Tanzania - Zanzibar Public Expenditure Review 2003 - Laying the foundations for improved public expenditure management

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 Currency Unit: Dominican Peso (DR$) Exchange Rate US$l.OO = DR$16.35 (as of March 22,2000) 2. Fiscal Year January 1 - Dec 3 1 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 Stabilization Institute 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  Table of Contents Preface.................................................................................................................................. i ... EXECUTIVE SUMMARY ............................................................................................... 111 Key Recommendations .................................................................................................. xv Summary of Recommendations .................................................................................. xvll .. PART 1: ............................................................................................................................... 1 COUNTRY BACKGROUND ............................................................................................. 1 1. The Macroeconomic Context....................................................................................... 2 Real Economic Growth.................................................................................................... 2 Sectoral Performance ....................................................................................................... 3 General Price Level Developments(Inflation) ................................................................ 4 Fiscal Performance........................................................................................................... 5 The External Sector .......................................................................................................... 7 Public Debt and Liabilities............................................................................................... 9 Implications of the macro-economicenvironment for fiscal policy and public expenditures ..................................................................................................................... 9 2 . Poverty Reduction...................................................................................................... 10 The Zanzibar Poverty Reduction Plan (ZPRP) .............................................................. 10 Progress in the Implementationo f the Zanzibar Poverty Reduction Plan (ZPRP) ........11 3 . The Public Sector in Zanzibar .................................................................................... 13 Overview ....................................................................................................................... -13 Functions Carried out by the Union Government.......................................................... 14 The Central Government................................................................................................ 16 Local Governments ........................................................................................................ 18 District and Regional Development Committees .......................................................... 19 Public Enterprises and Utilities ...................................................................................... 20 Other Government Institutions ...................................................................................... 25 PART 2: ............................................................................................................................. 27 PUBLIC RESOURCE AVAILABILITY IN ZANZIBAR ................................................ 27 4 . Domestic Revenue ..................................................................................................... 28 Institutional Set-Up for Domestic Revenue Administration and Collection ................. 28 Domestic Revenue Structure.......................................................................................... 29 Revenue PerformanceAssessment: Key Issues and Opportunity for Improvement....35 5 . External Development Assistance ............................................................................. 41 External Resource Inflows to Zanzibar .......................................................................... 41 Coordination, Monitoring and Management o f External Resource Inflows.................. 42 6 . Public Debt And Liabilities ....................................................................................... 46 Zanzibar’s Debt Situation .............................................................................................. 46 Conclusions and Recommendations .............................................................................. 49 PART 3 : ............................................................................................................................. 5 1 PUBLIC EXPENDITURES............................................................................................... 51 7 . Zanzibar’s Public Expenditures................................................................................. 52 Structure o f Zanzibar’s Public Expenditures................................................................. 52 Expenditures on Priority sectors .................................................................................... 59 Issues in Budget Management ....................................................................................... 66 8 . Government Pay and Employment in Zanzibar ........................................................ -73 The Wage Bill ................................................................................................................ 73 Public Sector Employment ............................................................................................ 74 Public Sector Salaries .................................................................................................... 76 Pensions ........................................................................................................................ -80  Wage Bill Management ................................................................................................. 80 Towards a Civil Service Reform Program..................................................................... 83 Employment and Salary Reform.................................................................................... 85 Salary Enhancements ..................................................................................................... 89 Data Annex ........................................................................................................................ 97  List o f Tables Table 1.1 : Zanzibar GDP Annual Growth Rates. Real 199.5-2000 (YO) .......................................... 2 Table 1.2: Volatility in growth......................................................................................................... 3 Table 1.3: Sector Contributions to GDP (Current Prices) 1995-2002 (Yo) ..................................... 3 Table 1.4: Revenue and Expenditures in Zanzibar, 1998/99-2001/02 ............................................ 6 Table 1.5: Volatility o f Fiscal Variables and Correlation with GDP 1981-2001............................. 7 Table 1.6: Trade Balance and Re-Exports, 1995/98-2001102 (US$ million) .................................. 8 Table 1.7: Zanzibar Import-Export Trade 1997-2000 (T.Shs. Mill.)............................................... 8 Table 1.8: Zanzibar’s External Debt (June, 2002) ........................................................................... 9 Table 3.1: Administrative Divisions and Number of Shehias in Zanzibar .................................... 13 Table 3.2: Share of Public Administration and Services in GDP .................................................. 13 Table 3.3: List of Budgetary Votes ................................................................................................ 16 Table 3.4: Subnational administrativestructure in Zanzibar ......................................................... 18 Table 3 .5. Compositionof Regional and District Development Committees............................ 2020 Table 3.6: Summary of Parastatals in Zanzibar, Position as o f FYOl ........................................... 24 Table 3.7: Semi-Autonomous Institutions: Subventions and Total Expenditures, FY02 .............. 25 Table 4.1 : Revenue Collection Costs of ZRB and TRA-Zanzibar ................................................. 29 Table 4.2: Zanzibar Revenue Structure (YO of Domestic Revenue) ............................................... 30 Table 4.3: Individual Tax Rates Applicable in Zanzibar 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.) .............................. 34 Table 4.7: Zanzibar Customs Duty Exemptions by Category of Beneficiaries (% of Total) ......... 38 Table 5.1 : Extemal Development Assistance 1997-2001 (US$million) ....................................... 41 Table 6.1: Zanzibar’s External Debt (June, 2002) ......................................................................... 47 Table 6.2: Domestic Debt, (June 2001) ......................................................................................... 47 Table 6.3: Zanzibar’s Domestic Debt (June, 2002) ....................................................................... 48 Table 7.1 : Government Expenditures as % of GDP ...................................................................... 53 Table 7.2: Selected Subventions in Recent Years (current TShs . Million)................................... 54 Table 7.3: Summary of Domestic Debt Servicing (TShs. Million) .............................................. 54 Table 7.4: Government Expenditures by Purpose (% o f GDP) ..................................................... 55 Table 7.5: Unspecified Govemment Expenditures (Million TShs.) ............................................. 57 Table 7.6: Expenditure Shares by Sectors (% of Total) ................................................................. 59 Table 7.7: Participation and Internal Efficiency in Basic Education (YO) ...................................... 60 Table 7.8: Shares in Education Expenditures (%) ......................................................................... 62 Table 7.9: Education Resources by District ................................................................................... 63 Table 7.10: Expenditure shares within health (% of total health expenditures) ............................ 65 Table 7.1 1 : Differences Between Budgeted and Actual Expenditures .......................................... 67 Table 7.12: Major Gainers and Losers during Budget Execution in FY02.................................... 67 Table 8.1 : Expenditures on Wages and Salaries as a Percentage of GDP and Recurrent Expenditures (FY02 or latest fiscal year available) ....................................................................... 74 Table 8.2: Government Employment (as Percent of Population) for Selected Regions ................ 75 Table 8.3: Employment by Salary Category .................................................................................. 76 Table 8.4: Average Base Salary by Job Category, 1999-2002 ...................................................... 77 Table 8.5: Budgetary Expenditures on Base Salaries and Allowances, FY99-FY03 ...................77 Table 8.6: Allowances ................................................................................................................... 77 Table 8.7: Base Salary and Allowances for Highest and Lowest Paid Salary Categories, FY02 .. 78 Table 8.8: Comparison of Estimates of the Average Salaries for Zanzibar And Mainland Tanzania, FY02.............................................................................................................................. 79 Table 8.9: Required Reduction in Government Employment to Reduce the Share of Wages and Salaries in Recurrent Expenditures ................................................................................................ 88 Table 8.10: Scenarios for the Coverage of the SASE Scheme ...................................................... 93 Table 8.11: Average Monthly Salaries (Incl . Allowances) And SASE Supplement, FY04-FY08 94 Table 8.12: Additional Annual Cost O f SASE Implementation.................................................... 95 Table A 1: Summary o f Central Government Operations .............................................................. 98 Table A2: Current Account o f Central Government o f Zanzibar-Administration . 1980/81- 2001/2002 ...................................................................................................................................... 99 Table A3 : Current Account o f Central Government o f Zanzibar-Administration .................... 101 Table A4: Current Account o f Central Government-Economic Classification......................... 103 Table A5: Current Account o f Central Government-Economic Classification........................ 105 Table A6: Classification o f Central Government Expenditures by Purpose................................ 107 Table A7: Classification o f Central Government Expenditures by Purpose................................ 109 Table A8: Government Expenditures by Ministries and Years ................................................... 111 Table A9: Actual Tax Collection by TRA-Zanzibar and Zanzibar Revenue Board.................... 113 Table A10: Composition o f Inland Revenue (% o f Total Collections by ZRB) .......................... 114 Table A 11: Composition o f Income Tax and Customs Revenue (YOo f Total Collections by TRA) ................................................................................................. 114 Table A 12: Health Sector Budget Summary................................................................................ 115 Table A 13 : EducationMinistry budget summary (T.Shs. billion) ............................................... 115 Table A14: Zanzibar Exports, 1995/96-2001/02 ......................................................................... 116 Table A15: Zanzibar Imports 1995/96-2001/02 .......................................................................... 117  List o f Figures Figure 1.1 : GDP and Agriculture Sector Performance. 1995-2002 ................................................ 4 Figure 3.1 : Zanzibar State Trading Corporation. Ratio o f Cost o f Sales to Sales. FY91-FY002222 Figure 3.2: ZSTC. Average Distribution o f Revenue. FY91-FYOO .............................................. 23 Figure 4.1 : Structure o f Revenue (% of Total Revenue), FY92-FY02 ......................................... 30 Figure 4.2: Central Government Revenue by Collection Authority (TShs. Billion) .................... 31 Figure 4.3: Revenue (% of GDP), FY92-FY02 ............................................................................ 35 Figure 4.4a: Revenue (% o f GDP), FY92-FY02 ........................................................................... 36 Figure 4.4b: Indirect Taxes by Source (% o f GDP), FY92-FY02 ................................................ 36 Figure 4 . 4 ~Other : Revenue by Source (% of GDP), FY92-FY02 ............................................... 37 Figure 7.1: Trends in Real Public Spending on Education............................................................ 61 Figure 7.2: Trends in Real Public Spending on Health (1991 prices)............................................ 64 Figure 8.1 : Share of the Wage Bill in Recurrent Expenditures and GDP, FY92-FY02 ...............73 Figure 8.2: Expenditures on Personnel Emoluments, Deviation between Budget and Actual ...... 82 Figure 8.3: Projection o f the Share o f Expenditures on Wages and Salaries in Recurrent Expenditures if the Number o f Government Employees I s Kept Constant ................................... 86 Figure 8.4: Projection o f the Share o f Expenditures on Wages and Salaries in Recurrent Expenditures with a Hiring Freeze................................................................................................. 87 Figure 8.5: Projections of Salary Increases in Line with Real Economic Growth......................... 90 Figure 8.6: Projections of Average Real Monthly Salaries by Job Category ................................ 91 Figure 8.7: Projections o f Average Salaries in a Scenario with Differentiated Salary Increases ..91 List o f Boxes Box 7.1: Experience with Stabilization Funds ............................................................................... 57 Box 8.1 : Tanzania’s Development o f an Integrated Human Resource System ............................. 84 Box 8.2: International Experience with Civil Service Censuses .................................................... 85 Box 8.3: The SASE Scheme in Tanzania ...................................................................................... 92  Zanzibar Public Expenditure Review Page i PREFACE I n May 2002, the Revolutionary Government o f Zanzibar (RGoZ) launched the Zanzibar Poverty Reduction Plan (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 o f 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 o f development assistance to Zanzibar following a political stalemate that resulted from disputed general election results in 1995 and 2000. The stalemate ended after the signing o f an agreement (“Mwafaka”) between the ruling party - Chama Cha Mapinduzi (CCM) and the leading opposition party - the Civic United Front (CUF) in 2001. At the launching of the 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 i s being supplemented by more specific work, such as a study on the clove industry, sectoral public expenditure reviews for health and education, and a private sector development study. This PER i s the first of its kind for Zanzibar and has been prepared in close collaboration between the RGoZ and the World Bank. The objective o f the PER i s two- fold. First, it aims to diagnose the fiscal and public expenditure issues which 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 decisions on future development assistance to Zanzibar. This PER report consists of three parts. The first part provides the 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 development assistance. The last part o f the study reviews expenditures and wage bill issues in detail. Sectoral public expenditures in health and education are reviewed in separate reports. The discussion of budget management and institutional issues 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 Robert Utz (overall coordination and report preparation and chapters on the Public Sector and Wage and Employment Issues). The members o f the World Bank team include Philip Mpango (Domestic Revenue), Praveen Kumar (Expenditure Analysis), Ben Tarimo (Macro-Economic Developments and Debt), and Hamisi Mwinyimvua (Zanzibar Poverty Reduction Plan). Benno Ndulu offered valuable guidance to the team. Emmanuel  Zanzibar Public Expenditure Review Page ii Mungunasi and Wilhelm Ngasamiaku provided able research assistance. Philip Courtnadge o f the UNDP contributed a 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 (Commissioner for Human Resource Development, MOFEA), Ame B. Ame (Accountant General, MOFEA), Hussein S. Khatibu (Commissioner for External Finance, 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 consultations and dissemination efforts in Zanzibar. A World Bank mission visited Zanzibar in September 2002 to discuss the draft concept paper and to agree on the focus of the study with the RGoZ. I n 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 incorporated into the final report. I n addition, in response to the deteriorating situation 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 expenditure needs. A round table was held in May 2003, during which the findings o f the various pieces of diagnostic work and emerging action plans were discussed. The task was carried out under the general guidance of Judy O’Connor, Country Director (AFC04) and Frederick Kilby, Sector Manager (AFTP2). Raj Nallari (WBI PR), Allister Moon (ECSPE), and Christos Kostopoulos (AFTPI) are peer reviewers.  Zanzibar Public Expenditure Review Page iii EXECUTIVE SUMMARY 1. Zanzibar finds i t s e l f currently in an extremely precarious fiscal situation, caused by the dramatic decline in government revenue during the past four years from 25 percent o f GDP in FY98 to 16 percent o f GDP in FY02. The decline i s primarily the result o f the harmonization o f tax policy and administrationbetween Tanzania Mainland and Zanzibar and as such i s likely to be permanent. Increasing revenue will require significant improvements in tax policy and administrationand i s likely to be a slow process. 2. I n the short to medium term, the key challenge for Zanzibar i s thus to adjust budgets and the level and the structure o f public expenditures to the new reality o f 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. Restoring the 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 i s to become the key instrument for the implementation o f Zanzibar’s poverty reduction plan. 3. To date, the burden o f adjustment has fallen exclusively on development and non- wage expenditures, which have been cut in line with the fall in resources. This has led to a severe structural imbalance in Zanzibar’s public expenditures, with expenditures on wages and salaries claiming 62 percent o f 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 comprehensive public sector reform program. Government’s capacity to ensure the integrity of the wage bill i s weak and thus creates the potential for leakages. Furthermore, the salary structure i s compressed which i s one o f the causes for a shortage o f professional, technical, and managerial staff while there i s overstaffing in the lower levels of the civil service. 4. In addition to fiscal and public expenditure issues, this public expenditure review also assessed the role o f the wider public sector in the Zanzibar economy. Here the most urgent need for reform i s with respect to the role of the Zanzibar State Trading Corporation, which currently enjoys a monopoly on the export of cloves, Zanzibar’s main agricultural product. Removing the monopoly and liberalizing the clove sector i s likely to have significant positive welfare implications for clove farmers and the Zanzibar economy.  Zanzibar Public Expenditure Review Page iv Context o f Public Expenditures Management in Zanzibar 5. The combination of a weak economic base and high vulnerability o f the economy to external shocks make sound fiscal policy and public expenditure management a formidable task in Zanzibar. Creating a supportive environment for economic growth has to be the key priority o f the Revolutionary Government o f Zanzibar (RGoZ). In addition, measures that reduce vulnerability to external shocks should also be considered. 6. Problems arising from the economic structure of Zanzibar are compounded by a fiscal management system that i s characterized by a high level o f discretion over both the 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 direction o f creating greater predictability in the system for spending units, while at the same time putting systems in place that allow more effective central control o f key expenditure aggregates. 7. The suspension o f 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 resumption of aid should take into account lessons learned in Mainland Tanzania and also the significant changes in aid management that have taken place in Mainland Tanzania since 1995. 8. Although the review o f fiscal policy and public expenditures over the past few years reveals significant shortcomings, it i s encouraging that RGoZ i s keen to address these problems with important steps being initiated at present. The main message o f this report i s thus to encourage RGoZ to continue the course o f public expenditure reform. Our report i s intended to provide ideas and options for such reforms. Overview o f Public Sector 9. The public sector includes central and local governments and the parastatal sector. Comparative indicators suggest that Government i s rather large, both in terms of resource use and employment, especially when Union services are taken into account. 10. The demarcation of responsibilities for government functions between the Union Government and the RGoZ i s defined in the constitution and various acts. However, the financial relationship between the two Governments i s much less clearly defined, and, in practice, Zanzibar has been reluctant to contribute financially for services provided by the Union to Zanzibar, including services 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 i s 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 Page v 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 o f their impact on the budget, PBZ (People’s Bank o f 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 w i l l be critical to ensure that PBZ i s 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 o f cloves i s 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 o f 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 Resource Availability 14. Secular Decline in Revenue. Domestic revenue declined from 25 percent o f GDP in FY98 to only 16 percent o f GDP in FY02. Tax harmonization with Mainland Tanzania and the closing o f the so-called “Zanzibar Route” contributed to a significant relocation o f 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 turn made Government expenditure planning highly unpredictable. I Revenue (% of GDP), FY92-FY02 I I 1 35% 1 30% 25% % I Of 20% GD I 15% , ~ 1 10% I 5% I 1 0% I I FY92FY93FY94FY95FY96FY97FY98FY99FYOOFYOlFY02 , Direct Taxes Indirect Taxes Other Earnings 16. The main reasons for the high variability o f revenue include the fact that Zanzibar collects most o f i t s revenue from indirect taxes, which reflect instability in the performance o f 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 Public ExPenditure Review Pane vi rates and a standard valuation o f imports. However, measures to protect expenditures from the variability in revenue will be key to reduce the negative impact of high revenue variability. 