state-owned enterprises, and the Central Bank- was estimated to be muchhigher at about 17percent o f GDP in 1997.' The high overall public sector deficit has inpart continued to provide the structural basis for high inflation inthe country. Table 2: Zambia: MacroeconomicIndicators, 1970s- 1990s (in percent) CPI Inflation Rate 10.2 36.1 70.9 35.2 45.2 24.5 24.4 26.8 30.1 21.7 22.2 Domestic SavingdGDP 33.2 14.0 7.1 12.2 5.3 9.4 3.9 -0.9 3.1 9.8 3.8 Investment/ GDP 30.2 16.2 14.1 15.9 12.8 14.6 16.4 17.9 18.3 20.0 18.0 InterestRate (lending rate) 7.8 16.0 54.7 45.5 53.8 46.7 31.8 40.5 38.8 46.2 45.2 CurrentAccountDeficit/GDP -6.6 -10.8 -4.7 -4.2 ,-3.8 -8.0 -10.9 -8.6 -13.7 -13.0 -9.6 (incl. Grants, percent) Exchange Rate (KwachaAJSD) 0.7 4.8 903.1 866.0 1,207 1,315 1,862 2,388 3,110 3,608 4,398 __ External Debt/ GDP 64.5 171.9 204.1 200.7 215.5 170.2 212.0 209.7 216.8 168.8 168.1 2.8 The macroeconomic performance since 2000 has been encouraging. In the aftermath o f the sale o f ZCCM, the economic outlook for the country brightenedand a strong performance by the mining, manufacturing and service sectors led to a GDP growth rate o f 3.5 percent in 2000, 4.9 percent in 2001, and about 3.0 percent in 2002. Fiscal and monetary performance continued to be satisfactory, and inflation, which stood at about 30 percent by end-2000, came down substantially and reached about 18.7 percent at end-2001 on account o f prudent financial policies, decelerated food prices, and appreciation o f the Kwacha. However, inflation rose to about 26.7 percent in 2002 on account o f a rise in food (in particular maize) prices due the region-wide food crisis causedby drought. 2.9 Since 2000, Zambia faced three major challenges. First, the deterioration in the global economic environment due to the September 11 tragedy and the general slowdown in the world economy depressed the world demand for Zambia's major export '1997 i s the latest year for which estimates o f quasi-fiscal deficits are available. -7- commodities and tourism services, and reduced the foreign exchange earnings. Second, on January 24, 2002, Anglo-American Corporation (AAC) announced that it would not continue operating the Konkola Copper Mines (KCM), its joint -venture, which comprised the major part o f ZCCM's assets when it was privatized in 2000. This decision introduced considerable economic uncertainty in Zambia, as K C M produced about two-thirds o f Zambia's total copper production. In August 2002, A A C and the Government reached agreement on a financing package that finalized AAC's withdrawal from the copper sector. The government is now searching for a strategic partner to run K C M on a commercially viable basis. Finally, Zambia's economic difficulties were compounded by a serious regional food crisis which began inMay 2002. The food crisis was caused by drought in the previous production season, which severely reduced crop production and ledto a significant shortfall (over 600, 000 tons) o f maize, Zambia's main staple crop, which had to be imported by the Government. Despite these difficulties, overall economic performance has remained favorable since 1999. 2.10 The overall macroeconomic picture that emerges is clear: while Zambia saw a significant progress in the area o f structural reforms during the 1990s, full macroeconomic stability and sustainable growth have remained elusive. On the one hand, apart from areas like agriculture and energy, most major market distortions have been eliminated, the role o f the government in commercial activities has been reduced, prices are largely market determined, and some o f the previously state-owned enterprises have been restructured and divested to the private sector. On the other hand, Zambia's economy continues to exhibit weak and at best uneven growth, high inflation, very low savings rate, high real interest rates, a very weak extemal position, a high degree o f dependence on copper production and exports, and a high degree o f vulnerability to shocks. Needless to say, these have had adverse implications for trends inthe poverty and social indicators o fthe country . Progress in Poverty and Social Indicators 2.11 The trends inpoverty and social indicators inthe 1990s provides a useful context inwhich to discuss what public expenditures have or have not been able to achieve in Zambia during the same period, and what, if any, might be neededinthe future. A brief review o f Zambia's progress in poverty, education, and health indicators is provided below. 2.12 Poverty. Zambia's principal household surveys - the Social Dimensions of Adjustment Priority Surveys o f 1991 and 1993, and the Living Conditions Monitoring Surveys o f 1996 and 1998 show that, in general, poverty levels increased during the 1990s. Poverty incidence in 1998 was at 72.2 percent (versus 69.2 percent in 1996). Rural poverty was higher (82.7 percent) and was closely associated with geographic isolation, but poverty inurban areas was on the rise, at 54.8 percent (versus 46 percent in 1996) as a result o fthe economic deterioration inthe second halfo f the nineties. 2.13 Both health and education indicators have deteriorated over the 1990s (Table 3). Infant mortality, adult illiteracy, and malnutrition remain extremely high. Furthermore, the rate o f HIV/AIDS infection is high and has contributed to a sharp decline in life -8- expectancy and a deterioration in health standards (Figure 1). Indeed, the HIV/AIDS epidemic i s affecting all aspects of the social and economic fabric. Table 3 summarizes the progress inpoverty and social indicators duringthe 1990s. 2.14 Education. Zambia's enrollment rates for children o f primary school age are below 70 percent in the rural areas. This i s 10 percent lower than a decade ago. The lowest attendance rate for children o f primary school age occur in the Eastem (48 percent), Northem (60 percent), Luapula (61 percent), Westem (63 percent) and North Westemprovinces (65 percent).' The decline ineducation standards over the past decade and a half has resulted in low competency levels among the children and high illiteracy levels among the adult population with wide gaps between male/female and rural/urban sectors. Low enrollments o f school-age children are largely due to the inability o f their parentdguardians to meet the cost o f schooling; perceived high opportunity costs of children as they are engaged in income generating activities or looking after sick family members; and poor learning environments due to poor infrastructure, inadequate supply of furniture and leaming materials, and unmotivated and sometimes unqualified teachers. 1998 I 58 I Extreme Poverty-Rural/ Urban Rural 81 71 36 Extreme Poverty-Province Central 56 63 Copperbelt 44 47 Eastern 76 66 Luapula 73 69 Lusaka 19 34 Northem 76 67 North-Westem 65 63 Southern 69 I 60 Western 76 78 Education ' Net Primary School Enrollment Rate 83 72 Female 82 72 Male 84 73 Secondary School Enrollment Female 36 35 Male 55 49 HealthandNutrition Life expectancy at birth(years) 44 Under-5 mortality rate (per 1000live births) 191 51 197 Under-5 stunting(1ow height-for-age) I 39 53 ~~ 3IV Infection( percentof adults) -- 20 'Living Conditions Monitoring Survey Report; 1998, Pg. 42. -9- 2.15 Health and Nutrition. The key features o fthe health sector inZambia are: (i) high infant and under 5 mortality rates (Figure 1and Table 3). As o f 1998, the mortality rates were estimated at 114 per 1000 live births for infant mortality and 197 per 1,000 live births for under-5 mortality. Adult mortality rates stand at 597 per 1,000 for female, and 607 per 1,000 for male. These rates have recently been rising due to increased vulnerability to water-borne and infectious diseases, in particular HIV/AIDS, malaria, and diarrohoea. (ii) High malnutrition - according to the 1996 LCMS, malnutrition in rural Zambia had been 47 percent. In 1998, it worsened by almost 20 percent. Between 1991-98, stunting levels increased by 14 percent. The worst affected regions are the Northern, Eastern, Central, and Luapula provinces. These provinces are also characterized by high poverty incidence. Levels o f wasting are higher in urban areas owing to poor sanitation, crowding, as well as the absence o f an extended support system for childcare. Women and children appear to be disproportionately affected as a result o f higherbiological nutrientneeds. According to the 1998 LCMS, 10percent o fwomen are stunted due to protein energy malnutrition especially during pregnancy and lactation when physiological needs increase. (iii) The ruralpopulation exhibits a higherproportion o f illnesses thanthe urban areas. Figure 1: Zambia: Trends inHealth andEducationIndicators, 1990-2000 1993 1999 Source: World Development Indicators Database. 2.16 HIV/AIDS. The escalating HIV/AIDS pandemic, coupled with other infectious diseases like malaria and TB, i s an urgent issue that impacts all sectors o f the economy and society. In addition to their impact on health, education, and social safety nets, HIV/AIDS and other infectious diseases such as malaria and diarrhoea now constitute a key obstacle to hture rapid growth in Zambia. These diseases are expected to significantly reduce skilled labor supply and productivity, savings and investment, and public revenue while at the same time increasing public expenditure on health. A preliminary analysis indicates that HIV/AIDS could depress Zambia's GDP growth by -10- 1.5 - 2.3 percent a year on average." The key challenges inthis area are prevention o f firther increases in the prevalence rate o f HIV infections, care for those already infected and mitigation o f the impact o f HIV/AIDS on vulnerable groups especially orphaned children. 2.17 Thus, in summary, over the last decade, most o f Zambia's social indicators deteriorated. Poverty increased, particularly in urban areas; Zambia's enrollment rates for children o f primary school-age decreased; illiteracy rates among adult population increasedwith wideninggaps between male/female and ruraVurban sectors; andmortality rates, both o f children and adults, rose due to vulnerability to water-bome and infectious diseases, inparticular HIV/AIDS and malaria. B. PUBLICEXPENDITURE, POVERTY AND SOCIAL OUTCOMES 2.18 Macroeconomic stability and growth are essential for making progress inpoverty and social outcomes. While macroeconomic stability i s often a precondition, growth is a necessary condition for improving social outcomes. Inthe case o f Zambia, the lack o f sustainable growth above the average rate of population growth underpins the decline in social indicators. Over the past three decades, Zambia's real GDP has stagnated, representing a decline o f about 3 percent a year in per capita terms and a significantly weaker performance than some countries inthe region. This stagnation has been one o f the major causes o f increase inpoverty and deterioration insocial indicators. 2.19 However, while necessary, growth alone is not sufficient to reduce poverty and improve social indicators. Among factors determiningthe rate at which growth translates into poverty reduction and improvements in social indicators are: patterns o f growth, income distribution, public expenditures, institutions and service delivery mechanisms, regional equity and political representation, household income and demographic characteristics, social policy, andthe extent o funforeseen shocks affecting households.l2 Trendsin Social Expenditure 2.20 Public expenditure is the connecting thread between macroeconomic stability and growth on the one hand and improvements in social outcomes on the other. Without macroeconomic stability and sustainable growth, it would be difficult for any government to sustain a level o f public expenditure that would be adequate to finance key social services. ``IMF, "The Implications ofHIV/AIDS for the Zambian Economy," October 2001. "According to the 1998 LCMS, o f all children under age 20, about 16 percent have lost one or both parents. Among children of primary school age, the share of maternal or double orphans is projected to reach 14percent in2002, 16percent in2006, and 19percent in2010. 12See,for example, World Development Report 2000/2001: Attacking Poverty. World Bank, Washington, D.C. -11- 2.21 InZambia, social expenditures, defined in a specific way,13 are protected in the budget andby the conditions o fthe World Bank's lendingoperations at 36 percent o fthe total domestic discretionary budget. As a result o f the introduction o f such protection in the mid-l990s, social sector expenditures increased from 22 percent in 1991to 36 percent in 1996, and have remainedmore or less at that level since (Table 4). Two points should be noted with respect to the size o f social sector expenditures in the 1990s. First, although the Government has maintained the nominal 36 percent share inthe second half o f the 1 9 9 0 ~social sector expenditure have fallen inreal terms because o f the relatively ~ highinflation rate in Zambia (around 25 percent in 2002). Second, rural infrastructure investments - which are particularly closely related with growth and income opportunities for the poor - have fallen in the 1990s due to both the inability o f the government to make long-term commitments as well as the significant shortfall o f actual to budgeted resources (Zambia PER, 2001). Table 4: Zambia: Share of Social Expenditures, 1996-2001 (percent of domesticallyfinanced discretionary spending) 1996 1997 1998 1999 2000 2001 Educationandtraining 18 18.3 17.4 18.5 23 20.2 Health 13 12.9 13.1 13.7 10.6 13.1 Social security andwelfare 1.3 1.6 1.1 1.2 0.6 2.9 Water and sanitation 3.1 1.6 2.4 2 0.6 1.4 Other social exDenditure 0.3 0.1 0.2 1.1 0.2 0.5 Total social expenditure 35.8 34.4 34.2 36.4 35.0 38.1 Source: Ministry of Finance and National Planning; Zambia PER, 2001. EfficiencyandEquityof Social Expenditures 2.22 However, improvements in social outcomes depend not just on the size o f resources allocated, but also on the efficiency and equity o f social expenditure. In this respect, the conclusions o f the recently completed Zambia Public Expenditure Review (2001) offer a number o f important insights. 2.23 Efficiency of Social Expenditures. The findings o f the recently completed Zambia Public Expenditure Review (PER, 2001) indicate that the delivery o f public services in health and education facilities has been seriously affected by the cash rationing system, which is characterized by substantial unpredictability and volatility of expenditures (in the sense that the quarterly allocations confine the planning horizon for programs and activities ex- ante, and outtums significantly deviate from budget ex-post). In addition, there exists lack o f transparency and expenditure wastage. Proficiency and procurement l3Social sector expenditure is defined as current and capital expenditure on health, education, social safety net, water and sanitation, and disaster relief. Domestically financed discretionary expenditure i s defined as total expenditures, less foreign-funded expenditure, domestic interest payments, the allocation for arrears clearance, the civil service wage adjustments, the contingency reserve, civil service retrenchment costs, payments to the Public Service Pension Fund, net lending to ZCCM, and court awards made against the Government. -12- guidelines are there but oftentimes flouted. For example, transparency in the procurement o f pharmaceuticals and fertilizers remains a serious problem, and there i s a wasteful duplication o f functions and staff between parallel agencies - for example, between the Ministryo f Health and the Central Board o f Health. There are also weak fiscal controls and financial and procurement management, which reduce the money that gets to the final consumer. 2.24 BeneJit incidence of social expenditures. Again, the findings o f the Zambia PER suggest that (i)spending on primary education i s generally pro-poor. Secondary education shares this pattern, though the effect i s not as pronounced as in primary education. The beneficiaries o f the substantial subsidies to university education are not the poor; and (ii) The beneficiaries o f health expenditures are generally more complex and less clear. The largest group usingpublic primary health facilities tends to be inthe middle o f the income distribution, not at either end. In addition, expenditures on hospitals almost certainly benefitthe non-poor more than others. 2.25 Yet public provision o f social services constitutes only part o f the equation in improving social outcomes. Demand or household factors are often as critical. Principal among factors determining households' investment intheir own health and education are behavior, income, service quality, physical access, price, gender, and supply in related ~ectors.'~While our empirical knowledge on the demand side constraints in Zambia is rather limited, available evidence indicates that low income, cost recovery schemes, poor physical infrastructure, gender inequalities and income distribution within a household, and factors outside health and education sectors - such as availability o f food, clean water and sanitation services, and the environment - all constrain the poor in accessing health and education services. Data from the Demographic and Health Survey reveal that about 20 percent o f Zambian adults engage in behaviors known to be highly risky for HIVtransmission. This has contributed markedly to poorhealth outcomes. Through the growth o f the orphan population, it also has had an impact on the education, health, and nutrition outcomes for Zambian children. 2.26 In summary, Zambia made some effort in the 1990s in increasing and subsequently maintaining the nominal shares o f education, health, social welfare, and other social expenditure inthe domestically financed discretionary budget. The share o f the combined social sector expenditures increased from 22 percent in 1991 to 36 percent in 1996, and has remained more or less at that level since. However, due to lack of growth and full macroeconomic stability, social expenditures have fallen in real terms. The decline in real financing may have partly contributed to the deterioration in social outcomes and increased poverty witnessed inthe 1990s. However, it is notjust a decline in real financing that is at the root of the latter; evidence shows that poor public expenditure management and financial accountability, low public sector efficiency and 14 A good summary o f the role of these factors in the context of some African countries is provided in "Public Social Spending in Africa: D o the Poor Benefit" World Bank Research Observer Vo1.14, No.1, pp.49-72 by J. Castro-Leal, L.Demery, and K.Mehra (1999). -13- poor quality o f services, and the HIV/AIDS pandemic have all considerably contributed to the worsening trends insocial indicators andpoverty. Inaddition, demand side factors acted to exacerbate this trend by limitinghouseholds' investments intheir own health and education. c. THESCOPE FORINCREASINGGOVERNMENT REVENUE Structureand Compositionof Revenue 2.27 The two major components o f Central Government's revenue in Zambia are domestic revenues and external grants. Table 5 provides historical data on the trends and components o f Government revenue in Zambia. Revenue as percent o f GDP averaged 19.3 percent o f GDP in the 1990s, and grants accounted for about 8.7 percent of GDP. Foreign grants accounted for the bulk o f the total share, showing Zambia's high level o f dependency on external financing. The decline intax revenue and the fluctuation o f non- tax receipts (which includedincome from parastatals) were always matchedby offsetting andrising grants, which rangedfrom 4.3 to 14.7 percent o fGDP. 2.28 Zambia's main sources o f tax revenue are income tax (consisting o f both personal and corporate tax), VAT/sales tax, excise tax, import duties, and other taxes. In 1990s, total tax revenue averaged 18.4 percent o f total revenue, which was above the average for the countries in sub-Saharan Africa (SSA). The SSA average, excluding South Africa, was 15.6 percent. Tax revenue accounted for about 95.7 percent o f total revenue in Zambia. Of the total tax revenue, income tax accounted for 35.3 percent; trade taxes for 31.5 percent; VAT/ sales tax for 15.8 percent (currently at 17.5 percent); and excise taxes for 16.5 percent. 2.29 IndividualIncome Tax. Zambia's personal income tax system is progressive, with tax rates o f 10, 20, and 30 percent depending on individual income. According to a recent IMF study (2001b), not only have the values o f the tax threshold and rate bands been progressively eroded but also the threshold and rates o f the individual income tax are inequitable. 2.30 Business Income Tax. The general rate o f corporate income tax i s 35 percent, but many companies are liable to different rates depending on sector or activity: companies listed on Lusaka Stock Exchange pay 30 percent; banks pay 35 percent iftheir income i s less than K250 million and 45 percent if their income i s more; and farming and non- traditional exports to 15 percent. These multiple tax rates under the business income tax distort resource allocation, distort investment and capital flows, and increase the burden o f tax administration. In order to reduce these distortions, it may be necessary to unify the business income tax rates, except for the tax on the financial sector. 2.31 Value-addedTax (VAT). The VAT inZambia i s mainly a consumption tax, with a zero rating for exports, some agricultural products, food, medical supplies and drugs. Introduced in 1995 with a 20 percent rate, the VAT replaced an earlier sales tax which was leviedat 23 percent. However, since its introduction, the VAT base has been eroded, and its administration weakened. This has arisen from zero rating previously exempted -14- products, exempting additional ones, and registering a large number o f taxpayers with a turnover below a very low threshold. The five broad groups of zero-rated supplies under the VAT in Zambia are foods and agriculture, export o f goods, supplies to privileged persons (e.g., president, diplomats, donors, etc.,) medical supplies, and tourism sector and hotels inthe Livingstone district. The current VAT level i s 17.5 percent. Table 5: Zambia: Central GovernmentRevenue,1990-2000 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Average Central Govt Revenueas % of GDP Revenueandgrants 24.6 32.9 33 26.3 30.2 29 26.8 25 25.4 25.9 27.9 27.9 Revenue 20.3 18.7 18.4 15.9 20.1 19.9 20.7 19.9 18.8 17.9 21.6 19.3 Tax revenue 19.9 18.3 17.5 15.3 18.7 18.2 19 18.8 18.1 17.4 21.3 18.4 Incometaxes 7.8 6.6 6.5 5.5 6 6.3 6.2 6.4 6.6 6.5 7 6.5 Excisetaxes 2.9 3.6 2.7 2.2 3.1 2.8 3.2 3.3 3.5 3 3.1 3.0 Value-addedtaxes (VAT) 1.9 2.1 2.2 2.3 3.4 3.6 3.5 3.6 3.3 3.3 2.5 2.9 Trade taxes 7.3 6.1 6 5.3 6.2 5.4 6.1 5.6 4.1 4.5 6.6 5.8 Clearanceoftax arrears 0 0 0 0 0 0 0 0 0 0 2 . 1 0.2 Non-taxrevenue 0.4 0.4 0.8 0.5 1.4 1.7 1.6 1.1 0.6 0.5 0.2 0.8 Grants 4.3 14.2 14.7 10.5 10.1 9.2 6.1 5.1 6.6 8.1 6.4 8.7 Structureof Central Govt Revenue Revenue and grants, % of Total Expenditures, 74.8 81.9 88.7 73.5 79.2 88.3 98.1 99.0 88.6 92.6 96.6 87.4 Revenue, % ofRevenue andGrants 82.4 56.8 55.6 60.3 66.5 68.4 77.2 79.1 14.0 68.9 71.2 69.7 Tax revenue,% of Revenue 98.2 97.9 95.5 96.6 93.2 91.6 92.0 94.6 96.7 97.4 98.9 95.7 Incometaxes,%ofTaxRevenue 39.1 35.9 37.2 35.8 32.0 34.9 32.6 33.8 36.4 37.4 32.8 35.3 Excisetaxes, % ofTaxRevenue 14.5 19.5 15.6 14.6 16.8 15.5 16.9 17.4 19.3 17.2 14.4 16.5 Value-addedtaxes (VAT), % o f Tax Revenue 9.7 11.5 12.8 15.1 18.2 19.7 18.2 19.0 18.3 19.2 11.9 15.8 Trade taxes, %ofTax Revenue 36.8 33.1 34.5 34.5 33.1 29.9 32.3 29.8 26.0 26.1 30.9 31.5 Clearanceoftax arrears, % of Tax Revenue 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 9.9 0.9 Non-taxrevenue,% of Revenue 1.8 2.1 4.5 3.4 6.8 8.4 8.0 5.4 3.3 2.6 1.1 4.3 Grants, % ofRevenueand Grants 17.6 43.2 44.4 39.7 33.5 31.6 22.8 20.3 26.0 31.1 22.8 30.3 Source: MOFNP; Zambia PER, 2001. Scope for Increasing Tax Revenue 2.32 Inthe short to medium term, the scope for significantly increasing government revenue i s rather limited. Zambia's tax revenue, expressed as a percentage o f GDP, averaged slightly over 19 percent in the 1990s (see Table 5), which compared very well to countries in Sub-Saharan Africa. The SSA average, excluding South Africa, is 15.6 percent, Malawi 14.3 percent, Mozambique 11.O percent, and Uganda 8.3 percent (World Bank, PER, 2001). Furthermore, Zambia's tax effort i s good relative to countries in the region: its tax system i s well designed and managed by qualified headquarters staff. Raising the revenue effort any higher would therefore likely be difficult without addressing the growth o f the tax base. Without sustainable growth to help expand the tax base, and thereby increase the tax revenue, the capacity o f the government to increase social expenditure (in real terms) i s limited. A recent EMF study (2001b) concludes that -15- while some enhancements can be made to the Zambia tax revenue, efficiency gains fiom tax reform are likely to be limitedmainly to improvements intax administration." D. KEYAREASFOR MAXIMIZING IMPACT OF PUBLIC EXPENDITURE 2.33 Given the limited scope for increasing revenue, two broad areas remain central to increasing the impact on poverty through public expenditures in the years ahead. The first is to refocus domestic resources toward poverty reduction by deepening structural reforms, in particular reforms o f inefficient state-owned enterprises. Such reforms are fimdamental to promoting better utilization o f Zambia's otherwise extremely limited public resources. The second i s to ensure a more effective and transparent utilization o f existing resources. Each o f these i s briefly discussed below. Reducing the FiscalDrain of Parastatals 2.34 As noted above, Zambia has moved from a position o f state-dominated economy and government administered price structures and protective mechanisms in the early 1990s to a position today of an open economy characterized bymarket-determinedprices and far less state domination in the economy. Indeed, some o f the key state-owned enterprises have been restructured and divested to the private sector, including the copper company, Zambia Consolidated Copper Mines (ZCCM). 2.35 ZCCM's privatization in March 2000 marked a new era in Zambia's economic history for two main reasons. First, the transfer of copper assets into the private hands i s expected to lead to more efficiency and betterproduction inthe copper sector. There are already indications that the privatization has led not only to revival o f Zambia's copper production, but also to better investment in services, tourism, and construction in the Copperbelt area. Second, in view o f its huge financial losses, the sale o f the Z C C M constituted a major step toward reducing the flow o f overall public sector deficit. Yet, the privatization o f the ZCCM lefl behind a stock o f debt - estimated at about USD588 millionas o fMarch2000 -mucho fwhich stillneeds to be settled bythe Government. 2.36 However, although a major drain on public resources at the time, ZCCM was not the only parastatal claiming a significant amount o f public resources. Other parastatals contributing to the fiscal drain on public budget included the Zambia National Oil Company (ZNOC), Zambia Electricity Supply Company (ZESCO), Zambia Railways (ZR), andTAZARA and TAZAMA,. 2.37 Deepening o f the structural reforms inthe above areas i s important for three main reasons. First, a stable macroeconomic and fiscal environment i s not possible without mher reduction of the losses of major parastatals that continue to claim significant public resources, and thereby crowd out the much needed poverty-focused public spending in Zambia. Without structural reforms, not only will Zambia's growth record l5IMF(2001). Zambia: Reformo fthe Tax System, FiscalAffairs Department. -16- remain uneven, but also the high quasi-fiscal deficit16 will continue to provide the structural basis for high inflation and macroeconomic instability inthe country. Second, without macroeconomic stability and growth, it i s difficult to expect that the decline in real government expenditure seen in the 1990s in general and those on health and education inparticular will be reversed. Without a decent improvement inreal per capita spending on vital social services such as health, education, water and sanitation, and infrastructure, it is inturn difficult to expect improvements insocial outcomes. 2.38 It should be pointed out that several o f the needed structural reforms are already underway. In 2000, the Bank's FSCl initiated key reforms with respect to a number of key parastatals as well as the major pension finds in the country. The Government i s proceeding with the reform o f the electricity utility, ZESCO, through commercialization to ensure its efficient operation, protection from political interference as well as its independence from budgetary resources including commercial debt guarantees. Inthe oil sector, the Government has liquidated ZNOC and allowed the private sector to freely import crude oil, and agreed to the formation o f an oil price consortium for importation o f feedstock. However, it i s yet to offer for sale or long- term concession the Ndola storage tank; as well as the TAZAMA pipeline. Inthe railway sector, the concessioning is not yet completed. The immediate challenge is to complete these reforms. 2.39 Past pension reforms, in particular o f the National Pensions Scheme Authority (NAPSA), have promoted cost efficiency measures to reduce administrative costs, set up a transparent governance structure, and restructured the contribution and benefit parameters to ensure long-term sustainability. However, the Public Service Pensions Fund(PSPF) remains to bereformedto restore financial sustainability. Increasingthe EfficiencyandEffectivenessof Current PublicExpenditures 2.40 Countries show enormous variation in the efficiency with which public expenditures are used to improve poverty and social outcome^.'^ Insub-Saharan Africa, for example, while the per capita health spending o f Zambia and C8te d'Ivoire i s almost equal (USD24 per person inPPP terms in 1997), the outcome o f that spending interms o f a key indicator o f health - the infant mortality rate - i s significantly different for the two countries (113 per thousand for Zambia but only 87 per thousand for C8te d'Ivoire). The underlying factors for such significant variations in outcomes for similar levels o f spending are rather complex. However, broadly, the inter-sectoral and intra-sectoral `6Apreliminary ongoing work on estimation o f fiscal and quasi-fiscal deficits in Zambia shows that the overall public sector deficit o f Zambia, including deficits parastatals, pension funds, and extra-budgetary funds, stood at around 19 percent o f GDP in 1999 (World Bank, AFTP1). 17For example, a WHO report (2001) which compares health performances across the nations o f the world, finds that although the United States spends considerably more o n healthcare per person than any other nation, it ranks only 15th among countries o f the world on an overall health performance index. On the other hand, Oman, with USD334 inper capita health spending, ranks no. 1 inthe world in life expectancy for its investment. Chile and Lebanon have virtually the same per capita spending (about USD570 per year), but Chile ranks 33rd inoverall performance while Lebanon ranks only 91st. -17- allocations o f budget, the efficiency with which budgets are executed, whether there are clear and transparent financial management and accountability systems in place, and whether the procurement systems are efficient andtransparent are all critical ingredients. 2.41 As pointed out earlier, the delivery o f public services in Zambia has been seriously affected by the cash rationing system, which is characterized by substantial unpredictability and volatility o f expenditures - i.e., quarterly allocations which confine the planninghorizon for programs and activities ex ante, and outtums significantly which deviate from budget ex post. In addition, there is lack o f transparency and expenditure wastage at different levels o f public spending, weak fiscal controls and financial and procurement management, and lack o f financial accountability. All of these point to the need for improving the efficiency and effectiveness o f current public expenditure inorder to havebetter impact on poverty andother social outcomes. E. SUMMARY AND CONCLUSIONS 2.42 During the 1990s, Zambia made significant progress in the area o f structural reforms: major market distortions have been eliminated, the role o f the government in commercial activities has been reduced, prices are largely market determined, and some o f the previously state-owned enterprises have been restructured and divested to the private sector. However, full macroeconomic stability and sustainable growth have remained elusive. Zambia's economy still continues to exhibit weak and at best uneven growth, high inflation, very low savings rate, high real interest rates, and a very weak external position. 2.43 Zambia's social indicators mostly deteriorated during the same period. Poverty increased in urban areas; Zambia's enrollment rates for children o f primary school-age decreased; illiteracy rates among adult population increased with widening gaps between male/female and ruralhrban sectors; and mortality rates, both o f children and adults, rose due to vulnerability to water-borne and infectious diseases, in particular HIV/AIDS, malaria, diarrhea, TB andrespiratory diseases. 2.44 Zambia made considerable effort in the 1990s to increase and subsequently maintain the nominal shares o f education, health, social welfare, and other social expenditure in the domestically financed discretionary budget. In nominal terms, the combined share o f social sector expenditures increased from 22 percent in 1991 to 36 percent in 1996 and has remainedmore or less at that level since. However, inreal terms, the overall government expenditure as well as expenditure on health education, and other vital social services declined duringthe same period. The lack o f adequate financing was thus partlyto blame for the deteriorationinsocial outcomes andincreased poverty seenin the 1990s. Inaddition to the shrinkingrealper capita public expenditure, evidence shows that poor public expenditure management and financial accountability, low public sector remuneration and efficiency, and poor quality of services have considerably contributed to the worsening trends insocial indicators and poverty. 2.45 Without a sustained growth and an expansion in the tax base, the scope for increasing government revenue - and therefore the size o f public expenditure - in -18- Zambia i s limited in the short to medium term. Given the limited scope for increasing public expenditures, two broad areas are likely to remain central in this effort: (i) the need to refocus domestic resources toward poverty reduction by deepening structural reforms, and (ii) the need to ensure a more efficient, effective, andtransparent utilization o f existing public expenditures. These efforts can be accomplished by deepening the structural reforms that are already underway, as well as launching o f new reforms inthe areas o f budget preparation, allocation, and execution. Without better budget management, accountability and boosting o f value-for-money, the impact on poverty o f any amount o f resources will at best remain limited. 2.46 Better and improved budget management and accountability at the central level (e.g., ministry of finance) alone is no guarantee that inefficiency andwastage o fresources will disappear at the ministerial level and down-stream. Thus in order for public expenditure to generate improvements in social outcomes on the ground, effective budget management needs to be extended to the second (ministerial and provincial) and third levels (i.e., service centers) o f budget management. This in turn requires effective institutions, financial management and clear line o f accountability and transparency in resource management, an effective procurement system, and better service delivery. -19- 3. GOVERNANCEAND ACCOUNTABILITY INZAMBIA 3.1 The quality o f governance and the institutional structures that support it have a significant impact on the effectiveness o f public expenditures. While the bulk o f this report focuses on the dimension o f governance often referred to as "administrative" govemance, i.e., the effectiveness o f government in providing public services, political and economic governance are vital as well. Political governance reflects concerns over national security, stability o f the political system, rule o f law, separation o f power, and the democratic process generally. Economic governance includes economic policies, the state-market relationships, and the environment created by government for private sector activity. This chapter establishes the broader context inwhich public sector effectiveness must be viewed and provides an overview o f Zambia's govemance over the recent past. Though the data tends to paint a disappointing picture, it should lend strength and credibility to the current Government strategy to attack corruption, to enhance the rule o f law, and to reinvigorate the role o f Parliament in improving public accountability. The chapter concludes with a discussion o f specific steps that can be taken to strengthen public accountability and voice by addressing Parliament's role in economic management. With the Government having embarked upon a Constitutional review, a unique opportunity exists for the nation to build more effective mechanisms for improving transparency and accountability. A. TRENDS GOVERNANCE IN INDICATORS 3.2 Like the macroeconomic environment, the quality o f governance has generally declined in Zambia during the late 1990s. Although it i s difficult to come up with a concise quantitative measure of governance, one o f the best guides may be the World Bank's Governance Research Indicator Country Snapshot (GRICS). The GRICS examines Governance according to six indices (1) Political Stabilitymo Violence, (2) Voice and Accountability, (3) Government Effectiveness, (4) Regulatory Quality, (5) Ruleo fLaw, and (6) Control o f Corruption. As Table 6 shows, under allbut one o fthese dimensions, Zambia has been either steady or declining in its ratings. From 1997/98 to 2000/01, Zambia improved in terms o f Regulatory Quality and remained unchanged in terms o fRuleo f Law. Inall other dimensions, Zambia's scores declined over the period. 3.3 Zambia's worst governance scores have been in "Control o f Corruption" and "Government Effectiveness" respectively. On a scale from 0 to 100 (with 100being the best), Zambia's percentile rank in "Control o f Corruption" was only 19.9. In "Government Effectiveness" it was only 26.9. In contrast, in Regulatory Quality, where Zambia scored highest in2000/01, the percentile rankingwas 69.8. These statistics serve to highlight the importance o f PEMFAR because among the six dimensions o f -20- governance inthe GRICS, it i s the two on which Zambia scores worst that are the central focus o fPEMFAR. Table 6: Zambia GovernanceTrends POINT ESTIMATES(Worst:- PERCENTILE RANK 2.5, Best:+2.5) (Worst:O, Best:100) Source: World Bank, GRICSIt 2000/01: Governance Research Indicator Country Snapshot, 11 http://www.worldbank.org/wbi/governance/govdata2001.htm or httv://info.worldbank.ordgovernance/kkz/ B. UNDERLYINGFACTORS TRENDS GOVERNANCE FOR IN 3.4 Political Stability. External threats are a clear impediment to good governance because they promote political instability and encourage a diversion of scarce resources to the maintenance o f national integrity. Conflicts on the border with Angola and inDRC have inthe past contributed to a weakening inthis area. Internal conflict can also have a bearing on this dimension o f governance. Although Zambia's transition from one-party to multi-party rule in 1990 was peaceful, the political situation has not been entirely stable. A state o f emergency was declared in 1994; political tension also increased after the Constitution was amended to block the former President, Dr. Kaunda, from contesting the 1996 election; and a failed military coup attempt in October 1997 was followed by repressive measures against opposition parties. Then, inthe run-upto the 2001 elections, President Chiluba's attempt to run for a third term resulted in widespread factional struggles within and across the major parties. 3.5 Voice and Accountability. Election processes, civic participation, and decentralization all have a bearing on the concepts o f voice and accountability. Popular support for democracy is very high in Zambia despite limitations to full-fledged democratization. According to the Afrobarometer Survey in 1999, which is a comparative opinion survey conducted in Zambia and eleven other African countries, 74 percent o f Zambian respondents expressed that democracy is always preferable to any other system o f government. But ingeneral, the institutional linkage between the citizen and the state through formal political process appears to still be weak inZambia. This i s reflected in another Afrobarometer survey, which showed that Zambians were the most pessimistic among Sub-Saharan African countries surveyed18 about the ability o f their vote to change the government and to make their life better. On the other hand, non- "The other countries included in the survey are: Botswana, Ghana, Lesotho, Malawi, Namibia, Nigeria, South Africa and Zimbabwe. -21- government organizations (NGO's) are active and open, and are frequently consulted by Government. 3.6 Although many believe that institutions that are geographically closer to the citizens they serve will do a betterjob o f delivering services, this has not been the case in Zambia. Local governments remain extremely weak, with district councils lacking autonomy or capacity to raise much o f their own reven~e.'~The autonomy and discretion o f local government have also been undermined by continuous central control and interference by the centrally-appointed District Administrators. Perhaps, as a consequence o f this, the Zambian people least trust the local government to do the right things among other governmental institutions and there is a wide-spread dissatisfaction with performance o f the local governments as evidenced by the Afiobarometer Network survey (see Table 7). Table 7: PublicPerceptionabout GovernmentPerformance AfrobarometerSurvey Ouestions H o w well would you say the government is handlingthe following problems: Zambia 3otswana Cimbabwe Malawi ,esotho damibia i. Africa 77 67 53 51 89 20 31 38 47 10 eeping 84 92 57 51 80 FairlyNery Well 41 14 8 20 38 17 Source: The Afiobarometer Network, "Afiobarometer Round I : Compendium of Comparative Data from a Twelve- Nations Survey" Afrobarometer Paper. No. 11.Michigan State Universiv, East Lansing, MI. Table 2-3, p.23; 2002. l9Some o f the means for councils to raise revenue were taken away by central government - e.g., the sale of council houses, the collection o f market levies. Inaddition, local water and sanitation departments were commercialized. -22- 3.7 Government Effectiveness. Improving the effectiveness of the public sector i s the central theme o f the PEMFAR. Its importance i s underscored by a widespread dissatisfaction among Zambian citizens about government's performance. According to the Afrobarometer survey (Table 7), the percentage o f Zambians who are dissatisfied with government performance significantly exceeds those who are satisfied for all o f the areas surveyed. In general, people seem to think that the government has performed worse inresolving economic problems (e.g., employment and price stabilization) than for social services (e.g., security, health, and education). 3.8 Regulatory Rules. As noted inthe previous chapter, the Zambian government has been making a slow but steady progress in economic reforms, such as deregulation o f price controls, liberalization o f trade, and the privatization o f state-owned enterprises. Between 1993 and November 2002, 254 parastatals out o f a portfolio of 280 had been privatized. Privatization o f state-owned Zambia Consolidated Copper Mines (ZCCM) in 2000 is perhaps the most prominent achievement. Trade liberalization has been vigorously pursued since 1992, with all licensing and quantitative restrictions on imports and exports having been eliminated. The level and dispersion o f customs duty rates, which ranged from 0 to 100 percent prior to 1991, was narrowed to 0 to 25 percent. The tariff bands were reduced from 11 to 4, and the exchange rate regime was fully liberalized. Overall, businesses inZambia tend to face more obstacles intax policies and administrative regulation than intrade or licensingpolicies. 3.9 Rule of Law. The stattitory framework, human rights, and the judiciary system have each suffered some setbacks in the late 1990s (Table 6). The 1996 Constitution embodies the doctrine o f the separation o f powers, guarantees the independence o f the judiciary, and circumscribes the power o f the judiciary to pronounce legislation unconstitutional. It is widely agreed that the Zambianjudiciary has been able to preserve its independence fairly successfhlly. However, a 1996 amendment to the Constitution i s believed to have been motivated by political considerations. It has been observed that Zambia's constitutions "have been and remain flawed through the presence o f `savings clauses,' exceptions to each o f the enumerated rights, exceptions conferring wide executive and legislative discretion."20 They go on to express concem over freedom of the press and freedom from discrimination, especially for women. The new Government has initiated a constitutional review process, which could help to address the current weaknesses including public accountability. 3.10 With respect to the judiciary, its functional capacity is severely limited by a shortage o f trained jurists and infrastructure constraints. Poor salaries are the most cited reason for the Government's inability to fill the vacancies. Most courts also lack adequate facilities for courtrooms. Inthe World Business Environment Survey (WBES) conducted by the World Bank with private businesses, most respondents express moderately high trust in the integrity o f the judicial system (ie., faimess and honesty), 2%dulo MunaB.andRobertB.Kent (1998)"The Constitutionso f Zambia." Zambia Law Journal. Vol. 30: pp.26-27. -23- but overwhelming dissatisfaction with the timeliness o f the judicial process (i.e., quickness). Judicial under-capacity also results in a backlog o f cases in the criminal court system, which inturn results invery long pretrial detentions and possible violations ofhumanrights. 3.11 Control of Corruption. Transparency International ranks Zambia as 77th out of 102 countries in its Corruption Perceptions Index (CPI) for 2001. But what may be equally telling are Zambian's perceptions o f Corruption. According to an Afrobarometer survey conducted in 2002, it is widely believed that corruption was worse under the Chiluba government than that o f Kaunda (only 28 percent view corruption as having declined). The privatization program has been a particular source o f concern among Zambians, many o f whom feel that it was conducted in a way that permitted some political officials to amass personal fortunes. The ZCCM Luanshya Mine sale, in particular, is cited as one which lacked adequate transparency in the selection process. Likewise, surveys have shown the occurrence o f graft inpublic procurement. The WBES survey indicated that 70 percent o f business respondents say that firms in their industry customarily make additional or unofficial payments to secure government contracts. This was a higherpercentage than the survey found to be the case infive other southern Africa countries (Botswana, Malawi, Namibia, South Africa, andZimbabwe). Only Madagascar hadahigherpercentage o frespondents offering bribes. 3.12 To its credit, Zambia appears to be turning the corner. The new Mwanawasa government has made the fight against corruption a top priority, and has rejuvenated relevant institutions such as the Anti-Corruption Commission and the Prosecutions Division. However, it will take some time before this will be reflected in improved governance. 3.13 Inits PovertyReduction Strategy Paper (PRSP), the Government recognizes that good governance, especially economic and political governance, go hand in hand with effective poverty reduction. In the PRSP, the definition of poverty is expanded to encompass "exclusion from participation in institutions and processes that govern one's life, leading to voicelessness and powerlessness in these processes and institutions.'' Following from this, the Government identifies three key objectives for good governance that are part of its poverty reduction strategy. The first objective is to democratize decision-making through wider use o f the consultative process. The PRSP process has helped the Government enhance the normal checks and balances among branches, especially by empowering the legislative and the judiciary in preventing corruption. Priority i s also givento empowering local andtraditional authorities, and to strengthening institutional linkages between local authorities and communities. The second objective is to ensure efficient, equitable, and transparent management o f public resources. The third objective i s to guarantee justice for all by advancing human rights, improving the efficiency o f the legal system, and implementing constitutional and legal reforms. Since the development o f the PRSP, the Government has taken steps to beginthe constitutional review process, which in turn, could address a number o f the current weaknesses in public accountability. -24- c. THEACCOUNTABILITYFRAMEWORK INZAMBIA 3.14 One o f the most important issues facing Zambia i s the inconsistency between the legal framework for public finance and the actual structures and incentives in place to assure accountability. This is reflected first in the fact that the Constitution seems to grant an excessive amount o f discretion and authority to the Ministry o f Finance. Typically, in a system with democratic control o f public expenditure, Parliament approves the budget presented by the Executive as well as any major changes that are needed during the budget year. In contrast, under the current financial provisions inthe Zambian Constitution, the Ministry o f Finance i s granted the power to make sweeping changes in the budget allocations without any prior consent o f Parliament (see section below on constitutional issues). A review o f other Constitutions and Finance Acts in Anglophone Africa provides important insights on how Zambia might achieve a better balance. 