Report No. 26152-MU Mauritius The New Economic Agenda and Fiscal Sustainability June 30, 2004 Poverty Reduction and Economic Management 1 Southern Africa Africa Region Document of the World Bank TABLE OF CONTENTS INTRODUCTIONAND MAINFINDINGS ....................................................................................... i 1. PUBLIC EXPENDITURESHISTORICAL IN PERSPECTIVE .................................................. 1 DEVELOPMENT POLICIES ........................................................................................................ 1 PUBLIC FINANCETRENDSTHROUGH THEPARASTATAL THENEED CHANGE......................................................................................................... SECTOR..................................................................................................... 2000 ............................................................................. 37 FOR 9 2. THENEWECONOMIC AGENDA ...................................................................................... 11 THENEW ECONOMIC AGENDA: PROGRAM ........................................................................... 11 THECOSTOFTHE NEA AND ITSFISCALFRAMEWORK ......................................................... 15 IMPLEMENTATIONOF THE NEA TO DATE ............................................................................. 16 FUTURE IMPLEMENTATIONOF THE NEA .............................................................................. 20 3. A SUSTAINABLE FISCALENVELOPE ............................................................................... 21 OBJECTIVES OF THE ANALYSIS ............................................................................................. 21 STARTING CONDITIONS ........................................................................................................ 21 HowDEBT CANINCREASE ................................................................................................... 23 A LONG-TERMTARGET FOR CONSOLIDATEDPUBLIC SECTORDEBT ................................... 25 A FISCALFRAMEWORK THE CENTRALGOVERNMENT FOR ................................................... 28 CONCLUSION ........................................................................................................................ 30 4. POLICYOPTIONS ............................................................................................................. 31 EXPENDITURES ..................................................................................................................... 3.1 REVENUES ............................................................................................................................ 41 DEBTMANAGEMENT ............................................................................................................ 42 CONCLUSION........................................................................................................................ 43 5. REFORMING ECONOMIC MANAGEMENT INMAURITIUS .............................................. 45 ECONOMICMANAGEMENT INSTITUTIONS IN MAURITIUS ..................................................... 46 ANALYSISOFTHEMAURITIANECONOMICMANAGEMENT FRAMEWORK ............................. 47 RECOMMENDATIONS ............................................................................................................ 59 ANNEX 1 GOVERNMENTEXPENDITURESAND PARASTATALS: SELECTEDDATA ..................61 ANNEX 2 A HIGHER DEBT SCENARIOFOR THE CENTRAL GOVERNMENT AND THE COSTSOF ADJUSTMENT ........................................................................................................... 70 ANNEX 3 INCORPORATINGFISCALRISKSINTO THE FISCALFRAMEWORK ............................ 74 ANNEX 4 AN INSTITUTIONAL FRAMEWORK FORMANAGING FISCAL RISK ............................ 77 TABLES Table 1.1 Snapshot of Some Direct Costs of the Welfare System(percent of GDP) 3 Table 1.2 Summary of Budgetary Central Govemment Balances(percent of GDP) 5 Table 1.3 ParastatalSubsidies, Capital Transfers, Lending, and Govemment Repayments 8 Table 2.1 New Economic Agenda Objectives 12 Table 2.2 InvestmentsinNEA Flagship Sectors: Expendituresand Indicative Commitments (% 13 of GDP) Table 2.3 Illustrative NEA Fiscal Framework 16 Table 2.4 NEA Fiscal Indicators 2001/2 and2002/3: Budget and Actual (as % of GDP) 17 Table 2.5 NEA ImplementationExpenditure Composition 17 Table 2.6 ParastatalSubsidies,Capital Transfers, Lending, and GovernmentRepaymentsfor 18 2001/02 Table 2.7 Change inCentral GovernmentDebt:Decomposition ("3 of GDP) 19 Table 3.1 Public SectorDebt End-June2002 (Percentof GDP) 22 Table 3.2 Target Public Sector Debt Scenarios- Simulation of Debt ServiceRatios (4.5 percent 26 economic growth assumed) Table 3.3 Average 10-YearPrimary BalanceNecessary to ReduceTotal Public Sector Debtfrom 26 77 percentof GDP to 60 percentof GDP in 10years Table 3.4 Illustrative Average 10-YearPrimary Balanceso f the Public SectorNecessaryto 27 ReducePublic SectorDebt from 77 percent of GDP to 60 percentof GDP in 10 years Table 3.5 Indicative NEA Fiscal Framework (average GDP growth 4.5 percent) A Sustainable 28 ProgrammableExpenditure Envelope Table 3.6 ProjectedDebt of Selected Parastatals-Highest Riskto Central Govemment; Stock of 30 debt as ofJune 2002 Table 4.1 Breakdown o fthe Indicative SustainableExpenditure Envelope (2003/4-5/6) (As %of 32 GDP) Table 4.2 Capital Expenditure Projectionsfor a SustainableFiscal Envelope (as a percentof 33 GDP) Table 4.3 Illustrative Trade-offs Betweenand Within Sectors 40 CHARTS Chart 3.1 Public Sect Debt Stocks 29 Chart 4.1 Cost of Energy Sources and Average SalesPrice o f Electricity (1992=100) 34 Box 3.1 Why should parastataldebt be included in an analysiso f sustainability o f central 23 govemment expenditures? Box 5.1 Incorporating fiscal risk into the MTEF 49 Box 5.2 Public-Private Initiatives and Public Expenditure Reform inMauritius and South 51 Africa Box 5.3 Pilot Public SectorReforms inthe Ministry o f Civil ServiceAffairs 52 Box 5.4 Financeand Economic Development Ministries-Together or Apart? 57 FIGURES Figure 1.1 Revenue, Expenditures, Deficits as % GDP 3 Figure 1.2 Expenditures 4 Figure 1.3 Functional Expenditures 4 Figure 1.4 Recurrent Social Expendituresas a % of GDP, by sector 6 Figure 1.5 Social Capital Expenditures as a % o f GDP, by sector 6 Figure 1.6 Central Govemment Debt Stock 6 Figure 1.7 Central Govemment Total Debt by Maturity as % GDP 7 Figure 1.8 Guarantees & Central GovernmentDeficits 9 Figure 5.1 Featureso f a GeneralizedMTEF 47 PREFACE Mauritius has been heraldedatop economic performer, as it has arduously earnedits place among the uppermiddle income economies. Most notable has beenMauritius' particular approach to development: it's home grown approachto growth andits heavy emphasis on sharing the gains to the country's prosperity are being studiedworldwide: Today Mauritius' competitivenessis being challenged by globalization and other idiosyncratic factors. The public debate is about how to reorient the economy andto redefine the social contract betweengovernment, the private sector, and labor, while ensuring sustainability and growth. The Economic Agenda for the NewMillennium(or New Economic Agenda), introduced by the Government in2001, i s aresponseto this debate. This Agenda put education first, andinitiated investment inseveral sectorsto modernize Mauritius' infrastructure. It also initiated a reform ofthe public sector, to make itmore responsive to the country's challenges. This report contributes to the ongoing debate inMauritius about the role ofthe public sector. The report has a very simple "bottom-line" message, that Mauritius should be cautious about its large deficits inperiods when its growthprospects are below the historical average. The deficits canunderminethe very foundation o fMauritiuspast development success, macroeconomic stability, anddiminishthe country's credibility. The report suggests that a more systematicprioritizationof expenditures is needed, with more open debate about where public money is placed, andwhat eachMauritiangets from it. The report also says that, Mauritius needs to putback at the center, the policy debateswithingovernment and in the public arena, so costs andbenefits of current policies are discussed, as are trade-offs, at this time whenefficiency gainsare critically important. Already, during the 2003/4 fiscal year Government has actedto address many of the concerns raised inthis report andto implementpublic managementreforms, particularly those discussed inchapter 5, that have long been inthe planning. It is a privilege for the Bankto participate inthe dialogue on economic managementinMauritius, atthis stage ofthe country's development. ACKNOWLEDGEMENTS This Public ExpenditureReview (PER) was preparedbythe World Bankwiththe close collaboration ofthe GovernmentofMauritius. The Government team was leadby Mr.Krishnanand Guptar, Financial Secretary, MinistryofFinance andEconomicDevelopment. Other membersoftheteamincludeMr. Guy Wong So, Director-General, Ministry of Economic Development, Financial Services and Corporate Affairs, staffs the ministriesof Economic Development and of Financeprior to their recent merger, andstaff of the Bank o fMauritius. The preparation team benefited from extensive discussions with the Hon. Prime Minister, Paul Bdrenger, the Hon. Deputy Prime Minister Pravind Jugnauth, andthe Hon. Sushi1Khushiram, Minister of Industry, Financial Services, and CorporateAffairs. The team also benefitedfrom extensive discussionswith severalpublic enterprises andthe private sector. The Bankteam consistedofMaria-TeresaBenito-Spinetto (macro modeling), GuenterHeidenhof(public management), Christos Kostopoulos (Team Leader for the report and fiscal sustainability analysis), Liz Muggeridge (mediumterm expenditureframework), Hana Polackova-Brixi(contingent liabilities andparastatals), Liz Muggeridge (mediumterm expenditure framework), andFahrettinYagci (fiscal sustainability analysis andparastatals). Document processing was adeptly managedby Ckcile Wodon andEmmaFabienne Raharinjanahary. Jesko Hentschel advisedthe team extensivelyon the content and organization o f the document. Peer Reviewers for the report were Charles Humphreys, William Dorotinsky, and Michael Stevens. Other staffthat provided valuable comments to the report are Willem van Eeghen, DelfinGo, Jose Leandro (EU), PeterMoll, and participants inthe AFTP1andAfrica Region reviews. The report was preparedunderthe following managementteam: Jesko Hentschel was Cluster Leader for the Poverty ReductionandEconomic Management (PREM) team basedin Madagascar. Philippe L e Houerouwas Sector Manager through the GreenCover stage, and provided valuable guidance on all aspects ofthe report; Emmanuel Akpa i s Sector Manager at the Grey Cover stage; Hafez Ghanemi s Country Director, and has provided valuable comments on substantive issues coveredinthe report; and Alan Gelb i s Chief Economist of the Africa Region. INTRODUCTIONAND MAINFINDINGS 1. InJune 2001,the GovernmentofMauritiuspresentedanewfive-year program, the New Economic Agenda, to the nation. The Agenda came at a time when Mauritius faced a numberofworrying developments-lowered growth prospectsdueto unfavorable world market conditions and the phasing out of preferentialtrade agreements, unemployment at an all-time high of 10percent, and a rising fiscal deficits andpublic debt. The Agenda aims to transform the Mauritianeconomy through important structural reforms and an ambitious investment program ineducation, health, transport, andinfrastructure. Implementation of the Agenda would imply doubling public capital investmentsuntil2005/2006. .. 