17. Exemptions. Zanzibar grants various fiscal incentives aimed at promoting investment in tourism, and export-oriented activities. Although these exemptions have to a limited extent helped to attract foreign investment, they have been subject to abuse and have eroded the potential tax base. Thus, a thorough review and rationalization o f the exemptions regime i s imperative to counter abuse and improve revenue collection. Equally important, the RGoZ needs to review its investment promotion 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 i s 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, land lease, quality of labor, and the curtailing o f red tape. 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 o f Zanzibar, including climate, i t s 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) clamping down on tax evasion; and enhancing local capacity to offer quality services to tourists and increase taxable value added. 19. Growth-Based Expansion of the Zanzibar Tax Base. Zanzibar’s tax base remains very narrow, given its negligible manufacturing sector, and volatile mono-crop agriculture production. In these circumstances, expansion o f Zanzibar’s tax base needs to 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 of the strategy to reduce tax evasion and bring the informal sector into the tax net. 20. Sharing of Revenue between Zanzibar and Mainland Tanzania. The URT constitution provides for administration of taxes on incomes and international trade, and the sharing o f this revenue, Bank of Tanzania (BOT)seignorage, and concessional program loans and grants between Zanzibar and the Mainland. These arrangements are proposed to be overseen by a Joint Finance Commission appointed in June 2003. The existing revenue sharing arrangement involves the retention o f all revenue collected by 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 o f the 4.5 percent i s frequently questioned in Zanzibar. Revisiting these income-sharing arrangements i s likely to be high on the agenda of the Joint Finance Commission (JFC).  Zanzibar Public Expenditure Review Page v i i 21. Although taxes administered by TRA are supposed to be governed by the same legislation, there are s t i l l differences in both rates applied and valuation methods in Zanzibar compared to the Mainland. For example, the application of suspended duty in Zanzibar i s limited to only three commodities, compared to fourteen on the Mainland. Similarly, pre-shipment inspection only applies to imports into the Mainland. To resolve these issues, it would be important to get the JFC to address some of these concerns, such as possible compensation for the revenue loss suffered by Zanzibar as a result of import tariff unification. 22. Resource mobilization i s characterized by a large degree o f discretion, both by the RGoZ and by the Union, with regard to resources provided. This includes apparently arbitrary resource withdrawals from the parastatal sector, relatively large tax exemptions, and transfers from the Union, which mask fiscal deficits and seem to be related to the political cycle. 23. Tax administration i s weak with poor capture o f the revenue base, relying primarily on indirect taxes. Expenditures 24. Total central government expenditures in Zanzibar increased steadily during the second half o f the 1990s and almost doubled from 17 percent o f the GDP in FY95 to 33 percent o f the GDP in FYOO. Expenditures on wages and salaries increased from 6 to 12 percent o f GDP during this period. Since then, total expenditures have returned to more normal levels o f about 20 percent of GDP, but the share of wages and salaries has remained stable. 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 investment and has associated costs o f adjustment. The long-term solution to the problem of resource volatility i s a diversified tax base. In the short- to medium-term the key challenge, however, i s 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 Mainland Tanzania, notwithstanding Mainland’s financing o 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 during this period. 27. A large part o f recurrent expenditures i s claimed by the salaries and allowances of an overstaffed civil service. In FY02, salaries and allowances claimed more than 60% of recurrent expenditures, compared to 35% in the Mainland. The large size o f the 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 productivity o f existing public assets and service delivery. Second, it makes the overall demand for public funds very inflexible. This i s a problem when revenues display high variability, since it i s not possible to adjust expenditures on the wage bill quickly.  Zanzibar Public ExDenditure Review Page viii 28. Expenditures on subsidies and transfers were significantly reduced during the first half o f the 1990s and have since remained at a fairly low and stable level of around 1% of GDP. Main receivers of subventions are local government bodies, the Zanzibar Revenue Board, the Zanzibar Free Port Authority, the Zanzibar Investment Promotion Agency, and other non-profit bodies. While the level of cash subsidies i s low, there are sizable in- kind subsidies, mainly in the form o f subsidized water and power supply. 29. The Government runs a cash budget and has managed to keep the fiscal deficit in check. However there are several fiscal risks that the Government must contend with. The most important i s accumulated arrears and accrued liabilities. The Government does not collect information on arrears systematically, nor does it plan for them during budget formulation. Expenditures on Priority Sectors 30. In terms of expenditures by functions, government spending has undergone a major transformation in the last decade. The expenditure share on general administration has increased, while that on health, agriculture, transport and communication has decreased. The challenge before the current Government i s to restore the expenditures on social and economic sectors while cutting down the size of general administration. Expenditure Shares by Sectors (YOo f Total) Fiscal Year FY FY FY FY FY FY FY FY FY FY 92 93 94 95 96 97 98 99 00 01 Education IO 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 IO 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 government financing. 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 health and education are projected to increase. 32. Education sector 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 Mainland Tanzania and comparable with other countries in the region. Per capita annual public spending on education has been improving since FY95, and was around a reasonable U S $ l O in FY02, compared to about US$7 for Mainland Tanzania. 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 equipment are now expected to be provided by the community.  Zanzibar Public Expenditure Review Page ix 34. Government financing of the education sector remains inadequate in view o f the ZPRP objectives. Infrastructure i s highly inadequate and performance indicators are s t i l l low. Yet, it i s possible to achieve better results within the existing envelope through the 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 i s a strong case on equity and efficiency grounds for reviewing and reorienting expenditure patterns on a geographical basis. There i s a considerable opportunity for expanding access to basic education by redistributing existing teachers more efficiently, without having to hire more teachers. Finally, more could be done to encourage private provision o f primary and basic 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 improved to about 1.9 percent o f the GDP and 10 percent o f 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 i s below the levels achieved in the early 1990s. The sector shows tell-tale signs o f chronic under-funding. Health facilities are in a state o f disrepair, equipment i s scarce, and there i s an acute shortage of drugs and medical supplies. 36. As in education, while there i s a need for an increased overall level o f resources in health, there i s an opportunity for reducing administrative expenses and reorienting expenditures to give more attention to preventive care and to provide a greater share for drugs and medical supplies. Regional imbalances in spending also need attention. Budget Management 37. Judged against the desired outcomes o f 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 i s 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 Government plans to move to a multi-year framework for expenditures-the Medium-Term Expenditure Framework (MTEF). This will be a step in the right direction and should be a medium-term goal. I t will help the Ministry of Finance and Economic Affairs (MoFEA) to align the sectoral structure of expenditures with ZPRP priorities in the medium-term. As a first step, MoFEA should attempt to forecast aggregate expenditures for three years in the year 2003-04 based on some macroeconomicmodeling and then request ministries to submit estimates for these years. The Government i s likely to face capacity constraints in developing a medium-term approach to expenditure planning and should preferably start it in two to three sectors to begin with. 39. The unpredictability o f 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 o f limited transparency include:  Zanzibar Public Expenditure Review Page x P The expenditures from the vote ‘unspecified government expenditures’ have been fairly large and there are no clearly laid out criteria defining eligible contingencies 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 o f the central leadership o f the country. The Government has reportedly dropped this sub-vote from FY02 onwards. > The MoFEA issues lump sums to the ministries against wages and salaries. There have been instances when ministries have diverted them for other purposes. P The MoFEA operates a Community Development Fund for providing a government share towards community projects. Reporting on the use o f these resources and adherence to established criteria could be strengthened. 40. Apart from the above, lack of information (for example on arrears and accrued liabilities) makes the budget processes look opaque. I n order to generate confidence in the integrity o f budget processes, the Government needs to give top attention to transparency issues. Public Debt 4 1. Current Status: One o f the key problems in Zanzibar’s debt management i s 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 budgeting for debt service payments and can lead to unforeseen debt service requirements in the course o f 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 o f 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. 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 multilateral debt i s guaranteed by the United Republic of Tanzania (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 dividend due to Zanzibar. 43. Out of the total domestic debt of T.Shs. 30.4 billion, T.Shs. 13.6 billion (45.0 percent) i s owed to the Union Government. The People’s Bank of 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 o f T.Shs. 6.1 billion accounting for 20.0 percent of the total domestic debt. 44. Significant arrears exist, and systems need to be put in place to prevent the recurrence o f public arrears. A central payment system and a Local Purchase Order (LPO) system seem crucial to ensure better expenditure and commitment control.  Zanzibar Public Expenditure Review Page xi 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 o f MoFEA plays a role in the negotiation and contracting o f debt. The department maintains a database, but it i s considered quite incomplete, since project financing has been very difficult to capture. In addition, this department provides data to the Budget Section, but acknowledges that it i s also incomplete. Grants received from Union projects, for example, are often excluded from the estimates. There i s 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 Development Budget estimates are incomplete. This by-passing of 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) maintains a separate data set of Zanzibar’s debt position. 47. MoFEA acknowledges that the situation needs to be improved. I n view o f this situation, it i s 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 o f Debt: A centralized system giving legal authority to only one ministry-namely the Ministry of Finance-for contracting debt on behalf of any Government i s considered to be good practice. If such a discipline does not exist, then any official in the Government will contract debt, leading to a state of confusion whereby (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 lead to leakages and misuse o f the resources. By law, only MoFEA has power to contract debt on behalf o f the Government. However, the ineffectiveness and poor enforcement of such a requirement has created problems. There are s t i l l 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 contingent liabilities for the Government. In view o f this situation, it i s recommended to review the current legal status to provide for enforceable sanctions and, at the same time, strengthen the enforcement of legal rules. 49. The Need for a Debt Strategy: At present there i s 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. Such a strategy will also ensure that debt service i s 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. I n view o f this gap, the preparation of a Zanzibar Debt Strategy (ZDS) i s recommended.  Zanzibar Public Expenditure Review Page xii 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 understanding of the debt situation-both external and domestic-it i s 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 i s thus recommended that a Zanzibar Debt Task Force be constituted to bring the debt situation into a reliable and dependable position. The Bank of Tanzania (Zanzibar Branch) could provide assistance in this regard. Public Wage Bill and Civil Service Issues 5 1, International comparisons indicate that, with 62 percent of recurrent expenditures, the wage bill consumes a disproportionate amount o f government resources, mainly due to high staffing levels in Government. Overstaffing appears particularly prevalent in the lower cadres o f Government, while there i s 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 gratuities and pensions. 52. Simulations indicate that even under fairly optimistic assumptions with regard to economic growth performance, freezing staffing at the current levels while government revenue increases in line with economic growth would 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 reduce the share o f the wage bill below one-half o f recurrent expenditures after a five year period, again assuming sustained economic growth o 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 selective hiring 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 removal o f ghost workers, over-age workers, and workers with irregularities in their employment; (2) early retirement; and (3) voluntary and involuntary retrenchments. Appropriate safety net measures, including appropriate retrenchment packages, payment o f 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 lead to 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 remain competitive 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 o f real economic growth would provide an opportunity for faster salary increases for the higher cadres.  Zanzibar Public Expenditure Review Page xiii 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 increases make the adoption of 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 o f 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 related to merit and performance. I t i s thus recommended to 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 pensionable nor taxable, it will be important to carefully examine the fiscal impact o f the proposed consolidation. 56. The payment o f the wage bill i s delegated to the MDAs, which creates the potential for leakages of expenditures destined for wages and salaries in the form of ghost workers or redirection o f Personnel Emoluments (PE) funds to other uses. 57. There are internal measures in place to ensure the integrity o f the wage bill through reviews undertaken by the Civil Service Department (CSD), sectoral ministries, and the MoFEA. Nonetheless, an external audit of the wage bill and the establishment of a centralized control o f salary payments are recommended to provide increased confidence in Government’s wage bill management. 58. The development o f a comprehensive civil service reform program should be a priority for Government. Elements o f such a reform program should include the reduction in overstaffing, a rationalization of the system of allowances, decompression o f the wage structure, a medium-term program o f 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 i s 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 being developed. I t will be important to ensure that the refinement of the ZPRP and the sector programs are well coordinated in an interactive process, with the ZPRP process having authority in setting priorities and providing the overall framework. 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 plans need to converge quickly. There i s a need to ensure complete buy-in by all the sectors, most importantly by education and health.  Zanzibar Public Expenditure Review Page xiv Recent Reform Initiatives 63. Government i s aware o f many o f the problems and weaknesses in public expenditure management and strategic resource allocation outlined in this report. Initial steps to address some o f these problems are already underway and Government i s encouraged to pursue and deepen these reform efforts: P Adoption o f the MTEF principles in the government budgeting system, which started by raising awareness among politicians, and training directors of planning and administration and chief accountants from key sector ministries; P Establishment o f 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 procurement and wise use o f resources-draft regulations have been prepared in view o f operationalizingthe Central Tender Board; P Preparation and operationalization o f the financial administration and public enterprise act by putting in place relevant regulations; 9 Preparation o f the new investment policy, which also looks at the institutional setup for investment promotion in Zanzibar, which i s to be followed by the formulation o f implementationregulations; 9 Carrying out o f a series of studies to prepare the basis for further reforms. These include: a) A study on problems related to Zanzibar port operatiodtrade, ferryboat and dhow services between Zanzibar, Dar es Salaam, Pemba and Tanga in collaboration with the Tanzania Freight Bureau; b) A study on the assessment o f Zanzibar’s revenue potential; c) A study on clove production and marketing; d) A study on the procurement system; e) A country financial accountability assessment; and Q A Zanzibar Economic Analysis.  9 9 9 e Z 2 2 I z 3 2 3 3: c 3: 9 3: 9 B z r : E E a, Y $ .s  3 3 s 3 z 2 8 3: 5 .- 0 .- z e 0 P 3 z 2 I r f x C O L n 3 3 r: e 2 3 ? i c! cl r z L 3: e 3 2 2 2 r z e 3 r 2 3: 9 2 2 r z E E E 3 1? uu !i!i N m d  z 2 9 2 1% 3: c 9 E I I 3 3 2 L 2. z .- .- 13 0 t 3 c 2 3 1% i- t 3 U 0 M D 2 z Y 9 . 00 m  s 3 3 3 2 z s s s s 9 z - .- h .- b z 9 t z 4 4 3  b? 9 9 3 3 Z Z P 2 2 .- h i e I .- 0 3 2 k 5i I 5 e 0 & 0 Y P  3 3 s 2 * Y L z b 3 2 9 c 3 z r : * h '6 e z z .- 3 2 2 k 3 z e € 4 * b) % G v3 r r - N -\1 N r? N .- 2. .- 0 L E 3 m  Zanzibar Public Expenditure Review Page 1 PART 1: COUNTRYBACKGROUND  Zanzibar Public Expenditure Review Page 2 1. THE MACROECONOMIC CONTEXT ECONOMIC REAL 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 i s explained mainly by clove production (and, to a lesser extent, price instability) which demonstrates the significant influence that this sub-sector continues to 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 G D P 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 productive sectors, a fall in general productivity, low investment, and the impact o f 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 infrastructure services, which are the foundation of 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, reflected the underlying volatility in agriculture,  Zanzibar Public ExDenditure Review Page 3 manufacturing and construction sectors. Table 1.2 below shows measures o f absolute and relative volatilities by different sectors. Table 1.2 Volatility i n growth Industry Absolute volatility' Ratio o f Sectoral to Total G D P 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: Calculations based on data from the Office o f the Chief Government Statistician, Zanzibar. "Volatility i s measured by 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 i t s related sub-sectors), trade (including tourism) and services (which includes administration). Table 1.3 below shows that these sectors constituted almost 80 percent of the GDP in 2002. Table 1.3 Sector Contributions to G D P (Current Prices) 1995-2002 (YO) Source: Department of 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 been poor due to a number of factors, including (1) an agriculture policy which needs review and lacks both an implementation strategy and a well defined institutional framework, (2) low incentives to farmers, (3) government control on the marketing of cloves, and (4) a low level o f investment in agriculture-related  Zanzibar Public ExDenditure Review Page 4 infrastructure. Government has developed an agriculture sector policy which, however, lacked a strategy for i t s implementation. Figure 1.1 : GDP and Agriculture Sector Performance, 1995-2002 15 I O 5 0 -5 -10 Year +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 increasing their share in overall GDP from 19.8 percent in 1995 to 20.0 percent in 2002. The increasing trend i s the result o f robust trade activities within the period. Other fast-growing sectors include construction, transport, and utilities (energy/water) sectors, although they comprise a relatively insignificant share o f total output. P R I C E LEVEL GENERAL DEVELOPMENTS (INFLATION) 1.7 Monetary policy i s a Union matter implemented by the Bank of Tanzania. Between 1995 and 2002, the inflation rate in Zanzibar declined from 29.8 percent in 1995 to 6.0 percent by the end o f 2002, in parallel with the decline in the inflation rate in Mainland Tanzania, as a result of prudent monetary and fiscal policy pursued by the Government of Tanzania. Imported foodstuffs and a relatively stable exchange rate have also served to restrain the rate o f increase in the Consumer Price Index (CPI).' 1.8 The Zanzibar economy has liberalized its trade policy. Market prices are also liberalized with the exception o f the buying prices of cloves, which are s t i l l under state monopoly by the current law. ' The computation o f the CPI i s based on a basket o f goods and services that has not been reviewed since the last Zanzibar Household Budget Survey conducted in 1991. Structural reforms and policy changes, for example, mean that existing weights (e.g. 75 percent for food, 0.6 percent for health services and no weight at all for education) make the headline inflation rate an unreliable measure o 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 o f their incomes on health and education services alone.  Zanzibar Public Expenditure Review Page 5 FISCAL PERFORMANCE 1.9 Table 1.4 summarizes Zanzibar’s revenues and expenditures performance for the period FY00-02. There are several interesting features characterizing Zanzibar’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. I n FYO1, estimated revenue was 24.3 percent of GDP, while actual revenue collected was 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 o f GDP. As discussed in more detail in subsequent chapters, the decline in revenue i s likely to be structural rather than temporary, and it will be important that fiscal policy recognizes this change in the resource situation and adjusts accordingly. 1.10 Although expenditures were reduced in response to the revenue shortfall, financing requirements were typically larger than estimated. They amounted to about 3 percent of GDP during recent years and were covered through domestic borrowing in Zanzibar, and through the issuing o f T-bills by the Bank of Tanzania on behalf o f the RGoZ, leading to a significant increase in domestic debt in recent years. 1.1 1 There are significant data issues that arise in calculating revenues, expenditures, external financing and domestic financing. For instance, external flows 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 considering fiscal data.  Zanzibar Public Expenditure Review Page 6 Table 1.4: Revenue and Expenditures i n Zanzibar, 1998/99-2001102 199912000 200012001 200112002 2002103 bm Budget Budget Actual Budget Actual Estimates Estimates Estimates T.Shs. millions Recurrent Revenue 62,628 55,610 38,669 67,400 42,241 58,848 Recurrent Expenditures 62,468 53,671 42,272 62,500 46,626 60,073 Recurrent Budget -1,225 161 2,939 -3,603 -5,900 -4,385 DeficitiSurplus Development 36,596 599 21,985 1,019 24,800 1,326 Expenditures Overall Deficit (charges -37,821 -439 -20,047 -4,622 -30,700 -5,711 issued) before grants Grants 5 0 0 19,000 0 17,921 Overall deficit (charges -434 -1,370 -4,622 -11,700 -5,711 -19,900 issued)after grants Adjustment to cash and 0 6,736 -15,152 -11,843 other items(net) Overall Deficit (checks -19,900 6,302 -20,047 -7,221 -11,700 cleared) Financing 6,302 20,047 7,221 11,700 19,900 Foreign financing Domestic Bank 6,302 20,047 7,221 11,700 Non-bank 0 - 1444 GDP at market prices 193,543 229,195 229,195 271,414 271,414 273,392 % o f GDP Recurrent Revenue 32.4% 24.3% 16.9% 24.8% 15.6% 21.5% Recurrent Expenditum 32.3% 23.4% 18.4% 23.0% 17.2% 22.0% Recurrent Budget -2.2% -1.6% -0.5% 0.1% 1.3% -1.6% Deficithrplus Development 13.4% 0.3% 9.6% 0.4% 9.1% 0.5% Expenditures Overall Deficit (charges -13.9% -0.2% -8.7% -2.0% -1 1.3% -2.1% issued) before grants Grants 0.0% 0.0% 0.0% 7.0% 0.0% 6.6% Overall deficit (charges -7.3% -0.2% -0.6% -2.0% -4.3% -2.1% issued) after grants Adjustment to cash and 3.5% -6.6% -5.2% 0.0% other items (net) Overall Deficit (checks -7.3% 3.3% -8.7% -3.2% -4.3% cleared) 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 Public ExPenditure Review 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 declining rapidly, due to decrease in output and leakage in the form of smuggling. In recognition o f this, the Government has set out to rationalize the planning and budgeting system in order to match its expenditures with i t s revenues, and to diversify its revenue base, in addition to improving the tax administration. Table 1.5 Volatility o f Fiscal Variables and Correlation with G D P 1981-2001 Fiscal variable Correlation Std Dev (SD) of SD of growth o f coefficient with annual growth fiscal variableLSD of GDP rates (real) G D P growth Recurrent Revenue 0.95 32 3 Total Revenue 0.95 32 3 Recurrent Expenditures 0.95 42 4 Development Expenditures 0.49 41 4 Total Expenditures 0.95 37 4 GDP (Market Prices) 1 9 1 Source: Calculations based on 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 composition of exports shows high dependence on cloves, as stated earlier, the marketing of which i s 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 i s seaweed, the value of which doubled in the last five years. In 1998/99 i t s exports exceeded revenue from the export of cloves. 