Report No. 27504-TA Zanzibar Public Expenditure Review 2003 Laying the Foundations for Improved Public Expenditure Management June 2003 Revolutionary Government of Zanzibar Ministry of Finance and Economic Affairs Poverty Reduction and Economic Management II Country Department of Tanzania Africa Region Document of the World Bank 1. Currency Equivalents CurrencyUnit: Dominican Peso (DR$) ExchangeRate US$l.OO =DR$16.35 (as of March 22,2000) 2. Fiscal Year January 1-Dec 31 3. Acronyms and Abbreviations CB Central Bank CEA State Sugar Council CDE National Electricity Corporation FEyD Fundacion Economia y Desarollo FTZ Free Trade Zones GDP Gross Domestic Product INESPRE Price StabilizationInstitute LAC Latin America and the Caribbean ONAPLAN National Planning Office ONAPRE National Budget Office SPI Banking Sector Superintendency WTO World Trade Organization Vice President David de Ferranti Country Director Orsalia Kalantzopoulos Sector Director Guillermo Perry Task Manager John Panzer Preface.................................................................................................................................. Table of Contents EXECUTIVE SUMMARY ............................................................................................... ...i Key Recommendations.................................................................................................. 111 xv xvll .. PART 1:............................................................................................................................... Summary of Recommendations.................................................................................. COUNTRY BACKGROUND............................................................................................. 1 1. Real Economic Growth.................................................................................................... 23 The MacroeconomicContext....................................................................................... 1 2 4 FiscalPerformance........................................................................................................... 5 General PriceLevelDevelopments(Inflation)................................................................ Sectoral Performance....................................................................................................... The ExternalSector.......................................................................................................... Public Debt and Liabilities............................................................................................... 7 9 Implicationsofthe macro-economicenvironment for fiscal policy and public 2. expenditures Poverty Reduction...................................................................................................... ..................................................................................................................... 9 The Zanzibar PovertyReductionPlan (ZPRP).............................................................. 10 3. Progress in 11 Overview....................................................................................................................... -13 The Public Sector in Zanzibar.................................................................................... the Implementationofthe Zanzibar PovertyReductionPlan(ZPRP)........10 13 The CentralGovernment................................................................................................ FunctionsCarriedout by the UnionGovernment.......................................................... 14 16 District and RegionalDevelopmentCommittees .......................................................... LocalGovernments........................................................................................................ 18 Public Enterprisesand Utilities...................................................................................... 19 20 25 PART2: ............................................................................................................................. Other Government Institutions...................................................................................... PUBLIC RESOURCEAVAILABILITY INZANZIBAR................................................ 27 4. DomesticRevenue..................................................................................................... 28 27 InstitutionalSet-Upfor DomesticRevenueAdministrationand Collection.................28 DomesticRevenue Structure.......................................................................................... 29 5.RevenuePerformanceAssessment: ExternalDevelopmentAssistance ............................................................................. 41 Key Issues and Opportunityfor Improvement....35 ExternalResource Inflows to Zanzibar 6. Coordination,Monitoring and Managemento fExternalResourceInflows..................41 .......................................................................... Public DebtAnd Liabilities ....................................................................................... 42 46 Zanzibar's Debt Situation.............................................................................................. 46 PART 3:............................................................................................................................. Conclusionsand Recommendations.............................................................................. 49 PUBLIC EXPENDITURES............................................................................................... 51 51 7. Structure o fZanzibar's Public Expenditures................................................................. Zanzibar's Public Expenditures................................................................................. 52 52 Expenditureson Priority sectors.................................................................................... 8.The Issuesin BudgetManagement....................................................................................... 66 59 Government Pay and Employmentin Zanzibar........................................................ -73 Public Sector Employment ............................................................................................ 74 Wage Bill................................................................................................................ 73 Public Sector Salaries .................................................................................................... Pensions........................................................................................................................ 76 -80 Towards a Civil Service ReformProgram..................................................................... Wage Bill Management................................................................................................. 80 Employmentand Salary Reform.................................................................................... 83 85 Data Annex ........................................................................................................................ Salary Enhancements..................................................................................................... 89 97 ListofTables 2 Table 1.2: Volatility ingrowth......................................................................................................... 3 Table 1.1:Zanzibar GDPAnnual GrowthRates. Real 199.5-2000 (YO).......................................... 3 Table 1.4: RevenueandExpendituresin Zanzibar, 1998/99-2001/02 ............................................ Table 1.3: Sector Contributionsto GDP (Current Prices) 1995-2002 (Yo)..................................... 6 Table 1.5: Volatility of FiscalVariables andCorrelation with GDP 1981-2001............................. 7 Table 1.6:Trade Balanceand Re-Exports, 1995/98-2001102 (US$ million) .................................. 8 8 Table 1.8:Zanzibar's ExternalDebt (June, 2002) ........................................................................... Table 1.7: Zanzibar Import-ExportTrade 1997-2000(T.Shs. Mill.)............................................... 9 13 Table 3.2: Share ofPublicAdministrationand Services in GDP.................................................. Table 3.1:Administrative DivisionsandNumberof Shehias inZanzibar .................................... 13 Table 3.4: Subnational administrativestructure inZanzibar ......................................................... Table 3.3: List of Budgetary Votes................................................................................................ 16 18 2020 Table 3.6: Summary of ParastatalsinZanzibar, Positionas o f FYOl ........................................... Table 3.5. CompositionofRegionaland District Development Committees............................ 24 Table 3.7: Semi-AutonomousInstitutions:Subventionsand TotalExpenditures,FY02..............25 Table 4.2: Zanzibar RevenueStructure(YOof Domestic Revenue)............................................... Table 4.1:RevenueCollectionCosts ofZRB and TRA-Zanzibar................................................. 29 30 Table 4.3: Individual Tax Rates Applicable inZanzibar from February 2000.............................. 32 Table 4.4: Transfers from Government Enterprises,FY99-FY03 (T.Shs million) ...................... . 33 Table 4.5: Transfers by ZSTC to Government,FY93-FY02 ........................................................ 34 Table 4.6: Government Securities and Transfers from URT (Million TShs.) .............................. Table 4.7: Zanzibar Customs Duty Exemptionsby Category of Beneficiaries (% of Total).........34 38 Table 6.1: Zanzibar's ExternalDebt (June, 2002) ......................................................................... Table 5.1:ExtemalDevelopmentAssistance 1997-2001 (US$million)....................................... 41 47 Table 6.2: Domestic Debt, (June 2001) ......................................................................................... Table 6.3: Zanzibar's Domestic Debt (June, 2002) ....................................................................... 47 Table 7.1: Government Expendituresas %of GDP ...................................................................... 48 53 Table 7.3: Summary of Domestic Debt Servicing(TShs.Million) .............................................. Table 7.2: SelectedSubventions inRecentYears (current TShs. Million)................................... 54 Table 7.4: Government Expendituresby Purpose(% o f GDP) ..................................................... 54 55 Table 7.6: Expenditure Shares by Sectors (% of Total)................................................................. Table 7.5: UnspecifiedGovemment Expenditures(Million TShs.) ............................................. 57 59 Table 7.7: ParticipationandInternalEfficiency inBasic Education (YO)...................................... 60 Table 7.9: EducationResourcesby District................................................................................... Table 7.8: Shares inEducationExpenditures(%) ......................................................................... 62 63 Table 7.10: Expenditure shares within health(%of total healthexpenditures) ............................ 65 67 Table 7.12: Major Gainers andLosers duringBudget ExecutioninFY02.................................... Table 7.11: DifferencesBetweenBudgetedandActual Expenditures.......................................... 67 Table 8.1: Expenditureson Wages and Salaries as a Percentageof GDPandRecurrent Table 8.2: Government Employment (as Percentof Population) for SelectedRegions................75 Expenditures(FY02 or latest fiscal year available)....................................................................... 74 Table 8.3: Employmentby Salary Category.................................................................................. Table 8.4: Average Base Salary by Job Category, 1999-2002 ...................................................... 76 77 Table 8.5: Budgetary Expenditureson Base Salaries andAllowances, FY99-FY03 ...................77 Table 8.6: Allowances ................................................................................................................... Table 8.7: Base Salary andAllowancesfor Highest and LowestPaidSalary Categories, FY02 78 ..77 Table 8.8: Comparisonof Estimatesofthe Average Salaries for Zanzibar And Mainland Tanzania, FY02.............................................................................................................................. 79 Table 8.9: RequiredReductioninGovernment Employmentto Reducethe Share of Wages and Table 8.10: Scenariosfor the Coverage of the SASE Scheme ...................................................... Salaries inRecurrent Expenditures................................................................................................ 88 93 Table 8.11:Average Monthly Salaries (Incl. Allowances) And SASE Supplement, FY04-FY0894 Table 8.12: Additional Annual Cost Of SASEImplementation.................................................... Table A 1:Summaryof Central GovernmentOperations .............................................................. 95 98 2001/2002 ...................................................................................................................................... Table A2: Current Accountof CentralGovernment of Zanzibar-Administration . 1980/81- Table A3: Current Account of CentralGovernmentof Zanzibar-Administration .................... 99 101 Table A4: Current Account of Central Government-Economic Classification......................... 103 Table A5: Current Account ofCentral Government-Economic Classification........................ 105 Table A6: Classificationof Central Government Expendituresby Purpose................................ Table A7: Classificationof Central GovernmentExpendituresby Purpose................................ 107 109 111 Table A9: Actual Tax Collectionby TRA-Zanzibar andZanzibar RevenueBoard.................... Table A8: GovernmentExpenditures by Ministries and Years ................................................... 113 Table A10: Compositionof InlandRevenue(%of Total Collectionsby ZRB).......................... 114 Table A 11: Compositionof IncomeTax and CustomsRevenue 114 Table A 13: EducationMinistry budgetsummary (T.Shs. billion)............................................... Table A 12: HealthSectorBudget Summary................................................................................ (YOof Total Collectionsby TRA)................................................................................................. 115 Table A14: Zanzibar Exports, 1995/96-2001/02 ......................................................................... 115 116 Table A15: Zanzibar Imports 1995/96-2001/02 .......................................................................... 117 ListofFigures Figure 1.1: GDP and Agriculture Sector Performance. 1995-2002 ................................................ 4 Figure 3.1: Zanzibar State Trading Corporation. Ratioof Cost o f Sales to Sales. FY91-FY002222 Figure 4.1: Structure of Revenue(%of Total Revenue), FY92-FY02 ......................................... Figure 3.2: ZSTC. Average Distribution o fRevenue. FY91-FYOO .............................................. 23 30 Figure4.2: CentralGovernmentRevenueby Collection Authority (TShs .Billion) .................... 31 35 Figure 4.4a: Revenue (%of GDP), FY92-FY02 ........................................................................... Figure4.3: Revenue(% of GDP), FY92-FY02 ............................................................................ Figure 4.4b: Indirect Taxes by Source (% of GDP), FY92-FY02 ................................................ 36 36 Figure 4.4~:Other Revenueby Source (% of GDP), FY92-FY02 ............................................... 37 61 Figure 7.2: Trends inRealPublic Spending on Health (1991 prices)............................................ Figure 7.1:Trends inReal Public Spendingon Education............................................................ Figure 8.1: Share of the Wage Bill in RecurrentExpendituresand GDP, FY92-FY02 ...............64 Figure 8.2: Expenditureson Personnel Emoluments, Deviation betweenBudgetand Actual ......73 82 Figure 8.3: Projectionof the Share of Expenditureson Wages and Salaries in Recurrent Expendituresifthe Number of GovernmentEmployeesI s KeptConstant ................................... 86 Expenditureswith a HiringFreeze................................................................................................. Figure 8.4: Projectionof the Share of Expenditureson Wages and Salaries inRecurrent 87 Figure 8.5: Projectionsof Salary Increases inLinewith RealEconomic Growth......................... Figure 8.7: ProjectionsofAverage Salaries ina Scenario with Differentiated Salary Increases..91 Figure 8.6: Projections of Average RealMonthly Salaries by Job Category ................................ 90 91 List of Boxes 57 Box 8.2: InternationalExperiencewith Civil Service Censuses.................................................... Box 8.1 : Tanzania'sDevelopmentof an IntegratedHumanResource System............................. Box 7.1:Experiencewith Stabilization Funds............................................................................... 84 Box 8.3: The SASE Scheme in Tanzania ...................................................................................... 85 92 Zanzibar PublicExpenditureReview Pagei PREFACE In May 2002, the Revolutionary Government o f Zanzibar (RGoZ) launched the Zanzibar Poverty ReductionPlan (ZPRP), which put poverty reduction at the center of policy making in Zanzibar. ZPRP was prepared by RGoZ through wide consultations with all stakeholders, including all political parties, civil society organizations, donors and ordinary Zanzibaris. Subsequently, the ZPRP process has become the basis for renewed efforts to economic and social development and reengagement of donor nations and agencies to support Zanzibar's development and poverty reduction efforts. The reengagement o f donors came after almost six years o f suspension of development assistance to Zanzibar following a political stalemate that resulted from disputedgeneral ("Mwafaka") between the ruling party - Chama Cha Mapinduzi(CCM) and the leading election results in 1995 and 2000. The stalemate ended after the signingo f an agreement oppositionparty- the Civic UnitedFront(CUF) in2001. At the launchingofthe ZPRP it was agreed that the Government would undertake diagnostic work that would help guide development and poverty reduction efforts and that would provide stakeholders with basic information on the economic and financial situation o f Zanzibar. The core diagnostic work includes a study of the Zanzibar economy, a Country Financial Accountability Assessment, and a Public Expenditure Review(PER). This is beingsupplementedby more specific work, such as a study onthe clove industry,sectoralpublicexpenditure reviews for healthand education, and a private sector development study. This PER is the first of its kind for Zanzibar and has been prepared in close collaboration betweenthe RGoZ and the World Bank. The objective o f the PER is two- fold. First, it aims to diagnose the fiscal and public expenditureissueswhich will provide the basis for a public expenditure reform program. Second, the report will provide information to interested stakeholders on Zanzibar's fiscal situation and public expenditures, which can serve as an important input for decisionson future development assistanceto Zanzibar. This PER report consists ofthree parts. The first part providesthe backgroundon the Zanzibar economy, the ZPRP, and the public sector in Zanzibar. The second part reviews resource availability for the Government of Zanzibar, including domestic revenue, domestic and foreign borrowing, and developmentassistance. The last part o f the study reviews expenditures and wage bill issues in detail. Sectoral public expenditures in healthand education are reviewedin separate reports. The discussion of budget management and institutionalissues complements, rather than repeats, that in the CFAA. From the Government's side, the task was led by Julian Raphael, Principal Secretary, Ministry o f Finance and Economic Affairs. The World Bank team was led by RobertUtz (overallcoordinationand report preparationand chapters on the Public Sector and Wage and EmploymentIssues).The members o fthe World Bankteam includePhilip Mpango (Domestic Revenue), Praveen Kumar (Expenditure Analysis), Ben Tarimo (Macro-EconomicDevelopmentsand Debt), and HamisiMwinyimvua (Zanzibar Poverty Reduction Plan). Benno Ndulu offered valuable guidance to the team. Emmanuel Zanzibar PublicExpenditureReview Pageii Mungunasi and Wilhelm Ngasamiaku provided able research assistance. Philip Courtnadgeo fthe UNDPcontributeda chapter on development assistance. The counterpart team from Government that was led by Julian Raphael included Said M. Hussein (Commissioner for Public Investment, MOFEA), Yakuti H. Yakuti (Commissionerfor HumanResourceDevelopment, MOFEA), Ame B. Ame (Accountant General, MOFEA), Hussein S. Khatibu (Commissioner for ExternalFinance, MOFEA), Juma Juma (Deputy Commissioner for Budget, MOFEA), and Ali V. Ali (Director, Zanzibar Civil Service Department). The report has been subject to early and multiple consultationsand dissemination efforts in Zanzibar.A World Bank missionvisited Zanzibar in September 2002 to discuss the draft concept paper and to agree on the focus of the study with the RGoZ. In December 2002, at the closing o f the main PER mission, draft preliminary conclusions were discussed with the RGoZ. Subsequently, in February 2003, the first draft of the PER was discussed at a workshop held in Zanzibar. The team received extensive comments and suggestions from Government and other participants at the workshop, which were incorporatedintothe final report. In addition, in response to the deterioratingsituation in Zanzibar caused by the sharp drop in tourism following the September 11, 2001 terrorist attacks and subsequent tourist alerts, and the decline in clove prices, a subgroup of the Tanzania Development Assistance Committee was established to consider emergency support to Zanzibar, and the findings o f the report were used to brief the group on Zanzibar's fiscal situation and expenditureneeds. A roundtable was held in May 2003, duringwhich the findings o fthe variouspieces of diagnosticwork andemerging actionplans were discussed. The task was carriedout under the general guidance of Judy O'Connor, Country Director (AFC04) and FrederickKilby, Sector Manager (AFTP2).Raj Nallari (WBI PR), Allister Moon(ECSPE), and ChristosKostopoulos(AFTPI) are peer reviewers. Zanzibar PublicExpenditureReview Page iii EXECUTIVESUMMARY 1. Zanzibar finds itself currently in an extremely precarious fiscal situation, caused by the dramatic decline in government revenueduringthe past four years from 25 percent o f GDP in FY98 to 16 percent o f GDP in FY02. The decline is primarily the result o f the harmonizationo ftax policy and administrationbetweenTanzania Mainlandand Zanzibar and as such is likely to be permanent. Increasing revenue will require significant improvementsintax policy and administrationand is likely to be a slow process. 2. In the short to medium term, the key challenge for Zanzibar is thus to adjust budgets and the level and the structure o f public expenditures to the new reality of a considerably smaller resource envelope. Revenue estimates have not yet been adjusted downward. This has led to unrealistic budgets which could not be implemented. Decisions on public expenditures have thus de-facto moved from the budget process to a less transparent process of decisions by the Ministry o f Finance and more recently the budget ceiling committee, who decide on monthly cash releases to spending units in line with resource availability. Restoringthe credibility of the budget process through more realistic estimates o f revenue and expenditure needs thus to be a top priority in the short term, especially if the budget is to become the key instrumentfor the implementationof Zanzibar's povertyreductionplan. 3. To date, the burdeno f adjustment has fallen exclusively on developmentand non- wage expenditures, which have beencut in line with the fall in resources. This has ledto a severe structural imbalance in Zanzibar's public expenditures, with expenditures on wages and salaries claiming 62 percent of recurrent expenditures. This structural imbalance between wage and non-wage expenditures has direct implications for Government's capacity to deliver services, as staff often lack the means to carry out their functions. In addition to the economic need to make the level of wage bill expenditures consistent with available resources, there are also deeper seated problems in the civil service that will need to be addressed as part of a comprehensivepublic sector reform program. Government's capacityto ensure the integrity ofthe wage bill is weak and thus creates the potentialfor leakages. Furthermore,the salary structure is compressedwhich is one o f the causes for a shortage o f professional, technical, and managerial staff while there is overstaffingin the lower levelsofthe civil service. 4. In additionto fiscal and public expenditureissues, this public expenditurereview also assessedthe role o f the wider public sector inthe Zanzibar economy. Here the most urgent need for reform is with respect to the role of the Zanzibar State Trading Corporation,which currentlyenjoysa monopolyon the export of cloves, Zanzibar's main agriculturalproduct. Removingthe monopoly and liberalizing the clove sector is likely to have significant positive welfare implications for clove farmers and the Zanzibar economy. Zanzibar PublicExpenditureReview Page iv Context of PublicExpendituresManagementinZanzibar 5. The combinationof a weak economic base and high vulnerability o fthe economy to external shocks make sound fiscal policy and public expenditure management a formidable task inZanzibar. Creatinga supportiveenvironmentfor economic growthhas to be the key priority o f the RevolutionaryGovernment o f Zanzibar (RGoZ). Inaddition, measuresthat reduce vulnerabilityto externalshocks should also be considered. 6. Problems arising from the economic structure of Zanzibar are compounded by a fiscal managementsystem that i s characterizedby a high level o f discretionover boththe collection and allocation of funds by the Ministry o f Finance, while, at the same time, central controls for the appropriate use o f funds are weak. Our recommendations are broadly in the directiono fcreatinggreater predictabilityin the system for spendingunits, while at the same time putting systems in place that allow more effective central control o f key expenditureaggregates. 7. The suspension of most donor assistance due to political reasons since 1995 has had a severe negative impact on public service delivery in some o f the key sectors. The resumptionof aid should take into account lessons learned inMainlandTanzania andalso the significant changes in aid management that have taken place in Mainland Tanzania since 1995. 8. Although the review of fiscal policy and public expenditures over the past few years reveals significant shortcomings, it is encouraging that RGoZ is keen to address these problemswith important steps being initiatedat present. The main message o f this report is thus to encourage RGoZ to continue the course o f public expenditure reform. Our report is intendedto provide ideasand optionsfor such reforms. Overview of Public Sector 9. The public sector includescentral and local governments and the parastatal sector. Comparativeindicatorssuggest that Government is rather large, both in terms of resource use and employment,especiallywhen Union services are taken into account. 10. The demarcationof responsibilitiesfor government functions between the Union Government and the RGoZ is defined in the constitutionand various acts. However, the financial relationshipbetween the two Governments is much less clearly defined, and, in practice,Zanzibar has been reluctantto contributefinancially for services providedby the Unionto Zanzibar, includingservices such as police and defense, Zanzibari debt serviced by the Union Government, or services from parastatals such as TANESCO (the Tanzania Electric Supply Company). 11. The rationale of government involvement in carrying out certain functions and delivering public services is not always well established, including the activities of the parastatal sector. 12. The vertical structure of the central government includes four tiers-the central government, regional administrations, district administrations, and shehias. Local government starts at the district level and includes two tiers-district, municipal and town councils and wards. There appears to be a significant opportunity for rationalizing Zanzibar Public Expenditure Review Pagev this government structure by eliminating overlapping functions and clearly defining the responsibilities o f each level. 13. The parastatal sector has seen dramatic change over the past two decades with about eight operating parastatals remaining. However, the two largest parastatals in terms of their impact on the budget, PBZ (People's Bank of Zanzibar) and ZSTC (Zanzibar State Trading Corporation) are in need o f reform. Many o f PBZ's problems are due to weak management and financial relations with other parastatals and Government. Government i s currently considering options for privatizing PBZ, and it will be critical to ensure that PBZ is removed from the direct influence o f Government. ZSTC has a monopoly on the export o f cloves, but the rationale for this monopoly i s not evident, and liberalization o f the export of cloves is likely to have a significant positive impact on poverty reduction. A study on the clove sector that has recently been completed proposes a gradual liberalization of clove marketing over a three to five year horizon. While the proposed measures to support the clove industry should be considered, it might be advisable to allow for private participation in clove marketing as soon as possible. Public ResourceAvailability 14. Secular Decline in Revenue. Domestic revenue declined from 25 percent o f GDP in FY98 to only 16 percent of GDP in FY02. Tax harmonization with Mainland Tanzania and the closing o f the so-called "Zanzibar Route" contributed to a significant relocation of economic activity from Zanzibar to Mainland Tanzania. 15. Variability of Revenue. Zanzibar Government revenue has been variable over the past decade, fluctuating between 15 percent and 33 percent o f GDP (Chart 1). This has in turnmade Government expenditure planning highly unpredictable. I Revenue (% of GDP), FY92-FY02 I I 1 35% 1 30% 25% % I Of GD 20% I , ~ 15% 1 10% I I 1 5% 0% I I FY92FY93FY94FY95FY96FY97FY98FY99FYOOFYOlFY02 , Direct Taxes Indirect Taxes Other Earnings 16. The main reasons for the high variability of revenue include the fact that Zanzibar collects most o f its revenue from indirect taxes, which reflect instability in the performance of imports and production o f cloves. Possible options to reduce the variability o f revenue include (i)diversification o f the tax system to reduce dependence on taxation o f trade, and (ii)reform o f the indirect tax regime to have more predictable Zanzibar PublicExPenditure Review Panevi rates and a standard valuation of imports. However, measures to protect expenditures from the variability in revenue will be key to reducethe negative impact of highrevenue variability. 17. Exemptions. Zanzibar grants various fiscal incentives aimed at promoting investmentin tourism, and export-orientedactivities. Althoughthese exemptions haveto a limited extent helpedto attract foreign investment, they have been subject to abuse and have eroded the potential tax base. Thus, a thorough review and rationalization of the exemptions regime is imperative to counter abuse and improve revenue collection. Equally important,the RGoZ needsto review its investmentpromotion strategy to create a more favorable investment climate, particularly focusing on addressing infrastructural impediments-notably to improve roads and communication infrastructure, electricity, water, land lease, quality of labor, and curtailing red tape. There is also a need to enhance coordination among the various institutions (ZIPA, ZAFREZA, ZRB, TRA, and MoFEA) that are responsible for administering and approving different categories of fiscal incentives for investment. Recent efforts toward the development of a Zanzibar Investment Policy also include strategies to address problems related to infrastructure, landlease, quality of labor, andthe curtailingo fredtape. 18. Better Handle on Taxation of Tourism. The tourism sector has a strong potential of contributing to higher economic growth in Zanzibar. Tourist arrivals and earnings have shown an upward trend over the past decade. The special island features of Zanzibar, including climate, its rich culture, and long history, make Zanzibar one o f the most attractive tourist destinations in East Africa. To tap the revenue potential of the increased tourism activities, RGoZ needs to have a better handle on taxation o f this sector. Measures for adequate taxation of tourism that would spread the tax burden equally without creating negative incentives for the sector include (a) establishing an improved data base on businesses in the sector drawing on the recent establishment survey carried out by the government statistician; (b) clampingdown on tax evasion; and enhancing local capacity to offer quality services to tourists and increase taxable value added. 19. Growth-BasedExpansion of the Zanzibar Tax Base. Zanzibar's tax base remains very narrow, given its negligible manufacturing sector, and volatile mono-crop agricultureproduction. Inthese circumstances, expansion o fZanzibar's tax base needsto be growth-based. Specifically, there will be a need to embark on economic reforms to achieve sustained macroeconomic stability and spur investment and higher growth. Removal o f nuisance taxes and reduction of the multiplicity of taxes are important aspects ofthe strategy to reduce tax evasion andbringthe informal sector intothe tax net. 20. Sharing of Revenue between Zanzibar and Mainland Tanzania. The URT constitution provides for administration of taxes on incomes and internationaltrade, and the sharing of this revenue, Bank ofTanzania (BOT)seignorage, and concessional programloans and grants between Zanzibar and the Mainland. These arrangements are proposed to be overseen by a Joint FinanceCommissionappointed in June 2003. The existing revenue sharing arrangement involvesthe retentiono fall revenue collectedby TRA (the Tanzania Revenue Authority) at source. With regard to the distribution of program loans and grants, Zanzibar receives 4.5 percent of the total based on a temporary arrangement, and the basis of the 4.5 percent is frequently questioned in Zanzibar. Revisiting these income-sharing arrangements is likely to be high on the agenda of the Joint Finance Commission(JFC). Zanzibar Public ExpenditureReview Pagevii 21. Although taxes administered by TRA are supposed to be governed by the same legislation, there are still differences in both rates applied and valuation methods in Zanzibar compared to the Mainland. For example, the application of suspended duty in Zanzibar is limited to only three commodities, compared to fourteen on the Mainland. Similarly, pre-shipmentinspectiononly applies to imports into the Mainland. To resolve these issues, it would be importantto get the JFC to address some ofthese concerns, such as possible compensation for the revenue loss suffered by Zanzibar as a result of import tariff unification. 22. Resourcemobilizationis characterized by a largedegree o f discretion,bothby the RGoZ and by the Union, with regard to resources provided. This includes apparently arbitrary resource withdrawals from the parastatalsector, relatively largetax exemptions, and transfers from the Union, which mask fiscal deficits and seem to be related to the political cycle. 23. Tax administration is weak with poor capture of the revenue base, relying primarily on indirecttaxes. Expenditures 24. Total central government expenditures in Zanzibar increased steadily during the second half o f the 1990s and almost doubledfrom 17 percent of the GDP in FY95 to 33 percent o f the GDP in FYOO. Expenditureson wages and salaries increasedfrom 6 to 12 percent o f GDP during this period. Since then, total expenditures have returnedto more normal levels o f about 20 percent of GDP, but the share of wages and salaries has remainedstable. 25. Aggregate expenditures in Zanzibar mirror the high degree o f volatility in revenues and transfers from the parastatal sector, the Union Government, and external assistance. High volatility in expenditures results in disruption in public investmentand has associated costs o f adjustment. The long-term solution to the problem of resource volatility is a diversified tax base. In the short- to medium-term the key challenge, however, is to de-link government consumption from the volatility of the resource stream. 26. In FY02, recurrent expenditures were 97% o f the total, compared to 80% in MainlandTanzania, notwithstandingMainland's financingo f several state functions such as police, defense etc., which are Union matters under the constitution. A high level o f recurrent expenditures has led to an almost complete squeezing out of domestically funded development expenditures. This has had a serious impact on public investment, since donor finances were also absent duringthis period. 27. A largepart o f recurrentexpenditures is claimedby the salaries and allowances of an overstaffedcivil service. InFY02, salaries and allowances claimedmorethan 60% of recurrentexpenditures, compared to 35% inthe Mainland. The largesize o fthe wage bill makes expenditures management difficult on two counts. First, given limited resources, it forces cuts in operations and maintenance expenses, thereby adversely affecting the productivityof existing public assets and service delivery. Second, it makes the overall demand for public funds very inflexible. This is a problem when revenues display high variability, since it is not possibleto adjust expenditures on the wage bill quickly. Zanzibar Public ExDenditure Review Pageviii 28. Expenditureson subsidies and transfers were significantlyreducedduringthe first halfo fthe 1990s andhave since remainedat a fairly low and stable levelof around 1%of GDP. Main receiversof subventions are local government bodies, the Zanzibar Revenue Board, the Zanzibar Free Port Authority, the Zanzibar Investment Promotion Agency, and other non-profitbodies. While the level of cash subsidies is low, there are sizable in- kind subsidies, mainly in the form o f subsidizedwater and power supply. 