Report No. 27467-DOM Dominica OECS Fiscal Issues Policies to Achieve Fiscal Sustainability and Improve Efficiency and Equity of Public Expenditures June 25, 2005 Caribbean Country Management Unit Poverty Reduction and Economic Management Unit Latin America and the Caribbean Region Document of the World Bank DSS Dominica Social Security EC Eastern Caribbean ECCB Eastern Caribbean Central Bank ECCU Eastern CaribbeanCurrency Union ECD Early childhood development ECEMP EasternCaribbeanEconomic ManagementProgram EDF EuropeanDevelopment Fund EIB EuropeanInvestment Bank ETF Education Trust Fund EU EuropeanUnion FAA Finance (Administration) Act FCP Fiscal Consolidation Program FERP Fiscal and Economic Recovery Project GDP Gross domestic product GFS Government Financial Statistics HFLE Health and Family Life Education FAD International Fundfor Agricultural Development ILO InternationalLabour Organization IMF International Monetary Fund IMR Infant mortality rate IOCR Institutional and Organizational Capacity Review IPPF International PlannedParenthoodFederation JPS Junior Secondary Program MDGs MillenniumDevelopment Goals MIS Managementinformation system MoF Ministry of Finance MoH Ministry of Health MTESP Medium-Term Economic Strategy Paper NAO National Authorizing Officer NCB National Commercial Bank NDC National Development Corporation NFPS Non-financial public sector NGO Nongovernmental organization NSP Needy Students Program OECS Organization of Eastern Caribbean States OPEC Organization of the Petroleum Exporting Countries PAC Public Accounts Committee PAHO Pan-American Health Organization PFM Public Financial Management PMCU ProgramMonitoring and Coordination Unit PPP Purchasingpower parity PRGF Poverty Reduction andGrowth Facility PRSP Poverty Reduction Strategy Paper PSIP Public Sector Investment Program SBA Stand-By Arrangement SFP School Feeding Program ... 111 SIF Social Investment Fund SIGFIS Standardized Integrated Government Financial Information System SIGTAS Standardized Integrated Government Tax Administration System SOEs State-owned enterprises STABEX Stable Export EarningsProgram STEP Short-Term Employment Program STI Sexually transmitted infection UNESCO UnitedNations Educational, Scientific, and Cultural Organization VAT Value-added tax WAEMU Western Africa Economic and Monetary Union WFP World FoodProgram i v Acknowledgments This report is based on the contributions of a team consisting of Lorraine Blank (social protection), Andreas Blom (education), Patrick Grady (public sector investment program), David Lindauer (public sector employment and compensation), Ted Paterson (budget management), Helen Saxenian (health), and Claudia Sepdlveda (fiscal sustainability). As Task Manager, Claudia Sep6lveda was responsible for the overall preparation of the report. Sharokh Fardoust and David Rosenblatt were peer reviewers of the report. The team would like to thank Antonella Bassani, Mauricio Carrizosa, Camille Nuamah, and Andrew Mason for their helpful comments and their invaluable collaboration to improve this report. The report was edited by Dianne Stamm and benefited from excellent assistance of Patricia Holt, Anne Pillay, and Hazel Vargas. The assistance, cooperation, and input of many Government departments in Dominica-too numerous to mention individually-is gratefully acknowledged. In particular, extensive assistance was received from the Ministry of Finance. The report also benefited from the assistance and cooperation of the Eastern Caribbean Central Bank, the Caribbean Development Bank, and the International Monetary Fund. Without their help, preparation of this report would not have been possible. V TABLE OF CONTENTS ACKNOWLEDGMENTS ACRONYMS ABBREVIATIONS AND EXECUTIVE SUMMARY .................................................................................................................... 1 FISCAL DEFICIT. PUBLIC DEBT. AND FISCAL SUSTAINABILITY................................. 21 A. DominicaFiscalPerformance. FY1993/94-FY2002/03 .................................................. 23 B. Does Dominica FaceaFiscalSustainability Problem?.................................................... 26 C. Fiscal Consolidation inDominica's Public Sector........................................................... 31 BUDGETMANAGEMENT....................................................................................................... 37 A. ConstitutionalandLegalFramework............................................................................... 41 B. Budget Preparation. Execution. andMonitoring.............................................................. 44 C. Budget Accountability...................................................................................................... 55 THEPUBLIC SECTOR INVESTMENTPROGRAM............................................................... 59 A. InstitutionalOrganization andPreparation....................................................................... 61 B. Execution and Monitoring................................................................................................ 66 PUBLIC SECTOR EMPLOYMENT AND COMPENSATION ................................................ 71 A. Trends in Employment duringFY1991/92-FY2001/02 .................................................. 72 B. Trends inCompensation duringFY1991/92-FY2001/02 ................................................ 76 C. Public Sector Reform....................................................................................................... 79 THE HEALTHSECTOR ............................................................................................................ 83 A. Health Outcomes in Dominica......................................................................................... 85 B. The HealthSystem: Financing, Coverage, and Delivery ................................................ 88 C. Uses of Health SpendinginDominica, FY1991/92-FY2002/03 ..................................... 93 THEEDUCATION SECTOR..................................................................................................... 99 A. The EducationalSystem inDominica............................................................................ 101 B. EducationOutcomes inDominica.................................................................................. 102 C. Uses of EducationExpenditures in Dominica................................................................ 107 SOCIAL PROTECTIONPROGRAMS INDOMINICA.......................................................... 119 A. Dominica SocialProtection Strategy and Programs ...................................................... 121 B. Composition and Evolution of SocialProtection Spending inDominica. FY1995/96-FY2000/01 .................................................................................................. 128 C. Social Protection and Risk Management in Dominica................................................... 130 BIBLIOGRAPHY .............................................................................................................................. 133 ANNEX.............................................................................................................................................. 136 vii LIST OF TABLES Table E.S.1.Summary of Policy Recommendations........................................................................... 15 Table 1.1. Dominica Net Non-Financial Public Debt Structure as of June 30. 2003........................... 26 Table 1.2. Medium-TermFiscalEffort Required to Achieve Fiscal Sustainability, FY1994/95-FY2006/07.............................................................................................................. 28 Table 1.3. Annual RevenueForgone from Tax Concessions............................................................... 33 Table 3.1. Sectoral Composition of Actual Capital Expenditures for the DominicaPublic Sector.....45 Table 2.1. Dominica's Budget Overview, FY2000/01......................................................................... 67 Table 3.2. Sectoral Composition of Actual Capital Expenditures for Dominica Central Government................................................................................................................... 68 Table 3.3. Sources of Financingof Dominica Central Government Capital Expenditures, FY1995/96-FY2001/02.............................................................................................................. 69 Table 3.4. RealizationRate of Dominica Central Government Capital Expenditures, FY1994/95-FY2001/02.............................................................................................................. 70 Table 4.1. Established Posts in Dominica Central Government, FY1990/91-FY2001/02 ..................73 Table 4.2. Growth in Central Government EstablishedPosts by Ministry or Department, FY1997/98-FY2001/02......................................................................................... 74 Table 5.1. HealthIndicators inDominica Compared with other OECS Countries, Barbados, and Jamaica ...................................................................................................................................... 86 Table 5.2. Ministry of HealthFacilities and Functions........................................................................ 88 Table 5.3. User-Fee Revenue and Revenue as a Share of Princess Margaret Hospital Expenditure, FY2000/01(US$) ................................................................................................. 90 Table 5.4. User Fees at PrincessMargaret Hospital............................................................................. 91 Table 5.5. Comparative Estimates of Health Spending in Dominica Compared to other OECS Countries and Jamaica, 2001................................................................................. 94 Table 5.6. Average Payment Time, OECS Pharmaceutical Service, by Country, July 1, 2002-June 30, 2003 ....................................................................................................... 96 Table 6.1. Educational Institutions in Dominica by Ownership, 2001-2002 .................................... 102 Table 6.2. The Basics of the Educational System in Dominica, 2000 ............................................... 103 Table 6.3. Enrollment and Education Expenditure, by Quintiles....................................................... 111 Table 7.1. Social Protection Programs and Expenditures in Dominica, FY2000/01 ......................... Table 7.2. Incidence of Poverty, Indigence, and other Related Indicators, Dominica, 2002 .............126 131 Table 7.3. Dominica RiskIndicators by Age Cohort, ExistingPrograms, and SuggestedInterventions .................................................................................................... 132 Annex Table A1.Central Government Fiscal Accounts, Dominica, FY1993/94-FY2002/03 .......... 136 Annex Table A2.Central Government Fiscal Accounts, Dominica, FY1993/9&FY2002/03 .......... 137 Annex Table A3.Central Government Fiscal Accounts, Dominica, FY1993/94-FY2002/03 ..........138 Annex Table A4.Main Economic Indicators, Dominica, FY1993/94-FY2002/03 ........................... Annex Table A5.Rest of the Non-Financial Public Sector, Dominica, FY1993/94-FY2002/03 .....139 140 LIST OF FIGURES Figure 1.1. Dominica Central Government Fiscal Performance. FY1993/9&FY2002/03 ..................23 Figure 1.2. Dominica Central Government Primary Expenditure. Interest Payments. and Revenue FY1993/9&FY2002/03 ............................................................................................................. 24 Figure 1.3. Dominica Non-Financial Public Sector Debt and CGPrimary Surplus. FY1993/94-FY2002/03 ............................................................................................................. 25 Figure 1.4. Dominica Central Government Tax Revenue Composition, FY1993/9&FY2002/03 ......32 Figure 1.5. Dominica Central Government Expenditure Composition, FYl993/94-FY2002/03 ........35 V l l l ... Figure 1.6. Dominica Central Government Expenditure Composition. FY1993/94-FY2002/03 ........35 Figure 2.1. Dominica Government Structure ....................................................................................... 42 Figure 2.2. The Budget Cycle in Dominica.......................................................................................... 48 Figure2.3. Variance betweenActual and BudgetedRecurrent Expenditures inDominica, FY1994/95-FY2002/03 ............................................................................................................. 50 Figure 2.4. Variance betweenActual and Budgeted Primary, Recurrent Expenditures. andInterest Paymentsin Dominica. FY1994/95-FY2002/03 ................................................... 50 Figure2.5. Variance betweenActual and BudgetedRecurrent Expenditures by Ministries in Dominica. FY1994/95-FY2000/01 ................................................................. 51 Figure 2.6. Variance between Actual and Budgeted RevenuesinDominica. FY1994/95-FY2002/03.............................................................................................................. 51 Figure 4.1.a. Government Employment as Percentage of Population. 2001. SelectedCaribbean Countries .................................................................................................... 75 Figure4.1.b. Public Sector Wage Billas Percentageof GDP. 2001. SelectedCaribbean Countries .................................................................................................... 75 Figure4.2. Dominica's Real Salariesfor EstablishedPositions inthe Central Government. FY1993/94-FY2001/02 ............................................................................................................. 77 Figure5.1. Shareof Recurrent HealthExpenditures inGDPand Figure 5.2. Recurrent Health Spending inDominica. FY1991/92-FY2002/03 (inEC$95) ...............94 Total Recurrent Expenditures. FY1991/92-FY2002/03 ............................................................ Figure 5.3. Share of Health Programs inRecurrent Health Expenditures. FY1991/92-FY2002/03 ....95 95 Figure 6.1. Gross Enrollment and Completion Rates inPrimary Education in Caribbean Countries. SelectedYears ....................................................................................... 103 Figure6.2. DominicaTransition Ratesfrom Primary to SecondaryEducation................................. 104 Figure 6.3. GrossEnrollment andCompletion Rates inSecondaryEducation Figure 6.4. Share of Education Expenditures in GDP. Dominica. FY1995/99-FY2002/03 ..............105 inCaribbean Countries. SelectedYears ................................................................................... 108 Figure 6.5. Education Expenditures inDominica. FY1995/99-FY 2002/03 ..................................... Figure 6.6. Dominica Recurrent EducationExpenditures by Subsector. FY1995/96-FY2002/03 .....109 Figure 6.7. Pupil-Teacher Ratio and Costs of Education in the Caribbean Region. 2000 (inUS$) ..109 113 Figure 6.8. Exploiting Economies of Scale in Primary Education..................................................... Figure 6.9. Share of TrainedTeachersinPrimary and Secondary School. Dominica. 1999.............114 117 Figure 7.1. Social Protection Expenditures inDominica (as percent of GDP)................................... 129 Figure 7.2. Social Protection Expenditures in Dominica (in millions EC$ 1995).............................. 130 LISTOFBOXES Box 1.1. The Convergence Criteria inthe ECCUand WAEMU ......................................................... 30 Box 2.1. SelectedRecommendations from Dominica's Country Financial Accountability Assessment ........................................................................................................ 39 Box 2.2. RevenueForecasting inDominica......................................................................................... 52 Box 3.1. The Windsor Park Project and the New InternationalAirport .............................................. 66 Box 4.1. Non-Wage Compensation inDominica's Public Service: The Vacation and Study Leave Programs......................................................................................................... 78 Box 6.1. The Tourism Business inthe Caribbean .............................................................................. 106 Box 7.1. Ensuringthe Sustainability and ExpandingCoverage of the Dominica Social Security Scheme ..................................................................................................................................... 125 ix OECS FISCAL ISSUES: DOMINICA POLICIESTO ACHIEVE FISCAL SUSTAINABILITY AND IMPROVE EFFICIENCY AND EQUITY OF PUBLIC EXPENDITURES EXECUTIVE SUMMARY Overview Eastern Caribbean (EC) dollar, which has been pegged to the U.S. dollar since 1976 1. Dominica i s a small (750 square at EC$2.70 per US$l. The commitment to kilometers) and mountainous island in the a fixed exchange rate and a prudent Eastern Caribbean with a population of monetary policy under the ECCU umbrella about 73,000 inhabitants and a per capita has been a key element in maintaining an gross national income of US$3,300 in inflation rate close to international levels. 2003 (World Bank Atlas methodology). A former Britishcolony, it became a member 3. Notwithstanding a common central of the British Commonwealth in 1978. bank and a common currency, Dominica, This small, open economy has like the other ECCU member countries, traditionally depended on the production conducts fiscal policy independently. of a few commodities, in particular While limited progress has been made on a bananas and coconut-oil-based products. common fiscal policy in the region, the However, the decline inbanana production deterioration of the ECCU fiscal position because of increasing competition in the in recent years, notably in Antigua and banana market, as the preferential access Barbuda, Dominica, and St. Kitts and to the European Union (EU) eroded, has Nevis, has reinitiated discussion among its given way to an increasingly service-based members of the need to coordinate fiscal economy (mainly tourism and financial policies.2 services). 2. As a member of the Eastern Caribbean Currency Union (ECCU), Dominica shares a common central bank, Grenadines, in addition to three dependent the Eastern Caribbean Central Bank territories of the United Kingdom, Montserrat (full (ECCB), and a common currency, with the member), Anguilla, and the British Virgin Islands (associate members). other five independent members of the 2 Organization of Eastern Caribbean States The communiquk of the Special Meeting of the (OECS) and three dependent territories OECS Authority, Castries, St. Lucia, September 28, 2001, alludes to the need to examine in the (Anguilla, Montserrat, and the British medium term public sector reform aimed at Virgin Islands).' Their currency i s the increasing efficiency, and fiscal reform including matters related to revenue, expenditure, financing, 'The OECS consists of six independent members: and debt management, as well as common Antigua and Barbuda, Dominica,Grenada, St. Kitts approaches to taxation and incentives to be applied and Nevis, St. Lucia, and St. Vincent and the to encouragecruise shipping. 1 4. During 1994-2000, the Dominican Dominica. The national 2002 survey data economy grew at about 2.1 percent per indicate that approximately 4 in 10 year. However, during 2001-2003 the individuals are poor (that is, living below a economy contracted as a result of locally defined poverty line). Moreover, exogenous shocks (declining trade approximately 15 percent of Dominica's preferences for bananas, increased scrutiny population i s indigent; that is, they do not of the offshore financial services, the have sufficient resources to meet their September 11, 2001 events, and the global basic dietary needs.3 Although economic slowdown in the early 2000s), Dominica's indigenous Carib Indians and debt overhang. The economy declined account for only 7 percent of the poor, by 4.2 percent in 2001, 4.7 percent in about 50 percent o f them are indigent and 2002, and did not grow in 2003. This 70 percent are poor. Income insecurity economic contraction was accompanied b y and volatility at country and household rising public debt and debt service levels as a result of climatic and external obligations. The government, faced with economic shocks i s particularly lower revenues than expected due to the pronouncedinDominica, like inthe rest of protracted recession, access to domestic the Caribbean, due to the country's small and external financing only at highinterest size, limited economic diversification, rates, and increasing accumulation of high degree of openness, and its extreme arrears, agreed in August 2002 to a 12- vulnerability to hurricanes. The impact of month International Monetary FundStand- this extremely high volatility is especially By Arrangement (SBA). However, the negative for the poor and near poor who SBA went off track soon after approval. may not be able to rely on savings or on In July 2003, the first review of the SBA the government's social protection was completed and the program was programs to smooth consumption during extended through December 2003 times of hardship. following the commitment of the government to a strengthened policy 6. With respect to the Millennium framework to be implemented in two Development Goals (MDGs), the primary phases. Inthe first phase, under the SBA, education goal has been attained-primary the fiscal crisis was started to be addressed education is universal for boys and girls. through a mix of policy adjustment It is likely that the goal of eradicating measures and donor financing. The extreme poverty and hunger will be second phase, currently under achieved; the share o f people living on less implementation, includes a more than US$1 a day i s below 2 percent of the comprehensive reform program to foster population. With respect to the health growth and ensure debt sustainability in MDGs, the under-5 mortality and the context of the three-year Poverty HIV/AIDS goals are the most relevant to Reduction and Growth Facility (PRGF) Dominica. Malnutrition (defined as the arrangement financed by the IMF, the prevalence of underweight children), Economic Recovery Support Operation maternal mortality, and tropical diseases financed by the World Bank and other (other than dengue, which i s endemic) are donors' support. not major issues. The under-5 mortality 5. Despite high levels of per capita In 2002, the adult equivalent poverty and income and good social service provision, indigence lines were equal to EC$3,400 and poverty remains a persistent problem in EC$2,000, respectively. 2 rate in Dominica dropped from an policy needed to come from expenditure estimated 23 per 1,000 in 1990 to 14 per cuts, in particular a reduction in the 1,000in 2000. This is a decline of about 5 number of established and non-established percent per year. The MDG target-a positions, and focusing capital reduction of two-thirds from 1990 to expenditures on projects geared to growth 2015-would require that the rate fall to and poverty reduction and funded largely about 8 per 1,000 by 2015. According to by grants and concessional loam4 A trend projections, Dominica i s well on reduction in tax exemptions and track for achieving this MDG target. It i s discretionary concessions was also difficult to assess the likelihood of essential, as i s a restructuring of the reaching the HIV/AIDS targets, because revenue base through the introduction of baseline data on HIV prevalence and the value-added tax (VAT). While social incidence are poor. However, given the indicators are good for a country at this size of the epidemic in the Caribbean, the level of gross domestic product (GDP), target i s very relevant for Dominica, and there i s ample room for more efficient could be achievable if the country modalities of education and health service vigorously supports effective prevention delivery that can generate significant measures. savings and improve effectiveness. In addition, the government should strive to 7. To improve competitiveness, ensure that its social programs, in restore rapid economic growth, and ensure particular social protection, are delivered its medium-term sustainability in the in a well-targeted and effective manner in context of the currency union, the main order to provide more effective safety nets challenges to the Dominican government in the context of fiscal consolidation and are to (a) maintain a prudent fiscal policy; continued vulnerability to external shocks. (b) lessen the debt to GDP ratio by Specific recommendations on how to completing the debt restructuring exercise achieve fiscal sustainability, improve and limit new borrowing in line with a social service delivery, and reform social sustainable fiscal policy; (c) increase the protection are provided inTable E.S.1. efficiency of public investments and public service delivery; (d) strengthen 9. The report i s organized as follows: regulation and efficiency of public utilities Chapter 1 discusses fiscal sustainability in and sea/air transport; and (e) promote Dominica and presents options for fiscal education and skills development to consolidation; Chapter 2 discusses the prepare the population, notably the poor, budget management system in place and to take advantage of new opportunities in its effect on budgetary outcomes. Chapter the global environment. 3 examines the Public Sector Investment Program (PSIP); Chapter 4 discusses 8. This report concludes that during public sector employment and FY1993/94 to FY002/03 the fiscal policy implemented by the government was There are two types of government positions in unsustainable and posed a risk to the Dominica: established and non-established. The stability o f the currency board distinction between established and non- arrangement. Moreover, this report established positionshas a legal basis. Established recommends that most of the adjustment positions refers to those positions specified in the required to achieve a sustainable fiscal Public Service Act of 1991 and its subsequent amendments. Non-established positions are those not specified inthe legislation. compensation; Chapters 5 and 6 examine percent of GDP cumulative fiscal effort public expenditures and outcomes in the programmed under the PRGF, was about health and education sectors; and Chapter 0.4 to 2.8 percent of GDP, depending on 7 discusses social protection programs. the assumptions of real interest and growth rates. Thus, a significant reduction in the FiscalSustainability net present value of the public debt through debt restructuring or further fiscal 10. At the Central Government level, adjustment beyond the PRGFprogrammed the fiscal position of Dominica weakened effort was needed to regain fiscal from FY1997/98 to FY2001/02, and it sustainability. The CG guarantees the debt improved slightly in FY 2002/03. After o f public enterprises, but the government primary fiscal deficits (after grants) of would not need an additional adjustment about 3 to 4 percent of GDP during to ensure the fiscal sustainability of the FY1993/94-FY 1994195, the Dominica CG public sector because the rest of the NFPS followed a cautious fiscal policy from generates a small primary surplus. FY1995/96 to FY1996/97. During that period, the CG ran a primary surplus 12. The ECCB's Monetary Council between 0.3 and 0.6 percent of GDP. meeting that took place in February 2003 However, in FY1997/98 the C G ran a reconfirmed the agreed regional fiscal primary deficit of 1.1percent of GDP that framework that includes ceilings on debt increased progressively until it reached 7.5 and fiscal performance o f the C G (see percent of GDP in FY1999/2000. In Chapter 1, paragraph 1.19). The aim of FY2000/01, the primary deficit was the ECCB in setting these fiscal contained to about 6 percent of GDP, was convergence criteria i s to ensure long-term reduced to 3.3 percent of GDP in fiscal sustainability and the stability of the FY2001/02, and was reduced even more, currency union by limiting ne,gative to 1.6 percent of GDP in FY2002/03. The spillovers. Based on its fiscal ratios in main cause of this deterioration was FY2002/03, the deadline for fiscal increases in capital expenditures, the convergence (a four-year horizon) financing of which the government relied required a fiscal adjustment higher than on both external and domestic borrowing, the one discussed above if the objective mostly at commercial terms. As a result, was to meet the ECCB-proposed 60 the non-financial public sector (NFPS) percent debt-to- GDP ceiling, and would debt-to-GDP ratio increased from 61 require an even faster public sector reform percent of GDP in FY1997/98 to 111.5 and gearing up of the social protection percent of GDP in FY2002/03, and the programs. interest expenditures increased from 2.6 percent of GDP to 5.5 percent o f GDP in FinancialManagement the same period. 13. Dominica initiated in 1997 a 11. The medium-term fiscal budget reform to introduce changes to the sustainability exercise shows that during budget structure and processes. These FY1993/94 to FY2002/03 the fiscal policy changes, however, were not fully implemented by Dominica was implemented and did not translate into unsustainable. The projected size of the improved resource allocation and effective fiscal adjustment then required to achieve expenditure controls, because o f the lack fiscal sustainability, in addition to the 5 o f firm resource envelopes at the 4 aggregate and ministry level underpinned including the World Bank, to put in place by a medium-term economic strategy. The an effective system of cash management result i s that the budget has been driven by and expenditure control, as a matter of short-termconsiderations. priority, and to outline a medium-term program to further improve financial 14. The Dominica PRSP is expected to management along the recommendations fillthis void by establishing a medium- of the recently completed Country term economic framework that addresses Financial Accountability Assessment the following issues: (a) estimates of (CFAA). On July 21, 2003, the authorities aggregate resources available for public issued an Administrative Order to ensure expenditure consistent with these problems do not reoccur. The macroeconomic stability, (b) bottom-up Administrative Order established estimates of the cost of carrying out procedures to strengthen the cash ongoing and new policies based on management system and penalties for line reliable assumptions, and (c) reconciliation ministries that do not follow these of the aggregate resources with the cost procedures. The system now in place has estimates of the ongoing and new policies. been successful in limiting expenditure commitments to collected revenues. 15. The budget provides an However, it i s critical that this discipline accountant's view of government be continued in the medium term to avoid transactions rather than a presentation of the build-up o f arrears. the government's fiscal policy. The budget adheres to the traditional 17. The debt management function i s administrative, program, and economic the weakest component of the financial structure format. However, this management information systems. This i s presentation i s insufficient to assess the split between a small unit within the fiscal position of the government and to Budget, Debt, and Fiscal Management follow up on key programs supposedly Department, which uses the linkedto policy priorities. The Ministry of Commonwealth Secretariat Debt Finance does, however, have the capacity Recording and Management System (CS- to present the budget by standard DRMS) to maintain data on foreign loans, functional classification, and it i s and the Accountant General's Department, recommended that it do so. which uses Excel to maintain records on domestic debt (Treasury Bills, debentures, 16. The cash management system in local loans). The CS-DRMS has a module place until FY2001/02 was too weak to for domestic debt, and the Treasury monitor Dominica's cash shortage. In accessed this via modem to input domestic October 2002, with Caribbean Technical debt data until late 2000, when the modem Assistance Center (CARTAC) assistance, connection was discontinued at the time the government made an effort to improve the Standardized Integrated Government its cash management capacity by Financial Information System (SIGFIS) introducing a system of spreadsheets to was installed. The government i s in the anticipate, control, and manage process of reestablishing the CS-DRMS expenditure and revenue. Moreover, to module for domestic debt and making tighten control over spending and regain domestic and external debt data computer fiscal sustainability, the authorities have accessible to both the Ministry of Finance beenworking with a consortium of donors, 5 (MoF) and Treasury. Moreover, the investigations if he considers it necessary, Government has established a Debt he has not done so. Finally, following the Committee to oversee the signing of new 2000 election, the government appointed a debt contracts and the management of its Public Accounts Committee (PAC). It has debt. However, the government needs to met on occasion, most recently to consider consolidate in the near term the debt the accounts for FY1999/2000, but it has management function in one unit that not closed the accountability cycle b y maintains domestic and external debt reporting its findings to Parliament. records. The unit will need to be staffed Moreover, the PAC has never reviewed with qualified personnel that can keep the accounts of any statutory body. adequate debt records and advise the government on debt management issues. 20. The formulation o f the budget receives inputs from the social partners, 18. The Audit Department does a but this undoubtedly could be made more commendable job. Audit Reports are systematic. The government consulted current for the CG, and the Audit with representatives from various sectors Department examines the account of Local and interest groups in preparing its recent Governments. However, while in budgets, and these consultations were principle independent because of authority extensive for the FY2003/04 budget, with specified in the Constitution, the Audit widespread and open debate of the Department has no guaranteed operating proposed fiscal measures. The recent budget or staffing complement, and must agreement by OECS countries to establish obtain approval from the Minister of National Economic Councils with public Finance via the annual Estimates process. and private sector members will strengthen Additional factors that have affected its this trend. efficacy are the automation of critical financial management functions without Public Sector Investment Program adequate input from the Audit Department in the design and implementation stages, 21. On average, from FY1995/96 to the nonfunctioning of the Public Accounts FY2001/02 the CG spent about 9 percent Committee, and limited training for staff. o f GDP on capital expenditures. Total public sector capital expenditures were on 19. The financial statements of average about 11 percent of GDP in the Dominica present a partial view of same period. This level of capital government activities. Significant fiscal expenditure i s quite high compared to operations are excluded. The operations other OECS countries, and i s similar to of statutory bodies, including state-owned that of St. Kitts and Nevis, a country that enterprises (SOEs), are not covered, and was hit by three hurricanes between 1995 several contingent liabilities are not and 2001, with resulting high investment presented in the financial statements. reconstruction needs. Preliminary figures Most of the SOEs are significantly in for FY2002/03 show that capital arrears in the preparation of their financial expenditures were 5.1 percent of GDP, statements and the submission of these for and for FY2003/04 the planned capital audit and for parliamentary review. expenditures were about 7 percent of Although the Director of Audit has the GDP. This level of expenditure was right to require submission of SOE audit mostly financed through loans reports and to supplement them by further (commercial and concessional), and i s one 6 of the main reasons behind the difficult land acquisition for the construction of a fiscal position facing the government new international airport that was later today. deemed infeasible. Since FY2002/03, the government has halted several projects, 22. It i s difficult to assess the adequate whose impact on the economy was level of public capital expenditure for the questionable, and also has made an effort Dominican economy because capital to prioritize projects, taking into account expenditure figures reported by the their impact on growth and poverty. Government, as in most OECS countries, However, the government still has a include not only spending on fixed capital, difficult road ahead to systematize the but also some recurrent expenditures. If selection process of investment projects. one subtracts from the historical average C G capital expenditures the unproductive 24. The main focus of the Public capital expenditures (about 4 percent of Sector Investment Program (PSIP) unit GDP), and uses the evidence presented in until recently was the preparation of the this report for investment ineducation and annual capital budget. However, since health, the remaining level of capital FY2002/03 the PSIP; unit has also expenditures intended to acquire fixed prepared a three-year rolling PSIP. While capital of 2 to 3 percent of GDP i s in line this is an important improvement, the with the historical averages in the three-year PSIP remains weak due to: (a) Bahamas (2 percent of GDP) and the lack of a well-thought-out medium- Barbados (2.4 percent of GDP) during term country strategy that includes sectoral 1990-2001. This excludes capital plans to guide the PSIP process, and (b) expenditures that may be needed for weak administrative procedures for the exceptional reconstruction needs in development of the PSIP. The PSIP unit response to natural disasters. has received technical assistance from the World Bank, and the Government of 23. The government's capital budget Barbados. With this assistance, the until recently included investment projects authorities formulated the FY2003/04 that did not satisfy the stated government's investment budget and the three-year objectives of growth and poverty rolling PSIP, which tried to maximize the reduction. During FY1999/2000- use of available grant and concessional FY2000/01 about EC$8.9 million was financing. The PSIP now reflects realistic spent on the Windsor Park project to build estimates of project implementation and a a 10,000-seat stadium suitable for prioritization of projects in line with the international athletic and soccer government's objectives of supporting competitions (only the foundation has private sector investment and growth. been completed). During FY2000/01- Another step in the right direction has FY2001/02, the government spent been made with the circulation of the draft EC$33.2 million on the construction of a administrative procedures for the financial complex, which has been development of the medium-term PSIP completed, but still has significant prepared with technical assistance from vacancies due to the retrenchment in the the Barbados Government. In addition, a offshore sector. During FY1997/98- Projects Supervisory Committee, FY1999/2000, EC$36 million was spent comprising public officials from various on a feasibility study, detailed design, and key ministries, the Budget Director, and 7 the Debt Unit, has been appointed and PublicSector Compensation mandated to oversee project selection, financing, approval, and implementation. 27. Given the small size of its Quarterly budget execution reports are economy and population, Dominica has also being prepared. However, the limited resources to carry out the whole country still lacks a medium-term country range o f functions performed by any strategy and sectoral plans to anchor the government. As several studies have medium-term investment strategy. The shown, small countries tend to have bigger PRSP now in preparation hopefully will public administrations in response to the fill this void. higher cost of supplying public goods and the need to provide a stabilization role to 25. On average during FY1994/95- ameliorate the effect o f external shocks. FY2001/02, only half of planned capital However, in the case of Dominica, the expenditures were actually spent. public administration i s too large and However, this statistic hides substantial expensive for sound fiscal management. year-to-year variation in the rate of spending, which ranged between 24.5 28. The main finding of this report i s percent in FY1997/98 and 86.5 percent in that the high wage bill in Dominica (16 to FY2000/01. The realization rate, on 18 percent of GDP during FY2000/01- average, was in line with that of St. Lucia, FY2002/03 is employment driven (mainly was substantially higher than the 25 workers in established positions) rather percent rate of St. Kitts and Nevis, but than wage driven or GDP driven. The lower than the 64 percent rate o f Grenada. number o f established posts in FY1990/91 Low realization rates are attributed to was 2,615, and increased by 79 posts from resource constraints related to the FY1990/91 to FY1997/98, or 0.4 percent available human and technical capacity per fiscal year. In addition, from and other inadequacies at various stages of FY1997/98 to FY2001/02, 323 posts were the project administration cycle. Because added, or 2.9 percent per fiscal year. This of the importance of EU grant funding, i s a large increase, especially since in low implementation rates often have been FY2000/01 the primary deficit (after the result of slow EU disbursements. grants) was 6.0 percent of GDP, and 7.5 Preliminary figures for FY2002/03 percent o f GDP inthe previous fiscal year. presented a realization rate of 78 percent. Moreover, as a result of recent improved 29. In FY2001/2002 and FY2002/03, monitoring, the execution rate of the the Government imposed a freeze on new capital budget increased significantly in hiring and cut the wage bill as part of a the first quarter of FY2003/04 (to 93 strategy o f fiscal consolidation. However, percent). while adjustment of the pay scale and a hiring freeze have yielded some budget 26. The challenge for the government savings, further expenditure reductions i s to follow up by implementing the will have to come from reductions in procedures it has drafted and to keep its employment numbers. Because there is tight monitoring of the capital budget in little possibility for widespread job cuts the next fiscal years in the face of while maintaining basic services, a radical elections and the subsequent possible change in the existing mode of services pressures for capital spending. delivery will be needed as part of the public sector reform recently launched by 8 the Government. In particular, within the These include: road maintenance, airport new macroeconomic environment of weak services, laundry and catering services in economic growth and a heavy debt burden, hospitals, and cleaning and maintenance and slowing population growth, the services in public buildings. Outsourcing Government needs to adopt more efficient or privatization might realize efficiency and affordable modes of service delivery gains, although significant savings in inhealth, education, and security. government expenditures are unlikely. Finally, some public sector activities inthe 30. In 2003, the Government launched areas of social services and education a public sector reform program including could be transferred to private four components: (a) modernizing public organizations (profit and nonprofit), as is administration; (b) strengthening public being done to a limitedextent in St. Lucia. expenditure management and controls; (c) enhancing growth-supporting public Health services; and (d) rationalizing the delivery of social services and improving their 32. Dominica has a health system with targeting. This reform program needs to a strong primary health care focus. It i s be accompanied in the near term by the dominated by publicly financed and clarification of legal requirements for publicly provided health care. There are compensation that needs to be paid in case no private clinics or hospitals, but there of layoffs, and by the introduction and/or are physicians in private practice and a upgrade of information systems in order to private health insurance system that accurately cost its implications (pension appears to be growing. Recurrent public and severance payments) and monitor spending on health during FY199~92- public employment (established and non- FY2002/03 was about 4.2 percent of GDP, established positions) on a regular basis. on average. Public health spending in real It also will need to address the issue of the terms grew for the same period at about existing fragmentation and duplication of 3.4 percent per year. Given the tight fiscal work among ministries, departments, and constraints, this growth i s unlikely to be agencies, leading to a diffusion of maintained over the short to medium term. responsibilities, and the possibility of the Moreover, wages absorb about 74 percent government divesting some of its of all health spending in Dominica. These functions to the private sector. The reform figures are on the high side, and as a would need to be informed by a human consequence of the cash constraint resource management strategy adapted to experienced by the country lately, the the needs and challenges of a small-size supply of pharmaceuticals and medical administration, and a performance supplies, and the maintenance of appraisal system should be established equipment and buildings have been with pay or bonuses to reward good affected. Capital expenditures on health performance. are highly donor dependent, with a very large share from the EU. 31. There i s some possibility for the government to divest some of its functions 33. Public facilities appear to provide to the private sector. Some services where reasonably good coverage, with a good it is easy to specify output can be more level o f consumer satisfaction. efficiently performed b y the private sector. Nevertheless, the health system faces many challenges in the future. Most 9 immediately, the public system i s facing expenditures in FY2000/01. Facilities do strong budgetary pressures as part of not retain fees, so have little incentive to overall fiscal constraints. Dominica's put effort into collection. Fees revert to small size also poses many challenges to the consolidated account. Furthermore, the administration of public health although only those with public assistance programs, such as: (a) economies of scale, cards are automatically exempted from (b) health services mostly publicly payment of fees, in practice there are huge provided, (c) health financing mostly variations in collection rates from those through general revenue, and (d) a high not on public assistance. While fees may ratio of nurses to physician. The health seem low on a per-service basis, user fees system also faces the challenge o f for a catastrophic health problem could put adapting to a changing disease profile, an unacceptably high burden on a poor including the need to vigorously address household's financial situation, because HIV/AIDS. the exemptions are not well targeted. The user fee policies would benefit from a 34. With the aging of the population, review in terms of equity and efficiency. migration to urban areas, changes over It is recommended that the government time in the disease patterns, and with the also consider how fees might be goal of achieving greater economies of significantly increased to cover costs for scale, it may be possible to rethink the physicians who treat their private patients configuration of health facilities, primarily in public facilities, and for overseas b y reorganizing and consolidating some of visitors. the primary health center services and restructuring the two district hospitals. 