11. Thispublic expenditure report examineswhether the Agenda canbe financed ina sustainable way. The main finding of this report is that the Agenda's expenditure program needs to be phasedandprioritized. This necessity stems from the large public sector debt, which will require areduction of fiscal deficits to around 3 percent of GDP inthe medium term. PUBLIC EXPENDITURESINHISTORICAL PERSPECTIVE iii. Mauritiusisheraldedasanexampleof successful development.Mauritius experienced average annual GDP growth o f 5.5 percent over the past 20 years. Its per capita income tripledto US$3,800 andpoverty has fallen to about 10 percent o f the population, Mauritius has accomplished these gains through (i) political stability, (ii) fiscal good management, (iii) aggressive export promotionand extensive support ofthe private sector, and (iv) a generous social welfare system. Mauritius' approachto development included significant public sector intervention inthe economy, comprising the regulation of a core set of commodity prices, selective controls on commodity trade, and engagement into a number of productive sectors, most importantly sugar. iv. But at the turn of the millennium, severaltrendsemerged,putting into question the country's ability to grow at high rates. Three important trends became apparent at the end of the 1990s. First, the traditional growthpoles were runningout o f steam. Boththe sugar andtextiles sectorshadbenefitedfrom importantpreferential trade arrangementswhichhad secured exportsto the European market. But these preferenceswere beginning to be threatened-andthe process of phasing them out startedto expose Mauritian industries, with their comparatively lowproductivity, to world competition. Second, the welfare modelhad to berethought. Unemployment rates reached 10percent inthe year 2000, a large numberof secondary school students left school without their diploma, and it became increasingly clear that outlays for social welfare programs -most importantly the universal and civil service pensions-would likely to be unsustainableinthe future. Third, fiscal pressure mounted. In 2000/01, the fiscal deficit amountedto 5.7 percent o f GDP, a sharp increase from an average of 3 percentinthe 1990s. The highdeficit was causedbytotal government expenditures remaining stable while tax revenues hadbeeneroded. At the same time, the composition of i expenditureshadchangedinfavor oftransfers, includingthose to parastatalenterprises, leaving less room for investment expenditure. THENEWECONOMICAGENDA v. Mauritiusformulated a reform agendain response to the challenges. InSeptember 2000, Mauritians voted for a coalition government with a strong mandate for change. The new government draftedthe NewEconomic Agenda (NEA), afive-year reform framework to develop Mauritius into a high-income, high-tech service andknowledge economy. The NEA was introduced inJune 2001; the first year of its implementation was fiscal year 2001/2. vi. TheNEA outlines the need to heavily invest in three broad areas of the Mauritian economy.. .First, the Agenda focusesonimprovingthe environment for the private sector, particularly giventhe existing challengesto Mauritius' traditional export markets. Second, the Agenda outlines heavy investmentsinthe social sectors inorder to improve skills of the working population, better meetthe needs of an aging population, andmore effectively help the most marginalized insociety. Third, theNEA emphasizes steps to alleviate the multiple pressures on Mauritius' fragile environment, stemming from the hotel and textile industrybut also from deficient sewerageinprivate housing. vii. ...whileat the same time improving economic management. Inthe Agenda, the Government commits to reducingthe overall fiscal deficit to about 3 percent of GDP by the endofits mandatein2005/6. Further,the program outlines measuresto improve budget management by prioritizingexpenditures to meetNEA objectives and to gradually eliminate quasi-fiscal activities. The government would also redefine the role o fthe state so that the public sector gradually retreatsfrom the productive sectors as strategic relationships with the private sector are established. viii. The Government'sreform program includessubstantial increasesinpublic investment overpreviousyears InJune 2001, the Government's assessment of the total investmentcosts ofthe five-year NEA came to about Rs. 50 billion; this translates into 6 per percent of GDPper annum inpublic capital investmentsthrough 2005/6, which would doublethe pre-NEAlevels. These estimates are based on sector strategies and action plans. A detailed costing ofthe fullNEAnevertook place, however, andthe appropriaterole for the private sector remains flexible. ix. Thefirst year of NEA implementation witnessed an unfavorable external environmentand a highprogrammed deficit.. In2001/2 the economy performed more . poorly than hadbeen expectedwhenthe Agenda was drafted, owing to the global slowdown, the Madagascarcrisis, and a cyclone. Current and future growthprospectsremain dampened. The fiscal deficit was high (5.9 percent o f GDP), while only a limitedincreasein capital investmentsmaterialized. Although an overall decrease o f support to parastatalshad beenplanned underthe NEA, suchtransfers remainedhigh. X. ...whilea number offiscal risks became morepronounced. By the end of 2001/2, the Government was rolling over about 3 percent o f GDP a month inshort-term debt. At the same time, contingent liabilities were increasing as parastatals encounteredlosses and 11 .. accumulated debts; inaddition, the Bank ofMauritius engaged inquasi-fiscal activities to support the textile andsugar sectors. Further, central government debt increasedstrongly, largely due to liquidity managementoperations ofthe Bank of Mauritius. Mauritius' total public sector debt reached77 percent of GDP atthe endofthe fiscal year, an all-time high. xi. Today,Mauritius stillfaces the challenge offinancing theAgenda's investment program while at thesame time confronting rising social expendituresand transfers to the parastatal sector. The government is committed to undertake the proposed transformation of its public expenditures without endangeringthe stability of the public sector and the economy. Consequently, inimplementing theNEA, the Government has to answer three very basic andcritical questions: (i) muchcan government affordto spend?;(ii) How What should government prioritize?; and (iii) How to make good public management choices? HOW MUCHCAN GOVERNMENT AFFORD TO SPEND? FISCALSUSTAINABILITY xii. Mauritius'future success lies in the country's capacity to transform its economy while maintaining sustainablefiscal balances. Two factors are currently affecting Mauritius' fiscal stability, and its capacity to implementthe Agenda ina sustainable and fiscally responsible manner. First, Mauritiushas a high levelof debt andthe public sector debt carries important risks. Mauritius' total public debt stands at 77 percent of GDP, of which 55.3 percent of GDP comes from the central government and 22 percent are owedby parastatals. Central government debt is mostly domestic and short term, while parastataldebt i s mostly external and long term. Inaddition, rollover risks from the Government's short- term portfolio, contingent liabilities, and the risk o f further erosion of tax revenues also pose significant risks. xiii. Second, growth prospects are somewhat uncertain and interestrates are likely to increase inthe mediumterm. Currently, it does not seem likely that Mauritius can easily attain and exceeda GDP growth rate o f 5.5 percent per annum, its historical average, as preferential trade treatment of its exports is eroding, new sectors willtake time to pick up, andthe international environment is less favorable than inthe past. At the same time, global interest rates are at an all-time low, and are boundto increase. xiv. Mauritius may wish to set a long-term debt target of 60percent of GDPfor the consolidatedpublicsector... Although there exists no specific cut-off for what would be a sustainable debtrate for the public sector of a country, Mauritius may wish to aim for a long- term debt target of around 60 percent o f GDP. This would follow the example o fthe EuropeanUnionaccessioncountries. The proposedtarget impliesthat Mauritius' average primary balancesfor the public sector should be inthe neighborhood o f 0.9 percent o f GDP for the next 10years. There are various combinations o f attaining such positive primary balancesbutthe lower the government outlays to the parastatals, the higher its ability to finance the neededcapital expenditures ininfrastructure, education, and health. xv. ...and to improvethe quality of its debt. The quality ofthe debt couldbe improved if the maturity of domestic debt were lengthened(even ifthat would come at a cost), the operational framework o fthe parastatalswere strengthened(especially regarding price ... 