29. The Government runs a cash budget and has managedto keep the fiscal deficit in check. However there are several fiscal risks that the Government must contend with. The most important is accumulated arrears and accrued liabilities. The Government does not collect informationon arrears systematically,nor does it plan for them duringbudget formulation. Expenditures on Priority Sectors 30. In terms of expenditures by functions, government spending has undergone a major transformationin the last decade. The expenditure share on general administration has increased, while that on health, agriculture, transport and communication has decreased. The challenge beforethe currentGovernment is to restorethe expenditures on social and economic sectors while cuttingdown the size of general administration. Expenditure Shares by Sectors (YOof Total) FiscalYear FY FY FY FY FY FY FY FY FY FY 92 93 94 95 96 97 98 99 00 01 Education I O 11 15 12 13 13 11 13 9 14 Health 7 8 9 8 9 8 7 8 4 7 Agriculture 9 7 8 7 7 5 4 5 3 5 Transport and Communication I O 7 6 6 7 4 3 3 3 4 General Public Service 41 37 47 55 52 62 68 53 65 59 Other 24 30 14 12 13 9 7 19 16 11 Total 100 100 100 100 100 100 100 100 100 100 3 1, The Zanzibar Poverty Reduction Plan (ZPRP) has identified education, health, road infrastructure, agriculture, water, good governance, and combating HIV/AIDS as priority areas for governmentfinancing. The expenditures on priority sectors have been stagnating at around six percent o f GDP since FY97. Allocations for FY02 and FY03 show that expenditures on healthand education are projectedto increase. 32. Educationsector expenditures are about 2.7% o f GDP and about 15.5% o f total government expenditures (versus the ZPRP target of 20%), a level slightly higher than MainlandTanzania and comparable with other countries inthe region. Per capita annual public spending on education has been improving since FY95, and was around a reasonable US$lO inFY02, comparedto about US$7 for MainlandTanzania. 33. Almost all of the education expenditures are recurrent in nature, and within recurrent, salaries and allowances accounted for 93% in FY02. Development expenditures have stayed at less than one percent since 1995. Expenditures on "other charges" have been very limited too. School materials and teaching equipmentare now expectedto be providedby the community. Zanzibar PublicExpenditureReview Page ix 34. Government financing of the education sector remains inadequate in view of the ZPRP objectives. Infrastructureis highly inadequate and performanceindicatorsare still low. Yet, it is possibleto achieve better results within the existing envelope throughthe restructuring o f expenditures. One way to achieve better results would be to curtail expenses on wages and salaries and make more resources available for non-wage expenditures. There is a strong case on equity and efficiency grounds for reviewingand reorienting expenditure patterns on a geographical basis. There is a considerable opportunity for expanding access to basic education by redistributingexisting teachers more efficiently, without having to hire more teachers. Finally, more could be done to encourage privateprovisiono fprimary andbasic education. 35. Aggregate public spending on the health sector declined between FY95 and FYO1, both as a share of GDP (from 1.8 to 1.3 percent) and of total government expenditures (from 8.8 to 6.8 percent). FY02 saw a significant increase in the health budget, and spending improvedto about 1.9 percent of the GDP and 10 percent of total expenditures. The FY03 budget has plans for further increases in the health budget. However, it should be noted that, despite real growth in recent years, health sector spending is below the levels achieved in the early 1990s. The sector shows tell-tale signs o f chronic under-funding. Healthfacilities are in a state of disrepair, equipment i s scarce, and there is an acute shortage ofdrugs and medicalsupplies. 36. As in education, while there is a needfor an increasedoverall levelof resourcesin health, there is an opportunity for reducing administrative expenses and reorienting expenditures to give more attention to preventivecare and to provide a greater share for drugs and medicalsupplies. Regionalimbalances inspending also needattention. Budget Management 37. Judged against the desired outcomes of aggregate fiscal discipline, efficient allocation of resources, and ensuring operational efficiency and effectiveness, the performance of the budgetary systems leaves much to be desired in Zanzibar. While there is a semblance o f aggregate fiscal discipline, the predictability of resources for Ministries, Departments, and Agencies (MDAs) from year to year and within given years has been very low, which has severely impacted MDAs' operational efficiency. In the past three years, deviations o f actual outturns from the budgeted amounts have been of the order of 80% for development expenditures and between 40 and 70 percent for other charges. 38. The Governmentplans to move to a multi-year framework for expenditures-the Medium-TermExpenditureFramework(MTEF). This will be a step in the right direction and should be a medium-termgoal. It will help the Ministry of Financeand Economic Affairs (MoFEA) to align the sectoral structure of expenditureswith ZPRP priorities in the medium-term. As a first step, MoFEA should attempt to forecast aggregate expenditures for three years inthe year 2003-04 basedon some macroeconomicmodeling and then request ministriesto submit estimates for these years. The Government is likely to face capacity constraints in developing a medium-term approach to expenditure planningand shouldpreferablystart it intwo to three sectors to beginwith. 39. The unpredictability of resource availability and weaknesses in budget management result in a low level o f transparency with regard to resource allocation and the use of public resources. Examples of limitedtransparency include: Zanzibar PublicExpenditureReview Pagex P The expenditures from the vote `unspecifiedgovernment expenditures' have been fairly large andthere are no clearly laidout criteriadefining eligiblecontingencies for withdrawal. P In the past, large sums of money have been spent on items such as `Special Presidential Vote'. An expenditure o f T.Shs. 7.764 billion in FY99, and 5.168 billion in FYOl was made on this account. These resources reportedly cover emergencies such as epidemics, unforeseen expenditures, and operations o f the offices of the central leadership o f the country. The Government has reportedly > droppedthis sub-vote from FY02onwards. The MoFEA issues lump sums to the ministriesagainst wages and salaries. There havebeen instanceswhen ministrieshave divertedthem for otherpurposes. P The MoFEA operates a Community Development Fund for providing a government share towards community projects. Reporting on the use o f these resourcesand adherenceto establishedcriteriacouldbe strengthened. 40. Apart from the above, lack of information (for example on arrears and accrued liabilities) makes the budget processes look opaque. In order to generate confidence in the integrity o f budget processes, the Government needs to give top attention to transparency issues. PublicDebt 41. Current Status: One o f the key problems in Zanzibar's debt management is the absence o f comprehensive and consistent information on public debt. In addition to introducing uncertainty into the formulation of debt strategies and policies, this also creates problems in budgetingfor debt service payments and can lead to unforeseen debt service requirements in the course of the fiscal year. Available, but reportedly incomplete, debt statistics indicate that by the end o f June 2002, Zanzibar's total debt (external and domestic) was T.Shs. 85.7 billion (US$90.7 million) of which external debt was US$58.4 million, accounting for 64.5 percent, while the domestic component was 35.5 percent of the total debt. Of the external debt portfolio, multilateral creditors account for US$42.7 million, or 73.0 percent of the total external debt. Bilateralcreditor claims account for US$8.5 million, or 15.0 percent, while debt owed to commercial sources (suppliers' credit) amounted to US$7.1million, or 12 percent of the total external debt. All the multilateraldebt is guaranteedby the UnitedRepublicofTanzania (URT). 42. The Union Government and the Bank o f Tanzania externalize Zanzibar's debt. The Bank o f Tanzania typically deducts any debt service payments made on behalf of Zanzibar from the dividenddue to Zanzibar. 43. Out of the total domestic debt of T.Shs. 30.4 billion, T.Shs. 13.6 billion (45.0 percent) is owed to the UnionGovernment. The People's Bank ofZanzibar (PBZ) claims amounted to T.Shs. 7.2 billion (24.0 percent); Zanzibar State Trading Corporation (ZSTC) and Zanzibar Port Authority are owed T.Shs. 2.9 billion (9.0 percent); and T.Shs. 0.6 billion (2.0 percent) respectively. Other creditors claim a total o f T.Shs. 6.1 billion accountingfor 20.0 percent ofthe total domestic debt. 44. Significant arrears exist, and systems need to be put in place to prevent the recurrence of public arrears. A central payment system and a Local Purchase Order (LPO) system seem crucialto ensure better expenditureand commitment control. Zanzibar PublicExpenditureReview Pagexi 45. Debt Coordination and Management: At the moment there are various units which appear to be involved in debt management with a significant opportunity for improved coordination. The Accountant General's Office faces severe capacity constraints to perform its role in debt management effectively. The External Finance Department of MoFEA plays a role in the negotiation and contracting o f debt. The department maintains a database, but it is considered quite incomplete, since project financinghas beenvery difficult to capture. Inaddition,this department provides datato the Budget Section, but acknowledges that it is also incomplete. Grants received from Union projects, for example, are often excluded from the estimates. There is a need to clarify the roles of the departments so as to have a clear demarcation of their responsibilities. 46. There are, at the moment, very weak links between the line ministries and MoFEA on reporting receipts and expenditures o f project assistance, implying that the DevelopmentBudget estimates are incomplete. This by-passingof the central budgeting system also means that the External Finance Office and, more importantly, the Accountant General's Office are unable to ensure compliance on reporting and accounting for use o f funds. Furthermore, in such a situation, it becomes difficult for MoFEA to budget properly for the counterpart funding component. On the other hand, the Bank of Tanzania (Zanzibar Branch)maintainsa separate data set of Zanzibar's debt position. 47. MoFEA acknowledges that the situation needs to be improved. In view o f this situation, it is advisable to create an effective Debt Management Unit in MoFEA to coordinate all debt matters. The BOT (Zanzibar) has offered support in the area of debt management-especially on the Commonwealth States Debt Reporting System (CS- DRS). 48. The Contracting of Debt: A centralizedsystem giving legal authorityto only one ministry-namely the Ministry of Finance-for contracting debt on behalf of any Government is considered to be good practice. If such a discipline does not exist, then any official inthe Governmentwill contract debt, leadingto a state of confusionwhereby (1) the needed statistics will not be compiled/reconciledand (2) the debts may not be in line with the priorities o f the country. The situation may leadto leakages and misuse of the resources. By law, only MoFEA has power to contract debt on behalf o f the Government. However, the ineffectivenessand poor enforcementof such a requirement has created problems. There are still a number of ministries which borrow-although mainly domestically through the window of suppliers' credit while waiting for monthly releases from MoFEA. Some parastatals have also been doing the same, which creates contingentliabilities for the Government. In view o f this situation, it is recommendedto review the current legal status to provide for enforceable sanctions and, at the same time, strengthen the enforcementof legalrules. 49. The Needfor a Debt Strategy: At present there is no clear debt strategy for the islands. A debt strategy will guide the Government on the limits to borrowing, and specify the types of loans and their maturities, including an assessment of repayment capacity and the repayment schedule. Sucha strategy will also ensure that debt service is in line with the operations of the budget. In the absence of such a plan, borrowing and repayment will be quite erratic, and, in most cases, will be outside of the budget projections. Authorities have also already seen the need for a debt strategy. In view of this gap, the preparationof a Zanzibar Debt Strategy (ZDS) is recommended. Zanzibar PublicExpenditureReview Pagexii 50. The Need for a Debt Task Force. As stated earlier, there are various sources of the numbers on debt position and, unfortunately, they all differ. Without a clear understandingof the debt situation-both external and domestic-it is difficult for the Government to know its obligations so that it can, inter alia, prepare a realistic debt repayment schedule. The MoFEA has already started thinking of carrying out a debt census. It is thus recommended that a Zanzibar Debt Task Force be constitutedto bring the debt situation into a reliable and dependable position. The Bank of Tanzania (ZanzibarBranch)couldprovideassistance in this regard. Public Wage Bill and Civil Service Issues 5 1, Internationalcomparisons indicatethat, with 62 percent of recurrentexpenditures, the wage bill consumes a disproportionateamount o f government resources, mainly due to high staffing levels in Government. Overstaffing appears particularlyprevalent in the lower cadres of Government, while there is a severe shortage o f qualified manpower at the professional levels. Other potential causes for a bloated wage bill include the presence of ghost workers on the payroll and over-age employment due to the non- payment o f gratuitiesandpensions. 52. Simulations indicatethat even under fairly optimistic assumptions with regard to economic growth performance, freezing staffing at the current levels while government revenue increasesin line with economic growthwould only lead to a very gradual decline in the share of recurrent expenditures absorbed by the wage bill. Putting a complete hiring freeze in place would reducethe share ofthe wage bill below one-half of recurrent expenditures after a five year period, again assuming sustained economic growtho f five percent per year. However, a complete hiring freeze would be politically difficult to sustain and would prevent the civil service from acquiring and replacing critical skills necessary for the functioning o f Government. A more selectivehiring freeze that affects only the lower cadres would only have a marginal impact on the division of the budget between wage and non-wage expenditures. With regard to retrenchment options, our simulations indicate that retrenchments of about 25 percent across the board, excluding teachers, would be required to reduce the share of the wage bill in government expenditures to about 50 percent. Retrenchment options include (1) the removalof ghost workers, over-age workers, and workers with irregularitiesintheir employment;(2) early retirement; and (3) voluntary and involuntary retrenchments. Appropriate safety net measures, including appropriate retrenchment packages, payment of gratuities and pensions, training, or employment opportunities through subcontracting will need to be put in place to ensure that civil service restructuring becomes socially and politically feasible and does not leadto an increase in poverty. 53. Pay levels (including allowances) in Zanzibar are similar to pay levels in Mainland Tanzania. However, the salary structure appears to be more compressed, which provides poor incentives for attracting talent and extracting performance. At present, there are only limited formal employment opportunities in Zanzibar outside Government. However, to the extent that such opportunities expand, Government will have to make efforts to remaincompetitive in the labor market. Our simulations indicate that, once public sector staffing has been reduced, differentiated salary increases with positive real increases for the lower cadres which are below the rate of real economic growthwould provide an opportunity for faster salary increases for the highercadres. Zanzibar Public ExpenditureReview Pagexiii 54. We also estimate the cost o f introducing Selective Accelerated Salary Enhancements (SASEs) for senior civil servants. However, it seems that the cost, the administrative requirements, and the inherent risks o f a system o f advanced salary increasesmakethe adoptionof an SASE system not desirable at present. 55. Allowances add about 80 percent to the average salary of public sector employees. The system o f allowances in place reduces the transparency of the government salary structure, and, particularly for higher cadres, there appears to be a large degree of discretion for making decisions on the amount of allowances,which are not necessarily relatedto merit and performance. It is thus recommendedto consolidate basic salaries, all allowances, and other non-wage benefits (such as duty exemptions on the import of vehicles) into a consolidated and simplified salary structure. Since allowances are currently neither pensionablenor taxable, it will be important to carefully examinethe fiscal impact o fthe proposed consolidation. 56. The payment of the wage bill is delegated to the MDAs, which creates the potential for leakagesof expenditures destined for wages and salaries inthe form of ghost workers or redirectionofPersonnelEmoluments(PE) funds to other uses. 57. There are internal measures in place to ensure the integrity of the wage bill through reviews undertaken by the Civil Service Department(CSD), sectoral ministries, and the MoFEA. Nonetheless, an externalaudit of the wage bill and the establishment of a centralized control of salary payments are recommended to provide increased confidencein Government'swage bill management. 58. The development of a comprehensive civil service reform program should be a priority for Government. Elements o f such a reform program should include the reductionin overstaffing,a rationalizationofthe system of allowances, decompressionof the wage structure, a medium-term program of pay reform, and capacity-building measures. Realigning Public Expenditures with the Zanzibar Poverty Reduction Plan (ZPRP) 59. The ZPRP defines key needs that should be given priority in resource allocation and budget implementation. 60. The ZPRP is not sufficiently specific to guide expenditure allocations and will require further work at the sector level. 61. For some key sectors, sector strategies and programs are beingdeveloped. It will be importantto ensure that the refinementof the ZPRP and the sector programsare well coordinatedin an interactive process, with the ZPRP process having authority in setting priorities and providingthe overallframework. 62. Many Departments are revisiting their sectoral development plans in view of the ZPRP. However,links to the ZPRP in these plans are as yet weak and unclear; ZPRP and sectoral plansneedto converge quickly. There i s a needto ensure completebuy-inby all the sectors, most importantlyby educationandhealth. Zanzibar Public ExpenditureReview Pagexiv RecentReform Initiatives 63. Government is aware o f many of the problems and weaknesses in public expendituremanagement and strategic resource allocation outlined in this report. Initial steps to address some o f these problems are already underway and Government is encouragedto pursue and deepen these reformefforts: P Adoption of the MTEF principles in the government budgeting system, which started by raising awareness among politicians, and training directorsof planning and administrationand chiefaccountants from key sector ministries; P Establishment of a ceilings committee for the monthly review and allocation of resources; P Preparation of a Central Tender Board Act to establish a transparent system of procurementand wise use o f resources-draft regulationshave been prepared in view o f operationalizingthe CentralTender Board; P Preparation and operationalization of the financial administration and public enterpriseact by puttingin place relevant regulations; 9 Preparation of the new investment policy, which also looks at the institutional setup for investment promotion in Zanzibar, which is to be followed by the formulation o f implementationregulations; 9 Carryingout of a series of studies to preparethe basis for further reforms. These include: a) A study on problems relatedto Zanzibar port operatiodtrade, ferryboat and dhow services between Zanzibar, Dar es Salaam, Pemba and Tanga in collaborationwith the TanzaniaFreightBureau; b) A study on the assessmentof Zanzibar'srevenuepotential; c) A study on clove productionand marketing; d) A study on the procurement system; e) A country financial accountabilityassessment; and Q A Zanzibar EconomicAnalysis. 9Z 92 92 e I 3 2 z 3 3: 3: c 9 3: 9z B r : E E Y a, $.s 3 s 3 3 2 z 3: 85 .-.-0 0 P z 3 e I r f z 2 x C O L n 3 r: ? e3 i clc! 3 2r z L 2r 3: 2 e 3 z 2e 3 r 2 3: r 9 z 2E 2E E 3 1? u u !i!i N m d 2 z 92 c2 1 % 1 % 3: 9E II 3 3 L .-.-t z 2. c 0 13 3 2 3 i- t U 0M D3 2 Y z 9 . 00 m 2s 3 z s s3 s3 9s z .-.-tb - h z 9 z 4 4 3 b?Z 9Z 9 3 3 P 2 2 .-.-k i h e I 0 3 5i 2 I 5 e 0 & 0 Y P 3 s 3 2 * L 3 2 9 z cb Y 3 z r : '6* h .-k 3 e z z 3 2 z 2 e 4 * b) % G v3 r r - N -\1 N r? N .-.- 2. 0 L 3 E m Zanzibar Public ExpenditureReview Page 1 PART 1: COUNTRYBACKGROUND Zanzibar PublicExpenditureReview Page2 1. THE MACROECONOMIC CONTEXT REALECONOMIC GROWTH 1.1 During the recent past (1995-2002), the Zanzibar national output (GDP) increased from T.Shs. 80,737 million in 1995 to T.Shs. 176,000 million in 2002 (current prices). Real GDP growth fluctuated significantly in recent years, reaching 7.1 percent in 1996 and falling to only 0.5 percent in 1998. In the last two years (2001-2002), the GDP growth rate only averaged around 4.3 percent. The growth path is explainedmainly by clove production(and, to a lesser extent, price instability) which demonstrates the significant influencethat this sub-sector continuesto have on the economy. The best recent year for growth, for example, was 1999, a year when the agricultural sector grew at 7 percent following the boom in the clove sub-sector that year. Table 1.1: Zanzibar GDP Annual Growth Rates, Real 1995-2000 (YO) Source: Department o f Statistics, Zanzibar * Estimates 1.2 Real GDP growth has been weak due to the low growth of agriculture and other productivesectors, a fall in general productivity, low investment,and the impact of declining levels of external assistance to Zanzibar. Low levels o f public investment, partly due to declining levels o f external assistance, impacted on Government's ability to deliver essential social and infrastructureservices, which are the foundationof pro-poor growth. 1.3 Growth rates were highly volatile during 1981-2001, 50% more volatile than the average for Sub Saharan Africa (SSA), 33% higher than the average for the Middle East and North Africa (MNA), and 350% more volatile than industrialized countries. Growth volatility, in turn, reflectedthe underlying volatility in agriculture, Zanzibar PublicExDenditureReview Page 3 manufacturingand constructionsectors. Table 1.2 below shows measures o f absolute and relativevolatilitiesby differentsectors. Table 1.2 Volatility in growth Industry Absolute volatility' Ratio of Sectoral to Total GDP volatility Agriculture 14.9 1.6 Mining & Quarrying 23.6 2.5 Manufacturing & Handicrafts 46.9 5.0 Electricity & Water Supply 9.6 1.o Construction 43.2 4.6 Trade 17.8 1.9 Transport & Communication 21.5 2.3 Finance 21.0 2.2 Public Administration & 9.6 1.o Services Total G D P 9.4 1.o Source: Calculationsbased on datafrom the Office o f the Chief GovernmentStatistician,Zanzibar. "Volatility is measuredby the standard deviation o f real growth rates over 1981-2001. SECTORAL PERFORMANCE 1.4 The three major sectors that drive the Zanzibar economy are agriculture (and its related sub-sectors), trade (including tourism) and services (which includes administration). Table 1.3 below shows that these sectors constituted almost 80 percent ofthe GDP in 2002. Table 1.3 Sector Contributions to G D P (Current Prices) 1995-2002 (YO) Source: Departmentof Statistics,Zanzibar; BOT Bulletin * Estimates. 1.5 Agriculture, the main sector in the economy, accounted for 40.9 percent of the GDP in 1995 and declined slightly to 35 percent in 2002. The performance o f the sector has generally beenpoor due to a number of factors, including(1) an agriculture policy which needs review and lacks both an implementation strategy and a well defined institutionalframework, (2) low incentivesto farmers, (3) government control on the marketing of cloves, and (4) a low level o f investment in agriculture-related Zanzibar Public ExDenditure Review Page4 infrastructure. Government has developed an agriculture sector policy which, however, lackeda strategy for its implementation. Figure 1.1: GDP and Agriculture Sector Performance, 1995-2002 1 5 I O 5 0 - 5 -10 Y e a r +GDP Growth --r-AAric Sector G r o w t h I 1.6 The trade sector and its related activities, including tourism, have expanded steadily, marginally increasingtheir share in overall GDP from 19.8 percent in 1995 to 20.0 percent in 2002. The increasingtrend is the result o f robust trade activities within the period. Other fast-growing sectors include construction, transport, and utilities (energy/water)sectors, althoughthey comprise a relatively insignificantshare of total output. GENERAL PRICE LEVEL DEVELOPMENTS (INFLATION) 1.7 Monetary policy is a Union matter implemented by the Bank of Tanzania. Between 1995 and 2002, the inflation rate in Zanzibar declinedfrom 29.8 percent in 1995 to 6.0 percent by the end o f 2002, in parallelwith the decline inthe inflation rate in Mainland Tanzania, as a result of prudent monetary and fiscal policy pursued by the Government of Tanzania. Importedfoodstuffs and a relatively stable exchange rate have also served to restrain the rate of increase in the Consumer Price Index (CPI).' 1.8 The Zanzibar economy has liberalizedits trade policy. Market prices are also liberalized with the exception o f the buying prices of cloves, which are still under state monopolyby the current law. ' The computation of the CPI is based on a basket of goods and servicesthat has not beenreviewed sincethe last Zanzibar Household Budget Survey conductedin 1991. Structural reforms and policy changes, for example, meanthat existing weights (e.g. 75 percent for food, 0.6 percent for health services and no weight at all for education) makethe headline inflation rate an unreliable measureo f the cost o f living for most o f the population. Current estimates tend to suggest that many households spend more than 10 percent of their incomeson health and educationservices alone. Zanzibar PublicExpenditureReview Page 5 FISCAL PERFORMANCE 1.9 Table 1.4 summarizes Zanzibar's revenues and expenditures performance for the period FY00-02. There are several interestingfeatures characterizingZanzibar's fiscal policy in recent years. Most notable i s the large discrepancy between budget estimates and outturn. Revenue has been overestimated by a large margin. In FYO1, estimated revenue was 24.3 percent of GDP, while actual revenue collectedwas only 16.9 percent o f GDP. Similarly in FY02, estimated revenue was 24.8 percent of GDP while actual outturn was only 15.6 percent of GDP. As discussed in more detail in subsequent chapters, the decline in revenue is likely to be structural rather than temporary, and it will be important that fiscal policy recognizes this change in the resource situationand adjusts accordingly. 1.10 Although expenditures were reduced in response to the revenue shortfall, financingrequirementswere typically largerthan estimated. They amounted to about 3 percent of GDP during recent years and were covered through domestic borrowing in Zanzibar, and throughthe issuing of T-bills by the Bank of Tanzania on behalfof the RGoZ, leadingto a significantincrease in domestic debt inrecent years. 1.11 There are significant data issues that arise in calculating revenues, expenditures, external financing and domestic financing. For instance, externalflows are not captured by the Bank of Tanzania or the Ministry of Finance `flash' reports which have been used as source o f this information. Capturing and recording financial flows o f all types will be a major challenge in the future, while the likely understating of GDP further exacerbates data interpretation. Some caution must, therefore, be exercised when consideringfiscal data. Zanzibar Public ExpenditureReview Page6 Table 1.4: Revenue and Expenditures in Zanzibar, 1998/99-2001102 2002103 bm 199912000 200012001 200112002 Actual Budget Budget Estimates Estimates Budget Actual Estimates T.Shs. millions Recurrent Revenue 62,628 55,610 38,669 67,400 42,241 58,848 RecurrentExpenditures 62,468 53,671 42,272 62,500 46,626 60,073 RecurrentBudget DeficitiSurplus 161 2,939 -3,603 -5,900 -4,385 -1,225 Development Expenditures 599 21,985 1,019 24,800 1,326 36,596 Overall Deficit (charges issued)before grants -439 -20,047 -4,622 -30,700 -5,711 -37,821 Grants 5 0 0 19,000 0 17,921 Overall deficit (charges issued)after grants -434 -1,370 -4,622 -11,700 -5,711 -19,900 Adjustment to cashand other items(net) 6,736 -15,152 -11,843 0 Overall Deficit (checks cleared) 6,302 -20,047 -7,221 -11,700 -19,900 Financing 6,302 20,047 7,221 11,700 19,900 Foreignfinancing Domestic Bank 6,302 20,047 7,221 11,700 Non-bank 0 -1444 GDP at marketprices 193,543 229,195 229,195 271,414 271,414 273,392 %ofGDP Recurrent Revenue 32.4% 24.3% 16.9% 24.8% 15.6% 21.5% RecurrentExpenditum 32.3% 23.4% 18.4% 23.0% 17.2% 22.0% RecurrentBudget Deficithrplus 0.1% 1.3% -1.6% -2.2% -1.6% -0.5% Development Expenditures 0.3% 9.6% 0.4% 9.1% 0.5% 13.4% Overall Deficit (charges issued) before grants -0.2% -8.7% -2.0% -11.3% -2.1% -13.9% Grants 0.0% 0.0% 0.0% 7.0% 0.0% 6.6% Overall deficit (charges issued)after grants -0.2% -0.6% -2.0% -4.3% -2.1% -7.3% Adjustment to cash and other items (net) 3.5% -6.6% -5.2% 0.0% Overall Deficit (checks cleared) 3.3% -8.7% -3.2% -4.3% -7.3% Financing 3.3% 8.7% 3.2% 4.3% 7.3% Foreign financing Domestic Bank 3.3% 8.7% 3.2% 4.3% bon-bank 0.0% -0.5% Zanzibar PublicExPenditureReview Page 7 1.12 Fiscal variables during the period 1981-2001 show significant correlation with the GDP. They also show a high degree of volatility derived mainly from the volatility of GDP (table 1.5). Zanzibar's economy has for many years been characterized by heavy dependence on a single source of revenue-cloves. Revenue from this source has been decliningrapidly, due to decrease in output and leakage in the form of smuggling. In recognition o f this, the Government has set out to rationalizethe planningand budgetingsystem in order to match its expenditures with its revenues, and to diversify its revenue base, in addition to improving the tax administration. Table 1.5 Volatility of Fiscal Variables and Correlation with G D P 1981-2001 Fiscalvariable Correlation Std Dev (SD) of SD of growth of coefficient with annual growth fiscal variableLSD of GDP rates (real) GDP growth RecurrentRevenue 0.95 32 3 Total Revenue 0.95 32 3 RecurrentExpenditures 0.95 42 4 Development Expenditures 0.49 41 4 Total Expenditures 0.95 37 4 GDP (Market Prices) 1 9 1 Source: Calculations basedon data from MoFEA and Office o f the Chief Statistician, Zanzibar THEEXTERNAL SECTOR 1.13 Exports. Exports increased from US$4.4 million in 1995/96 to US$16.1 million in 2001/02. The compositionofexports shows high dependence on cloves, as stated earlier, the marketing of which is controlled by the Government. On average, cloves alone account for about 50 percent o f all foreign exchange earnings in Zanzibar. Second in terms of significance is seaweed, the value of which doubled in the last five years. In 1998/99 its exports exceededrevenue from the export of cloves. 1.14 Zanzibar imports almost doubled in value between 1995/96and 1999/00 from US$60.6 million in 1995/96 to US$119.5 million in 1999/00-see Table 1.6. However, in the more recent years, imports have started to decline to US$82.28 million in 2000/01 and are expected to decline further to around US$68.1million in 2001/02. The recent decline in imports is attributed to a period of enhanced exemptions in the late 1990s, followed by the withdrawal of these incentivesand the closure of loopholeswhich promotedre-exportactivities. The compositionof imports reveals that consumer goods have been the dominant category, accounting for about one-halfof the total import bill. It is estimated that by 2001/02 consumer goods will account for 46.0 percent ofthe total and, within this category, foodstuffs will continue to be the major import item. On the other hand, capital goods have been the second major import category, accounting for between 18.0 percent and 30.7 percent during the periodunder review. Transport equipment dominatedthis category in the period 1995/96to 1997/98 and, thereafter,machinerywas the dominantcapitalimport. 1.15 Trade Balance. The trade balance, as per Table 1.7 below, has been worsening due to the larger increase in importscomparedwith that of exports. This is attributedto the liberalized trade policy and, in part, to the decline in foreign inflows which mighthave boostedthe export sub-sector. Zanzibar PublicExpenditureReview Page 8 Table 1.6: Trade Balance and Re-Exports,from 1995/98to 2001/02 (US$million) I Exuorts I Imuorts /TradeBalanceI 1995196 4.4 I 60.6 I -56.2 1996197 II 6.1 I 65.8 I -59.7 1997198, -74.2 1998199 101.9 -97.9 1999100 35.7 119.5 -83.8 2000101 -76.6 2001102 16.1 68.0 -5 1.9 1.16 It should be noted that Zanzibar has not been compiling a full Balance o f Paymentsposition, mainly due to lack of data on financial and capital accounts. 1.17 Trade by Destination and Source. Approximately 30 percent of all imports are re-exported-mainly to the Mainland. Following a process of tadduty harmonizationwith the Mainland,this figure has startedto decline. Table 1.7: Zanzibar Import-Export Trade 1997-2000 (T.Shs. Mill.) Source; Ministry of Trade, Industry, Marketing and Tourism, 2001 1.18 Exchange Rate. Zanzibar shares an exchange rate policy with the Union Government. The Central Bank (Bank of Tanzania) pursues policies which ensure maintenance o f a competitive exchange rate, while building foreign exchange reserves. The exchange rate is freely determinedinthe Inter-BankForeignExchange Market (IFEM), with commercial banks being the key players. Non-bank financial institutions also participate, while the Central Bank's role is one of intervention (buyingand selling as deemed prudent). The main challenge facing the CentralBank is that o f maintainingstability of the Tanzanian Shilling against foreign currencies in order to win investor confidence. During the recent past, the Tanzanian Shilling has depreciated in nominal terms rather rapidly, mainly due to the poor performance of exports. This has very much affected the Zanzibar economy, through raising the import bill. ZanzibarPublic ExpenditureReview Page9 PUBLICDEBT ANDLIABILITIES 1.19 The statistics on the position of Zanzibar's debt are quite unreliable. Nonetheless, going by what is available so far, it appears that by end-June, 2002, Zanzibar's total debt stock (external and domestic) was T.Shs. 85.7 billion (US$90.7 million), o f which external debt was US$58.4 million, accountingfor 64.5 percent of the total debt stock, while the domestic component was 35.5 percent thereof. Of the external debt portfolio, multilateral creditors account for US$42.7 million, or 73.0 percent of the total external debt. Bilateral creditor claims account for US$8.5 million, or 15.0 percent, while debt owed to commercial sources (suppliers' credit) amounted to US$7.1 million, or 12 percent of the total external debt. All the multilateraldebt is guaranteed by the UnitedRepublicofTanzania (URT). Multilateral 42.7 73.0 Bilateral 8.5 15.0 Commercial 7.1 12.0 1.20 Zanzibar's contribution to the servicing o f its external debt has been limited over the past few years, mainly because o f disagreements over financial responsibilities with the Union Government. Zanzibar's constitution precludes financialcontributionsto the Unionuntil the matter is sorted out by the JFC. IMPLICATIONS OF THE MACRO-ECONOMIC ENVIRONMENTFOR FISCAL POLICY AND PUBLIC EXPENDITURES 1.21 Zanzibar's position as a least developed small island economy and the related macro-economic characteristics have a direct bearing on public expenditure management. Zanzibar's weak economic performance and relatively low economic growth rates represent a serious constraint on its ability to generate government revenue for economic development and poverty reduction. Creating an institutional and policy environment that is supportive o f economic growth thus needs to be a top priority, not only as a means to generate employment for the people of Zanzibar and to increase income, but also to raise the resource envelope for Governmentin order to allow it to fulfill effectively its core functions. 1.22 The high level of vulnerability typical for small islandeconomies is reflected in Zanzibar's high variability of economic growth, exacerbated by fluctuations in imports and exports. In an economy in which government revenue depends significantly on the taxation of trade, this translates directly into high variability of government revenue, which makes expenditureplanningmore difficult than in a more stable environment. In the medium to long term, Zanzibar will need to pursue measures to reduce vulnerability through diversification o f the economy, and to look for means that could provide a buffer against economic shocks. This could include the utilizationof stabilizationhnds, which could help to smooth incomes not only of farmers, but also of Government. Similarly, transfers from the Union Government and from abroad in the form of official development assistance couldalso be designed to compensate for economicfluctuations inZanzibar. Zanzibar PublicExpenditureReview Page 10 2. POVERTYREDUCTION THEZANZIBARPOVERTYREDUCTIONPLAN(ZPRP) 2.1 Poverty is persistentandwidespread, posinga major development challenge to Zanzibar, where over 50 percent of the populationis categorizedas poor. Most o fthe poor live in rural areas characterized by low income and expenditure, food insecurity, high vulnerability to diseases and natural disasters, low productivity, poor nutritional status, low educationalattainment, limited access to means and ways o f mobility, and exclusionfrom economic, social, and political processes. Within urbanareas, poverty manifests itself in terms of poor sanitation, overcrowding in slums, unemployment and underemployment, low earnings, and a general lack of necessary facilities for decent life. 2.2 To bring about economic growth and eradicate poverty, Zanzibar has prepared a DevelopmentVision 2020 that articulates, among others, a long-termfocus interms of poverty reduction, the major aim being eradication of absolute poverty by 2020. Accordingto the Vision, poverty reductionmeans increasingthe ability of the people to obtain the basic necessities, and improving democracy and social security. In reality, poverty reduction is construed to be synonymous with increasing access to employment opportunities, essential services, and income-generatingactivities, and providingsocial safety nets for the most vulnerable. 