1.14 Zanzibar imports almost doubled in value between 1995/96 and 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$68.1 million in 2001/02. The recent decline in imports i s attributed to a period of enhanced exemptions in the late 1990s, followed by the withdrawal o f these incentives and the closure of loopholes which promoted re-export activities. The composition of imports reveals that consumer goods have been the dominant category, accounting for about one-half of the total import bill. It i s estimated that by 2001/02 consumer goods will account for 46.0 percent of the total and, within this category, foodstuffs will continue to be the major import item. On the other hand, capital goods have been the second major import category, accounting for between 18.0 percent and 30.7 percent during the period under review. Transport equipment dominated this category in the period 1995/96 to 1997/98 and, thereafter, machinery was the dominant capital import. 1.15 Trade Balance. The trade balance, as per Table 1.7 below, has been worsening due to the larger increase in imports compared with that of exports. This i s attributed to the liberalized trade policy and, in part, to the decline in foreign inflows which might have boosted the export sub-sector.  Zanzibar Public Expenditure Review Page 8 Table 1.6: Trade Balance and Re-Exports, from 1995/98 t o 2001/02 (US$ million) I Exuorts I Imuorts /Trade Balance I 1995196 I 4.4 I 60.6 I -56.2 1996197 I 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 I t should be noted that Zanzibar has not been compiling a full Balance o f Payments position, mainly due to lack o f 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 harmonization with the Mainland, this figure has started to 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 i s freely determined in the Inter-Bank Foreign Exchange Market (IFEM), with commercial banks being the key players. Non-bank financial institutions also participate, while the Central Bank’s role i s one of intervention (buying and selling as deemed prudent). The main challenge facing the Central Bank i s that o f maintaining stability 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 o f exports. This has very much affected the Zanzibar economy, through raising the import bill.  Zanzibar Public Expenditure Review Page 9 PUBLIC DEBT AND LIABILITIES 1.19 The statistics on the position o f Zanzibar’s debt are quite unreliable. Nonetheless, going by what i s 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, accounting for 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 o f 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 o f the total external debt. All the multilateral debt i s guaranteed by the United Republic of Tanzania (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 financial contributions to the Union until the matter i s sorted out by the JFC. IMPLICATIONS OF THE MACRO-ECONOMIC ENVIRONMENT FOR FISCAL POLICY AND PUBLIC EXPENDITURES 1.2 1 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 i s 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 Government in order to allow it to fulfill effectively its core functions. 1.22 The high level of vulnerability typical for small island economies i s reflected in Zanzibar’s high variability of economic growth, exacerbated by fluctuations in imports and exports. I n an economy in which government revenue depends significantly on the taxation o f trade, this translates directly into high variability of government revenue, which makes expenditure planning more difficult than in a more stable environment. I n 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 utilization of stabilization hnds, which could help to smooth incomes not only of farmers, but also o f Government. Similarly, transfers from the Union Government and from abroad in the form of official development assistance could also be designed to compensate for economic fluctuations in Zanzibar.  Zanzibar Public Expenditure Review Page 10 2. POVERTY REDUCTION THEZANZIBAR REDUCTION POVERTY PLAN(ZPRP) 2.1 Poverty i s persistent and widespread, posing a major development challenge to Zanzibar, where over 50 percent o f the population i s categorized as poor. Most o f the 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 educational attainment, limited access to means and ways o f mobility, and exclusion from economic, social, and political processes. Within urban areas, poverty manifests i t s e l f 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 Development Vision 2020 that articulates, among others, a long-term focus in terms of poverty reduction, the major aim being eradication of absolute poverty by 2020. According to the Vision, poverty reduction means increasing the ability of the people to obtain the basic necessities, and improving democracy and social security. In reality, poverty reduction i s construed to be synonymous with increasing access to employment opportunities, essential services, and income-generating activities, and providing social 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 i s 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 improvements in human capacity, survival and well-being; and infrastructure for accessibility and lower costs of production. Promotion o f investment, strengthening o f the financial sector, enhancing social stability, and good governance are 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 Good Governance, and combating HIV/AIDS.  Zanzibar Public Expenditure Review Page 11 2.5 ZPRP i s an operational plan that incorporates strategies to be implemented based on utilization o f domestic financial resources, both public and private, and providing a climate for attracting external resources to support prioritized expenditure plans. The plan evolved through a broad participative and consultative process and i s 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 strong commitment to supporting ZPRP, and underlined the need for capacity building, mobilization o f resources to fill the resource gap, cost-effective and efficient management o f economic programs, aggressive promotion o f investments, control o f HIV/AIDS, and meeting the MDGs. Zanzibar accepted these challenges and indicated its determination to build up a firm macroeconomic foundation that would be able to absorb shocks and encourage various players to participate in economic activities. PROGRESS I N THE IMPLEMENTATION OF THE ZANZIBAR POVERTY REDUCTION PLAN(ZPRP) 2.6 Results on implementing the ZPRP are contained in the ZPRP First Progress Report. The report highlights progress in terms o f policy reform and technical and physical implementations of the ZPRP in the last year. As regards policy reform, the RGoZ undertook various policy and structural reforms 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 expenditure controls via the cash budget; (iv) positive trends in levels o f external inflows in the current fiscal year and improved aid coordination mechanism were achieved; (v) resource allocation to ZPRP priority sectors improved during 200 1-02, with the budget being allocated as follows: education 19 percent compared to 8.6 percent in FYOO; health 11 percent compared to 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 private sector in almost a l l sectors of the 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 Financial Accountability Assessment (FAA), and the analysis of Zanzibar’s economic situation. A review o f procurement systems i s 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 i s 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 o f public funds. The studies are expected to provide a common point of reference for the Government and i t s 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 Expenditure Review 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 opportunities for diversification and growth of non-traditional exports will be conducted in 2003. 2.8 The ZPRP progress report also lists preparatory steps that have been undertaken in determining the required financial resources envelope and its allocation. These include: sensitization o f senior government leaders on the PER and MTEF processes; initial training of technical staff on the PER/MTEF process; passing of and assent to the Financial Administration Act; and institutional measures to improve financial administration, including expansion o f the domestic tax base, improvement o f tax compliance through public education and awareness, and establishment o f Financial Administration Institute. As of now, resource requirements are determined at the sector ministries, based on priority activities identified for implementation. Unit cost analysis i s only being initiated for the Ministries of Health and Social Welfare and Education, Culture and Sports to guide more professional estimates of resource requirements, but for none o f the other sectors. 2.9 Although some progress has been made in the one year of implementation o f the ZPRP, more needs to be done in terms o f improving financial and public expenditures management, increasing resource 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 Public Expenditure Review Page 13 3. THE PUBLIC SECTOR IN ZANZIBAR OVERVIEW 3.1 Zanzibar gained independence in 1964 and soon after united with Tanganyika to form the United Republic o f Tanzania. The newly formed Union Government took responsibility for administering the activities o f 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 o f ministers (the cabinet), a legislature (the House o f Representatives), and a judiciary. The RGoZ i s 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 o f four main elements-functions carried out by the Union Government (GOT), RGoZ, local governments, and the parastatal sector. The RGoZ i t s e l f consists o f a legislature, a judiciary, and an executive. This report i s mainly concerned with the executive arm o f Government. Table 3.1: Administrative Divisions and Number of Shehias in Zanzibar Unguja Island Pemba North Region Urban West South Region North Region South Region 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 i s divided into five regions, each o f which contains two districts. Each district i s further divided into shehias. Table 3.2: Share o f 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 administration and service in GDP has increased from about 15 percent at the beginning of 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 Expenditure Review Page 14 FUNCTIONS CARRIED OUT BY THE UNION GOVERNMENT 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 Constitution of Tanzania and the Government of the United Republic b) Foreign Affairs c) Defense and Security d) Police e) Emergency Powers f) Citizenship g) Immigration h) External borrowing and trade i) Service in the Government of the United Republic j) Income tax payable by individuals and by corporations, customs duty and excise duty on goods manufactured in Tanzania collected by the Customs Department k) Harbors, matters relating to air transport, posts and telecommunications 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) Industrial licensing and statistics n) Higher education 0) Mineral oil resources, including crude oil and natural gas p) The National Examinations Council o f Tanzania and all matters connected with the functions of that Council q) Civil aviation r) Research s) Meteorology t) Statistics u) The Court o f Appeal of the United Republic v) Registration of political parties and other matters related to 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 responsibilities i s sometimes not entirely clear. For example, even though higher education i s 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 i s the area o f statistics, in which Zanzibar maintains i t s 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 i s 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 computation of GDP. 3.7 The financial relationship between the RGoZ and the Union Government i s less clearly defined than the distribution o f responsibilities. The constitution provides for the establishment o f a Joint Finance Commission which has, inter alia, the responsibility for:  Zanzibar Public Expenditure Review Page 15 a) analyzing revenuedexpenditures related to Union matters and making recommendations concerning the net contribution of each Government to the Joint Finance Account in the Union ConsolidatedFund; and b) scrutinizing fiscal relations between the two Governments. 3.8 The RGoZ has authority to set tax rates and tariffs on non-Union taxes, but in 1998 Zanzibar and Mainland Tanzania substantially harmonized their tax regimes. There are s t i l l differences in exemptions granted. Tax collection in Zanzibar i s divided between the Tanzania Revenue Authority (TRA), which collects import-related taxes, and the Zanzibar Revenue Board (ZRB), which i s responsible for the collection of domestic taxes in Zanzibar. All revenue collected by the TRA and ZRl3 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 i s no system for the distribution of project and other development aid. Inclusion of Zanzibar in project support extended to the Union Government i s currently made on an ad-hoc basis, in the form of negotiations between the Union Government and the donor. Zanzibar also attracts direct assistance from some donors, such as China. 3.10 Since 1996, Zanzibar i s also entitled to and has received 4.5 percent o f the BOT dividend. Responsibility for contracting external debt i s vested in the Union Government, which also guarantees debt contracted on behalf of Zanzibar. In the past, the RGoZ has often failed to make i t s debt service payments on external debt; such debt service was typically made by the Bank of Tanzania. I n turn, the BOTdeducted debt service made on behalf o f Zanzibar from the transfer of Zanzibar’s share of the BOTdividend. 3.1 1 Zanzibar does not make any contributions to 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 expenditures for 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 o f 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 designing an appropriate and feasible system of burden- sharing, the fiscal sustainability of such an arrangement needs to be taken into account. Given the severely constrained fiscal situation in Zanzibar caused by the decline in revenues and an unsustainably high wage bill, normalizing financial relationships will need to be part o f a medium term fiscal reform strategy. A burden-sharing formula will also need to take into account Zanzibar’s diseconomies o f scale in its government functions in order to avoid crowding out poverty-reducingand growth-supportingexpenditures.  Zanzibar Public Expenditure Review Page 16 THECENTRAL GOVERNMENT 3.13 The executive arm o f the RGoZ comprises four tiers: a) ministries, departments, and agencies (MDAs); b) the regional administration; c) the district administration; and d) the shehias. a. Ministries, Departments and Agencies 3.14 At present, the central government comprises 12 ministries. At the head o f each ministry i s 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 o f the ministries. While all ministries have their headquarters in Stone Town, each ministry also maintains an office in Pemba led by 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 reducing the number of ministries. As a result, two ministries (the Ministry o f 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 o f two ministerial, principal secretary, and deputy principal secretary positions. Table 3.3: List o f Budgetary Votes President’s Office Ministry o f Education, Culture, and Sports President’s Office Revolutionary Council Ministry o f Health and Social Welfare President’s Office: Regional Administration Ministry o f 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 House o f 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 o f 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 A dm in istratio n 3.16 Zanzibar i s divided into 5 regions, three in Unguja and two in Pemba. Each region has a regional commissioner, who has the rank of a Minister, appointed by the President. The functions o f the regional commissioner include:  Zanzibar Public Expenditure Review Page 17 0 monitoring, supervising, and assisting in the execution of the hnctions o f the Government in his region; 0 assuring that the policies, plans and directives of the Government are observed; 0 maintaining law and order in the region in collaboration with law enforcement agencies; and 0 assuring that resources, both material and manpower, are used for development in the economy to enhance welfare. 3.17 Regional Commissioners are also Members o f the House o f Representatives. A regional administrative officer, appointed by the President, i s in charge of the day-to-day running o f the government business in the region. H e also i s the head of public officers posted in the region and the accounting officer in the respective region. 3.18 The Minister o f Local Government and Regional Administration 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 officers are answerable to the Regional Commissioner. c. District Administration 3.19 Each region i s divided into two districts; Le., there are six districts in Unguja and 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 i s a district commissioner appointed by the President, and there i s 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 ministries assigned to the district. d. Shehias There are 249 Shehias in Zanzibar, which are at the lowest level of public administration in Zanzibar. Each Shehia i s administered by a Sheha, who i s a government employee and appointed by the Regional Commissioner. The Sheha i s answerable to the District Commissioner o f that area. H i s functions include: 0 implementing all the government laws, orders, policies, and directives for 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 of that area and wisdom; 0 keeping records o f a l l documents relating to the registration o f marriages, divorces, births and deaths, the transportation of crops and livestock, charcoal permits, and so forth, as directed from time to time by the institutions concerned; 0 controlling immigration in his shehia and keeping records thereof; 0 receiving notification for convoking all public meetings; 0 undertaking all other tasks that are legal and that have been assigned by the District Commissioner.  Zanzibar Public Expenditure Review Page 18 3.20 In every Shehia, there i s also an advisory council consisting o f 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 i s 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 f the shehia. e. Assessment 3.21 The most striking feature o f Zanzibar’s government structure i s 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 defining more clearly the role o f regional and district administration, which would go a long way in avoiding the overlap and duplication o f staffing. 3.22 In addition to the review o f the appropriateness o f the vertical structure o f the central government, there i s also an opportunity for reviewing the division o f responsibilities between (a) the de-concentrated levels o f the 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 i s 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 o f 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 i t s area o f jurisdiction. 3.25 Town Councils are assigned certain specific functions including: (i) the cleaning o f trunk roads; (ii)the regulation and conduct o f public hire vehicles; (iii) street lighting; and (iv) the naming o f streets and numbering o f buildings. The functions o f a District Council are more general and include responsibility for: (i) formulating, coordinating and supervising the  Zanzibar Public Expenditure Review Page 19 implementation o f 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 i t s jurisdiction. 3.26 The responsibilities o f the Zanzibar Municipal Council include: P controlling all public roads and streets within the municipality; P naming of streets and numbering o f buildings; k establishing and maintainingrecreation grounds; k implementing public health initiatives as required by the Minister (responsible for local government administration); P construction, equipment and operation of drainage and sewerage works; and P the administration o f public markets. 3.27 Borrowing by the Council i s only permitted with the approval o f the Minister, while external financing (outside Zanzibar) i s 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, Construction and Environment; and (v) Social Services Affairs. DEVELOPMENT DISTRICTAND REGIONAL COMMITTEES 3.28 Zanzibar law foresees also the establishment of development committees for each region and district. The functions o f these development committees are defined as follows: (a) to supervise the implementation o f government policies, to identify the problems, and to advise Government on the best way to solve these problems and to promote development in their areas; (b) to monitor and assist in the formulation of policies for local government authorities in their areas and advise on the best implementationstrategies; (c) to mobilize people to participate and contribute in a l l ways possible to assist in the efficient uses o f resources and the protection of environment for sustainable development and in all activities of national development; (d) to ensure that implementation strategies correspond to relevant policies and to create awareness among the people in their areas on the importance o 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) i s composed of staff from the de-concentrated levels o f the central government and from the local governments and their officers.  Zanzibar Public Exuenditure Review Pane 20 Table 3.5: Composition of Regional and District Development Committees Regional Development Committee I District Development Committee I PUBLIC ENTERPRISES AND UTILITIES 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 f the 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 parastatals that were to remain under government control; P Non-sensitive parastatals that were to be privatized on a joint venture basis; and P Parastatals that were to be completely privatized. 3.32 Subsequently the number o f parastatals declined from 25 in 1994 to 12 in 2002. It i s 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. O f 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 Public Expenditure Review Page 21 3 -33 The remaining operating parastatals include the 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 o f these parastatals i s weak. This i s due to low efficiency in operations, financial relationships with other parastatals that are characterized by high levels o f arrears and debt, presumptive taxation that i s 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 o f many o f the parastatals i s even worse than indicated by their financial accounts. 3.34 At present, the financial relationships among 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) i s the Government, with arrears o f 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 i t s retained profits o f T.Shs. 1 1.6 billion in government debt, including T.Shs. 2.5 billion in Government stock and a loan to the Treasury of T.Shs. 0.7 billion. The indebtedness o f GoZ to the People’s Bank o f 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 i s also currently paying more than market rates o f interest for large deposits from parastatal companies, which contributes to PBZ’s losses. 3.35 There i s also no, or a very limited, economic rationale for government involvement in 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 i s likely to have severe negative economic consequences. Similarly, in cases in which the parastatal i s the single provider o f specific services, such as the Zanzibar Port Authority or the Zanzibar Fuel and Power Corporation, poor service delivery has had a negative impact on economic growth and poverty reduction. Government has acknowledged weaknesses in the parastatal sector and the need for privatization o f many o f the remaining entities. However, efforts have so far been cautious, and privatization has been slow. I t i s recommended that 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 requirements due 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 Public Expenditure Review Page 22 Figure 3.1: Zanzibar State Trading Corporation-Ratio o f 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: Based on data provided by ZSTC 3.36 The monopoly over the export of cloves-Zanzibar’s main agricultural commodity- held by the Zanzibar State Trading Corporation, i s currently under review, and a study on the clove sector, with options for liberalizing clove 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 high world 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 remitted to the Government. The remaining one-half o f the surplus i s retained at ZSTC. However, ZSTC reserves typically have been used to provide loans to government and other parastatals and to purchase government securities. FYOO was an exceptional year with extraordinarily high clove prices which were not passed on to farmers. However, even if FYOO i s excluded from the calculation of the average cost shares, the surplus still comes to 5 1 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 o f Revenue, FY91-FY00 M a n a g e m e n t Cost 12% ost of Sales 14% Repair 2Yo 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 o f clove exports handled by ZSTC out o f total clove production i s 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 had to borrow heavily from PBZ to be able to purchase the harvest. Even so, it still ran out o f hnds and had to halt the purchase o f cloves from farmers. 