36. Public recurrent spending on The government should also be able to hospitals in Dominica has been fairly achieve greater economies of scale by constant over time, at 56 percent. It is better coordinating health services with its recommended that: (a) changes be OECS neighbors and other countries inthe introduced to improve hospital efficiency, region, including specialized laboratory and (b) chronic disease prevention and tests and procurement o f medical supplies. early diagnosis and treatment programs be (An excellent example of exploiting strengthened to keep people healthy and economies of scale at the regional level out of hospitals. On the other hand, that can be extended to other services i s pharmaceutical expenditures have been the Eastern Caribbean Drug Procurement severely underfunded in recent years due Service.) In addition, substituting nurses to budgetary pressures. with nurse assistants where possible would contribute to improving efficiency over the Education short to medium term. All these efforts should be pursued in the context of the 37. During FY1995/96-FY2001/02, public sector reform discussed earlier. Dominica spent in recurrent and capital expenditures about 6.6 percent of GDP for 35. Public health financing i s mainly education services. From a regional through the consolidated fund. Revenue perspective, this i s a very high share of from user fees i s minor. User fees are public spending on education. Moreover, charged only for services at the Princess this level of public spending has not Margaret Hospital, and amounted to about always been accompanied by an 10 percent of the Hospital's recurrent 10 improvement in education outcomes, given that a non-negligible share of highly equity, and efficiency. educated individuals migrates abroad, the public loses a corresponding share of the 38. The fiscal constraints that the returns to this public investment. country faces limit expenditure levels in Dominica i s the only OECS country that all sectors, but this should not preclude the has introduced a tuition fee for nationals government from completing its enrolled in tertiary education. The fee educational agenda by making better use amounted to EC$250 per semester in of the available resources and increasing 2002, and constitutes an important source the nongovernmental provision in of revenue for the Community College (13 education, for example, at the tertiary percent of total expenditure). But to level. Except for pupils in preschool accommodate the expected expansion of education, Dominicans receive education tertiary education, it i s recommended that predominantly from public institutions. fees be increased (coupled with scholarships for needy students) and 39. DuringFY1995/96 to FY 2002/03, regional cooperation among colleges be administrative costs and secondary and strengthened to ensure efficiency tertiary education took an increasing share improvement through specialization. of recurrent education expenditures. Given the large fixed costs involved in 41. From FY1995/96 to FY2002/03, education administration-for instance the share of primary education development of exams, information expenditures in total recurrent education systems, strategies, and curricula-it i s expenditures in Dominica decreased. In expected that smaller states allocate a the mid-l990s, 6 out of 10 education EC higher share to admini~tration.~However, dollars went to primary education. At the there are considerable savings to be gained start of the new millennium, 4 out of 10 from deeper OECS cross-country reached this subsector. Inthe same period, cooperation, which, moreover, would the share o f recurrent expenditure to stimulate quality improvements through secondary education increased from 26 international sharingof best practices. percent to 34 percent. 40. The share of tertiary education 42. The most important educational (including teacher training) in recurrent barriers for low-income families are education expenditures increased from 7.0 limited access to, and inconsistent quality percent in FY1995/96 to 10.2 in of, the education system. The School FY2002/03. In a Latin American context, Feeding Program i s traditionally perceived this share i s below average, but so is as an expenditure line that benefits low- coverage of tertiary education in income students. However, in Dominica Dominica. The combination of limited the School Feeding Program was access to tertiary education and a low cost- downscaled to 7 primary schools in 2002 recovery ratio for tertiary education makes from 22 schools and 40 daycare centers in this budget line inequitable. Further, 1996. No formal impact analysis has been undertaken. The Education Trust Fund The average cost of administration in other (ETF) provides assistance to needy education systems in the Caribbean reaches 10 students for transportation, uniforms, percent, twice the Latin American average of 5 books, and examination fees. The Bank percent. 11 estimates that the ETFprovides benefits to The lack of qualifications i s a severe and less than 1 in 10 needy students. The lack long-term barrier for increased efficiency of targeting for the ETF i s worrisome. inthe education sector. It is recommended The government could hence either limit that in the future only certified teachers be the School Feeding Program and ETF to eligible for open positions in order to low-income families only, or introduce improve teacher qualifications. fees for higher-income families. Social Protection 43. Salaries are, by far, the largest spending item, taking up about 92 percent 45. Social protection programs in o f all recurrent education expenditures in Dominica emphasize risk-coping FY2001/02. At the primary level, there i s strategies, that is, programs to help especially reason for concern because the households cope with the effects of risk wage share of total primary education (especially poverty) rather than prevention spending was about 96 percent. This and mitigation strategies. Moreover, the leaves exceptionally little money for other existing social protection programs are not crucially important learning inputs, such well poised to respond to either significant as writing paper, textbooks for poor economic downturns or natural disasters. students, and school maintenance. This Social protection expenditures in skewed allocation of expenditure toward Dominica increased from 4.2 percent of salaries unnecessarily reduces the GDP inFY1995/96 to about 5.6 percent of effectiveness of public resources and GDP in FY2000/01. Although slightly results in inefficiently low learning. The above the average for the Latin American main reason for the large share o f salaries region (4.7 percent of GDP), any effort to in recurrent education expenditures is the strengthen social protection cannot include low pupil-teacher ratio. Dominica spends costly new programs, given the difficult the most per student per year in primary fiscal position that the Dominican and secondary education among the government i s facing. Caribbean countries after St. Kitts and Nevis, U S 5 4 7 and US$615, respectively. 46. Nevertheless, in moving forward, In primary and secondary education there the challenge for Dominica is to: (a) are, on average, 20 and 18 pupils per strengthen the impact of existing social teacher, respectively. The inefficiency assistance programs through involved in this poor teacher deployment improvements in targeting, monitoring, i s significant. It is recommended that and implementation capacity, and expand these indicators be raised to 25 and 23 coverage o f certain programs taking into pupils per teacher in primary and account existing budget constraint; (b) secondary education, respectively. increase the capacity of social assistance programs, such as the Road Maintenance 44. Quality o f instruction, and and the Education Trust Fund, to be scaled especially teacher qualification, is key for up or down in the face of economic increasing effectiveness of learning. downturns or natural disasters; and (c) Teacher qualifications are inadequate in expand coverage and ensure the financial Dominica. Only 61 percent and 33 sustainability of the Dominica Social percent of the teacher corps inprimary and Security Fund. To achieve these secondary education, respectively, are challenges Dominica will have to adequately trained to perform their task. reallocate resources among programs and 12 possibly rethink existing and/or pipeline overburdened public sector. The lack of projects, especially community-based coordinated efforts across ministries and programs. departments has resulted in overlaps, duplication, inefficiencies, and 47. The government of Dominica has administrative waste. The skills-training adopted a social protection strategy programs in the Youth Division, Adult designed to pursue economic growth Education Division, Women's Bureau and complemented by pro-poor policies and Cooperatives Division, and the Social policies that reduce inequalities and Protection Program, to the extent that they promote human capital development. serve the unemployed, target the same Thus, social strategies focus on ensuring group. In addition, there is inadequate access to health care and primary labor market information to guide program education and on expanding access to development, lack of standardized content, secondary and tertiary education. The and uncertain labor market outcomes. All government also attempts to ensure that programs provide entry-level other programs and services (including reemployment skills training-despite the income support, food, clothing, housing, fact that international experience suggests and utilities) are available for the poor. that the economic returns to this type of This stated, the existing social protection training are generally low. Moreover, all strategy i s characterized by a lack of programs have different systems for strategic vision and coordination among identifying beneficiaries. This increases programs. With the exception of the administrative costs and time costs for Social Security Program, programs are beneficiaries. There is a need to administered by different ministries. strengthen the planning and executing Although the government has expressed its capacity for these programs and to put in intent to better coordinate its social place mechanisms for their fiscal protection programs, it has yet to establish prioritization. a mechanism to oversee the development of an integrated social protection strategy 49. Several programs on social to address critical vulnerabilities. Low protection are not well targeted. Reducing and volatile incomes are a source of risk in program leakage and introducing better Dominica. However, social protection targeting would considerably ease pressure programs are not well suited to address on the budget. Divestment of programs to income volatility and transient poverty. private providers (profit or nonprofit) The result i s that social protection could also result in cost savings. priorities are not clearly identified, and However, much-needed capacity building fiscal prioritization of programs i s lacking. and institution strengthening, including This is reflected in the budgetary process development o f targeting, management during which budgets are essentially information systems (MIS), and survey determined by previous budget allocations capacity, would add development and and not in response to changing needs and administrative costs. Expanding program priorities. coverage with provision of benefits conditional on activities that promote 48. The large number of programs, human capital also come with their own each with separate administrative systems investment costs. The net fiscal impact of and procedures, strain an already 13 the above reforms would need to be strengthened. Several programs, including carefully assessed. cash transfer and labor market programs, are not identified in the government's 50. The Road Maintenance Program budgets as separate programs. This limits that accounts for about 14 percent of total effective planning and weakens fiscal government expenditures on social accountability. Elevating social protection protection i s a good example of where activities to program status would mean better targeting (through the use of some that expenditures, including expenditures minimum living standard below the by object code, could be easily tracked market wage), increased rotation of over time and evaluated with respect to workers, and implementation of systems to program inputs and outcomes. ensure that the program can be scaled up and down in the face of external shocks 53. Inadequate management would strengthen the social protection information systems (MISS) and capacity function of the program. constraints with respect to data collection, monitoring, and evaluation are common 51. Timely and reliable information on across the OECS. Management poor and vulnerable groups i s essential for information systems need to be the design and implementation of social computerized. In addition, regional safety net programs. The Caribbean initiatives to support MIS development Development Bank (CDB)-financed and monitoring and evaluation capacity Poverty Assessment that was conducted in building would address the human 2002 provided useful information; resource constraints in individual countries however, data on poverty and labor and would be more cost-effective than markets need to be collected on a regular developing systems on a country-by- basis. country basis. 52. Budgeting and management information systems also need to be 14 E 3 3 Ee, Y 35 B E Ee, Ee, Y Y Y Y 4 L E: c c 8 0 in v) a 2 0 e, C m itC b0 +I 1 m I .-x L Y 2C A 3 C C C F 7 C E E E v e, fe, .i- 0 0 Y VI C -2Y 4i 2 Y 0 Q CI x R .3 m 2 3 R 0- a, fa, 5 S a, a, > 5 M . Eb U n a SVY -* : 9 Y E s 3 Y E Y 3 c 8 m E .-C 0 v, .3 > a2 3 x Y Su ! W x * 5w mc Q 4 1 . -.-.-> 0 r o I.9x i) E e, .3 g0 e, ? EE U Ea, Y .- E % E I 4 1, C C c C 7 C c E 0 a $8 x u c.e Y a m 0 a G- 0 E .C c) m .-OD E 0 s B3 YQ UgY e, > EE H 3 ? ? * e, rns .-C 9 u e e e I Ea, c) - OD C 0 ..ec U Y 7e o e c EEE P 4 ; 2e .L ab c e E E 6 F U F;i %F a 1. FISCALDEFICIT,PUBLICDEBT,ANDFISCAL SUSTAINABILITY 1.1 Dominica i s a small (750 square kilometers) and mountainous island in the Eastern Caribbean with a population of about 73,000 inhabitants and a per capita gross national income of US$3,300 in 2003 (World Bank Atlas methodology). A former Britishcolony, it became a member of the British Commonwealth in 1978. The economy i s predominantly agricultural and has traditionally depended on the production and export of a few commodities, in particular bananas and coconut-oil-based products. However, the decline in banana production because of increasing competition in the banana market, as the preferential access to the European Union (EU) eroded, has given way to an increasingly service-based economy (mainly tourism and financial services). 1.2 Dominica, as a member of the Eastern Caribbean Currency Union (ECCU), shares a common central bank, the Eastern Caribbean Central Bank (ECCB), and a common currency, with the other five independent members of the Organization of Eastern Caribbean States (OECS) and three dependent territories (Anguilla, Montserrat, and the British Virgin Islands).6 Their currency i s the Eastern Caribbean (EC) dollar, which has been pegged to the U.S. dollar since 1976 at EC$2.70 per US$1. The commitment to a fixed exchange rate and a prudent monetary policy under the ECCU umbrella has been a key element in maintaining an inflation rate close to international levels. 1.3 Notwithstanding a common central bank and a common currency, Dominica, like the other ECCUmember countries, conducts fiscal policy independently. While limitedprogress has been made on a common fiscal policy in the region, the deterioration of the ECCU fiscal position in recent years, notably in Antigua and Barbuda, Dominica, and St. Kitts and Nevis, has reinitiated discussion among its members of the need to coordinate fiscal p ~ l i c i e s . ~ 1.4 Dominica's fiscal position deteriorated sharply from the mid-1990s to FY 2002/03. InFY1995/96 and FY1996/97 the Central Government (CG) ran on average a small primary fiscal surplus (after grants) o f about 0.3 to 0.6 percent o f GDP. However, from FY1997/98 to FY 2002/03 the CG ran a primary fiscal deficit that fluctuated from 1.1percent of GDP in FY1997/98 to 7.5 percent o f GDP in FY1999/2000. In FY2002/03, the C G ran a primary deficit of 1.6 percent of GDP.8 As a result, total non-financial public sector (NFPS) debt increased from about 60 percent of GDP at the end o f FY1997/98 to 111percent of GDP in FY2002/03. The OECS consists of six independent members: Antigua and Barbuda, Dominica, Grenada, St. Kitts and Nevis, St. Lucia, and St. Vincent and the Grenadines, in addition to three dependent territories of the United Kingdom,Montserrat (full member), Anguilla, and the British Virgin Islands(associatemembers). The communiquC of the Special Meeting of the OECS Authority, Castries, St. Lucia, September 28, 2001, alludes to the need to examine in the medium term public sector reform aimed at increasing efficiency, and fiscal reform including matters related to revenue, expenditure, financing, and debt management, as well as commonapproaches to taxation and incentives to be applied to encourage cruise shipping. * Startingfrom FY 2002/03, the IMF is computingthe primary balance using the overall deficit from below the line (financingfithat is, according to financing sources rather than governmentrevenuesand expenditures. 21 1.5 The main reasons for this fiscal deterioration were a combination of large unproductive investment projects and a shortfall in revenues since 2001 (due to a protracted recession). On average, the Dominican economy grew at 2.1 percent per year during 1994- 2000. Real GDP contracted by 4.2 percent in 2001, and 4.7 percent in 2002 and did not grow in 2003. The government-faced with lower revenues than expected due to the protracted recession, access to domestic and external financing only at high interest rates, and increasing accumulation of arrears-agreed in August 2002 to a 12-month IMF Stand-By Arrangement (SBA). However, the SBA went off track soon after approval. InJuly 2003, the first review of the SBA was completed and the program was extended through December 2003, following the commitment of the government to a strengthened policy framework to be implemented in two phases. Inthe first phase, under the SBA, the fiscal crisis was dealt with through a mix of policy adjustment measures and donor financing. The second phase began in2004 and includes a more comprehensive reform program to foster growth and ensure debt sustainability in the context of a three-year Poverty Reduction and Growth Facility (PRGF) arrangement currently under implementation. 1.6 In light of the preceding discussion, the main objective of this chapter is twofold: first, to evaluate the fiscal sustainability of the Dominica Central Government and public sector in the short and medium term, and second, to propose various options for the adjustment required to achieve fiscal sustainability. 1.7 The main findings and recommendations can be summarized as follows: 0 The fiscal policy implemented by the government during FY1993/94 to FY2002/2003 has been unsustainable. The projected size of the fiscal adjustment required to achieve fiscal sustainability, in addition to the 5 percent of GDP cumulative fiscal effort programmed under the ongoing PRGF, i s about 0.4 to 2.8 percent of GDP, depending on the assumptions o f real interest and growth rates. 10 0 Thus, a significant reduction in the net present value of the public debt through debt restructuring or further fiscal adjustment beyond the PRGF programmed effort will be needed to regain fiscal sustainability. The authorities have launched a debt restructuring effort in 2004 which i s ongoing. Following the preparation of this report, the Dominican authorities successfully implemented a program of macroeconomic stabilization and structural reforms with the support of the IMF under the PRGF, the World Bank under the Economic Recovery Support Operation and other donors. Also, the authorities successfully undertook a debt restructuring exercise. Fiscal adjustment, debt restructuring and a recovery in economic activity in 2004-2005have contributedto positioningDominicaon the pathto restoringfiscal sustainability. loThe fiscal effort already programmedby the Governmentunder the IMF-supportedthree-year PRGF amounts to a cumulative 5 percent'of GDP during FY2004/05-FY2006/07, (from a projected primary surplus of 0.2 percent of GDP in FY2003/04 to 3 percent of GDP in FY2006/07, with intermediate targets of 0.5 percent of GDP in FY2004/05 and 2 percent of GDP in FY2005/06), and will consist of a further cut in the wage bill through a reduction in the size of public employment (yielding 1.2 percent of GDP in FY2004/05), and some tax measures. 22 0 If the rest of the NFPS is included in the exercise, the required size of the adjustment will be similar to the one discussed above, because the C G i s chiefly responsible for the difficult fiscal position of the NFPS. A. DOMINICAFISCAL PERFORMANCE,FY1993/94-FY2002/03 1.8 At the Central Government level, the fiscal position of Dominica weakened from FY1997/98 to FY2001/02, and it improved slightly in FY 2002/03. As Figure 1.1 shows, after primary fiscal deficits (after grants) of about 3 to 4 percent of GDP during FY1993/94- FY1994/95, the Dominica CG followed a cautious fiscal policy from FY1995/96 to FY1996/97. Duringthat period, the CG rana primary surplusbetween 0.3 and0.6 percent of GDP. However, from FY1997/98 to FY2001/02, the CG ran a primary deficit every fiscal year. In FY1997/98, the CG ran a primary deficit of 1.1 percent of GDP that increased progressively until it reached 7.5 percent of GDP in FY1999/2000. In FY2000/01, the primarydeficit was contained to about 6 percent of GDP, was reduced to 3.3 percent of GDP in FY2001/02, and was reduced even more, to 1.6 percent of GDP in FY2002/03. This last figure should be used with caution because FY2002/03 was an unusual year: grants (budgetary) received by the Government were extraordinarily high and capital expenditures were below their long-term level. Figure 1.1.Dominica Central Government Fiscal Performance, FY1993/9&FY2002/03 (aspercentage of GDP) 4 0 2 0 0 0 -2 0 -4 0 -6 0 -8 0 -10 0 -12 0 -14 0 J 0CurrentFiscalBalance p*9PrimaryFiscal Balance (affer grants) +Overall Fiscal Balance (affer grants) Source: IMF 1.9 The main cause of the deterioration of CG finances since FY1997/98 has been an increase innoninterest expenditures, predominantly capital expenditures. As Figure 1.2 shows, noninterest expenditures increased from about 31percent of GDP in FY1993/94to an average of about 34 percent of GDP during FY1994/95-FY1997/98. In FY1998/99, noninterest expenditures reached 39 percent of GDP, and increased steadily until reaching its peak of 44 percent of GDP in FY2000/01. Since then, they have been contained to about 33 percent of GDP in FY2001/02 and FY2002/03. At the core of the observed increase in CG expenditures i s a surge in capital expenditures, mainly in unproductive investment projects 23 such as an infeasible new international airport, an unfinished stadium, and a financial complex that remains largely unoccupied because of the retrenchment in the offshore sector; whereas, total revenue increased from an average of about 26 percent o f GDP for FY1993/94-FY 1995/96 to an average of 29 percent of GDP for FY1996/97-FY2001/02. In FY2002/03, total revenue was 28 percent of GDP. Figure 1.2. Dominica Central Government Primary Expenditure, Interest Payments, and Revenue FY1993196FY2002103 (aspercentage of GDP) 50 0 40 0 30 0 20 0 1 0 0 0 0 1 0 0 0CurrentPrimaryExpenditure -capital Expenditure oInterestPayments =Total Revenue -20 0 I Grants -Primary Fiscal Balance (after grants) Source: IMF. 1.10 The overall balance of the rest of the NFPS has been a small surplus." The financial operations of Dominica Social Security (DSS), local governments, the National Development Corporation (NDC), and the public enterprises, the Dominica Water and Sewerage Company (DOWASCO), the Dominica Export and Import Agency (DEXIA), the Dominica Port Authority (DPA), and the Dominica Broadcasting Corporation (DBC), presented on average an overall surplus of about 0.8 percent of GDP during FY1993/94- FY2001/02 (the last year for which data are available). DOWASCO has run an overall deficit duringFY 199.996 to FY2001/02 that at the NFPS level has been compensated for by the overall surplus of DSS12(see Table A5). 1.11 Thus, due to the weakening of the CG fiscal position, the government turned to borrowing to finance its fiscal deficit. The ratio of NFPS debt to GDP increased from 62.6 percent of GDP at the end of FY1993194 to 111.5 percent in FY2002/03. As Figure 1.3 " The NFPS is defined as the CG plus Dominica Social Security (DSS), local governments, the National Development Corporation (NDC), and the public enterprises. '*DOWASCO is a public enterprise, but it has been privately managed since 1989. 24 shows, the ratio of NFPS debt to GDP showed a negative trend, reaching its lowest value in FY1996/97 (55.3 percent of GDP). A year later, in FY1997/98, the debt-to-GDP ratio increased to 60.4 percent and kept rising to peak at about 111 percent of GDP in FY2002/03. As a result, interest expenditures increased from 2.2 percent of GDP in FY1993/94 to 5.5 percent o f GDP in FY2002/03. As will be illustrated in the next section, to maintain fiscal solvency, the C G will need to generate a sizeable primary surplus to pay the interest on the debt. Figure 1.3. Dominica Non-Financial Public Sector Debt and CG Primary Surplus, FY1993194-FY 2002103 (aspercentage of GDP) -200 J I FiscalBalance(affergrants) Primary 0InterestPayments -cNon-FinancialPublicSectorDebt Source: IMF. 1.12 Moreover, the additional borrowing of the CG has been contracted, mainly with external creditors, at commercial terms. This pattern is the opposite of St. Kitts and Nevis, in which the largest share of the NFPS debt is domestic. As Table 1.1 shows, external debt represented around 75 percent of total NFPS debt as of June 30, 2003. The largest individual holder of Dominica external debt i s the Caribbean Development Bank (CDB), with 23.0 percent of the total NFPS debt, followed by the Royal Merchant Bank, with almost 11percent of the total NFPS debt. The largest individual holding of domestic debt i s the National Commercial Bank (NCB), with 7.0 percent o f the total NFPS debt. The CG i s responsible for 78 percent of the total NFPS debt. 25 Table 1.1. Dominica Net Non-Financial Public Debt Structure as of June 30,2003 Stock inMillions Share of Total Debt Share of GDP EC$ (%) (%I Total 760.14 100.0 111.5 Domestic Debt 186.6 24.5 27.4 Commercial Banks 127.8 16.8 18.7 NCB 53.2 7.0 7.8 Others 74.6 9.8 10.9 ECCB 14.2 1.9 2.1 Private 44.6 5.9 6.5 External Debt 573.6 75.5 84.1 Mu1tilateral 266.8 35.1 39.1 CDB 176.8 23.3 25.9 EIB 14.7 1.9 2.2 World Bank 58.5 7.7 8.6 IMF 7.8 1.o 1.1 Othersa 23.6 7.7 8.6 Official 108.6 14.3 15.9 France 32.6 4.3 4.8 Venezuela 3.0 0.4 0.4 Taiwan 31.6 4.2 4.6 Kuwait 9.9 1.3 1.4 Grenada 2.7 0.4 0.4 St. Vincent and the Grenadines 2.7 0.4 0.4 Barbados 13.5 1.8 2.0 Trinidad & Tobago 10.0 1.3 1.5 Belize 2.7 0.4 0.4 Commercial 182.4 24.0 26.7 Citicorp 45.5 6.0 6.7 RoyalMerchantBank 82.6 10.9 12.1 Societe Generalb 30.2 4.0 4.4 IntercommercialBank 6.8 0.9 1.o Others 17.3 2.3 2.5 Debentures 16.1 2.1 2.4 Treasury Bills 1.2 0.2 0.2 Arrears 15.9 2.1 2.3 a. IncludesEIB, IFAD, and OPEC. b. Loanguaranteedby the U.K. government. The guaranteehas beenexercised. Source: IMF. B. DOES DOMINICAFACEFISCALA SUSTAINABILITY PROBLEM? 1.13 The figures on primary deficit and debt presented above raise serious concerns about the fiscal solvency of the government. These concerns are addressed using an intertemporal approach to fiscal solvency. This approach, broadly speaking, defines a fiscal policy as sustainable if the present value o f all future primary surpluses i s enough to repay the current outstanding debt. It i s important to stress that this type of debt sustainability 26 analysis does not address the optimal level of indebtedness, its composition, and political economy considerations.' 1.14 The fiscal policy implemented by the government from FY1993/94 to FY2002/03 has been unsustainable. The projected size of the fiscal adjustment required to achieve fiscal sustainability, in addition to the 5 percent of GDP cumulative fiscal effort programmed under the ongoing PRGF, i s about 0.4 to 2.8 percent of GDP depending on the assumptions of real interest and growth rates. 1.15 Thus, a significant reduction in the net present value of the public debt through debt restructuring or further fiscal adjustment beyond the PRGF programmed effort will be needed to regain fiscal sustainability. The authorities have launched in 2004 a debt restructuring effort, which i s ongoing. Central Government 1.16 The fiscal policy implemented by the CG from FY1993/94 to FY2002/03 has been unsustainable. As Table 1.2 shows, the medium-term fiscal effort required to achieve a stable debt-to-GDP ratio has been positive and sizeable in every fiscal year under reasonable assumptions for the discount rates (Scenarios A to B).14For instance, if the FY1998/99 CG fiscal sustainability were to be assessed by taking into account the government spending and revenue path inthe next three years, the estimates would show that the CG should have cut expenditures or increased taxes between 7.4 to 8.6 percent of GDP in order to have maintained the debt-to-GDP ratio at the end of FY2000/01 at the same level as at the beginning o f FY1998/99.15 1.17 Using the projections for primary surplus targeted by the Government under the three-year PRGF (FY2004/05-FY2006/07), the CG fiscal policy (expenditure and revenue targets) alone will not suffice to converge to a stable debt-to-GDP ratio by the end of the PRGF. As Table 1.2 shows, under scenarios A to C, the three-year fiscal adjustment required to achieve a sustainable fiscal policy in FY2004/05 using the projected primary expenditure and revenue and grants under the PRGF is positive. That is, the CG needs to cut expenditures or increase revenues more than it has been assumed in the PRGF. Indeed, for discount rates o f 2, 3, and 4 percent, the additional fiscal effort required i s about 0.4 to 2.8 percent of GDP. Inother words, the CG will need to cut expenditures or increase revenues by an additional 0.4 to 2.8 percent of GDP to the fiscal effort assumed inthe PRGF. l3 This methodology includes seignorage as a source,of revenue; however, the apportioning of seignorage to national economies has not been discussed in the ECCU, and it is not taken into account in the fiscal sustainability exercise at the country level. However, any fiscal sustainability exercise at the ECCU level should incorporate seignorage as revenue. l4Scenario A (2 percent discount rate) assumes the nine-year average for GDP growth (0.5 percent), inflation (1 percent), and nominal interest rate (3.5 percent, the average interest rate paid by the CG). Scenario B (3 percent discount rate) assumes a positive shock of one standard deviation (1.4) to the nominal interest rate, and scenario C assumes a positive shock of two standard deviations to the nominal interest rate. l5The projected levels of expenditure, revenue, and grants during FY2003/04-FY2006/07 correspond to the ones agreed by the Government with the IMFunder the SBA and the PRGF. 27 x ';t -9 zO Rm Q .1 'c? 2 m Y F ? 2 Q - F 9 c? m m m 2 x -c? W 0 These projections assume capital spending of 7 percent of GDP, which accommodates a possible investment in fixed assets of 2 to 3 percent of GDP plus current expenditures normally included in Dominica's capital budget. It also assumes a gradual adjustment in the wage bill from 17 percent of GDP inFY2002/03 to about 13 percent of GDP inFY2006/07. 1.18 In light of the previous discussion, and to ensure a stable debt situation and adequate financing, the Government publicly announced on December 10, 2003 its intention to undertake a debt restructuring exercise under the framework of the PRGF.'6 The aim of the exercise is to allow the country to eliminate the debt overhang and foster a resumption of growth. FiscalRulesinthe ECCU 1.19 The ECCB's Monetary Council that took place in February 2003 reconfirmed the agreedfiscalframework that includesceilingson debt andfiscalperformanceof the CG. The aim of the ECCB in setting these fiscal rules is to ensure long-run fiscal sustainability and the credibility of the currency union by limiting negative spillovers. However, as discussed in paragraph 1.3, the ECCB has no mandate over fiscal policy in any member country, and the fiscal rules are not binding at this moment for the member states. Box 1.1 presents the convergence criteria in the ECCU and the Western Africa Economic and Monetary Union (WAEMU). Like the ECCU, the WAEMU comprises developing countries with a currency pegged to only one single currency, the Euro. 1.20 In FY2002/03, Dominica did not meet any of the four debt and fiscal performance ceilings proposed by the ECCB's Monetary Council. Every year during FY1993/94 to FY2000/03, the C G has run a current account below the target of 4 to 6 percent of GDP. The target for overall deficit (after grants) has been reached in only 2 years in this period. If grants are not taken into account, in this period the CG has not met the target of overall deficit since FY1993/94, the target debt service to current revenue has not been met during FY2000/01 to FY2002/03, and the target debt-to-GDP ratio has not been met duringFY1997/98 to FY2002/03. 1.21 At the time of the reconfirmation of the fiscal framework agreement, the ECCB's Monetary Council regarded 2007 as a critical date for fiscal consolidation. This deadline for convergence seems unrealistic basedon the fiscal performanceof Dominica in FY2002/03, comprising: (a) a current deficit of about 5 percent of GDP, (b) an overall deficit (after grants) of 7.1 percent of GDP and 11.7 percent of GDP before grants, (c) a debt- to-GDP ratio of 111.5 percent, and (d) a debt service to current revenue o f 20 percent. Indeed, reducing the debt to 60 percent of GDP by 2007 would require a fiscal adjustment higher than the one discussed in the previous section. Moreover, the WAEMU experience meeting the primary criteria in a three-year span was unsuccessful, and it was extended three additional years (to end-December 2005). l6IMF staff estimate that, in the absence of debt restructuring that would substantially reduce the net present value of debt, financing gaps of about US$20 million, or about 8 percent of GDP, would persist for the foreseeable future. 29 Box 1.1.The Convergence Criteria in the ECCU and WAEMU Eastern Caribbean Currency Union (ECCU) The ECCB has put forwarda proposalthat comprisesthe following four targets to be achievedin 2007: A CentralGovernment current accountsurplusof 4 to 6 percent of GDP An overall government budget deficit of no morethan 3 percent of GDP A total outstandingCentral Government debt of no more than 60 percentof GDP Debt service paymentsby the Central Governmentof no more than 15 percentof currentrevenue. WesternAfrican Economicand MonetaryUnion(WAEMU)* The regional Convergence, Stability, Growth, and Solidarity Pact adopted in December 1999 by the WAEMU, has four primary convergencecriteria and four secondary criteria. The norms establishedby these criteria have to be met by 2002. The primary criteria are: The ratio of the basic fiscal balance (defined as non-grant revenue minus expenditureexcluding foreign- financed investment) to GDP mustbe zero or positive The ratio of outstandingdomestic andforeigndebt to nominalGDP must not exceed 70 percent The average annual inflation cannot be more than 3 percent a year Nonaccumulationof domestic andextemal payment arrears inthe currentfinancial period. The secondary criteria are: The ratioof the wage bill to tax revenuecannot exceed 35 percent The ratioof domesticallyfinancedpublic investmentto tax revenue must be at least 20 percent The ratio of current extemal deficit, excludinggrants to nominalGDP, cannot exceed 5 percent The tax-to-GDPratio mustbe 17 percentor more. However, because of the limited progress achievedby the member countries in meetingthe convergence criteria at end-December 2002, the WAEMU Commission proposed to the Heads of State to extend the timetable to end- December 2005. * The eight members of the WAEMU are Benin, Burkina Faso, CBte d'lvbire, Guinea-Bissau, Mali, Niger, Senegal. andTogo. 1.22 The successful achievement of debt and fiscal ceilings needs the simultaneous introduction of a system of incentives to encourage compliance and penalties againsl noncompliance, and a mechanism for monitoring and surveillance of the fiscal performance. The ECCB has not yet specified any system or mechanisms to ensure compliance with the proposed fiscal ceilings. Moreover, the WAEMU example suggests that surveillance of the fiscal ceilings and close monitoring against annual targets during the convergence phase are key to achieving fiscal convergence. 1.23 The ECCB-proposed fiscal ceilings cover only the CG; public enterprises and government-guaranteed debt have been excluded. However, the public sector position can be critical and should be included in any proposal of deficit and debt ceilings. Moreover, the inclusion of the public sector in the fiscal framework would ensure that the government would produce accurate and timely information on the financial position of public enterprises and their debts. 1.24 The deficit and debt ceilings do not take into account the cyclical nature of the fiscal indicators in small countries subject to frequent external shocks. In that regard, i t i s recommended that the ECCU move from debt and fiscal performance ceilings to a countercyclical fiscal policy based on a simple fiscal rule for the structural balance. This rule 30 has been applied successfully in Chile, a country as vulnerable to external shocks as Dominica. c. FISCALCONSOLIDATIONINDOMINICA'S PUBLIC SECTOR 1.25 In the previous section, we showed that the public sector primary surplus needs to be increased infuture years to ensure fiscal sustainability. We propose that the fiscal adjustment rely primarily on expenditure cuts, in particular in the wage bill. Tax increases are seen as only a small fraction o f the adjustment. As discussed below, this type of adjustment i s based on the observed pattern of government expenditures and on the assumption that a tax reform will yield an increase in tax collection of about 2.8 to 3.0 percent of GDP, and on several studies of how the composition of the fiscal adjustment influences its success.17 Tax Revenues 1.26 The main characteristics of Dominica's tax structure can be summarized as follows (see Figure 1.4): Current revenue (tax and non-tax revenue) has remained constant at about 29 percent of GDP during FY1997/98-FY2002/03 compared to 26 percent o f GDP during FY1993/94-FY 1996/97. About 80 percent of tax revenue i s derived mainly from two types of taxes: taxes on international trade and taxes on income and profits. Taxes on international trade have constituted about 50 percent of C G total tax revenue in recent years. Import duties are levied at rates from 0 to 20 percent, with higher rates applicable to a limited set of goods. A 2 percent customs service charge and a 25 percent consumption tax are applied to imports. An environmental surcharge i s also applied to a range of goods. 0 Taxes on income and profits represented 28 percent of tax revenue during FY1993/94FY2002/03. Corporate tax is levied at a rate of 30 percent, and personal income tax i s levied ranging up to 40 percent, with a stabilization levy applicable at a rate of 4 percent inFY2002/03. Taxes on domestic goods and services increased as a percentage of tax revenue from 17 percent in FY1993/94 to 20 percent in FY2002/03. This category includes a number of excises and licenses on specific goods and services; the largest ones are the tax on hotels and services, the gasoline levy, and the stamp duty. The gasoline levy, contrary to other taxes on domestic goods and services, "See, for example, the studies by Alesinaand Perotti (1996), and Perotti(1996) for the Organisationfor EconomicCo-operation and Developmentcountries. 31 i s the residual of the retail price (fixed by the government) minus the cost, insurance, and freight (cif.) price of the imported product and the fixed margins for importers and retailers. This net tax can be negative and fluctuates with the c.i.f. price if the retail price i s not adjusted. Dominica introduced in 2003 an automatic pricing and taxation mechanism for adjusting petroleum prices in line with international price movements. Property taxes on rental income from houses and on land contribute a modest 1.5 percent of the total tax revenue intake. Figure 1.4. Dominica Central Government Tax Revenue Composition, FY1993/94FY2002/03 (as percentage of total tax revenue) 100% 80% 60% 40% 20% 0% 1993194 1994/9j 1995/96 1996197 1997198 ~ Y ~ S / Y Y 1999100 Z O O O / O ~ 2001/02 2 0 0 1 / 0 3 Taxes on InternationalTrade E3 Taxes on Income & Profits Taxes on Domestic Goods & Services ITaxesonProperty Source: IMF 1.27 During FY1999/2000-FY2002/03, the revenue forgone by the Dominica Central Government in customs duties and consumption taxes, due to tax concessions, was on average 5 percent of GDP on a fiscal- year basis. A key feature of the present tax system in Dominica and other OECS countries is the pervasive use of tax concessions.'8 As Table 1.3 shows, hotels and industry received the highest proportion o f tax credits (on average, 1.9 percent of GDP), alongside tax credits based on the Customs Control and Management Act (1.9 percent of GDP). In the extreme case, for example, where all exemptions based in the Customs Control and Management Act were repealed, a revenue gain of 1.9 percent of GDP would be expected, assuming zero elasticity of imports with respect to tariffs and consumption tax. l8An elaborate system of import duty exemptions provides for very generous concessions for taxes on imports for the development of industry, tourism, agriculture, social, educational, cultural, and other purposes. The main exemptions are provided under the Customs Act, the Fiscal Incentive Act, and the Hotel Aid Act. Also, some investors benefitfrom income tax holidays. 