2.3 As a first stage in the implementationof the Vision, the RGoZ has prepared a Zanzibar Poverty Reduction Plan (ZPRP). The Zanzibar Poverty Reduction Plan, which was launched on May 2, 2002 and has a three-year span up to 2005, focuses on poverty reduction, with the aims of reducing income poverty; improving human capabilities, survival, and social well-being; and containing extreme vulnerability. ZPRP's strategic interventions target the reduction of both income and non-income poverty mainly through improved growth and service delivery. The target is to improve the growth o f the economy to 5 percent during the first year of ZPW, 5.5 percent during the second year, and 6 percent during the third year. To realize both high and pro-poor growth, efforts would be directed to selected sectors: agriculture for reduction o f income poverty; tourism and trade for higher growth; education, health and water for improvementsin human capacity, survival and well-being; and infrastructure for accessibility and lower costs of production. Promotion of investment, strengtheningo f the financial sector, enhancing social stability, and good governanceare other areas o f strategic intervention. 2.4 To succeed in halving the proportion o f population living in extreme poverty between 1990 and 2015 in line with the Millennium Development Goals (MDGs), Zanzibar has placed a greater emphasis on tackling the challenges in agriculture, which provides livelihoods for the majority o f the poor. Apart from Agriculture, the other priority areas for ZPRP include: Education, Health, Infrastructure, Water, Democracy and GoodGovernance, and combatingHIV/AIDS. Zanzibar Public ExpenditureReview Page 11 2.5 ZPRP is an operational plan that incorporates strategies to be implemented based on utilization of domestic financial resources, both public and private, and providinga climate for attractingexternalresourcesto support prioritizedexpenditure plans. The plan evolvedthrough a broadparticipativeand consultativeprocess and is highly regarded as an important tool for tackling poverty issues and bringing about social and economic stability. During the launching of the ZPRP, the stakeholders showed strongcommitmentto supportingZPRP, and underlinedthe need for capacity building, mobilizationo fresourcesto fillthe resource gap, cost-effectiveand efficient management o f economic programs, aggressive promotion o f investments, control of HIV/AIDS, and meeting the MDGs. Zanzibar accepted these challenges and indicated its determinationto build up a firm macroeconomic foundation that would be able to absorb shocks and encourage various players to participate in economic activities. PROGRESSIN THE IMPLEMENTATIONOF THE ZANZIBAR POVERTY REDUCTION PLAN(ZPRP) 2.6 Results on implementingthe ZPRP are contained in the ZPRP First Progress Report. The report highlights progress in terms o f policy reform and technical and physical implementationsof the ZPRP in the last year. As regards policy reform, the RGoZ undertookvariouspolicy and structuralreforms with the objective of achieving stable macroeconomic conditions and the sustainable economic environment necessary for higher growth, low inflation, a stable currency, and a low cost of provision of services. The measures undertaken yielded the following results: (i) GDP grew by 4.0 percent in 2001, compared to 3.2 percent in 2000; (ii)inflation declined from 8.3 percent in 2000 to 3.0 percent in 2001; (iii)budgetary performance improved due to better resource mobilization following tax administrative reforms and expenditurecontrolsvia the cash budget; (iv) positive trends in levels of external inflows in the current fiscal year and improved aid coordination mechanism were achieved; (v) resource allocationto ZPRP priority sectors improved during 2001-02, with the budget being allocated as follows: education 19 percent compared to 8.6 percent in FYOO; health 11 percent comparedto 4.0 percent in FYOO; agriculture 7.0 percent compared to 3.1 percent in FYOO; and transport and communication 6.8 percent compared to 2.9 percent in FYOO; and (vi) increased participation of the privatesector in almost all sectors ofthe economy was encouraged and facilitated. 2.7 The launching o f the ZPRP has necessitated updating of the statistical database for Zanzibar to improve its consistency, quality, and reliability. The database provides a common basis for future analytic work. The diagnostic work in support o f the Poverty Reduction Support Project (PRSP) includes the Public Expenditure Review (PER), the FinancialAccountability Assessment (FAA), and the analysis of Zanzibar's economic situation. A review of procurementsystems is also being carried out, with the objective of reforming them so as to achieve more efficiency and transparency. The main objective of the diagnostic studies is to identify strengths and weaknesses in financial accountability, public expenditures management, and procurement arrangements in the public sector with the aim of ensuring effectiveness, accountability and integrity in the use of public funds. The studies are expected to provide a common point of reference for the Government and its development partners in understanding the current situation and in drawing up plans for the future. They will also help identify benchmarks against which progress Zanzibar Public ExpenditureReview Page 12 can be measured, using international and national standards and best practices. Another survey aimed at assessing the private sector in Zanzibar and identifying issues and opportunitiesfor diversificationand growth of non-traditionalexports will be conducted in 2003. 2.8 The ZPRP progress report also lists preparatory steps that have been undertaken in determiningthe requiredfinancialresourcesenvelope and its allocation. These include: sensitization o f senior government leaders on the PER and MTEF processes; initial trainingof technicalstaff on the PER/MTEF process; passing of and assent to the Financial Administration Act; and institutional measures to improve financial administration, includingexpansion of the domestic tax base, improvement of tax compliance through public education and awareness, and establishment of FinancialAdministration Institute. As of now, resource requirementsare determined at the sector ministries, based on priority activities identified for implementation. Unit cost analysis is only being initiated for the Ministries of Health and Social Welfare and Education, Culture and Sports to guide more professional estimates of resourcerequirements, but for none o fthe other sectors. 2.9 Although some progress has been made in the one year of implementationo f the ZPRP, more needs to be done in terms o f improving financial and public expenditures management, increasingresource mobilization from both domestic and external sources, prioritization of resource allocation, increasing efficiency in resource utilization, improving investment and business environment for private sector development, enhancing and sustaining macro stability, and ensuring good governance and the rule o f law. The diagnostic studies are expected to identify strengths and weaknesses in various public management systems and to suggest areas and ways o f making effective interventions by the Government, development partners, and other stakeholders. Zanzibar PublicExpenditureReview Page 13 3. THE PUBLIC SECTOR INZANZIBAR OVERVIEW 3.1 Zanzibar gained independencein 1964 and soon after unitedwith Tanganyikato form the United Republic o f Tanzania. The newly formed Union Governmenttook responsibility for administering the activities of Tanganyika after the creation of the Union. Zanzibar retained its own government structure and transferred only certain functions-including foreign affairs, defense and security, and monetary policy-to the Union Government. Tanzania's constitution recognizes that Zanzibar should have its own constitution, which provides for the Zanzibar presidency, a council of ministers (the cabinet), a legislature (the House o f Representatives), and a judiciary. The RGoZ is fully responsible for running all government activities in Zanzibar which are not explicitly identified as Union matters in the constitution. 3.2 The public sector in Zanzibar consists of four main elements-functions carried out by the Union Government(GOT),RGoZ, local governments, and the parastatal sector. The RGoZ itself consists o f a legislature, a judiciary, and an executive. This report is mainly concernedwith the executivearm of Government. Table 3.1: AdministrativeDivisions and Number of Shehiasin Zanzibar Unguja Island Pemba North Region Urban West South Region North Region SouthRegion Region NorthA North B Urban West South Central Michweni Wete Mkoani Chake District District District District District District District District District District 28 23 40 29 18 38 13 18 22 19 3.3 Administratively, Zanzibar is divided into five regions, each of which contains two districts. Eachdistrict is further dividedinto shehias. Table 3.2: Share of Public Administration and Services in GDP 1980-84 1985-89 1990-94 1995-99 2000 2001 Public Administration & Services 15.1 15.7 17.2 23.1 24.9 23.0 Public Administrati on 0.0 11.3 13.0 16.4 19.3 18.3 Community & Social Services 0.0 4.4 4.2 6.6 5.6 4.7 3.4 Government statistics indicate that the share of public administrationand service in GDP has increasedfrom about 15 percent at the beginningof the 1980s to almost 25 percent in 2000. The figure for value added includes both union services in Zanzibar and the activities o f the RGoZ itself. Zanzibar Public ExpenditureReview Page 14 FUNCTIONS CARRIEDOUT BY THE UNIONGOVERNMENT 3.5 The division of functions between the Union Government and the RGoZ are defined in the constitution o f the United Republic of Tanzania. The functions of the Union Government include: a) The ConstitutionofTanzaniaandthe Government ofthe UnitedRepublic b) ForeignAffairs c) Defense and Security d) Police e) EmergencyPowers f) Citizenship g) Immigration h) Externalborrowingandtrade i) ServiceintheGovernmentoftheUnitedRepublic j) Incometax payable by individualsand by corporations,customs duty andexcise duty on goods manufacturedin Tanzania collectedby the Customs Department k) Harbors,matters relatingto air transport, posts andtelecommunications 1) All matters concerning coinage, currency for the purposes o f legal tender (including notes), banks (including savings banks) and all banking business; foreign exchange and exchange control m) Industriallicensingand statistics n) Higher education 0) Mineral oil resources, includingcrude oil and naturalgas p) The National Examinations Council of Tanzania and all matters connected with the functionsof that Council q) Civil aviation r) Research s) Meteorology t) Statistics u) The Court ofAppealofthe UnitedRepublic v) Registrationof politicalpartiesand other matters relatedto political parties. 3.6 All matters that are not defined as Union matters are under the sole authority of the RGoZ. However, even with respect to Union matters, the practical division o f responsibilitiesis sometimes not entirely clear. For example, even though higher educationis designated as a Union matter, Zanzibar covers the cost of Zanzibari students in Mainland colleges and abroad. With the liberalization o f education, Zanzibar has also established a Zanzibar State University. Another example is the area of statistics, in which Zanzibar maintains its own Bureau o f Statistics and, as a consequence, has been excluded from important statistical surveys such as the Household Budget Survey carried out in Mainland Tanzania in 2000/01. Carrying out such a survey is now considered a priority by the RGoZ to be able to harmonize the statistics o f Zanzibar and Mainland Tanzania and to create a sounder basis for the computationofGDP. 3.7 The financial relationship between the RGoZ and the Union Government is less clearly defined than the distribution o f responsibilities. The constitution provides for the establishment o fa Joint FinanceCommissionwhich has, inter alia, the responsibilityfor: Zanzibar Public ExpenditureReview Page 15 a) analyzing revenuedexpenditures related to Union matters and making recommendations concerning the net contribution of each Government to the Joint FinanceAccount in the UnionConsolidatedFund; and b) scrutinizingfiscal relationsbetweenthe two Governments. 3.8 The RGoZ has authority to set tax rates and tariffs on non-Uniontaxes, but in 1998 Zanzibar and MainlandTanzania substantially harmonizedtheir tax regimes. There are still differences in exemptions granted. Tax collection in Zanzibar is divided between the Tanzania Revenue Authority (TRA), which collects import-relatedtaxes, and the Zanzibar RevenueBoard(ZRB), which is responsible for the collection of domestic taxes in Zanzibar. All revenue collectedby the TRA andZRl3 in Zanzibar accrue to the RGoZ. 3.9 With respect to external assistance, Zanzibar receives 4.5 percent of program support provided to the United Republic o f Tanzania. There is no system for the distribution of project and other development aid. Inclusionof Zanzibar in project support extended to the Union Government is currentlymade on an ad-hoc basis, in the form of negotiationsbetween the Union Government and the donor. Zanzibar also attracts direct assistance from some donors, such as China. 3.10 Since 1996, Zanzibar is also entitled to and has received 4.5 percent o f the BOT dividend. Responsibility for contracting external debt is vested in the Union Government, which also guarantees debt contractedon behalfofZanzibar. Inthe past, the RGoZhas often failed to make its debt service payments on external debt; such debt service was typically made by the Bank of Tanzania. In turn, the BOTdeducted debt service made on behalf of Zanzibar from the transfer of Zanzibar's share ofthe BOTdividend. 3.1 1 Zanzibar does not make any contributionsto the cost of the Union Government. In 1977, a committee of the ruling party involving the leadership from Zanzibar and Mainland Tanzania established a formula for determining Zanzibar's share in financing the Union Government. A high level committee established in 1992 (the Shellukindo committee) confirmed that Zanzibar should contribute 3.9 percent o f the cost of expendituresfor Union matters, but so far, it has not done so. Zanzibar has also accrued significant arrears in its relationships with parastatal enterprises on the Mainland. Most significant among these are arrears accumulated by the Zanzibar Fuel and Power Corporation for the purchase of electricity from TANESCO, the Tanzania Electric Supply Company. 3.12 Given these facts, getting the Joint Finance Commission to address these issues should be a priority, to enhance the predictability and transparency of the fiscal relationship between the two Governments. In designingan appropriate and feasible system of burden- sharing, the fiscal sustainability of such an arrangement needs to be taken into account. Giventhe severely constrainedfiscal situation in Zanzibar caused by the decline in revenues and an unsustainably high wage bill, normalizingfinancial relationshipswill needto be part of a medium term fiscal reform strategy. A burden-sharingformula will also need to take into account Zanzibar's diseconomies o f scale in its government functions in order to avoid crowdingout poverty-reducingand growth-supportingexpenditures. Zanzibar PublicExpenditureReview Page 16 THECENTRAL GOVERNMENT 3.13 The executivearm o fthe RGoZcomprises four tiers: a) ministries, departments, and agencies (MDAs); b) the regionaladministration; c) the district administration;and d) the shehias. a. Ministries, Departments andAgencies 3.14 At present, the central government comprises 12 ministries. At the head o f each ministry is a politically appointed minister. Five ministries-agriculture, health, communication and transport, water, and education-also have an assistant minister. Principal Secretaries are responsible for the running of the day-to-day business of the ministries. While all ministries have their headquarters in Stone Town, each ministry also maintainsan office in Pembaledby an Officer in Charge. 3.15 The structure of the central government has undergone limited reforms. Following the elections in 2000, its structure was reviewed with the objective o f streamlining government and reducingthe number of ministries. As a result, two ministries (the Ministry of State Planning and Investment, and the Ministry of Information, Culture, Tourism and Youth) were merged with other ministries. However, all staff members were retained in the merged ministries, and only marginal savings were gained from the abolition of two ministerial,principalsecretary, and deputy principalsecretary positions. Table 3.3: List of BudgetaryVotes President's Office Ministry of Education, Culture, and Sports President's Office Revolutionary Council Ministry of Health and Social Welfare President's Office: Regional Administration Ministry of Water, Construction, Energy, and Land Auditor General's Office Ministry o f Communication and Transport Chief Minister's Office Ministry o f Information, Culture, Tourism, and High Court o f Zanzibar Youth Attorney General's Office Ministry o f Youth, Employment, Women, and Houseof Representatives Children Development Economic Brigade Anti-Smuggling Unit Ministry of Finance and Economic Affairs Tourism Commission Ministry of Agriculture, Natural Resources, Fire and Rescue Force Environment, and Cooperatives Ministry of State (P.O.) Constitutional Affairs and Ministry o f Trade, Industry, Marketing, and Good Governance Tourism People's Militia Unit Land and Environment Commission Public Debt Prisons Department b, Regional Administration 3.16 Zanzibar is divided into 5 regions, three in Unguja and two in Pemba. Each region has a regionalcommissioner,who has the rankof a Minister, appointedby the President. The functions o fthe regionalcommissionerinclude: Zanzibar PublicExpenditureReview Page 17 0 monitoring, supervising, and assisting in the execution of the hnctions of the Government in his region; 0 assuringthat the policies, plans and directivesofthe Government are observed; 0 maintaining law and order in the region in collaboration with law enforcement agencies; and 0 assuringthat resources, both materialand manpower, are used for development inthe economy to enhancewelfare. 3.17 Regional Commissioners are also Members o f the House of Representatives. A regional administrative officer, appointed by the President, is in charge of the day-to-day running o f the government business in the region. He also is the head of public officers posted in the regionand the accountingofficer inthe respective region. 3.18 The Minister o f Local Government and RegionalAdministration appoints a planning officer and a community development officer to each region. Sectoral ministries also assign officers and staff to the regions to carry out the responsibilities and duties of the sector ministries. In practice, each ministry appoints at least one officer to every region. These officersare answerable to the RegionalCommissioner. c. District Administration 3.19 Eachregion is divided into two districts; Le., there are six districts in Ungujaand four districts in Pemba. The staffing and functions o f the districts are similar to that of the regions. At the head o f each district administration is a district commissioner appointed by the President, and there is also a district administrative officer in charge of the district administration. In addition, every district has a planning officer and a community development officer, and officers from sectoral ministriesassignedto the district. d. Shehias There are 249 Shehias in Zanzibar, which are at the lowest level of public administration in Zanzibar. Each Shehia is administered by a Sheha, who is a government employee and appointed by the Regional Commissioner. The Sheha is answerable to the District Commissioner o f that area. His functions include: 0 implementingall the government laws, orders, policies, and directivesfor maintaining law and order; 0 reconciling and settling all social and family disputes arising in that area in accordance with the cultural and customary values ofthat area andwisdom; 0 keeping records of all documents relating to the registration o f marriages, divorces, births and deaths, the transportationof crops and livestock, charcoal permits, and so forth, as directedfrom time to time by the institutionsconcerned; 0 controllingimmigrationinhis shehia and keepingrecords thereof; 0 receivingnotification for convokingall public meetings; 0 undertakingall other tasks that are legal and that have been assigned by the District Commissioner. Zanzibar Public ExpenditureReview Page 18 3.20 In every Shehia, there is also an advisory council consisting of not less than 12 members who are appointed by the Sheha, and who perform their functions on a voluntary basis. The function o f the advisory council is to advise the sheha on matters related to the maintenance o f law and order in the Shehia and on other matters which are beneficial to the well-being o fthe shehia. e. Assessment 3.21 The most striking feature o f Zanzibar's government structure is the apparent duplication o f functions and staffing at each level o f government. Staffing and fimctions o f regions and districts as defined in the Regional Administration Authority Act, 1 o f 1998 are virtually identical. There appears to be a significant opportunity for streamlining the vertical government structure by definingmore clearly the role o f regional and district administration, which would go a long way in avoiding the overlap and duplication ofstaffing. 3.22 In addition to the review of the appropriateness of the vertical structure of the central government, there i s also an opportunity for reviewing the division o f responsibilities between(a) the de-concentrated levels o fthe central government and (b) local authorities. At present, the central government retains responsibilities for all aspects o f service delivery, leaving only very limited functions to the district authorities. While this division o f responsibilities seems appropriate for Zanzibar, measures to enhance transparency and accountability toward citizens and local authorities should be explored. LOCAL GOVERNMENTS 3.23 Zanzibar's local government structure comprises 9 district councils, one municipal council, and three town councils. Each council is subdivided into wards. Each ward elects one council member and the RGoZ nominates three council members in each council. r Unguja Pemba Total Regions 3 2 5 Districts 6 4 10 District Councils 5 (West, North A, North 4 (Chake Chake, Mkoani, 9 B, Central, South) Micheweni, Wete) Municipal Council 1(Zanzibar) 1 Town Councils 3 (Chake Chake. Mkoani. Wete', 3 3.24 District councils are assigned the following functions: (i)to formulate, coordinate and supervise the implementation o f plans for economic, commercial, industrial, and social development; (ii)to ensure the collection and proper utilization o f the revenues of the Council; (iii)to make by-laws applicable throughout its area o f jurisdiction; and (iv) to consider, regulate, and coordinate the development plans, projects, and programs o f villages and township councils within its area o fjurisdiction. 3.25 Town Councils are assigned certain specific functions including: (i)the cleaning o f trunk roads; (ii)the regulation and conduct of public hire vehicles; (iii)street lighting; and (iv) the naming o f streets and numberingo f buildings. The functions o f a District Council are more general and include responsibility for: (i)formulating, coordinating and supervising the Zanzibar PublicExpenditureReview Page 19 implementation of plans for economic, commercial, industrial and social development; (ii) passing o f by-laws; (iii) ensuring that revenues are collected; and (iv) considering, regulating and coordinating the development plans, projects and programs o f villages and townships within itsjurisdiction. 3.26 The responsibilitiesofthe Zanzibar Municipal Council include: P controllingall publicroads and streets within the municipality; P namingof streets and numberingofbuildings; k establishingandmaintainingrecreationgrounds; k implementingpublic health initiatives as required by the Minister (responsible for local government administration); P construction, equipment andoperationofdrainage and sewerageworks; and P the administrationofpublicmarkets. 3.27 Borrowing by the Council is only permitted with the approval o f the Minister, while external financing (outside Zanzibar) is only permitted with the approval of the Minister o f Finance. Both the Municipal and Town Councils operate a committee system. The Municipal Council has five committees and five associated departments. These are: (i) Finance and Economic Development; (ii)Town Planning; (iii)Law and Order; (iv) Labor, ConstructionandEnvironment;and (v) Social Services Affairs. DISTRICTANDREGIONAL DEVELOPMENT COMMITTEES 3.28 Zanzibar law foresees also the establishment of development committees for each regionand district. The functionsofthese development committees are defined as follows: (a) to supervise the implementationo f government policies, to identify the problems, and to advise Government on the best way to solve these problemsand to promote development in their areas; (b) to monitor and assist in the formulation of policiesfor local government authorities intheir areas and advise on the best implementationstrategies; (c) to mobilize people to participateand contributein all ways possibleto assist in the efficient uses o f resources and the protection of environment for sustainable development and inall activitiesof nationaldevelopment; (d) to ensure that implementation strategies correspond to relevant policies and to create awareness amongthe people intheir areas on the importanceo f both; and (e) to ensure and establish understanding, cooperation and coordination among government agencies, local government authorities, non-governmental organizations and the people in creating an enabling environment for sustainable development. 3.29 The membership o f regional and district development committees (see Table 3.5) is composed of staff from the de-concentrated levels o f the central government and from the local governments and their officers. Zanzibar Public ExuenditureReview Pane 20 Table 3.5: Composition of Regional and District Development Committees RegionalDevelopmentCommittee IDistrict DevelopmentCommittee I PUBLICENTERPRISES UTILITIES AND 3.30 During the first two decades after independence, state-owned enterprises were seen as the principal instrument to achieve economic development. Consequently, the period from 1964 to 1985 saw a rapid growth o f the parastatal sector. State-owned enterprises had a monopoly position in many sectors o f the economy, including the import and export o f all goods. The parastatal sector contributed as much as 17 percent o f GDP, provided 13 percent o f employment, and about 60 percent o f the total recurrent revenue o fthe RGoZ. 3.31 However, starting in the early 1980s, the performance o f the parastatal sector started to decline, and in 1993 the RGoZ recognized the need for the reform o f the parastatal sector. Government categorized parastatals into three categories: P Strategic parastatalsthat were to remainunder government control; P Non-sensitiveparastatals that were to be privatized on ajoint venture basis; and P Parastatalsthat were to be completely privatized. 3.32 Subsequently the number o f parastatals declined from 25 in 1994 to 12 in 2002. It is worth noting that many parastatals turned out to be not economically viable without the support o f the Government, and these parastatals ceased to exist as commercial entities, rather than becoming part o f the private sector. Table 3.6 provides an overview o f the remaining parastatals. Of the twelve remaining parastatals, six have either ceased their operations or are operating on a very limited scale. Divestiture o f these parastatals has proven to be difficult due to lack o f interest by investors, large liabilities, unclear procedures for dealing with redundancies, and unrealistic expectations with respect to privatization proceeds. Zanzibar PublicExpenditureReview Page21 3-33 The remaining operating parastatals includethe Zanzibar State Trading Corporation, the Peoples Bank o f Zanzibar, the Zanzibar Insurance Corporation, the Zanzibar Ports Corporation, the Zanzibar Shipping Corporation, and the Zanzibar State Fuel and Power Corporation. Without government support, the financial position and economic viability of most of these parastatals is weak. This is due to low efficiency in operations, financial relationships with other parastatals that are characterized by high levels of arrears and debt, presumptive taxation that is not based on realized profits, overstaffing, and weak management. Capital investments are typically financed from the government budget, and represent significant subsidies to the parastatals. Consequently, the economic viability of many o fthe parastatalsis even worse than indicatedby their financial accounts. 3.34 At present, the financial relationshipsamong parastatals, and between the parastatals as a group and the Government, are characterized by a complex set o f debts and arrears. For example, the main debtor of the Zanzibar State Fuel and Power Company (ZSF&PC) is the Government, with arrears of approximately T.Shs. 6 billion, with the water utility being the largest debtor within Government. In turn, ZSFPC has built up huge accounts payable to creditors, owing TANESCO about T.Shs. 26.3 billion. On the other hand, the Zanzibar State Trading Corporation (ZSTC) has invested a large share o f its retained profits of T.Shs.11.6 billion in government debt, includingT.Shs. 2.5 billion in Governmentstock and a loanto the Treasury of T.Shs. 0.7 billion. The indebtedness of GoZ to the People's Bank of Zanzibar (PBZ) currently stands at about TShs. 9 billion. At the same time, parastatal deposits at PBZ amount to about T.Shs. 9 billion, and PBZ is also currently payingmorethan market rates of interest for largedeposits from parastatalcompanies, which contributesto PBZ's losses. 3.35 There is also no, or a very limited, economic rationale for government involvementin the activities o f these parastatals. In cases in which the parastatal functions on the basis o f a government-grantedmonopoly, such as the Zanzibar State Trading Company, the monopoly is likely to have severe negative economic consequences. Similarly, in cases in which the parastatal is the single provider of specific services, such as the Zanzibar Port Authority or the Zanzibar Fueland Power Corporation, poor service delivery has had a negativeimpact on economic growth and poverty reduction. Government has acknowledgedweaknesses in the parastatal sector and the need for privatizationo f many o f the remainingentities. However, efforts have so far beencautious, and privatizationhas been slow. It is recommendedthat the restructuring and privatization of the remaining parastatals be given priority. For some parastatals, especially for the People's Bank o f Zanzibar (PBZ), divestiture will involve significant financial requirementsdue to the negative net worth of these entities, which will require an appropriate fiscal strategy in parallel with the adoption of the privatization strategy. Zanzibar PublicExpenditureReview Page22 Figure 3.1: Zanzibar State Trading Corporation-Ratio of Cost of Sales to Sales, FY91-FY00 40% 35% 30% 25% 20% 15% 10% 5% 0% FY91 FY92 FY93 FY94 FY95 FY96 FY91 FY98 FY99 FYOO Source: Basedon data provided by ZSTC 3.36 The monopoly over the export of cloves-Zanzibar's main agricultural commodity- heldby the Zanzibar State Trading Corporation, is currentlyunder review, and a study on the clove sector, with options for liberalizingclove marketing, has been recently completed. In the past, ZSTC's cost o f sales was typically less than 50 percent o f sales (Figure 3.1). In years with particularly highworld prices for cloves, such as FYOO, the cost of sales was even as low as 7 percent of sales. The recent study on the Zanzibar clove industry (EM, 2003) indicates that heavy administrative costs are a main factor constraining the efficiency and effectiveness o f ZSTC. Figure 3.2 shows that, during the period FY91-FY00, the cost of sales, i.e., payments made to the farmers, were only 14 percent of revenue. Management costs claimed another 12 percent of revenue, and 70 percent of revenues were surplus.2Out o f ZSTC's surplus, one-half i s remittedto the Government. The remainingone-half of the surplus is retained at ZSTC. However, ZSTC reserves typically have been used to provide loansto government and other parastatalsandto purchasegovernment securities. FYOO was an exceptionalyear with extraordinarily high clove prices which were not passed on to farmers. However, even if FYOO is excluded from the calculation of the average cost shares, the surplus still comes to 51 percent over the period FY91-FY99 while cost of sales are 21 percent and management cost are 22 percent o f revenue. Zanzibar Public Expenditure Review Page 23 Figure 3.2: ZSTC-Average Distribution of Revenue, FY91-FY00 Management Cost 12% ost of Sales 14% Repair 2 Yo Financial Cost 2% 3.37 The large difference between world prices and prices paid to farmers has led to significant smuggling activities, and the share of clove exports handled by ZSTC out of total clove production is estimated to have declined significantly, despite government efforts to curb smuggling. In 2002, world prices fell below the price announced by ZSTC; as a result, the ZSTC hadto borrow heavily from PBZ to be able to purchasethe harvest. Even so, it still ran out of hnds and hadto halt the purchaseof cloves from farmers. 3.38 Abolishing the ZSTC monopoly is likely to leadto increasedincomes for farmers, and work is currently underway to study the effects of liberalization of clove marketing. The study recommends measures to strengthen and liberalize the clove industry over a three-to- five-year horizon. However, private participation should be allowed as soon as possible. Since the ZSTC has in the past been an important, though highly volatile source of revenue for government, it will be important to put in place an appropriate taxation system for the clove industry, once clove marketing is liberalized. Zanzibar Public ExpenditureReview Page 24 Table 3.6: Summary of Parastatals in Zanzibar-Position as of FYOl Company Principal Staff Revenue Profit/(Loss) Net Current Privatization Status Activities (TSh. (T.Sh. Assets1 Million) Million) (Liabilities) Zanzibar State Monopoly on 375 1,272 (506) 4,666 Study for Trade export of cloves liberalization of Corporation clove exports (ZSTC) underway Zanzibar Insurance 70 1,23 1 206 Divesture through Insurance business in joint venture with Corporation liberalized strategic partner (ZIC) market planned Zanzibar Ports Operation of 525 3,725 Restructuring and Corporation Port improving facilities (ZPC) Zanzibar State Monopoly on 484 6,400 Restructuring and Fuel and Power the generation, building up Corporation transmission, capacities (ZSF&PC) * and distribution of power in Zanzibar People's Bank Commercial 222 3,251 (4,234) Options for of Zanzibar banking privatization under (PBZ)* considerationby government Zanzibar Shipping agency 203 1,200 428 270 The business has Shipping and shipping liberalized but Corporation line Government is (ZSC) operating its own fleet. Motor Trade Agency for 72 2,700 833 Seeking joint Corporation motor vehicles venture (MTC) Zanzibar Tour operator 121 2,000 10.299 Inthe long run, Tourist corporation needs to Corporation be privatized (ZTC) Mahonda Sugar Ceased operations, Industry to be divested Bwawani Hotel Zanzibar Vessel fishing 21 43,500 1,920 Privatizedto African Fisheries activities Fishing Corporation Corporation and now under liquidation Source: MoFEA and annual reports 'various parastatals *Data for 2000 Zanzibar PublicExPenditure Review Page25 OTHERGOVERNMENT INSTITUTIONS 3.39 The public sector in Zanzibar also includes numerous semi-autonomous government institutions. These include the Zanzibar Institute o f Financial Administration (ZIFA); the Export Processing Zone (EPZ); the Zanzibar Investment Promotion Agency (ZIPA), the Zanzibar Freeport,and the Zanzibar Social SecurityFund(ZSSF). Most of these government institutions receive subsidies from the government to augment resources from their own revenue. Table 3.7: Semi-AutonomousInstitutions: Subventionsand Total Expenditures,FY02 1Institution 1Employment ~. 1 GovernmentSubvention 1 Total Budget - 1 (T.Sh. Million) (TSh. Million) ZIFA 44 154.8 154.8 ZIPA 52 180.0 220.4 ZSSF 31 48.0 183.0 ZRB 109 913.5 1,498,731.0 ZAFREZA 63 135.2 ~ ZCT I90 485.4 ZFPA 23 150.0 104.4 3.41 Beneficiaries are entitled to receive benefits only after contributing for five years to the Fund. Thus, payments to members includedrefunds to members who had died, retired, became incapacitated, or had left the country for good before sixty months of contributions. Benefitpaymentsby ZSSF will commence in July 2003. Untilthen, pensions for government and parastatalemployees are paiddirectly from the government budget. An actuarialstudy is currently beingcarried out to assess the future obligations of the ZSSF. Having to take over pension and gratuity payments for all government employees will put significant demands on the ZSSF, especially since there seems to be a tendency to make use of early retirement (at the age of 55) since the establishment ofZSSF. 3.42 As the ZSSF plays an important role in providinga social safety net for the people of Zanzibar, professional and prudential investment o f the contributions needs to be the top priority for the ZSSF. Ensuringthat the staff and the board of trustees have the appropriate expertise will be crucial. Since this expertise might at present not be available in Zanzibar, havingthe investmentfunds of the ZSSF managed by a well established financial institution with the required expertise should be considered, while ZSSF can be responsible for administeringand collectingcontributions and payingbenefits. 