111 adjustments of petroleum and electricity products), and the central government became the exclusive entity that could issue its own debt. Furthermore, Mauritius' debt management capacity could be strengthened. xvi. The original NEAfiscal deficit targets are sustainable, but it would be very diffult for government tofinance planned expendituresand meet deficit targets at the same time, The "EA framework formulated inJune 2001provided for anoverall fiscal deficit o fabout 3 percent o f GDP by 2005/2006. Assumingthat this i s attained and such deficits maintained, the outlined target o f a consolidated public sector debt of about 60 percent would be attained in10years. However, the originally plannedexpenditurelevels ofthe Agenda, especially the doubling of capital investments until2005/2006, are not consistent with attaining such deficit levels. Inpart, this stems from the expected slower growth rate o f the economy and the possibility that governmentmighthaveto continuously support parastatal companies in the next years. xvii. Mauritius should raiserevenues and keep expendituresat historical levels. The fiscal adjustment necessary for a gradual reduction o fthe fiscal deficit to about 3 percent o f GDP could come largely fiom enhancing revenue collection rather than expenditure reduction. The medium-term fiscal framework indicates that total central government expenditure can be maintained at the historical level o f 25 percent o f GDP, but that the revenues will have to be increased from the present 20 percent to 22 percent o f GDP. The government has already taken steps inthis direction. The rate o f VAT has beenraisedtwice inthe pasttwo years. The follow uprevenue measureswould include streamliningtax exemptions and other tax incentives and improving the tax and customs administration. WHAT ARE GOVERNMENT'S OPTIONS FOR MAINTAINING FISCAL SUSTAINABILITY? xviii. Mauritiushas many optionsfor implementingtheAgenda. Mauritius can increase revenues, reprioritize expendituresandphase investments. Mauritius i s also able to achieve the Agenda's objectives by re-examiningunderlyingpolicies inthese areas andbuildingup capacity at the institutional levels. xix. Although revenuegains can be madefrom broadeningthe tax bases... Mauritius may wishto broadenthe tax base for direct andindirect taxes. Mauritius may also desire to quantify tax expenditures, andto develop a more even-handed tax policy between exporting and import-competing sectors. xx. ...expenditure reprioritization,phasing of investment,and efficiency gains willstill be required. Simple linear projections o fexpenditures suggest that Mauritius will probably have to achieve average annual savings o f about 1.5 percent o f GDP with respect to current expenditures inorder to meet the target sustainable expenditure ceiling o f 25.7 percent of GDP for the period 2003/4 - 2005'6. With respect to capital expenditures, we estimate that only about halfofthe plannedAgenda investment plan can be financed within a sustainable fiscal framework. It would be possible for Government to increase capital spending inthe short run, however, ifequivalent savings are made with respect to current expenditures, iv going beyond the 1.5 percent cuts already needed. There are several illustrative cases or options: xxi. Outlays toparastatals (utilities and importing monopoly). Historically Mauritius has made outlays t o parastatals through a variety o f channels, namely, direct subsidies, capital transfers, direct lending, and loan guarantees. It has especially used these channels to subsidize electricity and petroleumproducts, to a large extent benefiting industry. Mauritius may wish to reconsider the rationale o f subsidizing electricity prices and petroleum products at a time when scarce public resources are neededto carry out the muchneededAgenda investments ininfrastructure, education, and health. Mauritius may also wishto consider gradually, but steadily and credibly, reducingthe subsidy to electricity and petroleum products. xxii. Pensions. Demographic pressureswill leadto a doubling o f basic pensionpayments over the next 20 years, which already amount to 3.9 percent o f GDP today. Mauritius might find it important to limit the pension expenditures byeither targeting the basic pensionto the most needy, or by moderating pensionpayments to all and introducing complementary assistance programs for the elderly andvulnerable. xxiii. Housing. Most o fMauritius' housingexpenditures through 2000 were geared towards the middle class, while since then, expenditures have focused on very low income housing. Mauritius can reduce housing expenditures that do not directly support very low income groups. Mauritius may wishto reconsider its polices o f subsidizing interest on housing loans and guaranteeing debt o f housingparastatals. Inaddition, it may wish to intensify monitoring and evaluationo fthe very low housing schemes formulated inrecently under the NEA, which are aimed exclusively at the very poor. xxiv. Educationandhealth care. Mauritius' development strategy, which rests on building upthe country's humancapital, andMauritius' aging populationsuggest that the demandfor education andhealth services i s boundto increase. Therefore, the government may wishto (i)accompany development programs withprecise plans for cost saving; (ii) its citizens ask to pay for an increasing share o ftertiary level education andhealth services, and (iii) consider that the public sector's role might switch from provider to regulator for selected services. xxv. As a next step to identvying expenditurepriorities, the Governmentmay wish to assess the extent to which the current set of expenditureprograms correctly reflects its strategicpriorities. This would include (i) clearly identifying the policy objectives currently supported by the budget and by quasi-fiscal activities; (ii) undertaking a transparent and comprehensive costing o f existing policies; and (iii)identifying the beneficiaries o fthe support programs. Such a general, backward looking program review would also be a key ingredient for development o f a Medium-Term Expenditure Framework. xxvi. On debt management, thegovernmentmight consider to develop a debt managementstrategy that takesinto accountcentral government andparastatal debt and establishes debt level and risk benchmarks. Mauritius may wishto try to reduce risk rather thanminimize expected cost. Mauritius would needto buildupits capacity to implement V thisnew strategy. Mauritiusmay also wishto keep central government debtissuance as a prerogative to the central government and allow the central bank to benefitfrom the issuance o f its own debt. Similarly, the government may also wishto benefit from the deepening o f the secondary market for government paper. HOW TO MAKE GOODEXPENDITURE CHOICES? INSTITUTIONSFORBETTERBUDGETINGAND RESULTSFOCUS xxvii. I n order to achieveits sustainability andprioritization objectives, and in order to improveefficiency in itsprograms, the economicmanagementsystem would need to be modernized. Mauritius' existingeconomic management system has demonstrated its good capacity, over the past 20 years, to maintainfiscal discipline at the central government level. However, Mauritius needs to be better able to monitor its fiscal framework, link allocations to priorities, engage incomprehensive costing and risk assessment o f the public sector, engage inmore detailed expenditure planning, and improve capacity for program design, xxviii. TheMedium-Term Expenditure Framework is a useful toolaround which prioritization of expenditurescan takeplace. .. Mauritius needs an adequate framework to plan, manage, andmonitor the implementationo fgovernmentpolicies andthe allocation of resources. This framework should support strategic decision-making andthe prioritization o f activities inline with overall government policies. These government priorities should be adequately reflected inthe government budget. Budget implementation should be monitored against these targets. The Medium-Term Expenditure Framework (MTEF) can be a useful tool to plan and manage public resources. The MTEFwould enable the Government to strengthen the focus on results and to improve service delivery and efficient use o fpublic resources. This will be achieved by moving from a focus on inputs to assessing outcomes in line with the NEA's strategic priorities, following the principles o f results-based management (" xxix. ...butit needs to be comprehensiveto have value-addedfor Mauritius. Mauritius uses a broad set o fpolicies and instruments to achieve its development objectives. The policies range from price supports on energy usage, to incentives for growth sectors, and to free education and health care. The instruments include tax exemptions, preferential interest rates, loan guarantees, government loans, public private partnerships, and direct government expenditures to central government andparastatal entities. An MTEF for Mauritius needs to include all these financing mechanisms (and risks) to present a thorough costing o f existing programs. It also needs to facilitate the evaluation o ftradeoffs betweencompeting policy objectives. xxx. The ministriesof Finance and Economic Developmentcould bring their skills together more effective&. The Mauritius' Ministryo fFinance has developed expertise in runningatight budget preparationprocess, based onthe existingbudgetary system, and excellent expenditure monitoring. However, it falls somewhat short on development policy and sectoral program design, as well as monitoring and evaluation o f outcomes. On the other hand, the Ministry o fEconomic Development has ample experience inpolicy analysis, and in evaluation o f capital investments. A comprehensive approach o fthe government to vi expenditure management will require that the talents of bothministriesbe used intandem. The government may evenwishto consider merging the ministries. xxxi. However, improvementsinpublic managementrequire more than the MTEF. In order to improve fiscal sustainability andthe efficiency ofMauritius' public sector, Mauritius needs to make a cross-sectoralassessment of its system of economic management. Mauritius needs to also address capacity gaps inpublic administration. Capacity must exist at the technical level to translate government priorities into specific, time-bound activities and monitorable outcomes to enable governmentto assess the implementation o f its policies. Technical capacity i s also important to evaluate implementation progress inpriority areas. Management capacity is requiredto make necessary adjustments inthe course ofthe implementation of government policies. THIS PUBLICEXPENDITUREREVIEW xxxii. The present report forms part ofthe analytical and advisory services of the World Bank to the Government ofMauritius. The report looks at historical trends inpublic expenditures(chapter l), foundation