3.15 Secondly, even where the financial provisions are appropriate, they hinge upon a system o f enforcement and accountability that has not functioned inpractice. The current public finance rules are based on an assumption that the Ministry o f Finance plays a primaryrole o f stewardship or guardianship over the public purse and that it enforces an adherence to the rules by line ministries. However, Ministry o f Finance has not held Controlling Officers accountable for their management o f public resources, nor has the Parliament generally been ina position to holdthe Executive,accountable. As subsequent chapters o f this report will show, without these checks and balances the system simply breaks down. 3.16 This section of the report provides a brief overview o f the legal framework, describes some o f the major weaknesses that exist within the financial provisions o f the Constitution, and discusses ways in which the role o f Parliament could be enhanced to provide an adequate system o f oversight. Description of the Legal Framework 3.17 The legal framework in Zambia is a three-tiered system with the Constitution at its apex. The Constitution 3.18 The Constitution, last revised in 1996, articulates the respective roles and responsibilities o f the Legislature, the Judiciary, and the Executive in all matters, including financial matters. Articles 114 - 121 o f the Constitution give the Parliament powers to establish taxes and to authorize expenditures. The Articles define the financial management roles o f the President as the head o f the Executive, the minister in charge o f finance (M-MFNP), and the Auditor-General o f the Republic (AG). They set the framework for estimates, government expenditures, and other charges to the general revenues. The Articles provide for annual appropriation acts, supplementary and "excess" estimates, and supplementary and "excess" expenditure acts. Article 122 defines the AG's tenure o f office and the procedures for his removal. -25- TheFinance Act- Chapter 347 of theLaws of Zambia 3.19 The Finance (Control and Management) Act defines the roles andresponsibilities for Financial Management within the Executive. It gives the Minister o f Finance the responsibilities for the management, supervision, control and direction o f all matters relating to the financial affairs o f the republic and allows himher to designate a Controlling Officer (CO) for each expenditure head provided for in the annual budget. The Act makes the CO the Chief Accounting Officer and charges hider with the duty of controlling the resources under the relevant head, subject to directions given by the Permanent Secretary (PS-MFNP). It also allows the CO to delegate some responsibilities to any accounting officer (AO) under hisher responsibility, and to provide the latter with necessary directions to secure the proper exercise o f the delegated functions. 3.20 Sections 5-9 o f the Act, inter alia, includes provisions such as: (i)certification by each A0 o f correctness o f accounts and propriety o f charges; (ii) obligation o f Aos to keep books and records andprepare accounts; (iii) procedures for opening bank accounts; and (iv) conditions for bank overdraft facility - that no bank shouldpermit an overdraft on any account unless authorized inwriting by the PS-MFNP. 3.21 However, the Act gives some flexibility and authority to the PS-MFNP. For examples: (i) deemed necessary, the PS-MFNP can establish hnds and working when accounts for different purposes; (ii) the PS-MFNP may authorize the issue, from general revenues, o f sums which may be used to meet payments due for public services which, for reasons the PS-MFNP deems sufficient, cannot presently be charged to any vote, and to make provision for any payment which has been made under any head o f expenditure and which has been disallowed as a charge against the head o f expenditure in the approved estimates; and (iii) the PS-MFNP may authorize that the outstanding balance o f any advance account be charged to appropriate expenditure. The Finance Act also makes a very briefreference to the role o f internal audit, without defining it. ThePublic Audit Act of 1980 3.22 The Public Audit Act contains provisions similar to those inthe Constitution with respect to the AG. It spells out clearly the AG's roles, responsibilities, and reporting obligations. It is noteworthy that the 1980 Act essentially gives the AG the authority to follow public money to anywhere it may go. It widens the power of the AG beyond the audit o f books, records, and reports o f Government departments and the statutory corporations described in the Finance (Control and Management) Act to include the audit, as she/he deems fit, o f every statutory corporation and every private institution that receives Government grant, subsidy, or subvention in any financial year. The AG i s empowered to request from independent auditors o f state-owned enterprises (SOEs) any document, reports, or information relating to the accounts o f SOEs. Also, every contract involving the Government or its agencies and enterprises was to include a clause allowing the AG to have access and examine all books relating to the contract. It also provides for the hiring o f agents or specialist consultants by the Minister o f Finance, at the request of the AG, to assist the AG inthe performance o fherhis duties. -26- Financia1Regulations 3.23 The Finance Act is complemented by a subsidiary legislation, cap. 600 entitled "Financial Regulations," which contains general financial rules and procedures, some Ministry of Finance/Treasury regulations (1985) and circulars, the accounting guides of 1992 (financed by ODA), andother procedures and instructions such as stores regulations o f 1969. Intemal audit function is explained inthe financial regulations (cap. 600). ImprovementsNeededinthe LegalFramework 3.24 While the legislation governing public finances i s reasonably comprehensive in nature, it gives flexibilities to the Minister o f Finance and the Permanent Secretary o f MFNP that could be easily misused inthe absence o f a proper Parliamentary oversight. In fact, the experience over the past decade has shown that this has proven true. The PEMFAR is not intended to represent an exhaustive review o f the legal framework, but insteadit highlightsmajor issues that shouldby considered carefully inthe context of the overall constitutional review that the nation has begun to undertake. The review should address inconsistencies and also consider which financial provisions are best suited for the Constitution and which may be more appropriate in the Finance Act. Likewise, the Financial Regulations will need to be brought up to date to be consistent with a new Constitution and Finance Act. 3.25 Submission of the Budget Estimates to Parliament. The Constitution permits the Executive to wait three months into the new fiscal year before tabling the budget inthe National Assembly (article 117(1)). This essentially obliges the Parliament to approve a quarter o f the year's spending retroactively. Under the current practice, the Government actually tables the budget earlier - on the last Friday of January. This more closely parallels the provision found in the 1964 (Independence) Constitution, which gives one month into the fiscal year for tabling o f estimates. However, even with only one month delay, Parliament still requires some time to debate the budget and vote on it; typically this process is not completeduntillate March, or three months into the budgetyear. The Constitution permits the Government to continue spending for up to four months while the budget i s being considered. Nevertheless, the delay undermines democratic accountability and impedes many ministries from fully undertaking activities they have plannedfor the year. 3.26 OECD has proposed what they consider best practice inthis area: The government's draft budget should be submitted to Parliament far enough in advance to allow Parliament to review it properly. The budget should be approved by Parliament prior to the start o f the fiscal year. (Source: OECD Best Practicesfor Budget Transparency,May2001.) Countries may vary on exactly when the budget must be presented and how much time Parliament should be given. But at a minimum, the budget should be submitted prior to the start o f the fiscal year. Inthe cases o f Malawi and Kenya, the Constitution specifies that the budget mustbe laid before the legislature before the new fiscal year begins. Yet, -27- it would be preferable for Parliament to be given at least two months to review it before the start o f the year.21 3.27 Authority to Spend when the Budget is Delayed. Article 115(2)b of the Constitution permits the President to authorize a provisional warrant to permit spending for up to four months ifthe budget has not yet been enacted. While it is prudent to make provision for continuing government services in case o f a delayed budget, countries usually set limits inthe law to prevent spendingfrom exceeding the prior year's approved budget. InZambia, the Constitution is silent on this issue - implyingthat Government is not necessarily constrained. Although paragraph 11o f the FinancialRegulations do have a provision that appears to limit spending, the fact that it is only an administrative rule means that the Minister of Finance could choose to change it at any time. A clearer provision that sets monthly limits based on the previous year should be established in law. 3.28 Supplementary Estimates. Once the budget is enacted, the Executive in most countries has the authority to execute the budget without interference from Parliament. Yet, this does not mean that Parliament has no role to play. Parliament has a responsibility to monitor the Executive's execution of the budget. Where the Executive requires changes, they are typically to be approved by Parliament through a supplementary appropriations bill. When such changes to the budget can be made without the prior consent o f Parliament, it de-legitimizes the Parliament's role in the budget process altogether. 3.29 Zambia's constitutional provisions for Supplementary Estimates are extremely lax -instead of encouragingdemocratic accountability, they undermine it. Article 117(4) essentially gives the Executive a free-hand to modify the appropriations act any way it sees fit, so long as a report i s eventually tabled before Parliament. Specifically, it states the Minister o f Finance has up to 15 months after the end o f the fiscal year to table a supplementaryappropriations bill. Thus, it is nothing more than a confirmation o f the spending which has already occurred. There are apparent inconsistencies in the legislative framework as well, which suggests that stricter terms may have been considered at one time. For example, article 115(2)c o f the Constitution could be interpreted to mean that approval of supplementary estimates by Parliament was to proceed the issuing o f the warrant by the President. Likewise, paragraph 17 o f the Financial Regulations asserts that supplementary provisions were to be approved by the National Assembly before a warrant i s signed by the President and before the Permanent Secretary issues Treasury Authority to the controlling officer. 3.30 Inpractice, the Executive has been submitting its supplementary appropriations bill at the end of the fiscal year, and its content is incorporated into the following year's budget document (the Yellow Book). However, the point remains that the spending 21The precise date can be specified as part o f the overall budget preparation calendar inthe Finance Act or the FinancialRegulations. A fuller discussion on the presentation of the budget i s found inchapter 4. -28- occurs prior to Parliament's intervention into the process. It i s only after -the fact that Parliament gets involved formally. Perhaps not surprisingly, this has lead to a situation where supplementary appropriations are frequent and large - for amounts sometimes surpassing even the original budget estimate. Table 8 provides some examples from the Year 2000 budget. While any specific example may have some explanation behind it (an emergency o f one kind or another), taken as a whole they show a pattern o f failed budgeting. In some cases, even with a supplementary estimate the actual expenditures appear to exceed the authorized level. Thus, the budget that i s implemented by the Executive deviates significantly from that which Parliament authorized. 3.31 To improve accountability, the financial provisions in the Constitution should be amended to require that as a general rule, supplementary estimates be presented to Parliament for approval before the expenditures have occurred.22 Table 8: IllustrativeExamplesof Use of SupplementaryAppropriations (Year 2000 budget) (millions of Kwacha) HomeAffairs Pes HQ 425 697 1,122 164 1,884 RDCs Immigration 942 1,075 2,017 114 1,147 Nat'l Registration 398 5,507 5,905 1384 1,865 Defense RDC HQ 618 2,412 3,030 390 3,054 Air Force 1,941 24,295 26,236 1242 46,423 Nat'l Service 2,793 0 2,793 n/a 13,814 Intelligence 183 1,176 1,359 643 1,498 Foreign Affairs RDCs HQ 2,228 4,842 7,070 217 5,511 Finance RDCs E M 115 186 301 162 219 Labor RDCs HQ 207 858 1,065 414 802 Source: GRZ, Annual Budget ("Yellow Book") andFinancial Reports. 3.32 Authority for Emergency Spending. Article 115(2)d provides for the President to issue a warrant authorizing expenditure from the general revenues in cases where "no provision exists for the expenditure and the President considers that there is such an urgent need to incur the expenditure that it would not be in the public interest to delay the authorization of the expenditure until such times as a supplementary estimate can be laid before and approved by the National Assembly." This provision responds to a legitimate need for the Executive to have the authority to deal with national emergencies as they arise and without having to convene Parliament. The problem i s that this can 22 Currently, the provisions inthe Financial Regulationsappear stricter on supplementaryappropriations thanthe law, but this still leaves a potentialriskto the accountability framework becausethe regulations canbe more easily changed. -29- easily be abused. Perhaps to mitigate this risk, the Zambian Constitution permits Parliament to prescribe other conditions and limitations (article 115(4)), and it also requires a supplementary estimate to be laidbefore Parliamentwithin four months (article 116). 3.33 Incontrast, neighboring countries have dealt with emergency spending through a voted Contingency Fund. For example, article 180 o f Malawi's Constitution states: "The National Assembly may make provision for the establishment of a Contingencies Fund and for authorizing the Minister responsible for Finance, if he is satisfied that there has arisen an urgent and unforeseen need for expenditure for which no other provision exists, to make advances from that Fund to meet that need..." Kenya's Constitution Art 102 has identical wording. Both go on to say: "Where an advance is made from the Contingencies Fund, a supplementary Appropriation Bill shall be introduced as soon as possible for the purpose of replacing the amount so advanced." The South African FinanceAct requires that the amount used for emergency expenditures be reported to Parliament and the Auditor General within 14 days, and that a supplementary or "adjustments" budget be submitted to the National Assembly within four months at the latest. 3.34 Authority for Unauthorized Expenditure. It is common for finance laws to contain some provision for accommodating spending that occurs in excess o f the appropriation and without the authorization o f Ministry o f Finance.23 In such cases, Ministry o f Finance i s assumed to take disciplinary action against the responsible officers for the improper expenditure. Yet, because the State may have had the obligation to liquidate the liability to the supplier, Parliament i s asked to approve the expenditure retroactively (which it has the right to refuse). Article 117(5) o f the Zambian Constitution contains provisions for this: "mere, in anyfinancial year, expenditure has been incurred without the authorization of Parliament, the Minister responsible for finance shall, on approval of such expenditure by the appropriate committee of the National Assembly, introduce in the National Assembly, not later than thirty months after the end of that financial year...a Bill to be known as the Excess Expenditure Appropriation Billfor the approval by Parliament of such expenditure. '' Some delay i s to be expected for Ministry o f Finance and/or the Auditor General to discover the unauthorized expenditure, however, the time period specified inthe current article o f the Constitution is excessively long. The provision should be revised to include a much shorter time period - consistent with the tabling o f the audited financial statements. The second concern i s that the provision as written could be abused by a Ministry o f Finance purposely to carryout expenditures which were not foreseen by the budget. Provisions such as this one are only viable with an effective system o f oversight that can deter abuse. The South Afr-ican Finance Act specifies that Parliament can reject the unauthorized expenditure, in which case "that amount becomes a charge against thefunds allocated 23Kenya provides for a "statement o f excess" and Malawi provides for a supplementary if it i s found that "any moneys have been expended for any purpose in excess o f the amount (if any) appropriated for that purpose by the Appropriation Act." -30- for the following or future financial years under the relevant vote." Still, the unauthorized spending should be followed by sanctions by the Ministry o f Finance against those involved. 3.35 Virements. The Executive can make some changes to individual line items without recourse to Parliament. The rules on transfers or virements contained in paragraph 16 o f the FinancialRegulations o f Zambia seem reasonable and consistent with those o f other countries: "Ifthe excess expenditure on one item can be metfrom savings on another item within the same sub-head on another sub-head, not being a sub-headfor personal emoluments,' an application for Treasury Authority will be submitted for approval to the Permanent Secretary." Under these conditions no consent from Parliament is required. The South African Public Finance Management Act (1999) has a similar provision but goes a step further byplacing a limit o f eight per cent on the amount o f transfer that can occur between main divisions within a vote (article 43). 3.36 Deadlinefor Financial Statements. A key instrumentinParliament's oversight o f the budget is the annual Financial Report of the Government. Article 118 of the Constitution expressly requires the Minister o f Finance to table before Parliament the Financial Report within nine months o f the end o f the year. Often Government has been late in submittingthe report, but inthe past couple o f years this has improved. Still, the ninemonth delay is too long to promote accountability. The financial report should be presented within six months after the end o f the fiscal year, and the Auditor General's report'within ninemonths. Inthis way, they can both be inputs into the following year's budgetdebate. 3.37 Summary of Recommendations concerning theLegal Framework: The proposed budget should be submitted to Parliament with sufficient time for them to review and enact it prior to the start o f the fiscal year. See more discussion o fbudget presentation inChapter 4. Limits should be fixed in law on the amount of spending that can occur before the budget has been enacted, based on the previous year's approved estimates. Supplementary estimates should require Parliamentary approval before the spendingoccurs. Emergency or urgent expenditures should be funded through a voted Contingency Fund, followed by a supplementary appropriation bill within a short time fiame after the spendinghas occurred. The time period for submittingan excess expenditure billto Parliament for unauthorized expenditures should be significantly reduced, and should be accompanied by disciplinary action by Ministry o f Finance against the responsible officers. -3 1- (0 Annual financial statements o f the Government should be submitted to Parliament within 6 months after the end o f the fiscal year, and the Auditor General's report within 9 months. Enhancingthe OversightRole of Parliament 3.38 In the preceding section, it was emphasized that Parliament needs to have the legal authority necessary to provide oversight o f public financial management. This section considers other factors that will have a strong bearing on Parliament's ability to carry out its functions. These include technical expertise, financial resource, and the quality of information it receives from the Executive. 3.39 A Transparent and Complete Budget Presentation. Parliament's capacity to review the budget depends to some degree on the quality o f the information provided. The structure or format o f the budget should communicate the policy priorities o f the government in a way that is clear and comprehensive. By comprehensive one means that it should include all govemment revenue sources as well as all expenditures. InZambia, budgets have traditionally presented a misleading picture o f the Government's quasi- fiscal position because critical information has been excluded. For example, Parliament lacks information about potential loans and loan guarantees that may obligate the government. (A fuller discussion o fbudget comprehensiveness i s found inchapter 4.) 3.40 Other elements that help facilitate the readability o fthe budget documents include the budget classification system, the availability o f comparative data on actual expenditure (from the prior periods), and information regarding the underlying assumptions. The current document provides little information with which Parliament can compare the estimates with actual spending trends. Nor does the classification system provide a clear picture o f the functional or programmatic priorities o f the Government. In fact, in many ways the classification is internally inconsistent. In practice, Parliament has had virtually no input into the presentation o f the budget, and i s challenged to analyze whatever they receive from the Executive inhardcopy. A certain degree o f analytical time i s lost to the task o f extracting the right information from the budget document. Chapter 4 discusses in greater detail how the budget presentation mightbeimproved. 3.41 Adequate stafJing for budget analysis. Professional staff are essential for members o f Parliament to be able to analyze the Government's budget thoroughly. The Budget Estimates Committee o f Parliament was only formed about two years ago and still lacks adequate technical staff. Lack o f resources contributes to the problem. Some degree o f autonomy i s also warranted to assure that pay scales and conditions o f service are sufficient to attract the right staff. Ultimately, staff o f the Parliament should be accountable to the Parliament and not to the Executive. -32- 3.42 Greater influence on the content and direction of the budget. The Institute for Democracy in South Africa (IDASA) has identified three categories that explain the authority o f Parliaments to amend the budget that i s presented for vote.24 (See Annex 3-1 for models used in Western Europe.) On one end are the "unrestricted" powers, more commonly associated with presidential systems such as the United States, where the Legislature can increase or decrease expenditures or revenues without permission from Government. On the other end are groups o f countries associated with the Westminster tradition that apply a "reductions only" approach. Under this tradition, Parliament may only reduce expenditure items, but may not propose additional spending. In some such countries, if Parliament were to pass an amended version o f the Government's budget it would be comparable to a vote o f no confidence.25 This i s closest to the tradition in Zambia. A third approach, in between these two extremes is whereby countries charge Parliament with responsibility to maintain the "budget balance'' between revenues and expenditures proposed by the Executive. The Fifth Republic o f France i s cited by IDASAas havingplayed a largerole inpromoting this 3.43 InZambia, Parliament is only permittedto propose savings and cannot propose additional spending for a particular head.27 Because the budget i s voted by head, one practical implication i s that Parliament i s not able to make broad functional reallocations within the overall envelope. Instead, its focus is on the allocations within each head. Moreover, since the budget does not generally contain programmatic information, the quality o f the debate about the estimates is naturally going to be limited. 3.44 Going forward, Parliament should assert its existing powers to demand that Government estimates are well presented.28 Furthermore, Parliament should be given an opportunity to provide input into the budget process before the budget i s presented in detail in the Yellow Book. Inchapter 4, sections D and E o f this report, we propose that discussions with Parliament begin earlier over the medium term macroeconomic framework and the broad functional allocations that can be expected from the Government's budget. 24 "Parliament and the Power o f the Purse: The Nigerian Constitution in Comparative Perspective," revised version o f a paper by Joachim Wehner (Budget Information Service, Institute for Democracy in South Africa), first published in the Journal ofAfrican Law, Vol. 46, No.2 (2002), pp. 216-231. Also drawn from an unpublished paper by the same author on "Executive-Legislative Relations inAfghanistan: Considerations inthe Design o f a Fiscal Constitution." (2003). 25Ibid. Citing documents from the Inter Parliamentary Union, Wehner notes that the reductions only tradition grew out o f the days when the House o f Commons was called upon to consider demands for subsidies made by the Crown. The role o f the Commons was to decide whether to comply, and if so, to what extent and by what means. 26Article 40 o f the Constitutiono f the Republic o f France, 1958, cited inWehner (2002). 27The Parliament, o f course, has the power to authorize additional spending by approving supplemental a propriations. "Some initial steps in that direction may have already been taken. In 2002 Parliament unexpectedly reduced the excise duty o f diesel hel, which compelled the Executive to make other changes inthe budget. -33- 3.45 External financing i s an important part o f the Zambian budget. Often these international agreements are not presented to Parliament concurrently with the budget proposal. Yet, because these agreements may have binding consequences on both the present and future Governments, their acceptance should not depend uponthe Ministryo f Finance alone. Rather, they should be submitted to the entire Cabinet, and ultimately to Parliament for approval. 3.46 Effective Monitoring of Budget Execution. Parliament's role i s not limitedmerely to authorizing the initial budget; it needs to be able to monitor the implementation o f the budget as well. Inother words, it should be receiving regular reports from the Executive showing the rate o f execution for each o f the votes during the year. This level o f transparency in budget operations is essential for Parliament to effectively exercise its oversight role - especially when there may be large variations. Unfortunately inZambia, the Executive's current ability to produce reliable and timely reports on actual expenditure by budget line item i s extremely limited. Yet, this does not negate the importance o f establishing this capacity if Zambia i s going to buildeffective mechanisms o f accountability. The OECD has made monthly reporting on expenditures an example o f best practice for transparency in public finance. As an interim step in Zambia, the Executive might rely upon cash release reports to inform Parliament and to signal when the budget estimates may need to be adjusted. However, over the longer term the MFNP should develop regular reporting based on commitments and actual expenditures. The frequency and accuracy o f the reporting may be limited initially, but the objective should be to begin the process and to make improvements over time. 3.47 A more effective Public Accounts Committee. As in most Commonwealth countries, the Public Accounts Committee o f Parliament i s primarily responsible for review o f the Financial Report and the Audit Reports. In Zambia, the committee is comprised o fninemembers. Inkeeping with Commonwealth traditions, it i s chaired by a member o f the opposition, while the Minister o f Finance is an ex-officio member. Though chairmanship by the opposition may help accountability, if membership is nevertheless dominatedby the rulingparty there is a risk that the audit reports will not get the attention they require. According to PAC, other issues undermining accountability include the unresponsiveness o f the Executive to queries from the PAC, and the failure of the committee's recommendation to be binding on the Executive. Parliamentary reforms are already under way in Zambia, and the role and authorities o f the PAC should be addresseddirectly through this process. 3.48 Summary of Recommendations concerning Parliamentary Oversight: (a) The format o f the budget should be improved so that it provides Parliament with greater information and in a more accessible format (See Chapter 4.) (b) Parliament should assure funding for professional staff or consultants to support the Budget Estimates committee and the PAC in review o f the budget. The funding should be tied to the program prepared by the committees and the related cash requirement. -34- (c) Parliament should be required to agree on the overall macroeconomic framework or parameters o f the budget and the broad functional allocations before the budget proposal i s tabled by the Executive. (d) The Executive should make periodic reports availableto Parliament onthe execution o f the budget during the year. Moreover, requests for any supplemental appropriations should be made before expenditure is incurred. (e) Consistent with the overall Parliamentary reforms currently under way, there should be a mechanism developed for ensuring that government responds promptly to the recommendations o f PAC. (0 Government should continue its policy o f investigating individuals who have beennamed inthe Auditor General reports or PAC reports. -35- PART 11.PUBLIC EXPENDITURE MANAGEMENT AND FINANCIAL ACCOUNTABILITY 4. BUDGET PREPARATION 4.1 Budget preparation includes a wide range o f issues, including the credibility o f the budget as an instrument to carry out government priorities, comprehensiveness o f the budget, the classification system used to identify the priorities, and the relationship between the budget and the overall planning and policymaking process. While it i s hard to separate issues o f budget preparation from those o f budget execution, this chapter focuses on the processes that affect strategic allocation o f resources and ultimately the execution o f the budget. The section provides an analysis o f the major problems and a recommended course o f action for the government to pursue for each o f the identified problems. A. CREDIBILITY OF THE BUDGET 4.2 The lack o f effective and credible budget preparation is one o f the most serious weaknesses in Zambian Public Expenditure Management (PEM), which had significant impact on the quality and efficiency of public sector services generally. For years now, the Zambian government has consistently approved budgets that are structurally over- committed and that lack the resources to meet the on-going obligations o f the government. The most obvious result has been that actual expenditures at the end o f the year bear little relationship with the original budget estimates. While the Ministry o f Finance assumes the lead role in preparing the budget, the failure to reconcile spending priorities before the start o f the year i s clearly a problem that transcends any single ministrywithin Government. Indeed, the decisions for restructuring andstreamlining the priorities o f the Government are not merely technical issues but political ones and can only be made at a political level. Rationalization o f the budget and the development o f realistic budgets - given the resources that are expected to be available - must be a first priority of the government. Analysis of the Problem 4.3 Developing a credible budget has in recent years been a major priority o f the Government. However, the rationalization o f government services that may be needed to achieve this has not yet occurred. Instead, government budgets have largely reflected only incremental changes from the prior year, without tackling the kind o f substantive -36- policy decisions needed to bring costs in line with resources. Some government infrastructure and services that were possible during an era o f high copper prices are simplyunaffordable today. On an aggregate basis, the budget appears to be inbalance - with sufficient revenues and financing to cover expenses. However, what is less visible are the imbalances that exist on a ministry-by-ministry level - i.e., budgets for some ministry activities are insufficient to cover the actual costs, which means that other activities are over-budgeted relative to the aggregate financing available. Imbalances also exist between central and local government levels. Ministry o f Finance officials readily acknowledge that some o f the activities o f government have essentially fixed costs associated with them, which have to be met in some form or another, whether the activities are fully funded in the budget document or not. (Examples o f these include civil service salaries andbenefits, foreign embassies, bursaries for current students, meals for those in prison, etc.) Ultimately, these activities may receive the funds they require throughthe process o f supplemental budgetsapproved byParliament. However, they can only do so by under-funding other activities that havebeen previously budgeted. 4.4 Consequently, most line ministries realize that the budget is not a reliable basis on which to plan activities since actual releases will vary substantially from the amounts indicated in the budget. Chapter 5 provides quantitative examples o f how much budget execution varies from budget estimates. For some ministries, the expectation is that they will receive less thanthey have beenallocatedinthe budget; while for others, they canbe reasonably assured o f being able to obtain more. A review o f any year's Financial Statements will show that in many cases the excess spending is accompanied by a supplemental appropriation. However, frequently, the Financial Statements indicate that excess spending occurs without a supplemental appropriation, or for amounts beyond what the supplemental provided. Unfortunately, under these circumstances ministries may conclude that there is little incentive to put substantial effort into budget preparation. 4.5 As indicated above, a conscious under-budgeting duringbudget preparationis one o f the primary causes behind the eventual variance between the budgeted and actual expenditure. There may be a combination o f circumstances that contribute to the under- budgeting, including a lack o f good cost data from line ministries. However, there are clearly cases where the under-funding o f activities i s or should be easily discemable to MFNP. Table 9 below examines the historical pattem o f budget estimates and actual expenditures for a few selected line items. For any individual case, there may be a sound explanation for the budget allocation decision that was made. Yet, together they show a common pattem in which budget estimates are often not adjusted to reflect the higher level o f spending that has occurred in past years. Although the budgets do increase somewhat, they often remain below the historical spending level. The confirmation that the relatively high spending was not an anomaly for just one year is seen inthe fact that the actual expenditures remain at the same or at an increasinglevel. -37- Table 9: A Comparisonof HistoricalSpendingwithApproved BudgetEstimates for Selected LineItems (millionsof Kwacha) Home Affairs RDCs HQ 469 944 400 1,020 620 1,553 1,653 Prisons 447 4,285 390 1,495 1,508 3,871 2,126 Defense PES Army 24,970 30,396 26,259 30,697 28,098 44,579 40,320 Defense RDCs Army 3,078 19,373 4,360 30,989 3,935 42,369 4,569 Air Force 2,468 8,584 2,063 27,602 1,941 46,423 3,889 Nat'l Service 2,247 5,257 2,794 8,286 2,793 13,814 3,099 MedServices 2,230 2,229 3,512 4,607 3,5 11 5,967 2,495 Foreign Affairs RDCs HQ 2,440 5,116 797 3,415 2,228 5,511 4,2 10 Lubumbashi 47 276 75 205 57 107 58 Washington 305 622 264 629 385 120 912 Cairo 260 589 223 641 661 644 597 Kinshasa 145 429 135 430 139 546 191 AddisAbaba 302 566 237 725 804 1,110 829 Abuja 160 608 124 474 168 360 358 Source: GRZ, Annual Budget ("Yellow Book"), FinancialReports. 4.6 Three examples from table 9 make this point about under-budgeting stand out clearly. First, for prisons, the spending on recurrent departmental charges (RDCS)~~ was about 4.3 billion Kwacha in 1998 compared to a budget o f 469 million. Yet, each o f the next two years budget estimates were only 390 million and 1,508 million. Secondly, for the Army RDC spending in 1998 was more than six times the original budget estimate (19.3 billion versus 3.1 billion). This level o f spending continued to increase substantially in each o f the next two years (to 31.0 and 42.4 billion). Still, the budget estimates even up through 2001 for Army RDCs were only a small fraction o f the 1998 level actual spending. Third, RDCs budgeted for foreign missions tend to be lower than their historical cost. For example, with the Kinshasamission, RDC spendingin 1998 was about three times the budget estimate (429 million versus 145 million). During the next two years the spending remained steady or increased, but the budget for 2000 had actually declined slightly. 4.7 Although the MFNP obviously does not have available to it any audited data on expenditures for the current year when it begins preparing budgets for the next year, it does possess other kinds o f information that can be informative about the projected level o f expenditures. Some o f this may come from the line ministries themselves, but the MFNP also has its data on cash releases made to the ministries for the first half o f the year and it has some information on expenditures from the prior year. In the examples 29Recurrent departmental charges are equivalent to operations and maintenance. -38- included intable 9, data for 1998 would have been a good indicator o f spending levels in 2000. Unless the budget organizations are obliged to make policy or programmatic changes that can lower the cost structure of the activities, there may be little reason to believe that keeping budget estimates artificially low will necessarily lead to lower spending by them. Thus, the actual spending level would tend to be a better indicator o f future costs than is relying upon the prior year's budget estimates. Although the budget reflects national policy priorities (and not merely prior spending), pre-existing commitments may bringwith them a cost structure thancannot be changed immediately. 4.8 Even the information available on supplemental appropriations could provide a good indicator to MFNP o f the potential discrepancies in their budget estimate. For example, in the 2001 Yellow B00kY3'an appropriation o f 14.1 billion Kwacha i s proposed for the two RDC codes covering "Road, Rail and Air Fares." This i s despite the fact that the total authorized budget (including supplementals) for the prior year was 87.7 billion or over six times that amount. Similarly, in 2002, only 13.4 billion was appropriated, despite the fact that the final budget in 2001 was 63.5 billion (about five times higher). While it is not unusual to have some variation because o f the element o f unpredictability, the magnitude o f the difference suggests a need for inquiry. Perhaps policies must be changed to enforce the lower budget or a higher number acknowledged. 4.9 Had MFNP actually raised its budget estimates for the line items above, it necessarily would have meant lowering the budget estimates for other entities and/or increasing the planned deficit. Although the potential reductions in budgets for such activities as rural development, clean water, and health may not have been popular, it could nevertheless achieve budgets that better reflect the actual pattern o f spending. At the very least, this would have made the trade-offs and policy choices o f Government more apparent and perhaps speeded the process o f restructuring government programs. The current practice o f under-budgeting, however, gives a false impression o f Government's overall policy priorities and objectives. In addition, it creates an environment conducive to the accumulation o f arrears, as ministries seek to find alternative means to meet their public service commitments. 4.10 It is also become apparent in recent years that the Government's process o f collective bargaining with the civil service unions contributes to inaccuracies in the budget. For example, the 2003 budget was tabled well before the Government had concluded its labor negotiations. Although MFNP had allocated a precise amount for 2003 salary awards, unions did not feel bound by this figure and held out for increases that may have exceeded it. This, in addition to awards granted by the Public Service Management Divisionto non-unionized workers, can easily force the wage bill to exceed the amounts previously budgeted. This suggests that the entire process o f wage adjustments and union negotiations needs to be revisited. The amount o f the salary 30The Yellow Book i s the name given to the Government's annual budget document because o f its yellow cover. -39- awards for the coming budget year needs to be known more reliably when the budget is submittedto Parliament. FutureAction Plan 4.11 Going forward, the Government must give the utmost attention to encouraging a credible budget. This would include policy decisions by the Government that better align the budget estimates with the historic costs o f the activities, or that make policy changes inthe programs inorder to reduce their costs andbringtheminline withbudget. This is not an exercise that can be undertaken by MFNP acting alone. Nor can many o f them be made within the context o f the annual budget process. Closing offices or embassies, changing policies on bursaries, or reducingthe size o f the civil service are not actions that could necessarily be implementedwithin a few months, but would require some multi- year planning to achieve efficiency. Policies decisions are made now, with the full implementation in a future year. An MTEF would be an excellent fiamework through which to debate and make these kinds o f trade-offs. However, the absence o f a full MTEF need not preclude the government from using other means to begin making transparent policy trade-offs andmore clearly identifying costs inthe budget. 4.12 Making budgetary trade-offs is ultimately a political decision that requires a collaborative effort on the part o f multiple institutions at the center o f government policymaking. However, there are some technical preparations that MFNP can do that could assist policymakers in generating informed decisions. Indeed, MFNP has a responsibility to do so. Key components o f MFNP's work should be to quantitatively assess the current variances between the budget estimate and the actual expenditure to understand where the biggest problems lay and to understand the underlying activities that drive the costs. The Bank and other donors can provide intermittent TA to assist MFNP if necessary. Ultimately, programmatic decisions have to be made as to how to curtail cost in line with available resources. But as the government makes progress in budget preparation, they should also be able to quantitatively show that actual expenditures for a given line ministry are more closely aligned to the original budget estimates. This i s obviously a shared responsibility o f both MFNP and the line ministries. Outlined below are more details on a proposed action plan that the Government should consider adopting. PossibleAction Steps: (a) Establish a baseline analysis by comparing the 2003 budget allocations withthe actual expenditures for 2002. (b) Identify the line ministries (and subunits) that have the largest variance between budget estimate and prior actual expenditure. (This should indicate potential areas o funder-budgeting in2003.) (c) Develop a programmatic understanding of what have been the major cost drivers for each o fthe line ministries (and subunits) identified above. -40- Develop a policymaking framework or action plan for reconciling the differences between costs and budget estimates. The PladFramework would clearly delineate the roles andresponsibilitieso fvarious institutions including MFNP, Cabinet, State House, and Budget Committee. Plan should address issues such as: How to establish a multi-year planning horizon within the context o f an MTEF so that decisions can be made today that provide at least one full budgetcycle for line ministries to implement. How will decisions be made about the appropriate ceilings for each line ministry, and whether those ceilings should cover current cost or requireprogrammatic cutbacks. How will follow-up be done with line ministries to ensure that programmatic changes needed to produce cost savings are indeed implemented duringthe coming budget year. How will future budget execution in the context o f an MTEF be monitored to ensure faithful budget implementation. Beginningwith the budget for 2004, give line ministries ceilings that are based on a percentage o f actual expenditure rather than the prior year budget. (Since 2003 expenditures are not knownbythe time ceilings need to be communicated, the ceilings could be based on either the 6-month cash releases or on the 2002 full-year expenditures or some combination o f the two. Supplemental budgets are not fully available yet, and are only a partial proxy.) Require line ministries to submit budget requests for 2004 that conform with the ceilings that are communicated above, and reach agreement on the programmatic savings that are expected from them for the next year (2005). During 2004 verify that programmatic changes agreed to are being implemented by line ministries, and repeat actions above to identify new round of line ministries and/or activities to target. Review how collective bargaining with public service unions can be improved to assure that the amount o f the salary award is known by the time the budget i s presented to Parliament. Reduce the use o f supplemental appropriations. B. COMPREHENSIVENESSOF THE BUDGET 4.13 Lack o f comprehensiveness o f public expenditures and fragmentation o f budget information are two major issues that contribute to undermining credibility of the budget -41- in Zambia. A number o f policies or actions initiated by government and its agencies generate indirect costs which are not explicitly recorded inthe budget but ultimately have to be paid for by the government. These are often termed quasi-fiscal activities. A recent estimate for Zambia shows that the consolidated public sector deficit-consisting both o f the non-financial public sector deficit (or government) and the quasi fiscal deficits o f the central bank-is far higher than that o f central government budget deficit alone (World Bank, 1998). Accounting for public sector's expenditures and deficits beyond those reflected in the current central government's budget becomes very critical to understanding the true fiscal position o f the country. Without such comprehensive budget formulation, it will be difficult to have good budget management, policy analysis, and formulation. 4.14 Accounting for overall public sector expenditures and revenues is important for two main reasons. First, the size o f the overall public sector deficit has consequences for macroeconomic stability. A common method for financing government deficit i s to rely on the central bank to raise the reserve money, which without a corresponding increase in the demand for money, will amount to inflationary financing. In many countries including Zambia, this i s often facilitated by the lack o f independence betweenthe central bank and the government. Second, from a fiscal policy standpoint, sustainability o f the fiscal deficit depends not only on the deficit o f the central government, but on the consolidated public sector deficit that includes the accounts of the Bank o f Zambia, parastatals, pensions funds, and other public entities. In the case of Zambia, the importance o f the latter in determining fiscal sustainability and solvency are thoroughly discussed and illustratedinWorld Bank (1998). 4.15 A consolidated and comprehensive public sector expenditures budget would need to go beyond the central government to include activities and expenditures of local governments, extra-budgetary funds, public enterprises, and the central bank. Inaddition, the budgetwould need to include other "off-balance sheet" items, such as guarantees and other contingent liabilities o f the public sector as well as the fiscal risks associated with such items. The omission o f these items may provide a misleading picture o f the financial position o f Zambia, especially due to the heavy losses and contingent liabilities coming from state enterprises and public sector pension funds. Analysis of the problem 4.16 Zambia's budget is confined to central government and lacks information on local government expenditures, state enterprises, pension funds, special funds, and quasi-fiscal activities o fthe Bank o f Zambia. Each o f these issues i s briefly discussed below. 4.17 Quasi-fiscal deficits of the BoZ. The deficits of the BoZ mainly include losses on foreign exchange transactions, losses due to loans and guarantees to parastatals, and losses associated with failure o f commercial banks. Inprinciple, the BoZ is entitled to compensation from the treasury on all foreign exchange losses arising from its operation, as well as on losses on loans and guarantees to parastatals. Inpractice, these losses are not allowed for in the budget, although these losses have substantially been reduced in recent years. -42- 4.18 Parastatals. Subsidies to and losses o f state-owned enterprises are not reflected inthe budget. With the significant progress made inthe 1990sinprivatizing most ofthe key parastatals, the Government has been able to avoid paying additional subsidies to these companies, which constitutes an important saving. However, there are still a few state enterprises (e.g., ZANACO, ZAMTEL, ZESCO, and INDENUTAZAMA) whose activities and losses have important budgetary implications, which would need to be reflected inthe budget to ensure comprehensiveness. 