32 1.28 The main purpose of Table 1.3. Annual Revenue Forgone fromTax Concessions tax concessions in Dominica, (as percentage of GDP) as in other OECS countries, is Fiscal Year 1999/2000 2000/01 2001/02 2002/03 to attract foreign investment. Hotel (a) 0.13 0.27 0.15 0.17 However, the empirical Education and Culture (b) 0.07 0.05 0.04 0.01 Health (d) 0.02 0.00 0.00 0.00 evidence suggests that taxes purposes(d) and are not amongof theforeign 0.26 0.26 0.48 0.36 determinants main Military forces (d) 0.02 0.02 0.03 0.09 Diplomatic mission (a) 0.07 0.04 0.12 0.04 investment. Foreign investors Movement of persons (personal effects) (b) 0.16 0.13 0.17 0.11 place more value on a stable Fisheries(b) 0.09 0.03 0.04 0.03 economic enVir0nment with Agricultural and forestry (b) 0.78 0.20 0.09 0.05 Mining and Quarrying 0.00 0.00 0.00 0.00 transparent rules than on tax 0.05 0.03 0.05 0.02 Other approved purposes Statutory Bodies (e) incentives." The most suitable (vehicles) (d) 0.23 0.18 0.07 0.04 approach to address the concerns Mo;';;;,l,p:rons of tax competition among OECS 0.23 0.20 0.18 0.15 Industry (a) 1.32 1.17 I.42 1.18 countries i s to support, more Govemment-a~~rovedspecified forcefully, a program of fiscal duty exemptions (c) I.64 2.06 I.95 1.30 Govemment-approvedspecified policy convergence that includes chz;;;;$;a;;;;z$i;;; 0.24 0.11 0.13 0.23 tax memberharmonization among 0.00 0.00 0.00 0.01 - states. Church fi) . , 0.02 0.03 0.07 0.04 Sports (b) 0.01 0.01 0.01 0.01 Shipping (b) 0.00 0.01 0.00 0.00 Aircraft (b) 0.00 0.00 0.00 1.15 1.29 A complete discussion Total 535 4.19 4.99 4.99 ~ of a medium-term tax policy Classification strategy is beyond the scope of ~?~r~i.~~~ment' 1.79 1.75 2.19 1.85 this report. The IMF Fiscal (a) 3.50 3.01 2.76 3.13 Established procedure Affairs followed (b) 1.37 0.66 0.61 I.56 produced two reports on the Department has Granted at the Minister of Finance's discretion' (c) 1.88 subject and estimates that the 2.18 2.08 1.52 , a r t ~ ~ ' ~ ~ { ~ ~ e ~ ~ ~ ~ c0.25s , t 0.18 0.07 0.04 revenue implications of its tax partly purely discretionary (e) 0.05 0.03 0.05 0.02 recommendations (short-term g::), 535 4.19 4.99 4.99 revenue stabilization and (c), and (d): Tax concessions granted under the Fiscal Incentive Act and the Hotel Aid Act. medium-term structural 1. Granted under the Fiscal IncentiveAct and the Hotel Aid Act. 2. Based on Cabinet decision. an reform measures) will result in increasein revenue 3. Based on Customs Control and Management Act. Source: Dominica authorities. collection of between 2.8 to 3.0 percent of GDP (Dos Santos 2003; Stotsky, Dieterich, and Bristow 2003). The main recommendations of these reports can be summarized as follows: Consumption and Sales Tax. Apply the legislated rates of consumption tax to petroleum products. Extend the sales tax, at the current rate and threshold, to a range of services including construction, public transport, professional services, electricity and water, and others. In the medium term, implement a value-added tax (VAT) to replace the consumption tax, the sales tax, and the hotel occupancy tax, and a number of minor levies (the petroleum tax on a specific basis and the others on an ad valorem basis). ~~ ~~ l9 See Zee, Stotsky, and L e y (2002) and the references therein. 33 0 Customs Duties. In the medium term, continue to reduce import duty rates, consistent with the safeguarding o f revenues and obligations under regional trade agreements and the World Trade Organization. 0 Income Taxes. Leave the stabilization levy in place until measures to broaden the base can be identified. With appropriate base-broadening measures, reduce the top marginal personal income tax rate of 40 percent to 30 percent, inline with regional trends, and eliminate the stabilization levy. Revoke exemption from income tax for profits from real estate sales and rentals, and tax interest on bank deposits. 0 Property Tax. Use site values developed by the valuation unit in the M o F to assess properties for the property tax. Levy a tax rate of 1.5 percent on property inurbanmunicipalities. Tax concessions. Award tax concessions only in accordance with a strict interpretation of the Hotel Aid Act and Fiscal Incentive Act. Repeal other acts authorizing tax incentives and cease granting tax incentives on an ad hoc basis through cabinet or ministerial decisions. D o not renew income tax holidays and do not grant new ones. With the replacement of existing direct taxes with moderate taxes, such as the VAT and excises, do not extend concessions under the old direct tax system to the new tax system (VAT). Central Government Expenditures 1.30 CG noninterest expenditures increased from about 31 percent of GDP in FY1993/94 to an average of about 34 percent of GDP during FY1994/95-FY1997/98. In FY1998/99, noninterestexpenditures reached 39 percent of GDPand increased steadily, reaching a peak of 44 percent of GDP inFY2000/01 (see Figures 1.5 and 1.6). Since then, they have been contained to about 33 percent of GDP in FY2001/02 and FY2002/03. As Figure 1.5 shows, the largest part of the increase was in capital expenditures and interest expenditures associated with the increase in borrowing. These increases have led to the need for a fiscal adjustment focused mainly on the wage bill in order to achieve fiscal solvency. 1.31 During FY1993/94-FY 2000/02, personal emoluments and wage expenditures remained stable at between 14.4 and 16.0 percent of GDP, but in the last two fiscal years the ratio increased to about 18 percent of GDP. However, the wage bill in constant EC dollars has increased at about 3.8 percent per year during FY1996/97 to FY2002/03. More notably, in FY1996/97 and FY2000/01, the wage bill in constant EC dollars increased by 11.O percent. At the core o f the increases observed in the wage bill in real terms (in constant 1995 EC dollars) i s an increase in the number of established and non- established positions in the CG.*' Because Government's real wages remained flat during 2oThere are two types of government employees in Dominica-established and non-established-between which there is a legal distinction. Established workers refer to those positions specified in the Public Service Act of 1991and its subsequentamendments. Non-established workers are those in positions not specified inthe legislation. 34 that period, and the Government had already cut the wages of workers in established positions by 5 percent inFY2002/03, Figure 1.5. Dominica CentralGovernmentExpenditure Composition,FY1993/9CFY2002/03 (as percentage of GDP) 40 0 30 0 20 0 100 0 0 1993/94 1994/95 1995/96 1996/97 1997/98 1998/99 1999/00 2000/01 2001/02 2002/03 ElPersonalEmolumentsandWages 0 OtherGoodsandServices 0 TransfersandSubsidies 0CapitalExpenditure InterestPayments Source: IMF. Figure 1.6. Dominica CentralGovernment Expenditure Composition,FY1993/9&FY2002/03 (in millions EC$1995) 350 00 300 00 250 00 200 00 150 00 10000 50 00 0 00 1993/94 1994/95 1995/96 1996/97 1997/98 1998/99 1999/00 2000/01 2001/02 2002/03 PersonalEmolumentsandWages Other Goods and Services 0 TransfersandSubsidies 0 InterestPayments HCapitalExpenditure Source: IMF. any effort to reduce the wage bill would have to focus on a rationalization of the number of public servants in the CG. The existing number of established and non-established positions is too large to be fiscally sustainable (see Chapter 4). 1.32 Capital expenditures (as percentage of GDP), on average, doubled during FY1994/95-FY1999/2000 compared to the rest of the period. In constant EC dollars, capitalexpenditurestripled. InFY1993/94, capital expenditures were about 4.8 percent of GDP (approximately EC$27 million in constant 1995 EC dollars). From FY1994/95 to FY2000/01 capital expenditures fluctuated around 8 to 17 percent o f GDP (approximately EC$49.0 to EC$122.0 million in constant 1995 EC dollars). The main reasons for this 35 increase were large unproductive investment projects such as an infeasible new international airport, an unfinished stadium, and a financial complex that remains largely unoccupied because of the retrenchment in the offshore sector. Therefore, tighter control of the PSIP i s crucial to improving the efficiency of the public investment and to helping return the government to a fiscally sustainable path (see Chapter 3). Expenditures on goods and services duringFY1993/9&FY2001/02 were stable at around 5.4 percent of GDP. However, in FY2002/03 expenditures on goods and services were compressed to only 4 percent of GDP. Transfers and subsidies have beenstable at 5.0percent of GDP. 1.33 Finally, interest payments as a percentage of GDP increased 2.5 times during FY1993/94-FY2002/03, and almost tripled inconstant EC dollars. This increase reflects two factors: (a) the higher indebtedness of the Dominican government, and (b) the increased reliance of the government on borrowing at higher costs. Already, interest payments are crowding out the other components of government expenditures. Social Security Scheme 1.34 It is critical that the Government maintain regular contribution payments, to the Dominica Social Security, diversify its portfolio into regional and international assets, and reduce its current concentration in public sector assets. The DSS i s a pay-as-you-go system. The benefits are financed by current contributions-that is, 10 percent of wage income of which 7 percent i s paid by the employer and 3 percent by the employee.21 In terms of its role in resource allocation, there are concerns that DSS's investment strategy i s not sufficiently insulated from Central Government and political pressure. Inparticular, the government has accumulated contribution arrears to DSS and relies heavily on borrowing from it to finance its budget deficits.22 Currently, DSS holds one-third of government domestic financing and a significant percentage of its recent bond issues. The Government resumed employer and employee contributions to DSS starting in July 2003, and i s working to clear accumulated arrears through a transfer o f landholdings and other assets. Moreover, as with many social security funds worldwide, there i s a need to ensure that the long-term investment strategy i s sufficiently diversified into regional and international assets and away from heavy concentration on public sector assets. To that end, the DSS Board of Directors recently adopted a policy that 20 percent of the DSS portfolio should be invested outside of Dominica. The Government has also instructed DSS to prepare a medium- and long-term investment strategy to achieve this diversification. 21For a more detailed analysis of DSS, see World Bank (2003) and a follow-up review of the status of Dominica's pension system conducted by ajoint IMF-World Bank mission in June 2005. *'Theaccumulation of arrears began at least seven years ago when the previous administration started a policy of withholding employer and employee social security contributions to DSS, which was later continued by the current government. By end FY 2001/02, the contribution arrears to DSS had amounted to 5.1 percent of GDP. 36 2. BUDGETMANAGEMENT 2.1 Budget procedures, which lead to the formulation, approval, and implementation of the budget and its impact on fiscal outcomes, have received considerable attention in the recent literature. Several empirical studies have provided evidence that budget institutions have an impact on fiscal outcomes in developed and developing countries. Von Hagen (1992) and Von Hagen and Harden (1995) found that European countries with strong institutional rules governing the budget process have lower expenditure, deficit, and debt ratios. Eichengreen (1992), Alt and Lowry (1994), and Poterba (1994), among others, have studied the effect of fiscal restraints on fiscal outcomes for the United States, and found that more restrictive state fiscal institutions are correlated with more rapid fiscal adjustment. Alesina and others (1996), and Stein, Talvi, and Grisanti (1999) have extended this line of research to Latin American countries. They found evidence that countries with centralized, transparent budget processes and strict formal or informal constraints have lower deficits and debt levels. Jones, Sanguinetti, and Tomassi (1999) obtained similar results analyzing public expenditures in the Argentine provinces. Lao-Araya (1997) corroborates this finding for Thailand, Gleich (2003) for transition economies, and Campos and Pradhan (1996) for several developing countries inAsia and Africa. 2.2 Consequently, a good public expenditure management system provides a government with the incentive and means to promote fiscal discipline and efficiency. However, Dominica's high levels of public debt, fiscal deficit, and arrears indicate that the budgetary processes in place have not been successful in this regard. In late 1997, Dominica initiated a budget reform program that led to some improvements, including for the first time in many years the adoption of the FY1999/2000 budget before the beginning of the new fiscal year. However, Dominica did not implement the full package of budget management reforms, with the result that the financial management system in place was not adequate to monitor budget outcomes. Today's difficult fiscal situation i s the combination of large debt incurred to finance unproductive investment projects, increasing interest payments, a shortfall in revenues during 2001-2003 (due to a protracted recession), and a weak financial management system, notably weak commitment controls, inadequate revenue forecasts, and poor cash management. 2.3 The government, faced with lower revenues than expected and access to domestic and external financing only at high interest rates, was unsuccessful in adjusting downward commitment limits in expenditures within the fiscal year, and expenditure allocations in the budget until some time in FY2002/03 (as part of the SBA agreement supported by the IMF) and then more significantly afterwards (as part o f the PRGF program also supported by the IMF) as required inresponse to the revenue shortfall. The result was that Dominica builtup sizeable arrears in the process. In October 2002, a simple cash management system was established with the help of the Caribbean Technical Assistance Center (CARTAC) to tighten controls over spending. Moreover, to tighten control over spending and regain fiscal sustainability, the authorities have been working with a consortium of donors, including the World Bank, to put in place an effective system of cash management and expenditure control, as a matter of priority, and to outline a medium-term program to further improve financial management along the recommendations of the recently completed CFAA. The system now in place has been successful in limiting expenditure commitments to collected 37 revenues. However, the system i s new and will require continuing improvement as experience i s gained. Moreover, the authorities need to be committed to reducing expenditure commitment limits within the fiscal year if revenues fall short of expected projections, in order to avoid the buildupof new arrears. 2.4 This chapter focuses on the rules governing the decision-making process that leads to the formulation of the budget, its passage through the legislature, and its implementation, execution, and monitoring, and suggests changes in the budgetary processes and framework that can help to achieve fiscal discipline. The main findings and recommendations can be summarized as follows: OnLegislation The Government intends to replace current legislation governing financial management with a unified framework under an updated Finance (Administration) Act. The new regulatory framework should be revised taking into account the recommendations of the 2003 Dominica Country Financial Accountability Assessment (CFAA). (See Box 2.1, recommendations 1 through 4.) In addition, the government should consider: 0 Introducing more stringent provisions governing the use of advances and deposits. 0 Eliminating the power of the Financial Secretary to transfer funds between programs in agencies other than the MoF. This is to safeguard program allocation at the sectoral level from arbitrarychanges at the central level. 0 Broadening the scope o f the FAA to a Fiscal Responsibility Law, in particular by including ceilings on fiscal outcomes such as debt-to-GDP ratio and fiscal balances. 0 Consolidating all Loans Acts. On Budget Preparation, Execution, and Monitoring 0 The Dominica PRSP now in preparation needs to be framed into a medium-term economic framework that addresses the following issues: (a) estimates of aggregate resources available for public expenditure consistent with macroeconomic stability, (b)bottom-up estimates of the cost of carrying out ongoing and new policies based on reliable assumptions, and (c) reconciliation of the aggregate resources with the cost estimates of the ongoing and new policies. Inaddition, corporateplansneedtoberevisedineachsectortofeedintothePRSP. It i s recommended that budget committees within each agency be established to lead the corporate planning process, prepare the budget submission, and monitor implementation. 38 Box 2.1. SelectedRecommendationsfromDominica'sCountry Financial Accountability Assessment LEGALAND REGULATORYFRAMEWORK Recommendation1. Update and modemize the Finance (Administration) Act (FAA).It is recommended that the Government take the following actions: Identify areas that needrevision Develop revised text Submittext for parliamentary approval. Recommendation2. Update the provisions of the Loans Act. Hire a specialist to review and recommend changes to the Act to increase its effectiveness in guiding and controlling Dominica's expenditure and debt strategy. Recommendation3. Update the provisions for the purchase and disposal of govemment lands and the provisions of the Land Acquisition Act. Prompt action should be taken to establish a formal land registry and to update the accounting practices and laws governing procedures to record, sell, and buy land, and how interest is accrued, recorded, and paid or collected. Recommendation4. Consolidate regulations and laws affecting Public Financial Management (PFM) intc one framework law. This will simplify compliance by allowing one single reference to govern PFM and provide a comprehensive financial legislation framework. CASH MANAGEMENT Recommendation5. Prepare and establish payment priorities and commitment limits. Institute an interilr policy requiring the Government to forecast cash availability and establish payment priorities anc commitment limits. This policy will control when expenditures should be paid, serve as an early-warning system when expenditures payments may exceed cash availability, and provide a guideline to the Accountant General for authorizing payments in the case of a cash shortage. Recommendation6. Improve effectiveness of cash management. Increase the government's capacity tc prepare accurate monthly revenue and expenditure estimates by procuring technical assistance (consultant). The consultant will assist the management teams of the line ministry in preparing the monthly reports of their actual and expected revenues and expenditures, in accordance with the requirements detailed in the Circular issuedby the Finance Secretary in November 2002. Recommendation7. Simplify cash management in govemment bank accounts by closing all accounts except: Where separate bank accounts are requiredby law or trust instruments For temporary payroll accounts that are cleared after each batch of payroll checks is cashed For imprest accounts operated by named officers, which have to be cleared at the end of each year. This will simplify the bank account reconciliation process, contribute to a more efficient allocation of cash resources, and allow investment of surplus funds. Recommendation 8. Issue guidelinesfor the management of public funds in government bank accounts by issuing a circular that: Explicitly establishes the criteria neededto open a govemment bank account Reminds ministriesto submit monthly reconciliations o f government accounts to the Accountant General Remindsministriesthat public funds belonging to the government should be made payable, by way of checks or currency, to the Accountant General, and deposited in the consolidated fund. I I Source: Dominica's Country Financial Accountability Assessment, August 2003. 39 Extend the budget cycle from 18 to 24 months to allow additional time for strategic analysis and for agencies to revise their budget proposals following their negotiations with the MoF. Complete the budget manual to document policies and procedures. The capital budget prepared by the Planning Unit and the development of a medium- term Public Sector Investment Program (PSIP) with growth and poverty at its core need to be an integral part of the PRSP. The preparation of the FY2003/04 capital budget and a three-year rolling PSIP, with the technical assistance of the World Bank and the Barbados Government, i s an improvement compared to previous capital budgets. However, the challenge for Dominica i s to further strengthen the three-year PSIP and ensure the successful preparation and implementation of its capital budget inthe coming years. A complete discussion of the PSIP and policy recommendations i s the focus of Chapter 3. The current program presentation in the budget i s insufficient to link expenditures to policy priorities such as growth and poverty reduction, a key element in a PRSP. Therefore, it i s recommended that expenditures be disaggregated at the program level in a way that will be useful for monitoring outcomes, such as inhealth, education, and social protection. Moreover, it i s recommended that the government adopt international best practices such as the ones contained in the IMF's Government Financial Statistics for its budget presentation. The debt management function i s the weakest component of the financial management information systems. The government i s in the process of establishing the Commonwealth Secretariat Debt Recording and Management System (CS- DRMS) module for domestic debt and making domestic and external debt data computer accessible to both the M o F and Treasury. Moreover, the Government has established a Debt Committee to oversee the signing and management of its debt. However, the government needs to consolidate the debt management function in one unit that maintains domestic and external debt records. Moreover, the unit needs to be staffed with qualified personnel that can keep adequate debt records and advise the government on debt management issues. The Standardized Integrated Government Financial Information System (SIGFIS) in Treasury has resulted in timely reporting. However, there are still shortcomings that curtail the effective operation of SIGFIS. Online access to the system needs to be extended to all line ministries with priority given to the commitments control system (fundscontrol module). Dominica uses the cash basis of accounting, in which Treasury prints checks for all payments to bring expenditures to account. In recent years, the government has not been able to meet all o f its payments; so many checks have not been issued to creditors and are held at Treasury. This has created a large divergence between expenditure records and bank account statements, and a security risk from holding large numbers of unissued checks. Reconciliation with the main bank account i s 40 manual and running two months behind. The government should adopt the recommendations in the 2003 Dominica CFAA in this respect (see B o x 2.1). In addition, it i s recommended that the Accountant General follow the St. Lucia system that avoids the need for printingchecks for all outstanding bills. The cash management system in place until FY2001/02 was weak in monitoring Dominica's cash shortage. Since October 2002, the government, with the help of CARTAC and a consortium of donors (including the World Bank) has put in place an effective system of cash management and expenditure control. The system now in place has been successful in limitingexpenditure commitments to collected revenues. However, it i s critical that the Government continue this discipline to avoid the build- up of arrears On Budget Accountability 0 The financial statements of Dominica should cover all government activities. The operations of statutory bodies, including state-owned enterprises, should be included inthe financial statements. 0 Weak Public Accounts Committees (PACs) are a common feature in the OECS countries. The recent Country Financial Accountability Assessment (CFAA) report for the OECS recommended a number of measures, including the following, to begin addressing this common problem: (a) Given that opposition parties are often insufficiently represented to form a functioning PAC, each country should consider enacting the appropriate constitutional or other changes to allow the appointment of nonparliamentary members. (b) Members of Parliament and other interested parties should participate in the Commonwealth Parliamentary Association, which, in conjunction with the World Bank Institute, has organized a Study Group on Public Accounts Committees and plans follow-up seminars. (c) OECS countries should draw on the advice of neighboring countries such as Jamaica, where the PAC i s a vibrant component of government oversight. (d) A media campaign should be developed to publicize PAC activities to the public and create demand from the public for strong public financial management. A. CONSTITUTIONALAND LEGAL FRAMEWORK 2.5 The Commonwealth of Dominica attained independence in 1978 as a Westminster-style parliamentary democracy. Dominica i s a unitary state with a President elected by the Parliament. Executive authority lies with the Prime Minister and the Cabinet. Figure 2.1 presents the government structure. 41 2.8 The current FAA should be revised in line with the model financial legislation developed for the OECS countries under the Eastern Caribbean Economic Management Program (ECEMP). The Government intends to replace the current legislation governing financial management with a unified framework under an updated Finance (Administration) Act. The new regulatory framework should be revised taking into account the recommendations of the 2003 Dominica Country Financial Accountability Assessment (CFAA). (See Box 2.1, recommendations 1 through 4.) In addition, the government should consider: Introducing more stringent provisions governing the use of advances and deposit accounts. Eliminating the power of the Financial Secretary to transfer funds among programs in agencies other than the MoF. A more detailed discussion is presented inparagraph 2.18. 0 Broadening the scope of the FAA to a Fiscal Responsibility Law, in particular by including ceilings on fiscal outcomes such as debt-to-GDP ratio and fiscal balances. The Brazilian Fiscal Responsibility Law can be a good model to follow. 2.9 Legal authority and the regulatory framework to contract external and domestic debt rely on at least five Acts. However, maximum ceilings on debt are set per year and only for domestic loans to finance certain development projects.24 The Loans Act authorizes the Government of Dominica to borrow money from approved sources (external) for the purpose o f general development and other matters. The power of authority rests with the Minister of Finance. The General Local Loans Act declares the terms and conditions applicable to local loans authorized to be raised by the Government of Dominica. It gives the Minister of Finance the overall mandate to issue bonds and securities and obtain local loans. The Local Loans Act confers power via the Minister of Finance to raise up to a maximum of EC$3 million per year to finance certain development projects. The Bonds and Securities Act authorizes the Government o f Dominica to issue bonds with respect to certain financial obligations of the State, to provide for the issue of National Savings Bonds, to make provision for the redemption or exchange of bonds and securities, and for purposes connected therewith. Bonds under this Act can be issued locally or overseas, and can be either general or National Savings Bonds. The power to borrow rests with the Minister of Finance, but can be delegated to the Financial Secretary. 24This paragraphis basedon the report, "Strengthening Debt Management," Oxford PolicyManagement,May 2002. 43 0 The Treasury BillsAct authorizes the Government of Dominica to borrow money for the purposes of Dominica. The power of authority rests solely with the Ministerof Finance, but may be delegated to the Financial Secretary. The authorities are in the process of revising the framework for government borrowing in order to strengthen controls and accountability. In this context, it i s recommended that a consolidation of all Loans Acts with a ceiling on total accumulated stock of debt be established. 2.10 Other legislation that affects financial managementinDominica includes: 0 Revenue legislation-which. defines taxes and fees-establishes tax bases and rates; assigns responsibilities for assessments, collections, and accounting; and sets penalties and means of enforcement. 0 Fiscal Incentives Legislation, mostly the Fiscal Incentives Act and the Tourism Incentives Act, which defines the conditions under which exemptions from taxes, customs duties, service charges, and fees are granted. 0 Acts, which establish statutory bodies (including state-owned enterprises). 0 The Audit Act of 1994. This requires the Director of Audit (DOA) to report to Parliament at least once a year on the work of the office and key findings related to the audit of the government's accounts and records. The D O A i s also empowered to make special reports. B. BUDGETPREPARATION,EXECUTION, MONITORINGAND Preparation 2.11 Dominica has a dual budget system. The Budget unit in the MoF is responsible for recurrent expenditures and for monitoring overall expenditures and revenues, while the Planning Division takes the lead with respect to capital projects. Dominica's fiscal year runs from July 1to June 30. Inthe past, the budget has rarely been passed before July 1. In FY2003/04, the budget was passed on July 8, 2003. Table 2.1 presents the FY2001/02 budget. As is discussed in paragraph 2.16, the budget is presented by administrative classification only, rather than by functional classification. The budget cycle i s an 18-month process, as presented inFigure 2.2. 44 Table 2.1. Dominica'sBudgetOverview, FY2000/01 (in mill. EC$) CurrentBudget Current Revenue 200.6 Current 230.1 Current Expenditure 230.1 (28.0)* Expenditure (32.1) Administrative Classification (32.1) Tax Revenue 169.6 Personal 116.2 blinistry of 18.9 (23.7) emoluments (16.2) Zommunications, Works (2.6) and wages mdHousing Taxes on 54.6 Transfers and 40.4 Ministry of Community and 9.3 income & subsidies Developmentand Gender profits (7.6) (5.6) (1.3) Affairs Taxes on 2.6 Other goods and 37.5 Ministry of Education, 41.6 property (0.4) services (5.2) Sports and (5.8) Youth Affairs Taxes on 28.6 Interestpayments 36.0 Ministry of Finance 92.4 domestic (12.9) goods and (4.0) (5.0) services Taxes on 83.8 Ministry of Healthand 32.2 international (11.7) SocialSecurity (4.5) trade Non-Tax 31.0 Others ** 35.6 Revenue (4.3) CapitalBudget CapitalRevenue 3.0 (0.4) Capital 120.2 CapitalExpenditure 120.2 Expenditure (16.7) Administrative Classification (16.7) Ministry of 15.7 Communications, Works (2.2) and Housing Ministry of Communityand 2.8 Development and Gender (0.4) Affairs Ministry of Education, 10.9 Sports and (1.5) Youth Affairs Ministry of Finance 25.2 (3.5) Ministry of Healthand 3.4 Social Security (0.5) Others * 62.3 *Figures in parentheses are percentage of GDP. **I&ludes President, Ministry of Legal Affairs, Immigration and Labor, Legislature, Audit, Elections, Ministry of Foreign Affairs, Trade and Marketing, Prime Minister's Office, Ministry of Agriculture and the Environment, Ministry of Tourism, and Ministry of Industry,EnterpriseDevelopmentandPhysicalPlanning. Source: IMFandEstimatesof the Commonwealthof Dominica FY2002/03. 45 2.12 A budget reform program initiated in late 1997, as part of the ECEMP-2 program, introduced the following changes: 0 Conversion of the budget from a line item format to a program format. 0 The development of Corporate Plans b y each ministry or nonministerial department (agency) outlining its mission and priorities, plus the objectives for each of its programs. 0 The publication of all Corporate Plan summariesinthe Annual Estimates. e Preparation at the start of each budget cycle of a Strategic Outlook with recommendations to Cabinet for fiscal targets, sectoral limits for the public sector investment program (PSIP), and preliminary allocations for recurrent spending. 2.13 The reforms led to some improvements, including for the first time in many years the adoption of the FY1999/2000 budget before the beginning o f the new fiscal year. However, Dominica did not implement the full package of budget management, such as: Extending the budget cycle from 18 to 24 months to allow additional time for strategic analysis and for agencies to revise their budget proposals following their negotiations with the Ministry of Finance. Establishing budget committees within each agency to lead the corporate planning process, prepare the budget submission, and monitor implementation. Establishing upgraded Finance Officer positions within each agency to strengthen budget planning and implementation and, more generally, financial management at the agency level. Ensuring adequate staffing in the Budget, Debt, and Fiscal Management Department. Strengthening coordination between the Budget Unit and the Planning Unit. Completing the budget manual to document policies and procedures. The implications of these omissions and other shortcomings are discussed below. 2.14 The budget, in line with the 1997 budget reform, is supposed to be prepared based on the Country Strategic Outlook, the current Medium-term Economic Strategy Paper (MTESP), the capital budget, and each ministry's Corporate Plan. However, in recent years most of these key documents have not been prepared, with the result that the budget has lacked a medium-term macroeconomic framework or, at best, it has been driven by short-term fiscal stabilization considerations. The Strategic Country Outlook (a yearly document that presents the economic prospects of the country for the next fiscal year) has not been prepared since FY2001/02, and the MTESP (a four-year economic plan that provides the basic information for the next budgets) was last prepared for the 46 Caribbean Group for Cooperation in Economic Development (CGCED) Meeting in 2000. For the meeting of June 2002, Dominica did not prepare an MTESP because discussions were taking place with the IMF for the Stand-by Arrangement (SBA) approved on August 28, 2002 to support fiscal stabilization. In mid-2003, Dominica further revised its macroeconomic framework for the FY2003/04 budget to address the continuing shortfall in revenues in response to the continuing recession and the consequent need to introduce additional measures to mobilize revenues and cut expenditures. The absence of a medium- term strategic economic framework, and with most of the ministries' Corporate Plans being nonexistent, out of date, or not able to provide a link among objectives, expenditure required, and outcomes, have impliedthat the FY2002/03 budget and the FY2003/04 budget have been driven by short-term stabilization considerations. The preparation b y Dominica of a PRSP and of a medium-term macroeconomic framework under an IMF-supported PRGF currently under preparation i s expected to overcome this shortcoming inthe budget formulation. 2.15 Until recently, the capital budget prepared by the Planning Division was an afterthought to the country budget instead of an integrated element of the budget. However, since FY2002/03,improvements have beenmade inthe capital budget. Before FY2002/03 the capital budget, which presents the current investment plan, and the PSIP, which presents the investment plan for future years, were a wish list of investment projects rather than a well-thought-out investment plan with growth and poverty reduction as main objectives. Improvements were made in the capital budget for FY2002/03, and more substantially for the FY2003/04 and follow-up years capital budget, with the assistance of the World Bank and the Government of Barbados. Now, the challenge for Dominica i s to strengthen its medium-term PSIP and at the same time ensure the successful implementation of its capital budget. A complete discussion of the PSIP and policy recommendations i s the focus of Chapter 3. 2.16 The budget provides an accountant's view of government transactions rather than a presentation of the government's fiscal policy. The budget adheres to the traditional administrative, program, and economic structure format, but no breakdown by department i s provided. Moreover, this presentation i s insufficient to assess the fiscal position of the government and to follow up on key programs reportedly linked to policy priorities. For instance, it i s almost impossible to obtain from the budget the amount allocated to crucial safety net programs and to follow them over time. The latter i s one of several key indicators that the PRSP should track to evaluate the success of the program; thus, the government should start to rethink its budget presentation to link government expenditures with the PRSP targets. 2.17 The Ministry of Finance does, however, have the capacity to present the budget by standard functional classification, and it is recommended that it start doing so. It is recommended that the budget presentation follow international best practices, such as the ones contained in the IMF's Government Financial Statistics (GFS). This will also facilitate investors in the Regional Security Market to more accurately assess the fiscal position of the OECS countries. 47 2.18 The Dominica budget scores well in comprehensiveness, but unlike other OECS countries, the Financial Secretary can transfer funds among programs inagencies other than the MoF. Dominica's budget i s comprehensive, generally covering all CG expenditures. However, the FAA empowers the Financial Secretary to vire, or transfer, funds between programs within the same supply vote (agency). This means that funds can be transferred to a completely new recurrent program within the same agency, or from one capital project to a completely new one without the obligation to obtain subsequent endorsement from the Parliament via a supplemental appropriation. For example, in FY2002/03 the government used funds budgeted for repairs inprimary schools to pay capital expenditures for the Windsor Park stadium. Intheory, any such virements, or transfers, and, ultimately expenditures, should at least be reported in the next year's Estimates Books. However, the Estimates Books do not always contain complete and accurate figures for revisions and prior year expenditures, particularly with respect to capital projects. This could mean that the government can legally spend funds on programs or projects without the Parliament ever becoming aware that such a program or project existed. The FAA should consider empowering only the Minister of the respective agency to transfer funds between programs up to a ceiling, and if the transfer exceeds the ceiling, the transfer will need the approval of the Parliament and the MoF. This i s to safeguard program allocations at the sectoral level from arbitrarychanges at the central level. Execution 2.19 In terms of budget execution and its procedures, the evidence presented below suggests a trend to budget expenditures beyond the program requirements (in the mid- 1990s) and to the revenue collected (in the early 2000s). As Figure 2.3 shows, in every fiscal year during FY1994/95 to FY2002/03, actual recurrent expenditures have been below budgeted expenditures. Until FY1998/99, this was accompanied by a current fiscal surplus. However, since FY1999/2000, actual expenditures have been below budgeted expenditures, at the same time that the country has run a current fiscal deficit of between 1and 7 percent of GDP and builtup sizeable arrears. This is evidence that until FY2002/03 the budget did not adequately reflect the resources available to the country and the program requirements. 2.20 However, actual interest payments consistently exceeded the budgeted amount. As Figure 2.4 shows, debt servicing was consistently underestimated every fiscal year during FY1994/95 to FY2001/02. This pattern i s consistent with the increasing reliance o f the government on domestic and external debt to finance its fiscal deficit, and the weak debt management function. Preliminary figures for FY2002/03 suggest that the SBA succeeded in a better budgeting of interest payments (see paragraph 2.28). 2.21 All line ministries, with the exception of the MoF and Elections, spent below their budgeted allocation. As Figure 2.5 shows, the persistent underexpenditure, at least until FY1999/2000, suggests that ministries budgeted spending requirements in excess of what they needed, and that the succesive Cabinets were not succesful in setting an overall expenditure allocation in line with actual expenditures. In recent fiscal years, the cash constraint drove actual expenditures, with the government being successful in cutting budgeted expenditures only inFY2003/04. 49 Figure2.3. Variance betweenActual andBudgetedRecurrent ExpendituresinDominica,FY1994/95-FY2002/03 (inpercentage) 0 00 PiCdYear -1 00 2E -200 3 d -300 5* a -400 -500 2 2z -600 B -700 2se4 5 04 -800 -900 -1000 Source: Author's calculations Figure 2.4. Variance betweenActual and BudgetedPrimary,RecurrentExpenditures, and InterestPaymentsinDominica, FY1994/95-FY2002/03 (inpercentage) 25.00 25.00 20.00 20.00 55:: v1 15.00 15.00 y k 0 I C W \ t 10.00 10.00 6 p $5 I 5.00 5.00 e ,x 3 2 - Year Z8 B a '-3 5 -5.00 9 103 m -5.00 D 2 m 2 3 -10.00 -10.00 -15.00 -15.00 PrimaryRecurrentExpenditures +-InterestPayments Source: Author's calculations. 50 Figure 2.5. Variance between Actual and Budgeted Recurrent Expenditures by Ministries inDominica, FY1994195-FY2000101 (inpercentage) X : 2 UY 3mz 2z (0 b CO m 7J 0 0 5 z 0) m co m m 0, 0 r z xz m 0 N --+- ...X ..Agriculture&the Environment -C- LegalMars lmmigratlon&Labour *Legislature Audit -X-Elections +Foreign Atfirs &Carib Attars +Rime Ministers Oftice - Finance&Planning __ Agrwlture &the Environment Education, Youth & Spons +-Communications Works & Hou\ing A CommuniryDev & Gender X Health&Social Secunry X Tourism &Trade Industry& Marketing Source: Author's calculations. Figure 2.6. Variance between Actual and Budgeted Revenues inDominica, FY1994195-FY2002103 (inpercentage) 1 10 0% i n 5 0% - t VT"" OO%; 2 01- 2 2002-03 1 -50% f% -10 0% - 0Non-taxrevenue 4 s -15 0% - =Domestic Goods & ConsumFrtion T Taxes onInternational Trade -20 0% - 0IncomeTaxes -2s 0% - -Total -30 0% ' Source: Author's calculations 2.22 On the revenue side, every year during FY1999/2000 to FY2002/03 actual revenue collected has been below budgeted revenue. As Figure 2.6 shows, the revenues collectedduring this period have been consistently below government forecasts. Moreover, 51 in each subsequent fiscal year the variance increased until it reached about 27 percent in FY2001/02. That is, in FY2001/02 collected revenues were 27 percent below the forecasted revenue. The main reason was the downturn experienced by the Dominican economy during 2001-2003, and the inability of the MoF to adjust its forecasts to the new economic environment and therefore present a realistic budget. Preliminary figures for FY2002/03 suggest that the SBA succeededin reducing the variance in revenue from about -27.0 percent inFY2001/02 to -10 percent inFY2002/03. (See Box 2.2 for a detaileddiscussion.) Box 2.2. Revenue ForecastinginDominica [n June 2000, Dominica presented a Fiscal Consolidation Program (FCP) as part of its Medium Term Economic Strategy Paper (MTESP) prepared for the Caribbean Group for Cooperation in Economic Development (CGCED). This acknowledged the unsustainable fiscal trend and proposed a variety of tax measures to raise an additional EC$26 million, which would stabilize government finances and support a modest, 6 to 8 percent of GDP, program of :apital expenditures. However, the budget presented the following month was markedly different from the FCP. The tax categories identified in the FCP [FCP Measures) were increased by less than EC$5 million rather than EC$26 million, while an entirely different set of tax measures (Budget Measures), which had not been mentioned in the FCP, were introduced to raise an additional EC$15.5 million. Subsequently, actual collections on the Budget Measures tax categories were 29 percent lower than budgeted, far worse than the amount collected for the FCP Measures (2 percent above budget) or for all Dther current revenue heads (19 percent below budget). Someof Variationin~WOl F?evewes curent RelativetotheFCP FcPMeanaes BucJg?tMeanres otheriterrs Source: Author's calculations The FCP Measures had been worked out over the months leading to the CGCED in June 2000, but the government then adopted a substantially different set of measures, which proved to be far less effective, in its budget delivered in July 2000. This suggests a remarkable disconnection at that time between MoF personnel working on the MTESP (who also prepare economic and fiscal forecasts for the budget) and the political directorate (who made the budget decisions). Source: Dominica MTESP (2000). 52 Monitoring 2.23 Although budget execution is monitored through several reports, the past practice of the Budget Unit to compile quarterly monitoring reports for review by the Cabinet was discontinued in 2000 and was reestablished only in 2002. Budgetary performance i s monitored by monthly reports on revenue and expenditure to the responsible departments, and quarterly reports by the M o F to the Cabinet. The Standardized Integrated Government Financial Information System (SIGFIS) in Treasury i s the key system for government accounting and budget monitoring. Until recently, the system resulted in timely reporting. However, between June and October 2002, the Accountant General could not generate reports because the latest version of Smart Stream installed was found to be incompatible with the version of Access software used.25 The Annual Accounts are submittedfor audit within the stipulated period (six months following the end of the financial year). 