andcosting ofthe NewEconomic Agenda (chapter the 2), fiscal sustainability intimes of risingpublic debt (chapter 3), necessaryexpenditure choices (chapter 4)' and finally the institutional set-upfor economic management(chapter 5). The report does not provide an in-depthsectoral expenditure analysis of, for example, the transport, education, health, and environment sectors, which are all core investmentareas of the NewEconomic Agenda. Such sectoral reports are plannedfor the near future andcan come together under a subsequentPublic ExpenditureReview. The present report also does not address broad financial sector issues because a Financial Sector Assessment Program (FSAP) report was completed inMarch2003. Some elements ofthe FSAP coincided with the fiscal risk work undertakeninthis report and are indeeddiscussed. vii 1. PUBLICEXPENDITURES IN HISTORICAL PERSPECTIVE 1.1 Since independence, Mauritius' economy has performed strongly and the standards o f living o f its populationhave improved significantly. The economy grew at an average o f 5.5 percent er annum from 1980, triplingper capita income of the populationto US$3,800 in 2001/2. Mauritius outperformed sub-Saharan African countries which grew at an average of P 3 percent per annum over the period 1981-2000and its historical growth record compares well with high-performing Asian economies, which grew at an average o f 6.5 percent per annum over the same time period.2 Importantly, while incomes increased, inequality fell and poverty, both absolute andrelative, decreased strongly: relative poverty decreased from 28 percent twenty years ago to about 10percent today. Similarly, literacy rates improved from 74 percent o f the population in 1980to 85 percent in2000, life expectancy increased from 66 years in 1980to 72 years in2000, andnon-communicable diseases have been eliminated. 1.2 Butat the turn o fthe millennium, several worrying trends emerged. Fiscal deficits beganto rise duringthe 1990s, with parastatal enterprises becoming more indebted. Domestic debt edged up, unemployment reached about 10percent o f the population, and preferential trade arrangements insugar andtextiles were about to be phased out. 1.3 The historical trends described inthis chapter gave riseto the formulation o fa new development strategy, the NewEconomic Agenda, which the Government o f Mauritius formulated inJune 2001. The chapter beginswith a briefdescription o fthe Mauritian development model, and then examines public finance and debt trends inthe 1990s.It concludes with a description o fthe parastatal sector, which has played a very important role inthe development o fthe country. DEVELOPMENT POLICIES 1.4 Mauritiuschose a unique pathto economic development. Thepathhas been characterized by a commitment to stabilization inthe early 1980s; a strong policy shift towards export orientation inthe mid-1980s; and reliance on a strong welfare system, which emphasized redistribution.The present section briefly reviews this path given its relevance to the public finance situation of Mauritius today. 1.5 After independence,the Government took important measures to stabilize the economy. At the time o f independence, Mauritius faced a number o f macro-imbalances: high fiscal deficits (stemming inpart from an overblown public sector), highexternal deficits, and an overvalued exchange rate that undermined competitiveness o f the local industry. To stabilize the external account, the Government o f Mauritius devaluedthe Rupeetwice, in 1979andin 1981.Beginningin 1983, the Government floated the Rupee by linkingitto an undisclosed basket o f currencies. To stabilize the fiscal account, the MauritianGovernment ' Inpurchasingpower parity terms, per capita income evenreachedUS$9,330 in2000. The average includesHong Kong SAR, Indonesia, Korea, Malaysia, Singapore, and Taiwan, 1 imposed a surcharge on sugar profits, cut public sector wages andtransfersY3and removeda number o f price controls. By 1983, Mauritius had successfully stabilized its economy and brought external and internalbalances under control. 1.6 Stabilizationpaved the wayfor adoption of export-promotionpolices.With the stabilization o fthe economy well underway inthe early 1980s, Mauritius sought to undertake structural reforms instrategic sectors inorder to promote exports and diversify its economy. The adjustment followed the realizationthat the country's small domestic market and import substitution polices initiatedin 1964 offered few opportunities for sustained growth. The Government implemented important reforms infour sectors: (i) agriculture: to improve profitability of the sugar sector export taxes were reduced, sugar processingwas rationalized, and as a consequence, the public sector labor force inagriculture shrank; (ii) export-oriented industrialization: the Government created a number o fpromotional institutions, double- taxation treaties were concluded, tariffs were rationalized: quantitative restrictions were eliminated, andprice controls were markedly relaxed(but a total o ffifty price-controlled commodities were left inplace); (iii) tourism: air links to Europe were created, the transport policy was reformed, and the development o f a domestic shopping service center increased expenditures per tourist; lastly, (iv) public management: the Government reorientedthe investment rogram, improvedits debt management, andstarted divestiture from selected parastatals. As part o fthe policies to promote the private sector, the Government also ? subsidized fuel and electricity prices. 1.7 The export promotion policies were highly successful because they aimed to support sugar and textile exports, which were protected through preferential trade agreements. Mauritius' membershipinthe Commonwealth came with a guaranteed sugar quota for exports to the European Community at a guaranteed price, which has turned out to be about 30 percent higher than the world market price.6 Similarly, the Multi-Fiber Agreement provided duty free access to the Europeantextile market, which shielded its Export Processing Zone from the competition from lower cost producers inAsia. As a consequence, sugar exports were maintained andtextile exports increased inthe second half o fthe 1980s. 1.8 Mauritius has invested in a comprehensive welfare system. The Mauritian welfare system i s closely modeled after the BritishBeveridge system. While individual welfare policies have changed over time, the basic components o f the system have remained unchanged over the past decade and comprise: (i) direct transfer schemes to households, most important among those are the basic (non-contributory) pension scheme which guarantees an income to the entire old-age population; (ii) the provision o f free education and healthcare at all levels, financed out o f general tax revenues; and (iii) an important (and over time growing) systemo f housing subsidies. Inparallel, the public sector (central government and public enterprises) has traditionally beenvery large, offering employment to 16 percent of Mauritiustried to removethe riceandflour subsidiesin 1979, but the LegislativeAssembly reinstatedthem. Fiscal, customs, and stamp duties and surcharges were consolidatedinto auniformtariff, A moreuniform tariff structurewas createdto lowerto the ERPand increasethe tariffs with negativeEms. 'The DevelopmentWorks Corporationdownsizedfrom 14,000 to 4,000 workers, employment inthe Tea DevelopmentAuthorityfell by 5 percent, andthe CentralHousingAuthority was closeddownandworkers were absorbedbyother sectors. This was the case from 1978-1987. 2 the work force (2001). The Government also usedextensiveprice controls and commodity subsidies for basic foodstuff (e.g., low quality of rice) andbasic services. 1.9 The costs of the welfare systemhave increasedsubstantially since the mid-1980s. Table 1.1compares the major componentsof the welfare systemin 1985 and 2000. Over these fifteen years, the total outlays for transfers, social expenditures, and subsidies have increasedfrom 10.7 to 15.1 percent o f GDP. Table 1.1 Snapshot o f Some Direct Costs o f the Welfare System (percent o f GDP) 1985186 2000101 Subsidies (on commodities) 0.3 0.6 Transfers to households 4.9 7.2 (including pension) Educationand healthcare 5.2 5.6 Housing and comm. amenities 0.3 1.7 Total 10.7 15.1 Source: Governmento fMauritius. PUBLIC FINANCE TRENDS THROUGH 2000 1.10 While Mauritius was able to limit its fiscal deficits to below 3 percent of GDP during much ofthe 1 9 8 0 this ~ ~ situation changedmarkedly over the last decade. The public sector deficit increasedsteadily, reaching an all-time highof almost 6 percent in2000. This was mainly due to declining revenuescoupled withexpendituresthat remained high, albeit a considerable change incomposition. As a consequenceofthe higher deficits, total Government debt increasedsteadily. Simultaneously, boththe composition o f debt (from foreign to domestic) andthe maturity (from long to short term) changed significantly. 1.11 Declining revenues and high expenditures led to increasing deficits. 30 Figure 1.I:Rev, Exp, Def as % GDP Figure 1.1 shows the development oftotal PI' public revenues and expenditures(bars) 5 and the resulting fiscal deficit (line) from 25 4 the mid-1980s to 2000. As can be 3 observed, the revenue-to-GDP ratio fell by almost 4 percent of GDP, from 23 percent 20 2 inthe mid-1980s to 18.2percentin 1 2000. While inthe early years this was due 15 0 - to a proliferation of exemptions, inthe late 86-88 89-91 92-94 95-97 98 99 00 1990sthis was a consequence of tariff L-iRev & grants Expend & NL liberalization inline with the free trade (-) Deficit (R) agreements Mauritius signed, such as COMESA. This large drop inrevenueswas Source: Government of Mauritius only partially offset by an increaseintaxes on domestic goods and services (see Table 1.2). At the same time, expenditurefluctuated only slightly andremained at around23 to 25 3 - percent o f GDP throughout the entire period. Consequently, fiscal deficits show arising trend over the fifteen years considered, reaching an all-time highin2000. 1.12 Recurrent and capital expenditures fluctuated littZe. Total, current, and capital expenditures were keptwithina range of 23-25, 19- Figure1.2:titpendiures 21, and 3-5 percent of GDP inthe late 1980sand throughout the 1990s. Figure 1.2 shows minor .