4.19 Pensions funds. The two most important pension schemes in Zambia are the Public Services Pensions Fund (PSPF) and the National Pension Scheme (NPS). The PSPF is a defined benefit scheme, funded by a 14.5 percent payroll tax split evenly between employees and the government. However, government has on several occasions failed to include its share o f pensions contributions and obligations in the budget. Such obligations would need to be adequately reflected in the budget to ensure comprehensiveness. 4.20 Local governments. Information on local government accounts is not adequate in Zambia. Anecdotal evidence shows that revenues o f sub-national governments are estimated to be extremely small as a proportion o f total government revenue - about 4.6 percent in 1997, for example - and have been contracting in real terms for the last six years or so (Saasa et. Al, 1999). Transfers from the central government constitute only a negligible part o f local governments' revenues as well - around 3 percent in 1997. There i s a need to obtain adequate information on local government budget, and include it inthe national budget to ensure comprehensiveness. 4.21 Contingent liabilities. Accounting for contingent liabilities is important because actual historical deficits represent too narrow a view o f the public sector's exposure to fiscal risk. While contingent liabilities are not directly associated with any existing budgetary program, a government's commitment to accept obligations contingent on fiture events amounts to a hiddensubsidy and may result in a major unexpected drain on government finances in the fiture. Several expenditure items may need to be accounted for as "contingent" liabilities faced by the public sector. Inthe case of Zambia, the key items are likely to be: GRZ guarantees on public pension finds, GRZ guarantees on loans to public enterprises, and BoZ guarantees associated with parastatalshtate-owned enterprises. The fiscal fragility o f the public sector, as reflected by the unrealized but yet still potential loss in each period, has been growing up to 2000, in large part due to the delay in ZCCM's privatization . These contingent liabilities are expected to continue to diminishwith phasing out of the GRZ ``paper" notes issuedto Bank o f Zambia and the expected improvements inthe implementation o fthe ongoing pensions reform. 4.22 Inaddition to the above, there are two other areas that may need emphasis. First, there i s strong evidence suggesting that revenues from non-tax sources (inparticular, fees for services) are substantially under-reported and generally withheld at the source rather than remitted to the government account at the BoZ. In addition, extra-budgetary funds, financed by non-central government grants, are not covered in the budget documents. Second, the budget i s not comprehensive in terms of coverage of donor activities. Donors often execute their own projects in order to improve outcomes and avoid -43- expenditure wastage and irregularities. However, to the extent that actual donor expenditures are not reported and differ from plans, the fill coverage o f public expenditure is not known. The inputs o f donor resources into a particular sector may also create imbalances inthe distribution o fresources across provinces ifnot coordinated well with Government. (See discussion in Chapter 9 on the Educational Service Delivery Survey.) ActionsTaken to Date 4.23 InZambia, untilrecently, the domestic and external budgets were not integrated well institutionally. The Director o f Budget in the MFNP concerned himself with domestic revenues and domestic expenditures, while the external resource mobilization department o f the MFNP andthe BoZ monitored balance o f payments (BOP) inflows and used them to pay external debt service directly. Ifthere were a shortfall inBOPsupport, the BoZ would cover the debt service requirements with its reserves and increase the balance o f the "forex bridge loan" from BoZ to the government correspondingly. Similarly, ifthere were an excess of BOPsupport, it would be used to reduce the balance of the forex bridge loan. Domestic debt was recorded by the Bank of Zambia, which reported to the MFNP, but MFNP did not have its own database o f domestic debt with which to monitor the overall exposure as well as integrate domestic and external debt management. 4.24 However, under the support o f Bank's Fiscal Sustainability Credit 1 (FSC1) o f 2000, the government was able to improve the database and management o f domestic and external debt. In particular, FSCl helped the BoZ and the MFNP to adopt a computer based data system for recording and managing domestic and external debt in an integrated manner. In addition, the agreement under the Credit that the BoZ should be compensated by the government for all losses on new guarantees helped control the quasi-fiscal deficit o f BoZ arising from such activities. The BoZ has not been called on to provide any guarantees inthe last two years. FutureAction Plan 4.25 A number o f actions are needed to address problems surrounding the lack o f budget comprehensiveness: 0 Enforce the estimating and reporting on the quasi-fiscal deficits o f the public sector, including deficits o f the BoZ, pension finds, and state owned enterprises, as part o fthe annual budget presentationto Parliament. 0 Integrate information on major contingent liabilities o f the Government into the budget. 0 Make a steady effort in collecting regular information on local government revenues and expenditures, and inintegrating such information inthe budget. -44- 0 Enforce regulations/controls on reporting o f all government receipts including fee income generated by line ministries. Establish a system to improve coverage o f donor-executed projects and expenditures inthe budget. C. BUDGETCLASSIFICATION 4.26 Because the budget structure i s limited to an administrative and very broad economic classification, it often obscures the major activities and programs being carried out by Government within and across sectors. In response, the MFNP has worked over the past several years to develop the concept o f Activity Based Budgeting(ABB). ABB i s intended to provide a detailed functional classification as the basis for budget management, andthereby improve the link betweenresource inputs and specific activities or outputs. Considerable work remains, given that ABB has thus far only been piloted in a few ministries and it i s not yet integrated into the cash release system and accounting system. Until this occurs, ABB is unlikely to have a substantial impact on the budget process. Realistically, it is difficult to see how ABB will be implemented meaningfully until the information systems are upgraded, thereby enabling the accounting system to accommodate it. In the meantime, more could be done to reinforce compliance with existing budget rules, including the proper classification o f expenditures in the budget. The misclassification o f expenditures across budget headings and economic classifications i s not uncommon and serves to obscure the real nature or intent o f public expenditures. Analysis of the Problem 4.27 The current budget structure lacks a functional or programmatic classification that would allow one to understand the policy priorities o f the government. While some o f the administrative headings - especially in education - do allow one to discern the functional purpose o f the expenditures, by and large the current administrative classification is not a substitute for a functional or programmatic classification o f the budget. Because o f this, it is difficult to identify what are the poverty-reducing expenditures in the budget, except in very broad aggregate terms. While the ministries that participate inpoverty-reducing expenditures are known, the activities for which they are responsible are often obscured in the current administrative classification. Even if budgets are initially prepared within the line ministry in a more detailed manner, these are not carried through in the cash release process or inthe accounting for expenditures. (For example, M O E has prepared detailed estimates o f the cost o f runninga school based on various characteristics o f the school, but the accounting system does not permit them to track the actual cost.) It i s in part for this reason that the HIPC process necessitated that the Government adopt a parallel system that preserved some o f this additional information in outtum reporting. Without this, it is only by classifying an activity as a "project" that one i s able to directly monitor spending on it. The trade-off i s that one loses the detail on the economic nature o f the spending, e.g., whether for salaries or supplies. -45- 4.28 The economic classification currently in use obscures in many ways the true nature o f expenditures and i s largely inconsistent with GFS standards. The Zambian budget groups expenditures into four broad categories - Personal Emoluments (PES), Recurrent Departmental Charges (RDCs), Grants, and Capital Expenditures. Under each o f these are more detailed categories - some o f which are displayed in the budget and others that were used only for initial budget preparation and internal use. But the inconsistencies in the budget classification quickly become apparent. Capital expenditures, for example, may consist of "movable objects" or "projects," the latter o f which may or may not include capital assets. Many o f the "projects" are recurrent in nature, and may be categorized there because o f the element o f donor hnding. 4.29 Likewise, the category "Grants" includes payments for salaries and for operational expenses. In some cases, salaries and wages are very deliberately budgeted for under the grants heading inorder to reduce the amount shown to be budgeted for PES. During the research for the PER, staff from the Ministry of Health were candid in explaining the incentive to do this for the 1997budget (i.e., to respond to donor pressure). But this type o f reclassification o fpersonnel costs as grants was still evident inthe 1999 budget. A review o f the 1999 Financial Statements (Table 10) shows that the budget for the Ministry o f Health called for PESof 5.6 billion Kwacha (at headquarters) and Grants o f 44.4 billion Kwacha (for provinces), but the report on actual expenditures suggested that the grants were once again intended for salary payments. The financial statements prepared by the Auditor General show expenditures on grants as only 10.8 billion Kwacha (or one-fourth the budget), but the amount spent on PESis 38.6 billion Kwacha (almost seven times the budget). When the PE and grant amounts are taken together, there i s little difference between budget and actual expenditures, it i s only the relative proportions betweenthe two that have changed. Table 10: Ministry of Health (billions of Kwacha): Budget line items through which personnel costs are paid L I -~ Source: GRZ, Financial Report, 2000. 4.30 Complicating matters further, personnel-related costs also appear hidden within the allocations for RDCs. One of the major subcomponents o f RDCs is "allowances." Though the RDC category i s generally the equivalent o f "operations and maintenance" (O&M) in other countries, in Zambia RDCs include employee allowances for subsistence, mileage, and meals.31 In the 2001 budget, allowances were set at 33.4 billion Kwacha, which is equivalent to about 5 percent o f the budgetedPESand about 5 31RDCsalso include training expenses, which in practice i s possibly being used to supplement wages to some degree as well. -46- percent o f the total RDCs. Though this may seem relatively small, in actual practice it may be considerably higher. The structure o f the Financial Report does not permit one to see the aggregate expenditures for this economic classification as a whole, but it does show the wide variance at some institutions. For example, allowances paid by the Office o f the Vice-president, Parliamentary Business were 4.95 billion Kwacha compared to a budget o f 4.1 million. The Judiciary-Industrial Relations Court allowances were budgeted at 6.9 million but actual spendingwas 66.1 million. (Though this looks like a relatively small variance compared to that for the Office o f the Vice-president, the difference i s nevertheless a magnitude o f 10, and there are other examples inthat range.) The overall point is that the budget classification as it i s structured and applied inZambia makes it extremely difficult to determine the true level o fpersonnel versus non-personnel related expenditures. The system, therefore, i s inneed o f substantial reform. ActionsTakento Date 4.31 Recognizing the limitations o f the current budget classification, MFNP has had a team working on the development o f ABB with assistance from the EU. The ABB aims to provide functional, economic, administrative, and spatial (or geographic) classifications o f the budget. Although earlier versions o f the ABB classification appeared to be incompatible with GFS, those issues have been resolved and the latest proposal i s generally consistent with GFS. 4.32 ABB is being piloted in several ministries However, it has not yet been incorporated into the mainstream o f public expenditure management. Thus far, ABB has primarily been a budget preparation tool used within line ministries. Although there has been discussion about the need to integrate it with the accounting system, one of the obstacles to this occurring has been the outdated computer system and software currently used by the Accountant General's office. The system is not conducive to making the kinds o f major overhauls that are required for ABB. Similarly, the ABB classifications are not usedto determine the cash releases to line ministries. Insummary, it means that even if line ministries use ABB to prepare their budgets, the budget approved by Parliament is not in that format, nor are the cash releases they receive, or the outturn reports requiredby the Accountant General. Indeed, as long as ABB is a parallel system for budgetpreparationit will be o f marginal value to the ministries. 4.33 Once ABB is integratedinto the internal control and accounting structures it could potentially be very helpful to improving the transparency of the budget by highlighting 32Thebudgets for four ministries (Local Government and Housing; Commerce, Trade and Industry; Works and Supply; and Land) were developed in 2000 using ABB for application in 2001. This pilot test was extended to an additional six governmental units (Office o f the Vice President; Public Service Management Division; Cabinet Office; Environment and Natural Resources; Agriculture, Food, and Fisheries; and Science, Technology, and Vocational Training) in2001 for application in2002. -47- specific activities or functions being carried out, including identification o f poverty- reducing spending.33 FutureAction Plan 4.34 Going forward, MFNP should take immediate steps to reduce the more blatant cases o f misclassification o f expenditures that occur under the existing system. Inother words, for the sake o f transparency, PESneed to be budgeted under PESand not under other economic classifications. Over the medium term, MFNP should be taking actions to expand the use o f ABB by line ministries and to integrate it with other parts o f the budget process and reporting. Although full implementation o f ABB may hinge upon implementation o f IFMIS, there are transitional steps that MFNP can begin taking as preparation. PossibleAction Steps (shortterm): Ensure that expenditures inthe next budget are categorized correctly under the current economic classification system. In particular, ensure that payments for personal emoluments to civil servants are not hidden inother categories such as grants. Make the integration o f ABB a more explicit part o f the planning for IFMIS. This includes addressing specific concerns that the Accountant General may have with the structure and ensuring that all coding needed for accountingpurposes are taken into consideration. Begin planning how (domestic) capital expenditures will be better integrated with the recurrent budget; in particular, how to ensure that it uses the same classification structure. Begin developing a crosswalk between the ABB and existing budget classification, and then generate supplemental expenditure reports that can be compared with the initial ABB budgets. PossibleAction Steps (mediumterm): (e) Use ABB coding inthe pilot rollout o f the IFMIS. (0 InformParliament o fthe presentation o fthe budget inABB format. 33There has been some concern expressed that the potentially 32-digit coding structure could be more complicated than i s necessary or practical. However, even if there i s scope to make modifications in the ABB, it is already a credible foundation on which to beginbuildingthe account coding. -48- (g) Ensure that cash releases, commitments, and expenditure accounting all use the same classification - rooted in ABB. (Some differences in the level o f aggregation may be appropriate however.) (h) After ABB in pilot ministries has been expanded to budget releases and accounting, then expand ABB inbudget preparationto all ministries. D. LINKS BETWEEN BUDGETING AND PLANNING 4.35 Given the difficulties that Zambia has had with executing the annual budget, integrating a medium term expenditure framework could pose some challenge. Indeed, some aspects o f an MTEF would be difficult to implement credibly until there i s greater consistency between the annual budget and the actual expenditures for the year. Yet there are parts o f the process o f developing an MTEF that could be critical to helping the GRZ to develop more credible budgets and to makingthe prerequisite downsizing o f the public service structure. The MTEF is obviously no panacea for improving the quality o f budget preparation. However, it i s a tool that could serve to reinforce the importance o f other reforms that are taking place. This section examines the linkages between planning and budgeting and outlines some incremental steps that could over time lay a foundation for better budgetplanning and implementation. Analysisof the Problem 4.36 The relationship between planning and budgeting in Zambia can be viewed from several perspectives. With each aspect there is a substantive impact on the quality and credibility o fbudgetpreparation. 4.37 Disconnect between medium term development objectives and the annual budget process. One o f MFNP's own critics o f the PRSP has been that it lacks a medium term fiscal framework. Without the link to budgetary constraints, the PRSP becomes little more than a wish list. Although some prioritization i s already reflected in the PRSP process, it i s hard to judge whether it i s sufficient without greater insights into the prospective hndingenvelop either on an aggregate basis or bymajor sector. 4.38 A consultant's January 2002 draft "Zambia PRSP: Resource Envelope, Sectoral Priorities, Financial Costs and Programming," was a first step taken by the government to compare the expected cost o f the PRSP activities with the expected resources. Considerable work remains, however, if this initial analysis i s to be mainstreamed into the budgetprocess. A couple limitations o f the exercise are that (i)focused exclusively it on the development budget and not the entire budget, and (ii)it was a stand-alone exercise and was not directly linked to the annualbudgetprocess. 4.39 New initiatives are not adequately costed. As new policies and programs are being developed, decision-makers should be aware o f the likely budgetary impact and consider it intheir deliberations. However, there has been little evidence inthe past few years that policy formulation in Zambia is closely linked to budget preparation. On the contrary, it appears that many new government activities are initiated without an adequate and comprehensive costing, and more importantly, without an explicit -49- indication o f where the off-setting resources would come from. It i s well-known that the budget is fully committed at the start o f the budget cycle, i.e., that there are no additional funds for new initiatives. Yet, the institutions that are at the center o fpolicy formulation are not necessarily obligated to consider the cost o f the activity or to propose off-setting savings inorder to accommodate it. 4.40 Prior to the start o f the budget year, there needs to be a means to translate the broad policy statements o f the Government into budget terms. For example, do the priorities outlined inthe annual budget speech really translate into concrete programs and activities that have discrete costs associated with them - both in the budget year and beyond? The relative roles o f Cabinet Office and MFNP in costing the Government's programs have traditionally been undefined, and often this activity does not occur. Although additional resources may be added to a ministries budget, there has not been a clear link to specific activities or policies. One example cited bythe IMF inits June 2000 report was the 1999 internationalconference inLusaka on AIDS, which althoughplanned inadvance, was not explicitly budgetedfor. In2001, the same canprobably be noted of the OAU conference hosted inLusaka, which requiredsubstantial outlays inexcess o f the initialbudget. 4.41 Too short of a planning horizon to implement some policy changes that require phasing. Because the annual budget i s structurally imbalanced, it will require the Government to make hard choices to curtail some public sector activities. Yet for some activities, programmatic changes needed to reduce the cost structure may not be easily implementable within a single year - even less so within the weeks between the presentation o f the budget to Parliament and its approval by Parliament. Some decisions such as closing embassies, changing benefits to university students, hospital reimbursements, or the size o f the payroll may require some lead time to be introduced. Ministries cannot wait until the approval o f the budget to begin implementing them for that same year. MFNP therefore needs to have a longer budget horizon to offer line ministriesinorder to assurethat programmatic changes are implemented. 4.42 Not only incases o f downsizing, but also inthe initiation o f new programs or new capital projects, a longer planning horizon i s needed. Before a new program or project is fimded initially, MFNP should be compelled to show how the progradproject will develop over time and its implications for the budget. This i s particularly crucial for capital projects, when multi-year funding is necessary in order to complete the project andto produce a usable asset. For example, ahalf-completed road or bridge would make a second year o f funding essentially an obligation, and therefore it should be factored into decision making for the sake o f transparency and for economically justifiable decisions. The recurrent costs of the projects once they are completed could also be more easily captured with a multi-year framework inplace. 4.43 Lack of a medium term budget horizon and budget priorities inhibits effective dialogue between Government, Parliament, civil society, and donors about the broad policy direction beingpursued. Experience inSouth Africa has shown that their Medium Term Budget Policy Statement (MTBPS) has increased transparency o f the budget process and improved discussion with Parliament as well as civil society about public -50- spending priorities. The MTBPS has been useful because it i s formally debated by Parliament, and because its aggregated format focuses the discussion at a more strategic level - before detailed line items are ever presented. In contrast in Zambia, the presentation o f the full budget, in all its detail, to the full plenary o f Parliament and only after the beginning o f the budget year, is not a process that i s conducive to strategic deliberations. To promote greater accountability and transparency for the budget, MFNP would benefit from presenting to Parliament an updated MTEF prior to the development of the Yellow Book estimates. This shorter, more strategic-level policy document could help focus discussion about relative functional priorities ina way that the current process does not. 4.44 Likewise, a medium term budget horizon would facilitate the Government's dialogue with cooperating partners and help to assure better congruence with national development objectives. As has been seen in Tanzania and Uganda, an MTEF that i s comprehensive in its coverage o f the budget provides a logical framework for dialogue between the government and its cooperating partners over spending priorities. Too frequently, GRZ is in a position in which it is reacting to the funding proposals o f cooperating partners. By focusing on the totality o f the budget, the annual developmenthpdating o f the MTEF would put government in a position to initiate and guide the discussions around funding needs and donor commitments for each o f the major sectors. Even if the estimations o f cooperating partners for future funding are imprecise, it nevertheless puts government in the appropriate position o f leading the discussions around fundingpriorities. Actions Taken to Date 4.45 Duringthe midto late 1990's MFNP didproduce an analytical document referred to as medium term fiscal framework (MTEF), however, it was never a very significant part o f the budget process. The document appeared to be little more than an internal planning tool for MFNP to consider aggregate resource levels and broad functional allocations. There was no Parliamentary discussion o f the document; nor was it binding on any agencies. The document was generated by MFNP and lacked bottoms-up input from line ministries or donors. 4.46 Recent efforts to launch development of an MTEF have hinged upon a restructuring o f MFNP and the creation o f a PlanningUnit. Restructuringwas completed at the beginning o f 2002 and the Director o f Planning was appointed . Although a local consultant had been available to the MFNP Budget Office under FSC to assist with MTEF throughout 2001, the Ministrycould not take full advantage at that time as the Department o fPlanning had not yet beenestablished. 4.47 Subsequently, through FSC support, the government sent three teams o f staff to review the work done on MTEF in South Africa, Uganda, and Tanzania. With that experience behind them, the MFNP Planning Department, developed a Budget Framework Paper (BFP) as a forerunner to the MTEF and as a guide to drafting the 2003 budget Call Circular and developing budget ceilings. The BFP i s intended to provide Zambia's political leadership with a full picture o f the financial liabilities, policy -51- commitments, and program costs o f the government, as well as the likely resource envelope. The BFP takes into account both domestic and non-project foreign financing expected in 2003. The process o f developing an MTEF for 2004-2006 has already commenced. FutureAction Plan 4.48 Experiences in other countries have shown that a well-functioning MTEF i s not realized ovemight, but instead it requires steady work on multiple fronts. Given Zambia's current challenges in institutional capacity, it should strengthen budget planningbytargeting several importantcomponents: (a) forward year program planning, (b) bottoms-up costing and cost projections, (c) aggregate resource forecasting, (d) rigid adherence to budgetceilings, and (e) consensus buildingon the functional allocation o fresources. 4.49 Forward year program planning. To successfwlly implement some o f the programmatic and policy changes that would help bring government commitments inline with resources, GRZ should have a multi-year planning horizon that is linked to the budget process. In other words, GRZ needs a decision-making framework or review process in place that enables it to propose policy and programmatic decisions whose implementation can extend beyond the immediate budget year. If budget preparation occurs in year x, the line ministry would have until the beginning of year x+2 to fully implement the policy or program changes. This represents a commitment by GRZ to (i) identify a few important target areas inwhich programmatic changes could be introduced to lower their structural costs; (ii)reach agreement with the responsible ministry that the changes will be studied and implemented; (iii) follow-up that they are implemented; and (iv) reflect their implementationinthe budget decisions for the next year. For the sake o f feasibility, the multi-year planninghorizon recommendedhere i s merely one year beyond the budgetyear. 4.50 Yet, this forward-year program planning is not exclusively the responsibility o f MFNP. Indeed, because it includes reviewing govemment wide policy priorities, the exercise should really involve close collaboration between MFNP and the Cabinet to establish cross-sectoral priorities. While MFNP is critical for providing a technical evaluation o f the full cost o f various government commitments, others would normally makethe decisions as to which o fthose commitments to reduce. The BFP for 2003 was a first step for MFNP in involving the political level o f government more deeply in the budget preparation cycle. Indeed, while the creation o f a planning department within MFNP should help, going forward Government may still need to clarify how this new department's work will relate to the activities o f Cabinet Office, State House, and other -52- policy coordinating offices o f the government. One consequence may also be the need for strengthening Cabinet Office inplanning andbudgeting. 4.5 1 Bottom-up Costing and Cost Projection. A good budget should have as one o f its core inputsbottom-up cost projections for maintaining the existing level o f services. This provides a baseline cost o f continuing public services at the current level o f quality and coverage. To do this requires line ministries to understand the factors that drive the costs o f their major programs and activities and to be able to summarize them. Because the Yellow Book structure i s not programmatic, it may take additional effort by line ministries to present information on their major activities. One o f the outputs should also be information about the workload or level o f service beingprovided. 4.52 MFNP and cooperating partners working together, possibly through the prism o f an annual Public Expenditure Review (PER), could help line ministries review their projected costs, their financing requirements, and their highest priority areas. The aim o f these cost projections i s to determine to what extent they are sustainable. And ifnot, then to facilitate the process o f making choices concerning what activities should bepreserved and which abandoned or curtailed. In the case o f capital projects, this may serve to reduce the quantity that are included in the Yellow Book - but increase the likelihood that those included can actually be funded. Currently, this information - for both recurrent and investment activities - seems to be largely lacking or not used by MFNP, since budget allocations frequently bear little relationship with the cost o f the program or activity. Gaining a deeper understanding o f the true cost o f government would be a substantial step toward establishing MTEF and for budget allocation decisions to be based on this information. Although IFMIS would facilitate the process, the existing records may be sufficient to provide initial guidance. 4.53 In addition to understanding the programmatic costs, the MFNP would want to also have a reliable inventory o f public sector liabilities (i.e., arrears) and projections o f the cost that will accrue for non-payment. This would serve as a starting point to assist the Cabinet developing a medium term planfor prioritizing and liquidating arrears. 4.54 Aggregate resource forecasting. MFNP should strengthen its medium range forecasting o f domestic revenues and improve dialogue with donors about expected mediumterm resources. This projection o f the overall resource envelope is essential if GRZ is to transform the PRSP into a useful policy document. Without the constraints beingreasonably evaluated, it is difficult to know how much prioritization should occur. Since there will always be some uncertainty around revenues, MFNP should also consider ways in which budget formulation can appropriately assess the risks and to reduce their potential impact on core activities o fthe govemment. 4.55 Rigid adherence to annual budget ceilings. GRZ can also prepare the terrain for an MTEF by more rigidly enforcing spending ceilings based on the annual budget. A prerequisite for an MTEFto have usefulness inplanningi s that the annual budget i s itself a credible and implementable document. MFNP and Cabinet Office may needto jointly share responsibility for ensuring that budget ceilings are respected. If additional -53- resources are needed, Parliament should be requested ex-ante to authorize a supplemental appropriation and offsetting sequestrations (rather than ex-post). 4.56 Consensus-building on the functional allocation of resources. GRZ should consider introducing into the budget cycle a stage during which resource allocations are debated at a broad, strategic level well before the budget i s presented to Parliament in detailed line items. Such a presentation would communicate the government's broad priorities among sectors or functions over a medium term horizon, including the shift o f resources between sectors/hnctions that i s implied. Doing so would have two main advantages for the Government. First, it would limit the amount o f information that needs to be absorbed by Parliament at one time, and thereby create a more focused discussion around major policy issues (e.g., should spending on health rise relative to spending on defense, should the overall budget increase by 5 percent or 25 percent.). This may be especially relevant in the current Parliament. Before there is any plenary discussion o f the Yellow Book, GRZ could present the broad parameters to Parliament in a separate document. This not only benefits Parliament and civil society, but it also focuses the Government's attentionto outlining its medium range objectives. 4.57 A second advantage o f the medium term allocation projections is that it can provide a basis for line ministries to conduct medium range planning, and for cooperating partners to align their interventions with government priorities. Even if the allocations are not binding on MFNP or the Government, they may still introduce a measure o f predictability for line ministries andhelp frame government's dialogue with donors. 4.58 The Bank can provide technical assistance ifnecessary to enhance capacity within MFNP to develop the parameters document. For example, while there are difficulties in predicting inflation and future revenues, the value o f the projections could be enhanced by including the allocations as a percent o f total spending and/or as a percent o f GDP. Likewise, categories o f spending could be left rather broad. The South Afiican MTEF, for example, shows allocations for four broad sectors, within which are eleven functional categories. The Bank's recent PRSP Fiscal Framework analysis also provided a potential functional categorization that could be used. Besides presenting a functional breakout o f spending, MFNP would be well-served to include distributions by economic classification (e.g., how will spending on PES change relative to spending on domestically-funded investment). The period covered could include spending for the prior year, estimated spending for the current year, and proposed spending for the budget year and the two years following. After some experience using the medium term projections, there should be growing incentive or ability to make them more bindingor to limit the level o f change. However, in the first couple of years GRZ would focus on beginningthe process, usingout-year estimates as only indicative. 4.59 To introduce these kinds o f projections, the relative roles o f MFNP, Cabinet, and State House may need to be updated. Indeed, arriving at a consensus within government around the mediumterm functional allocations may require more coordination duringthe budget cycle thanhas occurred inrecent years. MFNP staff should start now to work out a timetable and relative responsibilities o f each institution. One possible sequencing is the following: -54- Develop a macro-economic framework, budget guidelines, and potential sectoral allocations Submit draft macro-framework, budget guidelines, and sectoral allocations to Cabinet for review A GreenPaper is preparedfor Parliament and civil society to review Budget hearings between MFNP and line ministries, followed by draft budget and MTEFpreparationbyMFNP. Cabinet approves final budget ceilings for ministries MFNPfinalizes the budget Draft MTEF is submitted to Estimates Committee o fthe Parliament Finalbudget and MTEF are submitted to Cabinet. E. PRESENTATIONOF THE BUDGET The budget presentation's quality has a direct bearing on the level o f transparency inpublic finance. InZambia, the format and lack of comprehensiveness inthe budget presentation have in many ways limited the ability o f Parliament and civil society in exercising their rights to monitor the Government's policies. Some aspects o f this problem have been discussed in other sections o f this chapter. First, i s the fact that the budgetlacks completeness inthe expenditures which are covered-major items that have a bearing on fiscal management are left out o f the budget (Section 4-B). Secondly, the current budget classification system is unhelpful in delineating the distinct programs and activities o f the government, or the economic nature o f the spending (Section 4-C). This section builds on the earlier two to present a brief summary o f how budget presentation could be improved.34 Recommendations 4.61 Early agreement with Parliament on the overall macro framework and major functional allocation of resources. As described inthe preceding section o f this chapter, Parliament and Government should debate the overall framework o f the budget separately andprior to the submission o f the Yellow Book. This would enable analysis to remain focused on the economic assumptions and overall strategic direction o f the Government before presenting detailed line items. 34The Bank acknowledges that it has not reviewed all of the documents that are presented to Parliament as support material to the Yellow Book. Thus, some parts of the materials proposed here might already been given to Parliament as accompanying reference documentation. -55- 4.62 Historical data on expenditures. Good budgets will have historical data with which Parliament can compare the current budget proposal. At a minimum, MFNP should be expected to show a side-by-side comparison o f three periods - (a) actual expenditures for the prior year, (b) an estimate o f expenditures for the current year, and finally, (c) estimates for the budgetyear. The only historical data presented inthe current Yellow Book i s the current year's budget estimates, along with any supplementary appropriations. This is wholly inadequate for Parliament to evaluate the changes from year to year because the budget estimates bear no relationship to the actual expenditures. (Even with the supplementary appropriations, they are wide divergences from the actual expenditure.) 4.63 Prospective economic trends. As noted in the previous section, Ministry o f Finance should also provided a forecast o f future revenue and expenditure trends for the next two years. This should not be as detailed as the budget, but on a very broad aggregate basis - at least for the major expenditure categories - Parliament should receive economic projections along with the assumptions that underlie them. Ideally, this should be presented to Parliament as part o f a pre-budget package, which would include an estimate o f future sectoral allocations. Ministryof Finance should get acceptance on the aggregate resource and expenditure levels before allocation to the detailed budget estimates. 4.64 Status of Capital Expenditures. At least for domestically financed capital expenditures (and eventually for all capital expenditures), Government should indicate in supplementary budget documents the status o f each o f project. The status report should indicate the original cost o f the project and the year, the amount expended to date, and the future anticipated cost. It should also indicate the approximate rate o f physical completion. 4.65 Summary tables and analysis. The budget document should also present summaries o f major budget categories. For example, it should be possible to see at a glance the total budget for PESfor the Ministryo f Education, and to compare that with the total PESfor Ministry o f Health. But under the current format, PESfor Education are spread across several budget heads and are never totaled. At a minimum, the budget would be more usefbl ifit presented a comparison o f expenditures by ministryand major economic classification. In addition, it should provide tables comparing the major changes from year to year. 4.66 Authorized employment levels. The Yellow Book may be more useful if it showed not only the authorized budget for PES, but also the number o f authorized positions andthe actual headcount as o f the date o f its submission. 4.67 Changes in Debt Stock and other liabilities. The budget should disclose an estimate o f the current debt stock and other liabilities o f the Government, along with any changes that are forecast for the budget year. Additions to sovereign debt, inparticular, should not be incurred without the consent o f Parliament. -56- 4.68 Narrative on the Budget. In addition to data described above, the budget document(s) needs to provide a much better narrative o f the proposed policies. The "budget speech" that accompanies the tabling o f the budget is too generalized, while the Yellow Book itself i s virtually silent. Some issues which the budget document should explain innarrative include the following: 0 What are the major changes in the budget compared to the prior year - both in terms o f budget estimates and with respect to the policy or program itself - and what are the reasons for the changes. 0 Reasons for any major changes have occurred intax policy. 0 The major economic and programmatic assumptions. This might include any risks that would have a significant financial impact on the cost. 0 Some descriptive information on the objectives and purposes o f the proposed activities, including quantitative data on workloads. 0 Discussion o f changes in financial assets and liabilities, non-financial assets, employee pension obligations, and contingent liabilities. 4.69 Improved budget classification. As noted in section C o f this chapter, the budget documents should include a programmatic classification, as well as an improved economic classification. 4.70 An electronic format. In addition to the Yellow Book, transparency could be enhanced by making available to Parliament and civil society an electronic version of the budgetdata. 4.71 Incorporating all o f these changes in the presentation of the budget will not be feasible overnight. But these represent standards for improved transparency that the Zambian Government should be striving for, and which Parliament should over time be demanding. -57- 5. BUDGETEXECUTIONAND CONTROL 5.1 Problems inbudget execution and budget control lie at the root o f Zambia's fiscal failure in the past decade. This chapter discusses these issues, beginningwith the cash rationing system, which caused significant deviations between budgeted and actual allocations o f expenditures. The cash rationing system also led to a high degree o f uncertainty over the timing of those allocations. Together with other factors such as the structural under-budgeting at the start o f the year and ineffective control mechanisms, this system contributes to largebuild-upo f arrears. Actions to bringarrears under control need to include policies and procedures for limiting commitments as well as improved budget preparation measures. Two key areas that pose the highest financial risk to the govemment budget are payroll and capital expenditure controls and they are reviewed in detail in this chapter. Finally, a factor that contributed to the lack o f effective budget control in the past was the govemment banking arrangements; Annex 5.1 discusses progress inthis area. 5.2 Despite the substantial attention drawn to the technical aspects o f implementing and controlling the budget, there is a broader and overarching need for political commitment to enforce the existing rules and improve accountability. As section B o f this chapter highlights, the legal framework places the Controlling Officer at the center o f budget execution. However, he/she is appointed by the Minister o f Finance and accountable to the Minister for the proper execution o f each budget head. The system will fail to work when the Minister o f Finance i s not in a position to hold Controlling Officers responsible for their management o f the budget head. A. CASHRATIONING 5.3 The single most important characteristic o f Zambia's budget execution o f the budget has been its reliance upon "cash rationing" to determine resource allocation. Although an understandable response to macroeconomic crisis at the time o f its creation in 1993, the cash budgeting process has persisted until today in ways that are highly destabilizing to effective public expenditure management. Indeed, cash rationing has essentially rendered the Yellow Book unreliable as a predictor o f resource allocations to line ministries. Instead, actual expenditures by line ministries often bear little relationship to the budget initially approved by Parliament. The World Bank has attempted to moderate some o f the negative effects o f cash rationing and to encourage cash releases that are more consistent with budget estimates. These efforts have born some measure o f success, but there remain significant reforms needed if line ministries are to attain predictable and steady resource flows inorder to carry out their programs. In the future, MFNP's cash release decisions need to be based on an analytical assessment of cash needs and a rules-based process of allocations - bounded ultimately by spending limitsagreeduponinthe context ofameaningfulbudget. -58- Originsof CashRationing 5.4 Inthe 2000 study, "The Cash Rationing System inZambia," the Bank presented some o f the history and rationale for the current cash budgeting process in Zambia. Introduced in 1993, the cash budgeting system was part o f a fiscal adjustment program aimed at reestablishing financial stability and bringing down the runaway inflation. The system was based on the principle that no cash would be released to line ministries and other budget heads for payment of their expenditures before sufficient funds were available inthe Treasury's central account with BOZ to fully cover those payments. 5.5 The cash budget has evolved significantly since its inception, with one o f the most important changes occurring in 1997 when the daily funding was converted to monthly funding. However, the actual decisions about how much each ministry and other budget heads will receive each month in each expenditure category are still taken ad hoc by a committee within MFNP. The committee's aim i s to respect the quarterly benchmarks established with the IMF, while responding to the new and competing demands o f the line ministries. In the end, it i s the decisions o f this committee rather than the voted budgetthat determines the spending levels o feach o fthe line ministries. 5.6 Some general criteria guide the cash release decisions. Traditionally, the first revenue shortfalls, andpersonal emolument^.^^ The social sector has also been protected priority has been given to domestic debt service, contingency-set-aside for unexpected under an earlier agreement with IDA at 36 percent o f total domestic discretionary spending. In contrast, funding o f RDCs, grants, and capital expenditure (fur non-social sector activities) have been given lower priority, with domestically financed capital expenditures being little more than a residual. But across sectors, the data has shown a more varied story with some ministries clearly doing better than others inrespect to their allocations for RDCs and grants. Analysis of The Problem Impact on the Typeof Spending 5.7 Despite the semblance o f fiscal discipline it gave, the cash budget had a clearly adverse impact on the quality o f public management in line ministries. A distinguishing characteristic o f the MFNP's cash release policy has been that it fails to follow any consistent or predictable pattern. Without any basis to know how much they would receive (and when), line ministries have had little means by which to plan basic operational and management needs. Evenifreleases were going to be below the monthly budgeted amount, line ministries often felt that it would be essential to know in advance what that amount would be so that they could adjust accordingly. 35Until2002, servicing o f the foreign debt was financed overwhelmingly by foreign balance o f payments (BOP) support and as such remained largely outside the cash budget process. Beginning in2002, ifBOP contributions fall short o f the amount expected, however, it can have an impact on the cash budget. -59- 5.8 The Bank's 2000 study showed that the cash rationing system was responsible for significant dislocations in spending priorities relative to those that had been agreed to in the Yellow Book. The study categorized ministries into three sectors based on their primary function - general services, social services, and economic services - and examined the proportion o f budget each received. It found, for example, that releases to the general services sector and the social services sector were substantially over budget (128 percent and 113 percent respectively), while releases to the economic sector were only 62 percent o f budget. (The overall execution rate was about 92 percent.) Thus, there were large variances between the budgeted and actual allocations. Although social sectors did relatively well, this could be explained by the IDA agreements by which the government committed to maintaining social sector releases at a minimum o f 36 percent o ftotal domestic discretionary releases. 