2.24 There are shortcomings that curtail the effective operation of SIGFIS. First, besides Treasury and the Personnel Department, only the Ministries of Finance, Agriculture, and Communications and Works can access the system online, thus curtailing their ability to access timely information from a large number of agencies. Second, only the following six modules: general ledger, payables, purchasing, payroll, funds control, and the budget have been implemented. The commitment control system (funds control) needs to be extended to all ministries. Third, skilled and experienced personnel remain in short supply. To address the first problem, computers have been purchased but the plans to go online in all ministries have been deferred for lack of funds to cover the network costs. The inadequate information up to FY2002/03 implied that expenditure allocations were not being adjusted during the budget year in line with either revenue inflows, available cash, and/or financing. As revenue contracted, this lack of control resulted inthe build-up of arrears. 2.25 Concerning revenues, the Standardized Integrated Government Tax Administration System (SIGTAS, Inland Revenue) and Automated System for Customs Documentation and Administration (ASYCUDA) systems computerize most revenue accountingtasks for the two major revenue departments. Unfortunately, the two systems are not integrated, and making them so would facilitate the maintenance of a single account per taxpayer covering all taxes. 2.26 Dominica uses the cash basis of accounting, in which Treasury prints checks for all payments to bring expenditures to account. The Consolidated Fundi s made up of five different bank accounts to pay expenditures, and two accounts to receive revenues. Expenditure i s recorded from day to day on an accrual basis (invoices are entered as they are received), with checks printed out for all unpaid invoices. In recent years, the government has not been in a position to meet all its ayments, so many checks have not been issued to creditors and are held at the Treasury! This has created a large divergence between expenditure records and bank account statements and, as mentioned, a security risk from 25The SBA with the IMFwas signed in August 2002. 26The Financial Secretary and the Accountant General meet weekly to decide which checks will not be issued. 53 holding large numbers of unissued checks. Reconciliation with the main bank account i s manual and running two months behind. It i s recommended that the Accountant General follow the St. Lucia system that avoids the need for printing checks for all outstanding bills by recognizing expenditures at the point invoices are received for payment, rather than when the payment i s issued. As expenditures have already been brought to account, there i s no need to issue payments by printing checks. Moreover, it i s advised that the government follow the recommendations of the 2003 Dominica CFAA with respect to the consolidation of bank accounts. (See Box 2.1, recommendations 5 through 8.) 2.27 The cash management system in place untilFY2001/02 was too weak to monitor Dominica's cash shortage. The authorities, working with a consortium of donors, including the World Bank, have made successful improvements. In October 2002, with CARTAC assistance, the government made an effort to improve its cash management capacity by introducing a system of spreadsheets to anticipate, control, and manage expenditure and revenue. Moreover, to tighten control over spending and regain fiscal sustainability, the authorities have been working with a consortium of donors, including the World Bank, to put in place an effective system of cash management and expenditure control, as a matter of priority, and to outline a medium-term program to further improve financial management along the recommendations o f the recently completed CFAA. The system now in place has been successful in limiting expenditure commitments to collected revenues. However, it i s critical that the authorities continue this discipline to avoid the build-up of arrears. On July 21, 2003, the authorities issued an Administrative Order to ensure these problems do not reoccur by establishing procedures to strengthen the cash management system and penalties for line ministries that do not follow the procedure^.^^ 2.28 The debt management function is the weakest component of the financial management information systems. The debt management function is split between a small unitwithin the Budget, Debt, and FiscalManagement Department, which uses the CS-DRMS to maintain data on foreign loans, and the Accountant General's Department, which uses Excel to maintain records on domestic debt (Treasury Bills, debentures, local loans). The CS-DRMS has a module for domestic debt, and the Treasury accessed this via modem to inputdomestic debt data untillate 2000, when the modem connection was discontinuedat the time SIGFIS was installed. The government is in the process o f establishing the CS-DRMS module for domestic debt, and making domestic and external debt data computer accessible to both the M o F and Treasury. Moreover, the Government has established a Debt Committee to oversee the signing and management of its debt. However, the government needs to consolidate the debt management function in one unit that maintains domestic and external debt records. Moreover, the unit needs to be staffed with qualified personnel that can keep adequate debt records and advise the government on debt management issues. *'The procedures entail issuing monthly, six months ahead and revised monthly, cash-flow projectionsbased on inputs from the revenue collecting departments and line ministries. The MoF will communicate to line ministriesspending limits consistent with the resources available and programtargets. The MoF will reduce or delay monthly allocations to ministriesthat fail to providecomplete information on time. 54 C. BUDGET ACCOUNTABILITY SupremeAudit Institution 2.29 While independent by authority derived directly from the Constitution, the Audit Department, has no guaranteed operating budget or staffing complement, and must obtain approval from the Minister of Finance via the annual Estimates process. The President, acting on the advice of the Public Service Commission, appoints the Director o f Audit. Before tendering advice, the Public Service Commission must consult with the Prime Minister, but there i s no provision for input from the Parliament. These processes are inconsistent with the need for the Audit Department to be independent of the executive, and the status of the Director of Audit (DOA) as a servant of Parliament. 2.30 In addition, a number of other factors have reduced potential benefits accruing from the work of the Audit Department,including: 0 The automation of critical financial management functions (for example, the Standardized Integrated Government Financial Information System [SIGFIS] ,the Standardized Integrated Government Tax Administration System [SIGTAS], and the Automated System for Customs Documentation and Administration [ASYCUDA]) without adequate input from the Audit Department in the design and implementation stages. 0 Problems with SIGFIS in terms of generating reports and in keeping separate accounts for advances issued to public officers. 0 The inadequate functioning of the Public Accounts Committee (see below). 0 Lack o f training for staff, particularly on computer-assisted auditing tools, auditing of computer systems, and value-for-money auditing. 0 The D O A has no role in the recruitment of personnel, and its recommendations concerning promotions are not always followed b y the Public Service Commission. This means that new staff may lack the requisite skills and aptitudes, and also means the loss o f experienced and capable officers. 2.31 In spite of these problems, the Audit Department does a commendablejob. The D O A examines the accounts of 38 Village Councils, in addition to Roseau City Council, Canefield Urban Council, and Portsmouth Council. Audit Reports are current and the Department has undertaken a number of special studies at the request of the Minister of Finance. 2.32 The financial statements of Dominica present a partial view of government activities. Significant fiscal operations are excluded. The operations of statutory bodies, including SOEs, are not covered. In addition, although contingent liabilities are included in the financial statements, statements o f public debt show only domestic debt and not external debt. The D O A has pointed out this omission. 55 2.33 Mostof the SOEs are significantlyinarrears in the preparationof their financial statements and their submission for audit and, eventually,beforeParliament. Audits of SOEs are contracted to private accounting firms in accordance with their enabling acts. The Director of Audit has the right to require submission of their audit reports and to supplement these by further investigations if he considers it necessary. He has not done so. Parliamentandthe PublicAccounts Committee 2.34 Followingthe 2000 election,a PublicAccounts Committee(PAC) was appointed by the government,but the Committee has met infrequently. PAC duties, as prescribed inHouse Standing Order #72, are to examine the accounts andaudit reports referredto them by the Parliament, and to report its findings. The Committee has four members and i s appointed in each new Parliament from elected members and appointed senators. Unlike in other OECS countries, government ministers are not excluded from the Committee, which weakens its independence. It i s an accepted convention (though not legally mandatory) that the chairman should be the Leader of the Opposition. The government appointed a PAC following the 2000 election. It has met on occasion, most recently to consider the accounts for the FY1999/2000 financial year, but it has not closed the accountability cycle by reporting its findings to Parliament.28 Moreover, the PAC has never reviewed the accounts of any statutory body. 2.35 Weak PACsare a commonfeatureinOECS countries. The recent CFAA report for the OECS recommendeda number of measures, includingthe following, to begin addressingthiscommonproblem: Given that opposition parties are often insufficiently represented to form a functioning PAC, each country should consider enacting the appropriate legal changes to allow the appointment of nonparliamentary members. Members of Parliament and other interested parties should participate in the Commonwealth Parliamentary Association, which, in conjunction with the World Bank Institute, has organized a Study Group on Public Accounts Committees and plans follow-up seminars. OECS countries should draw on the advice of neighboring countries such as Jamaica, where the PAC i s a vibrant component of government oversight. A media campaign should be developed to publicize PAC activities and create demand from the public for strong public financial management. Informationto andfromthe Public 2.36 The formulation of the budget receives inputsfrom the socialpartners,but this undoubtedly could be made more systematic. The current government consulted with representatives from various sectors and interest groups in preparing its recent budgets, and ** Inthe past 10 years, only one annual report was sent to the House, and it was not debated. 56 these consultations were extensive for the FY2003/04 budget. This undoubtedly could be made more systematic, and the agreement by OECS countries to establish National Economic Councils might lead in this direction. However, the capacity of the social partners-most of which are small organizations-to go beyond their immediate concerns and address complex national policy issues remains limited. 2.37 The budget address is broadcast live each year and copies of the budget are available to all membersof the public. There are three weekly newspapers, and there is no evidence of any restrictions on press reporting on financial mismanagement. On the other hand, the media have rarely followed up on cases of alleged financial mismanagement. 57 3. THE PUBLICSECTOR INVESTMENT PROGRAM 3.1 The conventional wisdom of the economics profession is that countries need capital to grow, and that a direct relationship exists between capital spending and growth. Consequently, when economists evaluate the allocation of public resources between current and capital spending in government budgets, they tend to be critical o f countries that allocate a large share o f government expenditure to current spending, and to applaud countries that spend on capital. 3.2 This bias is embedded in the "golden rule," which says that only current expenditure needs to be balanced by local revenue, but a country can (within limits) run a fiscal deficit equal to the capital spending of the government. This rule i s implicit in many Organization of Eastern Caribbean States (OECS) budget presentations, where the focus i s on the current fiscal balance, even though for fiscal sustainability the key concept i s the primary fiscal balance (overall balance minusinterest payments). 3.3 Because most current spending by a government reflects entitlements or previous commitments, such as salaries, payments on public debt, and so forth, governments have, in the short term, limited scope to influence it. In contrast, the capital budget and its composition are highly discretionary. In formulating the capital budget, governments make the decisions that at the end will determine the size, composition, and in some cases, the geographical location, of the capital budget. 3.4 Why does this matter? Because one of the main reasons behind the difficult fiscal position facing the Dominican government today i s the significant proportion of public investment allocated to unproductive projects financed through loans (mainly commercial) at high interest rates. Recent examples of poorly conceived projects include an infeasible new international airport, an unfinished stadium, and a financial complex that remains largely unoccupiedbecause of the retrenchment in the offshore sector. 3.5 Responding to a difficult macroeconomic situation, in August 2002 Dominican authorities signed a one-year Stand-by Arrangement (SBA) with the IMF that was extended until December 2003 and was then followed by a three-year PRGF. The authorities recognize that inthe current macroeconomic environment it i s critical that Dominica raise the productivity of its public investments to support a resumption o f growth. They also recognize that the weak management of the public sector investment program has in the past contributed to the poor choice o f projects, financing instruments, and low execution rates. Since FY2002/03, the Dominican Government has halted a number o f projects the impact of which on the economy was questionable, and has taken measures to terminate the government's related contractual obligations. The capital budget since FY2002/03 reflects a better realignment of investment projects with the Government's objective to stimulate private investment and growth. Moreover, the Government is in the process of strengthening the medium-term Public Sector Investment Program (PSIP) that will help establish a better selection process for future investment projects and a platform for growth. Nevertheless, the challenge for the Government i s to anchor this medium-term PSIP to the Poverty Reduction 59 Strategy Paper (PRSP) in preparation for, and to strengthen implementation and monitoring of, the capital budget. 3.6 On average, from FY1995/96 to FY2001/02 the Central Government (CG) spent about 9 percent of GDP on capital expenditures. This level of capital expenditures i s quite high compared to other OECS countries, but is similar to St. Kitts and Nevis, a country that was hit b y three hurricanes between 1995 and 2001, with resulting high investment reconstruction needs, and i s almost triple the level for the Latin American and Caribbean Region. Total public sector capital expenditures were on average about 11percent of GDP in the same period. Figures for FY2002/03 show that capital expenditures were 5.1 percent of GDP; for FY2003/04 the planned capital expenditures were about 7 percent of GDP. It is difficult to assess the adequate level of public capital expenditure for the Dominican economy because capital expenditure figures reported by the Government, as in most OECS countries, include not only spending on fixed capital, but also some recurrent expenditures. If one subtracts from the historical average CG capital expenditures the unproductive capital expenditures (about 4 percent of GDP), and uses the evidence presented in this report for investment in education and health, the remaining level of capital expenditures intended to acquire fixed capital of 2 to 3 percent of GDP i s in line with the historical averages in the Bahamas (2 percent of GDP) and Barbados (2.4 percent of GDP) for the period 1990-2001. This excludes capital expenditures that may be needed for exceptional reconstruction needs inresponse to naturaldisasters. 3.7 The purpose of this chapter is first, to describe and draw attention to the limitations of the processes in place to prepare, implement, and monitor the PSIP in Dominica, and second, to assess the size and composition of the PSIP. 3.8 The main findings and recommendations can be summarized as follows: On Capital Expenditures and PSIP Preparation, Monitoring, and Execution 0 The capital budget and the three-year PSIP prepared by the PSIP Unit in the Ministry of Finance need to be made an integral part of the PRSP now being prepared by the Dominican Government. Moreover, the Budget unit should set the overall envelope for capital expenditures (consistent with macroeconomic stability) early in the year so that the PSIP unit can attempt to prioritize the projects from the start of the process instead of at the end. The government of Dominica needs to apply a selection criteria for investment projects (CG and the rest o f the public sector). First, only investment projects, independently of their source of financing, with all the required information (in particular projected recurrent costs) should be included in the capital expenditure budget and the PSIP. Second, only investment projects prioritized based on some type of cost-benefit analysis and in accordance with the overall resources available for capital expenditures should be included in the capital expenditure budget and the PSIP. Third, all projects with a cost above EC$2 million should be subject to a formal cost-benefit analysis before being included in the capital expenditure budget 60 or the PSIP. Fourth, given the high debt burden, only projects financed with concessional financing should be included. 0 The PSIP unit needs to be strengthened, including through the filling of current vacant posts. 0 At a technical level, staff needs to be trained in the procedures and tools of project analysis for preparing the PSIP. The PSIP unit should require the line ministries to report capital spending monthly and should stop disbursements to those projects that do not provide the required information at the end of the month. 0 Dominica needs to establish a central clearinghouse to coordinate donor assistance. On the Coverageand Transparencyof Capital Expenditures and the PSIP 0 Some investment projects currently included in the PSIP and as capital expenditures in the budget should be classified as recurrent expenditures. International practice should be followed inthis respect. 0 The government should ensure accurate representation of public sector investment by including the investment projects o f all statutory agencies in the PSIP. Investment projects implemented by private corporations but guaranteed by the government should be disclosed inthe government's financial statement as government liabilities. 0 The power of the Financial Secretary to transfer funds between capital projects in agencies other than the MoF should be eliminated. This would safeguard program allocation at the sectoral level from arbitrary changes at the central level. A. INSTITUTIONAL ORGANIZATION AND PREPARATION InstitutionalOrganization 3.9 Dominica has a dual budget system. The PSIP Unit in the MoF prepares the capital budgetandthe medium-termPSIP. The PSIPunit is headed by a senior economist and includes four economists responsible for broad sectors that encompass all the ministries. Their job i s to interface with these ministries in the preparation of the capital budget and the medium-term PSIP. However, as of today the unit i s underfilled, with three economist positions remaining vacant. There i s also a Macro Economic Unit, staffed with a senior economist (head) and an economist. This unit carries out the overall economic analysis that i s necessary to guide the process o f preparing the capital budget and the medium-term PSIP, including projecting Government revenues. The PSIP and Macro Economic Unit form the Planning Unit. The Budget Unit i s responsible for preparing recurrent expenditures in the budget. 3.10 Four different ministrieshave hadjurisdiction over the PSIPsince 1994. Inthe past, the PSIPwas prepared in the Economic Development Unit, which was originally part of 61 the Prime Minister's Office, but was moved to the Ministry of Finance, Industry and Planning in 1995. In 2000, the Economic Development Unit was transferred to the new Ministry of Agriculture, Planning and the Environment. Six months later, the unit was transferred to the renamed Ministry of Finance and Planning and restyled as the Planning Unit incorporating the current PSIP Unit and the Macro Economic Unit. A Permanent Secretary headed the Economic PlanningDivision, but inlate springo f 2000 the position was abolished. 3.11 The National Authorizing Officer (NAO) and a European Development FunWrogramme Monitoring and Coordination Unit (EDFPMCU) coordinate assistance from the European Union. Regrettably, it is not clear that the creation of the new EDFPMCU unit has resulted in a much faster disbursement of the financial resourcesset aside for Dominica by the EU. The N A O and the EDFPMCUUnit provide a central focus that should ensure a greater degree of coordination for EU assistance and ffexibility in planning public investments. While the N A O had been around since the 1975 Lom6 C o n ~ e n t i o n ,the~EDFRMCU was established as a dedicated support unit only in ~ October 2000. However, the slow pace of EU grant disbursements evidences that red tape curtails the process. It was only recently that the last tranche of the Stable Export Earnings Program (STABEX) 1995, which was signed in 1997, was disbursed. The N A O that heads the EDFPMCU reports to the Director General, Finance and Planning. A senior project analyst, two project analysts, an accountant, an information officer, a secretary, and an office attendanddriver staff the EDFPMCU. The funding instruments coordinated by the N A O include STABEX, the National Indicative Programmes, the Caribbean Regional Indicative Programme, the Structural Adjustment Facility, the Special Framework of Assistance to ACP Banana Producers, and the New Framework of Assistance resulting from the World Trade Organization banana decision. 3.12 It is recommended that Dominica establish a central clearinghouseto coordinate donor assistance. Currently, line ministries deal directly with donors and may bring in the M o F only at a later stage. Some donors, such as Japan, deal with the Ministry of Foreign Affairs. Others, such as the EU, deal with the NAO. A one-stop shop may help coordinate the assistance from all donors within the country's economic strategy and avoid future unexpected financial implications for the budget. Preparation 3.13 Until recently the main focus of the PSIP unit was the preparationof the annual capital budget. DuringJanuary and February, before the fiscal year begins, the MoF sends the call letter to line ministries asking them to submit their capital spending estimates. Recurrent expenditure implications of the projects are rarely submitted. After the capital submissions are received, discussions between the MoF and the line ministries take place. These discussions focus on their consistency with national and sectoral priorities. They also seek to confirm the availability of funding as evidenced by a financing agreement or letter of 29The LomB Convention is an agreement through which the EUprovides financial and technical assistance and some trade advantagesto countries in Africa, the Caribbean, and the Pacific. It was established in LomC, Togo in 1975 and has periodically been renegotiated. 62 intent. The Committee of Permanent Secretaries i s supposed to play a decisionmaking role at the technical level, but it has not been functioning as such in recent years. The capital estimates are finalized and submitted to the Cabinet for approval b y early June. The Cabinet i s responsible for approving all public investment projects and the Estimates Committee of the Cabinet is responsible for approving the capital budget estimates and the medium-term PSIP, and for decisions about priorities with regard to public investment projects. However, the Cabinet considers public investment projects on an ad hoc basis whenever a particular minister brings the project forward. 3.14 However, since FY2002/03 the PSIP unit has begun to prepare a three-year rolling PSIP. InAugust-September before the fiscal year begins, the Planning Unit reviews the various ministry PSIP proposals for the following fiscal year. This review i s based on the project information provided by the ministries in light o f their sectoral plans. Unfortunately, the connection between this exercise and the medium-term macroeconomic framework i s somewhat tenuous. Specifically, ministries do not always provide the requested information, and sectoral plans are often not prepared. Consequently, the net result of this exercise must be regarded as something less than a full-fledged medium-term PSIP since it i s far from comprehensive and does not succeed in really systematically prioritizing the projects for the PSIP. 3.15 For the most part, the weakness of the medium-term PSIP can be explained by the lack of a well-thought-out medium-term country strategy that includes sectoral plans to guide the PSIP process, and weak administrative procedures for the development of the PSIP. The PSIP unit has received technical assistance from the World Bank and during 2003 from the Government of Barbados, with the appointment of a long- term consultant exclusively dedicated to help the MoF in the formulation of a medium-term PSIP and the monitoring of the capital budget. With this assistance, the authorities formulated the FY2003/04 capital budget and the three-year rolling PSIP, which tried to maximize the use of available grant and concessional financing. The PSIP now reflects realistic estimates of project implementation, and a prioritization of projects in line with the government's objectives of supporting private sector investment and growth. Another step in the right direction has been made with the circulation of the draft administrative procedures for the development of the medium-term PSIP prepared with technical assistance from the Barbados Government. This draft addresses the shortcomings discussed above. Inaddition, a Projects Supervisory Committee, comprising public officials from various key ministries, the Budget Director and the Debt Unit, has been appointed and mandated to oversee project selection, financing, approval, and implementation. Quarterly budget execution reports are also being prepared. However, the country still lacks a medium-term country strategy and sectoral plans to anchor the medium-term investment strategy. The PRSP now in preparation hopefully will fill this void. 3.16 The challenge for the government is to follow up by implementing the procedures it has drafted, and to keep its tight monitoring of the capital budget in the next fiscal years in the face of elections and the subsequent possible pressures for capital spending. More specifically: 63 The capital budget and the three-year PSIP prepared b y the PSIP Unit in the Ministry of Finance need to be made an integral part of the PRSP now being prepared by the Dominican Government. Moreover, the Budget unit should set the overall envelope for capital expenditures (consistent with macroeconomic stability) early in the year so that the PSIP unit can attempt to prioritize the projects from the start of the process instead of at the end. 0 The government of Dominica needs to implement a selection criteria for investment projects. First, only investment projects, independent of their source of financing, with all the required information (in particular projected recurrent costs) should be included in the capital expenditure budget and the PSIP. Second, only investment projects prioritized based on some type of cost-benefit analysis and in accordance with the overall resources available for capital expenditures should be included in the capital expenditure budget and the PSIP. Third, all projects with a cost above EC$2 million should be subject to a formal cost-benefit analysis before being included in the capital expenditure budget or the PSIP. Fourth, given the high debt burden, only projects financed with concessional financing should be included. 0 At a technical level, staff needs to be trained in the procedures and tools of project analysis for preparing the PSIP. 3.17 The government is to be commended for the recent reclassification of several projects that in the past were in the capital budget as recurrent expenditures. Nonetheless, moving forward, the government needs to ensure that the distinction between recurrent and capital expenditures obeys international guidelines. OECS countries by and large include in their PSIP all projects for which they will seek external financing (loans and/or grants). Thus, when the capital budget i s prepared, all the projects in the fiscal year for which the budget i s being prepared are classified as capital expenditures. Nevertheless, this does not obey international and IMF guidelines. For example, training programs, maintenance of buildings, social protection programs, Poverty Assessment, Population Census and other expenditures incurred on a yearly basis should be recorded as recurrent expenditures independent of their source of financing. The government may keep all these programs in the PSIP with the purpose of seeking external financing from donors, but the recording in the budget should obey international guidelines. In FY2003/04, the government reclassified the following programs: Training, Scholarship and Higher Education; Scholarship Programme (EC$186,000 in FY2001/02); School Feeding Program (EC$242,000 in FY2001/02); School Text Books (EC$400,000 in FY2001/02); and Youth Environment Service Corps and Youth Skills Training (EC$lOO,OOO in FY2002/03) as recurrent expenditures. 3.18 The government's capital budget until recently included investment projects that did not satisfy the aim of growth and poverty reduction stated by the Government. Between FY1999/2000 and FY2000/01 about EC$8.9 million were spent on the Windsor Park project to build a 10,000-seat stadium suitable for international athletic and soccer competitions (only the foundation has been completed). DuringFY2000/01-FY2001/02, the government spent EC$33.2 million on the construction o f a financial complex, which has 64 been completed, but which still has significant vacancies due to the retrenchment in the offshore sector. DuringFY1997/98-FY 1999/2000, a total of EC$36 million was spent on a feasibility study, detailed design, and land acquisition for the construction of a new international airport that was later deemed infeasible. (Box 3.1 presents some of these examples in more detail.) Since FY2002/03, the government has halted several projects whose impact on the economy was questionable, and also has made an effort to prioritize projects taking into account their impact on growth and poverty. However, the government still has a difficult road aheadto systematize the selection process of investment projects. 3.19 The Dominican budget scores well in comprehensiveness, but unlike in other OECS countries the Financial Secretary can transfer funds among programs in agencies other than the MoF. Dominica's budget i s comprehensive, generally covering all CG expenditures. However, the FAA empowers the Financial Secretary to vire, or transfer, funds between programs within the same supply vote (agency). This means that funds can be transferred to a completely new recurrent program within the same agency, or from one capital project to a completely new one without the obligation to obtain subsequent endorsement from the Parliament via a supplemental appropriation. For example, in FY2002/03 the government used funds budgeted for repairs in primary schools to pay capital expenditures for the Windsor Park stadium. In theory, any such virements, or transfers, and, ultimately, expenditures should at least be reported in the next year's Estimates Books. However, the Estimates Books do not always contain complete and accurate figures for revisions and prior year expenditures, particularly with respect to capital projects. This could mean that the government can legally spend funds on programs or projects without the Parliament ever becoming aware that such a program or project existed. The FAA should consider empowering only the Minister of the respective agency to transfer funds between programs up to a ceiling, and require that if the transfer exceeds the ceiling, the transfer will need the approval of the Parliament and the MoF. This i s to safeguard program allocations at the sectoral level from arbitrary changes by the M o F at the central level. 3.20 Investment projects by statutory agencies, such as state-owned enterprises, do not follow the procedures of the PSIP and the capital budget. This i s a serious gap because these investment projects can represent a sizeable fraction of the PSIP (on average about 20 percent of total public sector capital expenditures), and can expose the government to serious fiscal liabilities. The government should ensure an accurate representation of the public sector investment program, by including in the PSIP the investment projects of all statutory agencies. Investment projects by private corporations but guaranteed by the government should be disclosed in the government's financial statement as government liabilities. 65 Box 3.1. The Windsor Park Project and the New International Airport The Windsor Park Project A 10,000-seat stadium was to be built in Windsor Park to provide a facility suitable for intemational athletic and soccer competitions. This was a location where local soccer and cricket matches were already being held. The estimatedcost of the project was EC$20.9 million, or almost 3 percent of GDP. The Govemment electedin 1995 proposed this project as part of its election platform. Consequently. it was never subjected to the analytical requirements of the PSIP process. Inparticular,no cost-benefit analysis was camedout. Ifit hadbeen, it would no doubt have shown that the likely revenues generatedby the project would have fallen far short of the project's originally estimatedcost. The Government electedin 2000began the projectafter the last election,even though no foreign donor was found to finance the project. A contract with a Venezuelan construction firm was signed in October 1999. The groundbreakingceremony for the project was held in January 2000. Spending on the project was financed by extemal borrowing on commercial terms, and continued even as the financial crisis deepened. By the end of FY2001/02,EC$8.9 million had been spent on the project and only the foundationwas completed. Work on the project was haltedin July 2001 because of a shortage of funds, but the Govemment was still makingpayments to the contractor. In FY2002/03, EC$240,000 was paid to the contractor as the result of an initial agreement to pay US$840,000 over 16 months. This reflects only part of the US$2.7 million total interim payment sought by the contractoras compensation for delays. Moreover, a final settlement could be much higher if the project is terminated. The initial paymentwas taken out of the allocationfor the DominicaGrammar School RoofRepairproject. The Govemment has recently obtained grant financing from China to build the Stadium whose construction i s plannedto start in FY2005/06. The New International Airport The prospect of a new intemational airport was first discussed in 1994. The United States Army Corps of Engineers had offered to build a runway at Crompton Point on the east coast, north of Melville Hall and further away from the ocean. Before the project could be implemented, however, an election was called and the airport becamea political issue due to concems about havingAmerican soldiers in the country during an election. After the electionin 1995,the newly electedGovemment proposedto build a new intemationalairport in Wesley- Marigot, which was also further inlandfrom the Melville Hall Airport, and was also flatter. There were a number of problems with that area. First, there was a secondary school already built on the land that would have to be tom down. Second, the land was a major agricultural area, and it would have been necessary to compensate the farmers for their lost income. Even though the Wesley-Marigot site was considered less desirable as an airport site than CromptonPoint, EC$26million was eventually allocated for landpurchases. The project did not proceed far before an election was called in 2000, during which the airport became an issue. The Govemment was defeated in the election and was replaced by a new Govemment that was committed to reviewingthe airport project. After a carefulreview, the Minister of Finance recommendedthat the Govemment not proceed with the project. By that time, EC$36million already hadbeen spent on it. B. EXECUTIOND MONITORING A N Execution 3.21 FromFY1995/96 to FY2001/02, capitalexpendituresof the CG were on average about 9 percent of GDP. Capital expenditures of the rest of the public sector were about 2 percent of GDP. As Table 3.1 shows, capital expenditures by the CG fluctuated from a low of 4.8 percent of GDP in FY1997/98 to a peak of 15.3 percent of GDP in FY2000/01 when work on the Windsor Park project started. Capital expenditures for the rest 66 of the public sector including the DSS and non-financial public enterprises accounted, on average, for about 20 percent of all public sector capital expenditures and fluctuated from a low of 0.7 percent of GDP in FY1998/99 to a peak of 6.1 percent of GDP in FY2000/01. This level of capital expenditures for the public sector is quite highcompared to other OECS countries and i s similar to that of St. Kitts and Nevis, a country that was hit by three hurricanes between 1995 and 2001, with resulting high investment reconstruction needs, and almost triple the level for the Latin America and the CaribbeanRegion. Table 3.1. Sectoral Composition of Actual Capital Expenditures for the Dominica Public Sector (as percentage of GDP)* Prel." Average FY 1995/96 1996/97 1997/98 1998/99 1999/00 2000/01 2001/02 95/9600/01 CentralGovernment 11.7 6.1 4.8 8.3 10.6 153 5.7 8.9 EconomicInfrastructure 3.1 2.7 1.5 1.7 3.4 3.9 1.1 2.5 Airport development 0.1 0.0 0.0 0.2 0.0 0.0 0.0 0.0 Roads andbridges 0.7 1.8 1.5 1.1 1.2 1.o 1.o 1.2 Sea defenses 0.0 0.0 0.0 0.2 2.2 2.9 0.0 0.8 Water and sewerage 2.0 0.1 0.0 0.1 0.0 0.0 0.0 0.3 Other infrastructure 0.4 0.7 0.0 0.1 0.0 0.0 0.0 0.2 ProductiveSectors 6.6 0.6 1.1 2.4 2.2 0.6 1.o 2.1 Agriculture 4.4 0.4 0.2 0.6 0.1 0.2 0.3 1.o Forestry 0.2 0.1 0.1 0.0 0.0 0.0 0.0 0.1 Microenterprisedev. 0.0 0.0 0.4 1.1 1.o 0.2 0.6 0.5 Tourismdevelopment 0.2 0.0 0.4 0.5 0.1 0.2 0.6 0.3 Other productivesectors 1.8 0.0 0.0 0.2 0.4 0.1 0.0 0.4 PublicAdministration 0.3 0.3 0.5 0.7 1.1 4.5 2.3 1.4 Adm. infrastructure 0.0 0.0 0.0 0.1 0.4 0.6 0.4 0.2 Env. management 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Informationsystemsdev. 0.0 0.1 0.0 0.2 0.1 0.3 0.1 0.1 Offshore investment 0.0 0.0 0.0 0.0 0.0 3.1 1.5 0.6 Public safety 0.1 0.0 0.1 0.1 0.6 0.6 0.3 0.3 Public sector modernization 0.2 0.2 0.3 0.2 0.0 0.0 0.0 0.1 Social Infrastructure 1.6 2.6 1.8 3.5 3.9 6.2 1.3 3.0 Communitydevelopment 0.2 0.4 0.6 0.5 0.8 0.7 0.2 0.5 Education 0.4 1.2 0.8 2.1 1.2 2.0 0.5 1.2 Health 0.5 0.6 0.1 0.4 0.1 0.3 0.0 0.4 Housingand settlement 0.5 0.4 0.3 0.3 0.9 0.3 0.2 0.4 Sportsand recreation 0.0 0.0 0.0 0.2 0.3 2.9 0.4 0.5 Rest of the PublicSector 2.3 1.2 1.2 0.7 1.3 6.1 2.8 2.2 Total PublicSector 14.0 7.3 6.0 9.0 11.9 21.4 8.5 11.2 * Sectoral figures for FY2002/03 are not yet available. **The FY2001/02total figure spent by Dominican authorities for productive sectors does not add up to the sum of its components. Source: Dominican authorities. 3.22 During this period, on average, about one-third of CG capital expenditures were insocialinfrastructure and about 30 percent were ineconomicinfrastructure. As Table 3.2 shows, the largest share within the subcategory of social infrastructure was education (14.1 percent of C G capital expenditures), and then community development (5.9 percent of CG capital expenditures). In addition, the CG allocated almost the same share of capital expenditures to sports and recreation as health (4.4 percent and 4.2 percent, respectively, of CG expenditures). C G capital expenditures in economic infrastructure were mainly allocated to roads and bridges (16.6 percent of CG capital expenditures) and sea defenses (6.1 percent 67 of CG capital expenditures). InDominica, the share of economic infrastructure in CG capital expenditures i s much smaller than inother OECS countries, which ran from around a third in St. Lucia to over a half in St. Kitts and Nevis. 3.23 Capital expenditures in productive services accounted for about 22.8 percent of total CG capital expenditures. Of this, agriculture accounted for almost 10 percent of C G capital expenditures. Public administration accounted for 15.1 percent of total C G capital expenditures, and within this, offshore investment 6.6 percent. The latter i s due to the construction o f the Financial Complex in FY2000/01 and FY2001/02. Table 3.2. Sectoral Composition of Actual Capital Expendituresfor Dominica Central Government* (aspercentage of total CG capital expenditures) Prel.** Average FY 1995196 1996197 1997198 1998199 1999100 2000101 2001102 95196-00101 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Economic Infrastructure 27.0 43.9 30.7 20.5 32.0 25.8 18.9 28.4 Airport development 0.6 0.86 0.0 1.9 0.0 0.0 0.0 0.5 Roadand bridges 5.9 30.1 30.4 13.7 11.1 6.9 18.2 16.6 Sea defenses 0.0 0.0 0.0 2.2 20.9 18.9 0.7 6.1 Water and sewerage 16.8 1.8 0.3 1.7 0.0 0.0 0.0 2.9 Other infrastructure 3.6 11.2 0.0 1.o 0.0 0.0 0.0 2.3 Productive Sectors 56.7 10.2 22.6 29.3 20.4 4.1 16.6 22.8 Agriculture 38.0 6.1 3.1 7.4 6.7 1.1 5.6 9.7 Forestry 1.4 2.3 2.8 0.3 0.1 0.0 0.0 1.o Microenterprisedev. 0.0 0.3 8.2 13.2 9.3 1.2 11.0 6.1 Tourismdevelopment 1.6 0.8 7.5 6.2 0.9 1.2 11.0 2.6 Other productivesectors 15.7 0.8 0.9 2.2 3.4 0.7 0.0 3.4 Public Administration 2.6 4.3 9.4 8.4 10.3 29.6 40.9 15.1 Adm. infrastructure 0.1 0.3 0.6 1.7 3.4 3.7 7.0 2.4 Env. management 0.0 0.0 0.0 0.0 0.0 0.0 0.7 0.1 Informationsystems dev. 0.3 1.3 0.6 2.1 0.9 1.9 1.2 1.2 Offshore investment 0.0 0.0 0.0 0.0 0.0 19.9 25.9 6.6 Public safety 0.6 0.3 2.2 1.7 6.0 4.1 5.8 3.0 Public sector modernization 1.6 2.6 6.0 2.9 0.0 0.0 0.0 1.9 Social Infrastructure 13.8 41.6 37.3 41.8 37.1 40.5 23.4 33.6 Community development 1.7 6.1 11.6 6.2 7.8 4.6 3.3 5.9 Education 3.2 19.6 16.6 25.0 11.5 12.9 9.6 14.1 Health 4.6 9.4 2.5 4.5 6.4 2.1 0.2 4.2 Housing and settlement 4.2 6.4 6.6 3.3 8.8 2.0 4.2 5.1 Sports and recreation 0.0 0.0 0.0 2.9 2.5 18.9 6.5 4.4 * Sectoral figures for FY2002/03are not yet available. **The FY2001/02 total figure spent by Dominican authorities for productive sectors does not add up to the sum of its components. Source: Dominican authorities. 3.24 In FY2002/03, revised estimates of capital expenditures were about 5.1 percent of GDP, below the 6 percent of GDP agreed in the SBA of August 2002. The Dominican Government was unable to use the capital budget as a key instrument to support growth in FY2002/03 due mainly to a lack of capacity in the MoF to implement and monitor the capital budget. 68 3.25 The budgeted capital expenditure for FY2003/04 was about 7 percent of GDP. The investment projects in the FY2003/04 budget are by and large growth-enhancing and poverty-reduction projects such as infrastructure, fisheries, banana restructuring, tourism development, and safety nets. These projects are financed b y a combination of grants (about two-thirds of the total), loans (about one-quarter of the total), anddomestic revenues. 3.26 It is difficult to assess the adequate level of public capital expenditure for the Dominican economy because capital expenditure figures reported by the Government, as in most OECS countries, not only include spending on fixed capital, but also some recurrent expenditures. If one subtracts from the historical average CG capital expenditures the unproductive capital expenditures (about 4 percent o f GDP) and uses the evidence presented in this report for investment in education and health, the remaining level of capital expenditures intended to acquire fixed capital of 2 to 3 percent of GDP i s in line with the historical averages in the Bahamas (2 percent of GDP) and Barbados (2.4 percent of GDP) for 1990-2001. This excludes capital expenditures that may be needed for exceptional reconstruction needs inresponse to natural disasters. 