\ fluctuations inaggregate and functional expenditures over this time period. Within current expenditures,however, amodestfunctional shift occurred as the share of wages was reducedwhile f - * - - - o h - - . - * - the importance of subsidies andtransfers increased P4 (see Table 1.2). Capital transfersto parastatalshave 86-88 89-91 92-94 9597 98 99 00 averaged about 1percent of GDP since the late 1980s, directly funding some o fparastatals' investmentandindirectly other items suchas debt service. Source: Government of Mauritius 1.13 During the 1990s,Mauritiusprioritized "18i ..--I social over economicspending. Mauritius Figure 1.3:Functional Expenditures allocated resourcesto the social sectors at a rate 1 4 , I muchfaster than economic growth, leadingto an increase inthe share of social sector expenditures K) while spendingon other sectorswas reducedin importance. Figure 1.3 shows the development of different functional expenditures, includingsocial 4 - --'0 expenditures,generalGovernment expenditures 2 (administration), and outlays for economic 86-88 89-91 92-94 95-97 98 99 00 services from the mid-1980s to 2000. Social -a-- GenGov -CommlSoc sector expenditures(definedas comprising -Econ Sew .-- -.- .Other outlays for health, education, housing, public pensions, anddirectedtransfer schemes) increased Source: Government o f Mauritius from 9 percent of GDP inthe mid-1980s to almost 13 percent in2000. Concomitantly, the Government strongly reducedexpenditures on economic services (comprising agriculture, transport, and communication) over the same time period. 4 Table 1.2 Summary of Budgetary Central Government Balances (percent of GDP) 1986-88 1989-91 1992-94 1995-97 1997198 1998199 1999100 2000101 Average Average Average Average Total Revenueand Grants 23.1 23.0 21.5 19.8 19.6 20.1 20.9 18.2 Tax revenue 20.5 21.5 19.4 17.3 16.7 16.9 18.1 16.1 Incomeand profit 2.3 3.3 2.8 2.7 2.6 2.6 2.6 2.4 Dom. goods, services 5.1 5.5 5.8 5.7 6.6 8.0 8.0 7.6 Imports 9.4 9.9 8.7 7.7 6.6 5.8 6.1 5.1 Nontax revenue 1.7 1.3 1.9 2.2 2.6 2.4 2.0 2.0 Grants 0.9 0.2 0.1 0.2 0.2 0.1 0.1 0.2 Total Expenditure 24.9 25.3 23.8 23.8 23.4 23.4 24.7 23.9 Current Expenditure 20.1 20.4 19.4 19.4 20.5 20.6 20.8 21.4 Wages and salaries 7.3 7.7 7.0 7.2 6.8 7.1 6.8 6.5 Purchase of goods, services 1.9 2.1 2.3 2.2 2.1 2.1 2.2 2.2 Interest payments 4.8 4.2 3.1 3.2 3.7 3.5 3.4 4.4 Subsidiesandtransfers 6.1 6.4 6.9 7.2 7.9 8.3 8.3 8.3 Capital Expenditure 4.9 4.8 4.5 3.8 2.9 2.8 3.9 2.5 olw Fixedassets 1.8 2.8 3.1 2.1 2.0 2.3 2.0 2.8 olw Capital transfers 2.0 0.9 1.o 0.8 0.7 1.o 1.2 1.2 olw Net lending (parastatals) 1.o 1.1 0.4 0.9 (1.9) (0.6) 0.1 (4.4) Other' 2.o 0.1 0.6 2.9 FiscalBalance (1.8) (2.3) (2.4) (4.0) (3.8) (3.3) (3.8) (5.7) Primary Balance 2.9 2.0 0.7 (0.8) 0.1 0.4 (0.3) (1-5) Financing 1.8 2.3 2.4 4.0 3.8 3.3 3.8 5.7 Domestic 0.9 3.1 2.9 3.1 4.1 4.4 4.2 3.5 Foreign' 0.9 (0.8) (0.5) 1.o (0.3) (1.1) (0.4) 2.2 Memo Items: Total Debt Stocks 59.9 54.9 41.3 44.9 48.4 49.2 50.4 48.5 Domestic 37.1 39.0 30.9 33.9 37.0 39.5 41.6 43.1 Foreign 22.7 16.0 10.4 11.0 11.5 9.7 8.8 5.5 Source: Governmentof Mauritius 5 1.14 Thecomposition of social spending changed markedly between Figure 1.4 RecurrentSocial Expenditures as the mid-1980s to 2000. Recurrent a % of GDP, by sector 1 social spendingincreasedfrom 8 12 1 I percent o f GDP inthe mid-1980sto 10 - nearly 11percent of GDP in2000. 8 - Figure 1.4 shows the composition of 6 - recurrent social spending. Social +*_,;IC--*--m 4 - assistance expenditures were the most *T.-.*.r.:.*.:. .. ......--....-......... 2 - important spending item within this & 0 Y Y Y Y I5 category and at the same time O + 86-88 89-91 92-94 95-97 98 99 00 responsible for practically the entire - increaseintotal social spending. In ---6---Health Total .-..-x-...-.SocialAssistanc Education Housing turn, the rise insocial assistance I expenditures itselfi s almost entirely dueto the rising outlays for the basic Source: Government o f Mauritius pension scheme which grew to 4.9 percent of GDP in2000. Recurrent expenditureson health and education stayed consistently atjust under 2 percent of GDP and slightly above 3 percent of GDP, respectively. 1.15 Social capital expenditures are small compared to social current Figure 1.5 Social Capital Expendituresas a expenditureoutlays, butthey have YOof GDP, by sector 2.0 7 I doubled inimportance (as apercentage o f GDP) since the mid-1980s. Figure 1.51 - 1.5 shows developments of total social capital expendituresas well as its main components (education, housing and health). Public housing and community amenities (environment andpublic 0.0 u Y 4 utilities) investmentwas the fastest 86-88 89-91 92-94 95-97 98 99 00 growing sector, enlarging from under -Total Housing 0.5 percent o f GDP inthe mid-1980s ...-...Education 4Health to 1.2 percent o f GDP in2000. Source: Government of Mauritius 1.16 Mauritius' centralgovernment debt has increasedsignificantlysince Figure 1.6:Central GoaDebtStock the early 1990s.Figure 1.6 compares the development ofthe (central government) debt to GDP ratio (left scale) with the GDP growthrate (right scale). The debt stock contractedby an average o f 1percent a year inreal terms between 1986-91, with GDP 86-88 89-9192-949597 98 99 00 growing at anaverage 6.6 percent. -Debt - - ----.GCPGr*owth(Fp Consequently, the debtdGDP ratio Source: Government o fMauritius 6 decreasedsignificantly from around 56 percent in 1986to 40 percent inthe 1992-1994 period. However, with increasing government deficits inthe 1990scame a reversal ofthis situation: inthis decade, the debt stock grew at rate of 7.5 percent, outpacing the growthrate of the economy (5.5 percent) inalmost every year. Hence, the most important debt indicator, the debt/GDP ratio, increased throughout the decade and was close to 50 percent of GDP in 2000. 1.17 The Governmentincreasingly used domesticsources tofinance rising deficits. Duringthe mid-l980s, the fiscal deficit was financed by domestic (60 percent) and foreign borrowing (40 percent). Motivatedpartly by the desire to avoid exchangerate risk, the government increasingly relied on domestic financing. As a result, the stock o f outstanding foreign debt fell from 22.7 percent of GDP inthe mid-1980s to 5.5 percent in2000/01, while the domestic debt stock rose from 37.1 percent of GDP to 43.1 percent inthe same period. Today, about 90 percent ofMauritius' central government debt is domestic, up from about 62 percent inthe late-1980s (see Table 1.2). 1.18 The maturity of debt became Figure 1.7 :CG Total Deb by Maturityas % much shorter. Inparallel with a shift in GDP the composition ofthe total debt stock K 60 towards domestic financing sources 50 40 came a marked shift towards shorter 30 maturities. While short-term debt 20 represented30 percent of total public 10 0 debt inthe mid-1980s, itrepresented 73 86-88 89-91 92-94 95-97 98 99 00 percent in2000 (Figure 1.7). This poses a substantial rollover risk for IC&JTotalDebt StockmShort TermULongTerm1 Government. Source: Government o f Mauritius THEPARASTATAL SECTOR 1.19 The parastatalsector has played anintegral part inMauritius' development, with public enterprises managing, at times, the airport, port, public housing, water and electricity services, andmeat processing. Parastatals inMauritius are usually founded (and their legal positiondefined) inseparate acts that also outline their "raison d'&tre." Parastatalsare in principlemanagedby associated"parent" ministries.The parastatals' legal status requires themto beaudited onan annualbasiswiththe audits submittedto the NationalAssembly for discussion. However, prior to a majorpolicy shift in2001, Mauritius has not systematically collected and centralized informationonthe performance of individualparastatals.This section attemptsto give a broad indication o fthe importance o fparastatalsprior to the formulation ofthe NewEconomic Agenda inJune 2001. Chapter 2 discussesdevelopments intheparastatal sector subsequently. 1.20 Theparastatal sector is extensive.Mauritius has thirty-four non-financialpublic enterprises, or parastatals, and employs about 16,000 workers, or about 3 percent o f the labor force. Their activities spreadover almost all economic sub-sectors. General information 7 about the largest sixteen enterprises i s provided inAnnex 1.7 The economic and financial performance o f public enterprises has significant implications for service delivery quality, government finances, andthe degree o f competition inthe economy.* 1.21 Parastatals benefitfrom substantial centralgovernmentfunding. Table 1.3 includes subsidies, lending, and capital transfers to parastatals inthe most important sectors, Over the three-year period 1998 to 2000, an average o f about 0.7 percent o f GDP per annum was spent on subsidies, about 1.O percent o f GDP per annum has been spent on capital transfers, and about 0.3 percent of GDP per annum has been spent on gross lending. Capital transfers and net lendingrefer to outlays to non-financial public enterprises, not extra-budgetary units or ministries. Taken together, the three categories have beento a large extent directedat housing and community amenities, 37 percent, followed by agriculture, 19 percent, and transport 17 percent. Interestingly, repayment o f lendinghas come from sectors that are not major recipients of funds in 1998- 2000 period. Table 1.3 Parastatal Subsidies, Capital Transfers, Lending, and Government Repayments (3-Yr average, 1998/99-2000/01, as a percent o f GDPand share o ftotal) Current Capital Share of Total Subsidies Transfers Lending (1)+(2)+(3)/ Repayment Sum (I,2,3) 3-year 3-year 3-year 3-year ParastatalSector average average average Percent average Gen. Public serv. 0.03 2 Education Housing & comm. amenities 0.03 0.6 0.1 37 0.07 Recreational, etc. 0.0 Fuel& energy Agricultural etc. 0.1 0.2 0.03 19 Miningetc. 0.03 0.07 5 Transportation etc. 0.1 0.2 0.05 17 1.4 Other econ. affairs 0.4 20 0.7 TOTAL 0.7 1.o 0.3 100 2.2 1.Lendingis project lending, not subsidies, to parastatals."-"denotes less than 0.005 percent. 2. Parastataloperations for defense, public order, health, social security and welfare, and other expendituresamountsto less than 0.005 percentofpayments. Source: Government ofMauritius 'Mauritius also has e&$t Public Financial Institutions (Bank o fMauritius, Development Bank o fMauritiusLtd, Mauritius Housing Company Ltd, Post Office SavingsBank, State Bank of Mauritius Ltd, State Insurance Corporation ofMauritius Ltd, Sugar InsuranceFund Board, Mauritius Cooperative Bank) and fom-five Extra BudgetaryUnits(EBUs) (see Digest ofPublic Finance Statistics2001, Central Statistical Office, Republic of Mauritius, April 2002). The EBUs function as detached government departments whose budgetsare financed entirely by the parentministries. EBUs operate primarily, but not exclusively, inthe education sector. Mauritiusdoes not possess centralized information or a database for public enterprisesinMauritius. The information on the major public enterprises was collected from enterprises individually with exceptional assistance from the Ministry o fFinance. 