3.38 Abolishing the ZSTC monopoly i s likely to lead to increased incomes for farmers, and work i s currently underway to study the effects o f liberalization o f 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 o f revenue for government, it will be important to put in place an appropriate taxation system for the clove industry, once clove marketing i s liberalized.  Zanzibar Public Expenditure Review Page 24 Table 3.6: Summary o f Parastatals in Zanzibar-Position as o f 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 o f cloves liberalization o f 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 o f 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 o f power in Zanzibar People's Bank Commercial 222 3,251 (4,234) Options for o f Zanzibar banking privatization under (PBZ)* consideration by government Zanzibar Shipping agency 203 1,200 428 270 The business has Shipping and shipping liberalized but Corporation line Government i s (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 In the 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 Privatized to African Fisheries activities Fishing Corporation Corporation and now under liquidation Source: MoFEA and annual reports 'various parastatals * Data for 2000  Zanzibar Public ExPenditure Review Page 25 INSTITUTIONS OTHERGOVERNMENT 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 Security Fund (ZSSF). Most of these government institutions receive subsidies from the government to augment resources from their own revenue. Table 3.7: Semi-Autonomous Institutions: Subventions and Total Expenditures, FY02 1 Institution 1 Employment ~. 1 Government Subvention 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 included refunds to members who had died, retired, became incapacitated, or had left the country for good before sixty months of contributions. Benefit payments by ZSSF will commence in July 2003. Until then, pensions for government and parastatal employees are paid directly from the government budget. An actuarial study i s currently being carried 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 o f early retirement (at the age o f 55) since the establishment of ZSSF. 3.42 As the ZSSF plays an important role in providing a social safety net for the people o f Zanzibar, professional and prudential investment o f the contributions needs to be the top priority for the ZSSF. Ensuring that the staff and the board o f trustees have the appropriate expertise will be crucial. Since this expertise might at present not be available in Zanzibar, having the investment funds of the ZSSF managed by a well established financial institution with the required expertise should be considered, while ZSSF can be responsible for administeringand collecting contributions and paying benefits. 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 Public ExPenditure Review Parre 26 These agencies appear to fulfill similar functions, and there appears to be significant overlap. Integrating these 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 costly to government; and avoiding conhsion among potential investors.  Zanzibar Public Expenditure Review Page 27 PART 2: PUBLIC RESOURCE AVAILABILITY IN ZANZIBAR  Zanzibar Public Expenditure Review Page 28 4. DOMESTIC REVENUE REVENUE INSTITUTIONAL SET-UP FOR DOMESTIC ADMINISTRATION AND COLLECTION 4.1 There are three institutions that are responsible for domestic revenue administration and/or collection in Zanzibar. The first i s the Tanzania Revenue Authority (TRA) which i s 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 i s remitted to the Zanzibar Treasury governed by the principle of retaining the revenue at source. 4.2 The second i s the Zanzibar Revenue Board (ZRE3) which administers inland revenue, composed of over 15 different tax sources. The major ones include: VAT-local at the rate o f 20%;3 petroleum product levy, which i s 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 i s 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 o f 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 Z RE 3 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, business licenses, and the entertainment tax. 4.3 The third i s the network o f 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 o f local government revenue include trade license fees, the crop cess (assessment), productiontaxes, property rents, and fines and penalties. 4.4 Four institutional concerns are worth mentioning. First i s the duplication of effort by the several licensing authorities, such as ZIPA, the Ministry of Industries, the Zanzibar Municipal Council, etc. This has resulted in a loophole, whereby would-be licensees choose at will where it i s more convenient to them to obtain licenses. RGoZ i s aware o f 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 island states, Zanzibar VAT on manufactured goods produced on the Mainland but destined for Zanzibar i s collected by TRA- Mainland and i s subsequently remitted to the Zanzibar Treasury.  Zanzibar Public Expenditure Review Page 29 has limited institutional capacity in terms o f skilled human resources, equipment and facilities. For example, in the case o f ZRB, only 32% of the 127 employees have at least an advanced diploma. There i s a shortage o f qualified tax assessors, collectors, and accounting and managerial cadres. The third related issue i s inefficiency in collection. A comparison o f the revenue collected by TRA and ZRB vis-&vis the corresponding recurrent costs (salaries and other charges) suggests that ZRB i s relatively cost-inefficient in collecting revenue. I t s 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% 200010 1 15,167.9 612.4 4.0% 23,405.9 499.2 2.1 Y o 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. I t i s interesting to note that special allowances alone paid to ZRB staff amount to over a quarter (26%-29%) of total annual recurrent expenditures o f ZRB. I n view of the above, there i s 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 i s 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 legislation in Zanzibar does provide for appeal machinery, it i s yet to be operationalized to deal with taxpayer complaints. This has partly to do with a shortage o f lawyers needed to serve on the Tax Appeal Tribunal. REVENUE DOMESTIC STRUCTURE 4.6 Revenue Structure by Type o f Revenue: Zanzibar depends mainly on taxes as the major source of revenue (Figure 4.1). I n the past five years, revenue from taxes contributed an annual average of 89 percent o f total revenue, Zanzibar’s revenue structure i s heavily skewed toward indirect taxes (Table 4.2). The major indirect taxes include import duties, VAT, and excise duty. These together contribute between 60 and 70 percent o f 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 contribution of direct taxes (pay-as-you- earn, company tax, individual income tax, withholding tax, and capital gain tax) to total domestic revenue in Zanzibar i s rather small. I n 1999/00, 2000/01 and 2001/02, direct taxes contributed only 3.6 percent, 7.1 percent, 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 i s 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 Page 30 Figure 4.1: Structure o f Revenue (YOof Total Revenue), FY92-FY02 1 Chart 1: Revenue ( O h o f Total Revenue) 90 70 +Direct Taxes g 50 1 --)-Indirect Taxes 1 1 Other Eamings 3n IO 0 DO0 1/02 199 I/921992/931993/941994/951995/961996/971997/981998/991999/002000/0 Yem Table 4.2: Zanzibar Revenue Structure (% o f Domestic Revenue) 4.7 Zanzibar Public Revenue by Collecting Agency: Until recently, revenue collected by TRA constituted the 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 dropped to about 50 percent in 2001102, with the collapse of imports and, correspondingly, o f customs revenue (Figure 4.2). Over the past three years, the contribution to inland revenue from most of the revenue sources administered by ZRB has been increasing (Annex Tables A2 and A3), except for the contribution from parastatals, which i s dominated by declining contribution from 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 harmonization was adopted, and more recently with the relocation o f two main importers to Tanzania Mainland where they also established industries.  Zanzibar Public Expenditure Review Page 3 1 Figure 4.2: Central Government Revenue by Collection Authority (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 o f i t s revenue sources to identify those that merit repeal, and potential new instruments, such as VAT on petroleum, which i s 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 i s the c.i.f. value of certain imported items, including import duty on the products. The base for VAT on imports i s the c.i.f. value of imports including import duty and excise duty (if applicable) on the products. There are four nominal tariff 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 East African Community (EAC) are charged at 20 percent o f 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 i s 20 percent and i s applied at the point o f importation and delivery of service, with a zero rate on exports. I n 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 i s charged on the remainder of the excisable goods. Although the tax base for international transactions i s similar to that used for Mainland Tanzania, there are s t i l l some differences in both rates applied and valuation methods 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 i s 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, imposition of suspended duties on needed imports would have increased the cost of living o f the people of Zanzibar. Similarly, while there i s pre-shipment inspection and valuation in the case o f imports into Mainland Tanzania, there i s no similar arrangement for Zanzibar. There are also differences in the standards for the goods, since those set by the Tanzania Bureau o f Standards (TBS) do not apply in Zanzibar. 4.10 Individual income tax rates applicable in Zanzibar since February 2000 (Table 4.3) fall into 1 1 bands ranging from 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 ExPenditure Review 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 f taxable income. Income Bracket (Monthly) T.Shs. Rate Payable 1 c)-2s.nnn - -_ >--- Nil 2,501-45,OOO 0 + 7.5% over T.Shs. 25,000 45,OO 1-65,000 T.Shs. 1,500 + 10% over T.Shs. 45,000 65,OO 1-85,000 TShs. 3,500 + 12.5% over T.Shs. 65,000 85,OO 1-105,000 T.Shs. 6,000 + 15% over T.Shs. 85,000 105,OO 1-1 25,000 T.Shs. 9,000 + 17.5% over T.Shs. 105,000 125,OO 1-1 45,000 TShs. 12,500 + 20.5% over T.Shs. 125,000 145,OO 1-1 75,000 T.Shs. 16,600 + 22.5% over T.Shs. 145,000 175,OO 1-205,000 T.Shs. 23,350 + 25% over TShs. 175,000 205,001-235,000 TShs. 30,850 + 27.5% over TShs. 205,000 235,001-265,000 T.Shs. 39,100 + 30% over TShs. 235,000 Above 265,000 T.Shs. 48,100 + 35% over T.Shs. 265,000 4.1 1 It i s quite clear from Table 4.3 that there i s 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 ensuring revenue neutrality. I t i s important to underline here 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 United Republic o f Tanzania (URT). Otherwise the situation i s likely to degenerate into competition for tax revenue and investment between the two sides o f the Republic. Thus, in all these areas, harmonization o f 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 i s s t i l l in i t s initial stage and has so far entailed only de-concentration (that i s 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 o f own revenue and applicable rates are specific to their areas o f jurisdiction (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 o f their total annual expenditures. In the case o f Zanzibar Municipality, only about 4 percent o f expenditures i s financed from i t s 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 i s a move in the right direction to foster economic decision making in public enterprises and enhance their viability. Table 4.4: Transfers from Government Enterprises, FY99-FY03 (T.Sh. million) Source: Estimates for Recurrent and Capital Revenues and Expenditures for Year 2001/2002 and 2002103 4.14 The high contribution to revenue from parastatals in FYOO i s 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 i s 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 o f 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 in the clove sector.  Zanzibar Public ExPenditure Review Page 34 Table 4.5: Transfers by 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 % o f T o t a l 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 i s port and airport charges such as departure taxes. The significance o f other earnings has declined in recent years. On average, other earnings contributed 5 percent o f total revenue during the last five years, whereas, they contributed 13 percent during the previous five year period. 4.16 Transfers from the Union Government. The Union constitution provides for the administration o f taxes on incomes and international trade, 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 involves the retention o f all revenue collected by TRA at source. With regard to the distribution of program loans and grants, Zanzibar i s supposed to receive 4.5 percent of the total. It was noted that the 4.5 percent subvention, though erratic, provides $8-$10 million to RGoZ annually. In addition, there are also hidden transfers from the Union Government 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’S decision to 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) s t i l l remain to be resolved. The 4.5 percent share o f program loans and grants i s 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 import tariff unification. 4.18 Domestic Borrowing. RGoZ has had to resort to domestic borrowing to finance i t s expenditures, mainly through issuance o f 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 FY97 to T.Sh. 4.5 billion in FY99, and further to T.Sh. 6.3 billion in FYOO.  Zanzibar Public Expenditure Review Page 35 PERFORMANCE ASSESSMENT: REVENUE 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 i s largely attributable to the removal o f external tariff differentials between Zanzibar and Mainland Tanzania, particularly since the closing o f the so-called “Zanzibar route” through improved tax administration by TRA. This collapse represents a major stress on Government’s capacity to finance its expenditure program. The secular decline was interrupted in 1999/00, when a combination of 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 taxation of revenue and profits from clove exports. Other measures could include curtailing exemptions and improvement in tax administration, particularly in the tourism sector. Figure 4.3: Revenue (% o f GDP), FY92-FY02 ~ 35% 1 30% , 25% 10% I 0% ~ FY92FY93FY94FY95FY96FY97FY98FY99FYOOFYOlFYO2 4.20 Variability o f 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 Government expenditure planning highly 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 o f sources. One of the main reasons for the high variability o f revenue i s the fact that Zanzibar collects most o f its revenue from indirect taxes, 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. I t i s interesting to note that transfers from the Union Government have to a certain extent offset changes in other sources of revenues. Diversification o f the tax base through the introduction of new domestic taxes such as VAT on petroleum and property taxes would help to reduce i t s 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 o f the Zanzibar economy, it i s likely that revenue variability will remain a key feature o f the economy. In order to be able to minimize the negative effects o f 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 o f central government transfers and donor resources in a way that would compensate for fluctuations that have domestic origins. 25% .I= p" Figure 4.4b: Indirect Taxes by Source ( O h o f GDP), FY92-FY02 I 12% 10% 1 +Import duties 8Yo 6 9'0 ~ -Other taxes - i- Excise duty 4y o , -Sales TadVA' (Local) 2% 0Yo  Zanzibar Public Expenditure Review Page 37 Figure 4 . 4 ~ :Other Revenue by Source (% o f GDP), FY92-FY02 9Yo 1 8% 7Y O .. .m.. . Contrib.fro 6% Pub.Enterprises 5 ?"o 4% 3% 2% 1Yo 0% 4.2 1 Inadequacy o f 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 revenue effort i s in line with international data from small island states which indicate that these have typically larger governments and higher revenue effort than most continental countries with similar economies. In the case of Zanzibar, there i s the additional caveat that GDP figures could be severely underestimated. I n spite of the apparent high tax effort, public revenue remains highly inadequate relative to the overall resources needed to successfully implement the Zanzibar poverty reduction plan (ZPRP) and to realize Zanzibar's Vision 2020. Zanzibar's tax base remains very narrow, given i t s negligible manufacturing sector, significant tax exemptions, and volatile mono-crop agricultural production. In these circumstances, expansion of Zanzibar's tax base needs to 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, presumptive taxation of the informal sector, and reduction of the multiplicity of taxes are important aspects of the strategy to reduce tax evasion and bring the 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 encouraging the production of a variety of other spices, high value 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 investment allowance, exemption from export profit tax, and tax holidays ranging from 2 to 5 years to exemption from import duty and VAT on the importation of capital goods, raw materials, and the personal effects of expatriate employees. Similarly, ZAFREZA administers a number of fiscal incentives to promote and facilitate private sector investment in export processing zones (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 received by the investors, and exemption from local taxes on goods and  Zanzibar Public Expenditure Review Page 3 8 services purchased or hired locally for use in EPZ enterprises. The main objectives o f the fiscal incentives regime are to (1) protect infant domestic industries, (2) attract foreign private investment, and (3) promote export-orientedmanufacturing and trade. 4.23 In the past three years, exemptions from customs duties alone amounted to about 19 percent o f actual annual customs collection (Table 4.5). Approximately 44 percent o f customs duty exemptions are those granted under the Zanzibar Investment Promotion Agency (ZIPA) Act, The other major beneficiaries o f 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 i s conservatively estimated by TRA to be in the range of 10-20 percent of total tax revenue collected each year. Table 4.7: Zanzibar Customs Duty Exemptions by Category o f Beneficiaries (YOo f Total) 4.24 Although these exemptions have to a limited extent helped to attract foreign , ~ have been subject to abuse and investment and encouraged some activities for e ~ p o r t they therefore have eroded the potential tax base. One of the major problems has been that there are several autonomous institutions that deal with the approval and management o f fiscal incentives. These include the Ministry of Finance and 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 collection agencies (TRA and ZRB). Apart from weak coordination and duplication o f 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 o f operational investments, employment and, foreign exchange yield. As of February 2002, only 103 projects promoted by 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, compared to the prospective employment figure of 696 Zanzibaris. There are now seven operating enterprises in the Export Processing Zone (EPZ), nine more are expected to be established, and five have recently closed or shifted. I n addition, most of the institutions dealing with the promotion o f investment, trade and tourism such as ZIPA, ZAFREZA, and ZTC s t i l l depend on subsidies from the central government, in addition to being allowed to retain their collections from licenses and fees. 4.25 Therefore, a thorough review and rationalization o f the exemptions regime i s imperative to counter abuse and improve revenue collection. RGoZ has already recognized some o f 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 processing zones doubled from $1.4 million in 1997 to $3.2 million in 2001. The total capital investment o f operational projects promoted by ZIPA and ZAFREZA stood at $ 1 14.5 million and $67.2 million respectively in 2002.  Zanzibar Public Expenditure Review Page 39 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 i t s investment promotion strategy to create a more favorable investment climate. I t should be emphasized that, based on international experience, generous taxation incentives per se do not attract large investment flows. I n the case o f Zanzibar and similar economies, both the Zanzibar Chamber o f Commerce and the Zanzibar Employers Association underlined that it i s 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 i s 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 loopholes for abuses of the system, such as the selling o f exemptions by those who qualify. Improving Tax Administration, 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 considerably and declined from 1999/00, partly reflecting the impact of the terrorist attack 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 of the clove sector starting in 2000/01. Until recently, annual contributions o f public enterprises to government revenue were set by the Ministry o f Finance and Economic Affairs (MoFEA) on a presumptive basis to counter the tendency of parastatals to under-report their 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 contributions to the Ministry of Finance and Economic Affairs. RGoZ has now changed this policy in favor of allowing public enterprises to operate commercially and be able to pay dividends. 4.27 There i s an opportunity to improve tax administration in Zanzibar and efficiency in revenue collection. As already noted, the RGoZ and URT Governments need to consider extending the Tax Administration Project (TAP) to cover Zanzibar. Some o f the issues that would need to be addressed include a review o f ZRB’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 upward trend 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 potential of increased tourism activities, RGoZ needs to have a better handle on the taxation o f this sector. A recent survey o f tourism earnings has been carried out and could be used for an improved estimate o f the tax potential and enhanced revenue generation from tourism. One way to maximize revenue from this source could include clamping down on tax  Zanzibar Public Expenditure Review Page 40 evasion and the introduction o f better tax handleshnnovativeways o f collecting revenue in the face o f the growing popularity of package tourism. In particular, special focus will need to be directed toward enhancing local capacity to offer quality services to tourists. RGoZ i s contemplating the 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 i s conscious o f the experience elsewhere, e.g., Poland in 200 1, where the introduction o f new regulations concerning the taxation of package tours tended to cause an increase in the price charged by tour operators and ushered in a decline in booked turnover. I n addition, a greater and more coordinated effort to collect more revenue from other sources and from petroleum products i s estimated by TRA to have the potential to raise revenue by as much as 10 percent. Improved coordination between RGoZ and the Union Government on fiscal issues would also go a long way to strengthen the Union and improve the predictability of resources on both sides of the United Republic of Tanzania. Clamping Down on Tax Evasion 4.28 As in many other countries, tax evasion and tax avoidance are widespread in Zanzibar. Tax evasion i s believed to be particularly high in the hotel sub-sector, and for petroleum products, which are subject to widespread smuggling. A quick comparison o f the collection and scheduled rates for some tax categories does point to significant erosion of the tax base via evasion and exemptions. For example, the collection ratio of VAT on imports for 1999/00 and 2000/01 i s about 14.3 percent and 12.3 percent respectively, compared to the applicable VAT rate of 20 percent, partly due to the exemption of imports by Government and for investment purposes. Similarly, the ratio of import duty revenue to imports c.i.f. over the same period i s 21.0 percent and 14.3 percent, against the unweighted-averagescheduled tariff rate o f 20 percent. 4.29 Officials o f TRA-Zanzibar also indicated that smuggling and mis-declaration and under-valuation o f 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. I t i s high time for RGoZ to consider extending pre-shipment inspection and valuation to Zanzibar as one measure to curb cheating. RGoZ i s also encouraged to consider paying taxes so as to close some o f the loopholes and to review the penalty rates in order to be more deterrent.  Zanzibar Public Expenditure Review Page 41 5. EXTERNAL DEVELOPMENTASSISTANCE RESOURCE EXTERNAL INFLOWS TO 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 development assistance that continued to flow, including the NGO support, represented a significant share of resources available to the Government for i t s development activities. For some time now external flows have been quite 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$30 per capita on the Mainland. Over 50 percent o f the development assistance in 2000 and 2001 was directed to the social sectors, with most o f the remainder being used for the development of infrastructure. Taking the various sources o f flows together, it i s not clear that aid flows have fallen drastically in the aggregate. Total support during that period came from the many non-OECD donors that stayed on, together with the 4.5 percent subvention from the URT which also started that year. Table 5.1 below summarizes the trend o f the external resource inflows to Zanzibar for the period 1997-200 1. Table 5.1 External Development Assistance 1997-2001 (US$ million) Source: MoFEA, Zanzibar and ZPRP 5.2 I t will be recalled that 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. I t will also be noted that, significant data concerns arise relating to commitments and disbursements of aid. An improved aid coordination mechanism must be established, not only to improve data reliability, but also to ensure that development projects are implemented and monitored in a coherent, timely and effective manner. This i s particularly important if the donors are to return to Zanzibar, in a full way, to assist in the implementation o f the ZPRP in a coordinated and orderly manner. A number o f donors are waiting for the Government to lead the way in this process. 