3.27 During FY1999/2000 to FY2001/02, CG capital expenditures were mostly financed through loans instead of grants and/or local revenues. As Table 3.3 shows, during this period, more than 50 percent of CG capital expenditures have been financed through loans (mostly commercial) compared to only less than 30 percent in most previous fiscal years. The increase in the use of loan financing i s a common pattern in the OECS countries and i s the result of the growing access to the Trinidad and Tobago capital market. Grant financing has declined from almost 60 percent in FY1995/96 to about 30 percent in FY2001/02, and was expected to increase inFY2002/03 and FY2003/4. Table 3.3. Sourcesof Financing of Dominica Central Government Capital Expenditures, FY 1995/96--FY2001/02* (aspercentage of total CG capital expenditures) Average Fiscal Year Prel. 1995196- 1995196 1996197 1997198 1998199 199912000 2000101 2001102 2001102 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Revenue 18.8 40.8 46.7 42.3 24.2 5.3 23.1 28.8 Grants 59.0 23.5 24.5 28.9 25.8 32.2 31.8 32.2 Loans 22.2 35.7 28.8 28.8 50.1 62.5 45.1 39.0 * Financingsourcesfigures forFY2002/03 are not yet available. Source: Dkinican aithorities. 3.28 The estimated figures for capital expenditures presented inthe budget have to be used with caution due to the low realization rate of planned capital expenditures and the high year-to-year variation inthe rate of realization. As Table 3.4 shows, on average during FY1994/95-FY2001/02, only half of planned capital expenditures were actually implemented. However, this statistic hides substantial year-to-year variation in the rate of realization, which ranged between 24.5 percent in FY1997/98 and 86.5 percent in FY2000/01. The realization rate, on average, was in line with that o f St. Lucia, was substantially higher than the 25 percent rate in St. Kitts, but was lower than the 64 percent rate in Grenada. Low implementation rates are attributed to resource constraints related to 69 the available human and technical capacity and other inadequacies at various stages of the project administration cycle. Because of the importance of EUfunding, low implementation rates often have been the result of slow EU disbursements. Preliminary figures for FY2002/03 presented a realization rate of 78 percent. Moreover, as a result of recent improved monitoring, the execution rate of the capital budget increased significantly in the first quarter of FY2003/04 (to 93 percent). Table 3.4. Realization Rate of Dominica Central Government Capital Expenditures, FY 1994l95-FY2001l02 Average 1994195- 1994195 1995196 1996197 1997198 1998199 1999100 2000101 2001102 2002/03* 2001102 (In mill EC$) Approved Estimates 128.1 104.0 165.8 183.6 245.1 321.9 121.3 131.4 44.6 Actual 77.2 66.3 64.8 45.0 82.0 93.3 104.9 68.0 34.7 (aspercentage) Realization Rate * Preliminary. 60.3 63.7 39.1 24.5 33.5 29.0 86.5 51.8 77.8 48.5 Source: Commonwealthof DominicaEstimatesfor the Year. Monitoring 3.29 The capacity of the PSIP unit to monitor the implementation of capital projects, which are carried out by line ministries, is weak. Line ministries are supposed to complete quarterly reports on project implementation to be sent to the PSIP unit, but compliance has been rather lax, which makes it rather difficult to monitor implementation by the PSIP unit. The Accountant General's Department o f the Ministry of Finance i s responsible for making the approved disbursements. Not all projects are included at budget time. Some can be introduced at midyear through supplementary estimates. This gives the government some flexibility to respond to urgent needs. But it also gives ministries an opportunity to short-circuit the whole PSIP process. 3.30 Until recently the PSIP unit did not track monthly expenditures in investment projects. Indeed, record of actual expenditures in some capital projects lagged behind in some cases by more than one year because line ministries did not submit the required information, and the PSIP unit did not enforce this requirement. However, new monitoring procedures were implemented this fiscal year and line ministries have being presenting monthly reports with expenditures. 3.31 Postevaluation of projects is usually done only for externally funded projects where it is required by the financing institutions. A good example i s the Basic Education Reform Project, which was financed by the World Bank and implemented between 1996 and 2001. An indicator of its success i s that the transition rate from primary to secondary school increased from 54 percent to 76 percent over the period. This was 12 percentage points better than the project target figure of 64 percent. The World Bank Project Completion Report for this project gave it a satisfactory to highly satisfactory rating. 70 4. PUBLICSECTOREMPLOYMENT AND COMPENSATION 4.1 With a total population of about 73,000 inhabitants, Dominica has limited resources to carry out the whole range of functions performed by any government. As several studies have shown, small countries tend to have bigger governments in response to the higher cost o f supplying public goods and the need to provide a stabilization role to ameliorate the effect of external shocks (Alesina and Wacziarg 1997; Rodrik 1996). However, even in a small country a government can be "too large" in several ways (Nunberg and Lindauer 1994): 0 The public service may be "too large" in the broad sense that the government sector i s overextended, possessing too many agencies and departments charged with too broad a range of responsibilities. The public service may be "too large" in the narrower sense o f employing workers in excess of the requirements of designated tasks. 0 The public service may be "too large" because government workers do not fulfill the tasks assigned to them or do so only partially, with great delays and at highcost. 0 The public service may be "too large" in the sense that the wage bill i s too expensive for sound fiscal management. 4.2 The first notion, that Dominica's government sector i s "too large," possessing too many agencies and departments, does not seem to be a generic problem. Duplication of functions across ministries does not seem widespread nor, as in some settings, i s there an obvious presence of superfluous government units. Certainly specific activities could be eliminated. 4.3 The second notion i s that the government i s "too large" in the narrower sense of employing workers in excess of the requirements of designated tasks. Concern over this question prompted authorities in the Establishment Department to conduct audits of the central administrative offices of various ministries and departments. This program, under the title Rationalization of Human Resources, did find some excess capacity. But as often as workers were identified as underloaded, there were others who were overloaded, often in the same department. The reports contain many useful suggestions for improving management and better allocation of staff within individual departments, but these detailed accounts do not convey any sense of systemic excess ingovernment employment. 4.4 The third notion i s that a government might be considered "too large" because the workers are too unproductive, do not fulfill the tasks assigned to them, or do so only partially. This i s easy to dismiss. Dominica scores high marks on education outcomes. Its health sector performs well. This does not imply that all civil servants perform admirably or even adequately. This i s highlighted by some of the recent personnel audits. Similarly, not all branches of government are serving the national interest well. The ports, organized as a state enterprise, frequently are cited as too expensive and as an impediment to private 71 investment and economic growth. Customs and investment promotion both need to become user-oriented. 4.5 The fourth notion, that the public service may be too large and expensive for sound fiscal management, i s the key challenge that Dominica has faced in recent fiscal years, and will have to continue to face in the future. In this chapter, we mainly focus on why the government's wage bill in Dominica constitutes too high a percentage of GDP, about 16 to 18 percent during FY2000/01-FY2002/03, for cost-effectiveness and sound fiscal management. The main findings and recommendations can be summarized as follows: Although real wages were flat during FY1990/91-FY2002/03, 323 established posts were added in the presence of a primary deficit (after grants) o f 6.0 percent of GDP in FY2000/01 and 7.5 percent of GDP in the previous fiscal year. Thus, the government will have to cut expenditures mainly by a downsizing of the civil service to at least the levels of the mid-1990s. 0 However, to implement a well-thought-out downsizing program, the government will have to obtain clarification on the legal requirements for redundancy benefits and introduce and/or upgrade its information systems in order to accurately cost its implications (pension and severance payments) and monitor public employment (established and non-established positions) on a regular basis. 0 Moreover, the adoption of a medium-term downsizing program will need to be accompanied by a human resource management strategy and development of new strategies to carry out essential tasks by leveraging scarce resources. Otherwise, rehiring staff in the medium term would overturn the fiscal gains obtained by reducing the size of the civil service. For example, a performance appraisal system should be established with pay or bonuses to reward good performance. In addition, the fragmentation and duplication of work among ministries, departments, and agencies, leading to a diffusion of responsibilities, should be addressed, as should the possibility of the government divesting some of its functions to the private sector. The use of allowances needs to be rationalized and reduced; although savings from such measures would be limited, a more transparent system would minimize the use of allowances as salary supplements, reserving them for compensation for those expenditures required to perform specific tasks. A. TRENDS EMPLOYMENT IN DURINGFY1991/9%FY2001/02 4.6 There are two types of government positions in Dominica: established and non- established. The distinction between established and non-established positions has a legal basis. Established positions refer to those positions specified in the Public Service Act of 1991 and its subsequent amendments. The majority of government employees fall under this act. Non-established positions are those positions not specified inthe legislation, and include postal agents, senators, substitute teachers, unskilled laborers, and watchmen. Established positions are permanent positions listed in the budget, and their compensation i s classified as Personal Emoluments. Employees in non-established positions are hired directly b y line 72 ministries, their numbers are not always presented accurately in the budget, and their compensation i s classified as wages. 4.7 Trends in governmentemploymentare more difficult to establishthan trends in compensation. This is because there is no systematic recording of total government employment. Good information is available on trends in the number of established positions, but the number of non-established positions, especially casual labor that occupies positions with hourly wage rates, i s harder to determine. Further complicating assessment of employment trends i s a lack of data on vacancies and on hours worked. What i s well documented i s the number of established positions. These are reported on a fiscal year basis inthe budget reports.30 4.8 The numberof establishedpostsincreasedduringFY1990/91-FY2000/01 by 402 posts, or 15 percent, which averaged 0.95 percent per fiscal year. This is the same percentage increase of real GDP over the same period, suggesting that established posts grew at the same rate as the economy. A closer examination of Table 4.1 reveals that between FY1990/91 and FY1997/98 the number of established positions grew modestly, by only 79 posts, or by a total of about 3 percent. Most of the increase in established posts occurred after FY1997/98: 323 posts were added during FY1997/98-FY2001/02. This is a large increase, especially in light of the fiscal problems that occurred toward the end of the period due to an unsustainable increase in unproductive capital expenditures. Table 4.1. Established Posts inDominica Central Government, FY1990/91-FY2001/02 19901 19911 19921 19931 19941 19951 19961 19971 19981 19991 20001 20011 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Established Positions 2,615 2,612 2,655 2,713 2,612 2,694 2,694 2,694 2,749 2,755 3,007 3,017 Annual Growth Rate -0.11 1.65 2.18 -3.72 3.14 0.00 0.00 2.04 0.22 9.15 0.33 NA = Not available due to mssing document. Source Allocation of Staff and Pay Scales inthe Public Service, Estimates 2001/02, Commonwealth of Domimca 4.9 Most of the increase in established posts occurred in FY1999/2000-FY2000/01, and it was concentratedin the ministries of Education, Sports and Youth Affairs, and Healthand SocialSecurity. Table 4.2 decomposes the most recent increase in established posts and determines that 255 of the 323 new positions (80 percent) were either in the Ministry of Education, Sports and Youth Affairs, or the Ministry of Health and Social Security. The remaining 20 percent are distributed across a number of ministries and reflect both the movement of tasks across ministries and the taking on of new functions. In health, 116 of the 150 new posts were designated for community health aides, nurses, and orderlies. In education, 60 percent of the new posts were for instructors, librarians, teachers, and principals; and 40 percent roughly were divided between junior and senior administrative 30 In August 2002,487 established posts out of a total of about 3,000 were reported as vacant. However, many (if not most) of these vacancies are filled by personnel "acting" in the position, that is, not yet formally approved to hold the post in a permanent capacity. According to government officials, the widespread practice of staff "acting" in established posts suggests that the number of truly vacant positions i s low, and that the numberof established posts can be used as a good indicator of actual employment. 73 positions. Some of these administrative positions were in apparently new functions, including divisions for Learning Support and for Planning and Development in the Ministry of Education. Table 4.2. Growth inCentral Government Established Postsby Ministry or Department, FY1997/98-FY2001/02 FY FY Ministry or Department 1997/98 2001/02 Change Ministry or Department President's Office 4 5 +1 President's Office Legal Affairs 58 75 +17 LegalAffairs Houseof Assembly 3 4 +1 Legislative Audit Department 21 21 _ _ Audit Department Election's Office 4 4 -- Election's Office Cabinet Secretariat & Security 636 639 +3 PrimeMinister's Office Department ExternalAffairs 14 15 +1 ForeignAffairs PersonnelOffice 33 49 +16 Establishment,Personnel& Training Tourism, Ports & Employment 28 10 -18 Tourism Finance, Industry & Planning 241 229 -12 Finance, Industry & Planning Agriculture 139 132 -7 Agriculture Trade & Marketing 12 14 +2 Trade, Industry& Marketing Education 850 955 +lo5 Education Communications& Works 109 149 +40 Communications & Works Community Affairs & Women's 62 88 +26 CommunityDevelopment& Affairs Gender Affairs Health 471 621 +150 Health Labor& Immigration 9 7 -2 Labor & Immigration Total 2694 3017 +323 Source: Allocation of Staff and Pay Scales inthe Public Service, Estimates2001/02, Commonwealthof Dominica; and Allocation of EstablishedPosts inthe Civil Service, Estimates 1997/98,Commonwealthof Dominica. 4.10 It is harderto documenttrends inthe number of governmentemployeeswho are not filling established positions. For FY2000/01, officials of the Ministry of Finance report total CentralGovernment (CG) employment of 5,352 individuals. This includes 3,007 establishedposts. Budget documents account for approximately another 730 workers innon-established positions whose pay is basedon a monthly wage rate. The majorjob titles of these non-established positions include Registering Officers in the Election's Office (129), Domestic Staff in hospitals and other health centers (120), Special Constables (72), Postal Agents (62), and Substitute Teachers (48). Beyond the 3,007 established posts and the 730 non-established positions remain another approximately 1,600 employees, enumerated by the Ministry of Finance, but whose jobs and hours are not detailed in budget documents. The Ministries of Agriculture and of Communications and Works engage most of these workers. 74 4.11 Because there Figure4.1.a. GovernmentEmploymentas Percentageof Population,2001, appears to be no time series SelectedCaribbeanCountries 16- - _ _ on these different categories of government employment, it is hard to determine how their ranks have grown over time. Data from two labor force surveys, one conducted in 1989 and the other in 1997, however, suggest limited growth in the number of government workers in non-established positions. Source. Mukheqee. de Tommaso, and Schiavo-Campo(1997) The labor force surveys identify Figure4.1.b. PublicSector Wage Billas Percentage of GDP, 2001, the number of Selected CaribbeanCountries workers who report the public sector as the source of Id1 { +Dominica their employment. In 1989 .I there were 5,300 such S t Vincent and the r S t Kimand Nevir Grenadines + workers and in 1997, there 0 L 12 !? +Grenada `St LUcia $Jamaica +Barbados were 5,460. These totals include employment in general government and in public enterprises, where employment refers to at least 2/ "one hour of activity in the 0 1 reference week." Since Source: Author's calculations. 1989, the C G has accounted for 95 percent of the wage bill of the consolidated sector. Applying this same ratio to the employment data from the labor force surveys suggests that C G employment was approximately 5,035 in 1989 and 5,190 in 1997. If added to this total are the roughly 300 established posts created since FY1997/98, one would predict total C G employment in FY2000/01 of just less than 5,500-an amount higher than but still close to the 5,352 reported by the Ministry of Finance. What this rough calculation suggests i s that the number of government workers in non-established positions probably has not expanded appreciably over time. 4.12 Inconclusion, inFY2000/01, approximately 7 in 100individuals worked for the Dominican Government inan established or non-established post. Although this figure i s similar to the one observed in other Caribbean countries, the wage bill as a percentage of GDP is one of the highest inthe Caribbean region. As Figure 4.1.a shows, Dominica's ratio of government employment to population (3.8 percent) is similar to the one observed in other Caribbean countries, with the exception of St. Kitts and Nevis. However, as Figure 4.1.b shows, the wage bill as a percentage of GDP (16 percent) cannot be afforded 75 by the Dominican government in the short and medium term, when its debt-to-GDP ratio i s 111percent at the end of FY2002/03. 4.13 In FY2002/03 and FY2003/04, as part of the SBA, the government agreed to limit the wage bill mainly through wage cuts. But as discussed earlier, the fiscal adjustment required to achieve fiscal sustainability will require a significant downsizing of the public sector. In FY2002/03 the government limited the wage bill, with mixed results, by eliminating overtime payments, reducing the number of temporary workers, imposing a freeze on hiring (except for teachers and health workers), and granting no across- the board wage increases. However, limited progress was made in reducing the wage bill due to weak control over the use of workers in non-established positions. InFY2003/04, the government reduced the salaries of workers in established positions at the CG level by 5 percent, and agreed to reduce work time of casual workers, close some temporary worker positions, and extend the freeze on hiringfor all permanent and temporary employees (except when absolutely unavoidable becauseof the importance of the position). B. TRENDS COMPENSATIONDURINGFY1991/92--FY2001/02 IN 4.14 The salary structure of the Central Government includes 40 salary grades covering all established positions in the CG. A separate pay scale, employing both hourly and monthly pay rates, covers employees in non-established positions. Another small group of government employees includes contract officers, most of whom occupy senior, and usually technical, positions. Contract officers are covered by another set o f compensation schedules and, at present, number 37. 4.15 The salary structure governing workers in established positions, measured in real terms, remained remarkably constant during FY1993/96FY2001/02.3' Figure 4.2 identifies three grades in the salary structure that roughly conform to the top (Grade D), middle (Grade 15), and bottom (Grade 33) of the salary scale. Grade D, for example, includes the Director of Trade and Specialist Medical Officers, Grade 15 includes Qualified Teachers and Senior Clerks, and Grade 33 refers to entry-level Messengers. Between FY1993/94 and FY2001/02, and after adjusting for price inflation, each salary grade increased by a total of only 2.8 percent. Over the same period, real GDP grew by almost 10 percent, indicating that government pay scales have declined relative to the rest of the economy. The uniformity of salary increases across grades also implies no significant change in the compressioddecompressiono f compensation in the overall pay structure. In addition, inFY2002/03 the Government imposed a 4 percent stabilization levy for public and private sector wages as part of the SBA. In FY2003/04, the Government cut civil servants' wages (established positions) by 5 percent, while reducing the stabilization levy to 3 percent, thus further reducing government pay scales relative to the rest of the economy and leading to expected savings inthe wage bill of 0.7 percent of GDP. 31 In 1993 the CG "regraded" the salary scales, complicating comparisons of salaries before and after 1993. This regrading exercise, coupled with a 2 percent across-the-board increase in salaries, caused a large increase inthe wage billduring FY1992/93-FY1993/94, from EC$77millionto EC$89.3 million. 76 Figure 4.2. Dominica's Real Salaries for EstablishedPositionsin the Central Government, FY1993/9&FY2001/02 (in constant EC$I994) 60000 - soooo - - Grade Q - m rl A 1 * * g h 40000 - I Y 2 30000 l o ~ o o ai- +$.- ---* ____-_*---.-- ~ --e Grade 33 Source: Author's calculation$ 4.16 Including allowances does not appear to change the analysis. According to budget documents, there are 18 different allowances awarded to workers in established positions. These allowances cover everything from monthly transportation costs, to entertainment budgets, to compensation in lieu of private practices. Changes in FY2000/01 inthe recording of allowances inthe annual budgetmake it difficult to compare the growth in allowances over time. InFY2001/02, the sum of all allowances totaled EC$9.3 million, and in FY2002/03, EC$9.8 million. It is recommended that these be rationalized and reduced, although savings from such measures would be limited.32 4.17 Only workers in established positions receive allowances, and within this group only a limited number are eligible, depending on the allowance. The current system lacks transparency and could be improved upon. For example, transport allowances, which are not a direct payment for mileage, might eventually be incorporated into base pay. Improvedjob descriptions should be able to minimize the need for responsibility allowances, which are costly and add a layer of unnecessary bureaucratic reporting. A more transparent system would minimize the use of allowances as salary supplements, reserving them for compensation for those expenditures requiredto perform specific tasks. 32The increased use of contract officers has raisedpayments on allowances for travel, to compensate for the fact that doctors are paid less in the government than they would earn in private practice, and for "special duty." However, growth in the Acting Allowance is a result of a decrease in promotions, which has reduced salary commitments, but which increased, to a lesser extent, the amounts paid to those "acting" in posts to which they have not been permanently appointed. 77 4.18 The salary structure, measured %ox4.1. Non-Wage CompensationinDominica's in real terms, governing workers in 'ublic Service: The Vacation and Study Leave non-established positions by either 'rograms hourly or monthly wage rates, remained Zomparisons of public and private pay differentials remarkably constant during ihould include all elements of compensation, including FY1991192-FY 2001l02. Since illowances, pension benefits, and non-wage benefits. FY1991/92 these rates, generally, have lust as with salary data, the available information i s been held constant in nominal terms in a nsufficient for such comparisons. However, in the area three-year cycle. For example, the If non-wage benefits, there is little question that the 5overnment sector offers especially generous leave maximum hourly rate for an unskilled Jenefits, both for Study Leave and Vacation Leave. laborer was EC$3.29 per hour from FY1991/92 to FY1993/94, EC$4.40 per The Study Leave program entitles permanent officers, hour for the next three years, EC$4.87 per ifter three years of service, to a year of Study Leave at hour for three years after that, and illpay. Thisbenefitcanbeappliedformorethanonce. 4pplications are subject to approval, and during 1998 to EC$5.02 per hour since FY2OOO/O1.33 2002, between 39 (2002) and 61 (1999) Study Leaves Because these wage rates often are fixed lave been granted annually. In 2002, the estimated cost for three years, price inflation erodes their in salaries was about EC$750,000. Prior to the current real values until the next adjustment. Fiscal crisis, officers might receive financial assistance to Starting from the beginning of one pay :over tuition payments as well. Although recipients are bonded, if they fail to return to their posts the cycle to the next, in this instance government has not pursued collection of bond amounts. FY1994/95 to FY2000/01, real wages of Given that most of the benefits of education and, hence, such non-established positions as hospital of Study Leaves, accrue to the individual, this program cook, laborer, plumber, and security seems overly generous and should be scaled back. guard, experienced total real pay increases Proposals to trim the Study Leave program include awarding lower amounts (for example, half of salary or of 4 to 6 percent. This is marginally less), and granting fewer leaves with strict identification higher than the experience of workers in of priority areas of training for government officers. established positions, but still below the rate of increase inGDP. The Vacation Leave program is equally generous. As stated in the General Orders, workers in established 4.19 Unfortunately, determining positions earn between 21 and 36 days of vacation leave per year. The amount depends on salary level and years whether government workers in in service, with higher levels receiving more leave Dominica are over- or underpaid Employees are permitted to accumulate large amounts oj compared to the private sector is not an vacation leave, and have done so. According to the easy task. Data on prevailing findings of the Dominica Employers' Federation compensation in the private sector are vacation leave in the private sector i s significantly less generous than in the public sector. B y internationa needed, and in a small island economy standards, the government's leave policy appear! like Dominica, where a sizable fraction exceedingly generous and costly. Excessive amounts 0' of the working-age population migrates, vacation leave resulted in a variety of expenditurc it is that much more difficult to define liabilities, including an increase in payments foi an individual's alternatives. Some Responsibility Allowances, when another officei temporarily replaces one on leave. A revision in the salary data are presented in the Digest of General Orders concerning Vacation Leave should be Labor Earnings in Dominica (4thedition) imdemented. 33 During the three-year period, FY1997/98-FY1999/2000, the hourly wage rate of laborers actually was EC$4.67 in the first year, and EC$4.87 in the latter two years. Similarly, some daily rated workers had a pay raise in the first and second fiscal years of this period, while others only got an increase above their FY1996/97 rates in FY1998199. 78 produced by the Dominica Employers' Federation. Pay differentials based on median entry- level salaries between the CG and other (presumably large) employers range from +/-lo percent, depending on which of a small number of comparable occupations are considered. Given the limited coverage of the labor force represented in the Digest, it i s hard to draw any firm conclusions from these data. (See Box 4.1 on Non-Wage Compensation). In the absence of reliable pay comparisons, turningto other market signals of the appropriateness of government pay can be considered. Persistent vacancies and high turnover often are signals of government pay that i s too low relative to market alternatives. Long queues for government positions or corruption in the allocation o f such posts can be associated with pay levels in excess of the reservation wage of workers. Again, systematic evidence i s lacking, but the impression given by the Establishment Department is that none of these problems is a feature o f the current situation in Dominica. This leads to the conclusion that pay-setting in government has been adequate and appropriate from a labor market perspective. C. PUBLICSECTOR REFORM 4.20 Following a consultative workshop in January 2003, the authorities adopted and are inthe early stages of implementing a Medium Term Public Sector Reform Strategy, aimed at achieving greater cost efficiency and effectiveness in the delivery of public services. The preparation of this strategy was supported with technical assistance from Department for International Development (DFID), the EU, and the World Bank. The strategy has four components: (a) modernizing public administration, (b) strengthening public expenditure management and controls, (c) enhancing growth-supporting public services, and (d) rationalizing the delivery of social services and improving their targeting. The Strategy builds on the broad public consultations undertaken for earlier work on the Integrated Development Plan and the Carib People's Development Plan, a recent Poverty Assessment and Growth Conference, and on other studies and reviews on public sector management issues in Dominica undertaken over the last few years. The Strategy also incorporates and will help to harmonize earlier reform efforts that were ongoing in different parts of the public sector. Strategy implementation is being overseen by a Public Sector Reform Task Force chaired by the Minister of Health and Social Security, and consisting of key public officials and representatives from the Public Service Union, the Dominica Teachers' Union, the Dominica Association for Industry and Commerce, and the National Association of Non-Governmental Organizations. 4.21 Under the first component, modemizing public administration, the authorities are taking measures to strengthen policy formulation, strategic and manpower planning, and Cabinet decisionmaking processes, and to introduce performance management arrangements for senior public officials. Supported by DFID's Fiscal and Economic Recovery Project (FERP), the authorities have held a series of retreats and workshops for senior public officials aimed at clarifying roles and responsibilities. In addition, the Establishment Department i s rolling out a manpower-planning model to assist line ministries in reviewing their staffing requirements, with the objective of reducing the wage bill inan orderly fashion. It is also completing a process to update the Public Service Acts and Regulations, which govern public service employment that was begun earlier. The second component of the strategy, strengthening public expenditure management and controls, includes implementation of the Financial Management Action Plan discussed in Chapter 2, reform of 79 the budget process being supported by DFID's FERP, and improvements in the management of the Public Sector Investment Program described in Chapter 3. Under the third component, enhancing growth-supporting public services, the authorities have established a Task Force on Public-Private Cooperation and appointed a senator with responsibility for private sector relations in the office of the Prime Minister to strengthen dialogue. The Task Force has produced its first report, the recommendations of which the Government i s studying, including the introduction of Customer Service Improvement Plans in key public agencies, such as Customs, and the development of a case management system to clear the backlog and expedite hearings in the court system. The final component, rationalizing the delivery of social services and improving their targeting, has been informed by the results of the Poverty Assessment recently completed by the Government and Caribbean Development Bank, the Carib People's Development Plan, the Social Protection Review completed by the Bank in July 2003, and early inputs from this Public Expenditure Review. The Social Protection Review contains recommendations for streamlining and improving the targeting of social protection programs and for expanding coverage of social security. The Ministry of Education i s working on a proposal to consolidate primary schools, which would simultaneously improve the deployment of teachers, raise the quality of instruction, and achieve some cost efficiency. Finally, the Ministry of Health will review a proposal to expand cost recovery mechanisms in the health service, to consolidate a number of health clinics and district hospitals, and to outsource certain nonessential functions. These measures are consistent with the preliminary findings of this Public Expenditure Review. 4.22 The public sector reform would need to be informed by the following recommended inputs. The legislationthat governs public sector employment has a void with respect to the potential redundancy benefits to be paid when a worker in an established or non-established position is terminated on account of redundancy. Dominica's legislation does not address the issue of the redundancy benefits plan. However, there i s a Protectionof Employment Act that governs employees inthe private sector and that can be used as a proxy to estimate the costs for the government of any downsizing of the public sector. Moreover, public employees that have worked more than 10 years receive a gratuity in case of retirement. Discussions with Treasury indicate that in case of redundancy the gratuity will have to be paid, but the employee will receive only the amount that i s higher between the redundancy benefit and the gratuity. The government needs to get clear legal advice on these issuesbefore starting the reform of its public sector. 4.23 The personnel information systems are fairly basic for the needs of Dominica. The information system that usually contains such staff particulars as date of hiring, date of increments and movements between salary scales, training received, performance evaluation reports, and the date of retirement is inadequate for a modem public sector. It does not contain the information required for manpower planning, professional development, and the design and implementation of a public service reform. 4.24 Moreover, the Dominican government lacks a human resource strategy. Job descriptions for established positions are inadequate. No performance appraisal system i s in place, which makes it impossible to assess whether public servants are actually doing their job efficiently and effectively, and to establish performance pay or bonuses to reward good 80 perf~rmance.~~Training in essential skills such as writing, office procedures, financial and personnel management, and basic management i s deficient. There i s no training plan for individual employees to enable them to advance along their career paths. Training i s often offered in an unsystematic way, and depending on the availability of donor financing. 4.25 Dominica presents some fragmentation and duplication of work among ministries, departments, and agencies, leading to a diffusion of responsibilities. The survey of public officials, carried out for the OECS Institutional and Organizational Capacity Review (IOCR) in 2001, noted a duplication of functions and that such overlap adversely affected the public service. This suggests that missions, objectives, and functions of the different ministries need to be reviewed to determine the most effective and efficient organizational structure. Opportunities to pool resources and realize lower unit costs through greater collaboration extend to almost all areas of public sector activity, including international relations, economic and social policy, and policing. The benefits of common organizational arrangements for tax administration, customs, audit services, police services, magistracy, and environmental protection were discussed in the OECS IOCR. 4.26 There is some possibility for the government to divest some of its functions to the private sector. Some services where it i s easy to specify output can be more efficiently performed by the private sector. These include road maintenance, airport services, laundry and catering services in hospitals, and cleaning and maintenance services in public buildings. Outsourcing or privatization might realize efficiency gains, but significant savings in government expenditures are unlikely. Finally, some public sector activities in the areas o f social services and education could be transferred to private (profit and nonprofit) organizations, as i s being done to a limited extent in St. Lucia. 34 Based on the experience in Grenada, even with a good performance appraisal system it may be difficult to get managers to document any deficiencies of their staff, given the close nature of personal relationships on the island. 81 5. HEALTHSECTOR 5.1 Dominica has invested in a health system with a strong primary health care focus. It has achieved impressive gains in health status, particularly in terms of maternal and child health. The Dominican health system i s dominated by public finance and public provision of care, although there are private physicians, and the private health insurance system appears to be growing. 5.2 Public spending on health in Dominica as a share of GDP i s the highest of all OECS countries. On average, duringFY1991/92-FY2002/03 the share of recurrent health spending of GDPwas about 4.2 percent. The health system that led to this level o f expenditure i s now challenged on many fronts. Most immediately, the public health system i s facing strong budgetary pressures as part of the overall fiscal crisis-pressures, particularly on pharmaceuticals and building and equipment maintenance-that could threaten the country's recent health gains. In addition, the health system i s challenged to adapt to a changing disease profile. With the exception of the HIV/AIDS epidemic, Dominica's disease profile i s increasingly dominated by noncommunicable diseases, such as diabetes and hypertension, and injuries. The health system continues to be challenged by emigration of trained staff, and to grapple with the added complication of how to provide specialized services given the country's small size. 5.3 There are several steps that the government could take over the short to long term to ensure that public spending on health has the highest impact, given it will likely be highly constrained over the medium term. These actions involve changes in how the system i s financed, organized, and delivered. This chapter reviews the status o f health in Dominica, and discusses how public health programs and health services are organized, financed, and delivered. 5.4 The main findings and recommendations can be summarized as follows: Improving and Sustaining Health Outcomes 0 The most important change that Dominica needs to consider in order to continue to improve health indicators and contain costs over time, i s to revitalize public health and primary care to meet the new challenges of the 21'' century-particularly from chronic diseases and injuries. This means focusing health promotion and disease control activities as effectively as possible on tobacco control, injury prevention, control of obesity through diet and physical activity, and screening and early treatment of diabetes and hypertension. This will avert much of the disabilities associated with and costly treatment of preventable heart disease, stroke, late-stage diabetes, and injuries. The other major pri0rit.y for public health is, of course HIV/AIDS, given the stage of the epidemic inthe Caribbean. Improving Eficiency in the Delivery of Service 0 As the population migrates to the urban areas and the disease patterns change over time, and with the goal of achieving greater economies o f scale, it may be possible to 83 rethinkthe configuration of health facilities inDominica. Inparticular: (a) reorganize and consolidate the 45 Type Iclinics and assess the role these clinics should play in noncommunicable disease prevention and control, (b) close or restructure the two district hospitals given that both hospitals have very low occupancy rates (between an estimated 20 and 30 percent), do not have laboratory, X-ray, or surgical capabilities, and it i s not clear how different their role i s from Type I11clinics, and (c) assess the ambulance service throughout Dominica. 0 The OECS Pharmacy Procurement Service has been a model of how the subregion can achieve efficiencies by pooling efforts-although as mentioned below, this model i s threatened by the arrears situation of Dominica, as well as other OECS countries. Over the medium term, there are other areas where shared services might introduce economies o f scale and efficiencies, including medical equipment, biomedical equipment maintenance (having a pool o f experts serve a group of islands), laboratory services, in-service training, and specialist services. 0 The government only charges user fees at Princess Margaret Hospital. These fees are not retained at the hospital level; instead, they revert to the consolidated account. The amount collected i s equivalent to about 10 percent of the Hospital's spending. The user fee policy needs to be thoroughly reconsidered, with changes desirable along the following lines: (a) Permit the Princess Margaret Hospital to retain fees (this may require a change in legal status), which could yield other long-term benefits, such as providing better incentives for collection and giving the hospital budgetary flexibility to improve services. (b) Increase collections from private insurance. (c) Increase fees for nonnationals (beyond the double charge) at the Princess Margaret Hospital. (d) Consider reducing or eliminating fees in some areas or for some groups o f people, in particular, for disease prevention purposes, HlV/AIDS and sexually transmitted infection (STI) tests because these communicable diseases threaten the population at large, and for equity reasons, nonselective procedures that are catastrophic in cost for poor households. Dominicans with a public assistance card are exempt from fees at the Princess Margaret Hospital, but not all poor households have public assistance cards. (e) Consider introducing fees at Type I11 clinics. The fees should be modest (except for nonnationals) and there should be exemptions both b y type of service (immunizations, for example) and by individual (those with public assistance cards). The fees should be retained by the health center to improve quality of care. 84 Uses of Health Spending 0 Wages absorb most of the spending in the health sector and raise questions as to whether there i s sufficient allocation for pharmaceuticals, medical supplies, and maintenance of equipment and buildings for the efficient delivery of services. There may be significant opportunities to generate cost savings without affecting the quality of services by substituting some nursing positions with lower-cost nursing (nurse assistants). 0 Princess Margaret Hospital absorbs about 55 percent of all public spending on health. There are many possible ways that its operation could be made more efficient. Many of these ideas have been discussed for years in Dominica, so strong political leadership will be required to implement them. In addition, some innovations will require up-front investments. 