5.9 Inadditionto the realignmentsacross sectors, the studyshowedhugechanges in the composition o f spending within the sectors - away from RDCs and towards personal emoluments. Inbrief, PESwent up inall three sectors due to a wage hike duringthe year, but RDCs increased only in the general services sector o f the budget - and there it skyrocketed to 177 percent o f budget. In contrast, RDCs for economic services were dramatically lower (-36 percent). Similarly, the analysis showed that releases for grants were very close to the original budget (97 percent). But this was the result o f two opposing trends - substantial over-spending by general public services, compensated by massive cuts for economic services (-47 percent) and minor cuts for social services. Giventhe apparent political pressures to increase general services while at the same time protecting social sectors, it is not surprising that economic services were left to bear the bruntofcuts. 5.10 Still, it is domestically financed capital expenditures that traditionally tend to suffer the worst. The study revealed that these were cut by nearly half - though not so much the purchase o f movable assets (cars, computers, etc.) (-9 percent) as cash releases for investment projects (-51 percent). There were also differences in finding o f capital expenditures across the sectors. Cash releases to the group o f general public services were largely intact, but slashed for social services and economic services (-3 8 percent). Impacton the Qualityof SpendingTimingof Resources 5.11 Not only has the cash budget impacted the annual amount that line ministries receive, but it has also led to a high degree o f uncertainty over the timing o f those allocations. While the actual releases may be made at the same time o f the month, the amount i s usually unknown to line ministries until they actually receive it. The Bank's 2000 study showed that aggregate figures on cash releases could mask the magnitude o f the swings being experienced by individual ministries. For example, Ministry o f Health RDCs should have averaged about K1.5 billion in 1998. But actual releases for M o H ranged anywhere from 0 to 4.6 billion per month. Grants for the same ministry ranged from 2.0 to 9.3 billion despite a monthly budget o f K4.8 billion. A similar story was found at the M o E and Ministry o f Agriculture. (See Table 11). Often shortfalls are compensated for by a largejump ina succeeding month. -60- Table 11: Cash Releasesto SelectMinistries for Key Expenditure Categories during 1998 K' billion Ministry of Health I Personal -- Ministryof Education RDCs Grants !Personal RDCs Grants Emolume. Emolumen. Expenditures 3.1 1.3 7.3 2.1 2.01 1.6 2.4 1.2 7.3 2.0 2.0 1.2 3.8 2.2 7.3 1.6 2.q 0.6 2.0 2.0 7.3 4.1 2.1 0.8 2.0 1.1 7.3 2.4 1.5 2.1 0.2, 2.0 0 7.3 0.3 0.7 3.7' 2.0 1.2 9.3' 7.5 2.2 1.8 5.8i 2.0 0.9 4.q 7.8 1.5 2.1 0.8, 2.0 0.4 8.8 2.6 1.1 0.5' October 2.0 4.6 10.8 3.0 1.8, 2.8 ovember 2.0 0.7 7.8 4.8 2.1 ecember 2.0 2.4 7.8 1.9 3.8~ Budget* 2.0 1.5 4.81 7.9 2.4 1.9 1.d ~ I 5.12 One might have expected to find some pattern o f predictability in monthly expenditure levels, but instead the Bank's analysis showed that the pattern was essentially random. For example, one hypothesis had been that there would be a link between monthly government revenues and monthly cash releases. However, a regression analysis for the years 1995-98 showed that only 44 percent in monthly variations in expenditures was explained by variations in revenues. In fact, government revenues fluctuated less than cash releases, particularly for RDCs, and those fluctuations were characterized by a seasonal trend - low during the first quarter, average during the second quarter, and gradually rising to a peak in December. The statistics further indicated that Government had been quite successful inprojecting annual revenues at the time the budget was prepared. During the first three o f these years actual revenues slightly exceeded budget projections and in the fourth year the actual shortfall was less than 0.5 percent. Cost to GovernmentProcurement 5.13 Inevitably, unpredictable funding will lead to higher costs o f procurement and inefficient operations as ministries try to adjust to their environment. For example, it may cause ministries to make purchases during periods or in sizes that are less economical than others. Bybuyingon supplier credit, many had to pay higherprices. Or they may simply make inefficient use o f their assets and staff because o f complementary inputs shortages, e.g., staff or assets may sit idle until other critical inputs can be purchased. 5.14 The cost o f unpredictable cash releases may be highest with capital projects. Because o f the structured work plans involved in carrying out physical projects, the -61- funding sources needto be able to support them. When contractors are notpaidandwork has to slow down, this obviously leads to higher cost, if not wastage or spoiling o f previously performed work. Within MFNP there have been various stories o f projects that have been impaired or cancelled because funding did not arrive when needed to support the physical progress on the project. Still, MFNP does no real cash planning in order to match cash needs throughout the year with the timingo f inflows. ActionsTaken to Date 5.15 The Bank has assisted the Government to reform the cash rationing process and to introduce greater predictability and transparency through the Fiscal Sustainability Credit (FSC). Under FSC, the Government agreed to develop a quarterly cash allocation plan to be communicated to line ministries prior to the start o f each quarter. The plan would detail the minimum release that ministries could expect for RDCs and Grants for each month of the quarter. This would apply to all ministries inevery sector, although it would have greatest significance for those in the social and economic sectors. As MFNP beganto gain sufficient experience with the quarterly plans, it was hopedthat they would eventually be able to move to a six-month allocationplan. 5.16 Through quarterly plans and subsequent correspondence to line ministries (LMs), MFNP agreed to guarantee ministries at least 80 percent o f their budgeted RDCs and grants for the quarter. The releases did not need to be the same each month, but MFNP needed to communicate what the amounts would be and assure that the sum was at least 80 percent o f budget. It was assumed that PESwould be released at 100 percent. By allowing RDCs and grants to be as low as 80 percent, it would provide MFNP with a measure o f flexibility to respond to unexpected demands or to meet seasonal needs that may arise for a particular ministry. Yet, it also provided a floor for line ministries to be able to continue to cover some basic operational needs. As a continuing measure to protect the social sectors, FSCl also required 36 percent o f domestic discretionary expenditures to be allocated to the social sectors. Hence MFNP aimed to fund the ministries o f Education and Health at 100percent o fbudget. 5.17 Since revenue collection might not follow the same pattern as expenditure needs withinthe quarter, the Bank helpedMFNPto establish a USD15 millionrevolving credit with BoZ. This was intendedto help cover the month-to-month fluctuations that would occur in revenue and still enable MFNP to meet its commitment to fund 80 percent o f RDCs and grants. FSCl also specified that ifa deficit were to be incurred inone quarter, then the next quarter's plan needed to be reduced in order to cover it. Moreover, it specified that if at any time MFNP were unable to respect the 80 percent commitment for ministries, that it should provide an explanation o f the reasons and the criteria used to reallocate funding it a different way. ExperienceImplementingthe Quarterly Plan 5.18 MFNPhas produced the quarterly cash plans since the third quarter o f 2000, but the actual implementation o f the plans has only partially fulfilled their purpose. Since the first quarter o f 2001, when the Bank first began to compare carefully the plan with actual -62- releases, the results have been fairly mixed. Discrepancies between planned and actual releases were generally greatest during the first quarter but have improved somewhat over the course o f the year. The first quarter plan proposed an across the board 80 percent guarantee for all ministries. Giventhat actual releases overall were substantially more than the total amount allocated inthe plan, many ministries received more than the planned amount. Yet, most o f those that received additional amounts were concentrated inthe general services sector, while some ministriesinthe economic and social sectors received amounts considerably lower thanplanned(Le., lower than 80 percent o fbudget). During the second and third quarters, however, the distributions o f cash releases were more balanced across sectors. 5.19 Analysis prepared by MFNP itself showed that for the full year 2001 there remained a bias in cash releases towards PESover other expenditure categories and towards administrative/general government sector over other sectors. (See Annex 3.) While the overall execution was 38 percent over the GRZ budget estimate, the largest increases were inPES,which were 48 percent over budget and the smallest in grants (15 percent). Across sectors the administrative / general government sector did considerably better than the social and economic sectors when it came to RDCs and grants. For example, RDCs for administrative ministries were on average 58 percent over the budget amount, but 20 percent below budget for economic sector ministries. With respect to grants the administrative sector benefited from a 38 percent surplus, while the social sector saw a decrease o f 3 percent. FutureAction Plan 5.20 The FSCl's recommendations have had two important purposes: (i) encourage to a more predictable stream o f funding to LMs, and (ii)to make that stream more consistent with the budget estimates. It has done so by encouraging development o f a quarterly plan that is communicated to LMs and by promoting transparency on the implementation o f the plan (through publication in the MFNP's Macro Indicators journal). As noted earlier, the results are mixed. Cash releases have become less volatile on a month-to-month basis. Ministries have received from MFNP more consistent cash release amounts even ifthey sometimes fell below the initial plan amount. On the other hand, the cumulative cash releases still show a considerable variance from the budget estimates. Inparticular, economic services ministries still tend to receive less and general services ministriesmore than budget. 5.21 Recommendations under FSCl should continue to have an important role to play inthe interimor transition period untilbudgetpreparation and execution are taken more seriously and fiscal discipline i s stored. In particular, the "requirement" to maintain hnding to LMs at a minimum o f 80 percent o f budget (for RDCs and grants) is a partial incentive to preparemore realistic budgets and to maintaintighter control duringthe year. Ultimately, LMs should receive closer to 100percent of their budget estimates, but this i s would most likely occur over a mediumterm horizon as a part o fbroader PEMreforms. -63- 5.22 Going forward, the Government needs to pay special attention to three broad areas for improvingthe cash release system: (a) Quarterly plans need to be based in an ongoing process of rigorous cash management for domestically generated resources. In other words, they should be driven not by a blanket 80 percent allocation for all ministries, but through analysis of the expected aggregate resources and the requirements o fministries . (b) MFNP should continue to encourage transparency in the cash release system bypublishingamounts received by each ministry. (c) MFNP should commit itself to fully implementing the planned budget, reinforced by the commitment o fthe political leadership to comply with it. 5.23 Annex 3 expands upon and discusses the challenges ineach o f these areas. Below is a summary o f the recommendations described therein. PossibleAction Steps: Assign institutional responsibility for cash management, including the development and updating o f quarterly cash plans in collaboration with revenue departments, Investment and Debt Management Unit, Economic and Technical Cooperation Unit, Accountant General, Planning and Economic Management, and line ministries. Develop quarterly cash allocation plans that are based on comprehensive forecasting o f the expected cash inflows for the quarter, analysis o f line ministries' cash requirements for the quarter, and Government's immediate policy priorities. The quarterly plans should function as expenditure ceilings for line ministries. Incorporate capital spendinginthe quarterly allocationplanbased on input from line ministries concerning contractual cash requirements. Demonstrate to line ministries that the plan will be implemented each month, i.e., actual cash releases will be consistent with the planamount. Provide some minimum guaranteed level o f monthly cash releases for RDCs andgrants (e.g., 70-80%) Develop procedures for how and when the quarterly planis updated (e.g., whether it should be a rolling quarter without revisions to the prior plannedmonths). Continue to publish information on cash releases in newspapers, or in an alternative media that assures access to the information by civil society. -64- Information should compare budget versus actual release for selected ministries and/or subheads. (h) Continue to monitor and report on deviations between the quarterly plan and actual amounts. B. REVENUE CONTROLS 5.24 Revenue collection i s one o f the few areas in the Zambian public finance system that works very efficiently. With a 1993 Act o f Parliament, the Zambia Revenue Authority (ZRA) was created and entrusted with the responsibility for collecting all government tax revenues, including income taxes, VAT, and custom and excise duties. The ZRA i s supervised by a Board whose members are appointed by the Government from the public and private sectors. The Board reports to the Minister o f Finance who i s accountable to Parliament for the performance o f the Authority. The Authority prepares annual financial statements which are finalized within three months of the end o f the year, audited by an independent firm o f auditors and laid before Parliament within 6 months o fthe end o fthe year. Also, its accounts are examinedbythe AG. 5.25 MFNP and ZRAjointly set the target o f revenues to be collected. The estimated revenues are included inthe annual budget that is laid before Parliament by the Minister o f Finance. The revenue collection is monitored regularly by MFNP through daily reports on Control Account 99 from the Bank o f Zambia, monthly reconciliation between MFNP and Zambia Revenue Authority (ZRA), and quarterly and ad-hoc reports from ZRA. ZRA has put inplace arrangements to ensure that all revenues collected inLusaka andKitwe are bankedpromptly with the Bank o f Zambia, while revenues collected from remote areas are banked promptly with participating commercial banks for onward transmission to the Bank o f Zambia within 4 - 14 working days. All revenues collected by ZRA are deposited in the government general revenue account. The source o f the ZRA's funding is the Government budget. There are no ad-hoc exemptions from taxes. Every exemption requires Statutory Instruments, i.e. Act o f Parliament that provides appropriate instructions on the exemption. Actions Taken to Date 5.26 Some o f the important steps Government has taken to ensure proper collection o f revenues and adequate monitoring o f the revenue collection process include the following: 0 ZRA is adequately staffed. It was set up as a government agency outside the civil service recruitment and remuneration arrangements. It therefore operates on a different salary scale from the civil service and i s able to compete effectively with the private sector for skilledandexperienced personnel. 0 The Authority has adequate material resources to perform its functions, although much of the fundinghas so far come from donors. -65- 0 ZRA is relatively independent and able to apply the full letter o f the law in collecting government revenues. It is, therefore, not bedeviledby questionable exemptions- the exemptions granted are backed up by statutory provisions. 0 The accountability arrangements for ZRA are reasonably robust. They include a supervisory board, reporting to Parliament through the Minister o f Finance, external audit conducted byprivate audit firms, and audit by the OAG. 0 The arrangements for transmitting revenues collected to the government account i s fairly adequate, given the state o fZambia's banking network. FutureActions 5.27 The Government needs to capitalize on the current strengths o f ZRA and ensure that the Authority can continue to operate efficiently. ZRA is adequately resourced at the moment due to generous funding from donors. However, there is a need to put inplace a mechanism that will assure adequate funding for ZRA ifand when donor funding stops. C. EXPENDITURE CONTROL 5.28 The expenditure control system in Zambia i s built around the principle that Control Officers are responsible for execution o f the budget head, and that they are held accountable by the Minister o f Finance for proper management. Significant problems arise when the expenditure rules are not enforced, and when Controlling Officers are not held accountable. Legal Basisof ExpenditureControl 5.29 The Constitution and Finance Act stipulate that: (i) shall not be expended moneys from the general revenues o f the Republic unless the expenditure is authorized by a warrant under the hando f the President; (ii) warrants shall not be issued by the president unless the expenditure i s authorized by an Appropriation (iii)M-MFNP shall appoint Control Officers for each expenditure head, who shall be responsible for budget execution and who shall be the Chief Accounting Officer o f the relevant expenditure head, and such Accounting Officers shall keep such books andrender accounts as may be prescribed, or as may be directed by the PS-MFNP; and (iv) the M-MFNP is ultimately responsible to produce annual financial reports and present them to the Parliament within nine months after the end o f each FY. 5.30 Inconformity with the Constitution andthe FinanceAct, the Minister ofFinance and National Planning appoints the Controlling Officers (CO) for each o f the expenditure 36Exceptionsto those rules are contained in article 115 (1) and (2). Further exceptions to the above rules are contained inthe Finance (Control and Management) Act and substantial authorities are given to the PS- MOF. -66- Heads (51 at present consisting o f ministries, provinces, and a few other organizations). Following the budget approval by the Parliament in April o f each year, and issuance o f the general warrant by the President, the Secretary to the Treasury will issue sub-warrants to each CO. The record keeping of these credits is done by the accounting units at the appropriate level: (i) Ministries' H Q in Lusaka for RDCs o f HQs and offices in at Lusaka; and (ii) provincial offices (Provincial Cos and accounting Officers) for all at departments located outside Lusaka. These sub-warrants are the basis for cash release. 5.3 1 The responsibility for budget execution, including creating o f commitments, managing o f contracts, and accounting and reporting to the MFNP rests with the appointed C O or hisher designated Accounting Officer. The process and steps are clearly described in the instructions, which are reinforced with the provision o f appropriate forms and formats are provided for guidance. The institutional arrangements and instructions rely on the segregation o f duties as the basis for control. Furthermore, the secondment o f Internal Auditors (IA) to each CO i s a necessary control mechanism. A simple cash-based accounting system is maintained by each spending agency to form the basis for financial reports. Analysisof the Problem 5.32 It is difficult to separate expenditure control issues from other parts o f the budget process. For example, it i s impossible to maintain adequate budgetary control without good budget format, reliable estimates, and timely and reliable historical data (produced from the accounts). Conversely, poor budget execution impacts adversely on the quality o f accounts and reports produced. For this reason, the issues raised in this section overlap to some degree with those found inchapter 6 on accounting andreporting. 5.33 There are several problems with expenditure control in Zambia. First, ad-hoc systems such as controls via cash rationing discredits the budgetary exercise and undermines the legitimacy o f expenditures. The control o f expenditure via cash rationing, although sometimes necessary in the short-term, does not always achieve its intended purpose. It relegates parliamentary control and oversight, and creates conflict between the MFNP and the executing agencies. It often contributes to the lack o f discipline and non-compliance with rules and regulation. Furthermore, it promotes the end-of-year practice o f fitting the actual data into the Appropriation Acts or visa versa (supplementary estimates to match actual data). 5.34 Second, many spending departments inZambia operate as if standing instructions on the cash control mechanism did not exist, although inmost cases they stay within the budget envelope. Inmany cases, orders are placed and payment vouchers are prepared and kept (often unrecorded) until cash is available. Once funds become available, a selected number o f vouchers are sent for payment, enough to exhaust the cash. Other vouchers, ready for payment, remain pending for the next release o f fund. The bulk o f these vouchers form the Government arrears. Because, many of them are unrecorded in the ledgers, the exact amount o f arrears is unknown. Ad-hoc measures are required for their identification and estimation. However, even with those ad-hoc measures, there is -67- no certainty about the exact amounts o f arrears. (See section C o f this chapter for a furtherdiscussion o fcommitment control.) 5.35 Third, besides the fact that some arrears are not accounted, thepractices described above raise two issues: (i) late payment penalties charged by suppliers and the impact on prices quoted by suppliers and/or quality o f goods; and (ii) the probable consequence on transparency in the selection o f vouchers for payment. Late payment penalties are explicit for civil works certificates but implicit inother invoices. Although the small size o f economy and limited market for suppliers can be the reasons for availability o f suppliers' credits to the Government, delayed payments are likely to have an impact on the price (or quality) o f goods, particularly when the commercial rate o f interest i s as high as 40 percent. With respect to civil works contracts, and other competitive based contracts, the late payment charges are added to the invoices. For a few invoices examined, the latepayment interests andother additions resultedinan almost 100percent increase in the payment for the works done. When controls are weak, this could also be used as means for financing kickbacks by pre-dating invoices (submitting invoice before the work i s done). The lack o f transparency in selecting vouchers for payment creates a rent seeking opportunity and could create an avenue for corrupt practices. (See Section F inthis chapter for furtherdiscussiononcapital expenditure controls). 5.36 Fourth, several other irregularities have also been observed: (i) funds released by the Budget Office were earmarked and the contracts (and implicitly the contractor) for which funds were released were identifiedby the Budget Office in several cases; (ii) in one o f the spending agencies, checks were issued before delivery o f goods and those checks were delivered to the procurement staff (or stores) for exchange against goods; and (iii) in some cases overdrafts have been used by the spending agencies to pay for expenditures over and above the budget. Those cases relate to the influential spending agencies such as the National Assembly, Ministry o f Defense, the President's Office, and few other agencies. 5.37 Fifth, the control mechanisms are based on segregation o f duties, and depends on full compliance with the rules. Changes have occurred that made the controls ineffective, among which i s the increased volume o f transactions. Staff are not well trained and, although manual and guidelines exist, staff do not have copies for reference. Intemal audit i s weak. The main work performed by the IA consists o f pre-payment audit (or vouching). Inadequate systems audit or risk assessment takes place. While the quality o f IA staff at the MFNP seemed good, IA staff at the sector ministries' level seem to lack rigor in their work, includingreporting. For example, working documents used in their work were not available.37 (See chapter 6 section D for further discussion on Intemal Audit.) 370neo f the tasks that recently has been given to the I A s is to produce the amount o f arrears. Inone case that we verified, the amount given by the IA did not match the amount produced from vouchers by the accounting staff. There was no backing o f data produced by IA while the accounting staff had the full information about each voucher and the totals provided. -68- 5.38 Finally, the financial management staffing i s generally inadequate. A team o f well qualified and dedicatedmotivated staff needs to be inplace to ensure that rules are applied as designed and that adequate controls are maintained over expenditures. The incentive system needs to be improved to attract the right caliber o f staff. (See Chapter 8 for a fuller discussiono fhumanresource issues.) FutureAction Plan 5.39 The purpose of expenditure control is to provide reasonable assurance that public funds are spent inaccordance with parliamentary authorization andthat the set objectives are achieved. To this end, it i s crucial that measurable objectives are clearly set and that the approved budget is respected. In order to address the current weaknesses in the expenditure control process, the following measures should be considered: Strengthen and modernize the Internal Audit (IA) arrangements by adopting international standards best practices, and improving the capacity o f IA in all ministries. (See chapter 6 section D for further discussion on how to strengthen InternalAudit.) Carry out a comprehensive review o f business processes in order to improve the control system, inparticular the commitment control, and data integrity. Provide specific directives on intemal control to improve staff and strengthen staff training on internal controls, especially for restructured institutions so that the control framework remains intact. Enforce existing rules and sanctions. Prepare and execute aplanto repay all outstanding arrears. Tighten the legislation in order to prevent loopholes in executing expendituresoutside the budget approvedbyParliament. Strengthen parliamentary oversight and control by way o f regular reporting to Parliament by the Executive, and adequate arrangements by Parliament to evaluate and act on the financial reports. D. COMMITMENTCONTROLAND THEARREARSPROBLEM Analysis of the Problem 5.40 The accumulation o f expenditure arrears i s a persistent symbol o f the public expenditure management system's ineffectiveness. Inits 1997 report to the Government, the IMF devoted considerable attention to the problem o f arrears. It noted that MFNP had set aside in 1996 about K15 billion to meet the stock o f arrears assessed through 1995. It also went on to recommend an audit o f the existing stock of arrears and measures to control future accumulations. Yet, over time things seemed to worsen rather -69- than improve. In 1999, MFNP had to devote nearly 9 percent o f budget resources, or K121 billion, to the clearance o f arrears. Although the stock o f arrears fell for a brief period afterwards, they have again climbed dramatically. Estimates for end-2002 suggest that the total arrears may be approaching 400 billion kwacha. 5.41 There are a number o f factors that contribute to the commitment control and arrears problems, many o f which have been noted by IMF missions. First, there have beeninstitutional weaknesses and lack o fcapacity to fulfill accounting requirements. For example, there had been concem that many o f the accounting staff lacked clear understanding about the commitment process and what exactly constituted a commitment. Inresponse to this, an IMFresident advisor developed a training course for identifyingcommitments, accurately completing the monthly outturn reports, and general principals o f accounting. The training, which covered accounting staff in every line ministry, was viewed as useful and served to highlight some o f the deficiencies in accounting knowledge that had existed. The advisor also noted the lack o f appropriate writtenguidance on the policies andprocedures to be executed bythe accounting staff. 5.42 Secondly, some arrears may be a consequence o f the variation between budget amounts and the actual cash releases. The budget estimates are not intended to represent a guaranteed minimum but rather a maximum expenditure, and macroeconomic circumstances can dictate the need for adjustments. However, some line ministries may simply insist on committing the full amount authorized inthe budget, or they may have implemented work plans based on the full budgeted amount. The annual warrant does authorize ministries to spend up to the annualbudget amount, but the cash release process has beenMFNP's main tool for limitingspending. Giventhe ad-hoc nature by which the cash rationing process works, it i s not surprising that ministries may incur commitments that are within the budget estimates but in excess o f the cash release by MFNP. There has been no analysis to assess to what extent arrears may be created for recurrent expenditures that are within the original amount budgeted. 5.43 A third factor contributing to potential arrears is the structural under-budgeting for activities at the beginning o f the year. Ina 1997 report, the IMF noted that "In areas like education, health, and prisons, the need to provide food for stafflresidents has led to continued orders from suppliers even though the cash allocations were inadequate to meet them." Consistent with this, a Bank mission for preparation o f the 2001 PER met health ministry officials who suggested that this was likely to be occurring at the University Teaching Hospital as well. They explained that some necessities o f maintaining patient care meant that they had a cost structure that was rather inflexible and could not be reduced ovemight. Other ministries are also confronted with budgets that are below the fixed obligations that they have. Under these circumstances, ministries often rely upon supplier credits to continue to obtain goods and services despite their shortage o f cash. Because o f weak customer demand in the private sector, there are suppliers who are willing to take chances with the Government inorder to make a sale. 5.44 Fourth, over-commitments may reflect an attempt by some ministries to increase their budgets and to thereby initiate new activities that were never budgeted for. Rather than making difficult trade-offs, some ministries may choose to over-commit with the -70- anticipation that they can receive additional finding from MFNP to meet their requirements. Given the flexibility offered to MFNP bythe cash rationing process, over- commitment becomes a viable strategy for some ministries. Indeed, the past training seminars have suggested that large levels o f outstanding commitments by a given ministryoften had little to do with the qualifications or expertise o f its accounting staff. Instead, there were institutions with very capable staff who readily understood that they were entering into commitments that exceeded their budget. Moreover, some o f these excess commitments may never turn into arrears because MFNP responds to the pressure byincreasingthe cash release for that ministry. 5.45 Yet, the largest and most costly source of arrears tends to be from capital projects managed by the Ministry o f Works and Supply. It i s not difficult to understand how these arrears accumulate and the danger that lies in them. As noted earlier, the releases for capital expenditure have generally been given the lowest priority in the cash budgeting process. Moreover, the timing o f the releases is generally erratic, as there is essentially no cash planning done by MFNP. Consequently, arrears build up for nonpayment on contracts where work has been performed. Inaddition, the lack of payment leads to the accrual o fpenalties and interest that inturnmay create new arrears. Interest rates tend to be high and are generally negotiatedby the contractor, without any fixed parameters used by government. Although there are no up-to-date estimates as to what proportion o f Works and Supply's total arrears are due to these types o f interest, penalties, and other non-payment o f signed contracts, the IMF estimated it could be as highas 75 percent in1997. 5.46 The organizational structure o f Works and Supply contributes to the problem. Those responsible for payment o f contracts for the ministries various departments do not work with or review the contracts themselves. Thus, they are unaware o f the terms o f the contracts and the likely penalties that may apply. Consequently, even ifcash rationing i s necessary with the Works & Supply Ministry, there i s little basis by which to arbitrate how that cash should be used. Likewise, it would be difficult for the accounting staff to accurately calculate the extent o f its liabilities at any point in time when the contract terms and conditions are managed by the project staff who are organizationally separated. ActionsTaken to Date 5.47 The IMF and the Bank have been fairly persistent in recent years in proposing actions to address the problem o f over-commitment - as a step to preventing arrears. While one component has been training o f accounting staff, a substantial effort has also gone into recommending procedures to bring greater control over commitments. For example, pursuant to an earlier JMF proposal, MFNP created a Commitment Monitoring Unit(CMU). This unit, initiallycomprised ofstaff seconded fromthe BudgetOffice, the Accountant General's Office, and Internal Audit, was charged with reviewing monthly commitment reports from line ministries. 5.48 MFNP has also issued directives to line ministries clearly outlining limits on the amount o f commitments that they could enter into. Beginning in 1998 MFNP implementeda requirement that outstandingcommitments should not exceed 20 percent -71- o f the (monthly) budget provision without consent o f MFNP. Before the end o f that year, MFNP had tightened the rule to a maximum o f 10 percent. Monthly reports on expenditures were modified to permitmonitoring o fministry compliance. 5.49 Among the problems with MFNP's 10 percent rule was the lack o f enforcement. Although MFNP has at various times issued letters to controlling officers remindingthem o f the rule and the sanctions that could result from violation, the problems persisted. The most recent examples o f this have been inthe context o f the Bank's FSCl. Since the 4th quarter o f 2000, the Government has written controlling officers who were in specific violation o f the 10 percent rule and informed them that without a written justification, their future cash releases risked being withheld. Still, ministries continue to over- commit. Although officials from the Accountant General's office indicated that there have been cases where they have taken actions against Controlling Officers, these seem to be extremely rare. 5.50 Part o f the challenge in enforcing commitments is that the Expenditure Monitoring Unit (formerly the CMU) is highly dependent upon the quality andtimeliness o freporting from line ministries. For example, during2001 it was not unusual for reports to arrive late and containing numerous errors or gaps. The Accountant General's office is obligated to follow-up with those that have delayed or where there are inconsistencies in the reporting. Inaddition, those responsible for reviewing the commitment reports have been typically detached from the office deciding the cash release. For this and other reasons, the commitment reports have generally not been an input into cash budgeting decisions. 5.51 Some o f the most encouraging actions have occurred most recently, at least with respect to identifying the stock o f arrears. By improving the quality o f quarterly audits o f line ministries and having controlling officers sign "certificates o f acceptance," the Internal Audit Department has taken steps that could significantly improve the reliability o f arrears estimates. The certificates o f acceptance, for example, reduce the likelihood o f line ministries "finding" new bills at a later date. The Internal Audit department also had plans to construct a database o f individual arrears in 2002, including details on the original due date, name o f the payee, and a budget code to which it corresponds. By 2003, it was hoped the database would be fully operational and that it would include a process for recording the clearance o f individual arrears as MFNP releases funds to line ministries for that purpose. Though delayed, the arrears database has now been created and is inuse. FutureAction Plan 5.52 Actions needed to address arrears should target four areas: (a) inadequate policies andprocedures for limitingcommitments, (b) ineffective monitoring and enforcement, (c) flawed budgetpreparation, and -72- (d) inadequate monitoring o f liabilities as they are created for capitalprojects. 5.53 Although there are some immediate steps that can be taken, progress in each o f these areas will generally require a medium term horizon. Below we examine the kinds o f actions that would be appropriate for each area. 5.54 Policies and Procedures: Effective expenditure control depends inpart upon line ministries having timely information on the level o f commitments at the time they are entered into. A workable set o f operational procedures are a critical element in making that possible. The IMF began providing assistance to MFNP in 2002 to implement a commitment control system (CCS) in line ministries. Under this system, a requisition would be prepared and examined by the ChiefAccountant before a commitment could be entered into in order to confirm that sufficient uncommitted funds were available under the particular line item in the expenditure ledger. If funds are not available, the requisition is refised and no commitment is entered into. On the other hand, if uncommitted funds are sufficient, the Commitment Control Officer (normally the PS or hisher designate) would approve the requisition and enter the new commitment in the ledger. Payment vouchers would therefore only beprepared ifproceeded by an approved commitment requisition. 5.55 Over the mediumto long term, commitment control procedures can be facilitated through the IFMIS. Moving from a manual system, new commitments can be blocked automatically in the system once they reach a specified threshold. Likewise, payments would only be authorized inthe IFMIS ifproceeded by an approved commitment. 5.56 For either the manual or automated system, however, unambiguous expenditure limits need to be communicated to line ministries. These limits should not be limited to cash on hand, or to merely one-twelfih o f the Yellow Book estimate. Instead, MFNP must be able to provide some margin for planning that would enable line ministries to enter into commitments prior to a cash release but based on a clear expectation on the amount and timing o f that release. As detailed in the prior chapter, MFNP should develop a quarterly cash allocation plan that essentially becomes the short-term implementation agreement for the budget estimates. Whatever modifications in the budget- whether supplemental appropriations or sequestrations - can be reflected inthe cash allocation plan. But whatever the amounts in the plan, they should serve as the upper limits for how much line ministries can commit to during the quarter. So for example, if the plan provides for a ministry to receive K50 million per month from January to March, then as o f January lSt they may enter into commitments totaling K150 million. Notwithstanding the commitment amount, the actual delivery o f the goods or services might be more staggered during the quarter to take into account the fact that cash releases neededto make payment still occur monthly. 5.57 Monitoring and Enforcement: It is not unusual for Zambia to have good regulations/policies and poor compliance with them. To remedythis, it will be important for MFNP to maintain good and timely information on commitments and to enforce its policies limiting them to a pre-specified amount. A reinvigorated Expenditure Monitoring Unit (EMU) should be responsible for monitoring the receipt and accuracy o f -73- the monthly expenditure reports from line ministries. The EMU should also be responsible for presenting a summary report to the Secretary to the Treasury each month detailing where over-commitments have occurred and/or where reporting has been inadequate to determine over-commitments. 5.58 Effective enforcement o f policies and procedures i s the second critical step. Where commitments exceed the allowable limit, MFNP must be more aggressive in investigating the reasons, and where over-commitment is persistently unjustified, it should take action to sanction the ministry and/or controlling officer. Ifthe obligations o f a ministry are such that it will require more than is allocated inthe plan, then this should be addressed at the moment the quarterly plan i s formulated and approved. Likewise, virements between subheads would require prior written approval from the Secretary to the Treasury. Or if there are legitimate emergencies these should be addressed up-front, with appropriate Cabinet and/or Parliamentary authorization. 5.59 Budget Preparation: Where approved budgets have been unrealistic and failed to take into considerationthe ongoing obligations or fixed costs o f a ministry, it i s inevitable that this will put pressure on commitment ceilings. The structural over-commitments in the budget- discussed in detail inthe budget preparation chapter o f this report - will not be solved overnight. Yet, MFNP can still do more to reconcile ongoing commitments with budget amounts. For example, where there is under-budgeting this should be followed by amendments to the budget, and the commitment ceilings should be adjusted subsequently -upward for some ministfies and downward for others. 5.60 Capital Project Management: As was noted earlier, some arrears accumulate in the context o f investment projects because there i s a lack o f coordination between the physical progress on projects and the payment process. MFNP and the various departments o f Ministry o f Works and Supply need to coordinate to ensure that there is timely payment o f contractual obligations. Some responsibility lies with MFNP and can be improved once MFNP begins a serious effort at cash planning as discussed in the preceding chapter. However, within Ministry o f Works and Supply there must also be better coordination between those responsible for management o f the contracts and those responsible for payment. It should be noted that this coordination has recently begun to improve as a result o f implementation o f a commitment control system. The Bank's CFAA work in other countries has shown that there are often incentives for government officials to collude with suppliers to create arrears and the accompanying payment penalties. Although there i s no evidence to indicate whether this occurs inZambia or not, it is critical that the authorities take steps to establish firm policies on the management andpayment o fcontracts inorder to close any loopholes that may exist. PossibleAction Steps: (a) Continue to enhance implementation o f the commitment control system (CCS) so that commitment requisitions are submitted, approved, and recorded inan expenditure ledger before payments can be authorized. -74- Develop credible quarterly cash allocation plans, and establish a policy that commitments by line ministries will be strictly limitedto the quarterly amounts found in those plans. The EMU should be responsible for verifying this through the monthly expenditure returns and providing analysis to the ST on over-commitment levels. Take action to sanction ministries and/or controlling officers that persistentlyover-commit. Include in system requirements for IFMIS features that prevent payments from beingissuedwithout prior commitment andverification. The system should also set limits on the level o f commitments. As discussed in Chapter 4, begin developing budgets that are more realistic, and that take into consideration the on-going cost structure o f line ministries. Establish a procedure whereby commitment ceilings are adjusted to take into account supplemental appropriations and sequestrations that occur duringthe year. MFNP should establish a quarterly cash allocation plan that explicitly takes into consideration the contractual obligations that are likely to come due on investmentprojects duringthe period. (See Chapter 4.) E. PAYROLLCONTROLS 5.61 Because o f the sheer magnitude o f the budget devoted to personnel expenditures, they represent a significant area of financial risk if not managed well. In 2001, 37.5 percent o f budget releases were for personal emoluments (IC907 billion). The areas o f concern within payroll expenditures basically are o f two types. First, there are management and control issues that are associated with the actual transfer o f funds to pay individual civil servants across the country, i.e., how people receive their pay and whether the funds actually reach them. The second area is more upstream, and concerns management and control of the personnel registry to ensure that it is accurate and that only current employees are targetedto bepaid. Analysis of the Problem Payment of Personnel Emoluments 5.62 For a country with a relatively large landmass, and a relatively sparsely populated countryside, there are some specific challenges that arise for making sure that payments reach individuals. Zambia's financial administrative network covers an area o f some 750,000 square kilometers broken up into 9 provinces and 72 districts. The farthest provincial capital from Lusaka is approximately 1,100 kilometers, a 12-hour drive. In -75- some districts, there is not access to all o f the schools and health posts year round. A key element inthe implementation o fZambia's payroll controls i s the use o fbank accounts to transfer funds and to effect payments. However, out o f the 72 Districts that form the financial administration network, 26 districts do not have any banks. One consequence o f not all districts having bank accounts is that some ministries make payrollpayments in cash. 5.63 For employees working in provinces, MFNP releases cash directly to the provincial headquarters. The mechanism for making the transfers is based upon depositing consolidated sums in banks located in the provincial capital and having the bank facilitate the transfer to the district bank branches. The deposits are accompanied bypayroll schedules prepared by the Data Centre (a unit o fthe Ministryo fFinance). On this basis the bank transfers the salaries and allowances to the corresponding staff bank accounts. Of course, as noted above not all districts have banks, and consequently not all staff members have bank accounts.38 In such cases, the salaries and allowances are thus cashed by a designated person at the district and passed on to staff members. While this mechanism may not be problematic in and of itself, when coupled with inappropriate reporting controls and audits it may serve to encourage delay inreporting the termination o f a given staff member. 5.64 The challenges in executing PE payments are not limited to these 26 districts. Even though most o f the 5,000 pay points within the Ministry o f Education are located in Districts with banks, some o fthose would be still be far enough from a bank to justify the adoption o f cash payrollreleases. Monitoring and Control of Payroll 5.65 Regardless o f whether PE payments are made in cash or through the banking system, there must be an ongoing and effective system o f monitoring and reporting o f who has been paid. This i s not currently the case in Zambia. While the Data Centre provides payroll listings segregated by pay point, these are distributed only to the provinces. Consequently, payroll unit heads (senior officers at the level o f a pay point) have no way o f ascertaining who o fthe staff within their unit received salaries for a given month. A staff member, having left without notice, could well continue to receive salaries without the payroll unit head being aware o f it. Interviews with several line ministries, however, indicated that there i s nothing that would prevent the distribution o f payroll lists to the paypoint andthat this could be implementedimmediately. 5.66 The existing methods o f monitoring PE releases are ineffective in providing sufficient controls. For example, the monthly expenditure returns, with respect to the 38TheMinistry o f Agriculture transfers all personal emoluments directly into bank accounts, w h c h requires that some o f its staff have to travel long distances to collect their salaries. Apart from the high cost this introduces for employees because o f the large distances involved, it also negatively impacts productivity as staff take o f f work time to pick up their salaries. In the case o f the Ministry o f Education, payroll is collected incashby a representative inthose districts without bank accounts. -76- salaries, is merely a measure o f releases, determined as a sum o f the headquarters and provinces releases, rather than an accounting determination o f the actual expenditures made out to staff. A review o f some monthly expenditure returns showed flat releases each month to the line ministries. This is counter intuitive because for thousands o f staff in a given line ministries one would expect that terminations (retirements, death, resignations, terminations without notice, new hiring) will result in fluctuating monthly payroll expenditures. A flat release per month suggests that payroll releases are not informed by actual expendit~re.~' 5.67 Other reports, such as the monthly allowance requests and the Deviation Report, also fail to provide safeguards against "ghosts" or on unreported separations, which potentially constitute a major source o f loss. The latter, for example, focuses on flagging substantial deviations in individual pay level and even this i s limited in effectiveness by the complex and extensive system o f allowances, which make such variations a frequent occurrence. Monthly staff returns are submitted (sometimes only quarterly) by the Human Resources Departments, but often too late to be useful as a reconciliation document for accountingpurposes. 