3.43 The Zanzibar Government also supports numerous export and investment promotion agencies. These include the Zanzibar Free Economic Zone Authority (ZAFREZA), the Export Processing Zone (EPZ), and the Zanzibar Investment Promotion Agency (ZIPA). Zanzibar PublicExPenditure Review Parre26 These agencies appear to fulfill similar functions, and there appears to be significantoverlap. Integratingthese three agencies into a single agency might allow for more effective delivery o f services, since scarce human resources could be combined; making the operations less costlyto government;and avoidingconhsionamong potentialinvestors. Zanzibar Public ExpenditureReview Page27 PART 2: PUBLIC RESOURCEAVAILABILITY INZANZIBAR Zanzibar PublicExpenditureReview Page28 4. DOMESTIC REVENUE INSTITUTIONAL SET-UP FOR DOMESTIC REVENUE ADMINISTRATIONAND COLLECTION 4.1 There are three institutionsthat are responsible for domestic revenue administration and/or collection in Zanzibar. The first is the Tanzania Revenue Authority (TRA) which is responsible for all taxes and related fees that are under the jurisdiction of the Union Government. These include income taxes and taxes on international trade (import duty, excise duty on imports, trade levy, suspended duty, and withholding tax). However, all revenue collected by TRA is remittedto the Zanzibar Treasury governed by the principle of retainingthe revenue at source. 4.2 The second is the Zanzibar Revenue Board(ZRE3) which administers inlandrevenue, composed of over 15 different tax sources. The major ones include: VAT-local at the rate of 20%;3 petroleum product levy, which is charged per liter depending on the world market price and local price build-up; airport service charge at the rate of $20 for foreigners and T.Shs. 5,000 for nationals (compared to $30 for a foreign trip charged on the Mainland); parastatal contributions mainly by the Zanzibar State Trading Corporation (ZSTC) which is the sole buyer and seller of cloves; and collections by ministries and autonomous Government agencies, such as the Zanzibar Investment Promotion Agency (ZIPA), the Zanzibar Tourism Commission (ZTC), the Zanzibar Free Economic Zone Authority (ZAFREZA), and the Zanzibar Free Port Authority (ZFPA). With the approval of the Ministry o f Finance and Economic Affairs (MoFEA), various Ministries, Departments, and Agencies (MDAs) are allowed to retain for their own use most o f the revenue which they collect. However, they are required to report revenue collected to ZRB. Other revenue sources for ZRE3 include the hotel levy (15%), restaurant and tour operators levy (lo%), stamp duty (IS%), sea port service charges at the rate of T.Shs. 500 for nationals and $5 for foreigners, businesslicenses, andthe entertainment tax. 4.3 The third is the network of municipal and town/district councils, which are responsible for the administration and collection of local government revenue under the oversight of the Minister responsible for Regional Administration, Local Government and Special Forces. Principal sources of local government revenue includetrade license fees, the crop cess (assessment), productiontaxes, property rents, and fines and penalties. 4.4 Four institutionalconcerns are worth mentioning. First is the duplicationof effort by the several licensing authorities, such as ZIPA, the Ministry of Industries, the Zanzibar MunicipalCouncil, etc. This has resultedin a loophole, whereby would-be licensees choose at will where it i s more convenientto them to obtain licenses. RGoZ i s aware of this problem and has formed a committee to look into the harmonization and coordination o f licensing. Coordination between investment/tourism promotion agencies on one hand and tax institutions on the other also leaves a lot to be desired, particularly with respect to the scrutiny and granting of exemptions. Second, as in many other small islandstates, Zanzibar VAT on manufacturedgoods producedon the Mainland but destinedfor Zanzibar is collected by TRA- Mainland and is subsequently remitted to the Zanzibar Treasury. Zanzibar Public ExpenditureReview Page 29 has limited institutional capacity in terms o f skilled human resources, equipment and facilities. For example, in the case of ZRB, only 32% of the 127 employees have at least an advanced diploma. There is a shortage of qualified tax assessors, collectors, and accounting and managerialcadres. The third related issue is inefficiency in collection. A comparison of the revenue collectedby TRA and ZRB vis-&vis the correspondingrecurrent costs (salaries and other charges) suggests that ZRB is relatively cost-inefficient in collecting revenue. Its average cost-to-revenue share over the past two years has been almost twice the internationally recommended share of 2 percent (see Table 4.1). By contrast, TRA- Zanzibar's recurrent-costs-to-revenueratio has been just about right, averaging 2.2% for FYOl and FYOL4 Zanzibar Revenue Board Tanzania Revenue Authority-Zanzibar Year (1) (2) Ratio: (1) (2) Ratio: Collection Expenditure (2) / (1) Collection Expenditure (2) / (1) (T.Shs.Mill.) (T.Shs.Mill.) (T.Shs.Mill.) (T.Shs. Mill.) 1999100 20,721.2 461.4 2.2% 39,638.6 491.8 1.2% 2000101 15,167.9 612.4 4.0% 23,405.9 499.2 2.1Yo 2001102 20,937.7 756.1 3.6% 21,208.9 491.7 2.3% 4.5 Collection inefficiency in ZRB partly reflects capacity weaknesses and relatively higher operational costs. It is interestingto note that special allowances alone paid to ZRB staff amount to over a quarter (26%-29%) of total annual recurrentexpendituresof ZRB. In view of the above, there is a need to improve tax administration in ZRB, and one option in this regard could be to roll out the Tax Administration Project (TAP) to Zanzibar to support systems re-engineering, automation, and capacity building. The other is to review the economic justification for the continued coexistence o f the two institutions and to propose viable options looking forward, including the possibility o f ZRB contracting TRA to also collect inland revenue. Fourth, although tax legislationin Zanzibar does provide for appeal machinery,it is yet to be operationalizedto deal with taxpayer complaints. This has partly to do with a shortage of lawyersneededto serve on the Tax Appeal Tribunal. DOMESTIC REVENUE STRUCTURE 4.6 Revenue Structure by Type of Revenue: Zanzibar depends mainly on taxes as the major source of revenue (Figure 4.1). In the past five years, revenue from taxes contributed an annual average of 89 percent of total revenue, Zanzibar's revenue structure is heavily skewed toward indirect taxes (Table 4.2). The major indirect taxes include import duties, VAT, and excise duty. These together contributebetween 60 and 70 percent of total revenue annually. Non-tax revenue accounted for between 26 percent and 30 percent of domestic revenues over the past three years. The main sources of non-tax revenue include contributions from public enterprises, and levies and fees collected by tax departments and ministries, departments, and agencies (MDAs). The contributionof direct taxes (pay-as-you- earn, company tax, individual income tax, withholding tax, and capital gain tax) to total domestic revenue in Zanzibar is rather small. In 1999/00, 2000/01 and 2001/02, direct taxes contributedonly 3.6 percent, 7.1percent, and 8.6 percent to total domestic revenue, reflecting a very narrow company tax base in the isles. Part o f the difference in the cost o f collections is related to the fact that TRA collects from a few high-yielding sources, while ZRB collects from a large number o f low-yielding tax sources. Zanzibar Public ExpenditureReview Page30 Figure4.1: Structureof Revenue(YOof TotalRevenue), FY92-FY02 1 Chart 1: Revenue( O h of Total Revenue) 90 70 g +Direct Taxes 50 1--)-Indirect Taxes 1 1 Other Eamings 3n I O 0 199I/921992/931993/941994/951995/961996/971997/981998/991999/002000/0 DO01/02 Y e m Table4.2: Zanzibar RevenueStructure(%ofDomesticRevenue) 4.7 Zanzibar Public Revenue by Collecting Agency: Until recently, revenue collectedby TRA constitutedthe largest component of total revenue for the Zanzibar central government. Annual average collections by TRA amounted to 63 percent o f central government revenue from domestic sources over the period from1999/00 to 2000/01. However, this share droppedto about 50 percent in2001102, with the collapse of imports and, correspondingly,of customs revenue (Figure 4.2). Over the past three years, the contributionto inland revenue from most of the revenue sources administeredby ZRB has been increasing (Annex Tables A2 and A3), except for the contribution from parastatals, which is dominated by declining contributionfrom the Zanzibar State Trading Corporation(ZSTC) as a result of a sharp drop in clove output and world market prices. By contrast, all the major revenue sources (Annex Tables A2 and A4) administered by TRA-Zanzibar (import duty, VAT, and excise duty on imports) have collapsed, resonating the drop in imports through Zanzibar when tariff harmonizationwas adopted, and more recently with the relocationof two main importers to Tanzania Mainlandwhere they also established industries. Zanzibar Public ExpenditureReview Page 3 1 Figure4.2: CentralGovernment Revenueby CollectionAuthority (T.Shs. Billion) 4.8 Annex table 2 also reveals that there are a number o f revenue sources which contribute only negligible amounts to Zanzibar state coffers. This calls into question the economic justification for continuing to charge some taxes, such as the entertainment tax, capital gains tax, export duty, auction sales, sales o f stores, customs warehouse rent, customs agency fees, miscellaneous fees, withholding tax, and suspended duty. Therefore, RGoZ could consider undertaking a thorough review of its revenue sources to identify those that merit repeal, and potential new instruments, such as VAT on petroleum, which is already applied in the Mainland. 4.9 Rate Structure: Import duties are levied on the c.i.f. value of imports, while the base for excise duty is the c.i.f. value of certain imported items, including import duty on the products. The base for VAT on imports is the c.i.f. value of imports includingimport duty and excise duty (if applicable) on the products. There are four nominaltariff rates, including a zero rate on agricultural inputs and implements, capital goods, and raw materials; 10 percent for semi-processed inputs; 15 percent for fully processed inputs; and 25 percent for consumer goods. Imports from the EastAfrican Community (EAC) are charged at 20 percent of these rates. There are also suspended duties on a few items (tea, dairy products, and cigarettes) at the rate o f 10 percent up to 35 percent, primarily for the purpose of granting additional protection to local producers. The VAT rate is 20 percent and is applied at the point o f importationand delivery of service, with a zero rate on exports. In addition, there are eight product groups (soft drinks, beer, wine, spirits, tobacco, motor cars, gasoline, and mobile phones) that are subject to excise tax. Mobile phones, motor cars, and tobacco products are subjected to ad-valorem rates o f 5 percent, 10 percent, and 30 percent respectively, while a specific levy is charged on the remainder of the excisable goods. Although the tax base for international transactions is similar to that used for Mainland Tanzania, there are still some differencesin bothrates appliedand valuationmethods used in Zanzibar, compared to the Mainland. For example, suspended duty in Zanzibar applies to only three commodities (milk, cigarettes, and tea) against fourteen items for the Mainland. The argument is that while the Mainland imposes suspended duties primarily to protect domestic industries, Zanzibar has no similar industries and actually depends on imports of such commodities. As such, impositionof suspended duties on needed imports would have increasedthe cost of living o f the people of Zanzibar. Similarly, while there is pre-shipment inspection and valuation in the case o f imports into Mainland Tanzania, there is no similar arrangement for Zanzibar. There are also differences in the standards for the goods, since those set by the Tanzania Bureauof Standards (TBS) do not apply in Zanzibar. 4.10 Individual income tax rates applicable in Zanzibar since February 2000 (Table 4.3) fall into 11 bands rangingfrom 7.5 percent o f the amount in excess o f T.Shs. 25,000 to TShs. 48,100 plus 35 percent of the amount in excess o f T.Shs. 265,000. This contrasts sharply Zanzibar Public ExPenditureReview Page 32 with rates applicable on the Mainland, with only 4 bands that range from 18.5 percent over T.Shs. 50,000 to T.Shs.105,000 plus 30 percent over T.Shs. 540,000. Tax on the profits of companies i s levied at 30 percent o ftaxable income. IncomeBracket (Monthly)T.Shs. RatePayable 1 c)-2s.nnn - -_> - - - Nil 2,501-45,OOO 0 7.5% over T.Shs. 25,000 + 45,OO 1-65,000 TShs. 3,500 + 12.5% over T.Shs.65,000 T.Shs. 1,500 + 10% over T.Shs. 45,000 65,OO 1-85,000 85,OO1-105,000 TShs. 12,500 + 20.5% over T.Shs. 125,000 T.Shs. 9,000 + 17.5% over T.Shs. 105,000 T.Shs. 6,000 + 15% over T.Shs. 85,000 105,OO1-1 25,000 125,OO1-1 45,000 145,OO1-1 75,000 T.Shs. 16,600 + 22.5% over T.Shs. 145,000 175,OO1-205,000 TShs. 30,850 +27.5% over TShs. 205,000 T.Shs. 23,350 + 25% over TShs. 175,000 205,001-235,000 235,001-265,000 T.Shs. 48,100 + 35% over T.Shs. 265,000 T.Shs. 39,100 + 30% over TShs. 235,000 Above 265,000 4.1 1 It is quiteclear from Table 4.3 that there is room to reduce the number o f tax bands to enhance compliance. RGoZ should therefore consider reviewing the income tax brackets with the view to rationalizeheduce the number o f bands for both ease o f administration and revenue efficiency, while ensuringrevenue neutrality. It i s important to underlinehere that, since the allocation o f revenue from customs duties, income taxes and, excise duties between RGoZ and the Union Government i s based on the principle o f where the revenue i s derived, adherence to this principle hinges on tax administration and enforcement o f legislation being kept uniform throughout the UnitedRepublic o f Tanzania (URT). Otherwise the situation is likely to degenerate into competition for tax revenue and investment between the two sides o f the Republic. Thus, in all these areas, harmonization of rates, tax base valuation, and verification are equally important. 4.12 Local Government Revenue: Local governments in Zanzibar were established by Acts No. 3 and 4 o f 1995, which established the Zanzibar Municipal Council, and District/Town Councils respectively. However, the decentralization process in Zanzibar is still in its initial stage and has so far entailed only de-concentration (that is delegation o f authority) and much less in terms o f actual transfer o f authority and resources to regional administration and local governments. To date, there are no subventions from the central government to individual local governments to finance service provision. All financial resources from MoFEA for financing social services are transferred to the sector ministries, which are the ones that control and do the actual spending and subsequent accounting o f the hnds. Service delivery units (schools, dispensaries, and hospitals) receive in-kind supplies from the respective sector ministries and have virtually no direct funding from the center. However, all permanent staff o f district councils are employees o f the central government and are paid by the center from the general revenue o f the Government. Municipal and towrddistrict councils receive a subsidy from the Central Government and also collect revenue from their own sources mainly to provide support to community-driven development initiatives, and to pay for the wages o f casual laborers who are hired for certain activities such as refuse collection. Based on two local authorities visited (Zanzibar Municipal Council and the Unguja-West District Council), the main sources of own revenue and applicable rates are specific to their areas ofjurisdiction (rural or urban-based and type o f economic activities). The main sources o f revenue include trade licenses and fees, property rent, market dues, registration o f taxi cabs, licenses o f urban transport routes, fines, and crop cess. Although the relevant Acts give borrowing powers to Zanzibar Public Expenditure Review Page 33 the councils, and authority to obtain advances from the central Govemment and overdrafts from banks with approval by the Minister, these financing windows have not been used. Overall, local authorities' (LAs') own sources o f revenue account for less than 5 percent of their total annual expenditures. Inthe case of Zanzibar Municipality, only about 4 percent of expenditures i s financed from its own sources. In 2000-01 the Unguja-West District Council collected only T.Shs. 12.6 million from its own sources, o f which about 75 percent was used to support community-initiated development projects. The main problems faced by LAs in the area o f revenue mobilization include tax evasion, weak tax administration capacity, and inadequate revenue sources. 4.13 Transfers from Government Enterprises. The contribution o f transfers from the parastatal sector fluctuated between 1.4 percent and 9.0 percent o f revenue in the past four years. Table 4.4 shows the transfers o f individual parastatals during the past four years. The main contributors are the Zanzibar Ports Corporation and the Zanzibar State Trading Corporation. The decision by RGoZ to move away from presumptive taxation o f parastatals in favor o f dividends based on a profit-and-loss account o f each public enterprise is a move in the right direction to foster economic decision making in public enterprises and enhancetheir viability. Table4.4: Transfersfrom GovernmentEnterprises,FY99-FY03 (T.Sh. million) Source: Estimates for Recurrentand CapitalRevenuesand Expendituresfor Year 2001/2002 and 2002103 4.14 The high contribution to revenue from parastatals in FYOO is due to extraordinary profits by ZSTC as a result o f high clove prices in the world markets which were not passed on to the clove farmers. Given the current discussion on the eventual privatization o f ZSTC and the potential loss o f income to Government, we show in Figure 4.5 the annual contribution o f ZSTC to government revenue for a longer time period. It is interesting to note that during the period FY93-FY02, ZSTC only made payments to the Govemment in four years, and the average contribution to revenue over that period i s only 1.2 percent. The potential direct loss o f revenue from abolishing the monopoly for the marketing of cloves held by ZSTC i s thus likely to be small and likely to be more than offset by enhanced revenue from increased economic activity inthe clove sector. Zanzibar Public ExPenditure Review Page34 Table 4.5: Transfersby ZSTC to Government,FY93-FY02 1992193 1993194 1994/95 1995/96 1996197 1997198 1998199 1999100 2000/01 2001/02 T.Sh. Million 0 0 0 322 38 0 0 6,159 0 30 1 % ofTotal Revenue 0.0% 0.0% 0.0% 1.2% 0.2% 0.0% 0.0% 9.7% 0.0% 0.7% 4.15 Other Earnings. The key source o f other earnings is port and airport charges such as departuretaxes. The significanceof other earnings has declined inrecent years. On average, other earnings contributed5 percent of total revenue duringthe last five years, whereas, they contributed 13 percent duringthe previousfive year period. 4.16 Transfers from the Union Government. The Union constitution provides for the administration o f taxes on incomes and internationaltrade, and the sharing o f this revenue, Bank of Tanzania (BOT)seignorage, and concessional program loans and grants between Zanzibar and the Mainland. The existing revenue sharing arrangement involvesthe retention of all revenue collectedby TRA at source. With regard to the distribution of program loans and grants, Zanzibar is supposedto receive 4.5 percent of the total. It was notedthat the 4.5 percent subvention, though erratic, provides $8-$10 million to RGoZ annually. In addition, there are also hiddentransfers from the UnionGovernment in the form of Zanzibar's share of the cost o f operating the Union Government (3.9 percent o f total Union expenditures), and, until recently, the artificially low electricity tariffs charged by TANESCO for Zanzibar, which were typically not met. Data from the office o f the Chief Government Statistician shows that government securities and transfers from the Union Government to RGoZ have declined sharply in recent years, dropping from T.Sh.7.9 billion in FY96 to only T.Sh.0.7 billion in FYOO, and further to T.Sh.0.07 billion in FY02 (Table 4.4). The decline partly reflects BOT'Sdecisionto withhold Zanzibar's share o f BOTprofits to cover for outstanding debt service obligations. 7.868 I 5.09 I 1 1.935 I 1.200 700 I 1.000 72 4.17 There are a number o f outstanding contentious issues related to URT-RGoZ fiscal relations. In addition to non-uniform treatment in revenue administration and enforcement already alluded to, the division of tax revenue and obligations (such as the cost o f operating the Union) still remainto be resolved. The 4.5 percent share o f program loans and grants is also regarded by RGoZ as only a transitional arrangement. Therefore, it would be important for the recently established Joint Finance Commission (JFC) to address some of these concerns and other issues, such as possible compensation for the revenue loss suffered by Zanzibar as a result of importtariff unification. 4.18 Domestic Borrowing. RGoZ has had to resort to domestic borrowing to finance its expenditures, mainly through issuance of treasury bills and stocks. The issuance o f stocks was particularly large in 1999/00 (T.Sh.2.5 billion) and was used to finance the general elections. According to BOT, domestic bank borrowing increased from about T.Sh. 0.4 billion in FY97to T.Sh. 4.5 billion in FY99, and further to T.Sh. 6.3 billion in FYOO. Zanzibar Public ExpenditureReview Page 35 REVENUE PERFORMANCE ASSESSMENT: KEYISSUES AND OPPORTUNITY FOR IMPROVEMENT 4.19 Collapse in Revenue: Since FY98, total revenue collection has declined from 25 percent o f GDP to 16 percent of GDP in FY02 (Figure 4.3). Revenue from indirect taxes, mostly taxes on imports, plunged by 40 percent. This decline in revenue is largely attributable to the removal o f external tariff differentials between Zanzibar and Mainland Tanzania, particularly since the closing of the so-called"Zanzibar route" through improved tax administration by TRA. This collapse represents a major stress on Government's capacity to finance its expenditureprogram. The secular decline was interruptedin 1999/00, when a combinationof factors allowed a temporary increase in revenue to about 33 percent of GDP. A major source of this upward spike was a series of large transfers from ZSTC made possible by a sharp rise in clove prices during FYOO, which was not passed on to the farmers. One important measure to restore revenue will be to consider measures to reduce incentives for smuggling o f cloves through the liberalization of clove exporting and appropriate taxationof revenue and profitsfrom clove exports. Other measurescouldinclude curtailing exemptions and improvement in tax administration, particularly in the tourism sector. Figure4.3: Revenue(% of GDP), FY92-FY02 ~ 35% 1 30% , 25% 10% I 0% FY92FY93FY94FY95FY96FY97FY98FY99FYOOFYOlFYO2 ~ 4.20 Variability of Revenue: The Zanzibar Government's revenue has been variable over the past decade, fluctuating between 15 percent and 33 percent of GDP (Figure 4.3). This has, in turn, made annual Governmentexpenditureplanninghighly unpredictable. The more detailed analysis of revenue by source shown in Figures 4.4a - 4 . 4 ~indicates that the variability o f revenue stems from a variety of sources. One of the main reasons for the high variability of revenue is the fact that Zanzibar collects most o f its revenue from indirecttaxes, which reflect instability in the volume o f imports.Another important source of variability in revenue derives from transfers from parastatals and from the Union Government. It is interestingto note that transfers from the Union Government have to a certain extent offset changes in other sources ofrevenues. Diversificationofthe tax base throughthe introduction of new domestic taxes such as VAT on petroleumand propertytaxes would help to reduce its heavy reliance on customs revenue. Given the small size o f Zanzibar, property tax can be Zanzibar Public Expenditure Review Page 36 relatively easily harnessed as long as a good tax cadastre can be setup. However, given the structure of the Zanzibar economy, it is likely that revenue variability will remain a key feature of the economy. In order to be able to minimize the negative effects of revenue variability on public sector service delivery, Government should consider measures which would smooth revenue and prevent revenue variability from directly affecting government expenditures. Such measures could include saving in years with strong revenue performance to be able to compensate poor revenue performance in other years, or the use of central government transfers and donor resources in a way that would compensate for fluctuations that have domestic origins. 25% .I= p" Figure4.4b: IndirectTaxes by Source ( O hof GDP), FY92-FY02 I 12% 10% 1+Import duties 8Yo 69'0 -Other taxes ~ -i- Exciseduty 4yo ,-Sales TadVA' 2% (Local) 0Yo Zanzibar Public ExpenditureReview Page 37 Figure4.4~:Other Revenueby Source (%of GDP), FY92-FY02 9Yo 1 8% 7YO 6% ...m.. .Contrib.fro Pub.Enterprises 5?"o 4% 3% 2% 1Yo 0% 4.2 1 Inadequacy of Revenue for Implementing ZPRP and Need for Growth-Based Expansion of the Revenue Base: Zanzibar's revenue as a percentage o f GDP has fluctuated between 15 percent and 33 percent over the past decade (Chart 4.3). This revenueeffort is in line with international data from small island states which indicatethat these have typically larger governments and higher revenue effort than most continental countries with similar economies. In the case of Zanzibar, there is the additional caveat that GDP figures could be severely underestimated. In spite of the apparent high tax effort, public revenue remains highly inadequate relative to the overall resources needed to successfully implement the Zanzibar poverty reductionplan (ZPRP) and to realize Zanzibar's Vision 2020. Zanzibar's tax base remains very narrow, given its negligible manufacturing sector, significant tax exemptions, and volatile mono-crop agricultural production. In these circumstances, expansion of Zanzibar'stax base needsto be growth-based. Specifically,there will be a need to embark on economic reforms to achieve sustained macroeconomic stability and to spur investment and higher growth. Additional effort to reach all potential income taxpayers, removal of nuisance taxes, presumptivetaxation of the informal sector, and reductionof the multiplicity of taxes are important aspects of the strategy to reduce tax evasion and bringthe informal sector into the tax net. Based on the Mainland experience, RGoZ might want to consider creating a Large Taxpayers Department as part o f the tax reform drive. RGoZ should also devote increased attention to economic diversification, particularly by encouragingthe productionof a variety of other spices, highvalue crops, fruits, seaweed, and fishing. 4.22 Tax Exemptions: RGoZ offers numerous general and fiscal incentives for promoting tourism, trade and investment, mainly in the form of tax exemptions. For example, incentives offered under the ZIPA Act range from a provision for income tax exemption on reinvestment profit, a 25 percent investmentallowance,exemptionfrom exportprofit tax, and tax holidays ranging from 2 to 5 years to exemption from import duty and VAT on the importationof capital goods, raw materials,and the personal effects of expatriateemployees. Similarly, ZAFREZA administers a number of fiscal incentives to promote and facilitate private sector investmentin export processingzones (EPZs). These include exemption from import duties and sales tax on machinery, raw materials, equipment and spare parts necessary for investment in EPZs, export duty exemption on all goods from EPZs, exemption from income tax for an initial period o f 10 years on dividends, interest on shares, loans and any other type o f income receivedby the investors, and exemptionfrom localtaxes on goods and Zanzibar Public ExpenditureReview Page 38 services purchased or hired locally for use in EPZ enterprises. The main objectives o f the fiscal incentivesregime are to (1) protect infantdomestic industries,(2) attract foreignprivate investment, and (3) promoteexport-orientedmanufacturingandtrade. 4.23 In the past three years, exemptions from customs duties alone amounted to about 19 percent of actual annual customs collection (Table 4.5). Approximately 44 percent o f customs duty exemptions are those granted underthe Zanzibar InvestmentPromotionAgency (ZIPA) Act, The other major beneficiaries of customs duty exemptions included the Government (23 percent), private companies and individuals (17 percent) and parastatal organizations (13 percent). Overall tax revenue foregone through exemptions is conservatively estimated by TRA to be in the range of 10-20 percent of total tax revenue collectedeach year. Table 4.7: Zanzibar CustomsDutyExemptionsby Category of Beneficiaries(YOof Total) 4.24 Although these exemptions have to a limited extent helped to attract foreign investment and encouraged some activities for e~port,~they have been subject to abuse and therefore have eroded the potentialtax base. One of the major problemshas been that there are several autonomous institutions that deal with the approval and management of fiscal incentives. These includethe Ministry of Financeand Economic Affairs (MoFEA), Zanzibar Investment Promotion Agency (ZIPA), Zanzibar Tourist Commission (ZTC), Zanzibar Free Economic Zone Authority (ZAFREZA), Zanzibar Free Port Authority (ZFPA) and the tax collectionagencies (TRA and ZRB). Apart from weak coordination and duplicationof effort among these institutions, the tax exemptions granted by these different agencies have not always been consistent and have, by and large, not borne significant expected benefits in terms of operational investments,employmentand, foreign exchange yield. As of February 2002, only 103 projects promotedby ZIPA were operational, compared to the total of 261 approved projects. Similarly, as of November 2002, employment of locals in the EPZs was only 271, comparedto the prospectiveemployment figure of 696 Zanzibaris. There are now seven operatingenterprises in the Export Processing Zone (EPZ), nine more are expected to be established, and five have recently closed or shifted. In addition, most of the institutions dealingwith the promotion o f investment, trade and tourism such as ZIPA, ZAFREZA, and ZTC still depend on subsidies from the central government, in addition to being allowed to retaintheir collectionsfrom licensesand fees. 4.25 Therefore, a thorough review and rationalization o f the exemptions regime is imperativeto counter abuse and improve revenue collection. RGoZ has already recognized some of these weaknesses and has initiated a review o f the relevant legislation to minimize overlap and conflict. Some o f the measures taken have resulted in almost a 50 percent The value o f exports from the export processingzones doubledfrom $1.4 million in 1997to $3.2 million in 2001. The total capital investment o f operational projects promotedby ZIPA and ZAFREZA stood at $1 14.5 million and $67.2 million respectively in 2002. Zanzibar Public ExpenditureReview Page39 reduction in exemptions, particularly following the establishment of a committee to advise the Minister for Finance and Economic Affairs on granting exemptions. Equally important, RGoZ needs to review its investment promotion strategy to create a more favorable investment climate. It should be emphasized that, based on international experience, generous taxation incentives per se do not attract large investment flows. In the case of Zanzibar and similar economies, both the Zanzibar Chamber of Commerce and the Zanzibar Employers Association underlined that it is the lack of skilled labor, poor roads and communications infrastructure, high electricity tariffs, land lease and utilization, and legal and administrative obstacles to investment which are the greatest concerns of current and potential investors in Zanzibar. There is also a need to enhance coordination among the various institutions(ZIPA, ZAFREZA, ZFPA, ZRB, TRA, and MoFEA) that are responsible for administering and approving different categories o f fiscal incentives for investment. Finally, experience elsewhere indicates that VAT and the exemption of government purchases from import duty often provide an opportunity for abuse. Thus, removing the exemption on government purchases in line with that practice elsewhere should be considered. This also would close loopholesfor abuses of the system, such as the selling of exemptions by those who qualify. Improving TaxAdministration, Especially in the Tourism Sector 4.26 Inland revenue collection has in general been increasing since the Board started its work in 1998. However, revenue from VAT (local) and revenue from tourism fluctuated considerablyand declinedfrom 1999/00, partly reflectingthe impact ofthe terroristattack on the United States in September 2000, and the decline in contributions by parastatal organizations, especially by the Zanzibar State Trading Corporation (ZSTC) following the collapse ofthe clove sector startingin 2000/01. Untilrecently, annual contributionso f public enterprisesto government revenue were set by the Ministry o f Financeand EconomicAffairs (MoFEA) on a presumptive basis to counter the tendency of parastatals to under-reporttheir revenues and overstate their expenditures, as there was virtually no comprehensive year-end closing o f their books of accounts. However, this practice has proved to be inefficient and costly, as some o f the public enterprises even had to resort to overdrafts from the People's Bank of Zanzibar (PBZ) to meet their mandatory contributionsto the Ministry of Finance and EconomicAffairs. RGoZhas now changedthis policy in favor of allowing public enterprises to operate commerciallyandbe able to pay dividends. 4.27 There is an opportunity to improve tax administration in Zanzibar and efficiency in revenue collection. As already noted, the RGoZ and URT Governments need to consider extendingthe Tax Administration Project (TAP) to cover Zanzibar. Some o f the issues that would needto be addressed includea review o fZRB's cost structure,remuneration,facilities, exemptions, and taxation of the tourism sector. The tourism sector has a high potential o f contributing to higher economic growth in Zanzibar. Tourist arrivals (and earnings) have shown an upwardtrend over the past decade, rising from 42,141 travelers in 1990 to 97,165 in 2000, but dropping to 76,329 in 2001 following political tensions after the general elections and the aftermath of the terrorist attack on the United States in September 2001. The special island features of Zanzibar, including its climate, rich culture, and long history make Zanzibar one of the most attractive tourist destinations in East Africa. To tap the revenue potentialof increasedtourism activities, RGoZ needs to have a better handle on the taxation o fthis sector. A recent survey o f tourism earnings has been carriedout and could be used for an improved estimate of the tax potential and enhanced revenue generation from tourism. One way to maximizerevenuefrom this source could includeclampingdown on tax Zanzibar Public ExpenditureReview Page40 evasion and the introductionof better tax handleshnnovativeways of collecting revenue in the face o f the growing popularity of package tourism. Inparticular, special focus will need to be directed toward enhancing local capacity to offer quality services to tourists. RGoZ is contemplatingthe introduction o f a tax per tourist head per day on top of the usual tax on hotel services, similar to the practice in the Maldives and Mauritius, but is conscious of the experience elsewhere, e.g., Poland in 2001, where the introduction o f new regulations concerningthe taxation of package tours tended to cause an increase in the price charged by tour operators and ushered in a decline in booked turnover. In addition, a greater and more coordinatedeffort to collectmore revenuefrom other sources and from petroleumproducts is estimated by TRA to have the potentialto raise revenue by as muchas 10percent. Improved coordinationbetween RGoZ and the Union Government on fiscal issues would also go a long way to strengthenthe Union and improve the predictability of resources on both sides of the UnitedRepublicofTanzania. Clamping Down on Tax Evasion 4.28 As in many other countries, tax evasion and tax avoidance are widespread in Zanzibar. Tax evasion is believed to be particularly high in the hotel sub-sector, and for petroleumproducts, which are subject to widespread smuggling. A quick comparisonof the collectionand scheduledrates for some tax categories does point to significant erosionof the tax base via evasion and exemptions. For example, the collection ratio of VAT on imports for 1999/00and 2000/01is about 14.3 percent and 12.3 percent respectively, comparedto the applicable VAT rate of 20 percent, partly due to the exemption of imports by Government and for investmentpurposes. Similarly,the ratio of import duty revenueto importsc.i.f. over the same period is 21.0 percent and 14.3 percent, against the unweighted-averagescheduled tariff rate o f 20 percent. 