0 Permit the hospital to contract out services, such as laundry, plant maintenance, security, cleaning, and catering. Experience in other countries shows that contracting out, if done well, can lower costs and improve quality. The Hospital might start with one area, gain experience, evaluate it, and if successful, expand the approach. In addition: (a) Permit the hospital to retain user fees (mentioned above). (b) Introduce more hospital autonomy, giving the hospital more budgetary flexibility. (c) Invest in greater use of outpatient surgery as a substitute for more costly inpatient surgery. A. HEALTHOUTCOMES INDOMINICA 5.5 Maternal and Child Health. The government's expenditures in primary health care, and overall improvements in living conditions and educational levels, have generated impressive gains in maternal and child health. As Table 5.1 shows, the infant mortality rate i s low, at about 17 per 1,000 in 2000, with about 60 percent of all infant deaths occurring inthe first week of life. A trained health care professional attends almost all-over 99 percent-of deliveries in a health facility. While Type I11 health facilities and the country's two district hospitals are equipped to handle deliveries, this capacity i s underused because most women now prefer to be admitted to the country's one general hospital, Princess Margaret Hospital, to deliver. There, about 9 percent of deliveries are by cesarean section, a favorable rate by international standards. At the Princess Margaret Hospital, women are strongly encouraged to breastfeed their babies, although data are lacking on the proportion of women that initiates breastfeeding and the duration o f exclusive breastfeeding. Problems from diarrheal disease and respiratory infection are declining in young children in Dominica, but accidents are increasing. Immunization rates are very high, although there has been some slippage in recent years, and very recently, the budgetary crisis led to a shortage of vaccines. Obesity i s considered an increasing problem in children under age 5. A 85 micronutrient study in 1997 found that iron deficiency i s the only major micronutrient problem, with problems concentrated in young children, adolescents, andpregnant women. Table 5.1. Health IndicatorsinDominica Compared with other OECS Countries, Barbados, and Jamaica Infant Life Total GDPper Mortality Total Fertility Expectancy Population Capita, 2001 Rate (IMR) Rate at Birth 2001 (current EC$) 2000 2000 2000 Dominica 73,200 3,563 17 1.9 73 St. Kitts and Nevis 41,100 8,345 19 2.2 71 Antigua and Barbuda 68,500 9,729 15 1.7 75 Barbados 268,200 -- 17 1.8 75 Grenada 99,000 4,022 13 3.2 73 Jamaica 2,668,200 2,917 20 2.5 75 St. Lucia 158,100 4,356 13 2.0 71 St. Vincent and the Grenadines 115,900 2,914 20 2.1 73 Source: World Development Indicufors, World Bank. IMR and life expectancy for Dominica refer to 2001 and are from the Ministry of Healthreports. 5.6 Teen Pregnancy. While the share of births to teenagers has been declining, it was still fairly high at 15 percent in 1998, in spite of free contraceptives at clinics and the efforts of the Dominica Planned Parenthood Federation. This high share i s in part due to the very low age of sexual initiation in Dominica. Adolescent health surveys indicate that this i s as low as 10, but more frequently it i s age 14 to 15 for girls. A recent Adolescent Health Survey of 2,809 students from 40 primary and secondary schools indicated that almost half (47 percent) of the students who had sex indicated they never used contraceptives. About 14 percent of respondents reported they had been physically abused or mistreated. Sexual abuse was reported by about 10percent of the respondents. Of these 10 percent, 20 percent was by an adult living in the household, 44 percent by an adult in another household, 12 percent by a sibling, and 31 percent by other friends not in the same household. 5.7 HZV/AZDS. Although the prevalence rate is relatively low, the epidemic could increase rapidly, as it has in other countries. HIV/AIDS was first diagnosed in Dominica in 1987. Cumulative HIV-positive cases totaled 257 through mid-2002. The Caribbean Epidemiology Center (CAREC) estimates that the adult overall HIV prevalence rate was about 0.2 percent at the end of 2001. While the majority of the cases are still male, the female proportion i s increasing. These numbers must be interpreted with caution because there i s under- and incomplete reporting o f HIV/AIDS cases. The government supports condom distribution, voluntary counseling and HIV testing, prevention of mother-to-child- transmission, prevention and treatment of STIs, and education programs, but the programs need to be scaled up, with multisectoral leadership. A draft of the National HIV/AIDS Strategic Plan (2003-07) has been prepared and distributed for discussion. 5.8 Noncommunicable Diseases and Injuries. Given the aging of the population and the epidemiological transition, noncommunicable diseases have become the main cause of death in Dominica. The five principal causes of death are malignant neoplasms, hypertension, heart disease, diabetes mellitus, and other diseases of the respiratory system. 86 Transport accidents are now one of the top 10 causes of death. The level of diabetes and hypertension appears to be particularly high. Diabetes is the second-most-frequent cause of visits to a health center and the fourth principal cause of death. The Princess Margaret Hospital offers dialysis services. Factors behind the growing problems from diabetes and hypertension are complex. In addition to the genetic component, changes in lifestyle from reductions in physical activity (less of the labor force in agriculture) and changes in diet (the popularity of soft drinks, candy, and processed foods substituting for fresh vegetables and fruits) are major contributing factors. 5.9 Water and Sanitation. Access to water and sanitation have improved, but an estimated 20 percent of households still lack access to proper sewage disposal. The sewerage system in Roseau was recently improved, but it remains a problem for a significant proportion of especially poor households not currently possessing flush toilets. However, more than 95 percent of the population has access to potable water, either in homes or from standpipes. 5.10 Endemic Diseases and Substance Abuse. Dengue fever is still endemic in Dominica, with large outbreaks every few years. Malaria has been eradicated-the cases that occur are all imported. Substance abuse i s a problem in Dominica. This includes marijuana, alcohol, and cocaine, with marijuana and alcohol the most common problems. Currently, an adolescent Drug Education program has been established to address concerns about youth becoming involved in drugs. 5.11 Problems in the Carib Population. The Carib population, concentrated i n a reserve in two districts, Castle Bruce and Marigot, is one of the poorest communities in Dominica. This group has less access to potable water and solid waste disposal than the population at large. It has higher rates of tuberculosis than the rest of the country, and alcoholism, STIs, and heminthic (worms) infections are considered to be problems in this group. The Carib population totals about 4,500 inhabitants. 5.12 The under-5 mortality and HIV/AIDS targets of the Millennium Development Goals (MDGs) are the most applicable to Dominica of all the health-related MDGs. Malnutrition (defined as the prevalence of underweight children), maternal mortality, and tropical diseases (other than dengue, which i s endemic) are not major health issues in the country. The under-5 mortality rate in Dominica dropped from an estimated 23 per 1,000 habitants in 1990 to 14 per 1,000 habitants in 2000. This i s a decline of about 5 percent per year. The MDG target-a reduction of two-thirds from 1990 to 2015-would require that the rate fall to about 8 per 1,000 habitants by 2015, or about 4.4 percent per year. This MDG for Dominica in 2015 i s similar to the under-5 mortality rates observed in the United Kingdom and inthe United States today. On trend projections, Dominica i s well on track for achieving this MDG target. It i s difficult to assess the likelihood o f reaching the HIV/AlDS targets because baseline data on HIV prevalence and incidence are poor. However, given the size of the epidemic in the Caribbean, the target i s very relevant for Dominica and should be achievable if the country supports vigorous and effective HIV prevention measures. 87 B. THEHEALTHSYSTEM: FINANCING, COVERAGE,AND DELIVERY 5.13 Dominica's small size poses many challenges to the administration of public healthprogramsand clinicalservices, such as (a) mostly publicly provided health services, (b) unexploited potential economies of scale, (c) health financing mostly through general revenue, and (d) a high ratio of nurses to physicians. 5.14 Health care in Dominica, as in the other OECS countries, is predominantly publicly provided. All hospitalsand clinicsare public. At the central level, the Ministry of Health carries out policy formulation, operation o f the Central Medical Stores, and training, and it has small units focused on environmental health services, drug prevention, health information, and health promotion. The government, with donor assistance, invested in a strong primary health care model that consist of 45 Type Ihealth clinics (within 5 kilometers of every village) and strong links to the 7 Type I11health clinics. Health clinic staff make home visits and conduct outreach activities in the community. Each of the seven health districts has a Type I11health clinic with more specialized staff, including a resident doctor, who assists and supervises the outreach activities of the Type Iclinics. In addition, there are two small district hospitals that also provide inpatient services, although in practice these district hospitals are largely bypassed in favor o f Princess Margaret Hospital, which offers much more comprehensive services (see Table 5.2). Table 5.2. Ministry of Health Facilities and Functions Level of Care Function PRIMARY CARE Type IHealth Clinics (45 clinics) Provide health education, maternal-child health services, management of diabetes and hypertension, treatment of common diseases. Serve a population of 500 to 2,000 people. Staffed by primary care nurse. Type 111Health Clinics (7 clinics) Provision of more complicated primary health care. Resident doctor, family nurse practitioner, environmental health officer, pharmacist, community health nurses, registered nurse midwives, dental therapist, and support staff. District Hospitals at Portsmouth and 56 total beds. Little resolutive capacity (no surgery, X-ray, or Marigot Districts laboratory facilities). SECONDARY CARE Princess Margaret Hospital (General A 224-bed secondary hospital that provides inpatient services, Hospital) ambulatory, emergency, specialist, and diagnostic services, and dialysis, neonatology, and intensive care services. TERTIARY CARE No services in Dominica. Government refers some patients to overseas care and provides very modest assistance to those in need. Out-of-pocket and private insurance purchases overseas care, largely from neighboring islands. Source: Ministryof Health. 5.15 While the island has 45 Type Ihealth clinics, many are now nonfunctioningor only partly operationaland inneed of repair. Traditionally the government has depended on external assistance for such repairs; however, because external assistance has declined 88 over the years, the Type Iclinics have deteriorated. Type IIIclinics are in better shape, but do suffer from drug shortages and problems with repairs and equipment maintenance. 5.16 However, before new investments are made to repair Type I and I11 health clinics, the government should rethink its primary health care model. The health care model that Dominica developed has contributed greatly toward improved health indicators, but it is a costly model to maintain. With the aging of the population, migration to urban areas, and changes in disease profile, it might be possible for the government to: (a) reorganize and consolidate the 45 Type Iclinics and assess the role these clinics should play in noncommunicable disease prevention and control; (b) close or restructure the two district hospitals, given that both hospitals have very low occupancy rates (between an estimated 20 and 30 percent), do not have laboratory, X-ray, or surgical capabilities, and it i s not clear how different their role i s from Type I11clinics; and (c) assess the ambulance service throughout Dominica. Some rethinking of the primary care model might permit exploiting economies of scale and reducing the very large wage bill of the Ministry of Health (MoH), which i s currently at 78 percent, and increase efficiency, thereby freeing up resources for pharmaceuticals, buildingrepairs and maintenance, and equipment maintenance. 5.17 Princess Margaret Hospital had an occupancy rate of 74 percent in 2000- favorable by international standards-with an average length of stay of 6.6 days. The emergency room in the Princess Margaret Hospital i s very crowded, especially since a nearby Type I11Clinic in Roseau was closed because it was in disrepair and there were no funds to fix it. The hospital also suffers from drug shortages, problems with equipment maintenance (lack of trained staff and budget), difficulties in attracting doctors (many are from overseas), and lack of key services such as mammography screening. The hospital management i s also limited in its ability to innovate and seek efficiency improvements, such as contracting out, because of its legal status as a public hospital. St. Lucia, a neighboring island, i s making its two largest hospitals statutory entities, to give the hospitals more autonomy so that they can, among other things, contract out services, such as laundry, plant maintenance, security, and cleaning and catering, and retain user fees to reinvest in the services. Experience from other countries shows that contracting out, if done well, can lower costs and improve quality. The Princess Margaret Hospital might start with one area, gain experience, and evaluate and, if successful, expand, the approach. 5.18 The PrincessMargaret Hospital is attempting to introduce an innovation in the management of its new CT scan. Before it i s put into full use, it has requested the Ministry o f Finance to give the Hospital permission to retain the user fees for the CT scan, in a special account for that purpose. Although the user fees are not the full cost of the service, the Hospital hopes that the fees could ensure the long-term maintenance o f the unit. This could be a useful pilot in a transition to giving the Princess Margaret Hospital more autonomy in order to improve performance. 5.19 PrincessMargaret Hospitalcould also capture efficiency gains inother areas, for example, by substituting more inpatient surgery with outpatient surgery. Some efficiency gains will require up-front investments that the Hospital does not at present have. The Princess Margaret Hospital i s trying to set up a nongovernmental organization (NGO) called Friends of Princess Margaret Hospital to fundraise for the hospital. There may be 89 creative ways to attract donations from Dominicans living abroad, and from local residents and businesses. In response to the budget crisis, some Dominicans living overseas are sending contributions to their communities, often in-kind. The problem with these contributions i s that they may not be the items that the health system most urgently needs. In that case, monetary contributions that could be channeled to purchases through the OECS Procurement System would have more impact. 5.20 An excellent example of exploiting economies of scale at the regional level, and that can be extended to other services, is the Eastern Caribbean Drug Procurement Service. The St. Kitts and Nevis government uses a National Formulary to guide purchasing. The government may be able to achieve greater economies of scale by better coordinating other services with its OECS neighbors and other countries in the region, such as specialized laboratory tests and procurement of medical supplies. 5.21 The health system is predominantly publicly financed through the consolidated fund. All services are free in the primary care system (Type Iclinics), Type I11health clinics, and the district hospitals. Revenue from user fees is minor. The only fees charged are for services at the secondary hospital, Princess Margaret Hospital. People with public assistance cards are exempt from user fees. The main objective of the user fees i s to generate revenue. As Table 5.3 shows, in FY2000/01, 10.3 percent of the hospital's expenditures were collected from user fees. The user fee schedule was last amended in June 2000. The fees collected are passed on to the government's consolidated account. Only those with public assistance cards are automatically exempted from payment of fees, but in practice, there i s huge variation in collection rates from those not on public assistance, mainly by type of service. Laboratory tests, radiology, and the pharmacy have very high collection rates compared to other areas of the hospital. Collection rates for elective surgery are fairly good. Collections from obstetrical patients and for emergency surgery are very low. Table 5.3. User-Fee Revenue and Revenue as a Share of Princess Margaret Hospital Expenditure, FY2000/01 (US$) Revenue Collection by Type of Fee Hospital Fees 485,624 X-Rays at Hospital 112,219 Laboratory Fees 73,472 Dental Fees 24,740 Total 696,054 Total Expenditure, Princess Margaret Hospital 6,763,384 Total User Fee Revenue as Share of Hospital 10.3% Expenditure Source: 2002/03 Estimatesof the Commonwealthof Dominica. 5.22 While there has been an effort to increase the billing and collection procedures, there are a host of problems in collection. The Hospital's billing department reports that i t can be difficult to collect the posted fees, because people think the services at the hospital should be free. When patients cannot pay up front, the billing department gets people to sign promissory notes for user fees, but it i s then difficult to collect. It also signs people up for a 90 salary deduction from their wages, but some companies will not participate in that effort, and it is administratively costly to collect. The billing department is only open certain hours, so many patients never have contact with it. Privately insured patients might appear to be one straightforward source of revenue, but in practice collecting from the privately insured can be difficult. The administrative costs of collecting from privately insured patients are high-the hospital does not require privately insured patients to pay their portion o f the user fees up front. It first pursues collection of the user fees from the private insurance company, and then tries to collect the balance from the patient. 5.23 There is considerable possibility for increasing collection of user fees, but at the same time the user fee policy would benefit from a review in terms of equity and efficiency. Table 5.4 illustrates user fees for selected procedures, tests, and hospital stays. Some o f the charges appear to be too high-both for public health reasons and for financial protection. In terms of public health, one could argue that there should be no financial obstacle to HIV testing because of the seriousness of the epidemic and its threat to the community at large. In terms of financial protection, an appendectomy or cesarean section carries a considerable fee for the procedure (about US$300), which could be a catastrophic cost to a poor household. Not all poor households are exempt from fees because not all poor households have public assistance cards. Thought should be given to reducing or eliminating fees in some areas or for some groups of people-for disease prevention purposes (the HIV test, for example) and for equity reasons (nonelective procedures that are catastrophic incost Table 5.4. User Feesat Princess Margaret Hospital Procedure Charge (US$) Blood dialysis; repeat blood dialysis $149; $56 Drugs $2 per prescription Pediatric consultation-newborn, day 1 $30 Pediatric consultation, day 2-7 $19 Normal vaginal delivery $37 Forcepsdelivery $75 Cesareanprocedure $299 Appendectomy (acute with complications) $299 HIV 1& 2 test $15 Chest X-Ray $15 Accommodation, including routine nursing services 2-personroom, $56 per day; 4-6 personward, $19 per day; more than 6-person ward, $7 per day Source: Commonwealth of Dominica Ministry of Health Payment Schedule, 2000. Fees are double for nonnationals. to poor households). If the hospital could be permitted to retain fees to improve the quality of services, that would probably be incentive enough for the billing department to figure out ways to increase collections, especially from the privately insured. Fees could also be increased (beyond the double charge) for nonnationals. Very modest user fees might also be considered at Type I11health clinics, with exemptions by type of service (well-baby care, immunizations), and by individual (with a public assistance card), if those fees could be retainedby the clinic and reinvested inthe services. 91 5.24 The mandatory social security taxes for public sector and formal private sector workers could play a greater role in the long term in health financing, although it was not set up to contribute much toward health. But the system has many issues to address in the short to medium term before this could be considered. In terms of health spending, Social Security has of late used a small portion of its collections, US$1million to US$1.5 million, to offset loans it made to the government to build sections of Princess Margaret Hospital. It also provides some modest assistance in some years in helping members with overseas medical care. 5.25 The private health sector in Dominica can be classified into five categories: (a) NGOs, (b) private physician (many of whom are publicly employed as well) offering outpatient services, (c) one medical laboratory, (d) private pharmacies, and (e) private health insurance. There i s little formal public-private collaboration in health in Dominica at present, although there i s scope for developing greater collaboration to improve the quality and efficiency of care. NGOs are a small segment of the health sector, and are most prominent in family planning services and HIV education and prevention. Major NGOs active in health include the Dominica Red Cross and the Dominica Planned Parenthood Federation. There i s also a tiny educational association for diabetes-the Dominica Diabetes Association. The Dominica Red Cross does first aid andCPR training, disaster preparedness and response, and HIV peer education for youths and young adults. The association i s small, with three staff, and a total annual budget of about US$48,000. It has 14 branches throughout the island, with volunteer staff. The main sources of contraceptives in Dominica are from the Dominica Planned Parenthood Federation and private pharmacies. The Dominica Planned Parenthood Federation has only one clinic to serve the entire island, and it i s located in Roseau. Its annual budget i s about US$112,000, with 40 percent from user fees and 60 percent from the International Planned Parenthood Federation (IPPF) and other agencies. The clinic offers pregnancy tests, pap smears, and contraceptives. Because of the cheap prices it obtains, the Dominica Planned Parenthood Federation procures about 60,000 condoms annually (through IPPF) for the government. In addition to supplying the government with condoms and providing them to its own clinic patients, it supplies a network of shops and bars with the same very inexpensive condoms in plain wrappers. It has plans to expand sales by also offering a Caribbean-brand condom with fancier packaging and a higher price. The Dominica Diabetics Association operates from a small office in a closed health center and i s staffed by part-time volunteers. The Association holds talks on diabetes and has activities to observe World Diabetes Day. 5.26 The main for-profit actors in the health sector are pharmacies, private physicians offering outpatient services, and the L a Falaise House Medical Laboratory inRoseau. Pharmacies are not uniformly distributed across the country, and several health districts have no pharmacy. This means when the public health clinics are out o f drugs, people living in districts without a pharmacy may have to travel far to get drugs. Many publicly employed physicians also offer private services. The private medical laboratory has been in business 10 years. It gets clients with private insurance and clients who pay out-of- pocket for its services. While it gets patients referred by the government for tests that the Princess Margaret Hospital cannot perform, the government does not reimburse it for these tests-the patient pays directly. 92 5.27 In2000, there were an estimated 0.8 doctorsper 1,000 population, and about 4.8 nurses per 1,000 population, which results in a high nurse-physician ratio of about 6 to 1. Overall, there were 22 specialists and 38 general practitionersat that time.35 Mostwere in the public sector (17 specialists, 32 general practitioners), although many also had a private practice. Firms from the United States and United Kingdom travel to Dominica to recruit skilled nurses-usually after the nurses have gained a few years o f experience. Conversely, the government recruits physicians and other skilled personnel from overseas. Most Dominica-born physicians attend medical school at the University o f the West Indies or in Cuba. Dominica has a very large proportion of nurses relative to nursing assistants, and it could consider making efficiency improvements by substituting nurses with nursing assistants where possible. There i s an offshore medical school in Dominica, Ross University School of Medicine, mainly for U.S. students, but there i s little collaboration between the School of Medicine and the health system of Dominica. It provides one scholarship a year for its 18-month academic training program. 5.28 Many of the human resource problems inDominica stem from its small size and its difficulty in retaining skilled staff. It may have a need for only one biomedical engineer, and if that person leaves, there i s no backup. One way to lessen this problem would be to consider if more specialized services might be shared among the OECS countries. This could include specialist physician services, but also biomedical engineering, and even possibly health promotion. The emigration of skilled nurses is something that the government has little control over given the pull forces in the large markets trying to recruit these well-trained nurses. It can best respond, as it has done, b y training larger numbers of nurses. The country overall does benefit from the remittances these nurses send back to Dominica. c. USESOF HEALTH SPENDINGINDOMINICA,FY1991/92-FY2002/03 5.29 On average, from FY1991/92 to FY2002/03, the public recurrent spending on health as a share of GDP was 4.2 percent per fiscal year, with an upward trend. The share in total recurrent expenditures was 13.7 percent. As Figure 5.1 shows, in FY2000/01, public spending on health was 4.5 percent of GDP, or 12.8 percent of total recurrent expenditures (or US$167 per capita). The numbers budgeted for FY2002/03 are roughly similar (4.9 percent of GDP, 16.1 percent of total recurrent expenditure, or $170 per capita). However, since FY1998/99 the share of recurrent health spending showed an upward trend. As a share of total government recurrent expenditure, the M o H share has fluctuated between 12.5 percent and 16.2 percent. 35Pan-AmericanHealthOrganizationreport. 93 Figure 5.1. Share of Recurrent Health Expenditures inGDP and Total Recurrent Expenditures, FY1991/92-FY2002/03 - (percentage) , O O i - ---- ___---- - - - -- _ - __-_ ~ I - _ - t 0loo 00 1:: 1991192 1992193 1993194 1994195 1995196 1996197 1997198 1998199 1999100 2000/01 2001/02 2002103 as Yo GDP Fiscal Year +as Yo Total GovernmentRecurrentExpenAbxe Source: Dominicaestimatesfor the year. All figures are actual except for FY2001/02,which are revised, and for FY2002/03,which are estimates. 5.30 Public spending on health in Dominica as a share of GDP is the highest of all OECS countries. Table 5.5 presents comparative estimates prepared by the World Health Organization for public and private spending on health in the OECS countries and Jamaica. It also indicates that public spending on health is estimated to be about 71 percent of total spendingon health in Dominica. Table 5.5. Comparative Estimates of Health Spending inDominica Compared to Other OECS Countries and Jamaica, 2001 Total Spending Public Spending PublicSpending (Public and Private) on Health, on Health, on Health, as % of GDP Per Capita as % of GDP Antigua and Barbuda 3.4 $323 5.6 Dominica 4.3 $145 6.0 Grenada 3.8 $188 5.3 Jamaica 2.9 $80 6.8 St. Kitts and Nevis 3.2 $261 4.8 St. Lucia 2.9 $129 4.5 St. Vincent andthe Grenadines 3.9 $113 6.1 Note: The figures for Dominica differ slightly from those reported in the text of this chapter. The figures inthe text of this report are from the most recent government budget and expenditure books. Also, becauseno good estimates of private spendingon healthexist for most of these countries, these numbers should be interpretedwith care. Source: Preliminary estimates, World HealthOrganization. 94 5.31 In constant EC dollars, recurrentexpendituresin health increasedin the early 1990s, but have remained fairly flat during FY1998/99 to FY2002/03. As Figure 5.2 shows, from FY1991/92 to FY1997/98, recurrent Figure 5.2. Recurrent Health Spending inDominica, FY1991/92- expenditures in health FY2002/03 (in EC$95) increased by 5.9 percent per 3 5 0 1 year, mainly due to increases in the wage bill, compared to 0.5 percent from FY1998/99 to FY2002/03. 5.32 Wages represent the largest share of recurrent health spending (74 percent). On average, from FY1991/92 to FY2002/03, the share of wages in recurrent health spending was about 74 percent, with an upward trend since FY1997/98. As suggested in paragraph 5.14, Nore: All figures are actual, except for FY2001/02, which are revised, and FY2002/03, which are estimates. one of the for this Source: Dominica estimates for the year. large share i s the health care model that Dominica developed based on a large number of Type Iclinics and two district hospitals that have very low occupancy and are sidestepped by the population in favor of the Princess Margaret Hospital. In contrast, the share of supplies and materials in health expenditures declined from about 13 percent in the early 1990s to 11 percent inthe 2000s. 5.33 By program, Princess Figure 5.3. Share of Health Programs in Recurrent Health Margaret Hospital's share in Expenditures, FY1991/92-FY2002/03 recurrenthealthexpenditures (in percentage) has been fairly constant over 100% I O l h m 90% time, at 56 percent. But in OPme55 MargaRtHospital 80% OSt JosephHealthD&t Shtntal contrast, pharmaceutical expenditures have been MCsstleBmeHedthDutnct 60% severely underfunded in I L a P hHealthDrtnd 50% recent years. As Figure 5.3 0G d a yHealthDlstnct 40% shows, the share of IMargotHealthDlstnct 30% pharmaceuticals in recurrent 20% RPoltsmouthHealthDistnct lo% health expenditures decreased ox IRoseauHealthDetnct from about 10 percent in total F ~ X Y I~ V ~ OPHC, cenhallwel health expenditures in the early @@ @jQ 8' 4' 4'' e@&@#",$+" ,$+@ E#OperatmnofMe&alStores 1990s to percent in Nore All figures are actual except for FY2001/02, which are revised, and FY2002103, FY2002/03. About 95 percent which are emmates. of drugs and 5 percent of Source. Domnica estimates for the year medical supplies purchased by the government are done through the OECS Pharmaceutical Procurement Service, which has been in operation for about 13 years. The OECS 95 Pharmaceutical Procurement Service consolidates purchases o f the OECS countries and achieves excellent prices as a result. The government tries to maintain its purchasing of vital and essential drugs even under budgetary pressures. Contraceptives used to be purchased by the public sector but, because of budgetary pressures, were moved several years ago to the nonessential category. Now the only contraceptives purchased with public funds are condoms. The Princess Margaret Hospital and health clinics have experienced drug shortages in recent years due to budgetary pressures, but the situation has particularly worsened inthe past year. 5.34 Moreover, the government has a Table 5.6. Average Payment Time, OECS serious problem of arrears to Pharmaceutical Service, by Country, pharmaceutical suppliers; as of July 2002 it was about US$540,000 in arrears to Average Number them. This was brought down in September Country of Days To Pay Supplier Dominica 99 2002 to about US$400,000. Suppliers BritishV.I. 57 through the OECS Pharmaceutical St. Lucia 44 Procurement Service require payment St. Vincent and the 57 within 60 days of the shipment of goods. Grenadines Dominica averaged 99 days to pay suppliers Antigua and Barbuda 82 Grenada 53 during FY2002/03-the worst record of the Montserrat 72 nine countries participating in the scheme St. Kitts and Nevis 78 (Table 5.6). This problem i s not the only Anguilla 46 one that Dominica faces, because late Source: OECS Pharmaceutical Procurement Service. payments threaten the low prices achieved by the pooled procurement scheme. The OECS Pharmaceutical Procurement Service reported that the number o f vendors coming forward to bidfell sharply in2002, which it attributes, inlarge part, to the increasingly late payments by the countries in the scheme. 5.35 The drug shortages inthe government's health clinics and the Princess Margaret Hospital can have an impact on both health status and costs, especially for the poor. Health center staff report that they often cannot give full prescriptions to diabetics and hypertensives-the patients must either locate and purchase the balance on their own, or do without. This can result in serious medical complications that can worsen the condition and result in even costlier treatment. Health staff predict an increase in amputations and strokes from the shortages of diabetic and hypertensive drugs. Health staff from one Type I11clinic described in October 2002 the impact of shortages influphenozine on the schizophrenics in the community. Fhphenozine i s an injection used to calm schizophrenics for a period of three to four weeks, enabling them to function, hopefully, inthe community. With the recent unavailability of this drug, the community's several schizophrenics became so agitated that they had to be committed by the local police to the Princess Margaret Hospital psychiatric ward for inpatient care. 5.36 The arrears situation has also recently affected the country's vaccination programs, although not so dramatically. There were teinporary disruptions in the country's vaccination program, which has had a sterling record o f reaching over 99 percent of children with bacille Calmette-GuCrin (BCG, for tuberculosis), diphtheria-pertussis- 96 tetanus vaccine (DPT), and polio, because of payment problems. Vaccines are purchased through the Pan-AmericanHealth Organization's (PAHO's) revolving fund for vaccines. 5.37 Privatespendingis madeup of (a) out-of-pocketpaymentsfor userfees inpublic facilities; (b) out-of-pocket payments to private doctors, pharmacies, and labs in Dominica; (c) out-of-pocket payments for overseas care; and (4) private health insurancepremiums. User fees collected in the Princess Margaret Hospital amounted to about US$9.50 per capita in FY2000/01. No good estimates exist for the other sources of private spending. Private physicians are mainly in urban areas in the north and south of the island. With the exception of two districts that have a pharmacy, all other pharmacies are in the capital, Roseau. Five firms are registered to sell health and life insurance in Dominica, including the American Life Insurance Company (ALICO), British American Insurance, the Colonial Life Insurance Company (CLICO), Guyana and Trinidad Mutual Life Insurance, and Life of Barbados/Barbados Mutual. Health insurance tends to cover, after deductibles, Princess Margaret Hospital inpatient charges (the user fees), private physician outpatient care, and overseas care. Since private insurance only pays for the user fees at the Princess Margaret Hospital, it should be considered "supplemental" insurance, because the hospital gets most o f its funding from general tax revenue. One of the largest insurance companies, Life of BarbadosBarbados Mutual, reported that close to half of the claims paid are for overseas care. Airfare i s also reimbursed by some companies. It i s difficult to estimate the total population insured, but it i s probably between 15,000 and 25,000. There i s little oversight as of yet of the health insurance sector by the Insurance Regulation area of the Ministry of Finance. The current laws that regulate insurance are weak. A draft of a more comprehensive Insurance Act has beenprepared. 5.38 Capital expenditures on health have been almost entirely donor dependent in Dominica,with a very large sharefromthe EU. UntilFY2002/03, the government didnot prioritize projects within its Public Sector Investment Program. Since then, the authorities have made efforts to increase the focus of the PSIP, and are developing guidelines for enhancing project selection and implementation. Essential repairs are not being made to health centers, despite the consequences (such as the extreme overcrowding of the Princess Margaret Hospital), because of lack of external assistance. The government ultimately needs to start assuming this responsibility with its own revenue base, a task that i s not easy in this budget environment, and it also needs to develop ways to further prioritize these investments. 97 6. THE EDUCATION SECTOR 6.1 Education i s fundamental for economic and social development. The human knowledge and skills taught in the education system are a highly valued commodity in the labor market. InDominica, all indications suggest that education i s key to higher income and better jobs. According to the 2002 Poverty Assessment, a household head with primary or less education headed 90 percent of the poor households. 6.2 The Dominican education system has responded in a positive manner to society's pressures by noticeably improving access to secondary school. Large challenges lie ahead, however. As of 2000, less than 20 percent of the labor force in Dominica had completed secondary education, and less than 5 percent graduated from tertiary education. Indeed, the distribution of educational attainment inDominica resembles Barbados 40 years ago. 6.3 On average during FY1995/96-FY2001/02, Dominica spent 6.6 percent of GDP on recurrent and capital education expenditure^.^^ From a regional perspective, this i s an exceptionally high share of public spending on education. Latin American and Caribbean countries invest on average below 5 percent of GDP on education. In a Caribbean context, Dominica i s equally above average. If this allocation were used efficiently, it would imply a sizable investment inthe country's future growth potential. 6.4 However, as this chapter will show, the relatively high level of expenditures did not result in commensurate educational outcomes. While there have been significant gains in coverage, education expenditures suffer from serious efficiency and equity issues. The challenge for Dominica i s to improve outcomes and access in the context of a fiscal crisis in which fiscal savings are crucial. 6.5 This chapter reviews education outcomes in Dominica, discusses how education services are financed, organized, and delivered, and suggests interventions to improve outcomes, equity, and efficiency. The main findings and recommendations can be summarized as follows: ExpensiveDelivery of Education at the Primary and Secondary Level 0 Dominica has one of the most expensive education delivery systems of the OECS countries after St. Kitts and Nevis, mostly due to its high pupil-teacher ratio. The largest fiscal savings will come from increasing the pupil-teacher ratio in primary school to the Caribbean average of 25 pupils per teacher b y eliminating at least 170 teacher positions. This will generate savings o f about EC$5,000,000 (12.9 percent of recurrent education expenditures). Increasing the pupil-teacher ratio in secondary education to 23 pupils per teacher would generate fiscal savings of approximately EC$2,600,000 (6.7 percent of recurrent education expenditures), or provide additional secondary education to 1,000 students, or 70 percent o f what i s needed to achieve universal secondary education in 2002. Among the initiatives that the government can consider to increase the pupil-teacher ratio are: 36This figure refers to capitaland recurrent expenditures in the Ministry of Education, Sports and Youth. 99 Merging of primary schools. Inefficient use of teachers predominantly takes place at smaller primary schools mainly located in rural areas and experiencing a falling intake because of declining birth rates, urbanization, and migration. A merging of schools based on school mapping would exploit economies of scale inthe education system and reduce cost, while increasingquality. Mandating a minimum floor on class size. Many schools at the secondary level offer subjects with low demand, leading to small class sizes and highcosts. Mandating a minimumclass size of, for instance, 15-a rule that i s in use in many developed countries-would increase efficiency. Instituting a curriculum reform of secondary education. A common core curriculum for the first three forms of secondary education would reduce the excessive offering of subjects of a specialized nature, such as typewriting, auto mechanics, and metalworking, and lead to larger classes. Accelerating the phasing out of the Junior Secondary Program (JSP), in order to provide five years of secondary education to all students. This would free up resources and enable a more efficient use of the expensive facilities for vocational training. Changing funding of publicly assisted schools to one where the school receives a fixed amount per student (possibly an amount sufficient to cover teacher salaries for a pupil-teacher ratio similar to the one in public schools). This would liberate enough teachers to instruct 460 new pupils, a third of what i s needed to achieve universal secondary education in 2002, or savings of approximately EC$700,000 (1.7 percent of recurrent education expenditures). Rewarding well-performing schools. The Ministry of Education could provide incentives for each school to deliver education more efficiently by rewarding schools that commit to increasing the pupil- teacher ratio while retaining part o f the savings. Low Quality of Education Due to a Low Level of Teacher Training. The government could consider: Limiting only certified teachers for open teacher positions. This would prevent the lack of teacher qualifications from further deteriorating. Establishing a major retraining program o f the existing teaching staff, in particular in secondary education. However, this presupposes investments in teacher training courses, either through public or private investments. 100 Public Unit Costs in Secondary and Tertiary Education are too high. They can be reduced by: 0 Raising fees in secondary education. Currently, there exists a de facto policy of small fees. The government could formalize the de facto procedure by instituting a small fee, for instance EC$50 per year (US$18). The fiscal revenue associated with this small fee i s minor, EC$150,000 to EC$250,000, but the revenue could be a very important source for quality improvements if the fee was paid directly to the management of each school. 0 Raising fees in tertiary education. Nationals paid a fee of EC$250 per semester (US$93) in 2002, which i s payable directly to the community college. It constitutes an important source of revenue for the community college-I3 percent of total expenditure on tertiary education. Increasing this ratio to 25 percent would generate EC$l.2 million (2.8 percent of the recurrent education budget). Increasing regional cooperation between colleges, which in the medium term could pave the way for quality and efficiency improvements through specialization. Reducing Administrative Costs 0 Through deeper cooperation within the country and at the regional level with the development of exams, information systems, strategies, and curricula. This would reduce duplication of work and lead to quality improvements. A 25 percent reduction in administrative costs would imply economies of EC$l,OOO,OOO (2.6 percent of the recurrent education budget). Better Targeting of Merit Scholarships 0 In FY2001/02, the Ministry budgeted EC$522,600 for scholarships awarded to the best-performing graduates from primary education. However, it i s doubtful whether the program increases enrollment and completion of secondary education. The government could consider either immediately reducing the merit-based scholarships, or phasing out current commitments without granting new scholarships; alternatively, the government could change the awarding criteria to be need based, as i s the stated purpose inthe government budget. A. THEEDUCATIONAL SYSTEM INDOMINICA 6.6 Except for preschool education, public and publicly assisted institutions are the main providers of education services. As Table 6.1 shows, preschool education i s mostly provided by private institutions (66 schools). In the 19 publicly assisted institutions, the government only supervises and assists in training activities. Public and publicly assisted schools enroll more than 90 percent of the students inprimary school and above. Inprimary and secondary education, publicly assisted schools are mainly religiously affiliated schools. The government assistance consists of teacher salaries and the transfer o f a small per-pupil subsidy. 