8 ~~~~ 1.22 The Governmentmade extensiveuseof Figure1.8: Guaruntees andCentral loan guaranteestoparastatals,especially at GovsmmentDeficits timesof high deficits. At the endof June 2000, the Government heldparastatalloan guarantees amounting to about 14percent of GDP. Figure 1.8 shows new guarantees to the parastatal sector andthe fluctuation ofthe public sector I deficit from 1986 to 2000. The two curves show -84 I , , , , , , , , , , I , 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 a very strong negative correlation, suggesting -Signed Guarantees -- Eeficit that the Government issues more guarantees to parastatals at times when deficits are high Source: Government o f Mauritius compared to when deficits are low. Over time, the financing ofthe guaranteeshas shifted from external to domestic sources, similar to the borrowingpatternof the central government observedaboveeg 1-23 Nevertheless,thefinancialperformance of parastatals was disappointing.Inthe 1990s, public enterprises accumulatedlarge financial losses. In2000, the total revenues of the largest 16parastatalscame to only about 1percent of GDP, eventhough severalprice adjustmentshadbeenundertakenthat year. Consequently, in 1999/00 alone, the lossesof the three largest parastatalsalone, Central Electricity Board (CEB), State Trading Corporation (STC), and Developments Works Corporation (DWC), amountedto about 3 percent o f GDP. Short-term borrowing from the banking sector andthe accumulation of arrears, particularly to the central government, largely financedthese losses. The maincause for these public enterprise losseswas the failure to adjust a number ofkey prices (particularly oil, electricity, busfares, andwater rates) inline with increasing costs andworld marketprices. For example, electricity tariffs were not adjustedbetween1992-1999 andoilprices were kept constantbetween 1996-2000. Iti s not surprising to findthat Mauritianparastatalshad also accumulatedpayment arrears and overdrafts of about 2-5 percent of GDP bythe endo f 2001/2. THENEED CHANGE FOR 1.24 Mauritius' development strategy aimed to support economic growththrough export orientation, extensive government assistance, partnerships with the private sector, anda generous welfare system. This strategy was successful during muchof the 1980sand 1990s whenthe economy was growingdue to preferential market access inindustrialized countries, while increasing public expendituresinsocial sectors assuredthat the benefits from growth were widespread. 1.25 But at the turnofthe millennium, the country faced anumberofchallenges:First, growth sustainabilitv: Mauritius' export markets (and its growth) were threatened by the imminentdeterioration ofthe preferential treatment ofMauritius' exports andthe advent of globalization -bothof which make Mauritius' export markets more contestable. Second, welfare sustainabilitv: despite increasing welfare expenditures, a core part o fthe population The Government does not directly meet debt service obligations o f parastatals. However, the Government supports some parastatals through subsidies that enable them to meet their debt service obligations. 9 remained poor andunemployment reachedrecord highlevels of nearly 10percent. Finally, fiscal sustainabilitv: Mauritius' fiscal balancehadbeendeteriorating inthe 1990swhile debt levels increased. 1.26 TheNewEconomic Agenda, designedin2000/2001, set out to address these three challenges. 10 2. THE NEW ECONOMICAGENDA 2.1 A new Government was elected inMauritiusinSeptember 2000. Itcame to power at atime when the growth rate ofthe economy had fallen to 2.6 percent, down from an average of 5.7 percent over the previous two years, andunemployment hadreached 8 percent, the highest since 1986. Additionally, the growth rate for private investment had beendropping since 1997. Externally, petroleum prices hadreachedUS$30 per barrel, contributing to lossesof parastatals; the State Trading Corporation, STC,andthe Central Electricity Board, CEB, together accumulatedlosses ofUS$lOS million in2000 alone. 2.2 The new Government took early actionsto stabilize the public finances andavert a financial crisis in2001. Upontaking office inSeptember 2000, the Governmentcancelled a planned civil service wage increaseandimplementeda one-time adjustment inthe retail prices of petroleumproducts andelectricity to preventgreater lossesbeingamassedby the State Trading Corporation andthe Central ElectricityBoard. 2.3 Apart from addressingshort-term stabilization needs, the newGovernment also startedto formulate its strategy for dealing with the longer-term structural problems affecting growth, welfare, and fiscal performance outlined inchapter 1. A NewEconomic Agenda (NEA) was developed andpresentedto parliament inJune 2001,along withthe new Government's first budget. It followed areport on "The Present State o fthe Economy" (2001), which laid out the primarychallenges facing the country. This chapter examines in detail the reform program outlined inthe NewEconomic Agenda, its cost implications, and the underlyingassumptions requiredto finance the Agenda's reformprogram. This chapter also discusses the financing and functioning of parastatals, which, as we will see, play an important role inthe NEA's implementation." THENEW ECONOMIC AGENDA:PROGRAM 2.4 At its presentationto parliament inJune 2001, the Agenda's aim was to achieve high and inclusive economic growth inMauritius by addressingthe economic and social challengesthe country faced. The Government's primaryconcern was the reversal of the risingtrend inunemployment. TheNEA was designedto act as a set of policy guidelines that would leadMauritius to become a "high-tech andhighincome service andknowledge economy" ina five-year period. The Agenda has four pillars: (i) improving competitiveness, (ii)investing inhuman capital, (iii) preservingthe environment, and (iv) modernizing economic management. Table 2.1 presents the objectives and outputs or outcomes, as available, for pillars (i) through (iii) involve substantial costs. Pillar (iv), economic which management, is discussed indetail inchapters 3 through 5. The table illustrates the ambitious nature ofthe NEA, andthe government's commitment to lay the groundwork for restructuringthe Mauritianeconomy. lo The present chapter builds on the Budget Speech o f 2001/2 and that o f 200213, as the latter was used to refine the NEA. 11 Table 2.1 New Economic Agenda Obiectives Objective Outputs I Outcomes Overall Maintain growth at 5.5 percentper annum Reduce national poverty headcountindex by Reduceunemployment from 9 percent 2 percentage points between2002 and 2006 Private sector Reduceproductioncosts in sugar 8,000 workers retrenched;6 of 14 processingplants to close by 2008 Decrease connectivity costs inI T Telecom services liberalized Promote SME growth and employmentgeneration 600 SMEs to benefit from program over 4 years Restructuretextile sector and ensure maximum use Increasedexportsto the U S ofAGOA Maintain good road network Continue investmentand maintenance Reducecongestion on PL-C Corridor Introduce alternative mass transport Human capital Increaseenrollment insecondary schools Increaseenrolment of 12-16 age group to 100% by 2008 Improve quality and equity of educationsystem Achieve 60% CPE pass rates in low performing schools (ZEP) by2005, up from 40% in2001 Meet health needs of population, especially rising Introduce health insurancescheme demandfor interventions for non-communicable diseases Reducepoverty by extendingmicro credit schemes, Extensive coverageof micro credit schemes; as well as providinglimprovinghousing facilities construction o f 1500very low income housing units andtraining programs by 200516 Assure sustainability ofthepensionsystem Curtail expendituresbytargeting the basic pension, improve asset management, strengthensupervision and regulation of voluntary schemes Environment Protect public health and the environment Connecting housesto the sewerage system. Reduce environmental risk by better managing Construction of incinerator liquid waste Source: World Bank. President's Report: Public Expenditure ReformLoanto the Republic ofMauritius - Schedule 2 Mauritius' Reform Program2002-2004; April 3, 2002. 12 Table 2.2 InvestmentsinNEA Flagship Sectors: ExpendituresandIndicative Commitments(% of GDP) Investment ~ Actual Budget Remaining Total Type of 200112 200213 Projected Costs Expend.3 Private sector competitiveness - - - - 2.0 1.3 3.9 7.2 Air transport 0.0 0.1 0.1 0.2 Port 0.0 0.0 0.0 0.0 Roads& Traffic management 0.3 0.3 0.9 1.5 Light RailTransportsystem 0.0 0.0 1.7 1.7 Tourism 0.0 0.0 0.4 0.4 IT Sector 0.1 0.4 0.9 1.4 Textiles Modernization 0.6 0.0 0.0 0.6 Sugar Sector Modernization 0.9 0.5 0.0 1.4 Humancapital - - - - 1.5 2.3 5.O 8.8 Education(pre-prim, prim& sec) 0.2 0.9 1.2 2.3 Education(tertiary) 0.3 0.3 0.6 1.1 Hospitals 0.2 0.2 1.4 1.8 Housingvery low income+rehab 0.2 0.2 0.6 0.9 Housingother 0.4 0.4 0.6 1.4 Poverty 0.2 0.3 0.7 1.2 Environment - - - - 1.o 0.9 3.9 5.8 Water & Wastewater 1.o 0.9 3.9 5.8 SUBTOTAL 4.3 4.4 12.6 21.3 OTHER SECTORS (in 200213 inBudget) 1.4 1.6 9.0 12.1 INDICATIVENEA INV.COSTS 5.7 6.0 21.7 33.4 Memo: Electricity 0.1 0.1 0.2 0.4 Notes: This table usesthe current yearly GDP as denominator for 200112 and200213 andthe Remaining ' Balancewas assumedto be executedover 3 years (from2003104to 2005106). 'These are totalcosts since2001/2, the first budgetunderthe NEA. Someprojectshadincurredcostspriorto 200112, which are excludedfrom this table. 3TypesofExpenditure:M:Ministry investmentbudget, E:Equitypurchaseinparastatal,L:Loanto parastatal, S: Subsidyto parastatal,BOM:Bank ofMauritiusline ofcredit to commercialbanks, guaranteed by centralgovernment, not inbudget. Source: Budget200213; Bank o fMauritiusAnnual Report2001; Ministry ofFinance. Improving the competitiveness of theprivate sector 2.5 TheNEA describes important reforms that aim to increaseprivate sector competitiveness. The Agenda outlines improvements inthe private sector's operational framework and the functioning o f factor markets by revising company legislation, modernizing and streamlining public-private sector interfaces (such as the Board o f Investment), and developing a new supervisory structure for the private sector. The 13 Government also announcedthat itwould carry out an in-depthanalysis of the labor market to improve its flexibility. 