5.3 As a process, Zanzibar secures external resources through a number o f means. First, sectors may approach donors directly or ask the External Finance Department to negotiate on their behalf. Second, the External Finance Department may initiate the negotiations with  Zanzibar Public Expenditure Review Page 42 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-funded projects i s not undertaken in any structured manner and donors may undertake their own appraisals, but the Government has no equivalent independent mechanism. Sector ministries are encouraged to appraise programs, but have noted that they are not technically equipped to do so, while the Commissioner of the External Finance Department makes the final recommendation on all external projects. COORDINATION, OF EXTERNAL AND MANAGEMENT MONITORING RESOURCE INFLOWS i ) The Current Status 5.5 Currently there are about 57 projects/programsrecognized by the Ministry o f Finance and Economic Affairs in Zanzibar (as recorded in the 2002/03 budget), although the precise number i s subject to verification as part o f the ongoing database exercise. Based on known projects, infrastructure accounts for over 40 percent of all recorded development assistance to Zanzibar-this i s 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 largest recipients of external resource inflows. 5.6 The Government has 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, if not modalities, of desired assistance. This i s particularly the case for the development partners who are recently returning to Zanzibar, as they have been encouraged to use the ZPRP as their main point of engagement. The Vision 2020 i s thought to be too broad and with insufficient prioritization to serve as an indicator for external assistance. ii) The Monitoring o f External Resource Inflows 5.7 The recording of external flows i s the responsibility of the External Finance Department under MoFEA. A database i s maintained, but it i s 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- financed projects, disbursement applications and notices are sent directly to the implementing agents and are not always copied to MoFEA. Direct payments by donors to the providers o f goods and services are particularly difficult to capture. Where donor reporting i s applied, there i s 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 recently established, and regulations are being developed. 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 i s often not the case. 5.8 Physical monitoring o f project activities i s 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 i s no financial reporting conducted during these monitoring missions,  Zanzibar Public Expenditure Review Page 43 while the expertise and staffing o f the Department restricts the scope and degree of monitoring activities. The External Finance Department provides data to the Budget Section for the preparation of the budget, but acknowledges that it i s incomplete. Grants received from Union projects, 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, meaning that the Development Budget estimates 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 i s effectively no system in place through which external resources can be projected although sector ministries are requested to file returns on disbursements for retrospective recording-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 o f all resources available, thereby jeopardizing the consistency of domestic expenditures with external assistance, e.g., in ensuring that recurrent resources are set aside to safeguard the sustainability o f aid-financed projects. 5.10 This by-passing o f the central budgeting system also means that the External Finance 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 inefficiencies in budget execution, as resources available to the Government are not brought to the attention o f the MoFEA. In a situation in which counterpart contributions are required, this can affect the smooth implementation o f project activities. With regard to the ZPRP priorities, the ability of line ministries to source and negotiate their own grant funding, before presenting the proposals to MoFEA for endorsement, suggests undermining the overall prioritization in resource mobilization that the ZPRP requires. As such, there i s no guarantee that the project portfolio will reflect ZPRP priorities. 5.1 1 There i s a clear fiduciary risk in the current arrangements, when neither the MoFEA (External Finance) nor the Accountant General i s informed or involved in the management o f external resources. This particularly relates to the current arrangements for project procurement which, by the External Finance Department’s admission, are of serious concern. Until a Central Tender Board i s established, ministries will undertake these activities themselves with few external checks and balances. Sectors may comply with the donors’ procurement procedures but the main concern here i s the by-passing o f MoFEA to record the 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 o f the External Finance Department (EFD) to participate in this work. By its own acknowledgement, the EFD has no expertise in this area. This i s further exacerbated by the Department’s lack of awareness o f 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 implementation of the development budget. This also clearly undermines the ability to perform any financial monitoring. 5.12 The new Access-based database which i s currently being designed should provide a basis for addressing many o f these problems o f the debt position and i t s management. I t i s  Zanzibar Public Expenditure Review Page 44 intended that it will provide full information on the funding o f projects and will be accessible to a wide range o f Government institutions. More important, however, i s the need to ensure an improved flow of information, based on an improved institutional structure, so that data can be made available for recording, accounting, and monitoring purposes on the debt situation. 5.13 I t i s clear from the aforesaid that the exchange of information between the MoFEA and executing ministries i s not effectively organized or coordinated, meaning that the complementarity between Government and foreign resources cannot be assured. Given that line ministries may face a reduced allocation o f domestic support if the full extent of external assistance to them were known, there may be added incentives for project-executing agencies to delay or avoid reporting to the MoFEA. Similarly, NGO-executed projects, 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 o f Government resources both within and across sectors. I n a well organized system, NGO resources are complementary to Government efforts. 5.14 I n most cases, many development partners include a Zanzibar component in their assistance to the United Republic and this i s 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 o f loans, disbursements by the donor to the project are notified to the URT but not to the RGoZ. This presents problems not only for the MoFEA, which i s unaware o f fimds provided directly 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 implementation of the Joint Finance Commission (JFC) will solve part of 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 alternative options 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-level mechanism for reviewing and appraising all project proposals. Government institutions starved for resources may also accept, without question, the offer of project assistance as this allows them to ‘go-project’ their recurrent expenditures. The result i s that externally supported projects will not have their desired impact on poverty; rather they will be used to maintain the day-to-day operations of Government. This i s particularly significant, and of concern, for loan-financedprojects. Currently, debt management i s mainly the responsibility o 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 sustainability analysis. What i s really required in Zanzibar, in this regard, i s a National Debt Strategy. iv) Suggestionsfor Future Improvements 5.16 The Ministry o f Finance and Economic Affairs needs to take responsibility for establishing and institutionalizing a Government-wide aid coordination system to ensure full recording and accounting of 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,  Zanzibar Public Expenditure Review Page 45 Budget, and Planning); and, between RGoZ 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 refined and to become operational as soon as possible. I t i s important that the database be viewed in the context of this wider system for aid coordination to ensure that information flow i s 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 o f an institutional mechanism for information sharing and reporting, because the creation o f a database will not by itself guarantee improved transparency in the use of external resources. Currently the MoFEA i s doing its best, but may need to become more proactive in seeking out information and establishing routine data systems, both for improving financial accountability and for ensuring that projects are implemented as intended. 5.18 More strategically, and related to the establishment o f a reporting and accounting system discussed above, it i s 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 dialogue with all stakeholders on the design of programs and the funding o f the ZPRP i s 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 provide guidance 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 need to consider other relevant ZPRP priorities. 5.19 Capacity development appears necessary at all levels, and the mix o f technical assistance and development capital that i s 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 External Finance Department which can then consolidate this work into a coherent framework for external support to the ZPRP. Sector mechanisms, coordinated and 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 returning partners and ensure a coordinated approach to the establishment o f sector strategy, and promoting capacity development.  Zanzibar Public Expenditure Review Page 46 6. PUBLIC DEBT AND LIABILITIES DEBT ZANZIBAR’S SITUATION i) Current Status 6.1 The statistics on the situation of Zanzibar’s debt are quite unreliable. Nonetheless, going by what i s 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. O f 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 foreign commercial sources (suppliers’ credit) amounted to US$7.1 million, accounting for 12 percent o f the total external debt. I t i s well understood that a l l of the multilateral debt i s guaranteed by the United Republic of Tanzania (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 problem has been the unreliability of the numbers. Measures are being taken to reduce the debt, but the capacity o f the Government to meet its debt service obligations i s s t i l l very limited. In fact, for quite some time now the Government has not been serving i t s external debt-apart from that portion which i s guaranteed by the URT. There i s a need to develop a National Debt Strategy to guide the Government in i t s borrowing, and in making repayments which will be in line with budget operations. 6.3 By the end of June 2001, Zanzibar’s total debt was estimated to 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 Fund has 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. Of the 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 multilateral and US$8.5 million (or 15 percent) i s from bilateral sources. The remaining US$7.1 million (or 12 percent) o f debt i s from commercial sources. As stated above, the entire amount of the multilateral debt i s guaranteed by the Union Government. The RGoZ guaranteed both the bilateral and the commercial debt.  Zanzibar Public Expenditure Review Page 47 6.5 Since 1995/96, the Government has not been able to honor i t s bilateral external debt obligations. The stock of debt, together with scheduled debt service, i s therefore currently building up. In recent years, the Government has been allocating a token amount (less than 0.1 percent o f expenditures) from i t s budget for payment o f debt. Even this meager amount has not been paid, due to severe budgetary pressures. Table 6.1 below summarizes the external debt position. Table 6.1: Zanzibar’s External Debt (June, 2002) Creditor Amount (US$ Mil) % Multilateral 42.7 73.0 Bilateral 8.5 15.0 Commercial 7.1 12.0 6.6 Zanzibar’s contribution to the servicing o f i t s external debt has been limited over the past few years, mainly because of disagreements over financial responsibilities with the Union Government. Zanzibar’s constitution precludes financial contributions to the Union until the Joint Finance Commission (JFC) has been established and 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 Fund had been paid by December 2001. Table 6.2 below summarizes the situation regarding Zanzibar’s domestic debt by June 2001. Creditor T.Shs. billion YOo f total Union Government 12.7 57% Zanzibar State Trading Corporation 2.7 12% Zanzibar Ports Corporation 0.6 3% People’s Bank o f Zanzibar 7.2 32% Total 22.2 100% Social Security Fund/pensioners dues 4.7 6.8 However, by June 2002, the situation had deteriorated further. As Table 6.3 below shows, by June 2002, the domestic debt had increased from T.Shs. 23.2 billion (excluding the T.Shs. 4.7 billion owed to ZSSF) of June 2001 to T.Shs. 30.4 billion. Out o f the total domestic debt of T.Shs. 30.4 billion, T.Shs. 13.6 billion (45.0 percent) i s 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, accounting for 20.0 percent o f the total domestic debt.  Zanzibar Public ExDenditure Review Pane 48 "41 ' " ~~ Table 6.3: Zanzibar's Domestic Debt (June, 2002) ~~~ ~~ Creditor Amount T.Shs. Bil. Union Govt. 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 need to be put in place to prevent the recurrence o f public arrears. A central payment and Local Purchase Order (LPO) system seem crucial to ensure better expenditure and commitment control. I n 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 parastatal debt i s not yet verified, as the Government's Department o f Public Investments has not carried out an audit to get a clear picture of these debts. A good number of the parastatals however, claim that they are not in debt as such, but rather that it i s 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 i s recommended that an audit o f the parastatals be carried out to verify the position. iv) Preliminary Debt Sustainability Assessment 6.10 Analysis o f the available numbers indicates that, the situation notwithstanding, Zanzibar's debt i s probably not unsustainable. For instance, the total debt stock i s only at the level o f about 50 percent of GDP, as compared to over 90 percent of the URT. It appears Zanzibar i s not able to service its debt mainly because the payment o f Personnel Emolument (PE) alone takes up so much of i t s budget. PE i s estimated at an average of about 60 percent of the entire expenditures budget. Furthermore, i t s debt service ratio i s slightly less than 100 percent, a fairy comfortable position when compared to the HIPC (heavily indebted poor countries) threshold of 150 percent. In addition, rescheduling would also lower the NPV o f 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 i s also sustainable. However, given the unreliability of the numbers, it i s recommended that a detailed DSA be conducted in order to get a clear picture o f the sustainability of the debt position in Zanzibar. ii) The Contracting of Debt 6.11 Zanzibar has no authority to contract external debt (especially multilateral debt) as this i s 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 i s the case, the RGoZ i s involved in these negotiations. This function i s primarily performed by the External Finance Department o f the Ministry o f Finance and Economic Affairs. There are also cases o f loans contracted exclusively for Zanzibar and, once again, the RGoZ i s represented throughout these negotiations. However, RGoZ can not singly negotiate any Paris Club debt rescheduling or 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 Public ExpenditureReview Page 49 and, ipso facto, it i s responsible for servicing this 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 Organization of Petroleum Exporting Countries Fund (7 percent). Of the remaining debt not guaranteed by the Union, nearly all was contracted with the Government o f 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, but this i s yet to be finalized. 6.13 As stated earlier, since 1995-96, the Government has not serviced i t s external debt obligations and has not fully recorded interest arrears or penalty interest. The debt stock may therefore be somewhat higher than that recorded by the RGoZ Accountant General’s office. There i s no debt strategy for repaying. or negotiating the rescheduling or cancellation of debt. Further, a Debt Sustainability Analysis has not been conducted for Zanzibar. The RGoZ Ministry o f Finance and Economic Affairs, chiefly through the External Finance Department, i s able to directly mobilize grant funding, and has made use o f this facility to fund approximately one-third of its development budget-the remainder being foreign loans and a small domestic counterpart contribution component. 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 uncoordinated way. The unit at the Accountant General’s Office i s very weak and ineffective. The External Finance Department of MoFEA plays a role in the negotiations and the contracting of debt. The Department maintains a database, but it i s considered quite incomplete, since project financing has been very difficult to capture. Physical monitoring o f project activities i s the responsibility o f the Sector Policy and Research Department o f MoFEA. In addition, this department provides data to the Budget Section, but acknowledges that it i s also incomplete. Grants received from the Union projects, for example, are often excluded from the estimates. 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 Development Budget estimates are incomplete. 6.15 This by-passing o f the central budgeting system also means that the External Finance 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 o f Tanzania (Zanzibar Branch) maintains a separate data set o f Zanzibar’s debt position. In general, the situation i s very weak and uncoordinated. MoFEA acknowledges that the situation needs to be improved. In view of this situation, it i s advisable to create an effective Debt Management Unit in MoFEA to coordinate a l l debt matters. The Bank of Tanzania (Zanzibar Branch) has offered assistance, especially in the area o f using the computerized debt management system (CS-DRS).  Zanzibar Public ExPenditure Review 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 recommended by many countries. If such a discipline does not exist, then any official in the Government will contract debt, leading to a state of confusion whereby the statistics will 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 i s only MoFEA that has powers to contract debt on behalf o f the Government. The problem however, has been the ineffectiveness and poor enforcement o f such a requirement. There are s t i l l 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 i s a contingent liability to the Government. In view o f this situation, it i s recommended that the current legal status be reviewed to provide a solid basis for strengthening enforcement. iii) The Need for 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, including an assessment of the repayment capacity and repayment schedule. Such a strategy will also ensure that debt service i s in line with the operations of the budget. In the absence o f such a plan, borrowing and repayment will 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 i s strongly recommended that the Government should, as soon as possible, prepare a Zanzibar Debt Strategy (ZDS). iv) The Needfor 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 picture of the debt situation-both external and domestic-it i s difficult for the Government to know i t s obligations so that it can, inter alia, prepare a realistic debt repayment schedule. The MoFEA i s considering implementation of a debt census. It i s thus being recommended that a Zanzibar Debt Task Force be constituted to bring the debt situation into a reliable and dependable position. The Bank o f Tanzania (Zanzibar Branch) has offered its assistance in this area. Zanzibar Public Expenditure Review Page 5 1 PART 3: PUBLIC EXPENDITURES  Zanzibar Public Expenditure Review Page 52 7. ZANZIBAR’S PUBLIC EXPENDITURES i) Introduction 7.1 Zanzibar’s public spending i s characterized by a high level of recurrent expenditures and a very low level of capital expenditures. Within recurrent expenditures, the bulk o f the spending i s on wages and salaries for an overstaffed bureaucracy. A huge wage bill, which the Government must clear before making any other expenditures, makes the overall demand for funds very inflexible. Also, the Government operates a ”cash budget,” Le., it spends only what i s 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 receipts translates into a variable and uncertain availability 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 o f the barrel just 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 maintaining and operating the existing ones have become a casualty in the process. In order to reduce poverty and induce growth, the structure and management o f public expenditures will need to change. The Government i s making efforts to realign spending priorities with financing priorities identified in the ZPRP. However, much more remains to be done. The Government will need to move on three fronts: (i) creating space for increased developmental and 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) improving budget 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 doubled from 17 percent of GDP in FY95 to 33 percent of GDP in FYOO (Table 7.1). FYOO was not a normal year, as expenditures spiked due to elections-related expenses. Since then, expenditures have returned to more normal levels. During FY02, aggregate central government expenditures were around T.Shs. 50 billion, which i s about 20 percent of the GDP. Most o f the increase in the 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 expenditures on wages and salaries have come to stay and are s t i l l on the rise.  Zanzibar Public Expenditure Review Page 53 iii) Expenditures by Economic Classifcation Recurrent versus Development Expenditures 7.3 I t i s not the aggregate level, but the structure, of expenditures that i s worrisome in Zanzibar. Almost a l l 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. I n comparison, recurrent expenditures in Mainland Tanzania in FY02 were 80 percent o f 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 comparison would even be starker. A high level of recurrent expenditures has led to an almost complete squeezing out o f development expenditures-thus, Zanzibar’s development expenditures were less than three percent o f total spending in FY02. I t i s interesting to 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 i s all the more surprising since, during this period, foreign assistance was significantly reduced and it seems that there was an opportunity for domestic spending to substitute for foreign assistance. Table 7.1: Government Expenditures as a YOo f 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-Making Bodies 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 o f Finance and Economic Affairs, Zanzibar Subventions, Contributions, and Subsidies 7.3 Expenditures on subsidies and transfers were significantly reduced during the first half o f the 1990s and have since remained at a fairly low and stable level of around 1 percent o f GDP. There are practically no transfers to parastatals shown in the budget. However, there are hidden transfers, as capital investment for many o f the parastatals i s 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 contributes to the total of non-performing loans 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 Institute of Financial Administration, etc. Table 7.2 lists  Zanzibar Public ExDenditure Review Page 54 selected subventions that have been provided in recent years. I t should be noted that, while the level of cash subsidies i s low, there are sizable in-kind subsidies, mainly in the form of subsidized water and power supply. Table 7.2: Selected Subventions in Recent Years (current T.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 24 1 289 343 298 Chake, Wete and Mkoani Town Councils 132 152 127 202 204 179 Zanzibar Revenue Board 0 0 588 834 506 500 ZSSF 0 59 44 36 48 48 Zanzibar Free Port Authority 0 0 55 76 104 150 Source: Draft budget estimates for various years, Ministry o f Finance and Economic Affairs, Zanzibar Debt Sewicing 7.4 I n most years since FY94, Zanzibar has not paid any interest on i t s external debt. Nor were any expenditures made between FY96 and FY98 on servicing domestic debt. Table 7.3 provides a summary o f domestic debt servicing since FY98. Quite clearly, it has been erratic, and was the lowest spending priority o f the Government until recently, when the Government started devoting more resources to it. The Government can afford to treat debt servicing as a discretionary expenditure mainly because the entire domestic debt i s owned by parastatals. Table 7.3: Summary o f Domestic Debt Servicing (T.Shs. Million) FY98 FY99 FYOO FYOl FY02 FY03 Budget Actual Actual Actual Actual Approved Interest 0 50 1068 59 1 993 1333 Debt Redemption 0 198 500 0 4000 1444 Total 0 248 1568 591 4993 2777 Source: Draft budget estimates for various years, Ministry o f Finance and Economic Affairs, Zanzibar iv) Expenditures by Functional Classification 7.5 Table 7.4 shows the trends in central government expenditures by purpose as a percentage o f GDP over the decade of the 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 o f the increase in expenditures during the second half o f the 1990s occurred in general administration. From an economic point o f view, the expenditure structure has changed from more productive to less productive activities. The challenge before the current Government i s to restore expenditures on social and economic sectors while cutting down the size o f the general administration.  