5.68 The Data Centre, which is responsible for running the payroll, uses the Master Personnel Database File as a control file for payroll expenditure. The level o f inaccuracy o f the Master Personnel Database File, in terms o f reflecting precisely the staff strength, directly impacts the level o f losses in payroll expenditures. Accuracy is not merely a function o f well-designed, fully secure software, but also o f ongoing (monthly or at least quarterly) full assessmentso f the staffing strengths Le., a confirmation that the payroll list reflects all the active personnel, andonly the active personnel. 5.69 One o f the ways "ghosts" may be introduced is through inefficient recording o f personnel separations so that payments continue long after the person has ended their services with the Government. When terminations, study leave, retirements, deaths occur they are supposedto bereportedthrough a chainvia the district administration, provincial administration, and finally to the ministry's headquarters. Prior to deletion o f such a record off the payroll, the ministry's headquarters (or sometimes staff from the Data Centre) must first have the deletion approved by the Public Service Management Division. While this is a critical control step, the ad hoc nature o f reporting separations make applications for approvals come up on an ad hoc basis rather than as part o f a structured process anticipating 400 to 1,000 separation approvals per month, which adds further to delays. Given some o f the challenges outlined earlier with regards to the financial administrative network in Zambia and the number o f personnel in some o f the line ministries, it i s the inability to perform regular ongoing full assessments o f staffing strengths that contributes to substantial payroll losses. 391n interviews with line ministry staff they suggested that the flat payroll releases observed from month to month occurs because while the salary component does fluctuate the allowance which are in arrears far exceed the excess o f release over salary. Consequently the excess always goes to fund allowances. This suggestion that monthly allowances are only partially paid brings into question the very mechanism for paying and accounting for allowances and suggests that this is an area which requires further investigation. -77- 5.70 One o f the controls implemented by the Accountant General is to request for all banks to provide to line ministries lists of civil servants whose accounts have remained inactive for two consecutive months.40 In one example, a letter from the local bank branchto the Provincial Education requests that the ministry stop transferring salaries to a list o f ten names who are either deceased or do not maintain accounts with the bank. What i s revealing about this letter i s that the notification o f termination through death i s more efficient through the bank than through the internal payroll reporting system o f the Ministryof Education. This is despite the fact that the ministrypays burialcosts, giving families an incentive to notify the ministry incase o f death. Inaddition, the letter reveals that there are errors inthe direct transfer o f hnds to bank accounts since it points out that some of the names listed do not maintain accounts with the bank. Inall probability, the deficiencies identifiedinthis review are present throughout all o f the line ministries. 5.71 Currently, the payroll management system may be described as passive, in that the management system awaits notification inthe case o f employee separation or transfer. It has assumed a posture o f "no news is good news," ifthere are no reports o f separation then there must be no separation. It assumes that controlling officers can and do check the number o f active employees against the establishment register. But we note that the controlling officer without access to payroll lists cannot check the number o f people paid against the establishment record. In the current system, the Controlling Officer has no basis to operate the establishment control. 5.72 The internal audit process o f using head counts to check on the presence o f "ghost workers" i s hampered by having too few staff for an effective sample. For example in the case o f the Ministry o f Education, currently with only four internal auditors assigned to the provinces to cover all internal audit assignments, and with over 5,000 pay points, it is unlikely that more than 1percent or 2 percent o f the pay points can be audited each year. That level o f audit sampling rate provides little disincentive to persons determined to fraudulently milkthe payroll system. The internal audit process i s further hampered by the lack o f up to date information on the employment status at different pay points. Making a proper reconciliation o f head counts with payroll lists i s made extremely difficult ifsuch payroll lists are inaccurate and outdated. 5.73 The current practice o f relying on the banks for implementing expenditure controls ought not be considered a key control. While cooperation from the banks i s critical inmaintaining the control, such cooperation cannot be made a substitute for direct payroll expenditure controls. FutureAction Plan 5.74 The number of transactions (109,000) and the financial administrative network bottlenecks constitute a formidable challenge to effecting releases that are truly reflective 400f course, such a control is irrelevant to employees who arrange to continue to have withdrawals made. Nor does it impact those who are paid incash. -78- o f the employment services procured by the government. The aggregate releases to the provincial level currently in practice in Zambia, without proper timely feed back on actual expenditures o f these releases, assures a continual hemorrhaging o f government resources inpaying staff (who are no longer employed). It is important to recognize that this problem cannot be eradicated by undertaking a one time national head count. The source o f hemorrhaging is rooted in a poor system o f monitoring and reporting coupled with an inefficient reporting system o f staff who have separated, especially those who leave without notice. Therefore, given its systemic dimension, even after a successful one time massive nationwide "ghost" removal effort the problem would reemerge over time. 5.75 There are three principal recommendations: To ensure responsibility for checking and confirming the reconciliation between payroll lists (by pay point) and the people actually working on a regular basis, the senior pay point officer should be requiredto check and sign o f f on the payroll lists each month and return them to the provincial headquarters. These verified payroll lists would serve as the basis for preparing deletion and transfer lists to be submitted to the Data Center, after endorsement from the Human Resources Department, in order to update the payroll master file. This step would require information flows in a reverse direction over the financial administration network. It is apparent that information could flow well in both directions between the provinces and the headquarters; but is not obvious that reverse information flows between the pay points and the provincial headquarters on a regular and ongoing basis is currently feasible. Prior to implementing such a monitoring scheme, it shall be necessary to assess the feasibility o f submitting these signed payroll lists from all the pay points to the provincial headquarters. Use the signed lists above as the basis for performing internal audits. It would immediately facilitate holding the pay point senior officer responsible for any discrepancies found between head counts and signed payroll lists. The level o f accuracy would also permit reliable inferences on staff transfers, which i s currently often the presumed excuse for head count discrepancies. Currently, i s no direct establishment register control in Zambia. It is presumed that controlling officers will manually check and ensure that the employees do not exceed the establishment register. In practice though, the controlling officer does not have access to accurate, ongoing information on the precise status o f payroll lists to make such a check. It i s recommended that specific posts with unique identifying codes be developed for each and every single post within the ministry as a way o f implementing the establishment register as a control directly into the payroll master database. The budgetary process would include the approval o f these specific posts, which would then be filled. -79- Consequently, no new hires could be paid against a specific post while an existing post i s still being paid. This would ensure that the terminated employee would have to be taken o f f the payroll before a new one could be put in and in so doing would prevent the continual payroll upwards creep observed in Zambia. The creation o f new posts would have to be highly controlled and only permitted as part o f the budget preparation process. F. CAPITALEXPENDITURECONTROLS 5.76 Earlier sections discussed the problems in commitment control and arrears build- up in general, as well as the specific contribution that capital expenditures make to that. This section examines in greater depth some o f the characteristics o f financial management for capital expenditures, and highlights how they may impact the overall efficiency and effectiveness o f those expenditures. What is relevant here are the nature o f the contracts, the amount o f information and reporting that takes place, the physical verification, and the management o f payments. In summary, it was observed that the structure o f the contracts themselves coupled with the lack o f rigorous record keeping and management has significant financial penalties for Zambia. Cash management is highly detached from the commitment process and important information needed to inform the cash management process i s often missing. Budgets themselves are unrealistic for the level o f resources that are needed or actually released. Without better information management, stronger internal audit, and aggressive cash management approaches, capital expenditures will continue to suffer from inefficiency losses. Analysisof the Problem 5.77 Government experiences financial losses on capital projects when it pays more than it should or when work is not completed to the standards agreed to. While much o f capital spending i s managed directly by donors through independent project units, a significant portion o f capital works i s nevertheless funded by government. In 2001, budget releases for capital projects totaled K300 billion (or 12 percent o f all releases). Although only a small portion o f the total budget, what makes capital projects a concern for fiduciary purposes i s that the magnitude o f the losses on any given project can be quite significant. Some estimate that capital projects in Zambia may be costing twice as much as they should. 5.78 Two central problems faced with capital projects i s that budget releases are insufficient for the Ministry o f Works and Supply (MWS) to cover the obligations the Government has incurred, and the lack o f comprehensive and transparent accounting inhibits a careful prioritization o f cash needs. Moreover, the structure o f the contracts is such that if not managed aggressively the overall costs can skyrocket. There are several contributing factors. 5.79 First, the existing pipeline o f projects tends to have annual costs that exceed the amount that i s budgeted by MFNP. According to MWS, even if there were no new projects added (and there are a few) the budget allocations would be insufficient to cover -80- the work envisioned during the year by the project schedules. Consequently, M W S has to decide how to slow down progress on projects in order to fit within the available resources. Obviously, there are limits to how much a particular project can be slowed before it becomes very inefficient and wasteful. M W S indicates that insome cases, only enough money is budgeted for the advance payment to be made, but not for any work. The scarcity o f budget resources places an ever-greater importance on careful project prioritization. In other words, budget allocations themselves need to be informed by the specific structure o f the project and whether work can be slowed in a way that makes the project still economically viable. It i s not apparent that this i s always the case in the current budget preparationprocess betweenMFNP andMWS. 5.80 Secondly, as stressed earlier, budget releases often vary significantly from budget estimates, and domestic capital expenditures tend to be one o f the lower priorities during the year. As a result M W S may accumulate arrears or work maynot be performed as was scheduled. Because of the exploding level of arrears, the interest costs can be quite high. MWS has also said that the break down between contract payment and interest is approximately in a 1:l ratio. For some projects, Government i s only making interest payments andnot payments on the principal. 5.81 An important question is how the cash releases for individual projects are determined, and whether the process can be managed in a way that minimizes the impact on aggregate project costs. Under the current system, MFNP decides on cash releases to M W S with incomplete information to make economically optimal decisions. MWS sends to MFNP reports on the outstanding amounts due on projects, but other vital information are lacking. MFNP decisions should be informed by the amount o f current charges versus arrears so that an appreciation may be made o f the potential interest cost and penalties o f late payment. MFNP should also be able to know the upcoming cash needs for projects and the relative priorities as indicated by MWS. In some cases, contractors have opted to go directly to MFNP for securing payment so that when MWS receives a release, it is earmarked to that specific contractor. As a result, other needs which may have been o f greater urgency, get delayed. 5.82 Some o f the challenges indeveloping cash flow forecasts are consequences o f the information flows within the MWS. While MFNP obtains copies o f contracts that have been signed, the financial management units (FMU) within the line ministry are often unaware that such contracts have been entered into until the first request for payments o f interim certificates i s made. Within the line ministry, the project engineer typically keeps the copy o f the contract, and hence there i s no centralized repository to reference the financial elements o f a contract and to ensure that the FMUhas a comprehensive listing and up to date status o f all o f the capital expenditure contracts and variations. Thus, financial management within M W S i s limited to the monitoring o f payments o f interim certificates with no broader perspective on the outstanding contract works, rate o f work completion. Hence, they have absolutely no basis for preparingpro forma cash flows, or assessing the impact o f inflation, currency fluctuations, or late payments. Further, there i s only a limitedbasis upon which to apply controls for claims on price variations or late payment penalties, since the FMUi s often unaware o f the full terms o f the contracts. -81- 5.83 The contracts themselves are structured in such a way that unless the FMU i s carefully monitoring, there may be opportunities for abuse and for substantial overpayment. Typical capital expenditure contract includes clauses for: a Contract works variations (not to exceed 25 percent o f the total value o f the contract); e Price variations, which in an inflationary environment typically exceed 20 percent; e Currency rate fluctuation variations; and a Late payment penalties, which are applied at commercial bank interest rates. These are currently at 45 percent, thus this factor very quickly becomes a significant component o f capital expenditure. 5.84 The fact that Zambia incorporates both price variation clauses and foreign exchange clauses may be permitting some contractors to be over-compensated when price fluctuations are linked with foreign exchange rate changes. Moreover, by allowing price fluctuations to be awarded on a discretionary basis may open a window for abuse. 5.85 Finally, the monitoring and evaluation o f the physical progress i s also important to preventing financial loss. However, the internal audit department of the Ministryo f Works and Supplies, with four auditors responsible for well over 100 projects, i s understaffed to perform meaningful materiality audits o f the projects. Without any engineers assigned to the units, it is unlikely that any independent checks beyond the input from the project engineer on-site is achieved. The current level o fproject audits at the Ministry o fWorks and Supply is inadequate. FutureAction Plan 5.86 Inorder to addressthe weaknessesidentified above, the following actions need to be taken: (a) Develop a regular expenditure reporting mechanism between MWS and MFNPthat segregates: (0 PrincipaVOriginalContract Value (ii) WorkVariations (iii) PriceVariations/CurrencyFluctuations (iv) Late Payment Interest Penalty Institute regular cash flow forecasting from M W S to MFNP to inform cash release decisions. -82- (b) Develop a database and reporting structure that distinguishes for each project: (i) Total contract value, budgetedyears to completion (ii) Proportion o fproject completed, years already undertaken (iii) Proportion o fproject to be completedincurrent budget year, (iv) Future outstanding contract value, number o f years beyond current year to completion (c) Reviewthe practice o f allowing bothprice and exchange rate variations in contracts. (d) Ensurethat the FMUmaintains copies o feachproject contract. (e) Reinforce the capacity o f internal audit to conduct materiality audits. G. PUBLIC PROCUREMENT4' Background 5.87 Addressing the weaknesses inbudget execution andbudget control will go a long way inkeeping public expenditures under control and contributing to a sustainable fiscal policy. But, as the chapter on governance pointed out, Zambia needs to go beyond public expenditure management issues if it wants to restore both public confidence and the confidence o f the international community in accountability. Development o f a well- fhctioning procurement system, based on transparency, competition, economy and efficiency, fairness, and accountability is a critical step inthis direction. 5.88 A Country Procurement Assessment Review (CPAR) was conducted for Zambia inJanuary 2002 and inMarch2002 by a team comprising officials from the Ministry of Finance and National Planning, a multi-disciplinary Task Force representing various concerned organizations set up by that ministry, a group o f Government consultants (PLS RAMBOLL), and World Bank officials. The CPAR analyzes the existing public procurement system in Zambia and recommends suitable actions to improve the economy, efficiency, predictability, and transparency o f the procurement processes. It provides Government with tools to make changes and to improve the public procurement system and institutional framework. This chapter i s a summary o f the key aspects of the CPAR report. 41Thissection is a summary based on a larger report, World Bank (2002), "Zambia Country Procurement AssessmentReview(CPAR)". The CPAR is issuedas Annex 3 to this report. -83- 5.89 Importance of Public Procurement in Zambia. Public procurement plays an important role inthe Zambian economy. The total volume o fpublic procurementin2001 was estimated at USD485 million, equivalent to 15 percent o f GDP - this i s a higher proportion than the 10 percent estimated for most countries in Afiica. The Central Tender Committee (CTC) o f the Zambia National Tender Board (ZNTB) processed and approved procurement valued at 12percent o f GDP while another 3 percent o f GDP was procured at a decentralized level by ministries and provincial administrations. If even a conservatively estimated 10 percent o f this cost could be saved through better procurement policies, practices, and institutional arrangements, this would mean a yearly gain to the exchequer on the order o f USD5O million. Analysis of the Problem betweenGaps 5.90 in the Public Procurement System's Intentions and Practices. A gap intentions and practice pervades various aspects o f Zambia's public procurement. Despite a legal and institutional framework, weaknesses inits structure and content allow undesirable practices and procedures. While the ZNTB is expected to enforce procurement rules, inpractice, it is liberal inpermittingexceptions and allowing negotiations to replace clear procurement guidelines. A system to register suppliers for small-value procurement, for example, has flaws that could engender corrupt practices. While short-lists are drafted on the basis o f the registration list, each short-list i s subject to misuse to favor particular firms. Further, the registration lists are not systematically updated nor are the means available to check the data applicants submit in registration forms. Moreover, the system i s liberally used for larger value items than intended. The ZNTB also has a significant supervisory role, but its preoccupation with processing procurement prevents this. The existing system for handling complaints features a too- narrow scope and design, lacks independence from the ZNTB which i s involved in the procurement process, and it suffers from a perfunctory operation, causing supplier and contractor dissatisfaction to grow while malpractices continue. 5.91 The slow pace o f "delegation o f authority to procure" to the ministries and agencies is partly due to ZNTB's lack o f confidence intheir capacity. But there are also other reasons for persistent lack o f adequate interest in reform o f public procurement. The lack o f interest in ministries and agencies to actively pursue even the limited program o f delegation o f procurement authority is symptomatic o f underlying causes such as preference for a system o f diffused responsibility and accountability offered by the current lax systems andprocedures andinstitutional arrangements. Any tighteningo f the arrangements would affect the distribution o f rents, be it ZNTB, senior officials or politicians. Though difficult to pin down in any country, political interference in public procurement inZambia i s generally acknowledged. 5.92 This situation, combining poor legislation and procurement practices with political interference and untrained procurement staff, fosters corrupt practices. Though the government hadnot initiated (until September 2002) a survey on conuption financed bythe World Bank some time ago, the CPAR noted a common perception inZambiathat substantial corruption i s associated with public procurement. Yet, the Anti-Corruption -84- Commission (ACC) and the OAG, which have roles to play incurtailing corruption, have beenhamperedby funding and little effective attention to their reports. 5.93 Legal and Institutional Weaknesses.The legal framework4' lacks robustness and features structural, and content, inadequacies. Basic principles of procurement are not captured comprehensively. Further, aspects that should be in the Act are in the regulations or guidelines and vice versa. The ZNTB's role i s not limitedto oversight and policy functions, as it also must manage public procurement through CTC. Among the neglected oversight functions include: (i) monitoring and gathering statistics on public procurement, (ii)developing procurement capacities and capabilities, and (iii) implementing public procurement reform. 5.94 Weaknesses in Procurement Procedures and Practices. Two current practices also foster corruption and higher prices: using negotiations as an accepted procurement method and misusing the registration system for purchases from short-listed firms. Further, the ZNTB's capacity to resist ministry pressure promotes acceptance of less cost-efficient and corruption-prone procurement methods. Additionally, relatively low thresholds (USD300,OOO maximum for ministries in the highest category for decentralization through "delegation o f authority to procure") lead to more costly procured goods and civil works. These low thresholds cause ministries to split contracts into smaller sizes to avoid ZNTB involvement. The registration system i s then used to buyitems beyond the monetarythreshold for the registrationnormally permits. 5.95 Weaknesses in Procurement Management. In the few departments where they do exist, procurement files are often incomplete. Also, there are no instructions on how to structure an official procurement filing system and to maintain individual files. Without a proper filing system, supporting papers (such as performance guarantees) are inadequate for good contract management. Poor contract management produces avoidable losses to government. Also, procurement planning as a tool for conducting efficient and economic procurement i s largely non-existent. Vital aspects o f procurement planning(packaging, selecting appropriate procurement method and type o f contract) are not plannedearly inthe procurementprocess. Inevitably, substantial delays inprocessing and inefficient "urgent" procurement actions result. 5.96 Inadequacies in Budget Allocation and Funds Release System. As discussed in the Chapter on budget execution, due to the cash budget system, the funds allocated in the budget for each project and each ministry often are not fully released during the year43,and there are significant delays in the release o f small amounts. This seriously affects efficient project implementation, procurement, and contract management. 42Thelegal framework i s mainly based on the 1982 Zambia National Tender Board Act (ZTBA), which has limited scope and unclear enforcement mechanisms. It provides for the establishment o f the Zambia National Tender Board (ZNTB). 43 The practice of allocating a "nominal" 1Kwacha for "tentative" projects further strains the budget for other projects when more than the allocated 1Kwacha is finally released to such projects. -85- Consequences o f the budget system's weaknesses include: (i) contractors' bills being held up for several months, delaying projects; (ii) delayed payments incurring interest charges that are paid from a ministry's limited budget, further reducingfunds for product payments, and (iii)when contracts designated in foreign currencies have protracted payment delays, exchange rate changes cause larger amounts o f local currency to be needed than originally planned. All this contributes to higher procurement costs and avoidable losses to government. 5.97 Procurement Cadre. Even though a well-trained procurement cadre is crucial to efficient procurement, there has been no progress in efforts to set up such a cadre. Over 600 people have been trained in procurement at different levels, but most have left their jobs, leaving untrained people inmany positions. Trained staff leave because o f the lack o f a career path, an unattractive salary structure, etc. ZNTB salaries are even lower than the Zambia Revenue Authority's. The ZNTB's staff and those in the ministries do not have defined career paths or a department responsible for managing the cadre and its capacity development. 5.98 Capacity Building. Procurement training institution^^^ exist, but need to upgrade their offerings to feature mid-career and specialized procurement training programs to supplement the supplymanagement programs they currently conduct 5.99 Anti-Corruption Commission (ACC) and OAG Handicapped by Insufficient Funding and Salaries. The activities o f the Anti-Corruption Commission and OAG are critical if Zambia i s to achieve a well-functioning, fair, and transparent procurement framework. Though the President acted recently to strengthen the ACC to fight corruption by creating a special cell whose role would also include procurement issues, unattractive salaries and conditions o f service hamper its ability to attract and retain skilled personnel. O A G i s also under-funded and its reports remain inadequately acted upon. Actions Takento Date 5.100 Assistance to Procurement Reform. Reformo fthe public procurement system has been on the Zambian Government's agenda for almost a decade. This goal has resulted inanumberofdonor-funded projects, whichhave supported, andcontinueto support, the Government's public procurement reform agenda. In 1993, the World Bank and the Zambian government agreed on a Financial and Legal Management UpgradingProject (FILMUP), which among other things, focused on improving the Zambian National Tender Board's and various ministries' procurement performance. In the mid-l99Os, SIDA provided support for training and capacity building for improved public procurement. In ESAC I1 (the second Economic and Social Adjustment Credit), strengthening and streamlining public procurement was an important element. The 44For example, the Zambian Institute of Management (ZAMIM) and Zambia Insurance Business College Trust (ZBIC). -86- Public Service Capacity Building Project (PSCAP), which i s currently being implemented, seeks to strengthen the Office o f the Auditor General and the ZNTB in their roles o f ensuring sound management o fpublic funds. 5.101 Government Action. In 1996, the Zambian government adopted the Public Procurement Action Plan (1996), which aimed to develop the human resources o f the public procurement system along with decentralizing procurement to various ministries and agencies and strengthening audit and oversight functions. In 1997, a World Bank Country Procurement Assessment Review concluded that the action planimplementation had moved more slowly than expected, particularly with regard to the planned decentralization (through "delegation o f authority to procure") andthe establishment o f a cadre o f procurement professionals to support improved procurement activities. The current report concludes that, inmanyrespects, the objectives o f the 1996 action plan still have not been realized. A number o f steps have been taken, but in significant areas, requiredreforms have not beenimplemented effectively, or reform results have not been lasting or have been otherwise in~ufficient.~~ 5.102 The current CPAR concludes that the 1996 Action Plan's overall objectives have still not been realized, inter alia, due to insufficient efforts by ZNTB. Though a number o f steps were initiated, planned reforms in significant areas have not been efficiently implemented: e Improvement of procurement capacity in the Procurement and Supply Units (PSUs) has been slow. In the 1996 action plan, setting up the procurement cadre emerged as a crucial step for decentralizing procurement to PSUs. The cadre needs a defined career path and an attractive salary structure linked with a well-developed training system. These improvements were not initiated and the cadre was not created. The capacity inthe PSUs remains a problem except where ZNTB has seconded its staffto work insome ministries' PSUs. e Implementation of the Action Plan (1996) did not get adequate attention. Examplesare: (i) or no progress inde~entralization~~over 6 years, as little the ZNTB did not give a single ministry full authority to conduct its own procurement, (ii) partly because a procurement cadre was not created, procurement capacity building47was unsustainable, as recently trained 45However,the election o f a new political leaderslup in Zambia in December 2001, who appear determined to fight corruption, may open up new possibilities for the efficient implementation o f reforms 46Fulldecentralization o f procurement for all PSUs by 2002 did not materialize. Of 51PSUs, 24 are not yet certified; the monetary thresholds for the remaining 27 are very low. The ZNTB i s still involved in processing 80 percent o f procurement. 47Professionalprocurement staff i s required to support decentralized procurement. Where the ZNTB has seconded its staff to ministries to operate decentralized procurement units, PSUs procurement capacity i s better. -87- staff left for other jobs,.Preoccupied with managing procurement, the ZNTB didnot focus onthe 1996Action Plan. 5.103 Drafting of a Strategic Plan. Presently, in coordination with AfDB, the ZNTB has drafted a Strategic Plan (2002-2006) that envisages minor adjustments to legislation and decentralization limited to "certifying" all 51 PSUs over time, but not authorizing them to do all their own procurement. The strategy, coverage, and content o f this draft Strategic Plan could be enhanced as a result o f the current CPAR. FutureAction Plans 5.104 Central to any recommendations that emerge i s the need for a "champion" for the Reformprogram as it shouldnot be left to be an intemal program within ZNTB. Besides, such a program that needs to tackle the various vested interests inthe status quo can only be tackled bypolitical will. 5.105 Given adequate attention to the above, the actions described below are priorities for improving the Zambian public procurement system. (See also the fully developed action planinthe ACTION PLANMATRIX inannex at the end o fthis report). (i) Establish a Procurement Reform Task Force (High Level Committee). Set up a High Level Committee that will be responsible for initiating and overseeing implementation o f activities in the action plan developed by this CPAR. This Committee must have high-level members from, for example, ZNTB, MOFNP, ACC, and OAG, and members should not be allowed to appoint substitutes. The Committee should have a clear mandate and time frame for concluding its reform agenda. A Procurement Reform Implementation Unit (PRIU) should be created to manage day-to-day implementationo f reforms andto act as the secretariat to the Task Force. This Unit should be separate from the ZNTB and should report to the Permanent Secretary in the Ministryo f Finance. (ii) Re-establish the Policy and Supervisory Role of ZNTB Immediately. Without waiting for new procurement legislation (see below), refocus ZNTB to its policy and regulatory role already required under existing legislation. As the ZNTB Director General will manage the CTC until its dissolution in 3 years (see below), strengthen the secretariat o f ZNTB to support it in this role by enhanced staffing (and possibly splitting it o f f from the current secretariat). Provide an adequate budgetto support this objective. -88- (iii) Revise the Legal Framework 48 Ideally, a new procurement act whose content should be guided by the UNCITRAL Model Law andby current international best practices. Among other things, it should provide for a new procurement authority and a complaints system based in the procurement authority but with a separate, independent, part-time Appeals Body. A Cabinet Paper should be developed for the new legislation, which Parliament should enact. New regulations and guidelines should be developed on the basis o f the new act. Pursuant to its enactment, a new procurement authority should be established. A Director General and other staff for the new authority should be appointed to replace the ZNTB immediately on creation o f the Authority. The Authority's specialized b c t i o n s shouldbe described and developed indetail. (iv) Finalize the Plan for gradual delegation of procurement authority. A phased, 3-year plan for delegation o f procurement authority should be developed. The planwould give full authority to ministries and agencies to conduct procurement without reference to or approval by the ZNTB or CTC. In the short term, the plan should appoint one or two provinces as pilot provinces for full decentralization. Inthe long-term, upon full decentralization, CTC would cease to exist. Implementation o f decentralized procurement, however, i s contingent upon there being a strong accountability system in place, which would include use o f sanctions by MFNP for violators. Implementation i s also dependent upon adequate capacity at PSUs. (v) Establish a Professional Procurement Cadre. Define the composition o f the cadre to be established and supporting measures, including funding an adequately attractive salary structure. Start implementing such measures giving the cadre management function to the ZNTB until the new procurement authority (possibly called Zambian National Procurement Authority or ZNPA) comes into effect. The existing system o f secondment o f ZNTB staff to positions in PSUs directly managed by the ministries and agencies should apply for all the staff inthe procurement cadre. A human resource database should be established with both cadre members' curriculum vitae and those applying for membership. A training action plan should be developed. 48TheTerms o f Reference(T0R) for the review of the legal and institutional framework, which is currently being developed with the AfDB, should be revised to consider the recommendations o f the present report. The TOR should state clearly that revisions o f procurement legislation should build on the UNCITRAL Model Law. -89- (vi) Re-design Registration List System. A revised Registration System should be defined. New guidelines should be developed that restrict its use to small-value contracts only. (vii) Anti-Corruption Actions. The planned corruption survey (begun inOctober 2002) should be implementedquickly. Implementation should be coordinated with the Public Service Capacity Building Project (PSCAP). Anti-corruption clauses should be included in the standardbiddingdocuments. (viii) Introduce Procurement Planning and a New Filing System. A manual on procurement planning should be developed, containing clear instructions for introduction o f best practices. Release o f budget funds from the budgetary system should be suitably linked to realistic procurement plans. A separate procurement register should be defined. In addition, instructions and guidelines should be developed for PSUs on establishing and maintaining a procurement filing system. (ix) Strengthen Training Programs and Training Institutions. A capacity assessment o f training institutions should be carried out. The need for supplementary procurement training programs should be defined. Required training programs should be developed and delivered. -90- 6. ACCOUNTING, REPORTING,AND AUDIT 6.1 This chapter has two sections. The first section will examine Zambia's current system as well as capacity to provide timely, accurate, and ongoing reporting o f expenditures. It examines the current obstacles to reporting at central and line ministry levels, highlights reforms that have been carried out to date, and proposes measures which could be taken to address the remaining problems, including the implementation o f the Integrated Financial Management and Information System (IFMIS) as a medium to long-term option. The second section addresses budget auditing issues. It examines the institutional roles and relationships for the conduct o f internal and external audit, highlights the weaknesses that exist in internal and external audit, and lays out the progress made over the past few years andwhat remains to be done inthe future. A. LEGAL BASISFORACCOUNTINGAND REPORTING 6.2 The Zambian Constitution `defines4' the content o f the annual reports. The Subsidiary Legislation to the Finance Act and the Procedures Manual dated 1992 describe in detail accounting procedures and provide ample guidance to government agencies and officers: 0 A cash-based accounting system with simple ledgers for recording o f commitments and expenditures i s prescribed. Commitments are to be recorded when created and expenditures recorded (in a separate column) when payments are made. Monthlyreturns are to be produced from the books. 0 Monthly reconciliation o fbank accounts is required 0 Stores and assets management remain the responsibility o fthe Cos. 49Article 118 of the constitution prescribes that (1) the MOF (minister) shall cause to be prepared and shall lay before the National Asembly not later than nine months after the end of each FY a financial report in respect o f that year. (2) The financial report shall include accounts showing the revenue and other moneys received by the government in that FY, the expenditure o f the government in that FY other than expenditure charged by the constitution or any law on the general revenue, the payments made in the FY otherwise than for the purpose o f expenditure, a statement o f the financial position o f the republic at the end of the FY and such other information as parliament may prescribe. -91- 0 Checkbooks and issued checks are to be under the custody o f the Accounts Departments o f each spending agency, and issued checks are not to leave the department unless proof o f delivery o f goods and services has beenprovided. 0 Each CO is required to provide to the MFNP monthly returns and any other information as the PS-MFNP may instruct the agencies. Controlling officers must produce monthly returns for the Accountant General within 15 days after the end o f each month. The returns contain budget provision, total expenditure, outstanding commitments, total actual releases (including appropriations-in-aid), andarrears. 0 The monthly returns are to be accompanied bybank reconciliation statements. 6.3 The legislation, however, i s totally silent on the role and responsibilities of the Accountant General. The functions currently performed by the Accountant General are part o f the responsibilities assigned to the MFNP generally. Nevertheless, if fully observed, the existing provisions are adequate to ensure that proper accounting and financial data are maintained. B. THECURRENTSYSTEMFORFINANCIALMANAGEMENT AND REPORTING 6.4 Overall the systems used by MFNP for financial management tend to be either manual or rely on outdated technology with limited applications. Budget preparation, budget execution, and payroll management are all separate and independent applications within MFNP that do not interact with each other. In line ministries the situation is generally worse, with reliance upon basic spreadsheet applications where available. There are few personalcomputers available within MFNP (that are inworking condition) and many o f those are also outdated. Fifty personal computers were recently received and have been distributed for use throughout the GRZ. However, training on the new computers and networking o f these computers will be necessary before they are fully functional. 6.5 Budgetpreparationtends to be done through an in-house system called Computer- Aided Budgeting (CAB), which i s used for consolidation o f line ministry estimates and eventual production o f the budget document itself.50 The line ministries generally do not have access to this software, but instead either submit documents in hardcopy or on a diskette. Budgetreleases are managed separately and are not part o f the CAB. 6.6 The accounting function i s heavily reliant uponmanual record-keeping, especially in the line ministries. Line ministries and provinces submit their monthly expenditure reports on diskette to MFNP with the help o f a computer-based Financial Management System (FMS). The IMF has noted inits report that the FMS is "mainly a data-capturing application with few utilities and limitedreporting functionality." The FMS records the 50See IMF(2001) report onpublic expendituremanagementinZambia. -92- details o f transactions after the payments have been made. The MFNP data center consolidates the reports o f the various units using a Government Accounts (GA) application, though the IMF has noted that there i s usually considerable time lag between entering the data and its eventual consolidation. During the Bank's initial PER mission in 1999, the Accountant General's office also cited the extremely limited analytical and reporting capacity o fthe MFNP's COBOL-based computing system. 6.7 Payroll is managed in a slightly different manner, although the actual transactions are also recorded in the FMS by MFNP. Cabinet Offices maintain the establishment register and personnel records based on manual updates received from line ministries and provinces. The Data Center processes the updated payroll lists andtransmits them to the line ministries to execute payments. After check issuance, the payroll payments "are entered inthe FMS system inthe same way as non-wage expenditures." The weaknesses inthese procedures have been recognized, and a Human Resources System is currently being developed with implementation plannedin2003. Analysis of the Problem 6.8 There are two dimensions to the problems o f financial reporting in Zambia: (i) problems related to recording, accounting, and reporting; and (ii)problems with regardto the timeliness and accuracy o f financial reports. Each is discussed in turn. The observations are based on the procedures and processes inplace through the first half o f 2002. At that point, new software was introduced for expenditure and commitment by line ministries. As discussed later, this may make possible more accurate reporting inthe future. Untilthen, some o fthe deficiencies described below may still remain. AccountingandReporting 6.9 In exercising the powers conferred on the COS, officers in the ministries and provincial departments commit and spend and issue vouchers for payment. According to the formal rules, vouchers are issued by the spending unit and sent to the Accounts Department for recording and payment. The vouchers are used to record both the commitment and expenditure in the expenditure control ledgerkash book. The practice is, however, different. 6.10 First, in some restructured ministries, the commitment is recorded only when the check i s issued and at the same time that expenditure is recorded. This represents a serious breakdown o f the formal system - it undermines the distinction between commitment and expenditure, and distorts the financial information. Another departure from the prescribed rules i s a practice whereby checks are issued, and handedover to the procurement staff before the delivery o f goods and services. 6.11 Second, insome ministries, the relevant spending units are keeping the ledger for commitments while the accounting department keeps records o f expenditures and is not involved inthe commitment control process. -93- 6.12 Third, the above described shortcuts and deviations are found inrestructuredunits where experienced staff have been made redundant. Training for accounting staff needs to be prioritized. 6.13 Finally, Accounts Department produces accounts and monthly returns from the vouchers and their ledgers. AccG's office produces quarterly accounts and reports based on those returns. The ST can freeze a spending agency's bank account ifthe return i s not received on time. However, the sanction has not been applied even though several ministries are systematically late in their submissions. The assistance provided by the office o f the AccG to the agencies has to some extent improved the timeliness and the quality o f information. However, the quality of returns still remains uneven across government agencies. TimelinessandAccuracy of FinancialReports 6.14 Inconformity with the legislation (Article 118 o fthe Constitution), MFNPhas to produce the Republic's Financial Report for the year and lay it before the NA within nine months after the end o f each FY. Before the production o f consolidated accounts, the Office o f the Accountant General carries out some checks and reconciliation. For this purpose, in addition to the monthly returns, each CO is required to send a copy o f the cash book and a bankreconciliation statement to the MFNP. 6.15 For the first time in ten years, the audited report for FY 2000 was issued before the legal due date. This is an achievement, given the capacity and system in place. However, several problems remain. 6.16 First, the bank reconciliation statements do not include the beginning and ending balances o f the bank accounts. It is unclear as to how bank reconciliation can be done without those balances.51Possible discrepancies may arise from cancelled checks if the corresponding funds are not returned to the B o Z . ~Likewise, without the balances it ~ cannot be verified whether all o f the released funds are actually used. Moreover, it has been noted that some agencies have used overdrafts, which are expressly forbidden without written authorization from the PS-MFNP. (See below for further discussion on reconciliation o fbudget releases with actual expenditures.) 6.17 Second, other weaknesses were noted with respect to the completeness and accuracy o f financial information, which have implications for financial statements: 0 Ledgersdesigned to accurately account for commitments andpayments/ expenditures are not used properly. 51Infact, despite the time spent inthe relevant service, the mission was unable to receive fillinformation or hlly understandthe bank reconciliationpractice at MFNP. 52Any check outstanding for more than six months is cancelled automatically, but it is unclear ifthe finds are transferred fromthe ministry's mirror account back to the BoZ. -94- 0 Commitment and payment transactions are either posted at the same time - and this was sometimes done even before delivery o f goods and services - or sometimes, commitments are not posted. The latter practice was observed where commitments have been created without regard for standing instructions from the MFNP. 0 Incase ofa reorganized agency, we observed a total breakdown insegregation of duties and internal control mechanisms. The procurement unit performs incompatible tasks. These included control over the stores as well as receiving and keeping checks from accounting, in anticipation o f delivery o f goods and services by suppliers. 6.18 Because o f these irregularities, the books o f accounts are incomplete, and because the books are incomplete, the monthly returns and the subsequent financial reports based on these returns are most likely incorrect. 6.19 Third, it seems that the quarterly consolidated reports are incorrect due to omissions or double counting. While some ministries wait for information from provincial offices and consolidate the data from provinces, others send the HQ data alone. As we understood, the unit at the MFNP waits for and consolidates all data from all 51 Cos. Therefore, there is a highprobability for omission or double counting. This raises the question about the end o f year accounts and bank reconciliation as well. In addition, hidden vouchers are not necessarily included inthe monthly returns and it i s not clear if they are liquidated at the end o f the FY. On the other hand, some loans to individuals and private firms are included. These need further review and analysis. 