2.6 Plannedpublic investments target current andfuture growth sectors. The NEA plans for substantial public investments-either direct or via existing parastatalcompanies- inkeytransport infiastructure (maritime ports, airport, roads, lightrail system) inorderto lower transportation costs for the private sector. For the textile and sugar sectors, the Government plans to continue the preferential treatments currently accordedwith respect to the allocation o fcredit demands." Similarly, preferential treatment for income tax and import tariffs will be continued. To foster strategic and "future growth" sectors, financial incentives are plannedto attract investmentsinto the informationtechnology sector. The Agenda also includes plans for important equity investmentsand direct lending to Business Parks of Mauritius Limited(BPML), an informationtechnology project. Investing in human capital 2.7 TheNew EconomicAgenda seeks to increase service quality in the social sectors. One ofthe Government's priorities, as describedinthe NEA, is to focus on skilltraining (in order to close the gap betweenlabor market needs andthe existing skill level ofthe workforce) andeducation at all levels, with an emphasis on ensuring that all Mauritian children obtain a secondary education. Although the package has not been fully costed, the majority o f investment insecondary schools is expectedto be completed in2004/5, while teacher training, information technology, andtertiary education investmentsare ongoing. The NEA also focuses onthe needfor higher quality(and greater quantity) oftertiary healthcare services, which are inhighdemand as the disease and age profile of the population changes. The Government has decidedto introduce a contributory healthinsurancescheme to pay for the rapidly risingcosts, butthe costs andimplementationmodules have yet to beworked out. 2.8 The aging population will putpressureonthe pension system, alarge share of which i s publicly funded and managed. Reformplans, which are under discussion, would reduce pension expendituresoverall without reducing assistanceto the most needy. Finally, the NEA makesa concentratedeffort to reachout to the very poor with low-income housing schemes, micro-credit schemes, a special education program for poor areas, andthe existing welfare programs presentedinchapter 1. PreservingMauritius'fragile environment 2.9 TheNEA aims topromote developmentin an environmentally sustainablemanner. GivenMauritius' highpopulation densityand the importance o fthe country's pristine The Bank ofMauritius (BOM) made a special line of credit (LOC) available to the banking sector for on- lending to the EPZ inNovember 2000 (Rs. 500 million); the LOC was supplementedinJuly 2001 (Rs. 300 million). The LOC was designedto assist companiesexporting to the Euro zone (and at the time of a weak Euro) and companies seekingto improve their position inthe U.S. market under AGOA. BOM also introduced a specialLOC for the sugar sector inNovember 2001(Rs. 1.5 billion) to support the restructuring o fthe sugar industry inthe context ofthe Sugar Sector Strategic Plan. Inthe year 2001/2, about Rs.1.5 billionwas disbursed to both sectors. Ifthe overall fiscal balancefor 200112 had included the lines of credit under the BOM, it would have come to about 7 percentGDP, comparedto the actual o f 5.9 percent o f GDP. 14 environment to quality o f life andthe tourism industry, Mauritius i s improving regulation and supervision of the sector and also supporting an important waste management program. 2.10 The financing package for the environment sector i s to be supported by public funds, although consumers and industry will contribute more than they have inthe past. The Government currently underwriteslarge investment costs inthe environment sector for liquid waste created from the constructiono f housing connections. Disposal facilities for solid waste have beenpartially fundedby the Government, but new investment costs will be shared with the private sector. Inaddition, cost recovery for drinkingand wastewater from households and industrywill be set by parastatals and regulated by the utilities regulator, which Mauritius i s inthe process o f developing. Improving economic management 2.11 The Government set three priorities for economic managementunderthe NEA.First, stabilize the economy (because the budgetdeficit had reached 6.5 percent o f GDP in2000/1); second, improvebudgetmanagement; andthird, redefine the role o fthe state. 2.12 Attaining stabilization. When the Government introduced the NEA, itcommittedto reducing Mauritius' central Government deficit to about 3 percent o f GDP by 2005/6, which coincides with the endo fthe present electoral mandate. 2.13 Improving budget management. To achieve the NEA objectives ina tight fiscal framework, the Government aims to introduce a Medium Term ExpenditureFramework (MTEF) suitable for Mauritius that will helpbetter alignpriorities with expenditures and make expenditure trade-offs more transparent. The Government also aims to extend the scope o f budget managementbeyond expenditure controls to include results assessments and program evaluations. 2.14 Redefining the role ofthe state. The Agenda seeks to havethe public sector gradually retreat from the productive sectors as strategic relationships with the private sector are established andthe Government limits itself to the role o fregulator. Duringthis process, the NEA includes supportfor key parastatals such as the Central Electricity Boardandthe State Trading Corporation. THECOSTOFTHENEAANDITSFISCALFRAMEWORK 2.15 TheNEA includessubstantial increasesinpublic investment expendituresoverfive years. The Government's initial assessment o fthe total investmentcosts o fthe five-year '' NEA came to approximately Rs. 50 billion or about 6 percent o fGDPper annum. This report estimates, as shown inTable 2.2, that o f the Rs. 50 billion about two thirds would be inthe "flagship" NEA areas outlined above (infrastructure for private sector develo ment, human capital, environment) while about one thirdwould be for other expenditures. These programmed capital expendituresrepresent massive increases comparedto the pre-NEA levels. Inthe first two years o fNEA implementation capital expenditures rose by 32 percent '*The budget does not specify the precise time horizon for the expenditures, so the investments could be made before or afier the five-year mandate o fthe present government. 15 in2001/2, andare plannedto increaseby58 percentin2002/3. Ofthe three flagship areas, investments inhumancapital (education, health, andhousing) are expectedto bethe most costly, about 8.8 percent of GDP, followed by investmentsto foster private sector competitiveness, 7.2 percent, andinvestmentsinthe environment (communityamenities), about 5.8 percent (see Table 2.2). 2.16 Thefiscalframework underlyingtheNEA was contingent on several assumptions. For anumber o freasons, the fiscal framework underlying the NEA was basedon anumber of strong assumptions. Table 2.3 recaps the key ingredients ofthe initial fiscal framework in which Government aimedto finance significant increasesincapital expendituresandat the same time reduce the fiscal deficit from 6.5 to about 3.5 percent over a five-year period. First, the Government planned a significant increaseinfiscal revenues from 18.3 percent of GDP in2001/2002 to 20.5 percent in2005/06; however, concrete measures to realize this elevatedrevenue level were not outlined, although several options were available. Second, overall expenditureswere projectedto remain, more or less, at their pre-NEA levels of around 25 percent of GDP; this impliedthat a strong shift from current to capital expenditureswas requiredifNEA investment targets were to bemet.Third, the underlying growthassumptions were very optimistic given the economy's downturnat the onset of the NEA.Fourth, the fiscal projections assumed that parastatalswould be financially viable and impose minimumburdenon the public purse. Table 2.3 IllustrativeNEA FiscalFramework SupportingApprox. Rs. 50 Billion inCapitalInvestmentandDeficit Reduction 20011022002103 2003104 2004105 2005106 GDPgrowth 5.3 4.6 5.5 5.3 5.5 GrossDomestic Investment 25.1 25.8 26.0 26.3 26.3 Total gov. revenue 18.3 19.2 19.6 20.1 20.5 o/w new net revenueeffort .. 0.9 1.4 1.9 2.6 Total gov. expenditure 24.8 25.2 25.1 24.6 24.0 olw capitalandnet lending 4.9 6.6 6.6 6.0 5.8 Fiscalbalanceafter grants -6.5 -6 -5.5 -4.5 -3.5 Domestic debt 50.9 52.8 53.7 53.6 52.5 Externaldebt 6.0 7.4 8.5 8.1 7.7 CapExpend. (curr. Rs. bill.) 7 10 11 11 11.5 Source: World Bank.President's Report:PublicExpenditureReformLoan to the Republicof Mauritius;April 3, 2002 IMPLEMENTATIONOF THE NEA TO DATE 2.17 I n 200112, the economyperformed weaker than had been expectedin the initial NEAfiscal framework. Table 2.4 comparesthe planned (budget) and actual economic and fiscal indicators. The economy grew at 4.4 percent, about 0.9 percent less than expected, primarilybecauseof Cyclone Dina's devastatingimpact on sugar crops, the impact ofthe Madagascarcrisis on Mauritius' EPZ, andthe dampening impact of the September 11events 16 on international tourism. Concurrently, andvery importantly, investmentdidnot pick up as anticipated, but continued its declining trendfrom previous years, reaching 21.9 percent of GDP, a full 3.2 percent belowthe initialtargets formulated inthe NEA. Imports were also lower than anticipated, contributing to apositive external current account balance at the end of June 2002. The overall fiscal deficit came to about 6 percent of GDP, somewhat lower thanhadbeenplanned, owingto overall expendituresfalling short of plans. Table 2.4 NEAFiscalIndicators 200112 and 200213: Budget andActual (As % of GDP) 2001102 2001102 2002103 (Budget) (Actual) (Budget) GDP growth 5.3 4.4 4.9 Gross DomesticInvest. 