Zanzibar Public Expenditure Review Page 55 Table 7.4: Government Expenditures by Purpose (% o f 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 o f 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 o f Finance and Economic Affairs, Zanzibar Key Issues for Attention Expenditures on Salaries and Allowances 7.6 A large part of the recurrent expenditures i s claimed by 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 expenditures had steadily grown in the second half of the 1990s and have continued growing over the past two years, mainly due to increases in allowances, increased employment o f teachers, and salary increases for special departments. Compared to Zanzibar, Mainland Tanzania spends about 35 percent o f i t s recurrent budget on salaries. A large wage bill makes expenditure management 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 i s a problem when revenues display high variability, since it i s not possible to 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 maintaining them. Such neglect o f development expenditures can have serious consequences for growth and poverty reduction in Zanzibar. The Government recognizes that there i s an immediate need for higher development expenditures, 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 i s likely to move expenditures further away from operations and maintenance spending. There i s an urgent need to create space in the budget for development expenditures by first controlling the wage bill and then cutting it down over the medium term. Chapter 8, dealing with civil service, goes into the details of how this could be done.  Zanzibar Public Expenditure Review Page 56 Volatility of Aggregate Expenditures 7.8 Aggregate expenditures in Zanzibar mirror the high degree of volatility in revenues which, in turn, are highly susceptible in terms o f trade shocks. High volatility in expenditures results in disruption in public investment and has associated costs o f adjustment. The long- term solution to the problem o f resource volatility i s a diversified tax base. In the short- to medium-term the key challenge, however, i s 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 during times o f 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 could consider: 0 Creating an accounting-only design, 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. Restrictions would be placed on drawing down the fund’s resources for expenditures. Any draw-down o f government deposits, including the fund account would appear as deficit financing. Transfers to and from the fund could be explicit line items in the budget. 0 Creating an extra-budgetary fund as a separate entity with authority to undertake off- budget expenditures. 7.9 The first option has the advantage that it would fit in well with the existing budgetary procedures and would not need the creation of new institutional mechanisms. Such a revenue stabilization fund should serve to smoothen short-term volatility. However, given the difficulty o f distinguishing between short term volatility and secular revenue changes and the likely capacity constraints for administering such a fund, taking a more conservative approach to estimating revenues might be the most straight-forwardway to deal with revenue variability.  Zanzibar Public Expenditure Review Page 57 Box 7.1: Experience with 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 o f resource prices. An additional objective has been to save revenues from non-renewable resources 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 decisions to control expenditures or deficits outside the fund. Otherwise the Government could borrow or draw down its other assets outside the fund. Empirical evidence (based on a study o f 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 establishment of funds in four cases. Source: Stabilization and Savings funds for nonrenewable resources, IMF 2001 Fiscal Risks 7.10 Fiscal risks refer to unaccounted obligations/liabilities of the Government which are not provided for 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 activities financed under this head were 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 o f Finance and Economic Affairs, Zanzibar 7.1 1 The reserve i s 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 probability of 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 information about them, or may just 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 Government that pose fiscal risks to the government finances and need watching.  Zanzibar Public Expenditure Review 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) even though they are government liabilities. Major among these are outstanding government contributions to ZSSF (T.Shs. 1.8 billion) and cleanup liabilities o f collapsed parastatals (T.Shs. 2.9 billion owed to PBZ and 271 million to retrenched employees). The Water Department owes T.Shs. 2 billion to the Fuel and Power Corporation towards purchase of power for running i t s water supply pumps. Apart from these known arrears, ministries also acknowledged arrears in the nature o f unpaid bills of supplies already received, but their total volume could not be ascertained. Outstanding dues of ZSF&PC to TANESCO are T.Shs. 34 billion. 7.13 Arrears and accrued liabilities are now receiving Government attention. Outstanding dues o f ZSF&PC to TANESCO are being cleared in a phased manner. While the clearing o f old dues i s no doubt the first priority, it may not be sufficient. The Government will need to improve cost recovery in the water and power departments. I n the water department, fresh dues are accumulating at the rate o f T.Shs. 100 million per month due 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. Contingent Liabilities 7.14 Contingent liabilities are those that become government obligations,if a particular event occurs. A significant source o f such liabilities could be Zanzibar Social Security Fund (ZSSF). I n 1998, the Government established ZSSF, whose function i s to (1) receive and manage social security contributions from its members, and (2) pay old age, maternity, invalidity, and medical care benefits to them. All government employees, including those of the parastatals and local authorities, and private employees can become members o f the Fund. According to the Fund provisions, the employer i s expected to contribute 10 percent against the employee contribution of 5 percent o f 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 billion to the Fund. Apart from this, many parastatals, some of them that are either closed or are being closed, have yet to deposit their contribution with the Fund. The Fund could not confirm the amount of outstanding dues from such parastatals, but they are considered, for all purposes, to be government liability. Contributions from municipal bodies are also stated to be irregular. 7.15 Fiscal risks associated with the functioning of the Fund appear to be significant. The Fund i s expected to start paying pensions and other service benefits to 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 s t i l l be liable for providing benefits proportional to the period of service in excess o f five years. A broader issue i s that it i s not entirely clear that the Fund will remain fully funded in the long term. Apparently no actuarial calculations were carried out before the Fund was instituted to arrive at the package Water i s supplied free to residential consumers. The Water Department recovers about 12% of i t s costs. In FY03, at the end of November, the department’s revenue was T.Shs. 49.2 against running cost o f 382 million. Power tariffs are quite low too. Against a cost o f T.Shs. 130iunit only about T.Shs. 28iunit are recovered. Power generation i s very costly in Pemba, where it i s produced from diesel generators.  Zanzibar Public Expenditure Review Page 59 o f benefits and contributions. The Fund also needs expert advice on where and how to invest i t s funds. I t has put about T.Sh. 1.8 billion in real estate and i s 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 financial position of the People’s Bank o f Zanzibar i s 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 i s the main borrower, which i s under liquidation now; upon liquidation i t s 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. EXPENDITURES ON PRIORITY SECTORS 7.17 The ZPRP i s expected to provide the framework for sectoral allocation of expenditures. A budget that i s well-aligned with, and supports, the ZPRP, should be able to provide adequate financing in medium-term to 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 i s not clear, nor i s the pace at which expenditures would adjust in various sectors to reach the targeted levels. Therefore, ensuring alignment with the ZPRP boils down to seeing if finances available to various ministries overseeing priority areas are going up and if those for non-priority areas are going down relative to total expenditures. 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 i s a result o f increasingly reduced 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 ministries have gone up, which i s the right direction for adjustment. Table 7.6: Expenditure Shares by Sectors (YOo f 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 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 Source: Ministry o f Finance and Economic Affairs, Zanzibar  Zanzibar Public Expenditure Review Page 60 7.19 The Government i s encouraged to identify expenditure items in line with ZPRP priorities, which would receive priority during budget allocation and would be protected from 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-reducing expenditures would have to be defined quite narrowly in order to make effective protection feasible. Since, in a system in which certain expenditures are 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 and thus automatically also impact on priority expenditures. Expenditures on Education 7.20 Zanzibar follows a 7-3-2-2-3+ system of education. Compulsory basic education includes seven years o f primary education after the age of 6, and three years o f junior secondary ed~cation.~ Basic education i s followed by two years of second cycle lower secondary (Forms I11 and IV), two years o f third cycle advanced secondary study (Forms V and VI), and three years or more of higher education. Basic and secondary stages o f education thus encompass 14 years o f schooling. 7.21 Access to basic education has slowly improved in 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 o f school-age children do not yet attend school. The number of new school-age children i s growing rapidly due to high fertility rates. Wastage rates are high, and the quality o f education i s generally not satisfactory. Access to secondary education i s s t i l l very limited and the GER i s below 40 percent. There are significant regional disparities in education indicators. North A, South Unguja, and all districts in Pemba have below-average GERs for basic education. Table 7.7: Participation and Internal Efficiency in Basic Education (YO) Enrollment Rates (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: Education Sector Status Report Zanzibar. 2002. Legend: Gross Enrollment Ratio (GER), Net Enrollment ratio (NER), Drop-out rate (DR), Repetition Rate (RR), Male (M), Female (F), Total Male and Female (MF), Female-Male parity index (F/M) 7.22 The state of education infrastructure i s poor. There i s an acute shortage of class rooms-the average pupilklassroom ratio i s 82 and the average number of shifts i s 1.69 streams per class room. ’The junior secondary education consists o f one year o f orientation, and two years o f grade study called Form Iand Form 11.  Zanzibar Public Expenditure Review Page 61 Aggregate Expenditures 7.23 I n 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 level slightly higher than Mainland Tanzania and comparable to other countries in the region. While, in terms o f GDP, aggregate education expenditures 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 improving and was around a reasonable figure of US$l0 in FY02, compared to about US$7 for Mainland Tanzania (and US$10 for Kenya in 1993, and US$7 for Uganda in 1997). Figure 7.1: Trends in Real Public Spending on Education Trends in Real Public Spending on Education I ~ 2000 , 2 1500 .- I1000 I f 8 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 education look reasonable in a comparative sense, but one must allow for the fact that private provision o f education beyond the pre-primary stage i s very limited in Zanzibar, necessitatinga higher level of public financing. Furthermore, there has been very l i t t l e external financing since 1995, which means that local resources have had to meet the demand for development expenditures. This demand was not met and there i s a serious backlog of capital works that need to be carried out. For example, about 500 school rooms are incomplete where the communities have met their part but the Government has not been able to meet i t s share. And finally, as noted above, education performance indicators s t i l l have large room for improvement. 7.25 Almost all education expenditures 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 been very limited too. In FYOl the Government spent only TShs. 7 million on school materials and equipment. I n fact, school materials and teaching equipment are now expected to be provided by the community. 7.26 All the factors discussed above point toward the need for a s t i l l higher level o f financing. The Government i s 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 i s proposed in the FY03 budget. The level i s now close to the ZPRP target o f a 20 percent budget share.  Zanzibar Public Expenditure Review Page 62 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: Shares in Education Expenditures (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 o f Finance and Economic Affairs, Zanzibar 7.28 It appears that administrative and non-education expenses went up after FYOO. It i s partly due to the addition o f 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 in the budget i s around 97 percent. The bulk o f the increased expenditures in the previous two years has gone to salaries and allowances. Key Issues for Attention 7.29 As noted earlier, government financing of the education sector remains inadequate in view of the ZPRP objectives. Yet, at the same time, resources for education have been going up in real terms and per student. This increase in resources i s not reflected in much-improved performance indicators, which points towards the need for improving internal efficiency. Indeed, within the existing resource envelope, it i s possible to achieve better results through restructuring expenditures. Non-Salary Expenses 7.30 Improving the quality o f education i s 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 direction o f adjustment in education spending i s in the opposite direction. O f all the additional resources made available in FY02, three-fourths went towards wage bill increases. There was very l i t t l e increase in other expenses; in fact, the Government still spends negligible amounts on school supplies. Mainland Tanzania has a target o f providing T.Shs. 10,000 towards non-salary allocations per pupil in primary education. The Government in Zanzibar should also aim to move towards this level o f allocation. Allocative Efficiency Issues 7.31 Government spending on education appears to have inequities across districts and to be biased in favor of urban communities (Table 7.9). In the year 2000, almost 70 percent of the Ministry o f Education’s budget went to the two urban districts of Zanzibar city and Unguja West. There i s also a serious imbalance in the number of classrooms between the rural and urban districts. There i s a shortage of teachers in the rural areas, while the urban area schools have surplus teachers.  Zanzibar Public Expenditure Review Page 63 T a b l e 7.9: Education Resources 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 Government practice of encouraging the communities to cover some operational expenses for schools and parents to provide textbooks also puts poorer households at a disadvantage.' 7.33 There i s a strong case for reviewing expenditure patterns on a geographical basis and reorienting it to become more equitable. From an economic point of view, spending more in communities that have poorer education outcomes would give higher returns at the margin. There i s reason to believe that the regional distribution of resources will get even more skewed unless the Government gives it i t s full attention. For example, communities can collect part o f the funds for the construction o f a school building and approach the Government to provide the balance. Once the building i s in place, the Government i s committed to opening a school in it. Such an approach would lead to a concentration o f schools, and hence public expenditures, in communities that can afford to collect funds for new school buildings. Pupil/Teacher Rat io 7.34 The average pupil-teacher ratio i s about 3 3 : l in primary schools and about 30:l in secondary schools. The ratio for primary schools i s very low compared to a targeted 45:l in the Mainland Tanzania. There i s 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 i s limited at around 2.3 percent. The Government should encourage private provision o f post-primary education, which will allow it to leverage public spending. Government aid to private institutions, subject to regulations, would be the right approach. Expenditures on Health 7.36 Zanzibar has a substantial infrastructure for health services delivery, consisting o f 103 Primary Health Care Units, four cottage hospitals, three other hospitals and one referral ' * According to the Ministry o f Education guidelines, parents are required to provide annual contributions at rates varying between T.Shs. 3000 for primary to T.Shs. 7,500 for Higher Secondary education.  Zanzibar Public Expenditure Review 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 Unguja Island and eight on Pemba. Apart from this, there are 38 paramilitary units which provide medical services to uniformed forces and their families and nearby communities. Despite this impressive network of facilities, health sector outcomes are not very encouraging. The infant mortality rate i s estimated at 75.3 per 1000, under-five mortality i s 107.6 per 100, and the maternal mortality rate i s 3 14 per 100,000 live births. Malaria i s rampant, and about 340,000 cases were reported during 1995. Figure 7.2: Trends in Real Public Spending on Health (1991 prices) I Trends in real public spending on health (1991 prices) 1200 i 2 1000 Million) , .s w 800 e 600 400 I B 200 0-1 , , Fiscal Year i Aggregate Expenditures 7.37 Aggregate public spending on 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 i s whether Zanzibar i s 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 i s Mainland Tanzania, where health expenditures were 0.6 percent o f the GDP, and more than nine percent o f the total discretionary spending in FY02, excluding development expenditures separately financed by donors. 7.38 FY02 saw a significant increase in the health budget, and spending improved to 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 i s below the levels achieved in early 1990s. The sector shows tell-tale signs of chronic under-funding. Health facilities are in a state o f disrepair, equipment i s scarce, and there i s an acute shortage o f drugs and medical supplies. 7.39 As in education, the structure o f spending in the health sector i s 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 resources too have been negligible; in fact, they were zero in FYOO and FYOl. Social welfare department was added to the health ministry portfolio in FY02, 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 o f different components o f expenditures within the health sector in recent years. Table 7.10: Expenditure Shares within Health ( O h o f 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 o f Finance and Economic Affairs, Zanzibar 7.41 Several comments are in order: 1, Administrative expenses (which include salaries and allowances) far outweigh the expenses on provision o f health services. 2. Preventive services receive a very small share o f the non-wage resources. 3. The expenditures do not appear to be equitable. In regional terms, Pemba, which has about 40 percent o f the total population and lags Unguja in social indicators, has been receiving less than 30 percent o f the total budget. Pemba’s share o f non-wage resources for curative services i s even lower. Considering that Unguja has much larger private provision o f health services, the situation appears to be even more unfair. Key Issues for Attention 7.42 Key issues include the following: 1. There i s a need for an increased overall level o f resources in the health sector. The ZPRP also targets a higher level o f allocation to health. 2. There i s an opportunity for improving intrasectoral allocations 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 o f donor assistance and needs an infusion o f resources. c. Spending a greater share on drugs and medical supplies. At present there i s an acute shortage o f drugs in PHCs (primary health care units) and hospitals. Absence o f drugs hurts the poorest the most.  Zanzibar Public Expenditure Review Page 66 ISSUES IN BUDGET MANAGEMENT Budget Outcomes 7.43 An effective system for budget management i s one that i s 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 effectiveness o f programs in achieving development goals; and 3 , Ensure operational efficiency 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 incentives for productive efficiency in government operations. 7.45 Judged against the desired outcomes mentioned above, performance of the budgetary systems leaves much to be desired in Zanzibar. There i s a perception that aggregate fiscal discipline i s being maintained because the Government operates a cash budget, but the perception o f fiscal discipline i s illusory. The Government has sizable arrears, has been neglecting debt service, and has been reluctant to make agreed upon contributions 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 i s an attempt to align broad spending patterns (at the level of ministries) with ZPRP priorities. However, through discretionary releases o f cash during the year, MoFEA has been significantly deviating from ex-ante allocation priorities. Performance on the last outcome i s 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 operational efficiency. Budget Estimates vs. Actual Outturns 7.46 A comparison o f actual outturns with budget estimates i s a good indicator of the overall performance of a budget management system. I n 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 of actual expenditures at the end of the fiscal year to the ones planned at the time of budget formulation. On the other hand, a wide deviation between estimates and actuals i s a sign o f one, or many, of the following weaknesses: (i) inaccurate revenue forecasting-most often the revenue i s 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 underspending in other categories o f expenditure; (iv) inefficient cash flow management. 7.47 Significant deviations between resources budgeted and actually made available have serious costs. First, they erode the overall credibility o f 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 Public Expenditure Review Page 67 programs. Third, they may lead to unintended reordering of priorities during budget execution, thereby leading to the abandonment o f the plan and an abrogation o f the legislature’s sanction. 7.48 Table 7.11 below shows the average difference between budgeted and actual expenditures in the past three years in Zanzibar. The deviations are o f the 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 availability o f funds, other than for salaries. Table 7.11: Differences between Budgeted and Actual Expenditures 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 % o f (T.Shs. as % o f Mill.) budgeted Mill.) budgeted Mill.) budgeted Salary and 88 19 96 20 278 22 Wages Development 243 79 547 80 663 77 Other Charges 78 1 40 464 70 412 56 Source: Draft budget estimates for various years, Ministry o f Finance and Economic Affairs, Zanzibar Note: Averages taken over all votes 7.49 Table 7.12 below l i s t s the major gainers and losers from the reordering o f priorities during budget execution in FY02. Table 7.12: Major Gainers and Losers during Budget Execution in FY02 Vote Category o f Budget Actual expenditure (Million (Million TShs.) TShs.) M a j o r gainers President’s Office, Regular .Administration Other Charges 510 989 Ministry o f 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 M a j o r losers Ministry o f Health Other Charges 2017 1401 Economic Brigade Salary etc. 2909 1527 Ministry o f Education Other Charges 1623 623 Ministry o f Communication Dev. (local) 2500 63 8 Ministry o f Communication Other Charges 1321 317 Ministry o f Education Dev. (local) 1100 184 Ministry o f Water, Environment etc. Dev. (local) 690 0 Ministry o f State Women & Children Dev. (local) 893 0 Total budget 66,333 49,763 Source: Ministry o f Finance and Economic Affairs, Zanzibar Key Issues for Attention 7.50 Zanzibar follows a fiscal calendar similar to Mainland Tanzania. It has a dual budget system in which recurrent and development budgets are prepared independently. The budget  Zanzibar Public Expenditure Review Page 68 preparation begins in mid-November o f each year and i s 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 negotiations with the MoFEA. The CFAA provides further details about the process of budget preparation. Budget Planning and Resource Allocation Budget Realism 7.51 The revenue and expenditure estimates are not realistic, as evident from the wide deviations between 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 bringing more information to 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 preparation process. 