101 Number of Institutions Share of Enrollment(as %) Public E?:: Private Total Public Publicly Assisted private Sum Preschool 0 19 66* 85 NA NA NA NA Primary (including 55 5** 6 66 74.3 18.0 7.7 100 JSP) Secondary 7 7 1 15 62.7 34.2 3.1 100 Tertiary 1 0 O*** 2 100.0 0 O*** 100 6.7 Although common in a Caribbean context, the existence of a Junior Secondary Program (JSP) parallel to general secondary education is an expensive duplication. A pupil's results on the Common Entry Exam (CEE) determine whether he or she will enter the five-year general secondary education cycle or the three-year JSP. This system was designed to give pupils that failed to gain access to general secondary education an option to continue their education. However, as this chapter will show, this parallel system implies an expensive duplication of structures and a partial success in outcomes. 6.8 Provision of preschool education is an example where, if correctly managed, private provision can result in substantial savings for the public treasury. The virtues and vices of private provision of education continue to be intensely debated in many countries. However, Dominica's current fiscal situation implies that there i s no room for the expansion of the education system without tapping into private resources and achieving large efficiency savings. Clearly, increases in access to preschool and tertiary education cannot judiciously take place without further means of cost-recovery and/or private provision of education. 3. EDUCATION OUTCOMES INDOMINICA 6.9 In 2002 preschools enrolled about 7 out of 10 children aged 3 and 4. Preschool enrollment expanded from 2,100 students in 1994 to 2,600 in 1998, and then contracted to 2,144 in 2002. This pattern corresponds to the evolution of the cohort of 3-and-4-year-old children. Net enrollment in preschool education remained constant at about 70 percent from 1994 to 2000 (see Table 6.2). 102 Cycle Preschool Primary Exam Secondary Exam rl Tertiary Expected age 3~ 5 6 7 8 9 1 0 1 1 12 13 14 15 16 17-24 Grade K G1 G2 G3 G4 G5 G6 CEE 850 Enrollment 2,144 1.772 1,769 1,707 1.617 1,572 1.595 1,738 1.517 1.514 1.286 1,152 cxc Junior Secondary Exam Program (JSP) JSPl JSP2 JSP3 Jsp 304 425 540 exam Net enrollment 68% 99% 92%** 10% rate 6.10 Dominica, like most of the OECS countries with the exception of St. Vincent and the Grenadines, has achieved the Millennium Development Goal (MDG) of universal primary education. Dominica reached universal enrollment in primary education a decade ago. As Figure 6.1 shows, in FY1999/2000 the gross completion rate was 103 percent, with a completion rate of 107 percent for girls and 99 percent for boys. Figure 6.1. Gross Enrollment and Completion Rates in Primary Educationin Caribbean Countries, Selected Years (inpercentage) 120, 100 80 60 40 20 0 0 Gross EnrollmentRate (GER) Q Gross ComplehonRate (GCR) Note: The gross and completionrates for primary education correspond to the following years: Belize 2000/01, Dominica 1999/2000, DominicanRepublic 1999/2000,Grenada 2000/01, Guyana 1999/2000,Jamaica 1999/2000,St. Kitts and Nevis 2000/01, St. Lucia 2000/01, St. Vincent andthe Grenadines2000/01, and Trinidad andTobago 1999/2000. Source: MonitoringEducationPerformancein the Caribbean(forthcoming). 103 6.11 Secondary education underwent a remarkable expansion over the last decade, reaching a gross enrollment rate of 92 percent in 2000, and a completion rate of 56 percent (46 percent for boys and 65 percent for girls). As Figure 6.2 shows, in 1995, only three out of 10graduates from primary school were admitted to general secondary education. In2002, the transition rate increasedto 8 out of 10. This impressive development is due to two factors: (a) falling absolute numbers of graduates from primary education, which accounts for 55 percent of the increase in the transition rate; and (b) concerted efforts by the government and donors, the Department of InternationalDevelopment (DFID) and the World Bank, to increase capacity in secondary education. This expansion explains 45 percent of the increase. The progress in enrollment in secondary education in Dominica places the country among the top-performing Caribbean countries in enrollment from primary into secondary education. For comparison, in St. Vincent and the Grenadines only one in two graduates from primary education continues on to general secondary education. This expansion benefited especially the underperforming segments of the student population. In 1995, boys comprised only 40 percent of the intake into secondary education compared to 49 percent in 2002.37As Figure 6.3 shows, although the gross enrollment rate reached 91 percent in 1999, only 56 percent of the students completed secondary education. This low completion rate i s expected to increase in the next five years to 70 to 80 percent, when the new cohorts that have entered secondary education in 2000 and afterward will complete the educational cycle. Figure 6.2. Dominica Transition Rates from Primary to Secondary Education (as percentage) 100 90 I 199s 1996 1997 1998 1999 2000 2001 2002 Source: Author's calculations. 37A common problem in the Caribbean educational system is the large differencein learningbetween boys and girls. In most Caribbean countries girls consistently outperform boys. A case study of the reasons, implications,and policy solutionsfor this phenomenon is plannedfor St. Vincent and the Grenadinesas part of a pilot to mainstreamgender issues in Public Expenditure Reviews. 104 Figure 6.3. Gross Enrollment and Completion Rates inSecondary Education inCaribbean Countries, SelectedYears (inpercentage) 0 Gross EnrollmentRate (GER) 0 Gross ComplebonRate (CGR) 100 R Note: The gross andcompletionrates for primary educationcorrespondto the following years: Belize 2000/01, Dominica 1999/2000, DominicanRepublic 1999/2000,Grenada2000/01, Guyana 1999/2000,Jamaica 1999/2000,St. Kitts and Nevis 2000/01, St. Lucia 2000/01, St. Vincent and the Grenadines2000/01, and Trinidad andTobago 1999/2000. Source: Monitoring EducationPerformanceinthe Caribbean(forthcoming). 6.12 As secondary education expanded, a declining share of graduates from primary education continued into the vocationally oriented JSP, with a net enrollment rate of about only 20 percent in 2000. Given that the general secondary schools also have vocational subjects, the parallel existence of JSP and a general secondary program results in lower educational attainment for the pupils attending JSP, an expensive duplication of structures and an inefficient allocation of teacher staff. Moreover, only less than 60 percent of the students enrolled in JSP pass with grades Iand 11. Therefore, the government could consider accelerating the phasing out of the JSP, in order to provide five years of secondary education to all students.38 This would free up resources and make a more efficient use of the expensive facilities for vocational training. 6.13 The Dominican education system displays an exceptionally low drop-out rate in primary and secondary education, but a high repetition rate insecondary education. In 2002, the drops-out rates were 0.5 percent and 1.9 percent in primary education and secondary education, respectively. Repetition occurs at a moderate rate in primary education, 1 in 20 students repeats, but in secondary education, repetition has doubled from 6 percent in 1996 to 12 percent in 2002, simultaneously with the almost tripling in access to general secondary. The repetition rate i s the highest in the Caribbean. The doubling o f the repetition rate implies an increase (a) of 33 percent in the average number of school years 38 There were 21 schools with JSP in2001, and only 5 planned for school year 2002/03. 105 before graduation, (b) in the cost to the government of providing secondary education, and (c) in the internal inefficiency of the education system. The inability to instruct the students sufficiently stems from several factors, notably an insufficient learning foundation from primary education, and inadequate training among teachers. The latter i s addressed in paragraph 6.43. 6.14 In the Caribbean 3ox 6.1. The TourismBusinessin the Caribbean Examination Council exam (CXC-exam), a Caribbean-wide The tourism business drives a significant part of the Dominican exit exam for secondary xonomy; it accounts for about 18 percent of GDP, and has the iotential for future growth. In the tourism sector, client service education, Dominica compares s essential for competitiveness. For instance, the French hotel very favorably against other :hain Accor recently announced the closing of five hotels, a countries in the subregion. The :omplete pullout from the French Caribbean Islands, Martinique, CXC-exam measures learning ind Guadeloupe (Dominica's northern and southern neighbor outcomes for graduates of slands), as reported in the Financial Times (November 13, 2002). According to the investors, the reasons were "poor social secondary education, and hence Aimate" in general and "high operation costs and poor attitudes provides an indication of quality of o clients" in particular. instruction. On average, 72 percent 3iven the potential gains from improvements in client service, of students passed English and :an the government catalyze the neededchange in service culture math in 2000, while the regional and skills? Sector-specific service skills are important, but do average was 51 percent. Although, not belong to the set of fundamental skills that should be taught the results are encouraging, they in basic education. Service skills should hence be invested in by should be interpreted with caution ;ompanies and employees. From a national perspective, firms ;odd underinvest in training due to the risk of job separation because of a self-selection problem The same i s true for employees, because of lack of learning and comparability of the sample apportunities and no formal recognition of the acquired skill: across countries. Not all graduates Fromthe employer's side. participate in the CXC-exam, Chile sought to overcome these problems by the creation of i dropouts do not take the exam, and certification system called Competencia Laboral. The system in the case of Dominica, the exam provides learning guides, self-evaluation tests, verification oi i s open to individuals that did not learning outcomes, and certification of workers that meet tht requirements. A successful completion yields a diploma, whict participate in formal classroom enables the worker to document and market his or her skills. Fo teaching. the tourism sector, a worker can be certified in variou! professions, such as waitering. For participation, the program 6.15 In tertiary education, the focuses on low- and medium-income groups with prior community college currently experience in the sector. enrolls 850 students, an Although the program is still in a pilot phase, an impact is estimated 7.5 percent of the already observable. The litmus test for a certification process i s population of 18-to-24-year-olds. whether third parties attach a value to the obtained diploma. I n The government initiated a reform Chile, the experience shows that employers use the diploma in the hiring process. The public (low-cost) policy of certification in 2002 that consolidated the has hence led to an increased skill formation and higher value existing tertiary education added in productive sectors. Dominica could consider instituting institutions (nursingschool, teacher a similar process. (For more information, go to training institution, and the Clifton- www.competencialabora1.cl.) Dupigny College) into one administration. Taking into account enrollment at foreign tertiary institutions (UNESCO 2001), total enrollment increases to 1,180 students, and coverage to 10percent. This is half the average for the Latin America and the Caribbean region. This is an inadequate investment in the country's advanced human capital base. Further, the 106 pressure for entry into tertiary education i s likely to grow substantially in the next five years, as completion of secondary school rises from 50 percent to 70 to 80 percent. The challenge for policymakers will be to increase enrollment into tertiary education without further draining the public treasury. 6.16 Increasing learning opportunities for the adult population and recognizing skills learned outside of the formal schooling system can result in important human capital improvements, and raise labor productivity significantly. InDominica, contrary to other OECS countries, the secondary education exam (CXC-exam) i s open to both pupils in secondary school and to other individuals. More than 4 out of 10 examinees are individuals that did not participate in formal classroom teaching. Similar exams, or other types of certification mechanisms, could be introduced for skills needed in economic sectors that are critical for the Dominican economy. Box 6.1 explores the case of skills for the tourism sector. The certification mechanisms should follow the example of the CXC-exam and be sustained only by fees paid by the students. 6.17 In summary, the education system responded in a positive manner to society's pressures, but large challenges lie ahead. The efforts in secondary school resulted in noticeable improvements in access. Nevertheless, reform initiatives are necessary to address remaining unsatisfactory outcomes, such as: (a) the failure of preschool education to enroll a quarter of the 3-and-4-year-olds, (b) enrollment of 20 percent of graduates from primary to JSP, and (c) only 10 percent of a cohort attends tertiary education. Further, as the analysis will show, the system delivers education in a very expensive manner. c. USESOFEDUCATION EXPENDITURESDOMINICA IN 6.18 On average during FY1995/96-FY2001/02, Dominica spent about 6.6 percent of GDP per fiscal year on recurrent and capital education expenditures. From a regional perspective, this i s an exceptionally high share of public spending on education. Latin American and Caribbean countries invest on average under 5 percent of GDP. In a Caribbean context, Dominica i s equally above average. If this allocation were used efficiently, it would imply a sizable investment in the country's future growth potential. As this chapter will show, considerable efficiency gains and moderate fiscal savings can be made. 6.19 Education expenditures increased from about 5.3 percent of GDP in FY1995/96 to about 7 percent of GDP in FY2002/03. However, most of the increase was in capital expenditures. As Figure 6.4 shows, education capital expenditures increased from 0.2 percent of GDP in FY1995/96 to an average of 1.9 percent of GDP in FY1998/99 and FY1999/2000. Thereafter, education capital spending declined to 0.8 percent of GDP in FY2002/03. The increase occurred as a consequence of the construction of three secondary schools funded by grants and loans and the construction of the Windsor Park project. The three secondary schools were: (a) Northern Secondary Education School funded by the EU, (b) Grand Bay Secondary School financed by DFID, and (c) Castlebruce Secondary School financed by the World Bank. The Windsor Park project accounted for approximately 1.5 percent of GDP and, due to fiscal constraints, as mentioned, the government imposed a moratorium on construction, and only the foundation of the stadium was built. The 107 educational and economic merits of the Windsor Park project are questionable, and are discussed indetail inChapter 3. I Figure 6.4. Share of EducationExpendituresin GDP, Dominica, FY1995/99-FY2002/03 (aspercentage) 8 0% 7 0% 6 0% 5 0% 4 0% 3 0% 2 0% 10% 0 0% 1995196 1996197 1997198 1998199 199912000 2000101 2001102 2002103 Fiscal Year Note: Figuresfor all years except FY2001/02and FY2002/03are actual. In FY2001/02the figures are revised and for FY2002/03 they are estimates. Source: Author's calculations. 6.20 In real terms, recurrent education expenditures increased on average by 7 percent per fiscal year during FY1995/96-FY2002/03, and capital education expenditures increased by 39 percent per fiscal year. As Figure 6.5 shows, halting the rising trend of recurrent expenditures without deteriorating educational outcomes is one of the major policy challenges facing the government, and i s the topic of the next sections. Capital expenditures are expected to decline markedly, with the moratorium on the construction of the Windsor Park project and the completion of all the secondary schools neededfor universal enrollment. Expenditure by Subsector 6.21 During FY1995/96 to FY2002/03, secondary and tertiary education, in conjunction with administrative costs, has taken an increasing share of recurrent education expenditures. As Figure 6.6 shows, the recurrent education budget for FY2002/03 i s allocated in the following manner: preschool education, 0.2 percent; primary education-including J S P 4 4 percent; general secondary education, 34 percent; tertiary education-including teacher training-1 1 percent. This allocation is the result of four trends in budget allocations: (a) a reduction in the share o f recurrent expenditures to primary education, (b) an increase in secondary education, (c) an increase in tertiary education and teacher training, and (d) a rise in administrative costs. As will be argued below, the increase in administrative costs could be questioned, whereas (a) through (c) reflect the changing needs inthe education system. 108 Figure 6.5. EducationExpendituresinDominica, FY1995199-FY 2002103 (inmillions 1995EC$) 6 o o1 1995196 1996197 1991198 1998199 1999/2000 2000/01 2001102 2002/03 Fiscal Year ORecurrent BCapital Noret Figures for all years except FY2001/02 and FY2002/03 are actual. For FY2001/02 the figures are revised, and for FY2002/03 the figures are estimates. Source: Author's calculations. Figure 6.6. Dominica Recurrent Education Expenditures by Subsector, FY1995196-FY2002103 (aspercentage) 1I 100% 90% EO% 70% 60% 50% - 40% - 30% - 20% 10% 0% 1995/96 1996/97 1997/98 199W99 1999/2000 2000/01 2001/02 2002/03 Firsalycar w Pre-School Primary ~a Secondary nTertiary oTeacher training aa Administratlon Other Note: Figures for all years except NZM)1/02 and FY2002/03 are actual. For FY2001/02 the figures are revised, and for FY2002/03 they are estimates Source: Author's calculations. 109 6.22 The almost doubling in administrative costs-from 6.5 percent of recurrent education expenditures in FY1995/96 to 11.6 percent in FY2002/03-seems to be only partially warranted. The increase stems from greater emphasis on policy analysis, data collection, and implementation. Nevertheless, given (a) high fixed costs for policy development, (b) the small population, (c) similarity with other Caribbean countries, and (d) the existence o f a subregional education policy center, there exists a strong motive for subregional cooperation within policy development and technical administration. The administrative costs in Dominica exceed those of the average education systems in the Caribbean, which i s 10 percent, which in turn i s twice the Latin America average of 5 percent. Dominica could therefore emphasize subregional policy development and technical cooperation as a way to reduce the high administrative costs and stimulate quality improvements through international sharing of best practices. A 25 percent reduction in administrative costs would imply fiscal savings of about EC$1,000,000 (2.6 percent of recurrent education expenditures). 6.23 In accordance with Dominica's demographic and educational transition, resources were shifted from primary education to secondary education. From FY1995/96 to FY2002/03, the share of the education budget devoted to primary education decreased from 60 percent to 44 percent, while the share allocated to secondary education increased from 26 percent to 34 percent. This shift occurred as the demographic transition of the population reduced enrollment into primary education from 14,400 pupils in 1995 to 11,750 pupils in 2002. 6.24 Inaddition, an increase incost recovery could take place insecondary education through the introduction of minor fees that can be used for learning material and maintenance. Currently, there exists a de facto policy in public secondary schools of small fees, EC$5-EC$30 per semester. The government could formalize the de facto procedure by instituting a small fee, for instance EC$50 per year (US$18). The fiscal revenue associated with this small fee i s minor, EC$150,000 to EC$250,000 (0.5 percent of recurrent education expenditures). Nevertheless, the revenue could be a very important source for quality improvements if the fee was paid directly to the management o f each school. Each school principal has a small fund for learning materials and maintenance, and this fee could partially remedy the situation. Preconditions for introducing such a fee are: (a) transparency in the spending decision of each school management; and (b) the operation of a transparent and effective targeted public assistance system, whereby all poor households are exempted from payment. 6.25 The share of tertiary education in recurrent education expenditures increased from 4.7 percent inFY1995/96 to 7.6 percent inFY2002/03. Ina Latin American context, this share is significantly below average, but so i s coverage of tertiary education in Dominica. However, Dominica, like many middle- and low-income countries, will experience a surge in demand for tertiary education following expansion of secondary education. To accommodate a similar expansion, the government of Dominica could consider shifting, as it has already been doing since 2002, a larger part of the costs of tertiary education to its beneficiaries. 110 6.26 Dominica is the only OECS country that has introduced a tuition fee for nationals enrolled in tertiary education. The fee amounted to EC$250 per semester (US$93) in 2002, and i s payable directly to the community college. It constitutes an important source of revenue for the community college-13 percent of total expenditure on tertiary education. The additional resources have allowed the institution to increase enrollment and quality of teaching. From a regional perspective, users in Dominica pay around the level of the average Latin American student, which, however, i s low compared with fee levels in other regions of the world, notably North America. To accommodate the expected expansion of tertiary education, the government of Dominica could consider delegating the decision on the amount of tuition fee to the Community College. This could potentially allow for an increase in revenue to the institution. This would be in line with international best practice, and would improve the equity of public expenditure on education. A rise of fees should be accompanied by a suitable financing mechanism, such as an unsubsidized student loan scheme that would allow needy students to finance tuition, so qualified students from poor families would not be excluded from tertiary education. Equity of Spending 6.27 Dominica's public spending on primary and secondary school is progressive; that is, it is pro-poor. Spending on tertiary education is regressive. As Table 6.3 shows, based on the 2002 Poverty Assessment, the poorest quintile of Dominica's population received 26.9 percent and 20.1 percent o f public expenditure on primary and secondary education, respectively, but only 7.3 percent of tertiary education expenditures. In contrast, the richest quintile received 11.5 and 14.5 percent of primary and secondary education expenditures, respectively, but 32.5 percent o f tertiary education expenditures. The progressive spending on primary and secondary education i s driven by demographics and high enrollment rates. Moreover, these results support the recommended policy of an increasing transfer of tertiary education costs to the students. Quintile Primary Secondary Tertiary Enrollment Expenditure Enrollment Expenditure Enrollment Expenditure Poorest 126 26.9 61 20.1 9 7.3 2 110 23.5 77 25.3 24 19.5 3 98 20.9 64 21.1 21 17.1 4 81 17.3 58 19.1 29 23.6 Richest I 54 11.5 44 14.5 40 32.5 6.28 The most important educational barrier for low-income families is limited access to secondary and tertiary education?' International evidence overwhelmingly demonstrates that such barriers to entry into education disproportionately hurt the educational 39Unfortunately, the Poverty Assessment conducted recently by the Dominican government does not have an education module that allows evaluation of who are the main beneficiaries of the school feeding program, textbooks. and other education subsidies. 111 chances of children from poor and uneducated families. Consequently, equitable policies would: (a) complete the provision of universal secondary education and improve the quality of primary and secondary schools to assure all school leavers learn adequately enough to find meaningful employment in the labor market; and (b) overcome barriers in access to early childhood development and tertiary education. 6.29 The school feeding program could be an important incentive for school attendance of children from low-income families. Recently, however, it was downscaled. The school feeding program covered 22 schools and 40 daycare centers in 1996, but was downscaled and covered only 7 primary schools in 2002. Fiscal costs were reduced through an EC$l per day fee, which can be waived by school authorities, and reliance on parents to volunteer. It i s estimated that student contributions to the program cover less than 15 percent of total costs. No formal impact analysis has been undertaken. However, qualitative evidence suggests that the school feeding program promotes attendance and learning outcomes. Participating schools were selected based on the Ministry's assessment of which schools serve the poorest populations, and estimated distances that children walk to school. Policymakers could consider increasing funding, or at least safeguarding expenditures to this important low-cost program to cover additional schools in low-income areas, or exempt poor household from paying the daily fee. 6.30 The Education Trust Fund (ETF) provides assistance to needy students for transportation, uniforms, books, and external examination fees. Total expenditures on the trust fund were EC$28,869 in FY2000/01.40 Approximately 350 students received assistance from the ETF in FY2001/02. Over 95 percent o f beneficiaries attended secondary school. Based on school enrollments, the incidence o f poverty among children, and the number of ETFbeneficiaries, we estimate that the ETF provides benefits to less than 1in 10 needy students. Clearly, a scaling up of this program could improve educational attainment among poor households. However, there are no clearly established criteria for eligibility. Each board member of ETF decides eligibility in his or her district. Hence, a scaling up of the fund would require a more systematic selection o f beneficiaries to benefit the target group. 6.31 The merit-based scholarship program seems to be an expenditure line that benefits the relatively well off without an apparent impact on school attainment. In FY2001/02, the Ministry of Education budgeted EC$522,600 for scholarships awarded to the best-performing graduates from primary education. Anecdotal evidence, which i s backed by international evidence, suggests that students from high-income families in general win merit-based scholarships. It i s doubtful whether the program increases enrollment and completion of secondary education. The recipient student group would in most cases have attended and performed well in the absence of a scholarship. Thus, the public rationale for funding a large merit-based scholarship program i s weak. Policymakers could consider either immediately reducing the merit-based scholarships, or phasing out current commitments without granting new scholarships. Alternatively, the government could change the awarding criteria to need based, as is its stated purpose inthe government budget. 40Inaddition, the fundreceives contributions and raises money through community-based fundraising. 112 Efficiency of Spending 6.32 Across the subsectors, there is an urgent need to rethink the allocation of recurrent expenditures. Currently, salaries amount to almost all recurrent expenditures, 92 percent in FY2001/02. This i s particularly pronounced for primary education, where 96 percent of all expenditures are remuneration of personnel. This leaves inadequate resources to other crucially important learning inputs, such as writing paper, textbooks for poor students, and school maintenance. This skewed allocation of expenditure toward salaries unnecessarily reduces the effectiveness of public resources, and results in inefficiently low learning. 6.33 The pupil-teacher ratio is the key concept for enhancing efficiency of the education system. The low pupil-teacher ratios in Dominica make its education system among the most expensive in the Caribbean, after St. Kitts and Nevis, and strongly indicate room for savings. The costs of provision by the education system depend on two factors: (a) the average salary of the teacher staff, and (b) the number of pupils taught by each teacher. The latter i s measured by the pupil-teacher ratio. A low pupil-teacher ratio implies a greater number of teachers are employed, and the more expensive i s provision of education. Figure 6.7 shows just how closely this indicator i s linked to unit costs of education in the Caribbean. InDominica each teacher inprimaryeducation instructs on average 20 pupils. For secondary education, the ratio i s 18. This compares to 25 and 18, respectively, for the average Caribbean country. Hence, primary schools, in particular, have smaller classes and more expensive delivery than the average Caribbean country, which suggests that teacher deployment and class sizes can be managed in a more efficient manner. As developed below, the potential savings from an increase in the pupil-teacher ratios are non- negligible. Figure6.7. Pupil-Teacher Ratio and Costsof Educationinthe Caribbean Region,2000 (inUS$) PRIMARY EDUCATION SECONDARY EDUCATION lWO i S K N m i SKU `20? 1030 @ SL $BE SVG +GR +SVG r M tJA ,GR Y 200 l illi ~ 100i tDR SDR 0 J 3 1 15 20 21 30 35 40 0 5 10 I S 20 25 30 Pupil-Teacher Ratio Pupil-TeacherRatio Source: "Monitoring EducationalPerformance in the Caribbean," World Bank (forthcoming) 113 6.34 During 1998-2002,the pupil-teacher ratio in primary education declined due to falling cohorts and additional hiring. Raising the pupil-teacher ratio to an average of 25 pupils would imply moderate fiscal savings. In 1998, each primary teacher instructed on average 25 pupils. Enrollment into primary education declined from 13,700 in 1998 to 11,750 students in 2002. To have kept the same level of cost-efficiency as in 1995, the primary teacher staff would have had to contract by 141 teachers. However, between 1998 and 2002, the primary teacher corps expanded by 31 teachers. Since the demographic evolution of the country dictates that the absolute enrollment into primary education will continue to decline in the near future, it i s important for cost-efficiency that no new hiring takes place. Moreover, if the pupil-teacher ratio for primary education were raised to the average Caribbean level, 25 pupils per teacher in primary education, the required teacher staff in primary education would decline from 642 teachers in 2002 to 470. Shedding of the 172redundant teacher positions, either through attrition or layoffs, would generate savings of about EC$5,000,000 (12.9 percent of recurrent education expenditure^).^' 6.35 Inefficient use of Figure 6.8. Exploiting Economies of Scale inPrimary Education teachers predominantly takes place at smaller primary schools. A merging of schools based on school-mapping 351 exercises, and examining busing .-0 30 - options for children from small 26 c 25 - villages and future demand for J L schooling, would allow the .-Ir420 - a system to exploit the economies 15 - of scale in the education system, and reduce cost, while 10 - increasing quality. Figure 6.8 shows that larger schools manage more 7I ' teacher staff 0 4 0 effectively, especially the very 100 200 300 400 500 600 700 800 900 small primary schools-defined Pupilsin EachPrimary School as schools with less than 50 pupils-which have a very low pupil-teacher ratio of below 15. These pupils cost the public coffers twice as much in teacher resources per student compared to larger schools. These small schools are located in rural areas and experience a falling intake because of declining birthrates, urbanization, and migration. These demographic factors are expected to continue at least in the next decade; thus, the costs per student will continue to rise in the absence of policy action. 6.36 Based on exam results in primary education, lower pupil-teacher ratios do not necessarily increase learning outcomes in Dominica.42 It i s often argued that additional 41The savings do not include the cost of separation, including compensation packages. 42Regression analysis using the size of the school, the pupil-teacher ratio, the share of teacher staff with teacher training, and ownership of the school (private, publicly assisted, and government school) explains 38 percent, of the variation in the primary schools' average passing rate at the common entrance exam. Only the pupil- 114 teachers in the classroom increase student learning. However, in Dominica a higher pupil- teacher ratio i s statistically related to increased performance of pupils in primary education. The positive effect of a higher pupil-teacher ratio arises from a symbiosis of several related factors. The successful primary schools are larger urban schools with a higher pupil-teacher ratio, better-educated staff and, importantly, pupils from better-educated families. Unobservable factors such as motivation of teachers, involvement of parents in school management, and school leadership are likely to be important underlying factors for the finding as well. The result suggests that an increase in the pupil-teacher ratio could take place without decreasing quality of education. 6.37 Furthermore, efforts to raise the pupil-teacher ratio in public secondary education from the current 18.7 to 23 would increase the teacher capacity enough to provide teaching for 1,000 additional pupils. This corresponds to 40 percent of what was needed to provide universal secondary education in 2002.43 Alternatively, the increased efficiency would reduce the required teacher force by 90 teachers. A reduction of the teacher staff by this number could provide fiscal savings of EC$2,600,000 (6.7 percent of the recurrent education expenditure). 6.38 In addition, a different design for subsidies to publicly assisted secondary schools could increase the pupil-teacher ratio in secondary education and reduce the cost for the g0vernment.4~ Currently, the government pays for all teacher salaries plus a supplement (EC$45) per pupil to seven nongovernmental secondary schools (religiously affiliated). Given the funding method, the assisted schools have little incentive to be cost- efficient with teacher management. The government-assisted schools have the lowest pupil- teacher ratio, 15.6, which compares to 18.7 and 20.8 for government schools and the only private secondary school, respectively. The government could consider either: (a) Changingfunding of publicly assisted schools toper-pupilfunding, where the school receives a fixed amount per student (possibly an amount sufficient to cover teacher salaries for a pupil-teacher ratio similar to the one inpublic schools), or (b) Signing contracts with assisted schools to increase the pupil-teacher ratio to the level inpublic schools over a certain period of time. teacher ratio is statistically significant, and with a positive impact. Hence, the analysis would suggest that a higher pupil-teacher ratio leads to a better outcome. However, this could be an impact of omitted factors, such as the socioeconomic status of the children attending the school. With regard to the other variables, a higher share of certified teachers is related to a higher passing rate, but not statistically significantly so. The size of the school is insignificantly related to a higher passing rate, and private and publicly assisted schools are insignificantly related to a higher passing rate. 43An increase in enrolment into general secondary education would imply a decrease in enrolment into the JSP, and would thereby liberate teacher capacity at the JSP level. 44The same subsidies policy applies to publicly assistedprimary schools. However, the pupil-teacher ratio for public primary schools was 19.3 to 1 in 2001 compared to 24.0 to 1 for the five publicly assisted primary schools. 115 Raising the pupil-teacher ratio to the level in public schools would liberate enough teachers to instruct 460 new pupils, 30 percent of what was needed in 2002 to reach universal secondary education. Alternatively, it could reduce the needed number of teachers by 25, which would amount to a saving of approximately EC$700,000 (1.7 percent of recurrent education expenditures). 6.39 The government could consider several other policy initiatives to increase the pupil-teacher ratio in primary and secondary education. The following policies would, ifrightly managed, leadto higher efficiency andquality improvements: 0 Instituting a curriculum reform of secondary education. A common core curriculum for the first three forms in secondary education would reduce the excessive offering of subjects of a specialized nature, such as typewriting, auto mechanics, and metalworking, and would lead to larger classes. Further, such a reform could be combined with a modernization of offered courses that equip the pupil with more skills that are relevant for today's labor market. Besides improving efficiency through larger classes, a common curriculum within OECS could lead to further efficiency gains through common regional exams during the course of secondary education, collective textbook purchases, and sharing of teachers. Such a reform i s under way in an OECS context, but implementation has been lacking. 0 Mandating of a minimum floor on class size at the primary and secondary level. Many schools, in particular at the secondary level, offer subjects with low demand, leading to small class sizes and high costs. Mandating a minimum class size of, for example, 15-a rule that i s in place in many developed countries-would increase efficiency. 0 Rewarding well-petfiomzing public schools. The Ministry of Education could provide incentives for each school to deliver education more efficiently, possibly through establishing a contract with each school management, in which the school commits itself to increasing the pupil-teacher ratio, and as a reward retains a part of the savings. For such decentralization to function effectively, international experience shows that local and national school management often needs strengthening. 6.40 Besides the quality of instruction and number of pupils taught, the salary level of the teacher staff is the most important factor for efficient provision of education. The typical primary school teacher in Dominica receives EC$29,400 per year, which i s slightly above the OECS average of EC$27,800. When compared to the income level of each country, primary teachers in Dominica are paid relatively more than their colleagues from other OECS countries. Dominican teachers receive a salary corresponding to 303 percent of GDP per capita, while the average OECS teacher earns 252 percent of GDP per capita.45 45The same pattern arises when the salary of the average employee of the Ministry of Education in Dominica is compared to the salary of colleagues in other OECS countries. The absolute wage level in Dominica i s slightly above the OECS average (EC$33,700 compared to EC$30,600), while the wage relative to GDP per capita i s noticeably higher for Dominicans than elsewhere in the OECS, (348 percent of GDP per capita compared to 274 percent). 116 Importantly, this analysis does not necessarily indicate that teachers per se are more highly paid than other public employees with similar skill level. The higher salary i s a result of higher salaries to all public employees inDominica compared to other OECS countries. 6.4 1 Quality of instruction, and especially teacher qualification, is key for increasing effectiveness of learning and efficiency of investment. A major determinant of quality of instruction i s the education and training of the teacher corps. The Dominican teacher corps belongs to the least certified in the subregion. As Figure 6.9 shows, in 1999, 6 out of 10 teachers in primary education were certified. In secondary education, only 3 out of 10 teachers were ~ e r t i f i e d .Moreover, teacher quality in secondary education has deteriorated ~ ~ in the last five years as the secondary education system expanded without available trained teachers. The implication in terms of quality of instruction i s expected to be significant; international evidence shows that low teacher qualification leads to deteriorating learning outcomes. Furthermore, the lack of adequate training leads to lower efficiency since repetition increases and teachers become less flexible. Figure 6.9. Share of Trained Teachers in Primary and Secondary School, Dominica, 1999 (inpercentage) - PRIMARY EDUCATION SECONDARY EDUCATION 100 90 80 70 60 50 40 30 20 10 0 Source: "Monitoring Educational Performance inthe Caribbean," World Bank (forthcoming). 6.42 The government could consider stipulating that in the future only certified teachers will qualify for open positions in order to improve and strengthen teacher qualifications. This would prevent teaching qualifications from further deteriorating. Further, to seriously address the shortfall of training, a major retraining program of the existing teaching staff could take place. Provided efficiency gains take place in the system, time would be generated for teachers to attend retraining courses. This, however, presupposes public or private investments inteacher training courses. 46More recent domestic statistics (from 2000/01) indicate that the ratios are even lower; only 5 out of 10 teachers in primary education and 2.3 out of 10 teachers in secondary education are certified. This is the lowest percentage among Caribbean countries. 117 7. SOCIALPROTECTIONPROGRAMSINDOMINICA 7.1 Social protection programs are public interventions to assist the most vulnerable groups of society, with the objective of ensuring a minimumlevel of welfare for individuals, households, and comm~nities.~~ The most common social protection tools to address the social risks of vulnerable groups are social insurance and social assistance. e Social insurance consists of mechanisms that pool social risks across population groups, such as pension, unemployment, and health insurance programs. e Social assistance programs are designed to help the poor, vulnerable individuals and households cope with temporary or chronic poverty by providing income support and access to basic services. Social assistance programs aim to address the specific needs of groups at risk, and include a variety o f programs such as cash and in-kindtransfers, subsidies, workfare, public works, housing programs, and social funds. 7.2 Dominica implements a range of social insurance and social assistance programs that, in combination, address (or attempt to address) the critical risks and vulnerabilities faced by the population. Most of the social assistance programs in Dominica emphasize risk-coping strategies; that is, programs to help households cope with the effects of risk (especially poverty), rather than prevention and mitigation strategies. In particular, the existing social assistance programs are not well poised to respond to income volatility originated either from significant economic downturns or natural disasters. 7.3 Social protection expenditures in Dominica increased from 4.2 percent of GDP in FY1995/96 to about 5.6 percent of GDP in FY2000/01, or slightly above the average for the Latin America and the Caribbean region.48 Nevertheless, in moving forward, the challenge for Dominica i s to: (a) strengthen the impact of existing social assistance programs through improvements in targeting, monitoring, and implementation capacity, and expand coverage of certain programs while taking into account existing budget constraint; (b) increase the capacity of social assistance programs, such as the Road Maintenance and the Education Trust Fund, to be scaled up or down in the face of economic downturns or natural disasters; and (c) expand coverage and ensure the financial sustainability of the Dominica Social Security Fund. To address these challenges Dominica will have to reallocate resources among programs and possibly rethink existing and/or pipeline projects, especially community-based programs. 47 Itis important to note at the outset that very little social protectioneconomic sector work had been undertaken in Dominica prior to this report. Five different ministries and over 10 different departments within these ministries administer social protection programs. Program-level budget information is not available for all programs. This made programsdifficult to identify, and collectionof complete time series data on expenditures and beneficiaries impossible. Despite these constraints, considerable progress was made in identifying social protectionprograms, expenditures, and key issues. 48 These figures include recurrent and capital expenditures. In terms of only recurrent expenditures, social protectionexpendituresincreasedfrom 4.2 percent of GDP inFY1995/96to 5.1 percent of GDP in FY2000/01. 