4.29 Officials of TRA-Zanzibar also indicated that smuggling and mis-declaration and under-valuationof imports that are subject to customs duties are common and significant ways of denying RGoZ revenue. The report by the Controller and Auditor General for 1998/99 and 1999100 does reveal other forms o f tax recalcitrance, such as payment (even by RGoZ itself) without proper receipts or with receipts that do not have a sales tax stamp, thereby denying the Government revenue. It is high time for RGoZ to consider extending pre-shipmentinspectionand valuationto Zanzibar as one measure to curb cheating. RGoZ is also encouragedto consider payingtaxes so as to close some o f the loopholes and to review the penalty rates in order to be more deterrent. Zanzibar Public ExpenditureReview Page41 5. EXTERNALDEVELOPMENTASSISTANCE EXTERNALRESOURCE INFLOWSTO ZANZIBAR 5.1 Since 1996, official development assistance has fallen significantly, possibly with a corresponding impact on social indicators. However, a number o f NGOs continued and intensified their operations in Zanzibar. Nevertheless, the little developmentassistance that continued to flow, including the NGO support, represented a significant share of resources available to the Government for its development activities. For some time now external flows have beenquite erratic and the data are quite unreliable. Table 6.1 shows total external development assistance, including both loans and grants, provided directly to the RGoZ. These flows have been approximately US$10 million in each of the last two years, which represents about US$18 per capita as compared to US$30 per capita on the Mainland. Over 50 percent of the developmentassistance in 2000 and 2001 was directedto the social sectors, with most o f the remainder being used for the development of infrastructure. Taking the various sources of flows together, it is not clear that aid flows have fallen drastically in the aggregate. Total support during that period came from the many non-OECD donors that stayed on, together with the 4.5 percent subvention from the URT which also started that year. Table 5.1 below summarizes the trend of the external resource inflows to Zanzibar for the period 1997-2001. Table 5.1 External DevelopmentAssistance 1997-2001 (US$ million) Source: MoFEA, Zanzibar and ZPRP 5.2 It will be recalledthat following a 1996/97 agreement with the Union Government, 4.5 percent o f general program aid to the Union Government (bilateral and multilateral) i s provided as a subvention to the Zanzibar Government. It will also be notedthat, significant data concerns arise relating to commitments and disbursements of aid. An improved aid coordinationmechanism must be established, not only to improvedata reliability, but also to ensure that development projects are implementedand monitored in a coherent, timely and effective manner. This i s particularly important if the donors are to returnto Zanzibar, in a full way, to assist inthe implementationofthe ZPRP in a coordinatedand orderly manner. A number of donors are waiting for the Governmentto leadthe way inthis process. 5.3 As a process, Zanzibar secures external resourcesthrough a number o f means. First, sectors may approach donors directly or ask the ExternalFinanceDepartmentto negotiate on their behalf. Second, the External Finance Department may initiate the negotiations with Zanzibar Public ExpenditureReview Page42 donors based on identified ZPRP priorities. Finally, the External Finance Department receives a number of donor missions in which funding opportunities are presented and negotiated. 5.4 The appraisal o f grant-fundedprojectsis not undertaken inany structuredmanner and donors may undertake their own appraisals, but the Government has no equivalent independent mechanism. Sector ministries are encouraged to appraise programs, but have notedthat they are not technically equippedto do so, while the Commissionerof the External Finance Department makesthe final recommendation on all external projects. COORDINATION, MONITORING MANAGEMENT AND OF EXTERNALRESOURCE INFLOWS i) The Current Status 5.5 Currentlythere are about 57 projects/programsrecognizedby the Ministry o f Finance and Economic Affairs in Zanzibar (as recorded in the 2002/03 budget), although the precise number is subject to verification as part o f the ongoing database exercise. Based on known projects, infrastructureaccounts for over 40 percent of all recordeddevelopment assistanceto Zanzibar-this is mainly roads, but power and the airport runway also are part of this component. Education accounts for about 25 percent, while health claims 12.5 percent o f inflows. These three sectors are the largestrecipientsof external resourceinflows. 5.6 The Governmenthas no strategy akin to the Tanzania Assistance Strategy (TAS) on the Mainland, for example, for explicitly guiding the use o f external assistance. Instead, it makes reference to the ZPRP to indicate areas, ifnot modalities, of desired assistance. This i s particularly the case for the development partners who are recently returningto Zanzibar, as they have been encouraged to use the ZPRP as their main point of engagement. The Vision 2020 is thought to be too broad and with insufficient prioritization to serve as an indicator for externalassistance. ii) TheMonitoring of External ResourceInflows 5.7 The recording of external flows is the responsibility of the External Finance Department under MoFEA. A database is maintained, but it is felt to be incomplete with respect to coverage of all projects, and to data on individual projects. Agreements are also kept in the External Finance Department although in many cases, particularly for grant- financedprojects, disbursement applicationsand noticesare sent directlyto the implementing agents and are not always copiedto MoFEA. Direct payments by donors to the providers of goods and services are particularly difficult to capture. Where donor reporting is applied, there is not a common format and, aside from a short section in the Budget Speech, the MoFEA produces no regular summary report on the use of development assistance. The External Finance Department also participates in some project procurement exercises. A Central Tender Board has been recentlyestablished, and regulationsare beingdeveloped. In the meantime, ministries form their own procurement committees which are supposed to include External Finance representation. Due to staffing and capacity constraints, however, this is oftennotthe case. 5.8 Physicalmonitoringo f project activities is the responsibility o f the Sector Policy and Research Department of the MoFEA. Quarterly monitoring visits are made to projects to assess progress in implementation and a report i s then presented to the National Planning Commission. There is no financial reporting conducted during these monitoring missions, Zanzibar PublicExpenditureReview Page43 while the expertise and staffing of the Department restricts the scope and degree of monitoringactivities. The ExternalFinance Department provides data to the Budget Section for the preparation of the budget, but acknowledges that it is incomplete. Grants received from Unionprojects, for example, are often excluded from the estimates. There are currently very weak leaks between line ministries and the MoFEA for reporting on receipts and expenditures of project assistance, meaningthat the DevelopmentBudgetestimates are quite incomplete. iii) Assessment of the Current Situation 5.9 The most important issue to address with regard to the use of external assistance relates to the by-passing of the MoFEA in the disbursement of project funds. There is effectively no system in place through which external resources can be projected although sector ministriesare requestedto file returns on disbursements for retrospectiverecording-a practice with which sector ministries, by their own admission, comply only very loosely. Needless to mention, this practice undermines the ability o f the MoFEA to plan comprehensively based on the use of all resources available, thereby jeopardizing the consistency of domestic expenditures with externalassistance, e.g., in ensuringthat recurrent resourcesare set aside to safeguardthe sustainabilityo f aid-financedprojects. 5.10 This by-passingo f the central budgetingsystem also means that the ExternalFinance Department and, more importantly, the Accountant General's Office are often unable to ensure compliance on reporting and accounting for the use of funds. This also creates inefficienciesin budget execution, as resources available to the Government are not brought to the attentiono f the MoFEA. In a situationinwhich counterpartcontributions are required, this can affect the smooth implementationo f project activities. With regard to the ZPRP priorities,the ability of line ministriesto source and negotiatetheir own grant funding, before presenting the proposals to MoFEA for endorsement, suggests undermining the overall prioritizationin resourcemobilizationthat the ZPRP requires. As such, there is no guarantee that the projectportfolio will reflect ZPRP priorities. 5.1 1 There is a clear fiduciary risk in the current arrangements, when neither the MoFEA (ExternalFinance)nor the Accountant General is informedor involvedinthe managementof external resources. This particularly relates to the current arrangements for project procurement which, by the ExternalFinance Department'sadmission, are of serious concern. Until a Central Tender Board is established, ministries will undertake these activities themselves with few external checks and balances. Sectors may comply with the donors' procurement procedures but the mainconcern here is the by-passingo f MoFEA to recordthe nation's correct external obligations. Some degree of transparency should be lent to the process by MoFEA participation,but this does not work routinely due to the limited capacity of the External Finance Department (EFD) to participate in this work. By its own acknowledgement, the EFD has no expertise in this area. This is further exacerbated by the Department's lack of awareness of many project implementation activities, including procurement. With regard to physical monitoring, the absence of complete information on the existence of projects and the resources which they have received represents a major constraint on accurately assessing the progress that has been made in the implementationof the development budget. This also clearly undermines the ability to perform any financial monitoring. 5.12 The new Access-based database which is currently being designed should provide a basis for addressing many o f these problems of the debt position and its management. It is Zanzibar Public ExpenditureReview Page44 intendedthat it will providefull informationon the fundingo f projectsand will be accessible to a wide range of Government institutions. More important, however, is the needto ensure an improved flow of information, based on an improved institutionalstructure, so that data can be made available for recording, accounting, and monitoring purposes on the debt situation. 5.13 It is clear from the aforesaid that the exchange of information between the MoFEA and executing ministries is not effectively organized or coordinated, meaning that the complementaritybetween Government and foreign resources cannot be assured. Given that line ministriesmay face a reduced allocationo f domestic support ifthe full extent of external assistanceto them were known, there may be added incentivesfor project-executingagencies to delay or avoid reportingto the MoFEA. Similarly, NGO-executedprojects, which have become increasingly important in Zanzibar over the last decade, are not recorded meaning that these activities cannot be taken into account to promote an optimal allocation of Government resources both within and across sectors. In a well organized system, NGO resourcesare complementary to Government efforts. 5.14 In most cases, many development partners include a Zanzibar component in their assistance to the United Republic and this is often not explicitly identified in the Treasury (MoF) database of external assistance. Similarly, the RGoZ Ministry o f Finance and Economic Affairs finds it very difficult to record flows of this type of assistance, which are typically channeled directly to the project account by the donor. Most often, in the case of loans, disbursements by the donor to the project are notifiedto the URT but not to the RGoZ. This presentsproblemsnot only for the MoFEA, which is unaware o f fimds provideddirectly to project accounts, but also for the Union Treasury, which may overestimate resources available to it. Increased consultation between the URT Treasury and MoFEA has been established, but no systematic exchange of information has been implemented. Full implementationofthe Joint Finance Commission(JFC) will solve partof this problem. 5.15 At the project appraisal stage, there appears to be little capacity to assess relative priorities and to gauge the extent to which a proposal has taken into account the best alternativeoptions for effectively attaining the targets identified in the ZPRP. The efficient and transparent use of external resources, within and between sectors, cannot be assured in the absence of a higher-levelmechanism for reviewing and appraising all project proposals. Government institutionsstarved for resources may also accept, without question, the offer of project assistance as this allows them to `go-project' their recurrent expenditures. The result is that externally supportedprojectswill not have their desired impact on poverty; ratherthey will be used to maintain the day-to-day operations of Government. This is particularly significant, and of concern, for loan-financedprojects. Currently, debt management is mainly the responsibilityo f the Accountant General's Office, but the External Finance Department should play a role in establishing criteria for accepting new loans after a thorough sustainabilityanalysis. What is really requiredin Zanzibar, in this regard, is a NationalDebt Strategy. iv) Suggestionsfor Future Improvements 5.16 The Ministry o f Finance and Economic Affairs needs to take responsibility for establishingand institutionalizinga Government-wide aid coordination system to ensure full recordingand accountingof external resources. The system should also operate on at least five levels: between RGoZ and URT; between RGoZ and donors; within RGoZ (ministries, MoFEA, Accountant General, and Auditor General); within MoFEA (External Finance, ZanzibarPublic ExpenditureReview Page45 Budget, and Planning); and, betweenRGoZ and NGOs-in terms of where they are working and the sectors which they are supporting. 5.17 The ongoing database work needs to be refinedand to become operational as soon as possible. It is important that the database be viewed in the context of this wider system for aid coordinationto ensure that informationflow is done as a matter o f routine, and can then be used for accountability functions (links to the Accountant General's Department) and for budget work. The Government thus will need to give priority to the establishment of an institutional mechanism for information sharing and reporting, because the creation of a database will not by itself guarantee improvedtransparency in the use of externalresources. Currently the MoFEA is doing its best, but may need to become more proactive in seeking out information and establishing routine data systems, both for improving financial accountabilityand for ensuringthat projectsare implementedas intended. 5.18 More strategically, and related to the establishment o f a reporting and accounting system discussed above, it is recommended that the Government give increased thought to the role of external assistance in the implementation o f the ZPRP. To this end, a more general aid coordination mechanism, creating an interface for dialoguewith all stakeholders on the design of programs and the funding of the ZPRP is desirable. Comprehensive and carefully costed-out programs, based on the use of verifiable indicators, are required at both the sector and thematic levels in order to provideguidance on which to base external finance. Such a system will also provide more structure and coherence to the sectors themselves. A central and consultative forum for appraising project proposals and for assessing their consistency with ZPRP priorities will also need to be established to ensure an appropriate mobilization of external support. While sectors should be encouraged to prepare comprehensive plans to form the basis for external resource mobilization,this will be offset against the needto consider other relevantZPRP priorities. 5.19 Capacity development appears necessary at all levels, and the mix o f technical assistance and development capital that is required needs to be given careful consideration. In the first instance, these plans need to be developed at the sector and thematic levels, but they should be guided by the ExternalFinance Departmentwhich can then consolidatethis work into a coherent framework for external support to the ZPRP. Sector mechanisms, coordinatedand guided by MoFEA, will bring greater ownership to the ZPRP, which will reduce opportunities for sector ministries and donors to misinterpret the ZPRP priorities. Leadership at the sector and thematic levels will create the entry point for returningpartners and ensure a coordinated approach to the establishment o f sector strategy, and promoting capacity development. Zanzibar PublicExpenditureReview Page46 6. PUBLICDEBTAND LIABILITIES ZANZIBAR'SDEBTSITUATION i) Current Status 6.1 The statistics on the situation of Zanzibar's debt are quite unreliable. Nonetheless, going by what is available so far, it appears that by the end of June, 2002, Zanzibar's total debt stock (external and domestic) was T.Sh. 85.7 billion (US$90.7 million), o f which external debt alone was US$58.4 million, accounting for 64.5 percent, while the domestic component accounted for 35.5 percent o f the total debt stock. Of the external debt portfolio, multilateral creditors account for US$42.7million or 73.0 percent of the total external debt. Bilateral creditor claims account for US$8.5 million or 15.0 percent, while debt owed to foreign commercial sources (suppliers' credit) amounted to US$7.1 million, accounting for 12 percent o f the total external debt. It is well understoodthat all of the multilateraldebt i s guaranteedby the UnitedRepublic ofTanzania (URT). 6.2 The RGoZ has been borrowing both internally and externally to finance its development activities, especially development projects and programs. However, as it will be shown later in paragraph 6.9, a preliminary assessment of Zanzibar's debt situation may not be as unsustainable as it may appear. The main problemhas been the unreliability of the numbers. Measures are beingtaken to reducethe debt, but the capacity o fthe Governmentto meet its debt service obligations is still very limited. In fact, for quite some time now the Government has not been serving its external debt-apart from that portion which is guaranteed by the URT. There is a need to develop a National Debt Strategy to guide the Government in its borrowing, and in making repayments which will be in line with budget operations. 6.3 By the end of June 2001, Zanzibar'stotal debt was estimatedto be T.Shs. 91.8 billion (US$102 million), equivalent to about US$l05 per capita. Ofthis amount, T.Shs. 75.7 billion (US$83.9 million), or 83.3 percent of the total, was external debt, and TShs. 16.3 billion (US$18.1 million) was domestic debt. The domestic debt appears to be quite understated as it does not include interest arrears. The overall debt burden represented approximately 50 percent o f GDP. An additional T.Shs. 4.7 billion of arrears to the Social Security Fundhas also been identified. ii) External Debt 6.4 For quite some time, external debt (especially multilateral debt) has been guaranteed and serviced by the Union Government. Ofthe total US$83.9 million debt in June 2001, the Government had cleared US$25.5 million by June 2002, partly through the writing-off of a Chinese loan. Thus, by the end o f June 2002, total debt stood at US$58.4 million, of which US$42.7 million (or 73 percent) i s multilateraland US$8.5 million (or 15 percent) is from bilateralsources. The remainingUS$7.1million (or 12 percent) o f debt is from commercial sources. As stated above, the entire amount of the multilateral debt is guaranteed by the UnionGovernment. The RGoZ guaranteedboththe bilateraland the commercialdebt. Zanzibar Public ExpenditureReview Page47 6.5 Since 1995/96, the Government has not been able to honor its bilateral external debt obligations. The stock of debt, together with scheduled debt service, is therefore currently building up. In recent years, the Government has been allocatinga token amount (less than 0.1 percent o f expenditures) from its budget for payment o f debt. Even this meager amount has not been paid, due to severe budgetary pressures. Table 6.1 below summarizes the externaldebt position. Table 6.1: Zanzibar's External Debt(June, 2002) Creditor Amount (US$ Mil) % Multilateral 42.7 73.0 Bilateral 8.5 15.0 Commercial 7.1 12.0 6.6 Zanzibar's contributionto the servicing o f its externaldebt has been limited over the past few years, mainly because of disagreements over financial responsibilities with the Union Government. Zanzibar's constitutionprecludes financial contributions to the Union until the Joint Finance Commission(JFC) has beenestablishedand made operational. iii)Domestic Debt 6.7 In 2001, Zanzibar's domestic debt was estimated at 8.7 percent of GDP. Most of these debts are not currently being serviced. However, T.Shs. 893 million of the debt owed to the Social Security Fundhad been paid by December 2001. Table 6.2 below summarizes the situationregardingZanzibar's domestic debt by June 2001. Creditor T.Shs. billion YOof total Union Government 12.7 57% Zanzibar State Trading Corporation 2.7 12% Zanzibar Ports Corporation 0.6 3% People's Bank of Zanzibar 7.2 32% Total 22.2 100% Social Security Fund/pensioners dues 4.7 6.8 However, by June 2002, the situation had deterioratedfurther. As Table 6.3 below shows, by June 2002, the domestic debt had increased from T.Shs. 23.2 billion (excludingthe T.Shs. 4.7 billion owed to ZSSF) of June 2001 to T.Shs. 30.4 billion. Out of the total domestic debt of T.Shs. 30.4 billion, T.Shs. 13.6 billion (45.0 percent) is owed to the Union Government. The People's Bank o f Zanzibar (PBZ) claims amounted to T.Shs. 7.2 billion (24.0 percent); Zanzibar State Trading Corporation(ZSTC) and Zanzibar Port Authority are owed T.Shs. 2.9 billion (9.0 percent) and T.Shs. 0.6 billion (2.0 percent), respectively. Other creditors claim a total of T.Shs. 6.1 billion, accountingfor 20.0 percent o f the total domestic debt. Zanzibar PublicExDenditureReview Pane48 Table 6.3: Zanzibar's DomesticDebt(June, 2002) "4'1" ~~ ~~~ Creditor ~~ Amount T.Shs. Bil. UnionGovt. PBZ 24.0 ZSTC 2.9 9.0 I z-PortsCo. I 0.6 I 2.0 I Others 6.1 20.0 Total 30.4 1 100.0 6.9 In addition to the outstanding debt identified above, significant arrears exist, and systems needto be put in placeto prevent the recurrenceo f public arrears. A centralpayment and Local Purchase Order (LPO) system seem crucial to ensure better expenditure and commitment control. In fact, in addition to the direct public debt, there are also Public Enterprises which are in debt-which i s a contingent liability to the Government. The parastataldebt is not yet verified, as the Government's Departmento f Public Investments has not carriedout an audit to get a clear pictureofthese debts. A goodnumber ofthe parastatals however, claim that they are not in debt as such, but rather that it is the Government which has borrowed from them. They contend that, if the Government pays them, they will in a very solvent position. In view o f this impasse, it is recommended that an audit of the parastatalsbe carriedout to verify the position. iv) Preliminary Debt Sustainability Assessment 6.10 Analysis of the available numbers indicates that, the situation notwithstanding, Zanzibar's debt is probably not unsustainable. For instance, the total debt stock is only at the level o f about 50 percent of GDP, as compared to over 90 percent of the URT. It appears Zanzibar is not able to service its debt mainly because the paymento f Personnel Emolument (PE) alone takes up so muchof its budget. PE is estimated at an average of about 60 percent of the entire expenditures budget. Furthermore, its debt service ratio is slightly less than 100 percent, a fairy comfortable position when compared to the HIPC (heavily indebted poor countries) threshold of 150 percent. In addition, reschedulingwould also lower the NPV of the debt, and, more than likely, would keep Zanzibar below the sustainable Debt Service/ Exports ratio of 15 percent. Commonwealth ratios of Debt/Revenue are 200 percent for sustainability. Assuming revenue of T.Shs. 60 billion, clearly this is also sustainable. However, given the unreliability of the numbers, it is recommendedthat a detailed DSA be conducted in order to get a clear picture o fthe sustainabilityofthe debt positioninZanzibar. ii) The Contracting of Debt 6.11 Zanzibar has no authority to contract external debt (especially multilateral debt) as this is a matter ascribed to the Union under the 1977 Union Constitution. Many of the loans negotiated by the Union Government, however, have a program of activities to be implemented in Zanzibar and, where this is the case, the RGoZ is involved in these negotiations. This function i s primarily performedby the ExternalFinanceDepartment ofthe Ministry o f Finance and Economic Affairs. There are also cases o f loans contracted exclusively for Zanzibar and, once again, the RGoZ is represented throughout these negotiations. However, RGoZ can not singly negotiateany Paris Club debt reschedulingor cancellation. 6.12 All external debt contracted since 1993 has been guaranteed, and serviced, by the Union Government. The RGoZ had contracted some external debt in the years up to 1993, Zanzibar PublicExpenditureReview Page49 and, ipso facto, it is responsible for servicingthis component of the debt stock. For instance, o f the total US$58.5 million debt outstanding in February 2002, the amount of US$46.1 million (78 percent) was guaranteed by the Union Government. The main creditors are the African Development Bank (47 percent), IFAD (10 percent), IDA (9 percent) and the Organizationof PetroleumExportingCountriesFund(7 percent). Of the remaining debt not guaranteed by the Union, nearly all was contracted with the Government of China and various Chinese parastatals. However,the RGoZ had cleared US$24.5 million in early 2002 through the writing-off o f a Chinese loan. MoFEA has considered a World Bank-supported Debt Buyback (DBB) arrangement, butthis is yet to be finalized. 6.13 As stated earlier, since 1995-96, the Government has not serviced its external debt obligationsand has not fully recorded interestarrears or penalty interest. The debt stock may therefore be somewhat higher than that recorded by the RGoZ Accountant General's office. There is no debt strategy for repaying.or negotiatingthe reschedulingor cancellationof debt. Further, a Debt Sustainability Analysis has not been conducted for Zanzibar. The RGoZ Ministry of Financeand EconomicAffairs, chiefly throughthe ExternalFinanceDepartment, is able to directly mobilize grant funding, and has made use o f this facility to fund approximatelyone-thirdof its development budget-the remainderbeing foreign loans and a smalldomestic counterpartcontributioncomponent. CONCLUSIONS AND RECOMMENDATIONS i) Debt Coordination and Management 6.14 At the moment, there are various units which appear to be involved in debt management, albeit in a very uncoordinatedway. The unit at the Accountant General's Office is very weak and ineffective. The External Finance Department of MoFEA plays a role in the negotiationsand the contractingof debt. The Department maintains a database, but it is considered quite incomplete, since project financing has been very difficult to capture. Physical monitoring o f project activities is the responsibility o f the Sector Policy and Research Department o f MoFEA. In addition, this department provides data to the Budget Section, but acknowledgesthat it is also incomplete. Grants receivedfrom the Union projects, for example, are often excludedfrom the estimates. There are, at the moment, very weak links betweenthe line ministriesand MoFEA on reportingreceiptsand expenditures o f projectassistance, implyingthat the DevelopmentBudget estimates are incomplete. 6.15 This by-passingo f the centralbudgetingsystem also means that the ExternalFinance Department, and more importantly, the Accountant General's Office are unable to ensure compliance on reporting and accounting for the use of funds. Furthermore, in such a situation, it becomes difficult for MoFEA to budget properly for the counterpart funding component. On the other hand, the Bank of Tanzania (ZanzibarBranch) maintainsa separate data set o f Zanzibar's debt position. In general, the situation is very weak and uncoordinated. MoFEA acknowledges that the situationneeds to be improved. Inview of this situation, it is advisable to create an effective Debt Management Unit in MoFEA to coordinate all debt matters. The Bank of Tanzania (Zanzibar Branch) has offered assistance, especially in the area o fusingthe computerizeddebt management system (CS-DRS). Zanzibar PublicExPenditureReview Pane 50 ii) The Contracting of Debt 6.16 A centralized system giving legal authority to only one ministry-namely the Ministry o f Finance-for contracting debt on behalf of any Government has been highly recommendedby many countries. Ifsuch a discipline does not exist, then any official in the Government will contract debt, leadingto a state of confusion whereby the statisticswill not be compiled and reconciled, and the debts may not be in line with the priorities of the country. The situation may lead to leakages and misuse of the resources. Zanzibar actually has a requirement, by law, that it is only MoFEA that has powers to contract debt on behalfo f the Government. The problem however, has been the ineffectivenessand poor enforcement of such a requirement. There are still a number ministries which borrow-although mainly domestically through the window of suppliers' credit-while waiting for monthly releases from MoFEA. Some parastatals have also been doing the same, which is a contingent liability to the Government. Inview o fthis situation, it is recommendedthat the current legal status be reviewedto providea solid basis for strengthening enforcement. iii) The Needfor a Debt Strategy 6.17 At the moment, Zanzibar does not have a Debt Strategy. A debt strategy will guide the Government on the limits to borrowing, and will specify the types of loans and their maturities, includingan assessmentofthe repayment capacity and repayment schedule. Such a strategy will also ensure that debt service is in line with the operations of the budget. Inthe absence o f such a plan, borrowingand repaymentwill be quite erratic and, in most cases, will be far out o f the budget projections. Authorities have also already seen the need for a debt strategy. In view of this gap, it is strongly recommended that the Government should, as soon as possible, prepare a Zanzibar Debt Strategy (ZDS). iv) TheNeedfor a Debt Task Force 6.18 As stated earlier, there are various sources o f the numbers on debt position and, unfortunately, they all differ. Without a clear pictureofthe debt situation-both externaland domestic-it i s difficult for the Government to know its obligations so that it can, inter alia, prepare a realisticdebt repayment schedule. The MoFEA is consideringimplementationof a debt census. It is thus beingrecommendedthat a Zanzibar Debt Task Forcebe constitutedto bring the debt situation into a reliable and dependable position. The Bank o f Tanzania (Zanzibar Branch)has offeredits assistance inthis area. Zanzibar Public ExpenditureReview Page 51 PART3: PUBLICEXPENDITURES Zanzibar Public ExpenditureReview Page 52 7. ZANZIBAR'S PUBLIC EXPENDITURES i) Introduction 7.1 Zanzibar's public spending is characterized by a high level of recurrent expenditures and a very low level of capital expenditures. Within recurrent expenditures, the bulk of the spending is on wages and salaries for an overstaffedbureaucracy. A huge wage bill, which the Government must clear before makingany other expenditures, makesthe overall demand for funds very inflexible. Also, the Government operates a "cash budget," Le., it spends only what is available as cash. While this cautious fiscal stance allows the Government to maintain overall fiscal discipline in the short run, the cash budget system combined with highly variable revenue receiptstranslates into a variable and uncertainavailability of funds. Thus, we have a situation of high and steady demand for funds juxtaposed against a highly variable supply of funds, which finds the Government trying to scrape the bottom of the barreljust for clearing the wage bill in bad times, and splurging in good times-mainly on stepped up hiring and increased salaries and allowances. Expenditures on creating new public assets and on maintainingand operatingthe existing ones have become a casualty in the process. In order to reduce poverty and induce growth, the structure and management of public expenditures will need to change. The Government is making efforts to realign spending priorities with financing priorities identified in the ZPRP. However, much more remains to be done. The Government will needto move on three fronts: (i) creating space for increased developmentaland other expenditures in priority areas by first controlling, and then cutting down, the wage bill; (ii)improving the efficiency of spending in delivering public services-this would require reallocation of resources within sectors; and (iii) improvingbudget management. STRUCTURE OF ZANZIBAR'S PUBLIC EXPENDITURES ii)Aggregate Expenditures 7.2 Total central government expenditures in Zanzibar steadily increased during the second half of the 199Os, and almost doubledfrom 17 percent of GDP inFY95 to 33 percent of GDP in FYOO (Table 7.1). FYOO was not a normal year, as expenditures spiked due to elections-relatedexpenses. Since then, expenditures have returned to more normal levels. During FY02, aggregate central government expenditures were around T.Shs. 50 billion, which is about 20 percent of the GDP. Most of the increase inthe late 1990s was accounted for by increased spending on two categories: (i)goods and services, and (ii)wages and salaries, which doubled from six percent to 12 percent o f GDP between FY95 and FYOO. The increased spending on goods and services has been reversed in the past two years. However, large expenditureson wages and salarieshave cometo stay and are still onthe rise. Zanzibar Public ExpenditureReview Page 53 iii) Expenditures by Economic Classifcation Recurrent versus Development Expenditures 7.3 It is not the aggregate level, but the structure, of expendituresthat is worrisome in Zanzibar. Almost all of the Government's expenditures are of a recurrent nature, Le., spent on day-to-day expenses o f running the Government. In FY02, recurrent expenditures were 97 percent of the total, and accounted for about 20 percent of the GDP. In comparison, recurrent expenditures in Mainland Tanzania in FY02 were 80 percent of total expenditures and 15 percent o f GDP. If Mainland Tanzania's expenditures on police and defense in Zanzibar were to be added to Zanzibar's expenditures and subtracted from the Mainland's expenditures, the comparisonwould even be starker. A high level of recurrent expenditures has led to an almost complete squeezing out of development expenditures-thus, Zanzibar's development expenditures were less than three percent o f total spending in FY02. It is interestingto note that even as overall expenditures increased during the second half o f the 1990s, there was no increase in local development expenditures; in fact, during this period local development expenditures fell almost to zero. This is all the more surprising since, duringthis period, foreign assistancewas significantly reducedand it seems that there was an opportunity for domestic spending to substitute for foreignassistance. Table 7.1: Government Expendituresas a YOof GDP FY92 FY93 FY94 FY95 FY96 FY97 FY98 FY99 FYOO FYOl Recurrent Expenditures - - - - - - - - - - 36.3 28.3 15.0 15.3 19.1 19.8 25.3 20.3 32.3 18.4 Interest on Debt 1.8 5.2 0.1 0.0 0.0 0.0 0.0 0.0 0.6 0.3 Wages, Salaries and Allowances 5.7 10.0 7.4 6.1 7.9 9.3 11.8 10.8 12.0 10.0 Goods and Services 26.8 11.5 6.4 8.1 9.7 9.4 12.4 8.7 18.8 7.0 Transfers and subsidies 1.8 1.7 1.2 1.1 1.5 1.0 1.1 0.8 0.9 1.1 Pensions 0.2 0.2 0.1 0.2 0.3 0.3 0.3 0.3 0.3 0.4 Gratuities 0.6 0.7 0.5 0.4 0.6 0.3 0.2 0.1 0.3 0.2 Other Households 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Local Government 0.7 0.5 0.4 0.3 0.4 0.3 0.3 0.2 0.2 0.2 Non-Profit-MakingBodies 0.3 0.3 0.2 0.1 0.2 0.1 0.2 0.2 0.2 0.2 Development Expenditures 3.9 2.4 1.3 1.7 1.1 0.6 0.6 0.3 0.4 0.4 (Local) TOTAL EXPENDITURES 40.2 30.7 16.3 17.0 20.2 20.4 25.9 20.6 32.1 18.8 Source: Ministry of Finance and Economic Affairs, Zanzibar Subventions, Contributions, and Subsidies 7.3 Expenditures on subsidies and transfers were significantly reduced during the first half of the 1990sand have since remainedat a fairly low and stable levelof around 1percent o f GDP. There are practically no transfers to parastatals shown in the budget. However, there are hiddentransfers, as capital investmentfor many o f the parastatals is financed from the government budget. In addition, instead of receiving direct subsidies from the Government, parastatals have access to credit from the PBZ. In the case of insolvent parastatals, this contributesto the total of non-performingloans in the Bank's portfolio and increases the Government's contingent liabilities. The main recipients o f subventions are local government bodies, the Zanzibar Revenue Board, the Zanzibar Free Port Authority, the Zanzibar Investment Promotion Agency, and other non-profit bodies, such as the Wakf commission, the central library, the Instituteof FinancialAdministration, etc. Table 7.2 lists Zanzibar PublicExDenditureReview Page 54 selected subventions that have been provided in recent years. It should be notedthat, while the level of cash subsidies is low, there are sizable in-kind subsidies, mainly in the form of subsidized water and power supply. Table 7.2: Selected Subventionsin RecentYears (currentT.Shs. Million) FY98 FY99 FYOO FYOl FY02 FY03 Actual Actual Actual Actual Approved Budget ZIPA 47 56 56 75 1I 9 180 ZIFA 169 32 35 121 73 155 General Social Welfare Fund 118 1I 3 118 96 141 194 Zanzibar Municipal Council 233 227 241 289 343 298 Chake, Wete and Mkoani Town Councils 132 152 127 202 204 179 Zanzibar RevenueBoard 0 0 588 834 506 500 ZSSF 0 59 44 36 48 48 Zanzibar FreePortAuthority 0 0 55 76 104 150 Source: Draft budgetestimates for various years, Ministry o f Financeand EconomicAffairs, Zanzibar Debt Sewicing 7.4 Inmost years since FY94, Zanzibar has not paidany intereston its external debt. Nor were any expenditures made betweenFY96 and FY98 on servicing domestic debt. Table 7.3 providesa summary o f domestic debt servicingsince FY98. Quite clearly, it has beenerratic, and was the lowest spendingpriority o f the Government until recently, when the Government started devotingmore resourcesto it. The Government can affordto treat debt servicingas a discretionary expendituremainlybecausethe entire domestic debt is owned by parastatals. Table7.3: Summary ofDomestic DebtServicing (T.Shs. Million) FY98 FY99 FYOO FYOl FY02 FY03 Budget Actual Actual Actual Actual Approved Interest 0 50 1068 591 993 1333 DebtRedemption 0 198 500 0 4000 1444 Total 0 248 1568 591 4993 2777 Source: Draft budgetestimatesfor various years, Ministry o f Financeand EconomicAffairs, Zanzibar iv) Expendituresby Functional Classification 7.5 Table 7.4 shows the trends in central government expenditures by purpose as a percentage of GDP over the decade ofthe 1990s. The numbers depict a major transformation in the structure o f government spending during this period. Noticeable trends include fluctuating but generally increasing expenditures on general public services, and falling expenditures on health, agriculture, transport and communication. In fact, almost all of the increase in expenditures during the second half of the 1990s occurred in general administration. Froman economic point ofview, the expenditurestructure has changed from more productiveto less productive activities. The challenge before the current Government is to restore expenditures on social and economic sectors while cutting down the size of the general administration. Zanzibar PublicExpenditureReview Page 55 Table 7.4: GovernmentExpendituresby Purpose(% of GDP) Expenditures FY92 FY93 FY94 FY95 FY96 FY97 FY98 FY99 FYOO FYOl General Public Service General Administration 10.8 9.1 6.1 8.0 6.7 10.7 15.4 8.3 18.8 8.9 Public order 2.6 2.2 1.6 1.4 3.8 1.9 2.3 2.5 2.6 2.2 Sub-Total 13.5 11.3 7.7 9.4 10.5 12.7 17.7 10.8 21.4 11.0 Education 3.1 3.5 2.5 2.0 2.7 2.6 2.8 2.6 2.8 2.7 Health 2.3 2.3 1.4 1.3 1.8 1.6 1.7 1.6 1.3 1.3 Social Security Housing and Community Services 0.8 1.1 0.7 0.6 0.7 0.4 0.5 0.5 0.7 0.8 Recreational and Cultural Services 3.8 1.3 0.6 0.4 0.5 0.5 0.4 2.6 2.8 0.2 Agriculture 3.0 2.2 1.2 1.2 1.3 1.o 1.o 1.1 1.O 1.O Manufacturing and Trade 0.2 0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.1 Fuel and Energy 0.1 0.1 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.0 Transport and Communication 3.3 2.2 1.O 1.o 1.4 0.7 0.8 0.6 0.9 0.8 Other Economic Services 0.3 0.4 0.3 0.3 0.2 0.2 0.2 0.2 0.4 0.1 Sub-Total 16.9 13.4 7.9 7.0 8.8 7.1 7.7 9.4 10.3 6.9 Other Purpose 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Public Debt 1.8 5.2 0.1 0.0 0.0 0.0 0.0 0.0 0.6 0.3 Purchase of Financial Assets 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Pensions and Gratuities 0.8 0.8 0.6 0.7 0.9 0.6 0.5 0.4 0.5 0.6 Sub-Total 2.6 6.0 0.7 0.7 0.9 0.6 0.5 0.4 1.1 0.9 Tot. Rec. and Dev. Expenditures 33.0 30.7 16.3 17.0 20.2 20.4 25.9 20.6 32.7 18.8 Source: Ministry of Finance and Economic Affairs, Zanzibar Key Issues for Attention Expenditures on Salaries andAllowances 7.6 A large part of the recurrent expenditures is claimedby salary and allowances for an overstaffed civil service. In FY02, expenditures on salaries and allowances accounted for more than 60 percent o f the recurrent expenditures compared to about 40 percent in 1995. Wages and salaries expenditureshad steadily grown inthe second half of the 1990s and have continuedgrowing over the past two years, mainly due to increases in allowances, increased employment of teachers, and salary increases for special departments. Compared to Zanzibar, Mainland Tanzania spends about 35 percent o f its recurrentbudget on salaries. A large wage bill makes expendituremanagement difficult on two counts. First, given limited resources, it forces cuts in operations and maintenance expenses, thereby adversely affecting the productivity of existing public assets and service delivery. Second, it makes overall demand for public funds very inflexible. This is a problem when revenues display high variability, since it is notpossibleto adjust expenditures on wage bills quickly. 