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 expenditure targets, and work out borrowing requirements, if any, more realistically. It would also help MDAs develop more realistic estimates of expenditures if the budget guidelines convey budget implications o f expected macroeconomic developments such as the inflation rate, growth rate, and any significant demographic parameters for different categories o f expenditures. Development o f a successful macro-fiscal framework, however, would depend partially upon how well resource requirements o f the ZPRP are quantified (costed out). 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-functioning budget system, information on budget execution would become available in a timely manner through routine reporting and monitoring, but this does not happen in Zanzibar. I n short-term the MoFEA will need to make special efforts to collect information on actual outturns, and more importantly on arrears, before initiating the budget process. 7.54 MDAs develop expenditure estimates on the basis o f the previous year’s estimates o f input costs rather than discussing expenditure needs in a policy framework. Discussions with the MoFEA on expenditure cuts are also in terms o f input costs, such as travel expenses, fuel for vehicles etc. This focus creates an implicit prioritization in terms o f 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, including the broad parameters within which the next budget must be prepared. Examples include the forecast level of revenues, indicative aggregate expenditure constraints, expected growth rate, expected inflation rate for expenditure projections, demographic trends, etc.;  Zanzibar Public ExpenditureReview Page 69 0 Indicate Government’s fiscal priorities for 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, contracting out etc.); 0 Seek submission of new policy proposals separately from existing ones. The key question concerns how new policies would be funded; 0 Seek realistic estimates from the ministries in terms of spending programs and policies; and 0 Ask ministries to indicate the 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 contingent liabilities in a systematic way. This would require systematic collection o f information on (i)arrears and (ii)outstanding liabilities because of pensions and gratuities to government and parastatal staff. A schedule o f payments to ZSSF for pension benefits in excess o f five years should be drawn. I n 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 because o f the debt o f collapsed parastatals. Integration of Aid into Budgeting 7.57 At present, recurrent and development budgets are prepared through separate processes of formulation and review o f estimates. As in many other countries, this separation has historical and institutional reasons. I t stems from the intuitive idea that the recurrent budget i s meant for financing ongoing activities with Government’s own resources; while the development budget i s for financing “economic development”-mainly in the form o f capital expenditures with donor support. The development budget was earlier prepared by the Ministry o f Planning, now merged with the MoFEA, while the recurrent budget was prepared by the MoF. There are three main problems 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 inconsistencies between the two budgets. Most common are cases in which recurrent expenses are not provided in the recurrent budget for new capital projects, or the same item of expenditure i s included in the two budgets. (ii) There i s insufficient scrutiny of investment projects included in the development budget, because they are largely donor-financed. (iii) The development budget may often be assigned a priority status and lead to cutbacks on operational expenses for existing programs. 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 competitive allocation decisions, particularly regarding donor supported projects, through this ”unified process,” because those decisions are made outside the budget process, but the  Zanzibar Public ExPenditure Review Pane 70 unification will serve an important purpose o f integrating all budget estimates by sectors in one place and allow better allocation of government’s own resources, since resources are fungible. 7.59 An important prerequisite for this unification o f processes would be that MoFEA has up to date information on all donor-supported projects, both existing and planned. This i s 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 used by Mainland Tanzania. Multi- Year Framework of Expenditures 7.60 The Government plans to move to a multi-year framework for expenditures (the Medium-Term Expenditure Framework, or MTEF). Three major advantages would accrue in moving to a multi-year framework: 0 I t would force attention to the long-run expenditure implications o f new policies and investment projects. 0 I t would help in programming cutbacks (say, phasing out programs, or cutting down the wage bill) over a period o f time. 0 At the aggregate level, it would provide MoFEA with a medium-term view of inter- sectoral demands for resources and allow planning in order to align the structure of expenditures with the ZPRP priorities in the medium term. 7.61 Based on cross-country experience, several prerequisites need to be put in place for a successful introduction of the medium-term framework. Some o f these have been discussed above. I n the absence of a macro-fiscal framework and cost estimates o f implementing the ZPRP, MDAs will not be able to develop meaningful medium-term plans. Further, such plans will be of limited value if they leave out donor-provided funds. 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-term plans would overstretch their capacity. 7.62 I n view of this discussion, while a medium-term expenditure framework will be a step in the right direction, it should be a medium-term goal to follow careful strengthening o f 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 capacity for planning. Budget Execution 7.63 The Country Financial Accountability Assessment (CFAA) has a detailed assessment 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 Public Expenditure Review Page 71 e Monitoring budget execution: The Government should move toward regular monitoring and publishing at least quarterly reports on actual expenditures, rather than relying on cash release statements. Controlling arrears: Arrears can upset financial plans and be a source of fiscal indiscipline. One of the main reasons for arrears, in the decentralized payment system o f Zanzibar, i s a lack of effective MoFEA monitoring and control of commitments made by MDAs. There i s a need to institute either a mechanical system for such control, or, if a decision i s made to go in for an automated Integrated Financial Management Information System, then such control can be effectively exercised through that system. Transparency Issues 7.64 Enhancing transparency at all stages o f the budget process, including budget formulation and resource implementation, budget implementation, and budget reporting i s one o f the key challenges for Zanzibar. Developing more open and inclusive budget processes will play a key role in strengthening strategic resource allocation and enhancing the 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 natural disasters and epidemics, there are no clearly laid out criteria defining eligible contingencies for withdrawal. The contingency reserve should meet only truly unforeseen obligations. The Government should plan to meet quasi- contingencies through regular votes. e In the past, large sums o f money have been spent on subventions titled “Special Presidential Vote.” An expenditure o f T.Shs. 7.764 billion in 98/99, and 5.168 billion in 00/01 was made on this account. The Government has reportedly dropped this sub- vote from FY02 on. e The MoFEA issues lump-sum monies to the ministries for wages and salaries. There have been instances when ministries have diverted them for other purposes. A civil service census as suggested in the Civil Services chapter of this report should be able to provide Ministry of Finance with better control. e The MoFEA operates a Micro-Projects Finance scheme which provides government support for community projects. There are problems of coordinating plans for new schools or hospital construction with the concerned ministry. Operation of 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 o f how many community resources were 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 Government should consider preparing a statement on government finances each year to accompany the budget, and the same should be placed before the legislature. This statement should include the following information: 1. Outstanding debt and debt repayment profile over the next five years.  Zanzibar Public ExDenditure Review Page 72 2. Details of contingent liabilities-guarantees, parastatal loans, disputed liabilities, etc. 3. Details of accrued liabilities other than pensions. 4. Major policy decisions taken by the Government outside the budget process in the last year and their fiscal impact. 5 . A time-series on the number of employees in central government, local authorities, and other institutions, the burden of salary for which falls on the Government. 6. Pension liabilities o f the Government (five-year projections). 7. Tax expenditures and how they might have benefitedthe country. 8. Off-budget transactions, such as borrowing by the Government from parastatals. 9. Outstanding financial issues with Mainland Tanzania. 10. Details of expenditures made from unallocated budget retained in the MoFEA in the previous year.  Zanzibar Public ExpenditureReview Page 73 8. GOVERNMENT PAY AND EMPLOYMENT IN ZANZIBAR THEWAGEBILL 8.1 In FY02, expenditures on wages and salaries claimed 63.2 percent o f 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 l e f t to cover operation and maintenance and other development expenditures. Figure 8.1: Share o f the Wage Bill in Recurrent Expenditures and GDP, FY92-FY02 Share o f Wage in Total E k p x I w c8P 70% 60% 50% 15% 40% 30% B 10% .. 5% 1 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 o f government expenditures, and as a percentage o f GDP. From a share o f about 30 percent o f recurrent expenditures in FY93, it has continuously increased and reached more than 60 percent by FY02.l' A similar pattern can be discerned if the ratio o f wages and salaries to GDP i s 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 o f the wage bill was made possible by a steady increase in government revenue. This permitted an expansion o f the wage bill without requiring a concomitant reduction o f non-wage expenditures. However, as revenues started to decline in FY99, the wage bill could not easily be reduced, and most o f the decline in revenue had to be absorbed by a reduction in non-wage expenditures-a situation that consequently constrained the ability o f the public sector to deliver basic services. 8.3 To put Zanzibar's wage bill into perspective, we compare some o f the key indicators for Zanzibar to those o f other economies in the region and to other small island economies. It merits attention that, in addition to the employees o f the RGoZ, there are also a large number o f 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 IO The figures used in this 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 o f employees o f other government funded institutions such as the Zanzibar State University are embedded in grants and subsidies, the actual share o f wages and salaries in Zanzibar's expenditures i s even higher.  Zanzibar Public Expenditure Review Pane 74 Bank o f Tanzania. Government estimates o f the wage Estimated cost o f Union Government Staff in Zanzibar, 2002 . bill of Union Government employees deployed in Zanzibar run to around T.Shs. 10 billion. Adding this T.Shs. Billion to the wage bill covered by the RGoZ increases the Defense 4,473 ratio of the wage bill to GDP to 14.6 and the ratio of the wage bill torecurrent expenditures to 85.1 percent. I Foreign Affairs Police Immigration 20 4,846 174 Others 8.4 First, we compare the total wage bill o f (incl. TRA and BOT) 682 Zanzibar (Zanzibar Government plus Union staff in Total 10,197 Zanzibar) to that o f other African economies in the region: Kenya, Mozambique, Tanzania, and Zambia. Source: Office o f 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 Kenya at the high end. Zanzibar’s ratio of 14.6 percent i s clearly much higher than that of the comparable economies. Similarly, the share of recurrent expenditures being claimed by the wage bill i s also significantly higher than that of the comparable economies, which display ratios between 29.2 percent (Kenya) and 46.1 percent (Mozambique). 8.5 We also compare Zanzibar’s wage bill to that of other small island economies 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 i s typically much larger than that of larger countries as 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 performance o f a similar task for a small economy. Nonetheless, even in comparison to 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% 46.1 Yo Tanzania 4.0% 3 1.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 comprised more than 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 Expenditure Review Page 75 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 o f government employees has increased by more than 10 percent since 1994/95, when it was 23,156. 8.7 A useful indicator for international comparisons of the size o f the government labor force i s government employment as a percentage o f the population. In Zanzibar, 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 employed by Government i s 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. Industrialized countries including the OECD countries and the countries o f Eastern Europe and the former USSR on the other hand have a significantly higher share o f the population employed by Government. Table 8.2: Government Employment (as Percent o f Population) for Selected Regions Source: Zanzibar - CSD, others Salvatore Schiavo-Campo, Giulio de Tommaso, and Amitabha Mukherjee. Government Employment and Pay in Global Perspective: A Selective Synthesis o f International Facts, Policies, and Experience. Policy Research Working Paper No. 1771. The World Bank. 1997. 8.8 Government employment figures reveal that about 41 percent of the government workforce i s teachers and health workers, while 59 percent of the government workforce works in other areas. This distribution of government workers in Zanzibar i s surprisingly similar to the average for most other world regions, where between 37 percent and 44 percent of the government workforce i s employed in the health and education sectors. Only Eastern Europe and the former USSR show a significantly different distribution, with about 74 percent o f the government work force employed in the health and education sectors. 8.9 Government employees in Zanzibar in the general administration are classified in 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 education requirement for this service level i s form IV. (c) Most of the Government employees in Zanzibar fall under the general service scale (MUS). Employees in this salary grade are required to have completed Form V I or have completed Form I V and a two year training course. (d) Technical staff (MW salary scale) are required to have a first degree or an advanced diploma which i s almost equivalent to a f i r s t degree. Also, employees in the MUS  Zanzibar Public Expenditure Review Page 76 category with 15 years experience (continuous) in the same field of work and with the ability to discharge their duties effectively can be promotedto this category. (e) The highest category of government employment i s covered by the leaders scale (MV). Employees in this category include Ministers, Principal Secretaries (PSs), Deputy Principal Secretaries (DPSs), Commissioners, Directors, etc. Most o f employees in this category are Presidential appointees and staff who have been promoted from the M V 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, M- US19 8,63 1 9,275 9,52 1 MWI -MWl* 3,851 3,383 3,744 M V I - MV4 170 167 187 8.10 Senior government officials in Zanzibar expressed concern about overstaffing in the lower salary categories which goes hand and hand with critical skill shortages in the technical and managerial positions. 8.1 1 Initial surveys carried out by CSD, the MoFEA, and sector ministries did not reveal a 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 been diverted to 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 computerized central payroll system which Government intends to implement. 8.12 As in many other countries at a similar level o f development, employment in the formal sector i s also low. Government statistics derived from an employment and earnings survey carried out in 2001 indicate that the total number of employees in the formal sector i s around 64,250, out of which about 50 percent are employed by 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 o f the base salary and various allowances. Table 8.4 provides information on 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 i s fairly compressed, reflecting a policy that tries to minimize inequality. I t i s interesting to note that between 1999 and 2002 the wage compression ratio declined from 5.4 to 4.4. This i s due to the relatively smaller salary increases granted to higher salary grades compared to increases granted to lower salary grades. While the average salaries for the lowest grades increased between 1999 and 2002 by 48 percent, that o f the highest cadres increased by only 15 percent.  Zanzibar Public ExDenditure Review Page 77 Table 8.4: Average Base Salary by Job Category, 1999-2002 Source: CSD 8.14 I n addition to their basic salaries, government workers receive a variety o f 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) 1 11,155 I 11,552 I13,200 I 16,895 1 22,055 Allowances (TShs. ‘000) I 7,823 I 9,366 I10,714 I 16,372 I 18,164 I 18,978 I20,917 I23,914 I 33,267 1 40,219 I llowances as a % o f base salary 70% I 81% I 81% I 97% I82% 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 of these 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 o f T.Shs. 5000/= for lower to mid-salary grades. For higherisenior (Le. M W & MV) salary grades fuel allowance i s 2. Transport allowance paid. Only paid to presidential appointees ( M V ) ranging from T.Shs. 3. Entertainment allowance 80001= to T.Shs. 250001= Paid to presidential appointees (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% o f 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 o f the salary structure. Table 8.8 illustrates the magnitude of allowances and the impact on take-home pay for the lowest paid and for the highest paid salary groups.  Zanzibar Public ExDenditure Review Page 78 Table 8.7: Base Salary and Allowances for Highest a n d Lowest Paid Salary Categories, FY02 House Allowance 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. I t also should be noted that in addition to the allowances, government employees, especially the more senior cadres, also enjoy non-wage benefits such as the duty-free import of 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 i s therefore recommended. However, since at present allowances are neither taxable nor pensionable, a consolidation o f allowances and non-wage benefits into a single salary will require a careful assessment of the financial implications. This i s particularly important with respect to pensions, which, under the current rate structure, would significantly increase and add to the obligations o f 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 benefits under the present pension system. 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 i s important to note that the Union Government has consolidated base salaries and allowances into a single salary payment (Table 8.9).  Zanzibar Public Expenditure Review Page 79 Table 8.8: Comparison o f Estimates o f the Average Salaries for Zanzibar and Mainland Tanzania, N O 2 Ishare in Total Employment1 Average Salary I Zanzibar I Mainland I Zanzibar IMainland incl. base salary allowances Operational Service 13% 15% 34.210 44.882.44 56.633 C i v i l Service (incl. Health) 60% I 22% 1 52,927 1105,691 I 85,267 olice and Defense 0% I 15% ____ ____ I 75,669 Teachers 27% I 48% I 52,079 170,251 I 92,670 Source: Staff estimates 8.20 Average take-home pay (calculated by dividing the total wage bill by the number o f 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 Mainland Tanzania, differences in the distribution o f staff across employment categories and job grades need to be taken into account. In Zanzibar, teachers constitute only 27 percent o f government employment, while in Mainland Tanzania, they account for 48 percent. The share o f staff in the operational services i s 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 in the administration o f a larger entity. While the number o f teachers i s intimately linked to 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 i s 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 in recent years for most salary grades. 8.24 In general, l o w real wage levels in the 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 Reduced work effort, declining levels and expectations in performance; > Reduced incentives for improving work performance and willingness to accept greater responsibility;  Zanzibar Public Expenditure Review Page 80 k Weakening of accountability and control procedures/mechanisms; k Diminished ability to recruit and retain qualified managerial, professional, and technical manpower; k Reduced commitment to public service; k Moonlighting and “daylighting;”” and k Misappropriation, misuse of public institutions’ tangible or financial assets for personal gain. 8.25 Pay reform i s one o f the main elements o f Tanzania’s Public Sector Reform Program. Although the Mainland Government has committed i t s e l f to pay reform, progress i s 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 o f 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 i s 60 in Zanzibar. Early retirement with full benefits can be granted at the age o f 55. Individual pension benefits are calculated on the basis o f last salary drawn and number of years o f 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 o f 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 paid a 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 reached the retirement age are s t i l l on the payroll. An important measure for Government would 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 i s no central payroll system in Zanzibar. Each ministry maintains its own payroll on the basis o f which it receives the monthly release for wage payments from the Ministry of Finance and Economic Affairs. Wage payments are made by the ministries, regions, and districts. In the absence of a reliable central payroll system, there i s a significant opportunity for wage bill leakages. 8.29 The absence of a central payroll system also makes establishment control and budgeting for salaries difficult. An analysis o f the difference between budgeted and actual salary payments points towards serious weaknesses in the budgeting for wages and salaries. As figure 8.2 shows, during FY02 there were significant deviations from the budgeted salary 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 budget deviations were much smaller, although not insignificant. I’ That is, (remunerated) non-civil service work i s carried out during time for which civil service pay i s received.  Zanzibar Public Expenditure Review Page 81 8.30 The hiring process i s centralized. The authority to hire and fire civil servants in RGoZ and i t s institutions i s legally vested under the Civil Service Department and Civil Service Commission. The Civil Service will collect a l l requests to fill vacant positions and submit them to the Civil Service Commission for final selection o f 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 o f the civil service. 8.31 An employee can only be terminated from work for the following reasons: (i) completion o f the day’s work, for a daily employee; (ii)completion o f 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 o f misconduct; (v) if terminated under Presidential order/powers; and (vi) for purposes o f cost-cutting. Performance-related separation o f government employees has to be approved by the Civil Service Commission. Retrenchments undertaken for cost- cutting purposes have to be approved by the Revolutionary Council.  Zanzibar Public ExpenditureReview Page 82 Figure 8.2: Expenditures on Personnel Emoluments, Deviation between Budget and Actual Cash Crops & Fruit Comm. Ministry o f Information Land & Environment Comm. Ministry o f State Planning Ministry o f Health M i n o f Water, E.Environment Ministry o f Communication Tourism Commision Min.of State Women& Child. M i n . o f Agric.& L. Stock Chief Minister Office Ministry o f Education ' 0 FY02 House o f Representatives FYOl Attorney General Office 0 FYOO M i n o f Trade & Marketing President Office R. Council High Court M i n o f State & G. Governance Min. o f Finance E. Affairs President Office President Office R. Admin. Prison Department Ant. Smuggling Unit Fire & Rescue Force Economic Bridgade Peoples Militia Units I -100% -80% -60% -40% -20% 0% 20% 40% 60% 80% 100%  Zanzibar Public Expenditure Review Page 83 TOWARDS A CIVIL SERVICE REFORM PROGRAM 8.32 With the wage bill consuming a disproportionate share o f the government budget, wage bill management and civil service reform will need to be the top priority o f the Government if the performance and service delivery capacity o f the public sector in Zanzibar i s to be enhanced. The overall objective o f the Government should be to achieve a smaller, affordable, well compensated and efficient civil service. In that context, the Government needs to undertake structural reforms which could include: 9 Carrying out a civil service census to establish the integrity o f the 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 arrears with respect to gratuities and pensions, to avoid hardships; P Putting in place a central payroll system; 9 Consolidating base salaries, allowances, and non-wage benefits; 9 Adjusting the workforce to eliminate overstaffing and inefficiencies; 9 Adjusting pay scales to enhance the capacity o f the public sector to attract, retain, and motivate staff. Since there are critical skill shortages for professional and managerial positions, pay reform will also need to consider a decompression o f salary scales; and 9 Putting in place training and capacity-building measures. 