6.20 Fourth, as pointed out above, the exact roles and responsibilities o f the Accountant General are not defined in the existing legislation. Functionally, the AccG who produces the Republic's accounts on behalf o f the M-MFNP,has a "supervisory" role with respect to the accounting and financial information. However, he i s institutionally attached to the PS-FMA and reports to him. 6.21 Fifth, quarterly and annual financial statements produced by the Office o f the AccG reflect mostly transactions that flow through Control Account 99 (CA # 99). The Account is plagued with problems o f reconciliation and availability o f information in both inflows and outflows. The inflows are deposits from: (i) Revenue Authority Zambia (ZRA); and (ii) individuals or bodies depositing funds to the general revenue. However, the ZRA does N O T have any reporting relationship with the AccG, and therefore the AccG does not receive detailed informatiodexplanations as to how much i s deposited/ paid, by whom and why. The outflows, on the other hand, are executed by the Budget Office, which again has no formal direct relationship with the AccG. The lack o f efficient and effective communication with ZRA and the Budget Office has meant that the AccG does not receive adequate information to reconcile the accounts and the Budget Office i s not doing an adequate job. As a result, the consolidated accounts and aggregate statements are not reconciled. Besides the other factors mentioned above, this raises an issue regarding the completeness o f the annual reports. It i s worth noting that the Bank's and other donors' project resources and expenditures are not included in the Financial -95- Statements. In addition, the financial reports contain no information on contingent liabilities. 6.22 Table 12 below provides evidence of some of the anomalies that appear in the Financial Statements. For five ministries-Agriculture, Education, Health, Local Government, and Works and Supply - it compares the budget release with the actual expenditures reported in the Financial Statements. Where the variance is positive, it indicates that releases exceeded expenditure, and it suggests that funds would have been returned to the Treasury. It is unclear though whether this occurred, or instead, whether the funds were actually used for another -96- Table 12: Comparisonof BudgetReleasesandActual Expenditures Varianceon Budget finistry xpenditure Releases !ccounts pmount 1% 1998 PE 15,736,541,397 (1,829,93 1,136) -13 12,653,798,797 14,212,076,173 (1,558,277,376) -12 2,942,702,48 1 2,154,971,994 787,730,487 27 29,503,111,539 32,103,589,564 (2,600,478,025) -9 17,110,472,155 7,460,852,637 II 9,649,619,5 18 11 56 6,791,364,711 11,566,850,920 11,055,077,129 511,773,791 ~ 35,468,687,786 30,860,233,650 4,608,454,136 19,983,946,29 1 9,290,3 19,684 1 10,693,626,607 I 54 10,570,614,243 20,888,017,815 (10,3 17,403,572) -98 183,900,516 308,896,647 (124,996,131) -68 30,738,461,050 30,487,234,146 251,226,904 1 94,300,000,000 94,650,052,37 1 (350,052,371) 0 28,653,857,265 21,388,019,363 7,265,837,902 25 22,952,176,291 22,675,811,694 276,364,597 1 145,906,033,556 138,713,883,428 7,192,150,128 5 - 122,671,168,090 109,623,566,626 13,047,601,464 11 26,024,782,84 1 22,590,515,686 3,434,267,155 13 27,147,062,087 24,124,972,2 16 3,022,089,871 11 175,843,013,018 156,339,054,528 19,503,958,490 11 155,998,180,698 135,120,969,432 20,877,211,266 13 35,470,698,548 37,596,614,547 (2,125,915,999) -6 36,205,112,530 32,848,455,334 3,356,657,196 9 227,673,991,776 205,566,039,313 22,107,952,463 10 27,247,864,968 51,368,713,870 (24,120,848,902) -88.5 17,956,119,107 13,691,806,247 4,264,3 12,860 23.7 58,162,592,281 5 1,047,575,464 7,115,016,8 17 12.2 103,366,576,356 116,108,095,581 (12,74 1,s 19,225) -12 30,404,795,539 63,3 15,480,150 (32,9 10,684,6 11) -108.2 22,443,288,961 20,826,975,192 1,616,3 13,769 7.2 62,451,758,490 20,743,190,652 41,708,567,838 66.8 115,299,842,990 104,885,645,994 10,414,196,996 9 75,236,557,679 32,43 1,O 12,533 42,805,545,146 56.9 26,162,597,566 32,340,497,939 (6,177,900,373) -23.6 35,832,697,680 41,104,726,084 (5,272,028,404) 1 -14.7 137,231,852,925 105.876.236.556 II 31.355.616.369 423,255,282 836,774,385 (413,s 19,103) 97.7 4,381,735,696 3,129,275,879 1,252,459,8 17 8.6 -97- pdget Zinistry Year Budget Office xpenditure eleases Sub-Head Releases Grants& Other 10,012,668,836 5,443,488,292 4,569,180,544 .-~ 45.6 Total 14,817,659,814 9,409,538,556 5,408,121,258 -~36 1999 PE 1,111,602,165 1,016,184,041 95,418,124 8.6 RDC 2,656,746,358 2,592,487,824 64,25 8,s 34 2.4 Grants & Other 6,546,577,387 5,647,765,376 898,812,011 13.7 petal 1 10,314,925,910 - 9,256,437,241 1,058,488,669 10 000 /PE 1,608,549,697 1,472,800,020 135,749,677 8.4 4,606,762,416 4,457,445,278 149,3 17,138 3.2 I 14,356,867,434 1,266,607,595 8.1 21,838,787,142 20,287,112,732 1,551,674,410 -~ 7 11998 PE 1,626,982,795 2,738,775,354 (1,111,792,559) -68.3 3,7 10,975,699 (662,679,106) -21.7 216,521,803 4.0 -~ -36 34.9 Vorks and 2,666,794,749 362,865,418 12.0 upplies 645,258,227 107,995,424 ~-14.3 1,508,475,423 -~22 2,438,828,784 324,581,263--- 11.7 6,129,039,6 14 (765,153,863) -14.3 ketal 1 919,938,293 173,528,823 15.9 9,220,762,914 9,487,806,691 (267,043,777) -3 expenditure category, e.g., PE resources used for RDCs. Even though the Financial Regulations prohibit transfers between PESand other categories, there were ministries that acknowledged this occurred. The more curious cases are those where the variance i s negative, indicating expenditures exceeded the cash release. The magnitude o f the variances calls into the question the accuracy o f the reporting, and underlines the importance o f reconciliation needed between cash releases, expenditures, and banking data. RecentActions 6.23 In June 2002, MFNP implemented at the headquarters of each line ministry a Microsoft Access-based program for recording data on budget allocation, commitments, arrears, and expenditure payments. Provincial offices are expected to have the system ready by end-2002. The source document for enteringdata into the system i s the backing -98- sheet that is submitted to B o Z . ~Although the new Financial Management System ~ (FMS) provides for entering commitments, currently these are equivalent to the payment amount because they are entered into the computer at the same time. This practice may be explained in part by the fact that Accounting i s often not informed o f LPOs or contracts until the time at which payment is requested. Nevertheless, despite these limitations o f the system, the use o f the backing sheet to key in data almost certainly increases the accuracy o f the monthly expenditure reports to MFNP. It also creates an automatic reconciliation o f expenditure data with the bank accounts.54 FutureAction Plan 6.24 While the FMS and use o f the backing sheets constitute important steps toward improving the reliability o f expenditure reporting, further actions are still warranted. Modify the accounting procedures and chart o f accounts in order to capture commitments and arrears and to highlight interest charges on late payments. Modify the FMS process to include inputs on the LPOs, contracts, and all commitment documents. Reconcile monthly expenditure retums with monthly releases for each unit or department that is rolled up to the ministrylevel. Restrict funding for ministries that do not submit a monthly retum. Ensure that agency funds related to cancelled checks are retumed to the BoZ promptly. Impose penalties for overdrafts without written authorization from the Secretary to the Treasury. Clarify and clearly define the role o f the AccG in the legislation, so that he/she will directly report to the Minister o f Finance or the ST. This shift will also improve the coordination and communications with the Budget Office. Establish mechanisms for institutional review and vetting o f any organizational restructuring. Provide specific directives to guide staff and strengthen staff training on internal controls, especially for restructured institutions so that the control framework remains intact. 53The backing sheet is the list o f checks that the Ministry is issuing and serves to inform BoZ o f the amount that should be transferred to the Ministry's mirror accounts to cover the checks. 54Reconciliation i s still needed to account for checks that may not have cleared or additional funds that were deposited. -99- C. MOVING TOWARDS INTEGRATED INFORMATION FINANCIALMANAGEMENT SYSTEM (IFMIS) Need for I F M I S 6.25 Few will argue with the inadequacy o f the current information systems at MFNP to provide reliable control o f and reporting on public expenditures. As pointed out above, the current systems produce a number o f serious problems that must be addressed. First, expenditure reports are prone to errors and inconsistency that are difficult to find and to reconcile under the current system. Second, reporting tends to be significantly delayed because o f the various levels o f compilation required. Third, the focus on bookkeeping tasks distracts accounting staff from more meaningfbl analysis o f the data and trends. Fourth, the delays and inconsistencies greatly diminish the ability o f MFNP to provide sufficient oversight and control over the nature o f the expenditures. Because o f these inadequacies, the MOFNP has found it difficult to provide an accurate, complete and transparent account o f its financial position to parliament and other stakeholder, including the general public and donors. 6.26 While recognizing that computerization i s not a panacea, an IFMIS will serve as usefbl and efficient tool for sound financial management. In particular, the system can help improve reliability and timeliness o f financial data, strengthen financial controls, ensure better availability o f comprehensive financial information, and ensure enhanced transparency and accountability o f the Executive to the public. However, it should be emphasized that IFMIS is just a tool, and not a substitute for implementing the basic measures proposed in the preceding section to improve budget accounting and financial reporting. Indeed, IFMIS should be pursued as complementary to, and in parallel with, these basic measures. 6.27 There is wide recognition by the government and the donor community on the need for replacing the current manual accounting procedures and outdated technology at MFNP - which make it extremely difficult to obtain reliable and timely financial information and do not provide adequate safeguards for budgetary control - by a computerized information system. The challenge has beentryingto design a system that can meet international accounting standards, while at the same time being implementable in a medium term horizon. The key issues and challenges in IFMIS development in Zambia are provided inthe annex to this chapter. Actions Taken to Date 6.28 Several efforts have been made by donors to assist MFNP inplanning for a new IFMIS,although the primary assistance over the long term is expected to be through the Bank's PSCAP. During 2001, a Bank consultant helped the govemment with an assessment o f the needs and potential options for a new system, including a rough timeline and potential costs for each. During that year, the EU also f h d e d a consultant from KPMG to work with MFNP for several months to assess the needs for an IFMIS and to develop options to consider. This report, entitled "Project Management for the Integrated Financial Management Information Systems" was finalized in September -100- 2001. It included recommendations on the functionality o f an IFMIS, review o f the Tanzanian experience, and roadmaps for three possible options of IFMIS. The consultant suggested that total cost o f implementing IFMIS would range between USD8.3 and USD20.8 million depending on which option was adopted. The implementation period for each ranged from about six months to two and a halfyears. 6.29 An IMF mission inearly 2001 also provided recommendations on managing the IFMIS implementation process and on the core functionality o f the IFMIS. Their June 2001 report affirmed the MFNP's structure o f a Steering Committee, a Technical Committee (or working group), and a Project Manager. The report recognized the need that MFNP would have for external consulting advice, but stressed the importance o f in- house resources being fully involved in the project design, planning implementation o f hardware and software, and other areas. Perhaps even more importantly the report outlined helpful guidance on the scope and core functionality o f the IFMIS. Several highlightso fthe reportrecommendations have also beensupported byothers: e The reference to an "integrated" financial management information system (IFMIS) may be misleading by implying that all functions must be captured by a single system. Infact, "some o f the specialized information systems currently in place may still be required after the implementation o f the IFMIS." These specialized systems could then become part o f IFMIS through the use o f "integrating" software. e Even if all data is not captured within IFMIS, one nevertheless must anticipate the various interdependencies to ensure appropriate interfaces. e A modular approach is best for the development and implementation o f different components, with an initial focus on core functions. e Core functions would include budget recording, commitments and funds control, payment process, accounting for disbursements and receipts, maintenance o f the general ledger (GL), and generation o f fiscal reports. e The system would keep track o f financial transactions at various stages. 6.30 In 2001, the MFNP established two critical working groups to shepherd development o f the IFMIS. The first was a high level Steering Committee consisting o f very senior officers to make policy decisions relative to IFMIS when necessary. The committee has included representatives o f the Accountant General's Office, the Controller o f Internal Audit, the Budget Office, the Data ProcessingUnit, Investment and Debt Management, Economic and Technical Cooperation, Central Statistics, ZRA, BoZ, the Auditor-General's Office, and a selection o f ministryrepresentatives (especially from the largest such as Education, Health, Works and Supply, and Agriculture). It has been chaired by the Secretary to the Treasury to ensure that speedy progress is made in specifying the requirements o f the system. -101- 6.31 The second important group was the Project Team or Technical Committee, whose full-time responsibility i s to work with IFMIS consultants and to develop detailed specifications for the new system, among other tasks. Unfortunately, significant delays were experienced before the project team was appointed and before funding was found for them to begin work fwll-time. Ultimately, the team should be composed o f representatives from each o f the major functional areas-information technology, debt, budget, audit, accounting, and training - along with a Project Manager. The Project Team will take any policy decisions that need to be made to the Steering Committee for resolution. 6.32 Through PSCAP, the Bank had facilitated a study tour to Dar-es-Salem to examine first-handthe Tanzanian experience with IFMIS? Although it is not clear how applicable the Tanzanian approach will be for Zambia, it provided an important reference point in terms of functionality, cost, and timeframes for the team to factor in. During 2002, the three-person Project Team and a short-term consultant made progress on several fronts. Four pilot ministries and a province were selected, and stakeholder workshops planned.56 The Project Team also worked on the detailed system requirements in anticipation o f the arrival o f long-term consultants. Enhancements have been made to the existing FMS, with the improved version rolled out inJune 2002. This was expected to provide more timely information, enable the reconciliation o f expenditures against the budget, and establish a commitment/arrears database. Plans were also developed for in-service training needs and research was conducted on the future telecommunications infrastructure. FutureAction Plan 6.33 Important modules that should comprise the IFMIS include: a The Core Accounting Module (which would include linkages to the Activity BasedBudgetingModule currently under development). a A Cash Management Module (including a Statement o f Cash Flows) to address the problems in budget execution that are caused by the cash budgetand ad-hoc cash management decisions. a A Procurement Module to establish control over commitments. In the meantime, the manual procedures developed using the services o f a Commitment Control Office and the revised FMS for on-line commitment control will have to suffice. 55Thecomposition of the Project Team or Technical Committee has changed over time, so that those visiting Tanzania are not necessarilythe same as those now working full-time. 56As o f October 2002, the five pilot sites were expected to be the Ministries of Finance, Education, Works and Supply, Health, andNorthernProvince. -102- 0 A HumanResource Management Module is being developed separately under the DFID-fundedPayroll Management and Establishment Control (PMEC) project to keep track o f all employees, their grade and pay rate, and other pertinent personnel information. The system should ideally permit preparation o f payroll with appropriate deductions, as well as linkage to the General Ledger accounting system. 6.34 The current plan is for the Government to pilot IFMIS by mid-2004. However, the Bank recognizes that full implementationo f IFMIS is a medium- to long-term goal in Zambia. In its assistance to MFNP, the Bank has emphasized that whatever option is proposed to donors to be funded, needs to be manageable under the limited absorptive capacity o f the Zambian public sector. Moreover, the modules and functionality, while designed to meet international accounting standards for whole-of-government reporting inthe long term, also need to be able to produce more limitedbut tangible results inthe short term. One such tangible measure may be for MFNP to be able to provide comprehensive, accurate, and timely reporting on expenditures and commitments for some subset o f the overall budget (e.g., a few large ministries for spending in Lusaka province only). As MFNPbegins to build a series o f smaller successes with IFMIS,more andmore functionality canbe added, andone can expect greater donor support. 6.35 The Government has agreed that the best approach going forward i s to adopt a turnkey system and to draw heavily upon the procurement experience o f other countries in the region such as Uganda. Inthe short term, key actions related to IFMIS would include: (a) Pre-qualify potential bidders (b) Issuebiddingdocuments (c) Select andprocure the software and hardware (d) Continue stakeholder awareness workshops. (e) Planfor in-house training needs for users andprogrammers. (f) Pilot the IFMIS inselected ministries andprovinces. Related Issues for Success of IFMIS 6.36 Successful implementation o f IFMIS demands some parallel reforms in public management over the medium term: 0 Humanresource management and compensation are inadequate currently to recruit and retain more highly trained staff in financial management or in information technology. Ifbase salaries cannot be improved for the whole o f civil service, some other creative strategy may be required to assure that MFNP and other line ministries obtain the skills that they need -103- to perform these key support functions. (See the chapter on human resources inthis report for further discussion o fthis point.) 0 The legal and regulatory framework may also need changes inorder to be consistent with new and more modern accounting procedures. IFMIS procedures must be designed to comply with existing regulations, or those regulations need to be updated. All parties also agree that it is inefficient and wasteful to automate outdated procedures that were designed around a primarilymanual system. Yet, changes to the regulatory framework need to be managed well in order to prevent it from unnecessarily delaying IFMIS's development. 0 Government-wide I T policy development and support has been generally non-existent. IFMIS's development will introduce a new, larger infrastructure upon the government to maintain. And like more physical infrastructure, lack o f investment will result in rapid deterioration in usability. Consequently, the government may need to assign special responsibilities to assure that public sector IT policies are dealt with strategically and that funds are preserved for maintenance and upgrading. The fact that the IFMIS project team is already initiating a review and upgrading o f existing systems as a stop-gap strategy is a step in the right direction. 6.37 Sanctions must be enforced even as the new system introduces more discipline and transparency in public spending decisions. If audit findings or violations o f policy are ignored, the gains from the new systemwill be lost. D. INTERNAL AUDIT 6.38 The internal audit function is critical for helpingto ensure compliance with sound financial management. InZambia, the role o f the Internal Audit Department is set out in the Financial Regulations. The main internal audit functions are maintaining the pre- audit controls on expenditure, as well as assessing the overall adequacy o f the financial management systems and procedures in place. Inpractice, various factors impact the effectiveness o f internal audit, including the level o f independence, the scope and content o f its work plans, the level o f resources available to it, and the degree o f follow-up on its findings. InZambia, the Internal Auditors havebeen effective indiagnosingproblems in the financial management procedures, but their potential impact is still severely limited by the lack o f adequate resources to carry out their work and insufficient incentive for Controlling Officers to implement on their recommendations. Analysisof the Problem 6.39 The institutional relationships and structure o f Internal Audit appear to give it ample independence from the line ministries and other entities. However, there may be some vulnerabilities that stem from the funding mechanism for its work. The Internal Audit Department is anchored within MFNP, and headed by the Controller of Internal -104- Audits who inturnreports to the Secretary to the Treasury. Internal auditors are assigned to each o f the line ministries, but their reporting lines are still to the central ministry. Promotions, transfers, and personnel matters are addressed through the central ministry. Although Internal Audit has a corresponding budget line within the line ministry budget, the actual release mechanism for the funds is under the authority o f the Controlling Officer o f the line ministry. Therefore, the independence o f the internal auditors is not necessarily safeguarded by the current budgetary release practices. In practice, it has been reported that Internal Auditors have had full support from Controlling Officers, but there remains a risk that they could be penalized through the funding mechanism as a result o f their findings. 6.40 Annual audit work plans could be improved by developing cost estimates and resource allocations across specific activities. Currently, the annual Internal Audit Work Plan is prepared by the Internal Audit Department within the line ministries in consultation with the Controlling Officer. These work plans are then reviewed and consolidatedinto an Internal Audit Work Planprepared bythe Internal Audit Department o f the MFNP in December prior to the start o f the Budget Year in January. Generally, activities fall in two categories: (i)Ongoing routine pre- and post-audit activities o f expenditures files, procurement procedures, accounting ledgers, etc., and (ii) Specific audits o f targeted projects or functions. The work plan identifies the targeted areas o f activities, and develops cost estimates. However, only a portion o f the work plan can actuallybe implemented effectively duringthe year. 6.41 Financial resources and staffing constraints are important barriers to Internal Audit's effectiveness. Even ongoing routine audit activities suffer from the shortage in personnel. For example, the Internal Audit Department assigns three auditors to each o f the nineprovinces to perform internal audit for the line ministries. However due to work overload, these auditors are restricted to providing services to those departments that fall under the Office o f the President. The line ministries are responsible for providing the staff for their own internal audit requirements at the provincial and district level. This may be easier intheory than inpractice. For example, at the Ministry o f Agriculture an internal auditor hadbeen assigned for only one out o f nine provinces. At the time o f this assessment, four more had been hired for assignment to four more provinces but were still located at the headquarters. Staffing constraints also mean that payroll audits are performed only on an ad hoc basis. (See Chapter 7 on the broader difficulties in recruitment and retention o f qualified staff within the civil service.) Even with staff, there i s a need for computers and logistical support for them to be efficient in carrying out their duties, especially in the districts. Although some institutional strengthening o f MFNP has occurred already through PSCAP, it was not clear that the Internal Audit Departmenthadas yet benefited significantly from this. 6.42 Finally, one o f the most important factors in the effectiveness o f Internal Audit (IA) is the magnitude o f follow-up to its findings. When IA uncovers problems and recommends actions to be taken, it i s critical that there be follow-through. Without the legitimate threat o f sanctions, audits may be counter-productive by emboldening the behavior o f wrongdoers. The responsibility to take corrective action lies with the Controlling Officer, who in turn should be held accountable by the Secretary to the -105- Treasury. Unfortunately, effective follow-up often fails to occur. And though the Auditor General has authority to carry out audits to verify compliance with IA recommendations, severe budget constraints on the AG may prevent them from doing so. FutureAction Plan 0 Establish a mechanism to protect funding for Internal Audit activities within line ministries. 0 Increase staffing levels and RDC fundingo f Internal Audit 0 Monitor and report on whether recommended sanctions are pursued in cases o f misappropriation or fraud. 0 Prepare a quarterly report that tracks implementationo f the work plan. 0 Improve annual work plans by incorporatingthe following: 0 Clear segregation o f routine on-going activities from one-off ex- post audits. 0 Continue to identify the specific staffing and physical resource allocations necessary to carry out the activity, especially for those indistricts, to ensurethat all activities canbecompleted. Combine cost estimates with prioritization o f activities so that if budget resources are cut back, the work plan may be adjusted easily. Increase coordination with the Office o f the Auditor General in development o f the work plan. E. EFFECTIVENESSEXTERNAL OF AUDIT 6.43 Article 121 o f the Constitution provides for the appointment o f an Auditor General (AG) by the President, subject to ratification by Parliament, and declares it a public office. It assigns to the office holder the responsibility of: (i) ensuring that the provisions o f the constitution on public finance (i.e., Part X) are complied with, and (ii) auditing accounts relating to public revenues and expenditures. The Constitution grants the AG unimpeded access to all relevant public books, records, reports and documents. The AG is requiredunder the Constitution to submit herhis audit report to the President within 12months o fthe end o fthe financial year -the Presidenthas the responsibility for laying the report before Parliament within seven days o f receiving it. The Constitution stipulates that the AG shall not be subject to the direction or control o f any other person or authority inperforming their functions. It provides for the AG to hold office untilthe age o f 60 but s h e may resign before attaining that age. The AG can be removed from office for the following reasons: (a) inability to perform his hnctions due to infirmityof -106- body or mind; (b) incompetence; or (c) misbehavior. The Constitution requires Parliament to set up a Tribunal to look into the matter constituting a ground for removal, andto consider the report o fthe Tribunal before removing the AG. 6.44 Similar provisions were made in the Finance (Control and Management) Act o f 1969 and the Public Audit Act of 1980. The 1969 Act, amongst other things, empowered the AG to audit the accounts and reports o f statutory corporations for which the law has not provided for the appointment o f an auditor. It is noteworthy that the 1980 Act essentially gives the AG the authority to follow public money to anywhere it may go. It widens the power o f the AG beyond the audit o f books, records, and reports o f government departments and the statutory corporations described in the 1969 Act to include the audit, as shehe deems fit, o f every statutory corporation, and every private institution that receives Government grant, subsidy, or subvention in any financial year. The AG is empowered to request from independent auditors o f state-owned enterprises (SOEs) any document, reports, or information relating to the accounts o f SOEs. Also, every contract involving the Government or its agencies and enterprises was to include a clause allowing the AG to have access and examine all books relating to the contract. It also provides for the hiringo f agents or specialist consultants by the Minister o f Finance, at the request o fthe AG, to assist the AG inthe performance o f herhis duties. Analysisof the Problem 6.45 There are adequate statutory provisions concerning the appointmen.t, tenure, and auditees o f the AG. Given the democratic setting, the fbnctioning o f the OAG can be significantly enhanced ifthe identified weaknesses are addressed. 6.46 While the AG, as a public official, enjoys adequate independence based on statutory provisions, in practice the OAG's independence i s hampered by availability o f adequate human and financial resources to discharge the AG's statutory responsibilities. Financial resources available to the OAG are subject to restrictions imposed by MFNP while staffing i s subject to civil service regulations. 6.47 While the OAG has strived to adopt audit methodology recommended by the Intemational Organization o f Supreme Audit Institutions (INTOSAI) in practice, due to limitations incapacity the degree o f compliance i s low. 6.48 Although the AG's remuneration is a standing charge on the Government revenues as the remunerations o f judges, the remuneration i s inadequate, as it has not been reviewed for a long time. Inadequate compensation, especially when it involves lower pay compared to similar constitutional office holders can be demoralizing. 6.49 The AG does not have absolute control over staffing. OAG staff are recruited by the Public Service Commission (PSC). The service conditions o f O A G staff are determined by the PSC in a manner similar to other public servants. Consequently, the AG is unableto retainqualified staffdue to unfavorable conditions o fservice (Table 13). -107- Table 13: Staffing of the OAG Position StafJing level Authorized Actual Auditor General 1 1 StateAudit Secretary 1 1 Director ofAudits 5 4 AssistantDirector 9 9 PrincipalAuditor 18 16 Senior Auditor 26 21 Auditor 30 27 AssistantAuditor 30 18 Audit Examiner 35 8 AssistantAuditExaminer 40 2 Source: Office ofthe Auditor General. 6.50 Only one o f the staff is a professionally qualified accountant, but there are several staff (about 15) with a first degree or above in finance and other disciplines. Because of the difficulty in retaining qualified staff, he has not been able to use effectively the few training opportunities that become available. 6.51 The AG does not report directly to Parliament. The AG submits herhis report to President who then lays the report before Parliament. That the AG does not have reports through the overall head of govemment departments that she audits, diminishes herhis independence. More often than not, serious actions are not taken on the audit findings. Not only does this lower the morale o f O A G staff (and thus the quality o f audit), but more importantly, it provides an incentive for further violations o f the financial regulations. The primary responsibility for follow-up on the AG's report rests with the Minister o f Finance, which inturn should be held to account by Parliament. 6.52 The OAG is not adequately funded to enable it carry out properly all the AG's statutory functions. MFNP determines the final budget estimate for the OAG that is incorporated in the Government budget and presented to Parliament for approval. Moreover, the cash releases, which are controlled by MFNP, are generally lower than the approved budget (see Table 14). The result i s that the O A G i s under-funded andhas not been able to carry out all its planned activities. Table 14: Fundingof the OAG for the period 1998 2001 1998 1999 2000 2001 2002 I n million Kwachas AmountsProposed 6307 6100 AmountsAppropriated 2468 2123 3573 3954 4944 AmountsReleased 1868 1183 3798 3441 I n Percentages (%) AmountsReleasedas a % ofAmountsAppropriated 76 56 96 70 AmountsReleasedas a % ofAmountsProposed 30 19 Source: Office of the Auditor General; MFNP, BudgetOffice. -108- 6.53 The AG currently performs regularly financial compliance audits. This i s in line with existing statutory provisions, which requires the AG to audit all the accounts charged to the general revenues, and to satisfy herhimself that expenditures are in line with Parliament authorization. RecentActions 6.54 Inresponse to calls to strengthen the Office o fthe AG (OAG), the AG contracted three consultancies between 1996 and 2001. In 1996, a former AG o f Canada was hired from Cowater Accountability Group, Canada to conduct a study to redefine the legislative mandate and strengthen the OAG. In2000, a former Assistant AG o f Canada was contracted to formulate proposals for the strengthening o f the function o f the AG and to enhance the effectiveness o f the Public Accounts Committee. The World Bank funded both studies. In2001, the AG contracted the services o f Deloitte and Touche to assist in developing the necessary management and financial systems and procedures that would enable the AG to operate as an autonomous, transparent, efficient, and effective Supreme Audit Institution. 6.55 There have also been other attempts to strengthen the OAG. As part o fthe first o f the Public Service Reform Program, the Management Development Division o f Cabinet Office performed a management assessment o fthe OAG in 1996. Subsequently, the O A G held a Strategic Planning Workshop in May 1996 where its Mission Statement, Goal Statement, and Objectives were developed. The following year, Cabinet Office senior officials, some Parliament members and the O A G met to consider the recommendations contained in the 1996 consultant and management audit reports. On the basis o f the recommendations, a Cabinet Memorandum was prepared and presented to Cabinet. Although a similar Memorandum was again presented to Cabinet in 2000, no action has been taken on the implementation o fthe recommendations. FutureAction Plan 6.56 To enhance independence, effectiveness and efficiency o f the AG and his office, recommendations o f the previous studies shouldbe implemented andinparticular: 0 The OAG should be governed by an Audit Board (an oversight body). 0 OAG's budget should be submitted directly to the parliament (after approval by the Audit Board). 0 Staff should not be part o f the civil service and its hiring and condition o f service should follow different rules. e The AG's report should be presented to Parliament as provided for in the Constitution. -109- a Inorder to allow herhimto adequately audit the republic's finances, the AG should also carryout performance audits and shouldaudit the Central Bank as well. a Finally, no one should escape controls and scrutiny; the OAG should be subjected to an external audit -110- PART 111. COMPLEMENTARY POLICY MEASURESAND ISSUES 7. THE HUMANRESOURCE CHALLENGES FOR EFFECTIVE BUDGET MANAGEMENT5' 7.1 It is impossible to discuss the challenges to sound financial management in Zambia without understanding the human resource context inwhich it takes place. Even with the best information systems andregulations, sound financial management would be difficult to achieve without properlymotivated and skilled staff. Problems ininstitutional capacity, however, are not limited to the h c t i o n s o f financial management and information technology management; rather they cut across the whole o f the public administration. Many factors contribute to this including the low and declining real pay o f civil servants, the lack o f training and equipment, and the absence o f an effective incentive system. Duringthe 1990s Bank-sponsored projects such as the Public Service Reform Program (PSRP) attempted to address some o f the problems that existed, but major tasks remain unfulfilled. Furthermore, there has been a concem that unless these issues were intensively addressed over the short to medium term, they could jeopardize successful implementation of other reforms such as the IFMIS. It is inthis light that the Bank-supported Public Sector Capacity Building project (PSCAP) began targeting the government's training needs and remuneration policies as key components o f its overall public sector reform agenda. Within the framework o f PSCAP, the Government has set out to end the ad hoc system o f pay increases and over-reliance upon allowances by implementing a Medium Term Pay Reform Strategy (MTPRS). The new system would aim to decompress and simplify the salary structure so that it i s easier to retain and motivate professional andtechnical staff that will play key roles indriving reform efforts. 7.2 Low remuneration andpoor conditions o f service are major factors contributing to the problem o f poor productivity, motivation, recruitment, and retention o f employees in the public service. At a time when Government i s seeking efficiency improvements, in part by reducing the size o f the civil service, there exist significant staff shortages in a wide range o f professional and technical jobs owing to poor pay compared to that available in the private sector and within regional labor markets. While the major _____ "This chapter is based on the report "A Medium-Term Strategy for Enhancing Pay and Conditions o f Service in the Zambian Civil Service," prepared by Ted Valentine working with a GRZ Technical Committee, which included representation from the Cabinet Office and the Ministry o f Finance and National Planning, with funding from PSCAP o n behalf o f the Public Sector Capacity Building Projects (PSCAP) and the Zambian Management Development Division (MDD) o f Cabinet Office. -111- concern has been with regard to low and declining real public service salaries, the pay and incentives problems inthe Zambia public service go far beyondthe issue o f low pay. Low pay coupled with a poorly constructed internal grading structure and an increments systemthat is based on the lengtho f service rather than performance, stifles initiative and motivation. A firther problem arises from the discrepancies that have crept into wages and allowances paid for comparable jobs in the civil service and o f other sectors in the public service. It has been hoped that by addressing these pay issues it would bepossible to secure significant improvements in individual and collective performance and in the quality o f service. 7.3 This chapter starts off by providing a contextual background to understand how and why the human resource capacity o f the public sector has been declining. It then gives a description o f the principles and intentions behind the MTPRS initially developed by Government. It concludes with a brief update on how pay reform has been implemented thus far, and some o f the new challenges that have arisen. Inparticular, it highlights the concern that negotiations with public service unions and others may have led to an overshooting o f the wage bill resource and envelope, and postponement o f key policy measures. A. THEPROBLEMOFLOWOVERALLPAY 7.4 The decline in real wages for civil servants in Zambia generally mirrors the overall macroeconomic decline o f the country since the mid-1970s. For much o f the past 27 years the Zambian economy has suffered through economic crisis, volatility and decline. The downward trend inthe economy was precipitated initially by the fall inthe price and output o f copper. However, as noted in Chapter 2, the problems in the economy went far beyond the decline inthe mineral sector. Figure 2 presents an index o f real GDP per capita for the period 1975-2000. Here the declining tendency in the economy is apparent. By 1983, real GDP per capita had declined to only about 79 percent of its 1975 level. By 1990, it had declined to 68 percent o f its 1975 level. Between 1991 and 1995, as a result o f severe drought and continued decline in copper production, real GDP per capita declined by more than 20 percent. By 2000, real GDP per capita had declined to only about 55 percent o f its 1975 level. -112- Figure2: Index of Zambia's RealPer CapitaGDP, 1975-2000 110.0 1m.o +Index ofRcalPerCapitaGDP 90.0 w3.0 E 70.0 f BE t 11 MJ.0 =`O 40.0 Sources: Rakner, et al., 1999,Tables 4 and 5; StandardBank, 2002; and Republic of Zambia, 2001a. 7.5 As in other countries in sub-Saharan Africa, the declining real public service pay trends were the result o f several factors. First, real wage repression was attributable to a worsening public sector financial crises (Le., a central government's budgetary deficit crisis and a parastatal financial crisis) which reduced the government's capacity to cover aspects o f recurrent and public investment expenditure^.^' Second, was the "wage equalization policies" pursued by many countries, including Zambia, which were driven byvarious governments' desire for general payrestraint to narrow the urban-rural income gap, and which targetedpay restraint to reduce pay differentials between high-salary and low-wage employee^.^' Third, many governments showed willingness to trade-off real wage levels against higher levels o f public service employment. Fourth, central government wage bills became over-committed as the governments became over- extended. The central government wage bill had to cover not only central government administration, but all o f provincial and local administration personnel, as well as cover the shortfalls in the capacity o f state industries to meet the labor cost o f their personnel. Fifth, and finally, was weak public-expenditure management, which was partlyrelatedto the problems o f wage-bill planning and controls. These factors gave rise to a paradox where the civil servants were underpaid, but the wage bill became an increasing burden on the general citizenry. ?Jnder the circumstances facing many SSA countries, attempting to address fiscal crises by cutting government wages (and the wage bill) could lead and have actually ledto an increase inthe budget deficit. See for discussionvan Rijckeghem and Weder (1997) and Haque and Sahay (1996). 59Somesub-Saharan countries, including Ethiopia, Mozambique, and Tanzania, have followed socialistic wage equalizationpolicies. -113- 7.6 Figure 3 presents indices of real pay trends in the Zambia public service for various salary groups for the period 1975 through 2000. Three significant patterns emerge in public service real pay trends in Zambia. The first period i s 1975-1983, the second i s 1983-1991, and the third is 1991-2000. Figure 3: Indices of Real MinimumSalaries for Various Civil Services Salary Grades, 1975-2000 Period 7.7 Duringthe first period, realpublic service salaries declined sharply (by about 30- 40 percent for most salary groups between 1975 and 1979), before recovering much o f their value in 1980 (returning to 70-85 percent o f their 1975 value). Overall, the decline inreal salaries was not nearlyof the magnitude of the reductions inreal GDP. Between 1982/3 and 1991, all salary groups experienced substantial and uninterrupted decline in their real public service pay. By 1991, the real public service pay for most salary groups had declined to only 11-22 percent o f their 1975 value. 7.8 The thirdperiod covers the years from 1991- 2000. This period was markedby a change in government, the introduction o f a number o f economic reform initiatives and a comprehensive public sector reform program. Still between 1991 and 1995, the country experienced negative economic growth. Between 1995 and 2000, economic growth was low and volatile, averaging 1.9 percent per annum, while fluctuating drastically from year to year.. Given the economic growth situation, there was little scope for sustained improvement o f real salaries inthe public service. Real public service salaries fluctuated significantly, but for the most part moved in a downward direction for most salary groups. By 2000, most salary groups had experienced real public service salary declined o fthe magnitude o f 85-90 percent incomparisonto their 1975 value. 7.9 Overall during 1975-2000, real public service pay declined much faster than did real GDP per capita. Still the wage bill remained large relative to other government non- debt service recurrent expenditures and was perceived to have crowded out expenditures on public goods and services. -114- B. THEFRAGMENTED SALARY STRUCTUREAND OVERUSEOFALLOWANCE 7.10 The Zambian job grade structure is quite fragmented, with a large number o f different occupation-based job scales inuse for different staff categories. There are broad 26 job-grade scales in the current job-grade structure. This excludes grades associated with the Anti-Corruption Commission, the UTH Health Board, and Medical Health Boards, whose personnel are paid through the central government payroll, though they are semi-autonomous agencies. When the segmentation o f these occupationalgroups into "management", "unionized professional and technical", and "unionized classified personnel" are taken into consideration, the number o fjob-grade scales increases to about 40. 7.11 The job structure is populated with 185 job grades - far in excess o f what is considered manageable and desirable for the standpoint o f the capacity to access and maintain horizontal and vertical equity.60 Yet manyjob scales contain few grades (as few as one) and few personnelwithin them. Inaddition to the above scenario, a KPMGstudy revealed that career progression opportunities in the Zambian Public Service are very minimal. 7.12 The public service job-grade and corresponding salary-grade structures have not always been so fragmented. Before the problems inthe economy, the structure was fairly unified. But in the absence o f a coherent pay policy, over time, various unions have successfully negotiated for changes in salary, as well as terms and conditions o f service, inahopeto increase their nominal, ifnotreal, monetary compensation levels. 7.13 The proliferation o f salary scales and salary grades has made it very difficult to achieve and maintain horizontal and vertical equity in the salary structure. Salary differentials, both grade and incrementktep differentials, are far too small to motivate staff to improve performance and to move up the job ladder. In many cases salary increments are in the range o f 0.45 percent - 0.70 percent. Small salary differentials throughout the salary structure have led to a fairly compressed salary structure across all salary scales. This has reduced the value that the public service places on educational qualifications, skills, and experience. The result has been a reduction the competitiveness o f the service to attract andretain suitably qualified manpower. 7.14 The problems with public service compensation in Zambia go far beyond salary differentials and the salary structure. There are problems with the broader compensation structure itself. For some salary groups, the other forms o f compensation payment are far in excess of the salary and wages. Compensation inthe Zambia public service is very fragmented. Personnel on the December 2001 payroll could receive payments from as many as 139possible sources o f compensation. These include: 6oHorizontal equity refers to the comparability in pay for positions o f similar weight or responsibility across government. Vertical equity refers to the appropriateness o f pay from one grade to the next higher grade. In other words, positions with clearly different requirements and responsibilities should be compensated differently. -115- e 16 salary andwages payment categories, e 6 flexible allowances, e 47 fixed allowances, e 14 other emoluments payment categories, and e 56 arrears payment categories. 7.15 Among other things, the dependence on so many allowances to enhance compensation weakens horizontal and vertical equity. They also diminish the relationship between pay, performance, and responsibility, as in many instances lower- paid subordinates may receive higher total compensation than their superiors. At 10 percent of total monetary compensation, total monetary allowances as a proportion o f monetary compensation is not very large in comparison to the compensation mix currently experienced by the Kenya public service or that o f Tanzania before implementing pay reform. Yet the proliferation o f allowances i s quite problematic for salary administration. It contributes to fluctuations in the monthly wage bill and is difficult to plan for. This is one explanation for the large number o f categories o f arrears andthe amount o farrears inthe wage bill. c. THEIMPACT ONQUALITY OFSERVICES 7.16 As indicated above, the prolonged effects o f declining real pay and general conditions o f service in the public service negatively impacted work motivation, performance, capacity building and commitment to public service. Increasingly, with erosion o f real pay came the depletion o f scarce human and motivational capital in the public service, giving rise to: demoralization and demotivation of subordinate employees and their supervisors/senior civil servants; reduced work effort, declining levels and expectations inperformance; reduced incentives for improving work performance and willingness to accept greater responsibility; the weakening o f accountability and control procedures/mechanisms; diminished ability to recruit and retain qualified managerial, professional, and technical manpower; and reduced commitment to public service. 