119 7.4 This chapter identifies vulnerable groups, describes social protection strategy and programs in Dominica, reviews social assistance and social insurance spending, and suggests policy recommendations for strengthening social protection in Dominica. The main recommendations to strengthensocial protection programs can be summarized as follows: On Social Protection Reforms 0 Reallocate resources from programs with low returns (mainly training programs) to social protection programs that protect investments in children's education, and make these benefits conditional on school attendance. This reallocation should be in the context of a fiscally sustainable economic framework and may need to come from a rethinking of existing donor-fundedprojects. 0 Reorient the Road Maintenance Program, a public works initiative, to a program that can be quickly scaled up and down when external shocks occur by better targeting and more rotation of workers. 0 Improve management information systems and program monitoring and evaluation to promote greater fiscal accountability, and to allow programs to be scaled up when external shocks occurred. Ensure the long-term solvency of the Dominica Social Security Fund-to include reforms recommended by the International Labour Organization (LO) Actuarial Report, agreements identified in the 2002 Memorandum of Understanding between the Dominican government and the Dominican Social Security (DSS) scheme, and other reform strategies to be identified as part of analytic work being financed by the World Bank. On InstitutionalReforms 0 Develop a social protection strategy. This would require sector wide planning; that is, all ministries and agencies responsible for implementing social protection programs would come together to identify social protection priorities and coordinated strategies to address these priorities. 0 Develop a more effective and efficient system of identifying beneficiaries, including mechanisms to target households and communities (poverty maps), particularly for Public Assistance, the Education Trust Fund, and the Road Maintenance Program.49 The system should be easy to administer and should apply uniform and objective criteria, screening instruments, and procedures for approval. 49The wage rate for the Road MaintenanceProgram should not be based on the market wage, but on estimates of some minimum living standard below the market wage. This will ensure that only the most needy participate. Poverty maps may be useful in targeting geographic areas for implementationof the public works program. 120 0 Implement regular household surveys to assess poverty levels and the incidence and coverage of social safety net programs that could be used to assist inthe development of targeting systems. 0 Enhance fiscal accountability through introduction of program budgeting for social protection programs not currently identified as programs in the budget; for example, social public assistance and the road maintenance program. A. DOMINICASOCLALPROTECTIONSTRATEGYANDPROGRAMS Social Protection Strategy inDominica 7.5 The Government of Dominica's stated social protection strategy is to pursue economic growth, complemented by pro-poor policies and programs that reduce inequalities and promote human capital development. The government i s committed to provide safety nets that will enable people to meet their consumption needs and that promote income security, employment, and income generation. Resources have been mobilized to enable poorer communities to access basic social services, create short-term employment and training opportunities, and to empower women for community leadership and employment. A related goal is to ensure better targeting of limitedresources so that they reach the poorest and most vulnerable. Parallel social sector strategies focus on expanding access to and equity o f secondary education and improving education outcomes, expanding access to tertiary education, and ensuring accessto and strengthening health services. 7.6 Social protection programs are primarily targeted to indigent households, to needy children inthe formal school system, and to the unemployed. As part of its overall social protection strategy, the government provides free primary health care for all citizens at health clinics and district hospitals and fee waivers for public assistance beneficiaries at the Princess Margaret Hospital. The Social Security system targets employed and self-employed people, although coverage of the self-employed i s generally low. 7.7 To the extent that they are poor, banana workers and former banana workers are covered by the existing social protection programs, including the European Union (EU) Social Protection Program-s ecifically targeted to banana workers who were affected by changes in the industry! It is important to note here that, as indicated in the Social Protection Review prepared by the World Bank, the main consequences in poverty and employment of the loss of banana trade may have already occurred. This i s evidenced by the fact that the incidence of poverty in the banana industry i s relatively low compared to other economic sectors. This may be an indication that most o f the inefficient banana producers have already exited the sector, and those that remain are in a relatively stronger position. j0The loss of banana trade preferences has had a dramatic effect on the Dominican economy. Production of bananas for export declined from 60,000 metric tons to 70,000 metric tons in the late 1980s-early 1990s to under 30,000 tons in 2000, with the consequent decline in employment (the number of banana producers dropped from over 6,600 in 1990 to less than 1,200 in early 2002). 121 7.8 This stated, the existing social protection strategy does not address critical vulnerabilities. Low and volatile incomes are a source of risk in Dominica. However, social protection programs are not well poised to address income volatility and transient spells of poverty. Dominica would benefit from a social protection strategy (and programs) that address income volatility and loss of income due to short-term unemployment, while also focusing on providing income support to the poorest and most vulnerable; building and protecting human capital and human capital investments; and insuringagainst the loss of income inold age due to retirement, illness, or disability. 7.9 Further, coordination among programs is weak. There is a significant degree of overlap in existing programs and beneficiary groups. The most dramatic examples are in the area of skills training, where different ministries and agencies undertake similar types of skills training for overlapping constituencies. The same i s true with respect to community- based infrastructure development programs, such as the Caribbean Development Bank (CDB)-funded Basic Needs Trust Fund (BNTF), the Eastern Caribbean-funded Social Protection Program, and the pipeline EC Social Investment Fund (SF)-all of which have similar approaches and project menus. A clear strategy to guide the design and development of social protection interventions i s needed. This will require enhanced data collection, analysis, and planning capacity. 7.10 The absence of a coordinated social protection strategy makes fiscal prioritization impossible. This i s reflected in the budgeting process during which budgets are largely determined by previous budget allocations rather than in response to changing needs and priorities. Social Protection Programs and Expenditures inDominica: FY2000/01 7.11 Dominica has implementedthe following social protection programs: Social insurance programs: Dominica Social Security (DSS) scheme. 0 Social assistance programs: a means-tested income transfer program5'; a labor- intensive road maintenance program; education-based programs (a School Feeding Program and Education Trust Fund for needy students); skills training and adult education programs; donor-funded, labor-intensive, community-based infrastructure development activities, including the Basic Needs Trust Fund (BNTF) andthe Social ProtectionProgram; and others.j2 The key social protection programs are summarized in Table 7.1. "A meanstestisanappraisalofhouseholdincomeandassetstodetermine whether ahouseholdqualifies for program benefits. '2 Housing initiatives were not included because they are not targeted to the poorest or most vulnerable. Rental units, originally targeted to low-income households, are no longer rented to low-income households. The Dominica government is considering divesting these properties. The government's other housing program provides subsidized mortgages. However, the poor are excluded by virtue of mortgage, income, and collateral requirements. 122 Social Insurance 7.12 The Dominica Social Security (DSS) scheme accounts for the largest share of total social protection expenditures (61.3 percent). The scheme i s funded by contributions and returns to investments; covers private, public, and self-employed workers; and provides pension, disability, survivors, health, sickness, maternity, employment injury, and redundancy benefits (the latter available only to private sector workers). However, many self-employed workers do not contribute to the scheme, and those that participate declare low earnings. In addition, agricultural workers are generally not covered. The number of active insured increased from 14,400 in 1991 to over 21,000 in 1999, but declined to 18,200 in 2000-probably as the result of increasing unemployment. During the same period, the number of pensioners more than doubled-from approximately 1,000 to 2,350. 7.13 During 1995-2002, the Dominican government was in arrears to the DSS. Between 1995 and 2002, the government made few payments to the Social Security Fund. A Memorandum of Understanding, which seeks to arrest the accumulation of arrears and minimize future ones, was signed between the Dominican government and DSS in September 2002. EC$16 million of the arrears was cleared in December 2002, b y issuing a zero coupon bond with an interest rate of 7 percent and a 10-year maturity period. The remaining stock of arrears outstanding as of end-March 2003 was EC$36 million. However, the government did not stay current in its payments, and additional payment arrears of EC$2.7 million accumulated in the first quarter of 2003. The Government resumed regular payments to DSS in July 2003. The outstanding stock of arrears will be cleared by a barter arrangement for land and shares in the National Commercial Bank and other statutory agencies. 7.14 While DSS is financially stable in the short and medium term, its long-term sustainability is at risk due to several factors. First, regular contributions by the Government need to be ensured. Second, DSS needs to strengthen its asset management strategy to enhance returns on investments, notably b y diversifying its investments regionally and internationally. Third, the population of Dominica i s aging. In 1999, there were about 5 people supporting each old age pensioner, whereas by 2050, there will be only 2.5 people of working age for every old-age person. Under status quo conditions, total revenue (contributions plus investment earnings) will exceed expenditure until 2011. After 2011, a growing part of the accumulated reserve will have to be used, and the reserve will be exhausted by 2023. The 2001 L O actuarial report recommended that contributions for the long-term benefit branch increase from the current rate of 6 percent to 18 percent by 2040. The report also recommended a gradual increase in the retirement age from 60 to 65 over a 20-year period (see Box 7.1). Finally, a possible restructuring o f government's obligations to DSS, currently under discussion, could substantially reduce the value of DSS's reserves. Thus affecting its medium to long-term financial sustainability. 53A follow-up review of the status of Dominica's pension system was conducted by ajoint IMF-World Bank mission in June 20205. The review updated information on the status of the pension system in light of a possible restructuring of DSS's holdings of Government debt under discussion and proposed options for pensionreforms aimed at improving the accountability and efficiency of the pension system. 123 Social Assistance Programs 7.15 Labor market interventions, including employment and training programs, absorb the second-largest share of total social protection expenditures (21 percent). The Road Maintenance Program, a public works program, accounts for the lion's share of that- approximately EC$5.8 million, or 14 percent of total government expenditures on social protection for the wage portion of the program. The program employs approximately 900 workers annually to clean roadways and do light road maintenance work for periods of four to six weeks. Workers are paid the government minimum wage, which i s below the prevailing wage rate for unskilled labor. The Ministry of Communications, Works and Housing estimates that less than 10 percent of workers are female. Workers can be reemployed so that, in many ways, the program functions as a seasonal employment program rather than a revolving employment program. Parliamentary Representatives can nominate workers, and there i s a general perception-among the poor and among ministry personnel- that political party affiliation often affects whether a worker gets hired. Better targeting (through the use of some minimum living standard below the market wage), increased rotation of workers, and implementation of systems to ensure that the program can be scaled up and down in the face of external shocks, would strengthen the social protection function of the program. 7.16 Entry-level skills training programs targeted to different vulnerable groups are implemented by two ministries (and several divisions within one of those ministries) with considerable overlap intarget groups and subjects taught. This diversity of training programs, including programs implemented by the Youth Division, Adult Education Division, Women's Bureau, and Cooperatives Division, with funding from the government and from the BNTF and the Social Protection Program, means that limited training resources-both human and financial-are spread among multiple programs. In addition, there i s inadequate labor market information to guide program development, a lack of standardized content, and uncertain labor market outcomes. All programs provide entry- level reemployment skills training-despite the fact that international experience suggests that the economic returns to this type of training are generally low (Dar and Tzannotos 1999). Resources ought to be reallocated from training programs to social protection programs that protect investments in children's education, and these benefits should be made conditional on school attendance. 7.17 Income Support Programs represent the third-largest share in total social protection expenditures (10.2 percent). Public Assistance and financial support to NGOs are the primary income support programs for people who are poor, and account for approximately 8.2 percent o f social protection expenditures (EC$33.3 million). Eligibility i s subject to a means test, home visit, and an annual review. However, criteria for eligibility are not documented, and eligibility reviews are not conducted annually. In addition, fiscal constraints have resulted in irregular payment of benefits. There i s a waiting list for eligible beneficiaries. Information systems are inadequate and, as a result, the Ministry of Community Development and Gender Affairs could not produce basic operational information on how many people are served each year, or a breakdown of beneficiaries by age, gender, or other characteristics. The Ministry estimates that there were 1,750 people on the beneficiary list as of September 2002. Based on data from the Survey of Living 124 Conditions, we estimate that only 10 percent of indigent households, and 8 percent of households that are poor but not indigent, receive support from Public Assistance. Box 7.1. Ensuring the Sustainability and Expanding Coverage of the Dominica Social Security Scheme The Dominica Social Security (DSS) scheme i s administered by the Dominica Social Security Corporation and Funded through contributions and returns on investments. It provides pension, disability, survivors, health, sickness, maternity, employment injury, and redundancy benefits. The latter are available only to private sector workers. Over half of all DSS beneficiaries receive benefits under the short-term sickness component, while one-quarter of beneficiaries claim under the long-term Age and Retirement component. The DSS absorbs approximately 60 percent of social protection spending. With almost 10,000beneficiaries in FY2000/01,Social Security reachesthe largest number of beneficiaries of any of Dominica's social protection programs. Approximately 80 percent of the working-age population under 60 contributes to the DSS. DSS is financially stable in the short and medium term. It is expected to generate surpluses for at least the next 10 years, and has enough reservesto finance benefits for another 10 years after that. However, challenges exist on several fronts, including increasing the sustainability and expanding coverage of DSS. To ensure the long-term sustainability of the DSS, regular contributions by the Government will need to remain current. This will require a clear and sustained commitment to addressing the issue of Government arrears and to avoiding future arrears. Reform of basic parameters, such as contributions rates, the statutory retirement age, and/or income replacement rates may also be needed to ensure the financial sustainability of the system. The DSS also needs to strengthen asset management strategies to enhance returns on investments and financial stability in the face of high exposure to exogenous economic shocks and natural disasters. Coverage of the employed and self-employed is compulsory; however, an estimated 20 percent of the working-age population does not contribute. There are several groups that are affected by coverage problems: low-income employees (working in the formal or informal urban sectors), rural workers (employed or self- employed), and the urban self-employed who typically have low incomes. Employees that are not covered by the system are usually victims of inefficient collection and enforcement schemes. Policy recommendations vary for each of these groups, and their efficacy i s variable. Expanding coverage to reach near-universal coverage will not be easy, and will require careful analysis to detect who are the uncovered workers and why they are currently excluded. Expanding coverage must be weighted against the costs of increasing coverage. Source: World Bank, "Dominica Social Protection Review," July 2003. I 7.18 Community-basedprograms,includingthe Basic Needs Trust Fund (BNTF)-a CDB-funded community-based intervention-accounted for 4.5 percent of social protection expenditures in FY2000/01. BNTF provides funding for community infrastructure, larger infrastructure, skills training, and maintenance projects. The European Union i s providing support for a Social Protection Program with EC$3.4 million in funding from STABEX 96/97.54 The two-year project commenced in 2002 and i s not reflected in Table 7.I.Program components include skills training, a short-term employment program, and a credit facility for rural enterprise development. These programs are geographically targeted; however, targeting in the absence of reliable poverty maps i s difficult. A social investment fund project i s inthe EUpipeline. 54 The Stable Export Earnings Program (STABEX) is the EU's compensatory finance scheme set up under LomC Ito stabilize the export earnings of the African, Caribbean and Pacific (ACP) states, and to remedy the effects of the instability of export earnings. 125 Table7.1. SocialProtectionProgramsandExpendituresinDominica,FY2000/01 (inEC$) % inTotal m Recurrent Capital Total SocialProtection 70 Spending of GDP SOCIALINSURANCE 24.835.000 , , 61.31 3.41 Social Security 24,835,000 61.31 3.41 SOCIAL ASSISTANCE 12,211,161 3,458,831 15,669,992 38.69 2.15 IncomeSupportPrograms 4,122,914 4,122,914 10.18 0.57 Welfare-Administration and Supervision 448,257 448,257 1.11 0.06 Welfare-Public Assistance and Subventions to NGOs 3,336,946 3,336,946 8.24 0.46 Foster Care (Child Welfare) 337,711 337,711 0.83 0.05 Education-BasedPrograms 28,869 1,289,885 1,318,754 3.26 0.18 EducationTrust Fund 28,869 28,869 0.07 0.00 Needy Students Program (fundedby Taiwan, China) 1,289,885 1,289,885 3.18 0.18 SchoolFeeding Health-BasedPrograms 42,796 42,796 0.11 0.01 Overseas MedicalCare 42,796 42,796 0.11 0.01 Labor MarketPrograms 8,016,582 365,887 8,382,469 20.69 1.15 Youth Development 928,918 928,918 2.29 0.13 Youth Skills Training 277,331 277,33 1 0.68 0.04 Youth Environmental Service COT 56,804 56,804 0.14 0.01 Youth Technology Resource Centers 31,752 31,752 0.08 0.00 Adult Education 535,063 535,063 1.32 0.07 RoadMaintenance(short-term wages only) 5,809,534 5,809,534 14.34 0.80 Workshop for the Blind (Blind Welfare) 86,014 86,014 0.21 0.01 Gender Affairs 268,345 268,345 0.66 0.04 Cooperatives 388,708 388,708 0.96 0.05 Community-BasedPrograms 1,803,059 1,803,059 4.45 0.25 BNTFIV (CDB ) 1,803,059 1,803,059 4.45 0.25 TOTAL SOCIAL PROTECTION 40,504,992 100.00 5.56 Source: Author's calculations. 7.19 Despite the need to protect and promote human capital development, education- based social protection programs account for the smallest share of social protection expenditures (3.3 percent). The Education Trust Fund (ETF) and the Needy Students Program (NSP) funded in FY2000/01 b y Taiwan, China are means-tested programs that help to defray out-of-pocket educational expenditures. The ETF is targeted to secondary and tertiary students, while the NSP provides assistance to needy students at all academic levels. The ETF provided benefits to approximately 350 students in FY2001/02 at a cost of EC$28,869. An earlier report estimates that the Fundcovers an estimated 50 percent or less of the school costs. The NSP assisted 100 students that year. A seven-member Board of Trustees (appointed by the Minister of Education) administers the ETF and NSP. Each 126 Board member conducts an initial review and determines eligibility of applicants from his or her district. The Board reviews and gives final approval of these recommendations. The ETFand NSPhave separate application forms. Students cannot benefit from both programs, although households, which must apply separately for each school child, may receive ETF benefits for one child and NSP benefits for another child. Recipients reapply each year. Administrative systems are weak and there i s a perception that the programs stigmatize poor students and that this limits applications from needy students (World Bank 2001). Also, decisions on who receives benefits and who does not are not sufficiently transparent. 7.20 The School Feeding Program operates in 7 primary schools-down from 40 pre- primary and 22 primary schools during the period of World Food Program Support (1991-96).55 Schools were selected based on poverty estimates and distance that children walk to school. Between 900 and 1,100 students are served annually. Students are charged EC$1 per meal and are provided with a snack or hot meal. The fee may be waived, however; targeting criteria differ among schools, and there is concern that children may be stigmatized if they are unableto pay. Qualitativeevidence from the Ministry of Educationassessmentof the program and from discussions with parents and educators suggests that the School Feeding Program promotes attendance (particularly afternoon attendance). Further research i s needed to confirm these perceptions and to provide guidance on whether or not to expand the School FeedingProgram (Ministry of Education 2002). Planning and Executing Capacity for Social Protection Programs 7.21 The lack of coordinated efforts across ministries and departments has resulted in overlaps and duplication. Moreover, despite the fact that different programs target the same general population, they employ different systems for identifying beneficiaries and suffer from a lack of clearly established criteria for eligibility. There i s a need to strengthen the planning and executing capacity of these programs and to put in place mechanisms for their fiscal prioritization. Finally, existing systems would not support the rapid scaling up of programs in the face of shocks-either economic shocks or those resulting from natural disasters. 7.22 Management informationsystems for the Public Assistance and Education Trust Fundare inadequate. The Welfare Division cannot produce simple operating statistics. In the case of the Education Trust Fund and the Needy Students Program, data collection systems and procedures are dependent upon individuals and, when personnel change, both institutional memory and records may be lost. In general, there is a lack of clearly documented policies and procedures for social assistance programs. In their absence, 55 It i sdifficult to determine the actual cost of the School Feeding Program (SFP). The Dominicagovernment's Estimates Book for FY2000/01 reports that nothing was allocated to the SFP. However, the program was operational in that year. In FY2001/02, EC$26,040 was allocated-although the SFP Status Report for FY2001/02 lists expendituresof EC$59,40&with 94 percent going toward the purchase of food. This does not include the coordinators' salary or related administrative, storage, and delivery costs. In addition, World Food Program (WFP) project funds were used to advance purchase of a three-year supply of foodstuffs, and some of these supplies still remained in FY2001/02. Government estimates of expenditures and special account disbursements do not reflect the value of remaining supplies purchased with WFP funds, or contributions from the Christian Children's Fund for the feeding program at one school. 127 program implementation i s left to the discretion of managers and to political influences that reduce transparency. 7.23 Information on poor and vulnerable groups is not available on a regular basis. Household surveys are conducted intermittently and the capacity to implement and analyze survey data i s weak. The 2002 Country Poverty Assessment, funded by the CDB, i s the first quantitative assessment of poverty and provides very useful information. However, regularly updated poverty data will be required for effective planning, monitoring, and evaluation of social protection programs. 7.24 Regional initiatives to build information systems, data collection, and analysis capacity could address the human resource constraints and reduce the overall costs of development and implementation of monitoring and information systems. Inadequate MIS systems and capacity constraints with respect to data collection, monitoring, and evaluation are common across the OECS. Regional approaches to MIS development and monitoring and evaluation capacity building would address the human resources constraints in individual countries and would be more cost-effective than developing systems on a country-by-country basis. B. COMPOSITION AND EVOLUTION SOCIALPROTECTIONSPENDINGINDOMINICA, OF FY1995/96-FY2000/01 7.25 It is difficult to disaggregate the amount of the national budget that is allocated to social protection?6 Social protection programs cut across ministries and are integrated into a wide variety of types of programs and projects. Several of the programs traditionally defined as social protection programs (for example, school feeding programs and scholarship programs for needy students) are also represented in the budgets o f the education sector. Despite these difficulties, this report has attempted to estimate public expenditures on social protection. In FY2000/01, public expenditures on social protection (including social security) were estimated to be EC$40.5 million. 7.26 DuringFY1995/96-FY2000/01, social protection expenditures increasedfrom 4.2 percent of GDP to approximately 5.6 percent of GDP. As Figure 7.1 shows, this increase was distributed in equal parts by recurrent capital expenditures and social security. On average, in this period social protection expenditures were about 5 percent of GDP. This compares favorably with expenditures on social protection of 4.5 percent of GDP in Latin 56I t should be pointed out that the expenditures for the individualprograms for FY2000/01 and the expenditures for the time series analysis are slightly different. That is because, when looking at programs for a single year, we were able to get more program-specific data, while for the time series analysis we had to rely on spending data for ministerial departments that traditionally house social protection programs. Various ministries had portfolio responsibility for these departments in different years. Efforts have been made to track expenditures for departments no matter what ministry the department was housed in. Expenditure data do not include central ministry administration, because it was impossible to reliably apportion central ministry administrative costs to the different departments. Program-level expenditure data on programs housed in the Ministry of Health (feeding programs, fee waivers, and so forth) and the Ministry of Education (school feeding and student welfare programs) are also reported. Spending on microenterprise development, low-income housing, and contributory and non-contributory social security are reported. Given the number of agencies and programs involved, time series data could be provided for only FY1995/9&FY2000/01. 128 America and the Caribbean, and 1 to 2 percent in sub-Saharan Africa and South Asia. But it i s below the 11 to 12 percent of GDP spent on social protection in Western Europe, North America, and East and Central Europe.57 Figure 7.1. Social Protection Expendituresin Dominica (aspercent of GDP) 500 - T / 400 - 300 - Socid Rotechon Rotechon Eqendlturcs 2 00 100 0 00 95/96 96/97 97198 98/99 99/00 OOiOl 0 Recurrent SocialAssistance El Capital SocialAssistance 0 SocialInsurance ( Socd Security) Source Author's calculations 7.27 Inaddition, social protection expendituresfluctuated betweena minimumof 3.4 percent of total Central Government (CG) expenditures in FY1995196 and a maximum of 4.5 percent in FY1999/2000. During this period, the share of social insurance of total social protection expenditures remained above 60 percent. Although a moderate increase in the share of social assistance occurred during this period, those increases were largely the result of increases in capital expenditures, particularly externally financed capital projects, and not in response to income volatility. Indeed, contrary to expectations, the highest share of social protection expenditures in CG expenditures was achieved in FY1999/2000 when GDP increased by 1.5 percent, and not in FY2000/01, during which GDP decreased by 1.4 percent. 7.28 During FY1995/96-FY2000/01, expenditures on social protection increased in real terms by 8.6 percent per fiscal year. As Figure 7.2 shows, real expenditures on Social 57 I t is important to note that included are expenditures on administration of the Dominica Social Security (DSS) scheme and on benefit payments. However, the DSS i s funded from contributions to the scheme and not from the consolidated fund. As a result, although it is treated as public expenditure, it is very different from social assistance expenditures that are funded from general revenues. The picture is further complicated by the fact that the Government did not make any payments to Social Security for public sector employees between 1995 and 2002. 129 Security increased by 5.9 percent compared to a 9.4 percent per fiscal year in social assistance expenditures, the latter driven largely as a result of the increases in capital expenditures. The real increases in social insurance expenditures reflect the maturation of the system, with increasing numbersof workers becoming eligible for benefits. Figure 7.2. Social Protection Expenditures in Dominica (in millions EC$ 1995) 40 0 - --- - - - ~ - ______-_- 3 5 0 - 3 0 0 - 2 5 0 - social 2 0 0 - Protechon Emendures 15 0 10 0 5 0 0 0 95/9 6 96/97 97/98 98/99 99/00 O O i O l 0 Recurrent SocialAssistance [FdCapital SocialAssistance 0 Socialhsurance (Social Secunty) Source. Author's calculations. c. SOCIAL PROTECTION AND RISKMANAGEMENT INDOMINICA 7.29 According to the Country Poverty Assessment, approximately 39 percent of the population is poor.58 Further, approximately 15 percent o f the population i s indigent; that is, they do not have sufficient resources to meet their basic dietary needs.59 Although, the Carib Community accounts for only 7 percent of the poor, about 50 percent of the Carib Community i s indigent and 70 percent is poor. Poverty and indigence rates and other related indicators are shown in Table 7.2. As indicated, 55 percent o f the poor i s under age 25. Forty-percent of women and 38 percent of men are poor. People who were not educated beyond the primary level typically head poor households. Housing conditions among the poor are of concern. The poor often do not have indoor kitchen or bathing facilities or access to toilets or pit latrines. However, only less than 2 percent of the population lives below the 58The purchasing power parity conversion factor usedfor the 2002 calculation was 1.65. 59 In 2002, the adult equivalent poverty and indigence lines were equal to EC$3,400 and EC$2,000, respectively. 130 poverty line of one-dollar purchasing power parity (PPP) a day, and less than 3 percent lives below the poverty line of two-dollars PPP a day. 7.30 Poverty and income volatility are not the only sources of vulnerability in Dominica. We examine other sources of vulnerability by identifying risks indicators throughout the life cycle (Table 7.3). Information for this table was gathered through interviews with central ministry and project staff and a review of secondary documents. The matrix also identifies indicators not related to age and existing social protection programs and requirements. This exercise should be treated as preliminary. Statistical analysis of these risk indicators to determine prevalence, incidence, correlations with income, and household vulnerability could provide useful insights to help identify and prioritize appropriate social protection interventions, but will require updated and improved survey data. Table 7.2. Incidence of Poverty, Indigence, and other Related Indicators, Dominica, 2002 (percentage) Poor Individuals 39.0 Indigent Individuals 15.0 Poor Households 29.0 Indigent Households 11.0 Poor Under 25 55.0 Poor 65+ 8.0 Females as a Percent of All Poor 50.0 Incidenceof Poverty Among Females 40.0 Incidenceof Poverty Among Males 38.0 HouseholdHeads in ElementaryOccupations 25 .o HouseholdHeadNot EducatedBeyond Primary School 90.0 Householdswith Long-Term Sick 10.0 Poor with Outdoor Kitchens 41.0 No Toilet or Pit Latrine 29.0 Poor with Outdoor Baths 35.0 Source: Dominica Country Poverty Assessment and calculations by World Bank Staff. 7.31 Programs could do better at promoting human capital development initiatives. International experience suggests that a conditional cash transfer with benefits linked to health clinic or school attendance would be ideal, but given current capacity, it i s not at all clear that this should be the short-term priority. Alternative strategies, particularly the Education Trust Fund, could be redesigned to better promote human capital development initiatives-particularly makingbenefits conditioned on school attendance. Although it is in its incipiency, the St. Lucia model in which a grant from Taiwan, China was used to capitalize a fund with the interest earned used to finance student welfare programs, may serve as a model to be replicated. 131 Table 7.3. Dominica RiskIndicator! ige Cohort, Existing Programs, and Suggestec ,wentiom 4ge Group Risk Indicator Existing Program Risk Preventioflitigation Risk Coping {irth Low birth weight and blinistry of Healthprograms ncrease pre- and postnatal :CD programsto ensure cognitive 0 4 nutrition-relateddisorders ttendance,maternaland child ndphysicaldevelopment ealthprograms Not attendingearly :inancia1support (ETF) childhoodand/or preschool 'rivate pre-primaryprogram ncrease access to early onditional on schoolattendance mdpublic kindergarten hildhood development(ECD) xograms irogram i-9 Irregularschool attendance School Feeding ixpandSchoolFeeding amongpoor Veedy Students Program 3etter-targeted educationsubsidies vith benefits conditionalon school Ittendance Numtion-relateddisorders Ministry of Healthprograms 10-14 Not enrolledin secondary Initiativesto expand access tc nitiatives to expandaccess to 3etter-targetededucationsubsidies school secondary school ,econdary school with benefits conditionalon school ittendance Irregularschoolattendance EducationTrust Fund amongpoor Needy Students Program 15-24 Low level of humancapital Training. abor-intensive growth temedial education programs development Youth Skills Gender Division ,ifelong learningopportunities Adult Education hat are responsive to market CooperativesDivision lemands Basic Need5 Trust Fund Unemployment RoadMaintenanceProgram Labor-intensive public works STEP Program Teenage pregnancy, Programsin various ministrie Healthy lifestyle promotion reen mother and substance abuse substanceabuse among programs: HFLE, treatment programs young men 25-59 Low level of human capital Training. Labor-intensivegrowth Remedialeducationprograms development Youth Skills Gender Division Lifelong learningopportunities Adult Education that are responsive to market CooperativesDivision demands Basic Needs Trust Fund Unemployment RoadMaintenanceProgram Labor-intensivepublic works Shon-TermEmployment Program(STEP Program) No pensionor disability Expandcoverage of social insurance Dominica SocialSecurity security system Chronic diseases Healthy Lifestyles Promotion Targetedfee waivers Ministry of Healthprograms Programs 60 and over No pensioncoverage Dominica Social Security Expandcoverage of social Public Assistance Public Assistance security system Community-basedprograms Chronic diseases, disability. Ministry of Healthprograms social isolation Governmentassistance to N( Homes for the Elderly RisksNot Hurricanes DisasterPreparednessStrate, Housingrepair programs Post-humcane: Expandedpublic Relatedto Disaster Preparedness Strategies works, public assistance, Age Substandardhousing temporary food. water, shelter, Preventioflreatment Progri HIVlAIDS Education Program reconstructionprograms HIVIAIDS EducatioflevelopmenU Expandcoverage of Training Programs EducationDevelopmenU Disability Training Programs Source: Bar on author's discussionswith iual ministry and programsti Ireview of secondarydocuments. 132 BIBLIOGRAPHY Alesina, A., R. Hausman, R. Homes, and E. Stein. 1996. "Budget Institutions and Fiscal Performance in Latin America." National Bureau of Economic Research, Working Paper No. 5586. Cambridge, MA, May. Alesina, A., and R. Perotti. 1996. "Fiscal Adjustments in OECD Countries: Composition and Macroeconomic Effects." National Bureau of Economic Research, Working Paper No. 5730. Cambridge, MA, August. Alesina, A., and R. Wacziarg. 1997. "Openness, Country Size and the Government." National Bureau of Economic Research, Working Paper No. 6024. Cambridge, MA, April. Alt, J., and R. Lowry. 1994. "Divided Government and Budget Deficits: Evidence from the States." American Political Science Review, December. Blanchard, 0. 1990. "Suggestions for a New Set of Fiscal Indicators." Organisation for Economic Co-operation and Development, Department of Economics and Statistics, Working Papers No. 79. Paris, April. Blanchard, O., J. C. Chouraqui, and R. Hagemann. 1990. "Sustainability of Fiscal Policy: New Answers to an Old Question." OECD Economic Studies, (15) Organisation for Economic Co-operation and Development. Paris, Autumn. Buiter,Willem H. 1993. "Measurement of the Public Sector Deficit and Its Implications for Policy Evaluation and Design." In Mario I. Blejer and Adrienne Cheasty, eds., How to Measure the Fiscal Deficit. Washington, D.C.: International Monetary Fund. Campos, E., and S. Pradhan. 1996. "Budgetary Institutions and Expenditure Outcomes: Binding Governments to Fiscal Performance." Policy Research Working Paper No. 1646. Washington, D.C.: World Bank. Chouraqui, J. C., R. Hagemann, and N. Sartor. 1990. "Indicators of Fiscal Policy: A Reexamination." Organisation for Economic Co-operation and Development, Departmentof Economics and Statistics, Working PaperNo. 78. Paris. Crucini, M. 1997. "Country Size and Economic Fluctuations." Review of International Economics, May, :204-20. Dar, Amit, and Zafiris Tzannotos. 1999. "Active Labor Market Programs: A Review of the Evidence from Evaluations." Washington, D.C.: Work Bank. De Ferranti, D., G. Perry, I.Gill, and L. Serven. 2000. Securing Our Future in a Global Economy. Washington, D.C.: World Bank. Dos Santos, P. 2003. "Dominica: Strengthening the Inland Revenue Division." International Monetary Fund, Fiscal Affairs Department. Washington, D.C., January. Eastern Caribbean Central Bank. 2001. "Developing Sustainable Debt Indicators for the ECCB." Mimeo, July. . "The Targeted Public Sector Savings in the Member Countries of the ECCB." Mimeo, undated. 133 Eichengreen, B. 1992. "Should the Maastricht Treaty be Saved?" Princeton Studies in International Finance, No. 74. Princeton: Princeton University. Gleich, H. 2003. "Budget Institutions and Fiscal Performance in Central and Eastern EuropeanCountries." EuropeanCentral Bank,Working Paper No. 215, February. Head, A. 1995. "Country Size, Aggregate Fluctuations, and International Risk Sharing." Canadian Journal of Economics, November, :1096-1 119. ILO (International Labour Organization). 2001. "Seventh Actuarial Review of the Dominica Social Security Fund." Paris. Jayasuriya, R., and Q.Wodon. 2003. "Efficiency in Reachingthe MillenniumDevelopment Goals in LatinAmerica andthe Caribbean." Washington, D.C.: World Bank. Jones, M., P. Sanguinetti, and M. Tomassi. 1999. "Politics, Institutions, and Pubic-Sector Spending in the Argentine Provinces." InJames Poterba andJurgen von Hagen, eds., Fiscal Institutions and Fiscal Performance. Chicago: University of Chicago Press. Lao-Araya, K. 1997. "The Fiscal Constitution of a Developing Country: The Case of Thailand." Ph.D. Thesis. Bloomington: IndianaUniversity. Ministry of Education, Sports and Youth Affairs. 2002. "Status Report of School Feeding Program, 2OOl-2OO2." Project ManagementUnit. Mukherjee, Amitabha, Giulio de Tommaso, and Salvatore Schiavo-Campo. 1997. "Government Employment and Pay: A Global and Regional Perspective." Policy ResearchWorking Paper. Washington, D.C.: World Bank. Nunberg, B., and D. Lindauer. (Eds.) 1994. "Rehabilitating Government: Pay and Employment Reform in Africa." Washington. D.C.: World Bank. OECS (Organization of Eastern Caribbean States) Secretariat. 2002. "OECS Human Development Report 2002." St. Lucia. Oxford Policy Management. 2002. "Strengthening Debt Management." Report. Oxford, May. Perotti, R. 1996. "Fiscal Consolidation in Europe: Composition Matters." American Economic Review, Papers and Proceedings, May, :105-10. Poterba, James. 1994. "State Responses to Fiscal Crises: The Effects of Budgetary Institutions andPolitics." Journal of Political Economy, August, :779-821. Rodrik, D. 1996. "Why Do More Open Economies Have Bigger Governments?" National Bureauof Economic Research,Working PaperNo. 5537. Cambridge, MA, April. Stein, E., E. Talvi, and A. Grisanti. 1999. "Institutional Arrangements and Fiscal Performance: The Latin American Experience." In James Poterba and Jurgen von Hagen, eds., Fiscal Institutions and Fiscal Performance. Chicago: University of Chicago Press. Stotsky, J., C. Dieterich, and J. Bristow. 2003. "Dominica: Modernization of the Tax System." International Monetary Fund, Fiscal Affairs Department. Washington, D.C.. March. 134 UNESCO. 2001. "Good Neighbours: Caribbean Students at the Tertiary Level of Education", Institute for Statistics, Montreal, Quebec. Von Hagen, J. 1992. "Budgeting Procedures and Fiscal Performance in the European Communities." European Communities Commission, Directorate General for Economic andFinancial Affairs, Economic Papers No. 96, October. Von Hagen, J., and I.Harden. 1995. "Budget Processes and Commitment to Fiscal Discipline.'' European Economic Review, April, :771-79. World Bank. 2001. "Pre-Social Assessment Secondary Education Project for Dominica, Saint JSitts andNevis and St. Lucia." Draft. Washington, D.C. . 2002. "OECS Country FinancialAccountability Assessment." Washington, D.C. . 2002. "OECS Institutional Organizational Capacity Review." Washington, D.C. .2003. "Caribbean Economic Overview 2002: Macroeconomic Volatility, Household Vulnerability, and Institutional andPolicy Responses." Washington, D.C. . 2003. "Dominica Social ProtectionReview." July. . "Monitoring Educational Performanceinthe Caribbean." Washington, D.C. Zee, H., J. Stotsky, and E. Ley. 2002. "Tax Incentives for Business Investment: A Primer for Policymakers in Developing Countries." World Development, September, :1497- 1516. 135 m m N '? 3 2m r: 4 0 0 0 0 0 0 2 009 - \o 3 2 2 2 N 3 3'7- a m N N 0 2m 0 0 0 ?i 2N 2c;' : 0 0 r: 0 z '? P 0I :2 N 0 m m - ? v i r 4 c\1 3 3 E: m a 4 r?? N O c ; ' t a- - R m '9 e 0 r? - N '? m 0 \o m 3 -0 0 09 ? 3 - 3' I t I N * m 0 v! 0 m 9 "- m 2 2 : v! x PI m -6 0 " a m P ?v! 3w N 9 N m x r- a m - - N v l m 2 c? 9 2; 0 L - I , m 2 m m c?" - 0 d m d 2 : w m 2 9" r - N ;z r - w T T " :2 : N N m d C i r 2 m m r":? ? ? 0 2 x w 3 h r: CQ u7 u7 V M k .3 Mu k8 .I 43 Y YV Y m Y a m 8 0 8 2 2 % 0 0 - 0 0 8 3 0 2 % 0 0 - 0 s 0 8 2 0 0 2 0 8 0 9 0 8 0 0 8 0 ? 3 8 017 80 2 0 0 8 3 -c! 0 sr-0 a .-.e 0 0 8 0 2 a 3 0 0 2 s: 0 2 9n h u: ti 2a, rA UE Q