7.7 Zanzibar now spends around 35-40 times more on the wage bill than on creating public assets or operating and maintainingthem. Such neglect of development expenditures can have serious consequences for growth and poverty reduction in Zanzibar. The Governmentrecognizes that there is an immediate need for higher developmentexpenditures, and, in fact, there has been an increase in amounts released to MDAs towards development expenditures this fiscal year. However, an effort to increase development expenditures without first tackling the wage bill is likely to move expenditures further away from operations and maintenance spending. There is an urgent needto create space inthe budget for development expendituresby first controllingthe wage bill and then cutting it down over the medium term. Chapter 8, dealing with civil service, goes into the details of how this couldbe done. Zanzibar PublicExpenditureReview Page 56 Volatility of AggregateExpenditures 7.8 Aggregate expenditures in Zanzibar mirror the high degree of volatility in revenues which, in turn, are highly susceptible interms o ftrade shocks. High volatility inexpenditures results in disruption in public investmentand has associated costs of adjustment. The long- term solution to the problem o f resource volatility is a diversifiedtax base. In the short- to medium-term the key challenge, however, is to de-link government consumption from the volatility of the resource stream. In theory, this can be done if aggregate government expenditures are kept at a long-run sustainable level, while Government saves duringtimes of plenty and withdraws from savings in difficult times. Based on this rationale, some countries (for example Chile, Bolivia) have set up revenue stabilization funds that act as a buffer between revenues and expenditures. These funds also support fiscal discipline and provide greater transparency in the spending of revenue. There are two main options that Zanzibar couldconsider: 0 Creating an accounting-onlydesign, a `virtual fund' in which certain resources (for example, transfers from the Mainland or dividend payments from ZSTC) would be considered as belonging to the fund. These resources could be held in the main account of the Government or possibly in a separate account with the BOT. Restrictionswould be placedon drawing down the fund's resources for expenditures. Any draw-down of government deposits, includingthe fund account would appear as deficit financing. Transfers to and from the fund could be explicit line items in the budget. 0 Creatingan extra-budgetaryfund as a separate entity with authority to undertake off- budget expenditures. 7.9 The first option has the advantage that it would fit inwell with the existingbudgetary procedures and would not need the creationof new institutionalmechanisms. Such a revenue stabilization fund should serve to smoothen short-term volatility. However, given the difficulty o f distinguishingbetween short term volatility and secular revenue changes and the likely capacity constraints for administering such a fund, taking a more conservative approach to estimatingrevenues might be the most straight-forwardway to deal with revenue variability. Zanzibar PublicExpenditureReview Page 57 Box 7.1: Experiencewith Stabilization Funds Several countries with non-renewable resources like oil and other commodities have experimented with stabilization and savings funds. The basic objective has been to insulate domestic expenditures from the volatility of resource prices. An additional objective has been to save revenues from non-renewableresources for future generations. Stabilization funds work on accumulation and draw-down rules that lay down conditions when revenues would be deposited and withdrawn from the fund. These rules could be price-contingent (for example related to the price o f the resource commodity) or resource-contingent. Typically a reference level for the price or resource would be fixed that would trigger the deposit or draw-down. The accumulation o f assets in the fund may be subject to a cap determining the fund's maximum size. Since resources are fungible, achievement o f actual expenditure smoothing requires additional fiscal policy decisionsto control expenditures or deficits outside the fund. Otherwisethe Government could borrow or draw down its other assets outside the fund. Empirical evidence (based on a study of 12 countries) suggests two general results. First, some countries with a stabilization fund had a more limited expenditure reaction to changes in resource revenues than those without such funds. Second, the relationship between government spending and resource export earnings was not affected by the establishmentof funds in four cases. Source: Stabilization and Savings funds for nonrenewableresources,IMF 2001 Fiscal Risks 7.10 Fiscal risks refer to unaccounted obligations/liabilities of the Government which are not providedfor in the budget, but can arise over the year, thereby affecting aggregate fiscal discipline. The Government handles unplanned liabilities by keeping aside a part of the budget as a reserve within the MoFEA under a sub-vote called "Government Unspecified Expenditure." Table 7.5 shows expenditures from this account in the past few years. Details of specific activitiesfinancedunder this headwere not available. Table 7.5: Unspecified Government Expenditures (Million T.Shs.) FY98 FY99 FYOO FYOl FY02 FY03 Actual Actual Actual Actual Approved Budgeted 2,005 2,87 1 3,297 6,400 5,355 2,044 Source: Draft budget estimates for various years, Ministry of Finance and Economic Affairs, Zanzibar 7.11 The reserve is used for two categories of expenditures: (i)Pure contingencies, which are unexpected needs and emergencies, such as a cholera epidemic or a natural disaster; and (ii) quasi-contingencies, whose ex-ante probabilityof arising can be anticipated. The quasi- contingencies may arise during the mid-year due to new policy decisions, or may be liabilities that had already accrued at the time o f budget-making,but were not budgeted for due to inadequate informationabout them, or mayjust be contingent liabilities. Examples o f quasi-contingencies are mid-year salary increases, arrears, gratuity payments due to retired government/parastatal staff, and court awards. While the Government cannot anticipate pure contingencies, it must make every attempt to quantify and budget for quasi-contingencies because expenditures on unplanned quasi-contingencies disrupt other planned priority expenditures and are a symptom of a weak budgeting system. The following paragraphs describe some liabilities of the Zanzibar Governmentthat pose fiscal risks to the government finances and needwatching. Zanzibar PublicExpenditureReview Page 58 Arrears and other Accrued Liabilities 7.12 The Government has run up significant arrears in the past years, which do not show up in the government debt accounts (since they are not formally contracted) eventhoughthey are government liabilities. Major among these are outstanding government contributionsto ZSSF (T.Shs. 1.8 billion) and cleanup liabilities of collapsed parastatals (T.Shs. 2.9 billion owed to PBZ and 271 million to retrenchedemployees). The Water Departmentowes T.Shs. 2 billion to the Fuel and Power Corporationtowards purchase of power for runningits water supply pumps. Apart from these known arrears, ministriesalso acknowledgedarrears in the nature of unpaid bills of supplies already received, but their total volume could not be ascertained. Outstandingdues ofZSF&PC to TANESCO are T.Shs. 34 billion. 7.13 Arrears and accrued liabilities are now receiving Governmentattention. Outstanding dues o f ZSF&PC to TANESCO are beingcleared in a phased manner. While the clearingo f old dues is no doubt the first priority, it may not be sufficient. The Government will needto improve cost recovery in the water and power departments. In the water department, fresh dues are accumulatingat the rate of T.Shs. 100million per monthdue to poor cost recovery.6 Arrears also make government finances less transparent. Putting in place information systems that collect and report budget execution in a timely manner would be an important requirement (for example reporting expenditures at commitment stage). The Country Financial Accountability Assessment (CFAA) makes specific recommendations in this regard. ContingentLiabilities 7.14 Contingent liabilities are those that become government obligations,if a particular event occurs. A significant source of such liabilities could be Zanzibar Social Security Fund (ZSSF). In 1998, the Government established ZSSF, whose function is to (1) receive and manage social security contributions from its members, and (2) pay old age, maternity, invalidity, and medical care benefitsto them. All government employees, includingthose of the parastatals and localauthorities, and privateemployees canbecome membersofthe Fund. Accordingto the Fund provisions, the employer is expected to contribute 10 percent against the employee contribution of 5 percent of monthly salary to the Fund. However, the Government had been delinquent in paying its contributions in the past and still owes T.Shs. 1.8 billionto the Fund. Apart from this, many parastatals, some ofthem that are either closed or are beingclosed, have yet to deposit their contributionwith the Fund. The Fundcouldnot confirm the amount of outstandingdues from such parastatals, but they are considered, for all purposes, to be government liability. Contributions from municipalbodies are also stated to be irregular. 7.15 Fiscal risks associated with the functioningof the Fund appear to be significant. The Fund is expected to start payingpensions and other service benefitsto government employees who retire after June 2003. But the Fund would have received contributions for only five years o f service o f such employees; the Government would still be liable for providing benefits proportionalto the periodof service in excess o f five years. A broader issue is that it i s not entirely clear that the Fund will remain fully funded in the long term. Apparently no actuarialcalculationswere carriedout beforethe Fundwas institutedto arrive at the package Water is supplied free to residential consumers. The Water Department recovers about 12% of its costs. InFY03, at the end of November, the department's revenue was T.Shs. 49.2 against running cost of 382 million. Power tariffs are quite low too. Against acost of T.Shs. 130iunit only about T.Shs. 28iunit are recovered. Power generation is very costly in Pemba, where it is produced from diesel generators. Zanzibar Public ExpenditureReview Page 59 o f benefitsand contributions. The Fundalso needs expert advice on where and how to invest its funds. It has put about T.Sh. 1.8 billion in real estate and is considering investing in foreign currencies, without a carefulness assessment of the risks involved in such investments. 7.16 Other contingent liabilities could arise from parastatal debt and government guarantees. The financialpositionofthe People's Bank o fZanzibar is not sound. Apart from T.Shs. 7.2 owed by the Government to PBZ, other parastatals owe about 2.7 billion to PBZ. Zanzibar Import Corporation is the main borrower, which is under liquidation now; upon liquidation its liabilities to PBZ are likely to fall upon the Government. Information on guarantees provided by the Government could not be obtained, but apparently guarantees have been provided for parastatal borrowings in the past that could become government liabilities. EXPENDITURESPRIORITYSECTORS ON 7.17 The ZPRP is expected to provide the framework for sectoral allocation of expenditures. A budget that is well-alignedwith, and supports, the ZPRP, should be able to provide adequate financing in medium-termto priority areas identified in the ZPRP. The ZPRP has identified education, health, road infrastructure, agriculture, water, democratic governance and rule of law, and combating HIV/AIDS as priority areas for government financing. Since the ZPRP has not yet identified specific programs and projects in each of these priority areas, alignment can be monitored only at the level o f sectors and ministries. Again at this early stage of implementation,resource requirements to meet various objectives in the priority areas have not been worked out. Therefore the targeted structure of public spending by sectors is not clear, nor is the pace at which expenditures would adjust invarious sectors to reachthe targeted levels. Therefore, ensuringalignment with the ZPRP boils down to seeing iffinances availableto various ministries overseeing priority areas are goingup and ifthose for non-priorityareas aregoingdownrelativetototalexpenditures. 7.18 As evident in Table 7.6, expenditures on priority sectors have been stagnating at around six percent o f GDP since FY97. This is a result o f increasinglyreduced share of the priority sectors in total government expenditures (Table 7.5). In order to better align expenditures with ZPRP priorities, the Government could both let the priority sectors' financing grow at a faster rate than other sectors and make real cuts in non-priority sector funding. Allocations for FY02 and FY03 show that expenditures in health and education ministrieshave gone up, which is the right directionfor adjustment. Table 7.6: Expenditure Shares by Sectors (YOof Total) Fiscal Year FY92 FY93 FY94 FY95 FY96 FY97 FY98 FY99 FYOO FYOl Education 10 11 15 12 13 13 11 13 9 14 Health 7 8 9 8 9 8 7 8 4 7 Agriculture 9 7 8 7 7 5 4 5 3 5 Transport and Communication 10 7 6 6 7 4 3 3 3 4 GeneralPublic Service 41 37 47 55 52 62 68 53 65 59 Other 24 30 14 12 13 9 7 19 16 11 Total 100 100 100 100 100 100 100 100 100 100 Source: Ministry o f Financeand Economic Affairs, Zanzibar Zanzibar Public ExpenditureReview Page 60 7.19 The Government is encouraged to identify expenditure items in line with ZPRP priorities,whichwould receivepriority duringbudget allocationand would be protectedfrom expenditure cuts during budget execution. Since resources available for non-wage expenditures are subject to considerable variation and uncertainty from year to year, such poverty-reducingexpenditures would have to be defined quite narrowly in order to make effective protection feasible. Since, in a system in which certain expendituresare protected, the burden of absorbing shortfalls in resources falls disproportionately on unprotected expenditures, this burden will increase the share o f protected expenditures. If too large a percentage of non-wage expenditures were defined as pro-poor, effective protection might not be feasible, since resource shortfalls might exceed the total amount o f unprotected expenditures andthus automaticallyalso impacton priority expenditures. Expenditureson Education 7.20 Zanzibar follows a 7-3-2-2-3+ system of education. Compulsory basic education includes seven years of primary education after the age of 6, and three years o f junior secondary ed~cation.~Basic education is followed by two years of second cycle lower secondary (Forms I11and IV), two years of third cycle advanced secondary study (Forms V and VI), and three years or more of higher education. Basic and secondary stages of educationthus encompass 14years o f schooling. 7.21 Access to basic educationhas slowly improvedin the last decade and so have various enrollment indicators (Table 7.7). Yet there are serious challenges ahead. Both gross enrollment ratios (GERs) and net enrollment ratios (NERs) are low, and a large number of school-age children do not yet attend school. The number of new school-age children is growing rapidly due to high fertility rates. Wastage rates are high, and the quality of education is generally not satisfactory. Access to secondary education i s still very limited and the GER is below 40 percent. There are significant regional disparities in education indicators. North A, South Unguja, and all districts in Pemba have below-averageGERs for basic education. Table 7.7: Participationand InternalEfficiencyinBasicEducation (YO) EnrollmentRates(Participation) Wastage Rates (Efficiency) 1990 2000 1990 2000 GER NER GER NER DR RR DR RR MF 65.7 43.4 85.4 54.7 7.2 8.9 7.3 4.9 M 67.4 43.6 87.1 54.6 7.8 9.7 6.8 5.8 F 63.9 43.1 83.6 54.6 6.5 8.1 7.7 3.8 FIM 0.95 0.98 0.96 1 0.86 0.84 1.13 0.67 Source: EducationSector Status ReportZanzibar. 2002. Legend: Gross EnrollmentRatio(GER), Net Enrollmentratio (NER), Drop-outrate (DR), RepetitionRate (RR), Male (M), Female(F), TotalMale and Female (MF), Female-Maleparity index (F/M) 7.22 The state of education infrastructure is poor. There is an acute shortage of class rooms-the average pupilklassroom ratio is 82 and the average number of shifts is 1.69 streams per class room. 'Thejuniorsecondaryeducationconsists of one year of orientation,andtwo years of grade study calledFormIand Form11. Zanzibar PublicExpenditureReview Page61 Aggregate Expenditures 7.23 In FY02, the Government spent about T.Shs. 9 billion on education, which was about 2.7 percent of the GDP, and about 15.5 percent o f total government expenditures (against the ZPRP target of 20 percent), a levelslightly higherthan MainlandTanzania and comparable to other countries inthe region. While, interms of GDP, aggregate educationexpenditures have stagnated at around 2.7 percent since FY95, they have been growing in real terms at a rate higher than population growth during this period. As a result, per capita annual public spending on education has been improvingand was around a reasonable figure of US$l0 in FY02, compared to about US$7 for MainlandTanzania (and US$10for Kenya in 1993, and US$7 for Ugandain 1997). Figure 7.1: Trends in Real Public Spendingon Education Trends in Real Public Spending on Education I 2000 , ~ .-21500 I1000 I 8 f FY FY FY FY FY FY FY FY FY FY 92 93 94 95 96 97 98 99 00 01 Fiscal Year 7.24 Aggregate expenditures on educationlook reasonable in a comparativesense, but one must allow for the fact that private provision of education beyond the pre-primary stage is very limited in Zanzibar, necessitatinga higher levelof public financing. Furthermore,there has been very little external financing since 1995, which means that localresourceshave had to meet the demand for development expenditures. This demand was not met and there i s a serious backlogof capital works that need to be carried out. For example, about 500 school rooms are incompletewhere the communitieshavemet their part butthe Government has not been able to meet its share. And finally, as noted above, education performance indicators still have large room for improvement. 7.25 Almost all educationexpenditures are recurrent in nature, and within recurrent, most are accounted for by salaries and allowances (93 percent in FY02). Development expenditures have stayed at less than one percent since 1995. Expenditures on "other charges" have beenvery limited too. InFYOl the Governmentspent only TShs. 7 million on school materials and equipment. In fact, school materials and teaching equipment are now expectedto be providedby the community. 7.26 All the factors discussed above point toward the need for a still higher level o f financing. The Governmentis well aware o f the financing gap in the sector and has tried to improve funding in the past two years. In nominal terms, the allocation for the Education Ministry increased by 48 percent in FY02 and a further increase o f 38 percent is proposed in the FY03 budget. The levelis now close to the ZPRP target o f a 20 percent budget share. Zanzibar PublicExpenditureReview Page62 Allocations within the Sector 7.27 Table 7.8 shows the shares o f various sub-sectors in the Ministry of Education's budget over the past three years. Information on separate shares o f basic and secondary education was not available, since these are funded from the same sub-vote. Table 7.8: Sharesin EducationExpenditures(YO) Sub-sector FYOO FYOl FY02 FY03 Actual Actual Approved Budget Basic and Secondary Education 88 86 67 71 Higher education, S&T 15 13 Adult Education 2 1 2 Administrative and non-education exmnses 11 12 17 14 Source: Draft budget estimates for various years, Ministry of Finance and Economic Affairs, Zanzibar 7.28 It appears that administrative and non-educationexpenses went up after FYOO. It is partly due to the addition of sports, art, archives, festivals, and culture portfolios to the education ministry and partly due to the real increase in administrative and non-education spending. Within Basic and Secondary education, the share o f salaries and allowances inthe budget is around 97 percent. The bulk of the increased expenditures in the previous two years has goneto salaries and allowances. Key Issuesfor Attention 7.29 As noted earlier, government financing of the education sector remains inadequate in view ofthe ZPRP objectives. Yet, at the same time, resources for educationhave been going up in realterms and per student. This increase inresourcesis not reflectedin much-improved performance indicators, which points towards the need for improving internal efficiency. Indeed, within the existing resource envelope, it is possibleto achieve better results through restructuringexpenditures. Non-Salary Expenses 7.30 Improvingthe quality o f education is a key priority in the ZPRP and this cannot be achieved unless more resources are made available for operational expenses. One way to achieve better results would be to curtail expenses on wages and salaries and make more resources available for non-wage expenditures. However,the current directiono f adjustment in education spending is in the opposite direction. Of all the additional resources made available in FY02, three-fourths went towards wage bill increases. There was very little increase in other expenses; in fact, the Government still spends negligible amounts on school supplies. Mainland Tanzania has a target of providing T.Shs. 10,000 towards non-salary allocations per pupil in primary education. The Government in Zanzibar should also aim to movetowards this levelo f allocation. Allocative Efficiency Issues 7.31 Government spending on educationappears to have inequities across districts and to be biased in favor of urban communities(Table 7.9). Inthe year 2000, almost 70 percent of the Ministry of Education's budget went to the two urban districts of Zanzibar city and Unguja West. There is also a serious imbalance in the number of classrooms between the rural and urban districts. There is a shortage of teachers in the rural areas, while the urban area schools have surplus teachers. Zanzibar PublicExpenditureReview Page 63 Table 7.9: EducationResources by District District Budget Pupil/ Pupils per share ( YO Teacher classroom of total) ratio (basic) Zanzibar Town 45.1 30.6 77 West 21.7 32.6 82 North A 5.9 46.7 94 North B 1.6 30.6 87 Central 3.6 26.6 73 South 3.1 27.8 42 Micheweni 1.6 29.6 96 Wete 10.3 31.6 92 Chake Chake 5.7 33.3 93 Mkoani 1.3 40.7 82 Source: Education Sector Status Report Zanzibar. 2002. 7.32 The Governmentpracticeof encouragingthe communities to cover some operational expenses for schools and parents to provide textbooks also puts poorer households at a disadvantage.' 7.33 There is a strong case for reviewingexpenditurepatterns on a geographicalbasis and reorienting it to become more equitable. From an economic point of view, spending more in communities that have poorer educationoutcomes would give higher returns at the margin. There is reason to believe that the regional distribution of resources will get even more skewed unless the Government gives it its full attention. For example, communities can collect part of the funds for the construction o f a school building and approach the Government to provide the balance. Once the building is in place, the Government is committed to opening a school in it. Such an approach would lead to a concentration of schools, and hence public expenditures, in communities that can afford to collect funds for new school buildings. Pupil/Teacher Ratio 7.34 The average pupil-teacher ratio is about 33:l in primary schools and about 30:l in secondary schools. The ratio for primary schools is very low compared to a targeted 45:l in the Mainland Tanzania. There is a considerable opportunity for expanding access to basic education by redistributing existing teachers more efficiently, without having to hire more teachers. Private Provision of Post-Primary Education 7.35 Private provision of post-primary education is limited at around 2.3 percent. The Government should encourage private provisionof post-primaryeducation, which will allow it to leverage public spending. Government aid to private institutions, subject to regulations, would bethe right approach. Expenditureson Health 7.36 Zanzibar has a substantialinfrastructurefor healthservices delivery, consistingof 103 Primary Health Care Units, four cottage hospitals, three other hospitals and one referral ' * Accordingto the Ministry of Education guidelines, parents are required to provide annual contributions at rates varying betweenT.Shs. 3000 for primary to T.Shs. 7,500 for Higher Secondary education. Zanzibar PublicExpenditureReview Page 64 hospital. According to the Government, this makes a public health facility available to almost each Zanzibari within a five kilometer distance. Private medical facilities have been growing, and there are about 67 on UngujaIslandand eight on Pemba. Apart from this, there are 38 paramilitary units which provide medical services to uniformed forces and their families and nearby communities. Despitethis impressivenetwork of facilities, healthsector outcomes are not very encouraging. The infant mortality rate is estimated at 75.3 per 1000, under-fivemortality is 107.6 per 100, and the maternalmortality rate is 314 per 100,000 live births. Malariais rampant, and about 340,000 cases were reportedduring 1995. Figure 7.2: Trends inRealPublicSpendingon Health(1991 prices) Trends in real public spending on health I (1991 prices) 1200 i .swe 2 1000 Million) , 800 600 400 I B 200 0-1 , , Fiscal Year i Aggregate Expenditures 7.37 Aggregate public spendingon health sector declined between FY95 and FYOl-both relative to GDP (from 1.8 to 1.3 percent) and as a share of total government expenditures (from 8.8 to 6.8 percent). The question is whether Zanzibar is spending enough on public health. If not, to what level should the spending grow? These questions would be best answered once desired health outcomes in the ZPRP are appropriately costed out. Comparisons across countries are not very helphl, since there are large variations across countries concerning how much, and equally importantly, how they spend on health. The closest comparison is MainlandTanzania, where health expenditures were 0.6 percent of the GDP, and more than nine percent of the total discretionary spending in FY02, excluding development expendituresseparately financedby donors. 7.38 FY02 saw a significant increase in the health budget, and spending improvedto about 1.9 percent o f GDP and 10 percent of total expenditures.' The FY03 budget has plans for further increases in the health budget. It should be noted that despite real growth in recent years, health sector spending is below the levels achieved in early 1990s. The sector shows tell-tale signs of chronic under-funding. Healthfacilities are in a state of disrepair, equipment i s scarce, and there i s an acute shortage o f drugs and medicalsupplies. 7.39 As in education, the structure of spending in the health sector is skewed in the direction o f salaries and allowances. In the past several years before FYO1, salaries and allowances have typically claimed the bulk of the health services budget (80-85 percent), leaving little room for operational expenses. Development expenditures from government resourcestoo have beennegligible; in fact, they were zero inFYOO and FYOl. Social welfare department was added to the health ministryportfolio inFY02, which also led to an increase in the health ministry vote size. Zanzibar Public Expenditure Review Page 65 Allocations within the sector 7.40 Table 7.10 shows the shares of different components of expenditures within the health sector in recent years. Table 7.10: ExpenditureShares within Health ( O hof Total Health Expenditure) FY98 FY99 FYOO FYOl FY02 FY03 Actual Actual Actual Actual Approved Budget Administration 79.0 80.0 89.0 89.0 78.0 81.0 Head Office 53.0 54.0 61.0 56.0 50.0 56.0 Pemba 25.0 26.0 27.0 33.0 28.0 25.0 Preventive Health Services 1.9 3.2 2.7 1.o 6.8 4.3 Unguja 1.3 1.5 0.8 0.8 2.9 2.5 Pemba 0.6 1.7 1.9 0.3 3.9 1.8 Curative Health Services 20.0 17.0 9.0 10.0 16.0 14.0 Unguja 17.0 16.0 9.0 9.0 8.0 12.0 Pemba 3.0 1.o 0.0 1.o 7.0 2.0 Total 100.0 100.0 100.0 100.0 100.0 100.0 Source: Draft budget estimates for various years, Ministry of Finance and Economic Affairs, Zanzibar 7.41 Several comments are inorder: 1, Administrative expenses (which include salaries and allowances) far outweigh the expenses on provision of health services. 2. Preventive services receive a very small share of the non-wage resources. 3. The expenditures do not appear to be equitable. In regional terms, Pemba, which has about 40 percent of the total population and lags Unguja in social indicators, has been receiving less than 30 percent of the total budget. Pemba's share of non-wage resources for curative services is even lower. Considering that Unguja has much larger private provision of health services, the situation appears to be even more unfair. Key Issuesfor Attention 7.42 Key issues include the following: 1. There is a need for an increased overall level of resources in the health sector. The ZPRP also targets a higher level of allocation to health. 2. There is an opportunity for improving intrasectoralallocations by: a. Reducing administrative expenses. The expenditures on salaries and allowances are too high. b. Giving more attention to preventive care. Experience from other countries suggests that vertical `basic package' approaches like antimalaria programs could be more effective in achieving high impact with limited resources. The vaccination program has also suffered from the withdrawal of donor assistance and needs an infusion of resources. c. Spending a greater share on drugs and medical supplies. At present there is an acute shortage of drugs in PHCs (primary health care units) and hospitals. Absence of drugs hurtsthe poorest the most. ZanzibarPublicExpenditureReview Page66 ISSUES INBUDGET MANAGEMENT Budget Outcomes 7.43 An effective system for budget management is one that is able to produce the following three outcomes: 1. Enforce aggregate fiscal discipline over a medium term, restricting budget totals to explicitly decided levels; 2. Allocate resources according to government priorities and effectivenessof programs inachievingdevelopment goals; and 3 , Ensureoperationalefficiency and effectiveness. 7.44 These outcomes result from careful budget preparation, execution and monitoring. Weak budgetary processes lead to a loss of control over public spending, fail to channel spending into areas that are of social and economic importance,and fail to provide the right incentivesfor productiveefficiency in government operations. 7.45 Judged against the desired outcomes mentioned above, performanceof the budgetary systems leaves much to be desired in Zanzibar. There is a perceptionthat aggregate fiscal discipline is being maintained because the Government operates a cash budget, but the perception o f fiscal discipline is illusory. The Government has sizable arrears, has been neglectingdebt service, and has beenreluctantto make agreed uponcontributions to the cost o f the Union Government. On the second outcome, related to efficient allocation o f resources, some progress has been made over the last two years. The ZPRF' now provides a framework for prioritizing government expenditures, and, as discussed in previous sections, there is an attempt to align broad spending patterns (at the level of ministries) with ZPRP priorities. However,throughdiscretionary releasesof cash duringthe year, MoFEA has been significantly deviatingfrom ex-ante allocationpriorities. Performanceon the last outcome is the most disturbing. The severe resource crunch in the past years has made the flow of funds to the spending ministries completely uncertain, which in turn has severely impacted their operationalefficiency. Budget Estimates vs. Actual Outturns 7.46 A comparison of actual outturns with budget estimates is a good indicator of the overall performance of a budget management system. In a well functioning budget system, the budget provides a realistic estimate of resource availability, which ensures that allocated expenditures can indeed be funded. The credibility and effectiveness of the budget as an instrument for allocating resources in a strategic and planned manner depends on the closeness ofactualexpendituresat the end ofthe fiscal year to the ones plannedat the time of budget formulation. On the other hand, a wide deviation between estimates and actuals is a sign of one, or many, of the following weaknesses: (i)inaccuraterevenue forecasting-most often the revenue is overestimated; (ii) inaccurate expenditure estimates-most often, these are inflated in line with the overestimated revenue; (iii)changes in allocation priorities midway during the year, which results in overspending in some and underspendingin other categories o fexpenditure;(iv) inefficientcash flow management. 7.47 Significant deviations between resources budgeted and actually made available have serious costs. First, they erode the overall credibility of the budgeting process, thereby providing disincentives for MDAs to take expenditure estimations seriously. Second, they make it difficult for the spending units to plan and rationally implement their expenditure Zanzibar PublicExpenditureReview Page 67 programs. Third, they may lead to unintended reordering of priorities during budget execution, thereby leading to the abandonment of the plan and an abrogation of the legislature'ssanction. 7.48 Table 7.11 below shows the average difference between budgeted and actual expenditures inthe past three years in Zanzibar. The deviationsare o fthe order o f 80 percent for development expenditures, and between 40 and 70 percent for other charges. These numbers underscore significantly the uncertainty faced by the MDAs regarding the availabilityo f funds, other thanfor salaries. Table 7.11: DifferencesbetweenBudgetedandActualExpenditures FYOO FYOl FY02 Average Average Average Average Average Average absolute absolute absolute absolute absolute absolute difference difference difference difference difference difference (TShs. as % o f (T.Shs. as % of (T.Shs. as % of Mill.) budgeted Mill.) budgeted Mill.) budgeted Salary and 88 19 96 20 278 22 Wages Development 243 79 547 80 663 77 Other Charges 781 40 464 70 412 56 Source: Draft budgetestimates for various years, Ministry of Financeand Economic Affairs, Zanzibar Note: Averagestaken over all votes 7.49 Table 7.12 below lists the major gainers and losers from the reordering o f priorities duringbudget executionin FY02. Table7.12: Major Gainers and Losers duringBudgetExecutionin FY02 Vote Category of Budget Actual expenditure (Million (Million TShs.) TShs.) Major gainers President'sOffice, Regular .Administration Other Charges 510 989 Ministry of Communication Salary etc. 1145 1631 Chief Minister's Office Salary etc. 723 846 Ministry o f Water, Environment,etc. Salary etc. 1125 1183 Ministry o f Agriculture & Livestock Salary etc. 1908 2341 Major losers Ministry of Health Other Charges 2017 1401 Economic Brigade Salary etc. 2909 1527 Ministry o fEducation Other Charges 1623 623 Ministry of Communication Dev. (local) 2500 638 Ministry of Communication Other Charges 1321 317 Ministry of Education Dev. (local) 1100 184 Ministry of Water, Environment etc. Dev. (local) 690 0 Ministry of State Women & Children Dev. (local) 893 0 Total budget 66,333 49,763 Source: Ministry of Finance and EconomicAffairs, Zanzibar Key Issuesfor Attention 7.50 Zanzibar follows a fiscal calendar similar to MainlandTanzania. It has a dual budget system in which recurrent and development budgets are prepared independently. The budget Zanzibar PublicExpenditureReview Page68 preparationbegins in mid-November o f each year and is coordinated by the MoFEA. The MoFEA develops revenue estimates, establishes budget ceilings for each of the ministries, and issues budget guidelines in January. In the following months, Ministries submit their estimates, which are finalized following negotiationswith the MoFEA. The CFAA provides further details about the processofbudget preparation. Budget Planning and Resource Allocation BudgetRealism 7.51 The revenue and expenditure estimates are not realistic, as evident from the wide deviationsbetween budgeted and actual amounts. Revenue receipts are quite volatile due to heavy dependence o f the economy on a single export commodity. This feature of the economy cannot be changed in the short-term. There are, however, ways in which the expenditure estimates can be made more realistic by bringingmore informationto bear upon the estimation and allocation process and by creating the right incentives for the MDAs to estimate expenses realistically. The following issues require particular attention during the budget preparationprocess. 