7.17 In addition to reduced work effort, low and declining real pay may lessen goodwill, increase ill-will and motivation to engage in counterproductive behavior antithetical to increasedproduction, service delivery, good fiscal management, and sound -116- government resource allocation. These may give rise to both the incidence and tolerance for "deviant work behavior" in the public service. Here deviance i s viewed within the context of: a production deviance, which includes work slow downs, misuse o f paid work time to engage innon-work activities, or "time theft"; and/or e property deviance, which refers to the (mis)appropriation or misuse o f public institutions' tangible and/or financial assets for personal gain. 7.18 By combining the resources that they control internal and external to their place o f employment as best they can, adopting their income-maintenance strategies, public servants are perceived as responding pragmatically to the financial predicament in their attempts to maximize their incomes. 7.19 These income-maintenance strategies took on various forms. While some had no direct bearing on work performance and service delivery - though they might affect commitment to employment and work motivation - others did. Those strategies that had the greatest impact on service delivery fell into four categories: work sharing, where public service personnel agreed implicitly/informally between themselves to share work loads to allow workers to pursue other income maintenance activities during official work time, lowering acv~al work-time and generating personnel rotations at staffing levels far below those formally expected; bribery, where public servants solicited payments directly from customers/clients before providing (otherwise) free public services, and/or request payments in excess o f the formal service fees or user charges set bypublic institutions; revenue sharing, where revenue from formal user charges were misappropriatedor mismanaged for individual gain; and resource sharing, where public resources, supplies and equipment were misappropriated, pilfered, or mismanagedfor individual gain. 7.20 Employing such income-maintenance strategies has had significant negative implications for service delivery quality and accessibility, as well as on the govemment's fiscal situation and on capacity buildinginthe public service. D. THELINKTO CORRUPTION 7.21 The relationship between corruption and pay has invoked much debate in administrative and general civil service reform discussions. Not all behavior through which civil servants derive personal gain from the use of public resources need be lumpedinto the a single "corruption" basket. Nevertheless, the low-pay-petty-corruption relationship i s fairly clear drawing from evidence o f the efficiency-wage analyses of -117- economics and the deviant work behavior analyses o f organization behavior and human resource management literature. The pursuance o f many income-maintenance strategies generally leads civil servants into what the organization behavior and human resource management literature commonly refers to as deviant work behavior. But unless actions are taken to punish corrupt behavior, pay increases alone will do little to change the costhenefit calculation and corruption need not abate. Changes in compensation levels can be part of a package to reform civil servant behavior but other elements are essential to reduce corrupt practices. 7.22 A survey inTanzania found that the payment of fair compensation initselfwould not be enough to encourage increasing performance among some segments of the workforce. The survey raised two issues. First, the pay increase must meet the workers' minimum expectation o f fair compensation before they are willing and able to commit fully to their work. Second, improved pay in itself is no guarantee that the workers' choice o f a level o f work effort will match that desired by the government (as an employer). Increased pay may be a necessary though not sufficient condition for increased commitment and performance. Mechanisms, therefore, must be developed to improve accountability, monitor performance, and instituterewards andpenalties. 7.23 Salaries are not the only motivating factor for performance improvement. The lack o f availability o f complementary inputs, as well as institutional, environmental and managerial problems play major contributory factors in low service delivery and performance. These clearly need to be addressed. However, it i s only when the pay and incentive problems are adequately addressed that we can realistically expect the process towards the evolvement o f an effective, efficient and motivated public service to begin and capacity buildingefforts to become more effective. E. RECOMMENDATIONS Actions Taken to Date 7.24 The GRZ has not had an explicit public service pay policy. From independence untilthe early 198Os, the Zambia public service salaries and other terms andconditions o f service have been reviewed after intervals averaging four years through external Salary Commissions appointed by the Government in an attempt to motivate civil servants for increased efficiency and productivity. Salaries were generally adjusted on the basis o f government affordability, past inflation levels, prevailing economic conditions, and to maintain some degree o f parity with public enterprises for similar jobs. As salary adjustments were made in four-year intervals, after a large adjustment inthe initial year, public services had to forgo salary increases in intervening years. Unfortunately, these commissions, while reviewing public service remuneration and conditions o f service generally, offered only piece-meal solutions to addressing the problems o f the public service. This created further distortions across the various groups o f personnel employed bythe public service andwithin groupsthemselves. 7.25 As inflationary pressures persisted andbudgetary constraints became more severe, periodic wage determination by ad hoc salary commission gave way to direct collective -118- bargaining between government and respective unionized public servants. Over time, more and more personnel sought the protection o f unions, as they felt that they could better influence the wage determination process for their occupational group if they took collective action. The notion o f a public service salary structure and salary scales had given way to a collection o f unionized salary scales, where the overall salary structure was nothing more than a collection o f ad hoc derived salary scales. 7.26 As a result o f implicit GRZ pay policy, fringe benefit and monetary allowances have progressively been usedas a major vehicle for increasing compensation, particularly for upper-middle level and senior civil servants. The allowances have included: acting, special duty, hardship, responsibility, non-practicing, commuter/transport, risk, security, extraneous duty, field, overtime, honoraria, accommodation leave, transfer, entertainment, telephone, utility, mileage, subsistence, settlement, uniform among others. This has resultedina situation where allowances and fringe benefits: e have increased significantly in terms o f their number and in value, both relative to salaries andinabsolute terms; e are enjoyed disproportionately by various cadres and grades o f civil servants; and e have in some instances replaced pay as the major component of the compensation package. 7.27 Although initially conceived in 1993, the Public Sector Reform Program (PSRP) was revised in 1997 and jumpstarted then after years o f relative inactivity. The initial PSRP came on the heels o f a long series o f structural adjustment programs that failed to adequately address the government's budgetary situation or the problems in service delivery and access. The short-term objectives o f the PSRP were to reduce government's budget deficit by cutting down its labour force, to rationalize government structures and functions, to strengthen capacity, and to improve operations and maintenance facilities. The long-term objective was to improve efficiency andproductivity o fthe public service. 7.28 Given that the wage bill was and is large relative to government recurrent expenditures, to avoid further crowding-out o f operational and maintenance (O&M) expenditures, salary enhancement within an affordable wage bill was seen to necessitate staffreductions. As public service employment had increased substantially between 1985 and 1991 (Figure 4) (which reflected the willingness of the Government to trade-off increased real wages for higher levels o f employment), reducing the size o f the public service was perceived as a major priority o f the PSRP. The objective o f the Pay and Rightsizing component o f the PSRP was to ensure that the remuneration package was sufficient to cover the basic needs o f public servants and that provided sufficient compensation to attract andretain competent and efficient civil servants. 7.29 The PSRP pledgedto reduce public employment by 25 percent within three years and to improve the conditions o f service o f personnel who remained in an attempt to improve the competitiveness of the public service to attract and retain personnel with the -119- requisite skills and experience. However, no retrenchment took place within the proposed period. The main obstacles to the PSRP retrenchment exercise were two-fold. The first was related to the cost o fthe severance package. The second was the absence o f an objective and systematic assessmentof the employment requirements on a government ministry-by-ministry and department-by-department basis. Under the Emplop e n t Act the mandated retrenchment costs was the equivalent o f about 12 years salary for each retrenchee. Given the political cost and economic cost associated with retrenchment, perhaps the government decided that it was easier to retain workers rather than to retrench them. Figure 4: Public Service Employment Trends, 1978-2001 160,000 140.000 120.000 DI.loo'ooo 80,000 60,000 40,000 20.000 \+=+ ,4." e=++ebb e=+\+=+=+ \& \$+\# ,++ \+h ,++ \# 4%\f ,eQ p' Year Sources: Colcough 1997, Table 4.39. 73; International Monetary Fund, Figure, Number of Public Servants, 1989-99,page 5;Data Centre, MFNP. 7.30 Beginning in 1997, greater effort was made to control employment levels. But the initial reductions in employment numbers could be attributed almost exclusively to a reduction in the industrial class personnel, those with no permanent employment tenure in the public service. Employment levels were reduced through a retrenchment and voluntary early retirement scheme in 2000. Still employment has shown a rising tendency again. After averaging nearly 104,000 per month during the first nine months o f 2000, personnel numbers rose steadily on a month-to-month basis from October 2000 to August 2001, from about 105,800 to about 113,260, before declining to an average o f about 109,700 per monthduringthe last three months of 2001. -120- FutureAction Plan 7.31 The status o f the Government's pay reform strategy should be described in two phases: (1) development o f the overall strategy and principles, including tentative cost projections, and (2) updating and partial implementation o f year 1 in 2003. Under PSCAP, GRZ commissioned a short-term consultant to conduct a study o f pay and employment issues. The objective was to assist GRZ to develop a strategy for enhancing pay and conditions o f service in the public sector that laid out broad policy options, established links to major parameters o f the policy and identified the principles upon which the pay policy will be built. Development of the pay and conditions o f service policy will take time and its introduction and translation into pay awards will also take a numbero fyears. However, there has beenapressingneedto introduce measures quickly to address critical pay-related constraints to improved service delivery within the context o f a pay policy. Inparticular, immediate steps have been required to boost remuneration for managerial and qualified professional and senior technical personnel. The medium- term pay reform strategy (MTPRS) was a first step towards the formulation o f a Government pay policy, which would address all aspects o f work motivation and requirements for attracting andretaining personnelinthe public service. 7.32 Salary enhancement is only prudent and sustainable when undertakenwithin the context o f the likely resource envelope. Government recurrent expenditures must seek to strike an appropriate balance between the public service wage bill and expenditures on operation and maintenance. Regarding the size o f the public service, the proposed reductionshightsizing measures proposed in 2000 by KPMG aside, it is unclear whether there i s scope for additional employment reduction. With the vast majority of public servants being employed in the priority areas (75 percent being engaged in education, health, and the protective services), there is little scope for financing pay reform fi-om downsizing o f the public service alone. Though, there is still need to rationalize public service employment and to shift away fi-om a low skilled workforce to a more skilled, professional and experienced public service. A major constraint on salary enhancement inthe Zambia public service is the slow rate ofreal growth, which restricts the resource envelope. 7.33 Inthe absence of any other statement ofpay reform priorities, those presented in the revised PSRP document (Republic of Zambia, 1997) are still taken to be relevant to the current situation. The specific proposals on pay reform are: (a) Control the size and growth o f government employment; (b) Rationalise and enhance public service pay by: (i) eliminating the distortions and anomalies that have crept into the system; (ii) providing equal pay for equal work; and -121- (iii) paying salaries commensurate with skills experience and responsibilities. (c) Senior and middle ranking staff should have their pay progressively decompressed to the new target levels to increase their competitiveness with pay levels obtainable inthe non-government sector inZambia. (d) At lower levels, pay increases would be only sufficient to offset inflation. Pay increases only for inflation (i.e., no increase inreal pay levels) for all lower-ranking staff, (e) All employees should have their pay consolidated to include all allowances. Allowances should be progressively absorbed into basic pay, withjunior ranks receiving `pay protection' to ensure that they are never worse off. 7.34 With these goals inmind, the PSCAP consultant proposed four possible scenarios shown inTable 15. Ofthese, Scenario 4 was the preferred one by Government because it would do most to decompress the salary structure and make the civil service more competitive. Table 15 :Scenarios Assumptions Matrix , Scenario 1 The Wage BilUGDP ratio Size o f the public service N e w salary structure declines by 0.5% o f GDP remains constant at 2001 introduced. Salary from 2002 to 2003 and level. differentials (increment and remains constant grade differentials) remain thereinafter. constraint through the period. Same as Scenario 1. Size o f the public service N e w salary structure reduced inline with the introduced. Salary I KPMGreport differentials (increment and I I recommendations. grade differentials) remain constraint through the period. F T S a m e as Scenario 1 . Same as Scenario 2. N e w salary structure introduced. Salary I differentials (increment and grade differentials) rise I I S c e n a r i o 4 through the period. Same as Scenario 1. 1 Same as Scenario 1. introduced. Salary differentials (increment and grade differentials) rise through the period. Allowances are eliminated for -122- 7.35 A key feature in Scenario 4 was the partial consolidation o f allowances and a move towards a "clean" wage bill. It should also be noted that compensation could be enhanced even hrther with a reduction in the size o f the public service. But as noted indicated earlier, the scope for employment reductions in Zambia is likely to be much smaller than many analysts expect. 7.36 Inasmuch as donor-supported intervention complements the medium-term pay strategy it will facilitate improved performance and capacity building in the ublic service. Years o f reliance on ad hoc local cost compensation (LCC) arrangements67have evolved into a situation where the Zambia public service has to compete with donors for scarce managerial and qualified professional and technical personnel. While a stated intention o f donor salary top-up arrangements may be to increase the public services ability to attract and retain better-qualified professional and technical personnel, the ad hoc nature o f this type o f donor intervention has unintended(negative) consequences. It i s not yet clear on how best to move away from this entrenched practice. However, as one option, the Government (through the PSCAP consultant) has proposed that donors pool a proportion o f the resources that they would normally earmark for LCC and contribute to the establishment o f a scheme that will better complement the medium-term pay reform strategy. By doing so, ad hoc LCC arrangements could be transformed into a systematic scheme where donors collectively support the government's efforts to improve its capabilities to attract, retain and adequately motivate its personnel and build the requisite human resource capacity to facilitate improved service delivery and strategic outputs. This scheme i s referred to as the Selective Accelerated Salary Enhancement (SASE) scheme. It is proposed that through the SASE scheme donor-hnds support an accelerated pace o f salary enhancement for managerial, professional and technical personnel. As one option for addressing the urgent need to improve pay and incentives for key staff, the SASE scheme merits further review and consideration. Further explanation o f the justification and broad modalities o f SASE is found in the annex of this chapter. 7.37 Finally, it must be stressed that increased pay may be a necessary, though not sufficient, condition for increased commitment and performance. Mechanisms, therefore, must be developed to improve accountability, monitor Performance, and institute reward and penalties. Consequently, a new performance management system and employment policy i s also o f critical importance. Specific proposals conceming the development o f a performance appraisal system, along with complementary policies for performance-based 61Given the pay and incentives problems confronting the public service, donors sometimes provide incentive payments and salary supplements to key public service posts where incumbents are engaged on donor-funded projects and programs. Such payments are referred to as local cost compensation (LCC). These payments are perceived to attract, retain and motivate personnel, as well as to ensure higher levels o f performance and commitment from otherwise lowly paid and (potentially) demotivated public servants. LCC salary supplements are generally far higher than the basic salaries paid to personnel with comparable responsibilities, though these vary significantly from donor to donor, project to project, as well as within a project depending o n the funding arrangement. -123- pay rewards andjob evaluation and re-grading are found inthe report "A Medium-Term Strategy for Enhancing Pay and Conditions of Service in the Zambian Civil Service" prepared by Ted Valentine on behalf o f PSCAP and the Management Development Divisiono f Cabinet Office. 7.38 Initial Implementation of Pay Reform. In late 2002, it became evident to Government that the underlying assumptions on which the MTPRS was based were outdated and neededrevision. Inparticular, the estimated wage bill had risen from 6.3% o f GDP to over 8%. GRZ found that anticipated GDP growth in2003 would enable it to offer approximately K 142 billion for salary increases in 2003 before hitting its IMF- agreed upon ceiling o f 8%. This amount was recorded inthe 2003 Yellow Book. 7.39 Unfortunately, pressure from unions and other groups within the public sector, led to GRZ agreeing to pay increases for 2003 that may cause the wage billto exceed the 8% o f GDP target. Finding off-setting savings will become a important priority for the remainder o f the year. Finalizing the severance packages for those who have been identified for retrenchment will help;62while additional savings should be available from cleaning the payrolldatabase. 7.40 With the 2003 pay increases, GRZ began the process o f decompressing salaries and reducing the number o f separate salary scales. Pay increases were especially weighted toward mid-level technical and professional positions, where retention and recruitment i s especially difficult. Yet, negotiations with the civil service and teachers unions failed to achieve a consolidation o f allowances into base salary, and this will need to be pursued again in2004. 7.41 Insummary, the mainrecommendations arethat GRZ: (a) Complete the payroll data cleaning exercise and retrenchment packages, anduse the savings to off-set the cost o f the 2003 salary increases. (b) Over the short to medium-term, continue to target salary enhancement towards professional and technical personnel that the Government has most difficulty in attracting and retaining and who are most critical to efforts to raise public service efficiency and the quality o f service delivery; (c) Rationalize the pay structure by substantially reducing the proportion o f salary paid as allowances in2004; (d) Conduct a special assessment o fwhether SASE or another program can be developed to target immediate pay awards to critical staff based on performance or special skill-related job characteristics. 62About $10 million o f the severance package i s being funded through PSCAP. -124- 8. SERVICEDELIVERYINPRACTICE: LEAKAGESAND EQUITYINTHE FLOWOFFUNDINGTO EDUCATION INZAMBIA63 A. STUDY METHODOLOGY 8.1 This chapter summarizes the expenditure and service delivery survey (ESDS) for the basic education sector carried out as part o f the PEMFAR in 2002. Basic education was selected for being a priority sector for poverty reduction and was deemed a good candidate to test the linkage between budget design and execution on the one hand and outcomes/results on the ground, on the other. The main ESDS finding i s that allocating more central budgetary resources to the sector does not, by itself, guarantee better educational outcomes: budgetary resources (ifreleased by the Ministry o f Finance) may not reach the intendedbeneficiaries, and even ifthey do, these resources may not result in bettereducationaloutcomes. 8.2 While the Government o f the Republic o f Zambia (GRZ) has, from independence, recognized the importance o f education and increased budgetary allocations to the sector, the sharp decline from the mid-seventies onwards in copper prices has led to a commensurate decline in Government revenue and per capita incomes (which declined from USD590 in 1975 to USD300 in 2000). Because o f these declines, the education sector in Zambia faces several obstacles and, consequently, most education indicators have deteriorated over the last decade. To overcome some o f these obstacles, the Government and the donor community have been working closely to improve both educational inputs and outcomes. In this regard, it i s important to assess (i) the current record o f the educational sector in delivering services to the intended recipients; and (b) whether some refonndchanges undertaken by the Ministry o f Education have resulted into desired educational outcomes? 8.3 The ESDS sought to address these questions through a detailed evaluation and tracking o f educational expenditures earmarked and assess the extent to which such resources actually reach schools and the poor insociety (i.e., whether such expenditure i s progressive or pro-poor). The Zambia ESDS addressed four main issues: (a) What percentage o f resources i s spent at each level o f the administrative hierarchy? 63ThisChapter is a summary based onDas et.al. (2002). -125- To what extent i s variation in funding across districts and schools explained by `funding-formulas' that relate budgetary allocation to districtlschool characteristics such as enrollment or the number o f schools? To the extent that there i s variation infunding across districts and schools, what is the relationship between this variation and characteristics o f districts and schools (for instance, is it the case that richer districts/schools systematically receive more funds)? How important i s private funding in the provision o f education and how does this component of funding relate to equity in the overall funding o f education? 8.4 The flow o f funds from the Ministry o f Education Headquarters to schools in Zambia i s organized through a three-tiered administrative hierarchy involving provincial offices, district offices and schools. The normal flow of funds i s ina `top-down' manner- at each tier o f the hierarchy, funds may be allocated directly either from the Ministry Headquarters or from international and local donors. The ESDS classified the allocations o f funds into four different categories: rule-based allocations to schools, discretionary allocations funds to schools, rule-based allocations to teachers, and discretionary allocations to teachers. The characteristics o f each o f these four categories is described Rule-Based Allocations to Schools: This i s a fixed-grant allocation to schools, that is either USD600 or USD650, depending on the type of school. It i s important to note that this i s not a per-pupil allocation- the grant is independent of the enrolment and other characteristics o f the schools. Discretionary Allocations to Schools: This i s monies given to schools over and above the rule-based allocation. Such funds are allocated at the discretion of districts andprovincial education officers. Rule-BasedAllocations to Teachers: The payment o f salaries and monthly allowances i s administered through a centralized payroll in Zambia, whereby teachers directly receive their 'deposit-slips' without any money going through either provinces or districts. Such payment is regarded as beingrules-based. Discretionary Allocations to Teachers: In addition to regular salaries and monthly allowances, teachers are also entitled to time/event benefits, such as leave and transfer/relocation (paid when a teacher switches schools) and funeral benefits. Because of persistent inadequate funding from the Ministry Headquarters, the payment o f such benefits is left to the discretion o f the province and district level administration despite the fact that the terms and conditions o f service for teachers categorically specify -126- that they are eligible to be paid such allowances. The observed frequent withdrawal o f labor by the various teachers unions is, among other factors, as a direct result ofthis. 8.5 Inorder to improve education service delivery, the education sector is currently piloting decentralization through the creation o f district education boards. The ESDS divided provinces into those with district education boards and those without. Such a distinction is necessary since in the former provinces very little money flows from the province to the district; districts receive money directly from the Ministry both for rule- based and discretionary allocations. Inthe latter case however, while districts continue to receive money directly for rule-based allocation, all money for discretionary allocations is first allocated to the concerned province and from there onwards to the district. For thisreason, the Ministryterminology to refer to the former as decentralized provinces and the latter as centralized provinces was adopted by the ESDS. In the ESD sample there were two o f each: Lusaka and Copperbelt province are decentralized while Northern and Eastern are centralized. Figure 6 diagram below shows how funds are allocated across these two types o fprovinces. Ministry of Education Case IV donors (primarily the Program for Advancement o f I I Girls Education) I I v v I v Province I Province (Centralized) I I I (Decentralized) I I -.-.- I I I I v I Dishict I I District (Centralized) 4----.--I L .-.-.-.-.- .) (Decentralized) -r I I t YT Discretionary Households Schools Rule -Based B. EXPENDITURE TRACKING 8.6 The ESDS tracked all non-salary funding flows (recall that most payroll flows are allocated directly to teachers) through the administrative hierarchy and established the following characteristics: -127- On average about K28,OOO per pupil enters the educational system for the four provinces o f the ESDS. However this average amount hides a significant degree o f variation within these provinces with, for instance, Eastern Province (K44,300) receiving more than double the per-pupil fundingreceivedby Copperbelt (K19,OOO). Of this amount, discretionary Funds at the level o f the province and district account for 70 percent o f all funding while rule-based funds account for the remaining 30 percent: Thus rule-based funds allocated through the fixed school grant o f USD600 (USD650) account for less than one-third o f all the fundingreceived. Between one-sixth and one-third o f total funding inthe system eventually reaches schools: O fthis total amount between 14 percent (Lusaka) and 34 percent (Copperbelt) eventually reaches the schools as a combination o f rule-based and discretionary spending while the rest i s spent at the provincial and district levels. Decentralization has shifted spending fi-om the provincial to the district level, but has not resulted in greater disbursements to schools: There are important differences between centralized and decentralized provinces up to the level o f the district: centralized provinces spend more at the provincial level compared to those that have been decentralized. Thus in Eastern (38 percent) and Northern (18 percent) provinces a higher percentage o f funding is spent at the provincial level compared to Copperbelt (9 percent) and Lusaka (5 percent). However, the extra funding that reaches the districts indecentralized provinces just results in higher spending at the district level and i s not associated with greater fundingto schools inthose districts.64 8.7 These findings are summarized inthe funding-disbursement graphs below (Figure 7) where we plot Kwacha per pupil (K) funding at each level o f the administrative system. The amount that finally reaches the school is much less than the total amount available inthe system and this decline i s the sharpest inEastern province (despite having the largest average allocation per pupil inthe sample). Ineffect, the differences in funds available at the province level are due to the nature o f the rule-based component o f funding, whereby provinces with lower average enrollments receive greater per-pupil fundingthan provinces with the same number o f schools but higher average enrollments. 64This, initself, might not necessarily be a bad thing as we could be observing a shift of responsibilities from the provincial to the district level. For example, an increase in the number of vehicles at the district level in decentralized provinces would naturally increase levels o f expenditure; and an increase in school inspection visits by district officials might not be recognized as an increased "funding" by individual schools as the latter might only interpret increased receipts from the district by goods/funds received and not necessarily the services provided by the district education office. -128- Consequently, the low disbursement curves inthe decentralized districts when compared to the centralized district are as a direct result o fthe rule-based component. Figure 6: FundingDisbursementinZambian Education FundingDisbursement DecentralizedProvinces FundingDisbursement:CentralizedProvinces 50000 "1 P 40000 P LusakaProvince zp20000 30000 \ -..- B zc10000 0 04 7 Funds: Funds: Funds: Funds: Funds: Funds: Funds: Funds: Funds: Funds: Total Province District S W s Schools Total Province District Schools %hods (LBI (UBI (LBI (UBI AdministrativeLevel Administrative Level C. LEAKAGES 8.8 Following the methodology used inUganda (Ablo and Reinikka, 2000) we define leakage inthe Zambian educational system as the ratio o fwhat schools actually receive to what they were supposed to receive. The exact equivalent o f the per-pupil funding in Uganda i s the fixed-school grant inZambia. We thus define: 8.9 The Zambia ESDS went beyond an estimate o f leakage based on allocated funds. Specifically, it used an innovative approach to provide a complete picture o f the funding o f educational institutions that examines all sources o f funding for the school instead o f restricting attention to those components that are clearly defined with an allocation rule. Thus, in the case o f discretionary components where this methodology fails (since there i s no rule about the amounts that schools are supposed to receive inthe first instance, it i s not possible to determine what constitutes 'leakage' inthe system) the ESDS tracked the amounts that schools receive, and then directly examine the equity implications o f such allocations. The section on leakage establishes the following: -129- (a) For rule-based allocations to schools it was found that the Zambian educational system i s working efficiently and there i s little evidence to suggest that funds earmarked for disbursements are not reaching their intendedbeneficiaries. Specifically more than 90 percent o f all schools (95 percent in all provinces except Lusaka) had received the rule-based allocation at the time o f the survey and delays in disbursement rather than leakage o f funds was a more likely explanation in the case o f schools that had not received the grant (this grant was disbursed only one month prior to the survey). (b) For rule-based allocations to teachers, it was found that salaries and clearly defined allowances are disbursed efficiently. There i s some evidence o f delays in the updating o f the payroll system as well as significant arrears in the case o f allowances that are not clearly specified. Hence in the case of salaries 95 percent o f all teachers had no outstanding amounts. For allowances where there i s a clear specification based on location (hardship allowance given to teachers inrural locations) or status (teacher-trainee allowance), less than 15 percent of all who were supposed to receive the allowance were overdue by six months or more. However, inthe case of overtime allowances (which must be filed every term) and allowances that resulted from a change inthe status o f the teacher (such as added tasks with commensurate allowances) this percentage jumps to 50 percent. For all allowances there i s considerable evidence that lags in updatingthe payrollregularly result indelays o f 1-3 months inpayment. (c) Turning to discretionary allocations to schools it was found that the positive results obtained earlier no longer hold- less than 20 percent o f schools receive any funding at all from discretionary sources combined. The crucial importance o f rule-based funding at the level o f the school i s also highlighted by the change in the relative shares by sources as we move down the administrative hierarchy- at the level o f the province, the share o f rule-based funding ranges from 8 percent to 40 percent with a median o f 12 percent. Moving down to the districts, this share increases to between 19 percent and 63 percent with a median o f 44 percent. Finally at the level o f the school, this share ranges from 1.6 percent to 100percent with a median of 99 percent- more than 75 percent o f all schools have received cash resources only from rule-based sources inthe current year. (d) Similarly, for teachers there are substantial amounts overdue through one- time benefits and payments and in some province the overdue amount accounts for three times the monthly salary o fthe teacher. (e) The survey also found that the discretionary funding that is eventually disbursed to schools is disproportionately large compared to rule-based funding. Conditional on a school receiving such funding, it could account for as muchas 60 times what it would get under the fixed-allocation grant. -130- (f) This difference suggests that discretionary allowances may be earmarked for certain kinds of expenditures that require very large outlays. This i s confirmed through an analysis o f spending at the level o f the school it was found that most discretionary funding i s associated with infrastructure. A simple bivariate correlation confirms a high degree o f association (0.53) between per-pupil spending on infrastructure and discretionary funds received. (g) There are thus two potential explanations for the patterns o f discretionary funding and school expenditure based on a distinction between recurrent and capital expenditure. (i) Onepotential explanationis that schools are provided discretionary funds for capital expenditure on a rotating basis. Thus if a school were `rehabilitated' every 10 years we would only expect to find 10percent o f schools inany given year receivingmoney for capital expenditures. In addition, the amounts of such expenditures received would be very large compared to amounts received for recurrent spending. According to this hypothesis the pattern o f funding and expenditure that we observe is a reflection o f the 'lumpy' nature o finfrastructureinvestment. (ii) A second explanation could be that since rule-based funding is clearly defined with a simple allocation rule, it is extremely difficult for any political-economy considerations (such as `capture' o f funds by elites) to impact on school allotments. On the other hand, discretionary funding is not associated with any such rule. The pattern o f funding that we observe then is a reflection o f the difference between rules and discretion- the few schools that receive large amounts are `special' schools that have greater bargaining power with the higher levels o f the administrative structure. 8.10 The analysis o f the equity dimension o f education expenditure in section D discusses this concern and relies on the following observation: if the first explanation is correct it i s unlikely that the flow o f discretionary funds to schools will be correlated with wealth. If schools are provided with infrastructure funds on a rotating basis, it seems natural to assume that the sample o f schools receiving such funding will be a mix o f schools in rich and poor areas. Consequently, systematic differences in funding by wealth levels would lead us towards an explanationbased on the difference between rules and discretion, rather than one based on a sole distinction between capital and recurrent expenditures. -131- D. EQUITY 8.11 As with the previous sections, the ESDS examined funding-equityinthe Zambian educational system separately for rule-based and discretionary allocations and focused on the relationshipbetween funding and three related variables: (a) Urbanization: D o urbanschools/urbanized districts receive more than their rural counterparts? (b) Wealth: D o richer schools/districts receive more than their poorer counterparts? (c) Distance to Administrative Offices: D o schools that are closer to administrative offices receive more than those that are further away? 8.12 The main findings are summarized below: (a) Rule-Based hnding i s highly progressive with greater per-pupilamounts allocated to less urban andpoorer schools and districts. Giventhat there is no evidence o f leakage, one can see that the equity aspects o f rule-based funding will depend (both at the district and the school level) on the relationships between school size and these three variables (recall that the rule-based component o f funding i s a per-school grant o f USD600 or USD650). First, at the district level, those with smaller average enrollments will receive higher funding. To see this consider two districts each with 100 schools but average enrollments o f 50 (District A) and 100 (District B). Since the rule-based allocation i s a fixed per-school grant, both districts receive exactly the same allocation. However, since district A has a smaller number o f students than District Bythe average per-pupil grant will be much higher in District A compared to District B. At the school level, exactly the same reasoning applies- schools with higher enrollments will receive less per pupil fimding than those with low enrollments. Table 19 below shows how enrollment relates to the three variables o f interest: (b) The table shows that bigger schools tend to be richer (high enrollment schools are more than one standard deviation richer), more urban (82 percent o f all high-enrollment schools are urban compared to 5 percent o f all low-enrollment) and closer to administrative offices (62 percent o f all high-enrollment schools are less than 5 km from the district office). Hence, we would expect to find that their per-pupilrule-based allocation is lower than their rural and poorer counterparts. -132- Table 16: Enrollment, Urbanization, Wealth, and Distance Level Urbanization, Wealth and Low Distance Enrollment Schools District Average District Wealth -0.52 I.14 3.67 Level Schools in Urban Locations 2% 4% 66% (%) School Average Enrollment 275 719 1695 Level Urban Schools (%) 5% 25% 82% Schools less then 5 kmfrom 4.84% 22.11 62.3 district office (%) Schools less than 5 kmfrom 5.2 13.1 21.5 provincial office (%) Average Wealth Index -0.66 .0.07 3.75 Source: Education Service Delivery Data Thus the poor schools receive four times the rule-based allocation o f rich schools and rural schools receive three times the allocation of urban schools. Infact, the findings suggest a disturbingtrend: whentracking expenditure to the districts, provinces allocate discretionary funding only slightly more to poorer districts. As a result, for total receipts at the district level the progressive nature of fundingpattems i s retained, albeit smaller than what it would be underrule-based fundingonly. Any discretionary funding in rural schools has a very highprobability of being disbursed to wealthier schools, while within urban schools allocations are wealth neutral. Since the monies disbursed are orders o f magnitude higher compared to the rule-based allotment, the progressive nature o f the rule-based allocations disappears- at best, funding that flows from the district to the school is wealth neutral (with no significant differences in received amounts across rich and poor schools), and at worse it is regressive for rural andprogressive for urban schools. This arises primarily from higher class sizes andthe greater use o f trainee- teachers in rural locations. Hence, once we factor in staff compensation, the results show that the only progressive component o f the Zambian education system i s the rule-based allocation. Once per-pupil teacher funding is added in, the entire education funding system becomes regressive with poorer schools (K14,53 1 per pupil) receiving less than richer schools (K19,826 perpupil). 8.13 These findings are summarized in the graphs o f funding inequality (Figure 8) below. The horizontal axis inthe funding inequality graphs is the wealth distribution and the vertical axis is the share ofpublic fundingthat each group receives. Each curve inthe -133- graphs shows the share o f public funds that accrues to (for instance) the poorest X percent o f the population. The `line o f perfect equality' i s then the diagonal that shows fundingshares ifall funds were distributedequally amongthe population. Figure7: Inequalities inPublic FundinginZambian Education Inequalities in Public Funding:All - . . Inequalities in Public Funding:Within Urban - . . 0% Poorest Poorest Poorest Poorest All 0% Poorest Poorest Poorest Poorest All 20% 40% 60% 80% schools 20% 40% 60% 80% schools Wealth Wealth -Inequality . . in Public Funding:Within Rural 1 - 0.8- a - ......Shares:Rule-Based IShares: Discretionary U --Shares:StaffPayments -Perfect Equality 0% Poorest Poorest Poorest Poorest AII 20% 40% 60% 80% schools Wealth 8.14 These figures clearly demonstrate the findings o f this section: Rule-based fundingis always progressivewith higher shares for the poorer sections o f the population (the curve always lies above the diagonal so that (for instance) inthe entire sample, the poorest 20 percent o f the population receives 38 percent o f all rule-based funding). Moreover, such funding is progressive not only inthe entire sample o f schools, but also within urban and within rural schools. Staff-Compensation i s always regressive, with the curve lying below the diagonal for all three samples (the entire sample, and within urbadrural). -134- 8.15 Finally, discretionary funding i s regressive inthe entire sample as well. However once we look at such funding within urbadrural schools we find that within urban schools discretionary fundingi s progressivewhile within rural schools it is regressive 8.16 This then opens up the question o f household inputs into education- if there are huge funding differentials between schools (for instance, the rule-based component impliedthat per-pupil funding could vary from K1,889 to over K8,000 depending on the school), and some schools receive a lot less than others, do households adjust their own contributions to account for the level o f school receipts from the public funding system? If yes, how does this household spending impact on funding equity in the Zambian educational system? We present some preliminary results on this question from our household survey that complements our construction o fthe public expenditure system. E. PRIVATEEXPENDITURES 8.17 The issue o f school funding from household contributions is at an important crossroads in Zambia. Anecdotal evidence from a number o f studies during the 1990's suggested that schools had started charging high Parent-Teacher Association (PTA) fees and a concurrent decline in net enrollment during the same time period led to an association between enrollment and high PTA fees. Consequently, in April 2002 (3 months prior to the fielding o f the ESDS) PTA fees were abolished for primary and basic schools (although secondary schools could and do continue to charge fees) and the government reiterated its commitment to free education. With this background inmind, our key findings regardingprivate expenditures are as follows. (a) The announcement has had the desired effect with PTA fees decreasing to less than 30 percent o f their Year 2000 values, although this decline is concentrated primarily in the urban areas o f Lusaka and Copperbelt. In addition, the ESDS found that the gradient between PTA fees and school wealth declined sharply during the same period. (b) Consequently, for all provinces, but especially for Eastern and Northern, public funds are the most important source o f financial flows to the school; thus public funds comprise 96.3 percent o f the total funding o f the school in Eastern province (the most reliant on public funds) and this decreases only to 81.6 percent for Lusaka (the least reliant). (c) The decreased importance o f household contributions to school funding would suggest that the addition o f such sources o f funds should not alter or significantly exacerbate results regarding the regressive nature o f school funding obtained in the previous section. Thus, for instance, the difference between schools with poorer and richer households in private contributions to schools i s K1,300 per pupil, but for public funding this increases to K3,100. Similarly, schools that are closer to the district office raise K300 more per pupil, but this difference i s dwarfed when compared to the K8,200 difference in public funds within the same categories. Consequently, the addition o f private contributions to schools in -135- educational funding leaves unaltered our basic equity results and does not exacerbate differences in funding between rich and poor schools by any significant amount. (d) However, to examine only private contributions to school funds can be misleading if most inequality in private expenditure is at the level o f the household i.e., through private, household level non-fee spending on children. An examination o f such spending was based on a household survey carried out simultaneously with the ESDS. (e) The household survey results show that non-fee expenditures on children are seven times the corresponding expenditure on fees, and are the main source o f inequalities in private expenditure. The inclusion o f private expenditure worsens the regressive nature o f the public funding system: the poorer 50 percent o f the population sees a decline in its share o f educational expenditures from 40 percent to 34 percent once private expenditures are factored inthe analysis. (f) The examination o f the degree o f substitutability between private expenditures andpublic funding showed a strong relationship. Households decrease their own private expenditures whenpublic funding increases. 8.18 The main findings on inequality in private funding are summarized in the two graphs below. Figure8: Equity inPublic and PrivateFunding - 40000 -- Public non-salary funding *Salary Funding -.- / \ a . Non-Fee expenditure on 30000 P * * Fee expenditure on child child a Q z YQ 20000 .-0a . E 'El 5 10000 U 0 Poorest 20%-40% 40%-60% 60%-80% Top 20% 20% Quintiles of Wealth Distribution -136- 8.19 Figure 9 plots the funding per pupil from public non-salary funds, public salary funds andprivate fee and non-fee expenditure against wealth quintiles. As is clear from this figure, the major sources o f funding in this sample of schools are public non-salary funds and private non-fee expenditure. Figure 10 then plots the funding-shares graph for public and private funding, where the horizontal axis represents the share o f the population and the vertical axis the share o f funding as before. This figure demonstrates that while there is inequality in both public and private funding, inequality in the latter dominates that inthe former with the funding-share curve always lying further below the diagonal. Figure 9: LorenzCurvesof Public and PrivateFunding 1 - c m % 0.8- -m S - 3 0 6 - - Share of Private Funding Line of Perfect Equality 0 - 4- + -; v) Q/* ;$ ;$ ;$ s s 0 n88 Q 8 % Q 88 Q8 8 z 3a c Wealth F. DISCUSSION CONCLUSIONS AND 8.20 This ESDSmakes three contributions to our understandingo f educational funding in the Zambian context. The first is to show that funding characteristics are closely linked to the type of funding that i s disbursed. In the case o f rule-based funding, the Zambian educational system works efficiently and there is no evidence that such hnds are diverted from their stated purpose. For discretionary funding however, the majority o f such funds are spent at the district andprovincial levels, andthe rest i s allocatedto less than 20 percent o f all schools. 