8.33 In addition, the Government should embark on institutional reforms including: redefining the role o f the State, restructuring for organizational effectiveness if necessary, and building the capacity for management o f the economy, including establishing measures to attract and retain key technical, professional and managerial skills.  Zanzibar Public Expenditure Review Page 84 Box 8.1: Tanzania’s Development o f an Integrated Human Resource System Tanzania embarked on structural reforms in the mid- 1980s which included an initiative to establish appropriate manning levels to reduce the size o f 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 o f efforts led to the establishment of an Integrated Human 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 o f 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. I n 1994, the Government conducted a “National Pay Day Exercise” with the principal objective o f confirming valid employees and detecting any ghost workers in the payroll. This exercise revealed some 13,360 ghost workers. A second objective o f building a personnel database by capturing data on employees for entry into the payroll data base was not achieved. A civil service reform program was launched in 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 employers were requested to sign o f f against payroll-sourced lists for each employee whose presence they could personally attest. This exercise resulted in the deletion o f around 5,500 names from the payroll. I n 1998, a fresh data collection exercise was designed to capture the minimum required data for ensuring that 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 numbers revealed through the personnel verification exercises. The data o f the 1998 data collection exercise was merged with the payroll data and was the basis for the Integrated Human 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 Records Management Trust. London, UK. May 2001  Zanzibar Public Expenditure Review Page 85 Box 8.2: International Experience with Civil Service Censuses Civil service censuses are typically carried out in pursuit o f one o f the following objectives: P Cutting costs by reducing the number o f 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 requires establishing a detailed profile o f public employees-such as data on the mix o f skills, ages, and genders-to facilitate departmental or functional reviews. Methods used to conduct civil service censuses include: P Physical headcounts, which focus on the number o f staff employed and whether the names on a payroll belong to genuine employees. I t involves trained teams traveling to various parts o f the census area, where individual employees are required to present themselves, often with identification and sometimes with documentation, such as photocopies o f letters o f 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 o f database. P Payroll reconciliation involves reconciling the payroll against alternative data sources, such as individual personnel files or service books, the nominal roll or established register, 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 o f routine civil service controls, there has often been no mechanism to ensure that ghost workers do not get back on the payroll. Censuses have contributed to efforts to restructure civil 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 achievement of the targeted objective o f 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 o f a long-term investment in basic payroll and personnel systems. P There need to be appropriate incentives for compliance. For example, incentives may include the stopping o f salary payments for staff on the payroll who are not enumerated. P Coordinated action, top level support, and local ownership o f the exercise’s design and delivery are key to successful civil service censuses. Source: The World Bank. International experience with civil service censuses. PREM notes No. 62, January 2002 AND SALARY REFORM EMPLOYMENT 8.34 This section i s intended to illustrate the magnitude of the wage bill adjustment required to 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 and decompress current pay scales. 8.35 Throughout this section it i s 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 i s the ratio o f the wage bill to recurrent expenditures. Although there i s no scientific rule as to what an optimal wage bill to GDP ratio should be, the range o f data shown in Table 8.1 indicates that for most countries the wage bill consumes between 30 to 40 percent o f recurrent expenditures and that Zanzibar should aim at bringing its wage bill also within this range.  Zanzibar Public Expenditure Review 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 recurrent expenditures ratio projections as dependent on the real growth rate 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 o f the decade, ifno additional measures aside from freezing employment at the current level were to be taken. Figure 8.3: Projection of the Share o f Expenditures on Wages and Salaries as A Percent o f Recurrent Expenditures if the Number o f Government Employees I s Kept Constant 70% 60% ~ 50% 40% 30% 20% 10% 8.37 This first scenario shows that even under the assumption o f 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 o f the wage bill in recurrent expenditures i s to be achieved. 8.38 An alternative scenario would be one in which the Govemment implements a general hiring freeze. With an annual rate o f attrition of about 2 percent, if a hiring freeze were to be maintained until the end of the 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 Expenditure Review Page 87 Figure 8.4: Projection of the Share o f Expenditures on Wages and Salaries in Recurrent 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 i s higher, the ratio would drop below 40 percent earlier. For example, with a growth rate o f 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 i s to be lowered to more efficient levels. However, a hiring freeze i s 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 extended period o f time 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 adjustments would also leave little room for pay reform, which i s urgently needed. 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. I n this context, three scenarios are analyzed. I n the 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 affecting all levels of the civil service, excluding teachers. Teachers are excluded, since the PRSP targets increased enrollments. Even if there i s an opportunity for efficiency gains under the current enrollment and facility utilization rates, the likely need for more staff in the medium term makes staff reductions which would be followed by hiring in the medium term uneconomical. Finally, the third scenario looks at staff reductions affecting all government employees, including health workers and teachers. For each of these scenarios we examine the required reduction in staff if a certain wage bill to recurrent expenditures ratio i s to be achieved. The results of these simulations are presented in Table 8.1 1.  Zanzibar Public Expenditure Review Page 88 8.42 These calculations do not take into account the cost of removing government workers from the payroll. While removing ghost workers from the payroll has no cost aside from the conduct o f a civil service census, other forms o f reducing the wage bill can have significant cost attached to them. I n the past, lack o f funding for retrenchment packages 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 beyond the removal o f ghost workers. Table 8.9: Required Reduction in Government Employment to Reduce the Share o f Wages and Salaries in Recurrent Expenditures 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 i s about 24 percent o f the targeted grades. If the wage bill share were to be reduced to 40 percent, staff in the categories MS, OS, and MU S would have to be reduced by about 87 percent or by about 12,596 government workers. 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 reduction in staff by about 14 percent or 2,491 persons. If the target were a wage bill that consumes less than forty percent of recurrent expenditures, almost 50 percent o f the civil service would have to be laid off, Le., about 8,932 persons. 8.45 I f teachers were also included in the staff adjustment program, as illustrated under 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 spread over a larger number of public employees, the percentage reductions for each category o f government employees would be smaller. Reaching the target of reducing the wage bill to recurrent expenditures ratio to 50 percent would require cuts in every staff group by about 18 percent. Reaching the target o f 40 percent on the other hand would require a reduction in the labor force across the board by 34 percent.  Zanzibar Public ExDenditure Review Page 89 SALARY ENHANCEMENTS 8.46 The opportunity for salary enhancements in Zanzibar i s severely constrained by the large size of the civil service. Nonetheless, creating an appropriate incentive structure for c i v i l servants, which includes as one of i t s key components the level o f public sector wages, i s among the conditions that would help improve service delivery and efficiency in the public sector. In adjusting the public sector wage bill and re-equilibratingthe balance between wage and non-wage expenditures as part o f an overall program to reform the civil service, attention would need to be paid to enhancing salary levels for the categories o f employees that are more difficult to attract and to retain, and to reward well performing and well trained civil servants. If salaries remain low, in addition to poor motivation, there i s 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 misappropriation o f public funds and goods for private gain. An example of this found in many countries i s the widespread informal practice at health care units of charging for drugs which should be provided free o f charge and using the income to supplement salaries. 8.47 The following discussion analyzes the fiscal feasibility o f several salary enhancement scenarios that could be part o f an overall strategy to improve the service delivery of the civil service. As in the analysis of employment levels, our main indicator for fiscal sustainability o f wage bill measures i s again the ratio o f the wage bill to recurrent expenditures. I n each of these scenarios, we assume that the Government has created room for an active wage policy by reducing a l l 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. Projections o f average salaries are expressed in constant 2002 T.Shs. In addition to these real increases, it i s assumed that 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 analyzes three scenarios: (a) salary increases in line with real economic growth (b) differentiated salary enhancements (c) selective accelerated salary enhancements (SASE). Salary increases in line with real economic growth 8.48 After an initial reduction in the number o f government employees, their number i s kept constant, i.e., new hiring takes only place to replace government employees who retire or who leave the service for other reasons. If the number of government employees i s kept constant, real salaries can be increased at the rate o f economic growth without increasing the 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 o f Salary Increases in L i n e with R e a l Economic G r o w t h 180000 160000 140000 120000 3 8 100000 N 8 80000 0 I- 60000 40000 20000 0 8.49 At an annual growth rate o f 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 o f 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 o f the inherent problems o f Zanzibar's salary structure i s that pay levels for technical, professional and managerial staff are considered to be too low to attract, retain, and motivate sufficient numbers o f qualified staff. To overcome these problems, a pay reform in which salaries for these categories rise faster than for the other categories could be implemented as part o f an overall package o f reforms to improve the performance o f 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 o f the civil service. To illustrate the opportunity for higher salary increases for certain categories o f staff, we consider a policy in which the fiscal target i s to keep the ratio o f 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 o f 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 measures to restore transparency, accountability, and integrity in the public service.  Zanzibar Public ExDenditure Review Page 91 F i g u r e 8.6: Projections o f average r e a l monthly salaries by j o b category 900000 800000 700000 600000 -os1 - 0 S l O 500000 ~ -K U 1 K ~U 2 d v1 ~ M U S 1 -MUS19 400000 MWI - M W 1 2 300000 -M V I - M V 4 200000 100000 0 2002103 2003104 2004105 2005106 2006107 2007108 8.51 In this scenario, real wages for employees in the two highest job 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 o f staff benefit from fairly large salary increases, while other categories o f staff remain without any real salary increases might have negative effects on the motivation o f 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 o f staff, with higher increase o f staff in the categories M V and M W . Assuming an annual growth rate o f 5 percent, we examine the feasible increase for higher level staff if real salaries for lower cadres are increased by 2 percent annually and the wage bill as a percentage o f recurrent expenditures i s to be contained below 54 percent. F i g u r e 8.7: Projections of Average Salaries in a Scenario with Differentiated Salary Increases 700000 600000 500000 "os1 -os10 ' 400000 ~ r c 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 s t i l l receive annual real salary increases o f about 10 percent annually. This would result in a  Zanzibar Public Expenditure Review Page 92 cumulative increase in their real salaries o f 35 percent over a three year period and 64 percent over a five year period. Cost o f a selective accelerated salary enhancement (SASE) sche,me 8.53 Mainland Tanzania has adopted a system of selective accelerated salary enhancements (SASE) as part o f i t s public service reform program (PSRP). The objective o f the SASE program within the overall PSRP i s two-fold. The first objective i s to address distortions that are introduced through donor-funded local 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-funded projects rather than to the implementation of general government programs. In addition, with the large number o f donors and donor projects active in Tanzania, the system of LCC has become Box 8.3: The SASE Scheme in Tanzania The SASE scheme i s 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 M W 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 reach the 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 objective and transparent criteria were used to select nominees; > Salary scales are pegged to the medium-term pay reform strategy; > Each application i s supported by a work plan for each individual which sets out clear deliverables; and > The MDA i s still within i t s 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 o f the reform process, or the production o f strategic government output; > have an up-to-date job description, which i s detailed and specifies outputs, quality considerations, and timeframes for completion o f tasks, 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 sanctioning the proposals made by MDAs for SASE after a thorough review by a technical committee, the performance o f 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 o f SASE salary supplements. Experience in Mainland Tanzania shows that the implementation o f the SASE scheme i s administratively quite difficult, which has led to delays and only a limited roll-out o f the SASE scheme so far. I t i s to be noted that an employee who moves out o f a “SASE” position does not carry with him the salary enhancement. 1 unwieldy and has little transparency. Implementation of a SASE scheme was thus inten ed  Zanzibar Public Expenditure Review 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 i s to provide immediately enhanced salaries to key government staff who perform critical/strategic functions in the implementation o f the Government’s reform 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 supporting and 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. I t would be important to ensure that donor interventions are integrated into the public sector and contribute to a strengthening o f the performance and capacity o f the public sector, rather than drawing scarce human resources to the implementation o f donor-funded projects with only a limited impact on poverty reduction. 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 reform policy. 8.56 The basis for the analysis i s the pay reform program outlined in the previous section, in which real salaries o f the lower cadres increase by 2 percent annually, while real salaries of 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 immediately to the level which, under the gradual pay reform program, would only be reached after 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 introduction of a SASE scheme for three scenarios, which differ by the coverage of each job category in the SASE scheme. Table 8.10: Scenarios for the Coverage o f the SASE Scheme Grade 1 1 Scenario 1 Scenario2 I Scenario3 Share of staff benefiting from SASE 10% 20% 50% 30% 50% I 100% W 30% 50% I 100% MUS 762 1523 I 3808 MV 899 1498 1 2995 WW 45 75 150 8.58 Under the first scenario, we assume that 10 percent of staff in the salary grade M U S and 30 percent of the staff in the salary grades M V and M W participate in the SASE scheme. Under the second scenario, 20 percent o f staff in the salary grade MUS and 50 percent o f the staff in the salary grades M V and M W 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 M V and MW salary categories receive the SASE supplements.  Zanzibar Public Expenditure Review 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 MV 203,884 225,126 248,580 274,479 303,075 MW 44 1,03 8 486,987 537,724 593,746 655,605 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 MW 2 14,567 168,618 117,881 61,859 MUS 87,090 87,090 87,090 87,090 87,090 MV 303,075 303,075 303,075 303,075 303,075 MW 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 increased by 2 percent annually for the salary grade MUS and by 10 percent for the salary grades M V and MW. The salary enhancement for each year i s calculated as the difference between the projected salary for FY08 and the projected salary for the specific year. 8.60 The additional cost per year for each salary grade i s then calculated by 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 additional expenditures through SASE to annual estimates).  Zanzibar Public Expenditure Review Page 95 Table 8.12: Additional Annual Cost o f SASE Implementation 1 FY04 1FY05 I FY06 I FY07 I FYOS Additional Annual Cost (TShs. Million, at 2002 prices): Scenario 1 MUS 61 I 46 1 31 j 16 I 0 MW 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 I 303 1 230 1 155 1 78 1 0 MV I 3565 I 2802 1 1959 I 1028 I 0 IMW I 385 I 303 1 212 I 111 I 0 1 Total I 4253 I 3334 I 2325 I 1217 I 0 % o f recurrent expenditures I 8.3% I 6.2% 1 4.1% I 2.0% I 0.0% 8.6 1 Under each scenario, the additional spending due to SASE declines every year as the regular salaries converge towards the pay target. Under the first scenario, in which less than a third o f 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. Under the second scenario, in which 50 percent of the two highest salary categories and 50 percent of staff in the M U S 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 M V and M W and 50 percent of the staff in the 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 Examining the three scenarios shows that the additional cost for including M U S and M W categories in the SASE scheme i s comparatively small, since (a) for the MUS category, the amount of the salary supplement i s assumed to be low, and (b) for MW, the number o f staff in that category i s small. The cost and affordability of such a scheme thus depends critically on the treatment of M V staff in such a scheme. Whenever the coverage i s less than 100 percent of a salary’s categories, it i s also likely that our estimates understate the actual cost o f the scheme. This i s due to the fact that these calculations are based on average salaries for a job 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 leading to higher additional cost. 8.63 Experience in Mainland Tanzania points out the fact that SASE needs to be part o f an overall reform program that includes the introduction of a series of measures to enhance meritocracy. The experience so far i s limited, and the scheme, although showing potential to bring some order to the pay issue, needs to be carefully implemented. So far, the implementation o f the SASE scheme i s administratively quite difficult, and can only be  Zanzibar Public ExDenditure Review Page 96 implemented at the pace that donor resources are available for i t s implementation. Government has been reluctant to 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-term growth and revenue performance i s below the projected levels. Unless Government has restructured expenditures in a way that would create sufficient headroom to absorb shocks, the potential benefits from a SASE scheme are unlikely to warrant taking the inherent risks.  Zanzibar Public Expenditure Review Page 97 DATA ANNEX NOTE: Because of independent rounding of numbers, some totals shown in the following tables may differ slightly from the sum o f their components. Q\ -k 2 Q\ 0 ! z N 4 W P 2 c P N Q\ N 4 c e g N P x 2 P W L W N - w 4 N w g w 4 N - 4 N 0 0 0 0 0 Q\ w N - Q\ VI 2 w x 00 4 2 P N N P  a W 4 e - 4 P P W a e - P woo P 3 v N t4JN 0 0 E. 5 - f N N VIul c! P P 4 woo Q P P 0 i D Q - N VIM \ow c ul w\O P w w c 3 - 2? L - w w W i - h, P P Q \oa L . - ‘be “w z’o” w w m w o a w I woah,  P - 'b "w 03 - "4 o a P W 4 W w 4 w w o w m e w - - m e "P "0 4 ~ ~ N O P v l o o o o - 2 2 f ; 1. - - z 3 I. 4 w - N 0 p o w ’ 8 m a w 4 L 0 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 - 3 k 4 P O o Q \ w w uvl w Pin:$ Q14 uDp P W P p e woo  w * - w n o - 4 * L -- in& W P P “ 0 “ 0au.ON P  b 4 ?. Y 9 ? Y 2 z : m 2 3' W 3' n - N u m n w - P W P V I N W W W 0 0 V I w w 0 a c 3 U W n W - A N w m N L + A W 00 3 w 4 m N N W 4 N 4 VI c n o P h) r 4 N P - N n 01 0 0 4 3 0 0 0 A 2 9 N W t W N N W + w "VI w 4 P P W I W 00w  --e N " " P w 9 -"w"ww 4 w P V " a 4 ~ 0 0 m w N W " p w .-e" 0 0 0 0 0 0 0 l 0 0 + - V I - N U "w N P k " i 2 yd 4 \OPP " W "w VI 9 " "4 P 0 0 0 0 0 0 0 l 0 0 3 N + N " m " w"4 -"N-e"4 a - - i D w - 4 m m W N W9P3PoNr-P "Po 0 0 0 0 0 0 0 1 0 0 p ? 2 5 4 j : - 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 Expenditure Review Page 111 Table AS: Government Expenditures by Ministries and Years. 199912000 2000/2001 200112002 1 Budget Actual Budget 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 o f State Planning Salary &Wages 219 220 256 13 Other charges 178 93 Dev. (local) 525 115 :hief Minister Office Salary & Wages 398 397 464 465 723 846 Other charges 49 1 492 506 267 1,045 515 Dev. (local) Iigh Court Salary &Wages 362 416 43 6 412 506 482 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 25 7 90 177 48 Dev. (local) Iouse o f Representatives Salary & Wages 655 772 81 1 998 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 21 1 295 78 344 120 Dev. (local) 4in. o f 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 Public Expenditure Review Page 112 Table A8 (Cont'd): Government Expenditures by Ministries and Years. 999/2000 2000/2001 2001/2002 linistries Items Budget Actual Budget Actual Budget Actual tin. o f 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 23 0 89 300 149 t i n o f Trade & Marketing Salary & Wages 110 151 169 143 171 178 Other charges 176 72 160 44 24 1 115 Dev. (local) 35 95 18 and & Environment Comm. Salary & Wages 199 206 22 1 113 Other charges 110 96 113 14 Dev. (local) 20 40 finistry o f Education Salary & Wages 4,843 5,198 5,773 5,806 7,121 7,950 Other charges 68 1 280 627 175 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 23 0 300 87 fin o f Water, E. Environment Salary & Wages 675 72 1 810 899 1,125 1,183 Other charges 161 103 23 1 61 57 1 249 Dev. (local) 650 250 500 417 690 linistry o f Communication Salary & Wages 860 962 1,087 1,090 1,145 1,631 Other charges 870 203 96 1 120 1,321 317 Dev. (local) 1,000 348 2,500 29 1 2,500 63 8 linistry o f 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) :ash Crops & Fruit Comm. Salary & Wages 196 Other charges 340 Dev. (local) 'ourism Commission Salary & Wages 122 145 157 149 208 Other charges 61 1 339 580 370 307 Dev. (local) 10 ire & Rescue Force Salary & Wages 294 33 5 388 408 810 488 Other charges 58 64 83 42 295 65 Dev. (local) fin o f 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.  Zanzibar Public Expenditure Review Page 113 Table A9: Actual Tax Collection by TRA-Zanzibar and Zanzibar Revenue Board  Zanzibar Public Expenditure Review Page 114 Table A10: Composition o f Inland Revenue (% o f Total Collections by ZRB) Table A l l : Composition o f Income Tax and Customs Revenue (% o f Total Collections by T W  Zanzibar Public Expenditure Review Page 115 Table A12: Health Sector 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 Planning and 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: Education Ministry Budget Summary (T.Shs. billion) Sub vote Description FY98 FY99 FYOO FYOl FY02 FY03 Actuals Actuals Actuals Actuals Approved Budget 03 Pemba sub office 1.37 1.51 1.97 2.03 2.37 3.66 04 Planning and 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 0.3 1 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 T o t a l education vote 4.42 4.54 5.54 6.01 8.75 12.08 Source: Draft budget estimates for various years, Ministry o f Finance and Economic Affairs, Zanzibar 2 N 2 3 - ? d c C 5 c Y