7.52 The MoFEA does not generate a macro-fiscal framework that would describe the fiscal environment and therefore broad parameters within which the budget would be prepared. A macro-fiscal framework would allow the MoFEA to estimate revenues, set overall expendituretargets, and work out borrowingrequirements, if any, more realistically. It would also help MDAs develop more realistic estimates of expenditures if the budget guidelinesconvey budget implicationso f expectedmacroeconomic developments such as the inflation rate, growth rate, and any significant demographic parameters for different categories of expenditures. Development o f a successful macro-fiscalframework, however, would depend partially upon how well resource requirements of the ZPRP are quantified (costedout). 7.53 The MoFEA does not have good estimates of actual outturns and arrears in the previous year and works with information on cash releases made in the previous year. In a well-functioningbudget system, informationon budget executionwould become available in a timely manner through routine reporting and monitoring, but this does not happen in Zanzibar. In short-termthe MoFEA will need to make special efforts to collect information on actual outturns, and more importantlyon arrears, before initiatingthe budgetprocess. 7.54 MDAs developexpenditureestimates on the basis o f the previousyear's estimates of input costs rather than discussingexpenditureneeds in a policy framework. Discussionswith the MoFEA on expenditurecuts are also in terms o f input costs, such as travel expenses, fuel for vehicles etc. This focus creates an implicit prioritization in terms of inputs rather than in terms ofpolicies andprograms. 7.55 The budget guidelines are an important instrument for fostering greater realism in expenditure estimates. In view of the discussion above, the budget guidelines should specifically: 0 Describe the fiscal environment, includingthe broadparameters within whichthe next budget must be prepared. Examples includethe forecast level of revenues, indicative aggregate expenditure constraints, expected growth rate, expected inflation rate for expenditure projections,demographic trends, etc.; Zanzibar PublicExpenditureReview Page 69 0 IndicateGovernment'sfiscal prioritiesfor the year; 0 Provide guidance on areas in which spending agencies might seek to identify efficiency gains or implement policy initiatives, along with overall government employment and procurement policies with which ministries are expected to comply (examples are hiring freezes, management of vehicle fleets, building maintenance, contractingout etc.); 0 Seek submission of new policy proposals separately from existing ones. The key questionconcerns how new policieswouldbe funded; 0 Seek realistic estimates from the ministries in terms of spending programs and policies; and 0 Ask ministriesto indicatethe programs they would suspend if faced with a resource deficit of, say, 10 percent. 7.56 The Government should budget for quasi-contingencies,arrears, pensiondgratuities, obligations arising from policy decisions, and contingentliabilities in a systematic way. This would require systematic collection o f information on (i)arrears and (ii)outstanding liabilities becauseof pensions and gratuitiesto government and parastatal staff. A schedule of payments to ZSSF for pension benefits in excess of five years should be drawn. In order to plan for contingent liabilities, MoFEA should complete the actuarial evaluation of ZSSF, quantify government risks on guarantees, and estimate payments that might fall on the Government becauseo fthe debt of collapsed parastatals. Integration of Aid into Budgeting 7.57 At present, recurrent and development budgets are prepared through separate processes of formulationand review o festimates. As in many other countries, this separation has historical and institutional reasons. It stems from the intuitive idea that the recurrent budget is meant for financingongoingactivitieswith Government's own resources; while the development budget is for financing "economic development"-mainly in the form of capital expenditures with donor support. The development budget was earlier prepared by the Ministry o f Planning,now mergedwith the MoFEA, while the recurrent budget was prepared by the MoF. There are three mainproblems with this separation of processes: (i) Allocation decisions on the development and revenue budgets are made without reference to each other. This can lead to inconsistenciesbetweenthe two budgets. Most common are cases in which recurrent expenses are not provided in the recurrentbudget for new capitalprojects,or the same itemof expenditureis included inthe two budgets. (ii) There is insufficient scrutiny of investment projects included in the development budget, becausethey are largely donor-financed. (iii) The development budget may often be assigneda priority status and lead to cutbacks on operationalexpenses for existingprograms. 7.58 In view of these problems, the Government should move towards unifying the two budget processes. While the ultimate objective should be to merge the two budgets completely, it may not be achievable in the short- or medium-term. A middle-of-the-road approach would be to establish one ceiling for each spending agency, covering both categories, asking for one set o f estimates, and reviewing one set of estimates. The experience from the Mainland suggests that Government may not really end up making competitiveallocationdecisions, particularlyregardingdonor supported projects,throughthis "unified process," because those decisions are made outside the budget process, but the Zanzibar Public ExPenditureReview Pane 70 unification will serve an important purpose o f integratingall budget estimates by sectors in one place and allow better allocation of government's own resources, since resources are fungible. 7.59 An importantprerequisitefor this unificationo f processes would be that MoFEA has up to date information on all donor-supportedprojects, both existing and planned. This is easier said than done and needs the cooperation o f all concerned donors. To begin with, MoFEA should start collecting information from donors on expected disbursements during the next fiscal year in a format similar to that usedby MainlandTanzania. Multi-Year Framework of Expenditures 7.60 The Government plans to move to a multi-year framework for expenditures (the Medium-TermExpenditureFramework,or MTEF). Three major advantages would accrue in movingto a multi-yearframework: 0 It would force attention to the long-runexpenditureimplications o f new policies and investmentprojects. 0 It would help in programmingcutbacks (say, phasingout programs, or cutting down the wage bill) over a periodo ftime. 0 At the aggregate level, it would provide MoFEA with a medium-termview of inter- sectoral demands for resources and allow planning in order to align the structure of expenditures with the ZPRP prioritiesinthe mediumterm. 7.61 Based on cross-countryexperience, several prerequisitesneedto be put in place for a successful introductionof the medium-termframework. Some o f these have been discussed above. In the absence of a macro-fiscalframework and cost estimates o f implementingthe ZPRP, MDAs will not be able to develop meaningful medium-term plans. Further, such plans will be of limitedvalue ifthey leave out donor-providedfunds. Most important,budget processes currently lead to wide divergences between voted expenditures and actual outturns. These divergences have resulted in a serious loss o f credibility for the budget as an instrument of strategic allocations, and they create major disincentives for the MDAs to invest in generating medium-term plans. The divergences result both from an unrealistic budget and ineffective budget execution. Finally, there are serious capacity constraints in both MoFEA and the MDAs, and focus on generating medium-termplans would overstretch their capacity. 7.62 Inview ofthis discussion, while a medium-termexpenditureframeworkwill be a step in the right direction, it should be a medium-term goal to follow careful strengthening of budget processes. In the fiscal year 2004-05, it could be piloted in three ministries, for example, education, health, and transport and communication, which have better capacityfor planning. Budget Execution 7.63 The Country FinancialAccountabilityAssessment (CFAA) has a detailedassessment of budget execution in Zanzibar, including issues of cash flow management. Following the establishment o f a Ceiling Committee in FY03, the flow of hnds to the MDAs appears to have become more certain. From a fiscal point of view, the following issues are important and deserve Government attention: Zanzibar PublicExpenditureReview Page71 e Monitoring budget execution: The Government should move toward regular monitoringand publishingat least quarterlyreports on actualexpenditures, rather than relyingon cash release statements. Controlling arrears: Arrears can upset financial plans and be a source of fiscal indiscipline. One of the mainreasons for arrears, inthe decentralizedpayment system o f Zanzibar, is a lack of effective MoFEA monitoring and control of commitments made by MDAs. There is a need to institute either a mechanical system for such control, or, if a decision is made to go in for an automated Integrated Financial Management Information System, then such control can be effectively exercised through that system. TransparencyIssues 7.64 Enhancing transparency at all stages of the budget process, including budget formulation and resource implementation, budget implementation, and budget reporting is one o f the key challenges for Zanzibar. Developing more open and inclusive budget processeswill play a key role in strengthening strategic resourceallocationand enhancingthe efficiency of resource use. At present, there are many examples of practices that raise concerns about transparency. These include e The expenditures from the vote "unspecified government expenditures" have been fairly large and non-transparent. Ostensibly meant for meeting truly unforeseen situations, such as naturaldisasters and epidemics, there are no clearly laid out criteria definingeligible contingenciesfor withdrawal. The contingencyreserve should meet only truly unforeseen obligations. The Government should plan to meet quasi- contingencies throughregularvotes. e In the past, large sums o f money have been spent on subventions titled "Special PresidentialVote." An expenditureof T.Shs. 7.764 billion in 98/99,and 5.168 billion in 00/01 was made on this account. The Government has reportedlydropped this sub- vote from FY02 on. e The MoFEA issues lump-sum monies to the ministriesfor wages and salaries. There have been instances when ministries have diverted them for other purposes. A civil service census as suggested inthe Civil Services chapter of this report shouldbe able to provideMinistry of Financewith better control. e The MoFEA operates a Micro-ProjectsFinance scheme which provides government support for community projects. There are problems of coordinating plans for new schools or hospital constructionwith the concerned ministry. Operationof this fund requires more transparency. The proposals for community projects should also be cleared with the concerned sector ministry to ensure that they are in line with their existing plans. The MoFEA should also keep a record of how many community resourceswere mobilized. 7.65 In addition to the above, previous sections have listed certain areas in which lack o f information (for example, on arrears and accrued liabilities) makes the true fiscal position opaque both to the MoFEA and to internal and external stakeholders. In order to improve transparency, the Governmentshould consider preparinga statement on government finances each year to accompany the budget, and the same should be placed before the legislature. This statement should includethe following information: 1. Outstandingdebt and debt repayment profile over the next five years. Zanzibar PublicExDenditureReview Page72 2. Detailsof contingentliabilities-guarantees, parastatal loans, disputedliabilities,etc. 3. Details of accrued liabilities other than pensions. 4. Major policy decisions taken by the Government outsidethe budget process inthe last year and their fiscal impact. 5. A time-series on the number of employees in central government, local authorities, and other institutions,the burdenof salary for which falls on the Government. 6. Pensionliabilities of the Government (five-yearprojections). 7. Tax expenditures and howthey might havebenefitedthe country. 8. Off-budgettransactions, such as borrowingby the Governmentfrom parastatals. 9. Outstandingfinancial issueswith MainlandTanzania. 10. Details of expenditures made from unallocatedbudget retained in the MoFEA in the previousyear. Zanzibar PublicExpenditureReview Page73 8. GOVERNMENT PAYAND EMPLOYMENT INZANZIBAR THEWAGEBILL 8.1 In FY02, expenditures on wages and salaries claimed 63.2 percent of recurrent expenditures or 10.9 percent o f GDP. If salaries paid by the Union Government for police, defense, and other government staff deployed in Zanzibar are added, these ratios increase to 85.1 percent and 14.6 percent respectively. These figures indicate a severe structural imbalance in the Zanzibar budget towards personnel expenditures, with little funds left to cover operation and maintenance and other development expenditures. Figure8.1: Share of theWage Billin RecurrentExpendituresand GDP, FY92-FY02 Share o fWage in Total E k p x w I c 8 P 70% 60% 50% 15% 40% 30% B...10% 1 5% 0% 0% 8.2 As shown in Figure 8.1, expenditures on wages and salaries have risen sharply during the past decade, both as a percentage of government expenditures, and as a percentage of GDP. From a share of about 30 percent of recurrent expenditures in FY93, it has continuously increased and reached more than 60 percent by FY02.l' A similar patterncan be discerned ifthe ratio of wages and salaries to GDP is considered. Here one observes an increase from less than 6 percent o f GDP in FY92 to a peak o f about 12 percent o f GDP in FY98. Until FY98, the expansion ofthe wage bill was madepossibleby a steady increase in government revenue. This permitted an expansion o f the wage bill without requiring a concomitant reductionof non-wage expenditures. However, as revenues started to decline in FY99, the wage bill could not easilybe reduced, and most ofthe decline inrevenue hadto be absorbedby a reductionin non-wage expenditures-a situationthat consequently constrained the ability o fthe public sector to deliver basic services. 8.3 To put Zanzibar's wage bill into perspective, we compare some of the key indicators for Zanzibar to those o f other economies inthe regionandto other small islandeconomies. It merits attentionthat, in additionto the employees ofthe RGoZ, there are also a large number of staff from the Union Government deployed in Zanzibar. This includes the entire police force, the defense force, and other services, such as the Tanzania Revenue Authority or the I OThe figures used inthis analysis cover only expenditures on wages and salaries shown explicitly in the government accounts. Since the salaries and wages o f Zanzibar's local government workers and those of employees o f other government funded institutions suchas the Zanzibar State University are embedded in grants and subsidies, the actual share of wages and salaries in Zanzibar's expenditures is even higher. Zanzibar PublicExpenditureReview Pane74 Bank o f Tanzania. Government estimates of the wage Estimatedcost of Union Government bill of Union Government employees deployed in Staff in Zanzibar, 2002 . Zanzibar run to around T.Shs. 10 billion. Adding this to the wage bill covered by the RGoZ increases the T.Shs. Billion Defense 4,473 ratio of the wage bill to GDP to 14.6 and the ratio of Foreign Affairs 20 the wage billtorecurrent expenditures to 85.1 percent. I Police 4,846 Immigration 174 8.4 First, we compare the total wage bill of Others Zanzibar (Zanzibar Government plus Union staff in (incl. TRA and BOT) 682 Total 10,197 Zanzibar) to that o f other African economies in the region: Kenya, Mozambique, Tanzania, and Zambia. Source: Office of the Chief Government The wage bill to GDP ratio for these economies ranges Statistician from 4.0 percent of GDP for Tanzania at the low end to 8.1 percent for Kenyaat the highend. Zanzibar's ratio of 14.6 percent i s clearly muchhigher than that of the comparable economies. Similarly, the share of recurrent expenditures being claimed by the wage bill is also significantly higher than that of the comparable economies, which display ratiosbetween29.2 percent (Kenya) and46.1 percent (Mozambique). 8.5 We also compare Zanzibar's wage billto that of other small islandeconomies such as Comoros, Cape Verde, Sao Tome, and Principe and Mauritius in Africa, and island economies in other regions, such as Fiji and Trinidad and Tobago. For these small island economies, the ratio of the wage bill to GDP and to recurrent expenditures is typically much largerthan that of larger countriesas a consequence of economies o f scale that exist for many government functions. For example, it would be unlikely that the formulation of policies for large economies would require many more staff than the performanceo f a similar task for a small economy. Nonetheless, even in comparisonto this group of countries, Zanzibar's wage bill appears to be high. Wage BilVGDP Wage BilVRecurrent Expenditures Zanzibar 10.9% 63.2% Zanzibar (incl. Union services) 14.6% 85.1% Kenya 8.1% 29.2% Mozambique 6.0% 46.1Yo Tanzania 4.0% 31.OYo Zambia 5.5% 31.8% Cape Verde 9.9% 29.2% Sao Tome and Principe 9.6% 32.5% Mauritius 6.6% 31.2% Comoros 13.2% 45.1% Trinidad and Tobago 9.9% 39.2% Fiji 20.9% 41.2% PUBLICSECTOR EMPLOYMENT 8.6 Zanzibar's government work force comprisedmorethan 26,000 people in 2002. Out of the total workforce, around 8,000 are teachers and around 3,730 are health workers. This number does not include staff of the Union Government deployed in Zanzibar. The last Zanzibar Public ExpenditureReview Page75 major increase in the number of staff occurred in 2001, when the work force increased by more than four percent, Le., by more than 1000 employees. Looking a little further back, the total number of government employees has increasedby more than 10percent since 1994/95, when it was 23,156. 8.7 A useful indicator for internationalcomparisons of the size of the government labor force is governmentemployment as a percentage o f the population. InZanzibar, government employees account for 2.7 percent of the population. But again, in drawing these comparisons, it must be noted that the employment numbers for Zanzibar do not include Union staff deployed to Zanzibar. The share o f the population employedby Government is higher for Zanzibar than the average for Africa and Asia. Latin America and the Caribbean and the Middle East and North Africa show higher ratios. Industrializedcountries including the OECD countries and the countries of EasternEurope and the former USSR on the other handhave a significantlyhigher share ofthe populationemployedby Government. Table 8.2: GovernmentEmployment (as Percent of Population)for SelectedRegions Source: Zanzibar - CSD, others Salvatore Schiavo-Campo,Giulio de Tommaso, and Amitabha Mukherjee. Government Employment and Pay in Global Perspective:A Selective Synthesisof International Facts, Policies, and Experience. Policy ResearchWorking Paper No. 1771. The World Bank. 1997. 8.8 Government employment figures reveal that about 41 percent of the government workforce is teachers and health workers, while 59 percent of the government workforce works in other areas. This distribution of government workers in Zanzibar is surprisingly similar to the average for most other world regions, where between 37 percent and 44 percent of the government workforce is employed in the healthand educationsectors. Only Eastern Europe and the former USSR show a significantly different distribution, with about 74 percent o fthe government work force employedinthe healthand educationsectors. 8.9 Government employees inZanzibar in the general administrationare classifiedin five broad categories: (a) Drivers and messengers and similar staff fall under the OS category (Operations Scale). Employees in this category are required to have a minimum of form I11 education. (b) The lowest level (KU) o f government service comprises mainly clerks. The educationrequirementfor this service levelis form IV. (c) Most of the Government employees in Zanzibar fall under the general service scale (MUS). Employeesinthis salary grade are requiredto havecompletedFormVI or have completedFormIV and a two year trainingcourse. (d) Technicalstaff (MW salary scale) are requiredto have a first degree or an advanced diploma which is almost equivalent to a first degree. Also, employees in the MUS Zanzibar Public ExpenditureReview Page 76 category with 15 years experience (continuous) in the same field of work and with the ability to discharge their duties effectivelycan be promotedto this category. (e) The highest category of government employment is covered by the leaders scale (MV). Employees in this category include Ministers, Principal Secretaries (PSs), Deputy Principal Secretaries (DPSs), Commissioners, Directors, etc. Most of employees in this category are Presidential appointees and staff who have been promoted from the MV job category to this grade based on length of service and experience. Salary Grade 2000 2001 2002 os,-OSIO 3,239 3,3 19 3,472 KUI-KU2 1,1 15 1,173 1,479 MUS, MUS19 - 8,63 1 9,275 9,52 1 MWI -MWl* 3,851 3,383 3,744 MVI - MV4 170 167 187 8.10 Senior government officials in Zanzibar expressed concern about overstaffing in the lower salary categorieswhich goes hand and handwith critical skill shortages inthe technical and managerial positions. 8.1 1 Initial surveys carriedout by CSD, the MoFEA, and sector ministries did not reveala significant problem of ghost workers. However, a weak information base on employee turnover did lead to the continued issuance of salaries for names representing retirees or deceased. Such accumulated hnds have sometimes beendivertedto service other charges in circumstances in which allocations for other charges were minimal. However, as discussed below, a civil service census would be useful to enhance confidence in the payroll and to establish a firm base for a computerizedcentralpayrollsystem which Governmentintends to implement. 8.12 As in many other countries at a similar level o f development, employment in the formal sector is also low. Government statistics derived from an employment and earnings survey carried out in 2001 indicatethat the total number of employees in the formal sector is around 64,250, out ofwhich about 50 percent are employedby Government. PUBLIC SECTOR SALARIES 8.13 Pay scales in the public sector, including parastatals, are fairly uniform in Zanzibar. Public sector pay in Zanzibar consists of the base salary and various allowances. Table 8.4 providesinformationon the average base salaries by salary grades. While the average salary in the lowest grade was TShs. 34,210, the average salary for the highest salary grades was T.Shs. 148,833. The Zanzibar government wage structure is fairly compressed, reflecting a policy that tries to minimize inequality. It is interestingto notethat between 1999 and 2002 the wage compression ratio declined from 5.4 to 4.4. This is due to the relatively smaller salary increases grantedto higher salary grades comparedto increases granted to lower salary grades. While the average salaries for the lowest grades increasedbetween 1999 and 2002 by 48 percent, that o f the highestcadres increasedby only 15 percent. Zanzibar Public ExDenditureReview Page 77 Table 8.4: Average Base Salary by Job Category, 1999-2002 Source: CSD 8.14 In addition to their basic salaries, government workers receive a variety of allowances. At an aggregate level, allowances have added approximately 81 percent to the base salary during recent years (Table 8.6). FY99 FYOO FYOl FY02 FY03 (actual) (actual) (actual) (budget) (budget) Base Salary (TShs. `000) 11,155 I 11,552 III I 13,200 16,895 1 22,055 Allowances (TShs. `000) 7,823 9,366 10,714 16,372 18,164 18,978 20,917 23,914 III 33,267 I1 40,219 llowances as a % of base salaryI1II 70% III 81% 81% I 97% I 82% Source: RGoZ, Estimates for Recurrent and Capital Revenues and Expenditures for the Year 200112002 and 200012002. 8.1 5 Allowances paid by the Government include allowances for housing, transport, entertainment, electricity, telephone, and a teaching allowance for teachers. Table 8.7 shows the structure ofthese allowances. Table 8.6: Allowances Paid to all except those in government (RGoZ) houses & ranging 1. House allowance from 6% to 50% depending on salary grade Paid at a flat rate of T.Shs. 5000/= for lower to mid-salary grades. For higherisenior (Le. MW & MV) salary grades fuel allowance is 2. Transport allowance paid. Only paid to presidential appointees (MV) ranging from T.Shs. 3. Entertainment allowance 80001= to T.Shs.250001= Paid to presidentialappointees (MV) & ranging from T.Shs. 4. Electricity allowance 3000/= to T.Shs. 59000/= depending on salary grade Only paid to teachers at a rate o f 25% of the gross salary 5. Teaching allowance Only paid to presidential appointees (MV). The amount or the rate 6. Telephone allowance i s not defined. 8.16 Most of these allowances are differentiated according to the salary grade, whereby lower grades receive lower allowances than higher grades, which leads to a certain decompression of the salary structure. Table 8.8 illustratesthe magnitudeof allowances and the impact on take-home pay for the lowestpaidand for the highestpaid salary groups. Zanzibar Public ExDenditureReview Page78 Table 8.7: Base Salary and Allowances for Highest and Lowest Paid Salary Categories, FY02 HouseAllowance EntertainmentAllowance 25,000 Electricity Allowance 3,000 59,000 41.655 307.250 8.17 For the lower salary grades, allowances account for about 30 percent of the base salary. For the highest salary grades, allowances are more than 100 percent of the base salary, not even taking into account the telephone and fuel allowance. Thus, while the highest salaries (net of allowances) are only 4.4 times the salaries of the lowest salary group, the inclusion of allowances increases this ratio to 6.9 times. These calculations do not include telephone and transport allowances, which in many cases are a significant addition to the base salary for senior civil servants. It also should be noted that in addition to the allowances, government employees, especially the more senior cadres, also enjoy non-wage benefitssuch as the duty-free importof vehicles. 8.18 The widespread use of allowances and non-wage benefits also impinges on the transparency of the government salary structure, and wage benefits in the form of tax exemptions create an opportunity for abuse and significant revenue leakages. A consolidation o f basic salaries and allowances and non-wage benefits is therefore recommended. However, since at present allowances are neither taxable nor pensionable, a consolidationo f allowances and non-wage benefits into a single salary will require a careful assessment of the financial implications. This is particularly important with respect to pensions, which, under the current rate structure, would significantly increase and add to the obligations of the ZSSF. At the same time, pension individual benefits could become less equitable, since differences due to allowances would enter into the calculation of pension benefitsunder the presentpensionsystem. 8.19 An assessment of the level o f salaries in Zanzibar would not be complete without a comparison to those on the Mainland. Here, it is important to note that the Union Government has consolidated base salaries and allowances into a single salary payment (Table 8.9). Zanzibar Public ExpenditureReview Page 79 Table 8.8: Comparisonof Estimatesof the Average Salariesfor Zanzibar andMainland Tanzania, N O 2 Ishare in Total Employment1 I Average Salary Zanzibar I Mainland I Zanzibar I Mainland base salary incl. allowances Operational Service 13% 15% 34.210 44.882.44 56.633 Civil Service (incl. Health) 60% 22% 1 52,927 1105,691 85,267 olice and Defense 0% 15% _ _ _ _ _ _ _ _ 75,669 Teachers 27% III 48% I 52,079 170,251 III 92,670 Source: Staff estimates 8.20 Average take-home pay (calculated by dividing the total wage bill by the number of employees) appears to be higher in Zanzibar than in Mainland Tanzania: TShs. 73,353 in FYOl for Zanzibar (and increased to T.Shs. 94,102 in FY02), as compared to T.Shs. 78,878 for Mainland Tanzania in FYO1. 8.21 In drawing the comparison with MainlandTanzania, differences inthe distribution of staff across employment categories and job grades need to be taken into account. In Zanzibar, teachers constitute only 27 percent of government employment, while in Mainland Tanzania, they account for 48 percent. The share o f staff in the operational services is quite similar for Zanzibar and Mainland Tanzania. General administration including health services occupies about 60 percent o f government employees in Zanzibar, while in Mainland Tanzania the share i s only 22 percent. The structure o f public sector employment reflects to a certain extent economies o f scale inherent inthe administration o f a larger entity. While the number o f teachers i s intimately linkedto the size o f the population, the staffing requirements for general administration are to a much lesser extent related to the size o f the population. 8.22 Available data indicate that take-home pay for staff in the operational services (messengers, drivers, and so on) and for teachers i s lower in Zanzibar than in Mainland Tanzania. On the other hand, employees in general administration seem to command a higher take-home pay in Zanzibar than in Mainland Tanzania. 8.23 Even though salaries seem to be on average slightly higher in Zanzibar than in Mainland Tanzania, this does not imply that current salary levels are sufficient to attract, motivate, and retain qualified personnel. Between the early 1970s and the mid-l980s, real average pay in the Tanzania civil service has declined by about 80 percent. Since then, real average pay has recovered to about 80 percent o f the real average wage prevailing at the beginning o f the 1970s. Since some o f this recovery in average pay is due to the consolidation o f base salary and allowances, the recovery o f real wages was actually less pronounced and take-home pay has been declining inrecent years for most salary grades. 8.24 In general, low real wage levels inthe public service are likely to have the following effects: > Demoralization and de-motivation o f both subordinate employees and their supervisors/senior civil servants; 9 Reducedwork effort, declining levelsand expectations inperformance; > Reduced incentives for improvingwork performance and willingness to accept greater responsibility; Zanzibar Public ExpenditureReview Page 80 k Weakeningofaccountabilityandcontrolprocedures/mechanisms; k Diminished ability to recruit and retain qualified managerial, professional, and technical manpower; k Reducedcommitmenttopublicservice; k Moonlightingand"daylighting;"" and k Misappropriation, misuse of public institutions' tangible or financial assets for personalgain. 8.25 Pay reform is one o fthe mainelements of Tanzania's Public Sector ReformProgram. Although the Mainland Government has committed itself to pay reform, progress is slower than envisaged, due to fiscal constraints and other political imperatives. PENSIONS 8.26 The pension system in Zanzibar has undergone significant reform with the establishment of the Zanzibar Social Security Fund (ZSSF) in 1998. Civil servants pay 5 percent of their salary into the pension fund and Government pays 10 percent of the salary to the ZSSF. This entitles civil servants to receive a pension and a gratuity. The mandatory retirement age is 60 in Zanzibar. Earlyretirementwith full benefitscan be granted at the age o f 55. Individual pensionbenefitsare calculatedon the basis o f last salary drawnandnumber of years of service. Pensions range from T.Shs. 2,500 to TShs. 360,500 per month. There are 7,700 pensioners in Zanzibar with a total outlay of 130 million T.Shs., while the corresponding figures for Pemba are 1,950 and 54,903,250 T.Shs. per month. Thirty-two persons, on an average, are paida gratuity every month. 8.27 Government employees who have retired or have left the public service for other reasons since 1998 have not yet received their gratuities and are also not receiving their pensions. This leads to a situation in which many of those employees who have reachedthe retirement age are still on the payroll. An important measure for Governmentwould be to clear the backlog of pension payments, which would also be a pre-condition for removing over-age employees from the payroll. WAGEBILL MANAGEMENT 8.28 At present, there is no centralpayroll system in Zanzibar. Eachministrymaintainsits own payroll on the basis o fwhich it receivesthe monthly release for wage payments from the Ministry of Finance and Economic Affairs. Wage payments are made by the ministries, regions, and districts. Inthe absence of a reliable central payroll system, there is a significant opportunityfor wage bill leakages. 8.29 The absence of a central payroll system also makes establishment control and budgeting for salaries difficult. An analysis of the difference between budgeted and actual salary payments pointstowards serious weaknesses in the budgeting for wages and salaries. As figure 8.2 shows, during FY02 there were significant deviations from the budgetedsalary expenditures, both in the form of over- and under-spending on wages and salaries by spending units. The situation in FY02 also seems to be different from the situation in FYOl, in which budgetdeviationswere muchsmaller, althoughnot insignificant. I'That is, (remunerated)non-civil service work is carried out during time for which civil service pay is received. Zanzibar Public Expenditure Review Page 81 8.30 The hiring process is centralized. The authority to hire and fire civil servants in RGoZ and its institutions is legally vested under the Civil Service Department and Civil Service Commission. The Civil Service will collect all requests to fill vacant positions and submit them to the Civil Service Commission for final selection of people to fill the respective positions. In making decisions on additional staffing, adherence to budgetary limits appears to have been weak in the past, which contributed to the over-expansion of the civil service. 8.31 An employee can only be terminated from work for the following reasons: (i) completion of the day's work, for a daily employee; (ii)completion of contract or failure to abide to contractual terms, for an employee under special contract; (iii)failure to meet probation conditions, if newly employed and has to be under probation; (iv) penalized/fired from work for reasons of misconduct; (v) if terminated under Presidential order/powers; and (vi) for purposes of cost-cutting. Performance-related separation of government employees has to be approved by the Civil Service Commission. Retrenchments undertaken for cost- cutting purposeshaveto be approvedby the Revolutionary Council. Zanzibar PublicExpenditureReview Page 82 Figure8.2: Expenditureson PersonnelEmoluments, Deviation betweenBudgetandActual Cash Crops & Fruit Comm. Ministry o f Information Land & Environment Comm. Ministry o f State Planning Ministry o f Health Min o f Water, E.Environment Ministry of Communication Tourism Commision Min.of State Women& Child. Min. of Agric.& L. Stock Chief Minister Office Ministry o f Education House o f Representatives '0FY02 FYOl Attorney GeneralOffice 0 FYOO Min o f Trade & Marketing President Office R. Council HighCourt Min of State & G. Governance Min. of Finance E. Affairs President Office President Office R.Admin. Prison Department Ant. SmugglingUnit Fire & Rescue Force Economic Bridgade PeoplesMilitia Units I -100% -80% -60% -40% -20% 0% 20% 40% 60% 80% 100% Zanzibar Public Expenditure Review Page 83 TOWARDSCIVIL SERVICEREFORM A PROGRAM 8.32 With the wage bill consuming a disproportionate share of the government budget, wage bill management and civil service reform will need to be the top priority of the Government ifthe performance and service delivery capacity of the public sector in Zanzibar is to be enhanced. The overall objective of the Government should be to achieve a smaller, affordable, well compensated and efficient civil service. In that context, the Government needsto undertake structuralreforms which could include: 9 Carrying out a civil service censusto establishthe integrity ofthe public wage bill; P Removing ghost workers, over-age workers, and workers that show irregularities in their employment from the wage bill; P Clearing up arrearswith respectto gratuities andpensions,to avoid hardships; P Putting inplaceacentral payroll system; 9 Consolidatingbase salaries, allowances, andnon-wagebenefits; 9 Adjusting the workforce to eliminate overstaffing and inefficiencies; 9 Adjusting pay scales to enhancethe capacityofthe public sector to attract, retain, and motivate staff. Since there are critical skill shortages for professional and managerial positions, pay reform will also needto consider a decompressionof salary scales; and 9 Putting inplacetraining andcapacity-building measures. 