8.21 The second contribution is the analysis o f funding equity using the wealth o f pupils inthe school. The specific rule used inthe case o f rule-based allocations has ledto greater per-pupil funding for poorer and more rural schools. However, these allocations are the only progressive disbursements inthe ESD survey. Per-Pupilstaff allocations are higher in urban and richer schools. The evidence on discretionary allocations suggest that higher disbursements to richer schools within rural areas and wealth-neutral -137- allocations within urbanareas. Once all sources o fpublic funding are factored in, public school fundinginZambia i s regressive with almost 30 percent higher allocations to richer schools. 8.22 The third contribution i s to show how private expenditure at the level o f the household impacts on equity in educational funding. Non-fee expenditures incurred by households on children, rather than contributions to school funds through PTA and other fees, are the major source o f inequalities in the current environment. Moreover, households tend to decrease such private contributions when public funds to the schools increase. 8.23 These findings would suggest that a greater percentage o f all funding should be allocated on a rule-based basis to ensure that schools receive a larger share o f funds inthe system. It was initially thought that the process of decentralization would partially hlfill this need- since more money would flow directly to the districts; accountability and therefore disbursements would be higher. Unfortunately ESDS did not find strong evidence o f this happening. Decentralization seems to have only shifted spending from the province to the district, and in terms o f funding equity; it is precisely at the district level that richer schools are receiving higher discretionary funds than their poorer counterparts. 8.24 However even if rule-based funding were to be increased, there would be two subsidiary implications that would have to be carefully evaluated. First, the current rule- based allocation has the implication that schools would fare better in terms o f per-pupil funding ifthey could decrease their enrollments. The more common school fund rule is based on funding that is directly proportional to the number of enrolled children. There i s no guarantee that this would work as well as the current, unambiguous rule. One suggestion therefore would be to continue with the current rule (which also has the desired equity implications) but to monitor enrollment carefully through the regular data collected both administratively andperiodic surveys such as the Zambian School Census. 8.25 Second, when public funds to schools increase, this crowds out private spending. While the results presented here are preliminary, there i s evidence that this crowding-out can be fairly large. Thus, public funds may be far more effective at addressing inequalities across rather than within villages- this would suggest targeting at the level o f schools with greater funding to poorer regions. -138- PARTIV. SUMMARYAND A PROGRAM OF PRIORITY ACTIONS 9. A PROGRAMOF PRIORITYACTIONS 9.1 Thus far, the report has discussed specific actions to address issues related to public expenditure management, procurement, and financial accountability with a view to improve poverty reduction outcomes. While these actions are inter-related, their implementation has to be sequenced carefully taking into consideration the country's capacity, scarce resources, and the overall socio-political environment. This chapter discusses the vision that the Government aspires to in the medium term and, in that context, a program o f priority actions that would enhance Zambia's accountability, transparency and predictability inpublic spending. It lists specific policy actions for each o f the next three years, and builds upon measures proposed inthe HIPC assessment and action plan (AAP)in2001. It i s proposed that during the first year, donors will provide specific financial and technical assistance either through a pooled arrangement or individually. As GRZ PEMFA capacity increases, it is expected that financial aid could increasingly be channeled through the budget, and that by the end o f three years, a significant proportion o f donor funding could be provided inthe form o f budget support, inthe context ofthe PovertyReductionSupport Credit (PRSC). 9.2 Zambia is at a crossroads. The credibility o f the public spending is at stake both internally and externally. Internally, the cynicism regarding the budget and its lack o f impact on poverty is widely shared by the public, while externally the support o f donors is waning reflecting the doubts and difficulties that donors have had in tracking budget implementation. The new administration has recognized this excellent and unique opportunityto reverse this dangerous trend through embarking on PEMFARreforms. At a time o f diminishing resources, there are important financial gains from undertaking these reforms. As shown in the annex on procurement, the savings from procurement alone could add up to USDSO million. Other gains and increasing support from donors could add up to twice or three times that amount. A. MEDIUM TERMVISION FORZAMBIA 9.3 Before discussing in detail the recommended program o f priority actions, it i s important to specify what the outcome o f the PEMFA reform program would look like in the medium term (e.g., after five years o f reforms.) First, the budget would be prepared infullconsultationwith the technical ministries, and inamedium-term (e.g., three year) framework where implications on future recurrent expenditures o f the existing budget are made explicit. The budget would make explicit all transfers to other levels o f the government and other entities such as parastatals. The budgetwould include both current spending as well as capital spending, and projections on these would be supplied by -139- donors on a regular and accurate basis. The budget would reflect what has been agreed in the context o f the PRSP. It would be classified correctly, and one could trace where the money goes through the system. There would be a highlevel o fpredictability inthe flow o f funds to ministries and to spending units. Commitment control would be fully instituted and monitored on a weekly basis. Arrears would have been paid o f f and no new arrears would be emerging. The budget outtum would come very close (90 percent) to the plan, both inaggregate and by sector and subsector. Consequently, there would be zero, or very limited supplemental appropriation. Internal reports would be received within 4 weeks. Audited accounts would be presented within 6 months. Wages and non-monetary awards o f civil servants and public sector employees would all be consolidatedandthe wage bill would now accurately reflect, to a large extent, the explicit as well as the implicit compensation packages for the employees. Parliament and other development partners would be involved both in budget preparation and execution. Procurement would be undertaken efficiently and timely: up to 80-90 percent of procurement undertaken at the line ministries instead o f at the National Tender Board; bothprocessing time and corruptionwill be significantly reduced. 9.4 Achievement o f this medium term vision will require implementation o f all the recommendations inthis report. But it is important to sequence and prioritize the policy actions discussed below to achieve the above goals in accordance with the capacity and resources o f the country and this chapter therefore seeks to identify the priority actions over each o f the next three years that will be necessary to make steady and significant progress towards achieving the vision. Actions can be classified into the following broad categories: (i)measures to restore the credibility o f the budget through improving coverage, preparation, predictability and execution o f the budget; (ii)measures to improve budget reporting and monitoring; (iii) measures to improve procurement; iv) measures to improve financial accountability; and (v) supporting measures to ensure that PEMFAmeasures work. B. THETHREEYEARPROGRAM 9.5 At the PEMFARworkshop held in September 2002, it was proposed that ajoint GRZDonor Steering Committee be set up to direct the PEMFAR reforms and to ensure that these reforms will be implemented properly and in a timely fashion. It is proposed that the Steering Committee be chaired by the Secretary to the Treasury. The main champion for leading the implementing o f the reform could be thesecretary to the Treasury, MFNP. There will also be a Secretariat. This Secretariat should be incharge o f costing each o f the above measures and o f coordinating with donors in order to gather financial andtechnical support for implementation. 9.6 To achieve the above objectives, the following measures, grouped by broad categories, needto be taken over the three-year period. For convenience, these measures are separately presented for each o f the three years. -140- r Table 17: First-Year PriorityActions a) the- budget be submitted to Parliament with sufficient time [e.g., 60 days] to reviewandenact it beforethe start ofthe fiscal year, b) Parliament approve any supplementary estimates before the spending occurs, c) emergency expenditures be made through a voted Contingency Fund only, TBD d) the amount of time allowed for submitting an excess (unauthorized) expenditurebill bereduced, and e) the financial statementsbe submittedto Parliament within 6 months of the endof the fiscal year. StrengthenParliamentaryOversightover spendingby: (a) improving the budget presentation (as described below under budget MFNP, Parliament preparation), (b) presentingto Parliamentthe macro framework andfunctionalpriorities of the budget earlier inthebudget cycle - Budget Preparation Reconcilethe largevariancesbetweenbudget estimatesand actualexpenditures & for eachbudget headand subhead, and assurethat the proposedbudgetreflects MFNPEiudgetOffice `o either thehigher cost or specificpolicy actiontaken to reducefuture services. 2% -Developamulti-yearplanningframework(or MTEF) bybuildingontherecent budget frameworkpaperto (a) link nationaldevelopmentpriorities (TNDPPRSP) to annualbudgets, (b) to reconcilethe differencesbetweenbudgetestimatesand MFNP, Cabinet, actualprogramcosts., and (c) to engageCabinet andParliament earlier inthe Parliament r7 .C ,X 2 budget calendar. `0 Reduce the amountof supplementaryappropriations, so that they are usedonly in uE exceptionalcircumstances [e.g., 5% ofthe original budget]. MFNP I I Enforcethe correct economic classificationof expendituresunderthe existing system MFNP Improvebudget presentationby including: (a) columnswith comparative expendituredataby headand subheadfor the two years prior to the budget year, (b) summaryinformationon the status of capital projects(e.g., originalbudget amount,expendituresto date, andbalance), (c) summarytables of spendingby MFNP, Parliament category, and (d) narrativeon major policy changesreflectedinthe budget. a, E Continueto publishinformation on cashreleases as ameans ofpromoting MFNP -141- Implement Quarterly Cash Flow Plans that (a) are based on comprehensive revenue forecasting and an assessment o f seasonal expenditure requirements and (b) that provide an assuranceto line ministries as to the level o f commitments that MFNP they can enter into. Establish commitment ceilings for ministries based on the quarterly cash flow .E8 "O3 5 plans and enforce compliance with monthly outtum reporting. Apply sanctions in cases o f persistentnon-compliance. MFNPlBudget Office U Complete andmaintainthe domestic arrears database, anddevelop aplanfor U liquidatingpast arrears. MFNP Public Procurement Establish a Procurement Reform ImplementationUnitinthe MFNP. MFNP c) $23 Establish a procurement cadre underZNTB. ZNTB I Finalize the plan for delegation o f procurement authority to ministries/agencies in MFNP, ZNTB a I years. Accounting, Reporting,andAuditing I # 1Procure hardware and software for the IFMIS, pilot test, evaluate. MFNP 32 Remove the OAG's budget and staffing policies from control o f the Executive andplace themunder the responsibility of an independent Audit Board (an .% wx 44 oversight body), which intum reports to Parliament. Parliament, State House HumanResources Complete the payroll data cleaning exercise and retrenchment packages, and use the savings to off-set the cost o f the 2003 salary increases. PSMD -142- Table 18: Second-Year Priority Actions RecommendedAction Institutional ResDonsibilitv Legal Frameworkand ParliamentaryOversight StrengthenParliamentaryOversightover spendingby assuringthat Parliament providesadequate funding for the BudgetEstimatesand Public Accounts Committees. MFNP, Parliament Budget Preparation Develop a multi-year planning framework (or MTEF) by building on the recent budget framework paper to (a) link national development priorities (TNDPPRSP) to annualbudgets, (b) to reconcilethe differences betweenbudget MFNP, Cabinet, estimates and actual program costs., and (c) to engage Cabinet and Parliament Parliament earlier inthe budget calendar. Expand Activity Based Budgeting (ABB) to all ministries and integrate the classificationsystem into the cashrelease processandinto the accounting system, so that it becomes the primary basis of budgetpreparation, budget execution, and reporting. Use this to identify and monitor poverty-reducingexpendituresin the MFNP budget. Budget Execution Strengthen the Commitment Control System to (a) maintain timely and reconciledinformationregardingcommitmentsand payments, and (b) to prevent any payments from being made unless commitments were properly recordedin the expenditureledger. Apply sanctionsincases of persistentnon-compliance. MFNP/AccG Update payrolls more effectively by having the senior pay point officer check and sign off on the payroll lists each month and return them to the provincial headquarters. These verified payroll lists should serve as the MFNP basis for updating the payroll master file. Develop unique codes for each post inthe establishment registry and use PSMD Enhancethe regulatoryand supervisoryfunctions of ZNTB. MFNP, ZNTB Revisethe leeal framework. MFNP, ZNTB Introduce an improvedregistrationlist system. Improve procurement practices, throughgreater restriction of negotiationandinforming tenders. ZNTB -143- Accounting, Reporting, and Auditing -144- Table 19: Third-Year Priority Actions RecommendedAction Institutional Resnonsibilitv Legal Framework and Parliamentary Oversight StrengthenParliamentaryOversightover spendingby: (a) Submittingregular(e.g., semi-annual) reports on the executionof the budget accompaniedby any requestsfor supplementaryestimates; (b) Developingarrangementsto ensure govemmentresponseto PAC reports MFNP, Parliament Budget Preparation Developa multi-year planningframework (or MTEF) by building on the recent budgetframework paperto (a) link nationaldevelopmentpriorities (TNDPRRSP) to annualbudgets, (b) to reconcile the differencesbetweenbudget MFNP, Cabinet, estimatesand actualprogramcosts., and(c) to engage Cabinet andParliament Parliament earlier inthe budget calendar. Improve comprehensiveness o f the budget by reporting on quasi-fiscal activities including revenues and expenditures o f BOZ, state-owned enterprises, pension finds, and other extra-budgetary finds. Also begin MFNP reporting on major contingent liabilities. InstituteareportingmechanismbetweenMFNP andMWSthat providesdetailed financialinformation on eachcapitalprojectincluding the financialterms ofthe contract, work variations, pricefluctuations,interest penalties, etc., inorder to MFNP, MWS inform cashreleasedecisionsby MFNP. Assure that recurrent costs o f capital projects are adequately planned for inthe MTEF, as apre-condition to the project being started. MFNP Public Procurement Improve business practice through development o f a code o f ethics, training, and development of competition and fair trade regulation MFNP, ZNTB Improvetrade practicesthroughincreasedcapacity of customsdepartment ZRA Accounting, Reporting, andAuditing Roll out IFMISto all ministries andtrain users. MFNP Carry out a comprehensiveassessmentofthe control systemandrecommend correctiveactions. MFNP Carry out a comprehensiveassessmentof the control systemandrecommend correctiveactions. MFNP Expandthe OAG's responsibilitiesto includeperformanceaudits. Parliament, State House Human Resources Continue decompressing salaries to enhance pay for professional and skilled technical personnel that the Government has most difficulty in attracting and PSMD -145- C. PROGRESS MONITORING 9.7 To ensure that the above measures will be properly implemented in a timely fashion, progress will need to be monitored closely according to a number o f indicators proposed below. The indicators must be chosen on the basis o f the following criteria: (i) selectivity since the fewer indicators, the better; (ii)they focus more on outcome rather than process, i.e., indicators showing that performance o f the budgetary system has improved, rather than actions have taken place; and (iii) they can be readily measured. An indicative list o fperformance indicators is givenbelow. Indicator Possible sources of Notes data Credibilityof the budget MoFNP analysis o f This could be considered at : (Yellow Book) and actual expenditure. variances 1.Aggregate level 2. Functionallevel 3. Between economic catenories Supplementary budget 1 estimates. bills submitted to Parliament. Comprehensivenessof the budget 4. Level of off-budget expenditures. Linkingpoliciesandbudgets 5.Preparation o f background papers for Cabinet discussions o f policy andprogram tradeoffs as part o f the annual budget preparation. Predictabilityof funds flow MoFNP analysis o f Two levels could be considered pro-rata allocation o fthe total budget if variances. 1.Flow o f funds from MoFNP to cashflow plans are not available) and the line ministries. actual cashreleases (intiming and value). 2. Flow o f funds from the center through the provincial and/or district levels to service delivery units. I Expenditurecontrol 7. Level o f arrears 8. By sampling, percent o f contract Sampling payments that are either: a) unpaid due to lack o f commitment; orb) have interest penalty (because o f delays inpayment) o f 50 percent or above o f the original amount. -146- Indicator Possible sources of Notes data Compliance 9. percent of invoices paid without funds MoFNP having been committed previously. Auditor General's report Reporting 10.Timeliness ofthe submission ofthe Auditor General's audited accounts to Parliament report D. COSTINGAND SUPPORT BY DEVELOPMENT PARTNERS 9.8 Some of the policy measures above are essentially about the changing the way Government works, and do not require much financial support. Others, however, require technical and financial assistance from all development partners. Detailed costing o f the financial and technical requirements for implementation o f these reforms should be carried out jointly by the Government and donors. Once the program i s fully costed, a pooled-funding arrangement could be developedby donors to support the program. 9.9 In the past, support by development partners of Zambia have not always been effective due to the lack o f coordination and information sharing among donors. Therefore, it is proposed that the PEMFAR Secretariat be formed to support PEMFAR The Steering Committee, discussed above, should hold regular briefings meetings with donors inorder to report progress and to resolve any issues arising from implementation. Formal assessment o fPEMFARprogress, including a formal report, can be carried out on an annual basis and should involve all donors. -147- REFERENCES Afrobarometer Network, The (2002) "Afrobarometer Round I:Compendium of Comparative Data from a Twelve-Nations Survey" Afrobarometer Paper. No. 11. MichiganState University, East Lansing, MI.Table 2-3, p.23 Castro-Leal, J., L.Demery, andK. Mehra. 1999. "Public Social Spending inAfrica: Do the Poor Benefit," World Bank Research Observer, Vol. 14, No.1. Das, Jishnu, Stefan Dercon, James Habyarimana, Pramila Krishnan. 2002. "Rules vs. Discretion: Public and Private Funding in Zambian Basic Education." World Bank. Dinh,Hinh,Abebe Adugna andBemardMyers. 2002. "The Impact of CashBudgets on Poverty Reduction inZambia," World Bank, AFTP1. Dinh, Hinh, Heinz Bachmann, Abebe Adugna and MushibaNyamazana. 2000. "Cash Budget in Zambia: Stabilization versus Growth and Poverty Reduction," World Bank,AFTP1. Government ofthe Republic o fZambia, Financial Statements, Various issues. Hallerberg, Mark, Rolf Strauch, and Jurgen von Hagen. 2001. "The Use and Effectiveness o f Budgetary Rules and Norms in EU Member States." Prepared for the DutchMinistry of Finance. http://info.worldbank.org/g;overnance/kkz/ htt~://www.worldbank.ordwbi/g;ovemance/govdata200.htm. 1 Hughes, Jesse. 2001. "Strengthening Government Accountability," mimeo. IMF.2001a. "The Implications ofHIV/AIDS for the ZambianEconomy." IMF. 2001b. "Zambia: Reformofthe Tax System." Fiscal Affairs Department IMF. 2001c. "Zambia: Strengthening Public Expenditure Management," Fiscal Affairs Department. IMF. 1997. "Zambia: Improving the Control of Public Expenditure," Fiscal Affairs Department. IMF. 2000. "Zambia: Improving Public Expenditure Management," Fiscal Affairs Department. -148- Krafchik, Warren and Joachim Wehner, "The Role of Parliament in the Budgetary Process." Budget Information Service, Institute for Democracy in South Africa (IDASA). Revised version o f paper first publishedin the South African Journal o f Economics, Vol. 66, No. 4 (December 1998), pp. 512-541. Ndulo Muna B. and Robert B. Kent .1998. "The Constitutions o f Zambia." Zambia Law Journal. Vol. 30: pp.26-27. Valentine, Theodore. 2001. "A Medium-Term Strategy for Enhancing Pay and Conditions o f Service in the Zambian Civil Service." Crown Consultants International. Wehner, Joachim. "Parliament andthe Power ofthe Purse: The NigerianConstitution in Comparative Perspective." Budget Information Service, Institute for Democracy in South Africa. Revised version of a paper first published in the Journal of African Law, Vol. 46, No.2 (2002), pp. 216-231. World Bank, GRICS 11,2000/01: GovernanceResearchIndicator Country Snapshot, T I World Bank. 1993. "Zambia: Financial and Legal Management Upgrading Project (FILMUP)," StaffAppraisal Report. World Bank. 1997. "Zambia: Country Profile of Financial Accountability." World Bank. 1998. "Fiscal Management Report," No. 18552, AFTP1. World Bank. 2000. "Zambia President's Report on Fiscal Sustainability Credit", Report NO.P-7379-ZA World Bank. 2001a. "World Development Report 2000/ 2001:Attacking Poverty." World Bank. 2001b. "Zambia: Public Expenditure Review," Report No. 22543-ZA. World Bank.2001c. "Zambia: Aide Memoire," mimeo. World Bank. 2002. "Zambia: Country ProcurementAssessment Review." AFTQK. World Bank. 2003. "Zambia: Country Financial Accountability Assessment." AFTQK, forthcoming. -149- ANNEX,VOLUME I Annex to Chapter3 Annex 1: ComparativeReviewofthe Role of ParliamentinBudgetPassage The tables below are drawn from research by Hallerberg, Strauch, and von Hagen and show how Parliament's power to change the Government's budget varies across West European countries. In a separate commentary, Hallerberg has noted that amidst the variation, there i s an observable pattern Legislatures are most powerless in countries that traditionally have few social divisions. These countries also have one-party majority govemments. Greece and the United Kingdom are prominent examples. Legislatures play more prominent roles in societies with many social divisions. Germany and the Netherlands are examples. Countries with both weak parliaments can have good fiscal p e r f ~ r m a n c e . ~ ~ Where ParliamentaryPowersDiffer: Legislature'sAbility to Changethe Government'sProposedBudget European Union Countries, 2001 Austria Belgium Denmark Luxembourg Netherlands Portugal Spain Sweden Finland Germany Italy France Greece Ireland 65Unpublished paper prepared for the World Bank South Asia Region PREM, March2003, "Executive- Legislative Relationsandthe Budget: PromotingEfficiencyandAccountability." -150- Data from Hallerberg, Mark, Rolf Strauch, and Jurgen von Hagen. 2001, "The Use and Effectiveness o f Budgetary Rules and Norms in EU Member States." Prepared for the Dutch Ministry o f Finance. A country receives a score o f "1" on the parliamentary index for each o f the following: amendments are allowed; they are not not limited; they are not offsetting; and parliament may propose its own budget. Where ParliamentaryPowersAre Similar: SimultaneousTreatmentof ExpendituresandRevenues, Time Limits, andActualChanges to the Government's Budget b.The Rightsof Membersof ParliamentinBudgetaryMatters Rights Number of Countries 1. May reduce and increase expenditure andrevenue. 32 2. May reduce butnot increase expenditure. 17 3. M a y reduce expenditure, but only increase it with the permission of 4 the government. 4. May reduce and increase expenditure if alternative provisions are 13 made elsewhere. 5. Rights not specified. 15 6. Not applicable (Nicaragua). 1 Total 82 Source: Krafchik, Warren and JoachimWehner, Institute for Democracy inSouth Africa (IDASA), "The Role o f Parliament inthe BudgetaryProcess." (1998). Originally drawn from the Inter Parliamentary -151- Union, Parliamentsof the World. A ComparativeReferenceCompendium(2ndedition), Aldershot (Gower), 1986,Table 38A. Annex to Chapter5 Annex 2: GovernmentBankingArrangements Analysis of the Problem As recently as three years ago, there were roughly 1200 Lusaka-based govemment bank accounts kept mostly in commercial banks. Inaddition, there were another 300 accounts inthe provinces. Although the practice hadbeen initially encouraged by BOZ so that it could avoid over-involvement in retail banking, eventually it became clear that it was having very negative repercussions on the govemment, and indirectly on the BOZ itself. Since 1996 the IMF has made recommendations to the govemment to improve performance inthis area. One o f the first tasks proposedwas to conduct a census o f such accounts in order to determine how many existed. Inaddition to reducing the number o f such accounts, the Fundhas also been concerned to ensure that all cash is routed through the BOZ andthat receipts are credited promptlyto the BOZ account. The proliferation o f commercial bank accounts had been problematic at least in part because o f the large holdings o f government monies that were being heldinthe banks for days or weeks without interest. Infact, not only were govemment resources not eaming interest, but the govemment was paying fees and commissions on each o f the active accounts. Worse yet, with resources scattered over a large number o f accounts in different banks, some accounts would accumulate idle cash balances while others would be starved for cash. In the latter case, this would necessitate unnecessary borrowing by the Treasury. During this time ministries would have separate accounts for PES,RDCs, capital, and revenue, plus multiple donor accounts and revolving funds. The cost to the government was also increased by the fact that govemment revenues transiting commercial banks in route to the BOZ were not being credited promptly to the govemment's account. Insteadcommercial banks were enhancing their liquidityfrom the extended floats andthe government's interest-free money. Substantial progress has been made in consolidating multiple ministry accounts into a single bank account per ministry, but the total number o f govemment bank accounts still remains high. As the IMF noted during its 2001, the multiplebank accounts "essentially serve as a substitute for timely accounting." O f the various accounts that line ministries will have, the most important by far are the "mirror accounts" (used to pay RDCs and capital expenditure) andthe salaries accounts. The operating procedures vary for each o f these two categories o f accounts. Inthe case o f PES, when MFNP authorizes a transfer o f funds from control 99 account (at BOZ) into the control account o f a given line ministry (also at BOZ), the finds are immediately transferred to the line ministries' PE account at a commercial bank. Salary checks issued by the line ministry on the commercial bank accounts are typically encashed withinten days o f the transfer of funds from the BOZ. In contrast for RDCs and capital, line ministries must prepare a "cash backing sheet" each day, which lists the checks drawn. On this basis, BOZ transfers funds from the line ministries' control account to the line ministries' mirror account at the commercial bank. -152- FutureAction Plan As noted above, the IMF has taken the lead role among donors inproviding advice to GRZ in reforming government's banking practices. The IMF recommendations fkom their June 2001 mission basically fell into five categories described below. Lowering the cost to government and improving the cash management capacity o f GRZ are certainly two primary objectives, but not the only ones. Government should also ensure that the banks are financially sound, that there are uniform standards, and that the arrangements for monitoring the accounts are inplace. Centrally negotiate bunk arrangements. MFNP,jointly with the BoZ, should solicit bids and negotiate banking arrangements with the commercial banks to agree on uniform terms for government business. Terms o f such agreements should include a uniform fee schedule, a uniform rate o f interest on the minimum average balance, and the responsibility o f the banks for the issue o f bank statements on time. Of course, some o f the terms and conditions could vary across categories o f accounts. Such reforms are needed because currently, each line ministry makes its own arrangements with commercial banks, and most o f these arrangements are informal and are not defined in the form o f written agreements or contracts. Consequently, government fails to obtain terms and conditions that are responsive to its needs and cost-efficient for the size and scope o f its transactions with the banking sector. The new Accountant General has expressed an interest in2002 inactively pursuingreforms inthis area. Create a Treasury single account (TSA) at BoZ. During the Fund's 2000 mission it recommended that the GRZ establish a TSA with the BoZ and use commercial banks on a zero-balance basis. Operationally what this means is that BoZ would notify commercial banks about funding limits for each mirror account but without remitting any funds. The commercial banks would in turn honor government checks up to, but not to exceed, the funding limit. At the end o f each day, the commercial banks would submit details on government transactions to BoZ for settlement. The main advantage o f this approach is that all government hnds are concentrated in BoZ accounts, thereby facilitating government's cash management and reducing domestic borrowing needs. The implementation o f a TSA could be facilitated if it is incorporated into the current design andplanningo f IFMIS. Review andjustzjj the utilization of all bank accounts. MFNP should review the need for each o f the government's commercial bank account and consolidate them wherever possible. Monitor bank accounts. The Accountant General Department should increase capacity to monitor commercial accounts on a monthly basis. This would include maintaining a database with account details: the account number and bank branch, the names o f those authorized to operate the accounts, the balances, floats, and lags inreconciliation. -153- Review the definition ofa government bank account. MFNP should work inconsultation with BoZ and the commercial banks to agree on the list o f government accounts at each commercial bank. In particular, MFNP should continue to de-link community-based accounts, which are sometimes mistaken by the commercial banks as being government bank accounts. As a result o f this confbsion there can be large variations between the consolidated position o f government bank balances reported to the ag (by commercial bank) andthe consolidated positionreportedinthe monetary survey. Inaddition to the above concerning government use of commercial bank accounts, the IMF has recommended that government continue its review and rationalization of the accounts at the BoZ, including the instruments o f central bank credit to the government such as the forex bridge loan. Annex 3: Areas for Improvingthe Cash Release System Good cash management or financial planning i s critical to any organization of any size. Nevertheless, the Accountant General's office has cited some obstacles to this occurring in the public sector, including unimplementable budgets for some activities of government and the lack o f material support (computers) for staff to work with. The Bank maintains, however, that these obstacles can be overcome and that cash management should become a regular part o f the MFNP's activities. Cash management processes should be robust enough to deal with changing circumstances, including amended budgets. On the logistics side, PSCAP has already begunto procure computers to assist MFNP accounting staff intheir various tasks. Moreover, the Bank can provide technical assistance to help the Accountant General's Office draw up procedures for doing some basic cash management procedures and training staff at an appropriate level. But MFNP should immediately assign responsibilities for developing cash management fbnctions and working with the requisitetechnical advisors. Part o f the process for developing a cash management plan is soliciting input regarding the procurement and/or workplans o f LMs. These should be core inputs to MFNP in determining its cash release priorities for a given period (eg, Agriculture would get more duringplantingseasons). Furthermore, MFNP's goal shouldbe comprehensiveness inits cash management, so that it i s taking into consideration 100 percent o f the forecasted resources for the period. The current system is based on planning for a given a percentage o f budget - which may or may not bear any relationship to the forecast revenues for the period. Under the new system, hnding for RDCs and grants may vary each quarter based on the revenue forecasted for that particular quarter, as well as consideration o f debt servicing payments or contractual obligations on capital projects. Inits June 2000 report the IMF outlined characteristics of an effective cash system; the qualities noted then are fully consistent with the Bank's current proposals. It should: 0 Beanenablinginstrument o fexpenditure planning for spending ministries; 0 Be forward looking - anticipating macroeconomic developments, yet capable o f accommodating significant economic changes and minimizing their adverse affects; -154- 0 Be responsive to the cash needs o f spendingministries; 0 Be comprehensive, including all inflows o f cash resources. Improving transparency around cash release decisions has been a vital part o f FSC and will continue to be stressed by the Bank. Indeed, providing clear information to line ministries and to the public on cash releases, serves a means to increase the accountability o f MFNP Budget Office to execute the budget as it was intended by Parliament. If there are amendments to the budget approved by Parliament, or changes withinthe legal discretion o fMFNP to make on its own, those canbe disclosed as well in order to explain why cash releases vary from the original budget. MFNP has taken substantial steps in this area by publishing the cash releases each month in the local newspapers. Renewed commitment by the authorities to implement the quarterly cash allocation plan is also required. Cash release decisions in Zambia are subject to considerable influence from other institutions, including some that maintain a higher political stature than MFNP. Consequently, the commitment to execute monthly cash releases according to a plan- and to not subject them to further ad-hoc decisions -would require some consent and concurrence from these other institutions. The Bank believes that achieving agreement to be bound by a quarterly plan would be aided considerably by including other institutions in the formulation and/or approval o f the plan. Input from ministries regardingtheir needs should be taken into consideration more systematically (as outlined above in cash management). Furthermore, the final arbitration of which institutions should be given the priority during a given period o f the year transcends the authority o f MFNP alone to decide. Although the Yellow Book should ultimately determine the cumulative amount that LMs receive over the year, the timing o f those releases can be reasonably debated at Cabinet level prior to the quarter. Indeed, formulating a credible cash release plan will require input and commitment from other institutions besides MFNP. MFNP and Cabinet Office can beginimmediately working together on drafting procedures and guidelines for how that decisionmakinganddebate would take place. -155- Annex Table: Comparison of the 2001 GRZBudget and Actual Funding -156- m m 0m 0 0 0 0 hl N 0hl e 8 e8 E l e8 g 8 0 N 0 Eu l V - e, e, a, xa 3 z e, L 8E e 8 L 8> k 0 L d 0 0 N E 2 e n z I . I ri 3 Annex to Chapter 6 Annex 6. Key IssuesinIFMISDevelopment ObjectivesFor A New System The Public Sector Committee o f the Intemational Federation o f Accountants (IFAC- PSC), the oversight body responsible for establishing Intemational Public Sector Accounting Standards (IPSAS) for Government Operations, has recommended that over time all governments move toward the accrual basis o f accounting. The movement toward the accrual basis o f accounting has also been supportedby the IMF in the recent publication o f their Government Finance Statistics (GFS) Manual. Although most governments throughout the world control and report their Government Operations on a cash basis, this can present a misleading picture o f financial health since the time o f recording expenditures and revenues may differ significantly from the economic activities andtransactions to which they relate. Furthermore, it i s generally more difficult to assess solvency and future cash flows with the cash basis because information on arrears is missing. Although the implementation of the accrual basis is more difficult than the cash basis, the outcomes o f the accrual basis are more reliable and informative. It will probably take a number o f years for governments such as Zambia's that are currently on the cash basis o f accounting to fully implement the accrual basis o f accounting for the whole-of-government, as recommended by the IMF in the GFS Manual and IFAC-PSC. However, this long-term process needs to begin now. Moreover, the Government should be preparedto incorporate financial statements from Economic Enterprises as part o f its whole-of-government reporting, as recommended both by the GFS Manual and IPSAS. In Zambia, where state enterprises have had a history of draining away public resources without substantial public scrutiny, this broader view o ffinancial reporting will be muchneeded. MovingFromAssessment To DesignPhase IFMIS development projects go through at least four stages as detailed inthe box below: (1) assessment, (2) design, (3) procurement, and (4) implementation. Prior missions by the Bank and Fundhave generally concentrated on the assessment stage only. Duringthe assessment phase it was recognized that an IFMIS would needto be developed to satisfy the needs o f management. The Bank mission recommended that the IFMIS be based on advanced web-based technology and that it be implemented progressively starting with the MFNP and then rolled out to other ministries, cost centers, and local governments. The Ministries would roll out the system to the cost centers within ministries and local governments in co-ordination with the MFNP. The final objective is on-line processing o f all transactions at the source where the transactions are generated. PhasesOfIFMISDevelopment 0 Assessment -Analysis o f the current system and identification o f areas for improvement; proposal of an alternative systemto meet those needs -163- a Design- Identification o f function and user requirements; development o f conceptual design; sizing o f components in order to prepare the procurement specifications. a Procurement - Development of procurement strategies, e.g., off-the-shelf software purchase versus in-house system development a Implementation - Pilot testing o f system before implementation throughout the operation; extensive training to ensure all aspects o f system are performed properly. CodingStructure In order to implement the conceptual design there are key issues to be resolved in the adoption o f a coding structure and chart o f accounts. Central to the development o f any accounting system i s the adequacy o f the coding structure. Although the international oversight bodies do not specify a coding structure, it i s implied by the GFS tables. In other words, the coding structure should be robust enough to provide the financial statements identified by the IPSASs and specified inthe GFS The current coding structure used by MFNP inthe accounting system was developed in- house in 1993 and provides for control by budget organization and economic classification only. However, as noted in an earlier chapter, the system will need to be revised to provide cost information by fund, function, and program. To that end in part, an Activity Based Budgeting (ABB) system i s in the process o f being implemented. In some areas, the ABB coding structure may be more extensive than that needed for the accounting system and inother areas it may be lacking. However, it is important that the budgetary and accounting systems are well integrated and that the needs o f each are resolved duringthe design phase. Inorder to provide a good coding structure for the Chart ofAccounts, it isrecommended that one adopt the codes in the GFS Manual. The existing chart o f accounts should be aligned with the GFS Manual inthe short-term inorder to fully implement a double-entry partial accrual accounting system in the mid-term. Further, a look-up table should be used to convert the existing classifications to the classifications in the GFS Manual in order to maintain historical trend data. Inthe mid-term, a look-up table could be used for a few years to convert the classifications inthe GFS Manual to those used at the present time for historical trend information. Inaddition, EstimatedRevenue and Appropriation accounts should be established in the mid-term in order to record the budget in the double-entry accounting system. Also, Encumbrances and Reserve for Encumbrances accounts should be established in order to implement an encumbrance accounting system. Since the GFS codes are intended for extemal reporting, the proposed Chart o f Accounts 66These include the Statement of Government Operations, the Balance Sheet, the Statement of Sources and Uses o f Cash, and the Statement of Other Economic Flows. -164- may need to be expanded to meet the managerial needs o f decision-makers within the government. The adoption o f the GFS codes using the fund structure recommended simplifies the process o f reporting Government Operations since data collected during the year will be accumulated in report format rather than converted to some other format at report preparation time. To lessen the impact o f transition to the international classification structure, it would be possible to use both (current and proposed) classification schemes during a one or two year transition period. This could be accomplished by running the systems in parallel or through the use o f a look-up table. If it i s desired to summarize Economic Enterprises by industry (Le., extractive, conversion, etc.), separate codes could be established to assist inthe compilation o f this data for the Economic Enterprises. NetworkArchitecture and Support To successfully implement IFMIS, the strategy for an information technology infrastructure will need to be developed. This would include servers, software, and storage. Servers are the backbone o f the infrastructure and insure that sudden spikes in traffic can be handled as more users performing more complex transactions increase the demands on existing networks. Ultimately, a single network may connect thousands of devices from scores o f manufacturers. Much o f the work i s done behind the scenes, in the network, by bigger computers runningpowerful and complicated software to provide collaborative solutions that securely bring people and information together. To most efficiently provide for this infrastructure, the architecture should provide for storage devices that will group storage systems on their own high-speed networks. Initially, MFNP should consider operating a local area network through a web based architecture with central servers and unified databases maintained at the MFNP. Later, the local governments, Ministries and other Cost Centers (who may only need the Purchasing, Human Resources, Budget Management and Financial Reporting modules that are relevant to their operations) can be interfaced with the financial system maintained at MFNP. As these modules are brought on-line, these modules will be rolled out via the local governments, Ministries and Cost Centers where they are operational in nature. By implementing a web based architecture, the IFMIS will be installed and managed entirely on the servers in a central location that will eliminate the need for desktop software installations and upgrades. This will significantly reduce the costs associated with managing traditional clientherver systems and will make the roll out o f this solution to local governments, other Ministries and Cost Centers faster andmore cost effective. To implement systems such as IFMIS, the focus o f the Computer Center needs to change from its current COBOLhatch systems orientation to real-time, on-line applications with intranet and internet networking (through a Wide Area Net). This reorientation dictates that a highly secure facility be obtained with access limited to authorized individuals. This facility should have a built-in recovery plan (in case o f power outages or other disasters) that will permit the system to degrade gracefully. This would include back-up generators andback-up storage devices. -165- Capacity buildingshould be guided by a policythat clearly spells out the role ofthe Computer Center inprocuring, supporting, and training user departments on the web based architecture necessary to maintainthe information technology. These changes will dictate changes inthe organizational structure and the staffing policies. -166- Annex to Chapter 7 Annex 7. Description of a Selected Salary Enhancement Scheme (SASE) The justification for SASE i s premised on the belief that (a) relying on wage-bill savings fiom employment reductions and government resources are not likely to yield much in the way o f pay reform, and (b) whatever results they do yield will be too gradual and too meager to bring about the changes in work behavior required to lead to significant improvements inwork performance and service delivery. Thus, extra-budgetary support would lend strength to the GRZ's pay-reform effort by accelerating the pace o f salary enhancement and jump-starting the overall public sector reform effort. Furthermore, given the reality o f limited resources, the enhancements would have to be targeted to posts critical to improving service delivery, managing the reform process ,and provision o f quality strategic outputs. Through the SASE scheme the basic salaries o f personnel inkey positions are enhanced by bringingthe medium-term salaries levels forward. For example, if the GRZ adopts such a scheme, beneficiaries would be paid their medium-term target salary levels for 2007/08 upon the scheme's implementation. Then, each fiscal year, as the government makes salary adjustments in line with MTPRS, the gap between actual govemment pay levels and the target pay levels would narrow, thus, reducing the need for extra financial commitment, as the government's ability to pay competitive compensation out o f the general wage bill rises. With SASE, the long-term sustainability o f the salary levels should not be at issue, since the target pay structure will be based upon an affordable wage bill over the MTEFperiod, As annual salary levels converge towards the target levels, extra budgetary support for salary supplementation will be phased-out. Currently, SASE schemes are being implemented intwo countries that have reached the HIPC completion point, Bolivia and Tanzania, and under considered for implementation in a third, Mozambique. In each o f these countries, the targeted recipients are equal to about 3 percent o f the total civil service workforce. Objective and transparent criteria for the selection o f SASE-scheme funded positions is crucial to the success o f the scheme. There are various approaches to selecting SASE- scheme positions. They could include for example, targeting all senior managerial, professional, and technical positions; targeting positions based on how critical the position is to service delivery, managing the reform effort, and/or contribution to the strategic outputs o f the government; or targeting all incumbents currently receiving special retention allowance^.^'. Whatever the basis for selection, however, the supplementation should not be automatic. To ensure that SASE supports public service performance improvement, the positions should have up-to-date job descriptions with 67Retention allowances are paid to certain categories o f employees based on their educational qualifications. -167- clear performance criteria, incumbents should sign performance agreements, and their performance appraised annually. Implementing SASE-scheme interventions is not without its challenges, as much learning-by-doing i s required. However, ifstructured and implemented properly, it may offer significant benefits to addressing problems o f low motivation, lack o f competitiveness of the salary structure, andcapacity buildingconstraints inthe public service. The cost and risk associated with maintainingthe status quo (low pay, lack o f incentives which undermines performance and efforts to improve service delivery, weaknesses inprovision o f strategic output) are too highand are inconsistentwith the requirements for growth-oriented andmodernising economy. The total cost o f financing the SASE can be easily estimated since it is merely a function ofthe number o fpeople covered and the average salary supplement needed to reach the target pay level. Assumingabout 2,700 positions were selected (slightly morethanthe number currently receiving retention allowances), and compensation Scenario 4 were implemented, the annual cost o f SASE would be a maximum o f 42.3 billion kwacha inthe first year (2002/03), declining to zero by year 2007108. The total financing requirement over the period would be 126.3 billion kwacha (or about USD31.6 million). -168-