8.33 In addition, the Government should embark on institutional reforms including: redefining the role of the State, restructuring for organizational effectiveness ifnecessary, and building the capacity for management of the economy, including establishing measures to attract and retain key technical, professionaland managerial skills. Zanzibar PublicExpenditureReview Page 84 Box 8.1: Tanzania's Developmentof an IntegratedHuman ResourceSystem Tanzania embarked on structural reforms in the mid-1980s which included an initiative to establishappropriate manning levels to reduce the size of the civil service. A first attempt to address these issues was largely an unsatisfactory retrenchment exercise which highlighted an urgent need to institute more effective personnel administration and establishment controls. Over the next decade, a series of efforts led to the establishment of an IntegratedHuman Resource System. A Public Service Census was held in 1988 with the objectives o f (a) identifying "ghost workers" on the government payroll, (b) determining the exact size and composition of the civil service, and (c) providing a complete database with a profile of every government employee. The exercise identified some 16,000 ghost workers out o f a total work force of around 350,000 employees. However, attempts at reconciling the census data with the payroll were unsuccessful; it could not be demonstrated that ghost workers were indeed removed from the payroll. In 1994, the Government conducted a "National Pay Day Exercise" with the principal objective of confirming valid employees and detecting any ghost workers in the payroll. This exercise revealed some 13,360 ghost workers. A second objective of building a personnel database by capturingdata on employees for entry into the payroll data base was not achieved. A civil service reform program was launchedin 1995 which included a component to improve central systems for personnel control and data management. A payroll verification exercise took place in 1996, in which employerswere requested to sign off against payroll-sourcedlists for each employee whose presencethey could personally attest. This exercise resultedin the deletion of around5,500 names from the payroll. In 1998, a fresh datacollection exercise was designed to capture the minimum required data for ensuringthat an employee could legitimately be paid a salary or, in due course, a pension. This exercise resulted in the deletion of some 4,000 ghost workers from the payroll. In both the 1996 and 1998 exercise, the monthly volume of requests for deletion transactions rose dramatically once the exercise was launched, which suggests that the number of ghost workers eliminated from the pay roll was actually higher than the numbersrevealedthroughthe personnel verification exercises. The data of the 1998 data collection exercise was merged with the payroll data and was the basis for the IntegratedHuman Resource and Payroll System which was activated in 1999 and became fully operational in 2000. This system allows Government to know the exact nature of its workforce and wage bill on a monthly basis. Source: Neil McCallum and Vicky Tyler. International Experience with Civil Service Censuses and Civil Service Databases. International RecordsManagement Trust. London, UK. May 2001 Zanzibar PublicExpenditureReview Page 85 Box 8.2: International Experiencewith Civil Service Censuses Civil service censuses are typically carried out in pursuit o f one o fthe following objectives: P Cutting costs by reducing the number of ghost workers (fictitious, duplicate, or otherwise erroneous payroll entries). In some cases, civil service censuses were also carried out to provide rough data for downsizing exercises; and P Restructuring of departments and functions,redeploying staff, building capacity, and improving human resource management. This requiresestablishinga detailed profile of public employees-such as data on the mix o f skills, ages, and genders-to facilitate departmentalor functional reviews. Methods used to conduct civil service censuses include: P Physical headcounts, which focus on the number of staff employedand whether the names on a payroll belong to genuine employees. It involves trained teams traveling to various parts of the census area, where individual employees are required to present themselves, often with identification and sometimes with documentation,such as photocopies o f letters of appointment or birth records. These data are then checked off, usually against the payroll. P Questionnaires are used when more detailed data on human resources are needed for restructuring efforts. This approach typically involves the distribution of questionnaires to employees or employers who must submit the information back up the line, taking responsibility for its accuracy. The data are then compiled in some kind of database. P Payroll reconciliation involves reconciling the payroll against alternative data sources, such as individual personnelfiles or servicebooks, the nominal roll or establishedregister, and other databases. Civil service censuses have been implemented in a large number o f countries. Censuses have helped eliminate ghost workers in some countries, resulting in moderate or even significant savings. However, in the absence of routine civil service controls, there has often been no mechanism to ensure that ghost workers do not get back on the payroll. Censuses have contributedto efforts to restructurecivil services, but such exercises are typically incremental processes and need to be informed by regularly updated information on personnel management. Lessons o f experience for the achievementof the targeted objective of civil service censuses include: P Censuses are costly and must be planned strategically with clear objectives, structured relative to the local context, with well defined information requirements. P Censuses should be part of a long-terminvestmentin basicpayroll and personnelsystems. P There need to be appropriate incentives for compliance. For example, incentives may include the stoppingof salary payments for staff on the payroll who are not enumerated. P Coordinatedaction, top level support, and local ownership ofthe exercise's designand delivery are key to successful civil service censuses. Source: The World Bank. International experience with civil service censuses. PREM notes No. 62, January 2002 EMPLOYMENT SALARY REFORM AND 8.34 This section is intended to illustrate the magnitude of the wage bill adjustment requiredto establish a more efficient distribution o f resources between wage and non-wage expenditures in Zanzibar. It also looks at the fiscal consequences of pay policies which would try to enhance anddecompresscurrent pay scales. 8.35 Throughout this section it is assumed that the share o f government spending remains constant, i.e., government spending in real terms grows at the same rate as real GDP. Our key indicator is the ratio o f the wage bill to recurrent expenditures. Although there is no scientific rule as to what an optimal wage bill to GDP ratio should be, the range of data shown in Table 8.1 indicatesthat for most countriesthe wage bill consumes between 30 to 40 percent of recurrent expendituresand that Zanzibar should aim at bringing its wage bill also within this range. Zanzibar Public ExpenditureReview Page 86 8.36 'In the first scenario, staffing levels and real wages are kept constant. Under this scenario, hiring only takes place to replace civil servants who leave the service. We assume that the structure of the civil service remains the same, Le., that the number o f employees in each salary and job grade remains the same. Figure 8.3 shows the projected wage bill to recurrentexpenditures ratio projectionsas dependent on the real growthrate o f the economy. The faster the economy grows, the faster the ratio declines. With an annual real growth rate of one percent, the share of recurrent expenditures in this scenario would barely drop to below 60 percent by the end of the decade. Only if the Zanzibar economy were to grow annually by seven percent would the wage bill to recurrent expenditures ratio fall below 40 percent by the end of the decade, ifno additional measures aside from freezing employment at the current levelwere to be taken. Figure8.3: Projectionof the Share of Expenditureson Wages and Salaries as A Percentof Recurrent Expendituresifthe Number of GovernmentEmployeesI s Kept Constant 70% 60% 50% ~ 40% 30% 20% 10% 8.37 This first scenario shows that even under the assumption of relatively high economic growth, the restructuring of the budget from salary and wage expenditures to non-wage expenditures would proceed only very slowly. This scenario leaves very little room for increases in real salaries, if the objective of a lower share of the wage bill in recurrent expenditures is to be achieved. 8.38 An alternativescenario would be one inwhich the Govemment implementsa general hiring freeze. With an annual rate of attrition of about 2 percent, ifa hiring freeze were to be maintaineduntil the end ofthe decade, the number of government employees would fall from currently above 26,000 to a little above 22,000. The evolution of the wage bill to recurrent expenditures ratio would again be dependent on the rate o f economic growth as shown in figure 8.4. Zanzibar Public ExpenditureReview Page 87 Figure8.4: Projectionof the Share of Expenditures on Wages and Salaries inRecurrent Expenditures with a Hiring Freeze 70% 60% ~ 50% 40% 30% 20% 1 10% ~ 8.39 With a complete hiring freeze, real economic growth would have to increase by at least four percent annually if the wage bill to GDP ratio i s to fall below 40 percent by the end o f the decade. If economic growth is higher, the ratio would drop below 40 percent earlier. For example, with a growth rate of 7 percent, the share of the wage bill would drop to below 40 percent by 2007. These results illustrate that the hiring freeze would need to be maintained for a fairly long period o f time if the wage bill ratio is to be lowered to more efficient levels. However, a hiring freeze is likely to have negative implications for the capacity o f the civil service since there would be no new skills coming into the work force. A complete hiring freeze over an extendedperiodoftime would also be politically difficult to sustain. 8.40 The two scenarios discussed above show that relying purely on economic growth, possibly combined with a hiring freeze, would only bring very gradual improvements, and would always be contingent on solid economic growth performance. Relying on such gradual adjustmentswould also leave little room for pay reform, which is urgentlyneeded. 8.41 At the other extreme of measures to adjust the wage bill would be a staff reduction program including measures such as removing ghost workers, over-age workers, or workers with irregularities in their employment from the wage bill, early retirements, and retrenchments. Inthis context, three scenarios are analyzed. Inthe first scenario, we assume that staff reductions affect only the lower cadres (OS, KU, MUS), given that overstaffing appears to be most prevalent in these job categories. The second scenario looks at staff reductions affectingall levels of the civil service, excludingteachers. Teachers are excluded, since the PRSP targets increased enrollments. Even ifthere is an opportunity for efficiency gains under the current enrollment and facility utilizationrates, the likely need for more staff in the medium term makes staffreductionswhichwould be followed by hiring inthe medium term uneconomical. Finally, the third scenario looks at staff reductions affecting all government employees, including healthworkers and teachers. For each of these scenarios we examine the required reduction in staff if a certain wage bill to recurrent expenditures ratio is to be achieved. The resultsofthese simulationsare presented inTable 8.11. Zanzibar Public ExpenditureReview Page 88 8.42 These calculations do not take into account the cost of removinggovernment workers from the payroll. While removingghost workers from the payroll has no cost aside from the conduct o f a civil service census, other forms o f reducingthe wage bill can have significant cost attached to them. In the past, lack o f funding for retrenchmentpackages has derailed efforts to streamline the civil service. The development of a viable financing strategy will thus need to be part of any effort to reduce the size of the civil service that goes beyondthe removalo f ghost workers. Table8.9: RequiredReductionin GovernmentEmploymentto Reducethe Share ofWages and Salaries inRecurrentExpenditures 55% 50% 45% 40% 8.43 The first scenario, in which only the lowest grades are affected by staff reduction measures, requires the largest cuts, both in terms o f overall numbers and especially as a percentage of staff in the affected groups. To lower the share o f the wage bill in recurrent expenditures from the present 63 percent to 50 percent would require a reduction in staff by 3,512 which is about 24 percent o f the targeted grades. If the wage bill share were to be reducedto 40 percent, staff inthe categories MS, OS, and MUSwould haveto be reduced by about 87 percent or by about 12,596 governmentworkers. 8.44 Scenario 2 considers staffing cuts across the entire civil service, excluding teachers and health workers. Reducing the share o f the wage bill in recurrent expenditures would require a reductionin staff by about 14 percent or 2,491 persons. Ifthe target were a wage bill that consumes less than forty percent of recurrent expenditures, almost 50 percent of the civil service would haveto be laidoff, Le., about 8,932 persons. 8.45 Ifteachers were also includedinthe staff adjustment program,as illustratedunder the third scenario, the number o f staff that would need to be removed from the payroll would equal the number of staff reductions under the second scenario since average salaries for general administration and teaching are similar. However, since the staff cuts would be spreadover a larger number of public employees,the percentage reductionsfor eachcategory o f government employees would be smaller. Reachingthe target of reducingthe wage bill to recurrent expendituresratio to 50 percentwould require cuts in every staff group by about 18 percent. Reachingthe target o f 40 percent on the other hand would requirea reductionin the laborforce across the boardby 34 percent. Zanzibar Public ExDenditureReview Page 89 SALARY ENHANCEMENTS 8.46 The opportunity for salary enhancements in Zanzibar is severely constrainedby the large size of the civil service. Nonetheless, creating an appropriate incentive structure for civil servants, which includes as one of its key components the level of public sector wages, is among the conditionsthat would help improve service delivery and efficiency in the public sector. Inadjustingthe public sector wage bill and re-equilibratingthe balancebetweenwage and non-wage expenditures as part of an overallprogramto reformthe civil service, attention would need to be paid to enhancing salary levels for the categories of employees that are more difficult to attract and to retain, and to reward well performing and well trained civil servants. Ifsalaries remain low, in additionto poor motivation,there is also the real prospect that enhanced non-wage expenditures will not result in improved service delivery, but that nominal non-wage expenditures will be converted into salary enhancements through the misappropriationof public funds and goods for private gain. An example of this found in many countries is the widespread informal practiceat health care units of chargingfor drugs which should be providedfree o f charge and usingthe incometo supplement salaries. 8.47 The following discussionanalyzesthe fiscal feasibility of several salary enhancement scenarios that could be part o f an overall strategy to improvethe service delivery of the civil service. As in the analysis of employmentlevels, our main indicator for fiscal sustainability of wage bill measures is again the ratio of the wage bill to recurrentexpenditures. Ineach of these scenarios, we assume that the Government has created room for an active wage policy by reducing all employment in all categories of staff by 20 percent, and that staffing levels are subsequently kept at that level. Teachers are assumed not to be affected by these staff reductions. Projectionso f average salaries are expressed in constant 2002 T.Shs. In addition to these real increases, it is assumedthat all salaries are increased in line with inflation. For ease o f exposition, only real increases above the compensation for inflation are shown. The following discussion analyzesthree scenarios: (a) salary increases in linewith real economic growth (b) differentiatedsalary enhancements (c) selectiveacceleratedsalary enhancements (SASE). Salary increasesin line with real economicgrowth 8.48 After an initial reduction in the number o f government employees, their number is kept constant, i.e., new hiring takes only place to replace government employees who retire or who leave the service for other reasons. Ifthe number of government employees is kept constant, real salaries can be increasedat the rate o f economic growth without increasingthe share of recurrent expenditures claimed by expenditures on wages and salaries. Figure 8.5 illustrates the real increase in salaries for economic growth rates of 1, 3, 5, and 7 percent annually. Zanzibar Public Expenditure Review Page 90 Figure 8.5: Projections of Salary Increases in Line with Real Economic Growth 180000 160000 140000 3 120000 8 N 100000 0 8 80000 I- 60000 40000 20000 0 8.49 At an annual growth rate of 1 percent, real salaries would increase by only three percent over a three year period and by 7 percent over a five year period. However, if economic growth were sustained at five percent annually and real salaries increased in line with economic growth, real salaries would have increased by 16 percent over the three year period and by 28 percent over the five year period. With an economic growth rate of seven percent, real increases in salaries would be even more rapid: 23 percent over a three year period and 40 percent over the five year period. Differentiated salary increases 8.50 One of the inherent problems of Zanzibar's salary structure is that pay levels for technical, professional and managerial staff are consideredto be too low to attract, retain, and motivate sufficient numbers of qualified staff. To overcome these problems, a pay reform in which salaries for these categories rise faster than for the other categories could be implementedas part of an overall package of reforms to improve the performance of the civil service. Given fiscal constraints, high salary increases for all categories above the real growth rate are not feasible or would require very large cuts in the size of the civil service. To illustrate the opportunity for higher salary increases for certain categories of staff, we consider a policy in which the fiscal target is to keep the ratio of the wage bill to recurrent expenditures at the level attained after the initial reductions in government employment. With a cut in government employment by 20 percent, the share of salaries and wages in recurrent expenditures would fall to 54 percent. A possible policy would be to increase the salaries for higher level employees faster than for staff in lower salary grades in tandem with other measuresto restore transparency, accountability, and integrity inthe public service. Zanzibar Public ExDenditureReview Page 91 Figure 8.6: Projections o f average real monthly salaries by job category 900000 800000 700000 600000 -os1- 0 S l O -KU1~ KU2 ~ d 500000 v1 M U S 1 -MUS19 ~ 400000 - MW12 300000 -MWII M V - M V 4 200000 100000 0 2002103 2003104 2004105 2005106 2006107 2007108 8.51 In this scenario, real wages for employees in the two highestjob categories, M V and MW, would increase by about 50 percent over the three year period and by about 100 percent over the five year period. However, a policy in which only certain groups of staff benefit from fairly large salary increases, while other categories of staff remain without any real salary increases might have negative effects on the motivation of staff in lower cadres and might also be difficult to implement politically. We thus look at the option o f providing real salary increases to all categories of staff, with higher increase of staff in the categories M V and MW. Assuming an annual growth rate of 5 percent, we examine the feasible increase for higher level staff if real salaries for lower cadres are increasedby 2 percent annually and the wage bill as a percentageof recurrent expenditures is to be contained below 54 percent. Figure 8.7: Projections of Average Salaries in a Scenario with Differentiated Salary Increases 700000 600000 500000 "os1 -os10 ' 400000 ~ c r 300000 200000 100000 0 2002103 2003104 2004105 2005/06 2006107 2007108 8.52 Under this scenario, salaries for the lower cadres would increase by six percent over the three year period and by ten percent over the five year period. Higher cadres would still receive annual real salary increases of about 10 percent annually. This would result in a Zanzibar Public ExpenditureReview Page 92 cumulativeincrease in their real salaries of 35 percent over a three year periodand 64 percent over a five year period. Cost of a selective accelerated salary enhancement(SASE) sche,me 8.53 MainlandTanzania has adopted a system of selective acceleratedsalary enhancements (SASE) as part of its public service reform program (PSRP). The objective of the SASE program within the overallPSRP is two-fold. The first objective is to address distortionsthat are introducedthrough donor-fundedlocal cost compensation (LCC). Payment o f LCC has tended to impact on Government's capacity to deliver its own programs since it creates incentives for staff to devote their energy to the implementation of donor-fundedprojects rather than to the implementation of general government programs. In addition, with the large number o fdonors and donor projectsactive inTanzania, the system ofLCC has become Box 8.3: The SASE Scheme in Tanzania The SASE scheme is tied to the medium-term pay reform targets. For government staff that meet specified criteria, it provides salary enhancements that anticipate salary increases that would be only gradually reached through the pay reform program. For example, in the scenario shown in the previous section, the salaries for staff in the categories M V and MW would rise gradually between FY03 and FY08 by about 64 percent. Under the SASE scheme, certain staff would receive the entire salary increase already in FY04, while the remaining staff would only gradually reachthe same pay level. The broad parameters for the eligibility and participation o f ministries, departments, and agencies (MDA) in the SASE scheme include: P The MDA plays a leading role in change management and has significant impact on the socio- > economic well-being o f the average citizen; >> MDAs must demonstrate that objectiveand transparent criteria were used to select nominees; Salary scales are peggedto the medium-termpay reform strategy; Eachapplication is supported by a work plan for each individual which sets out clear deliverables; > and The MDA is still within its allocation for SASE. The allocation will be proportional to the number o f employees in the organization. Once an MDA determined to be eligible for participation in the SASE scheme, MDAs would nominate employees for inclusion in the SASE scheme based on certain criteria. At a minimum, the incumbent must: P be involved in activities that have the greatest impact on service delivery, the management of the > reform process, or the production of strategic government output; have an up-to-datejob description, which is detailed and specifies outputs, quality considerations, > and timeframesfor completion oftasks, and other objective criteria, and sign a performance agreement, which will serve as the basis for determining acceptable performance. In addition to a high-level committee sanctioningthe proposals made by MDAs for SASE after a thorough review by a technical committee, the performance of the incumbents in SASE scheme positions must be appraised on an annual basis, using an objective staff performance appraisal system. The appraisal will determine whether the incumbent's performance warrants the continued receipt of SASE salary supplements. Experience in Mainland Tanzania shows that the implementation of the SASE scheme is administratively quite difficult, which has led to delays and only a limited roll-out o f the SASE scheme so far. It is to be noted that an employee who moves out o f a "SASE" position does not carry with him the salary enhancement. 1unwieldy and has little transparency. Implementationof a SASE scheme was thus inten ed Zanzibar Public ExpenditureReview Page 93 to facilitate the phasing out of LCC and to have the government pay structure and SASE- supported pay levels gradually converge. The second objective of the SASE scheme in line with pay reform strategy targets is to provide immediately enhanced salaries to key government staff who perform critical/strategic functions in the implementation o f the Government'sreform program. 8.54 Adoption o f such a scheme could also be considered for Zanzibar in the context of a broader reform program. It could play an important role in supportingand motivating staff critical for the implementation o f a broad-based reform program. In addition, it could also serve as a benchmark and consolidating structure for local cost compensation that might become more prevalent if donor assistance to Zanzibar increases. It would be importantto ensure that donor interventions are integrated into the public sector and contribute to a strengthening ofthe performance and capacity ofthe public sector, ratherthan drawingscarce human resources to the implementationof donor-fundedprojects with only a limited impact on povertyreduction. 8.55 In the following, we provide illustrative estimates o f the cost of a SASE scheme in Zanzibar. Obviously, such as scheme would have to be part and parcel of a broader public sector reform program and would be contingent on the adoption of a medium-term pay reformpolicy. 8.56 The basis for the analysis is the pay reform programoutlined in the previous section, in which real salaries o fthe lower cadres increaseby 2 percent annually, while real salariesof the higher cadres increase by 10 percent annually. Over a five year period, real salaries for the lower cadres would increase by 10 percent, while those for the higher cadres would increase by about 64 percent. Under a SASE scheme, qualifying employees would receive a salary supplement that would increase their take-home pay immediatelyto the level which, under the gradualpay reformprogram, would only be reachedafter five years. 8.57 In our illustrative estimates, we assume that staff in the categories MUS, MV, and MW would be eligible for SASE. We estimate the annual cost of the introductionof a SASE scheme for three scenarios, which differ by the coverage of eachjob category in the SASE scheme. Table 8.10: Scenarios for the Coverage of the SASE Scheme Grade 1Scenario 1 1Scenario2 I Scenario3 Share of staff benefiting from SASE 10% 20% 50% 30% 50% 100% W 30% 50% II 100% MUS 762 1523 1 I 3808 MV 899 1498 2995 WW 45 75 150 8.58 Under the first scenario, we assume that 10 percent of staff in the salary grade MUS and 30 percent of the staff inthe salary grades MV and MW participate inthe SASE scheme. Under the second scenario, 20 percent of staff inthe salary grade MUS and 50 percent of the staff in the salary grades MV and MW participate in the SASE scheme. Finally, under the third scenario, 50 percent of all staff in the MUS category and 100 percent o f the staff in the MV and MW salary categories receivethe SASE supplements. Zanzibar Public ExpenditureReview Page 94 Table 8.11: Average Monthly Salaries (Incl. Allowances) and SASE Supplement, FY04-FY08 FY04 FY05 FY06 FY07 FYO8 MUS 80,458 82,067 83,708 85,382 87,090 M V 203,884 225,126 248,580 274,479 303,075 M W 441,038 486,987 537,724 593,746 655,605 Average monthly salary SASE supplement (T.Shs., at 2002 prices) MUS 6,632 5,023 3,382 1,708 MV 99,191 77,949 54,495 28,596 M W 214,567 168,618 117,881 61,859 MUS 87,090 87,090 87,090 87,090 87,090 M V 303,075 303,075 303,075 303,075 303,075 M W 655,605 655,605 655,605 655,605 655,605 8.59 Table 8.13 shows projections of the average monthly take-home pay for the salary groups MUS, MV, and MW. Real salaries are increasedby 2 percent annually for the salary grade MUS and by 10 percent for the salary grades MV and MW. The salary enhancement for each year is calculated as the difference between the projected salary for FY08 and the projectedsalary for the specific year. 8.60 The additionalcost per year for each salary grade is thencalculatedby multiplying the number o f staff benefiting in each salary grade by the average monthly SASE salary supplement and by 12 (to convert average monthly additionalexpenditures through SASE to annual estimates). Zanzibar Public ExpenditureReview Page 95 Table 8.12: AdditionalAnnualCost of SASE Implementation 1 FY04 1FY05 I FY06 I FY07 I FYOS AdditionalAnnual Cost (TShs. Million, at 2002 prices): Scenario 1 MUS 61 I 46 1 31 j 16 I 0 M W 116 91 63 33 0 Total 1246 977 682 357 0 % of recurrent expenditures 2.4% 1.8% 1.2% 0.6% 0.0% Additional Annual Cost (TShs. Million, at 2002 prices): Scenario 3 MUS 303 230 155 78 1 0 MV 3565 I I1 2802 1 11 1959 1028 I I 0 IMW 385 303 212 111 0 1 Total 4253 I 3334 I 2325 1217 I 0 % of recurrentexpenditures IIIII 8.3% I 6.2% 1 4.1% I1III 2.0% I 0.0% 8.6 1 Under each scenario, the additional spendingdue to SASE declines every year as the regular salaries converge towards the pay target. Underthe first scenario, in which less than a third of staff in the eligible salary grades benefits from SASE salary supplements, the additional cost during the first year would be TShs. 1.2 billion or 2.4 percent of total recurrent expenditures and then declines gradually until FY08. Underthe second scenario, in which 50 percent of the two highest salary categories and 50 percent of staff in the MUS category benefit from SASE supplements, the initial additional cost would be T.Shs. 2.1 billion or 4.1 percent o f recurrent expenditures. Finally, the third scenario, in which all staff in the categories MV and MW and 50 percent of the staff inthe MUS category receive SASE salary supplements, the initial cost would be T.Shs. 4.3 billion or 8.3 percent of recurrent expenditures. 8.62 Examiningthe three scenarios shows that the additional cost for includingMUS and MW categories in the SASE scheme is comparatively small, since (a) for the MUS category, the amount of the salary supplement is assumed to be low, and (b) for MW, the number of staff in that category is small. The cost and affordability of such a scheme thus depends critically on the treatment of MV staff in such a scheme. Whenever the coverage is less than 100 percent of a salary's categories, it is also likely that our estimates understate the actual cost o f the scheme. This is due to the fact that these calculations are based on average salaries for ajob category. However, under a SASE scheme, it would be likely that the upper salaries in each job category usually would be more strongly represented in SASE, thus leadingto higher additionalcost. 8.63 Experiencein MainlandTanzania points out the fact that SASE needs to be part of an overall reform program that includes the introduction of a series of measures to enhance meritocracy. The experience so far is limited, and the scheme, althoughshowingpotentialto bring some order to the pay issue, needs to be carefully implemented. So far, the implementation of the SASE scheme is administratively quite difficult, and can only be Zanzibar Public ExDenditureReview Page96 implemented at the pace that donor resources are available for its implementation. Government has beenreluctantto take on donor funding for pay reform. 8.64 However, given the high variability and poor predictability of economic growth and government revenue in Zanzibar, adoption of a SASE scheme would expose Government also to significant fiscal risks in case the medium-termgrowth and revenue performance is below the projected levels. Unless Government has restructuredexpenditures in a way that would create sufficient headroom to absorb shocks, the potential benefits from a SASE schemeare unlikely to warranttakingthe inherentrisks. Zanzibar PublicExpenditureReview Page97 DATA ANNEX NOTE: Becauseof independentroundingofnumbers, some totals shown in the following tables may differ slightly from the sum o ftheir components. -k Q\ 2 Q\ !z 0 N 4 W P 2 c P Q\ N N 4 c e g N x P 2 P W L W N - w 4 N w g w 4 N - N 4 0 0 0 0 0 w Q\ N - Q\ VI 2 w x 00 4 2 N P N P a W 4 4 e - P P W a P e - w o o P v 3 N t 4 J N E. 0 0 5 -f V I u l N N c! 4 P P w o o Q 0 i D Q P P - N V I M \ o w c ul P w \ O w w 2? c 3 - L - w w W i- h, Q \oa P P L.- w w m `be "w w o a h , w o a w I z'o" - 'b "w P - 03 "4 o a P W 4 W w 4 w w o w m e w - - m e "P "0N4 ~ v l o~o o o O P - 22f ; 1. --z 3 I. 0 8 4 w - p oaw 'N m w 4 0 L P 0 5 8 L 0 8 I. 0 8 L 0 8 5 8 i 0 8 - 0 8 L 0 - 8 z L - k3 O o4 \ wP Q w uvl w Pin:$ Q14 uDp P W P p e w o o w * - w n o -- - W 4 * L in& 0 a u . O N P "P 0 "P b 4?.Y 9 ? Y : m 2 2z3' W 3' n - u N n m w - w w P W W 0 0 V I P VW I N W 0 ac 3 U n W W - N w A m N L + A W 00 3 w 4 m W N 4 N N 4 VI c n o r P h) 4 N n 01 0 0 4 P - N 3 0 0 0 A 2 9 t N W W N N W + w 4 w P P"VI W I W 0 0 w w 9 P V " a 4 ~ 0 0 m w - " w " w w - - e "N "P 4 w N W " p w .-e" 0 0 0 0 0 0 0 l 0 0 ydNPk"i2 + - V I - N U 4 "w "VI "P w 4 " W 9 " 0 0 0 0 0 0 0 l 0 0 \ O P P 3 N + w " 4 "N "m - i D w - 4 m m -"N-e"4 a - W N W 9 P 3 P o N r - P " P o 0 0 0 0 0 0 0 1 0 0 p ? 254j : - n 9 P 2 r' a r' a 2 n B n 5 4 a n 4 Jl 3 a n a 4 n 4 r' fA 9 - 4 Zanzibar Public ExpenditureReview Page 111 Table AS: Government Expendituresby Ministries and Years. 199912000 2000/2001 200112002 Budget Actual Budget 1 I Actual Budget I Actual 'resident Office R. Council Salary &Wages 126 128 152 143 159 160 Other charges 72 57 92 31 113 59 Dev. (local) 'resident Office R. Admin. Salary &Wages 658 864 904 982 1,444 1,064 Other charges 480 240 483 191 1,444 1,064 Dev. (local) resident Office Salary & Wages 325 117 310 347 393 341 Other charges 416 542 403 507 502 163 Dev. (local) 4inistry of State Planning Salary &Wages 219 220 256 13 Other charges 178 93 Dev. (local) 525 115 :hiefMinisterOffice Salary & Wages 398 397 464 465 723 846 Other charges 491 492 506 267 1,045 515 Dev. (local) Iigh Court Salary &Wages 362 416 436 412 506 482 Other charges 85 83 128 44 153 82 Dev. (local) Lttorney General Office Salary & Wages 143 222 190 2,266 194 203 Other charges 285 149 257 90 177 48 Dev. (local) Iouse of Representatives Salary & Wages 655 772 811 998 1,047 Other charges 676 508 1,158 1,280 1,231 Dev. (local) conomic Brigade Salary &Wages 1,043 1,128 1,338 2,909 1,527 Other charges 207 105 814 190 1,572 Dev. (local) rison Department Salary & Wages 1,013 1,020 1,187 1,176 1,821 1,295 Other charges 299 211 295 78 344 120 Dev. (local) 4in. of Finance E. Affairs Salary & Wages 3,044 3,067 3,217 3,783 5,850 5,365 Other charges 7,211 31,854 - 8,794 14,890 12,044 8,705 1,195 155 896 395 Zanzibar PublicExpenditureReview Page 112 Table A8 (Cont'd):Government Expenditures by Ministriesand Years. 999/2000 2000/2001 2001/2002 linistries Items Budget Actual Budget Actual Budget Actual tin. of Agric.& L. Stock Salary & Wages 1,347 1,506 1,614 1,826 1,908 2,34 1 Other charges 322 162 393 161 1,182 587 Dev. (local) 140 121 230 89 300 149 tin of Trade & Marketing Salary & Wages 110 151 169 143 171 178 Other charges 176 72 160 44 241 115 Dev. (local) 35 95 18 and & Environment Comm. Salary & Wages 199 206 221 113 Other charges 110 96 113 14 Dev. (local) 20 40 finistry of Education Salary & Wages 4,843 5,198 5,773 5,806 7,121 7,950 Other charges 681 280 627 175 1,623 623 Dev. (local) 670 25 950 71 1,100 184 linistry o f Health Salary & Wages 2,322 2,454 2,637 2,507 3,146 3,497 Other charges 1,425 504 1,484 468 2,017 1,368 Dev. (local) 225 230 300 87 fin of Water, E. Environment Salary & Wages 675 721 810 899 1,125 1,183 Other charges 161 103 23 1 61 571 249 Dev. (local) 650 250 500 417 690 linistry of Communication Salary & Wages 860 962 1,087 1,090 1,145 1,631 Other charges 870 203 961 120 1,321 317 Dev. (local) 1,000 348 2,500 291 2,500 638 linistry of Information Salary & Wages 687 Other charges 367 Dev. (local) fin.of State Women& Child. Salary & Wages 119 155 135 180 234 295 Other charges 273 134 288 125 656 165 Dev. (local) 35 ,nt. Smuggling Unit Salary & Wages 1,401 1,455 1,57 1 1,559 2,640 1,828 Other charges 177 149 17 72 175 105 Dev. (local) :ashCrops & Fruit Comm. Salary & Wages 196 Other charges 340 Dev. (local) 'ourism Commission Salary & Wages 122 145 157 149 208 Other charges 611 339 580 370 307 Dev. (local) 10 ire & Rescue Force Salary & Wages 294 335 388 408 810 488 Other charges 58 64 83 42 295 65 Dev. (local) fin of State & G. Governance Salary & Wages 305 282 Other charges 632 167 Dev. (local) eoples Militia Units Salary & Wages 222 91 Other charges 18 18 Dev. (local) Source: Ministry o f Finance and Economic Affairs. ZanzibarPublicExpenditureReview Page 113 Table A9: ActualTax Collection byTRA-Zanzibar andZanzibar Revenue Board Zanzibar Public ExpenditureReview Page 114 TableA10: Compositionof InlandRevenue(% ofTotalCollectionsby ZRB) TableAll: CompositionofIncomeTax and Customs Revenue(% of Total Collections by T W Zanzibar PublicExpenditure Review Page 115 Table A12: HealthSector Budget Summary Sub-vote Description FY98 FY99 FYOO FYOl FY02 FY03 Actuals Actuals Actuals Actuals Approved Budget 03 Head-office Pemba 667 717 694 85 1 1,382 1,386 04 Planningand 1,396 1,499 1,556 1,464 2,461 3,093 administration 05 Social Welfare 0 0 86 130 06 Preventive(Unguja) 33 41 19 20 141 137 07 Curative (Unguja) 448 429 220 239 400 674 08 Curative (Pemba) 66 31 0 31 369 105 09 Preventive(Pemba) 17 47 48 7 194 101 10 Anti-drug unit (Unguja) 0 0 56 61 11 Social Welfare (Pemba) 0 0 53 12 Anti-drug unit (Pemba) 0 0 21 Total 2,627 2,764 2,537 2,612 5,163 5,759 Source: Draft budget estimates for various years, Ministry o f Finance and Economic Affairs, Zanzibar Table A13: EducationMinistryBudget Summary (T.Shs. billion) Sub vote Description FY98 FY99 FYOO FYOl FY02 FY03 Actuals Actuals Actuals Actuals Approved Budget 03 Pembasub office 1.37 1.51 1.97 2.03 2.37 3.66 04 Planningand finance 0.18 0.08 0.04' 0.08 0.15 0.09 05 Archives and documents 0.00 0.00 0.00 0.00 0.17 0.19 conservation 06 Festivals,culture and arts 0.08 0.09 0.05 0.13 0.15 0.19 07 Civil service and 0.13 0.10 0.12 0.13 0.34 0.31 administration 08 Education 2.3 1 2.41 3.01 3.3 1 4.27 6.01 09 Curriculum, exam and 0.19 0.21 0.20 0.24 0.10 0.13 training 10 sports 0.06 0.09 0.09 0.00 0.14 0.18 11 Private office Dy. C M - 0.04 0.01 0.01 0.01 0.00 0.00 12 Adult education 0.06 0.04 0.05 0.08 0.06 0.13 13 Inspection 0.00 0.00 0.00 0.00 0.06 0.07 14 Higher education, science and 0.00 0.00 0.00 0.00 0.94 1.12 technology Total education vote 4.42 4.54 5.54 6.01 8.75 12.08 Source: Draft budget estimates for various years, Ministry o f FinanceandEconomic Affairs, Zanzibar 2 N 2 3 - ? c C d 5 Y c