25.1 21.9 22.8 ExternalCAB 1.7 5.2 3.5 Total gov. revenue 18.3 18.5 20.0 Total gov. expenditure 24.8 24.4 25.9 olw capital+NL 4.9 4.1 5.0 Fiscalbal. after grants -6.5 -5.9 -5.5 Central gov. debt 50.9 55.3 57.3 olw external 6.0 6.1 6.6 Source: Govemmentof Mauritius 2.18 Implementation of theplannedprogramfell somewhat short of targets with respect topublic investmentsand transfers toparastatals. Table 2.5 shows that while the executed overall current expenditures (andtheir composition) were quite close to plans, Table 2.5 the investment program fell short of NEA ImplementationExpenditureComposition forecasted levels. Instead of the 2000/01 2001102 200213 programmed strong increase (of 30 Act. Bud Act. DifJ Bud. percent) incapital investment inthe Total Expend. 24.3 25.2 24.4 -0.8 25.9 "flagship" areas o fthe NEA, actual Recurrent expend. 20.5 20.2 20.3 0.1 20.3 capital spending decreasedslightly. Communily &social I O 10.3 10.5 0.2 10.1 Education 3.1 3.1 3.0 -0.1 3.0 2.19 This was mainly dueto Health 1.8 1.8 1.9 0.1 1.8 education investment -one of the Socialservices 4.5 4.6 4.8 0.2 4.6 centerpiecesof the NEA -falling Otherexpenditure 10.5 9.9 9.8 -0.1 10.2 considerably behindbudgeted amounts Capital exp. (excl. NL) 3.8 5.1 4.1 5.6 due to implementation delays. -I.o Community&social 1.9 3.4 2.5 -0.9 3.1 Education 0.4 1.4 0.5 -0.9 1.2 2.20 The secondimportant Otherexpenditure 1.9 1.6 1.1 -0.5 2.5 divergence betweenplan and realization inthe first year canbe foundinthe Source:GovemmentofMauritius parastatalsector. Subsidies andtransfers to public enterprises in2001/2002 were sizeable, despite the fact that the NEA hadprovided apath on whichsuch transfers would decrease and the parastatalswould become self sustaining. Table 2.6 indicates that the parastatal sector received atotal of about 1.6 percent of GDP in2001/2 insubsidies (0.6 percent), 17 capital transfers (0.4 percent), and direct loans (0.6 percent), or 1.4 percent o f GDP if repayments are nettedout. This representsabout one third to one half of the amountthat the Governmentinvestedin the acquisitionof fixed assets for the year. Inaddition, table 2.6 illustrates that by endJune 2002, the Government issuedoutstanding guaranteesof about 12.9percent o f GDP for parastatals. Table 2.6 Parastatal Subsidies, Capital Transfers, Lending, and Government Repayments for 2001102 Government Guarantees Subsidies Capital Government Outstanding, Transfers Lending Repayment June `02 Water andwastewater - 0.7 Education 0.0 Housing & comm. 0.2 0.1 0.1 2.9 Fuel & energy 0.1 - 2.5 Agricultural etc. 0.1 0.1 0.0 Miningetc. 0.1 0.0 Transportation etc. 0.1 0.0 0.2 0.0 3.1 Other econ. Affairs 0.5 0.1 0.1 Other expenditures - 3.7 TOTAL 0.6 0.4 0.6 0.2 12.9 1. Lending is project lending, to parastatals. "-"denotes less than 0.005 percent. 2. Parastataloperationsfor generalpublic services, defense, public order, health, social security and welfare, recreationaland other expendituresamounts to less than 0.005 percent of payments. Source: GovernmentofMauritius 2.21 Despite Governmentsupport, theparastatal sector is inpoor condition.The emergency measures taken inthe past years contributed to animprovedpositiono fthe parastatals. However, sixteen largestparastatals still posted operational losses in2000/1 and 2001/2. The historically poor managemento fthe sector i s still reflected inthe existing debts and arrear position ofthe parastatalsat the endo f 2001/2. The outstanding debt (excluding arrears) o f all parastatals stood at about 20 percent of GDP, while the outstanding debt (excluding arrears) ofthe sixteen largest parastatals at endJune 2002 stood at 10.5 percent of GDP (see annex table A.6). Inaddition, arrears for the sixteen largest parastatals stood at 1.5 percent o f GDP at the endof June 2002 (see annex table A.7). 2.22 Quasi-fiscalactivitiesweresubstantial.Apart from the direct transfers to parastatal companies, the Government continued to support the textile andEPZ sectors through Bank of Mauritius directedcredit programs (about 1.5 percent of GDP disbursedin2001/2). This subsidized credit had severalnegative spillovers as they may have reduced credit available to other sectors ofthe economy, ledto an underreporting o fthe true public deficit bythe central government, and blurredthe distinction betweenmonetary and fiscal policy which the Bank of Mauritiusis expectedto maintain. 18 2.23 Governmentdebt rose substantially in thefirst year of NEA implementation. The first year o f Agenda implementation was also markedby a substantial increase incentral government debt to 55.3 percent of GDP in2001/2002. This compareswith aplanned debt/GDP increasefrom 49.1 percent (June 2001) to 52.5 percent at the end o fthe NEA period (2005/2006). Table 2.7 shows that the stock of Central Government debt increased strongly, by 6.5 percent in2001/2. The table suggests that the risingdebt to GDP ratio is owed significantlyto deteriorating primary balances. The table also indicates that GDP growthhas helpedto suppress the debt relative to GDP; this suggests that a prolonged slowdown inGDP could have a significant adverse impacton the debt ratio, especially inthe presence of deteriorated primary imbalances. The Government financed the 2001/2 deficits by borrowing about 5.1 percent of GDP from domestic sources andabout 0.8 percentof GDP from external sources. However, central government debt increasedeven more strongly because the Bank ofMauritius issued Government of Mauritius Treasury bills as part of its open market operations to meet inflationtargets. Table 2.7 Change inCentral Government Debt: Decomposition (% GDP) 199718 199819 199910200011 200112 Stock o fdebt 48.4 49.2 50.4 48.8 55.3 Change'in debt 1.3 0.8 1.2 -1.6 6.5 Primary deficit 0.1 -0.1 0.4 1.5 2.6 Interest 3.7 3.5 3.4 4.4 3.3 Change ingov. DD .. 3.7 (-)Seigniorage 0.3 .. -0.1 0.1 0.1 (-)Privatization rev. ,. -5.7 (-)Other items 0.1 -0.5 -0.1 .. (-)Growtheffect -2.5 -2.3 -1.2 -3.3 -2.1 (-)Inflationeffect -1.9 -1.8 -2.3 -1.1 -2.5 (-)Revaluation effect 1.2 -0.7 -0.4 1.4 -1.2 Error in Prediction -0.2 -2.7 -1.4 -1.3 -1.8 Source: Bank Staffcalculations basedon Government o f Mauritius data 2.24 Three important short-termfiscal risks have emerged. At the end ofthe first year of NEA implementation, three important fiscal risks are present. First, with a large part ofthe (increasing) Central Government debt stock consisting of short-term domestic liabilities, the rollover and interest rate risk is substantial.Approximately Rs.4 billion of treasury bills are issued per monthto accommodate debt service payments. Second, a substantial credit risk stems from liabilities(both guaranteedand non-guaranteed) of parastatal companies, which have accumulateda large debt stock (around 18 percent of GDP) themselves. Lastly, a revenue risk exists as Mauritius has tappeda good part of its revenue sources at times when additional fiscal pressure is likely to arise inthe mediumterm as social expenditures are poised to increa~e.'~ l3 The possiblefiscal risk emerging from the financial sector inMauritius is being coveredbythe FSAPmission report. 19 FUTURE IMPLEMENTATIONOF THE NEA 2.25 Today, Mauritiusfaces the challengeof meeting the needfor a costly investment program necessary to underpinthe country'sfuture growth, while at the same time confronting rising social expendituretrendsowed to the existing welfare system. Mauritius mustundertake the implementationofthe Agenda without upsettingthe fiscal balances. Consequently, inimplementing the NEA, the Government has to answer three very basic and critical questions: H o w much can the Government afford to spend inthe coming. years? The Government's sustainable expenditure level will dependon the economy's expected rate o f growth, the starting level andquality o f existingdebt, tax revenues, and fiscal risks. Chapter 3 will assess the sustainable, feasible expenditure envelope, taking into account the new baseline after the first year o fNEA implementation. What should the Government prioritize? Inthe event that the Government cannot afford to spend the Rs. 50 billionthat is required to implement the NEA, itwill have to prioritize expenditures. Chapter 4 will look at the major tradeoffs the Government faces over the next several years inimplementing the NEA. H o w can the Government make good public management choices? Makingthe right tradeoffs involves having the necessary analytical and institutional capacity. Chapter 5 will examine the Government's decision-making capacity and identify areas for improvement. 20 3. A SUSTAINABLEFISCALENVELOPE OBJECTIVES OFTHE ANALYSIS 3.1 Mauritius' future success lies inthe country's capacity to transform its economy while maintaining sustainable fiscal andexternal balances. The public sector has a large role to play inthis transformation, partly because o ftradition, partly because o f the leadership o f the central government inimplementingprograms, andpartly because o fthe need to stabilize the parastatal sector. The critical questionfor Mauritius i s to determine how much its public sector can afford to spend over the next few years to implementthe NEA. Inrespondingto this question, itis importantto consider anenvelope that will comprise the entire set offiscal and quasi-fiscal activities. 3.2 This chapter develops a sustainablemediumterm fiscal framework for Mauritius, incorporating the central government and parastatal sectors. The expenditure envelope o f this framework is thenusedto assess the affordability o fthe"EA inthe nextchapter. The fiscal framework andthe resultingexpenditure envelope depend critically on the starting debt position o f Mauritius and on the growth rate o fthe economy. Other important factors are the real interest rate and the relative importance o f seigniorageinfinancing the fiscal deficit. 3.3 We adopt the approach that sustainability i s achieved when the levels o f domestic and external borrowing necessary to finance the budget deficit are not likely to leadto a debt crisis overtime and/or rising interest rates that could crowdout private investment and reduce growth, We try to determine the level o f debt-to-GDP ratio that would assure the above and look at the quality o fthe debt as well. While choosing a specific "cut-off' level for the debt/GDP ratio i s somewhat arbitrary, it does provide a useful benchmark against which policies can be assessed. STARTING CONDITIONS 3.4 Mauritius' totalpublic sector debt is relatively high. Mauritius' total consolidated public sector debt bothexternal anddomestic) amountedto about 77 percent o fGDP at the end o f June 2002.'` O fthis amount, about one quarter i s the external debt and the remaining three quarters i s the domestic debt. The central government owes almost three quarters o f the total consolidated debt, while the lion's share of the remaining one-quarter belongs to the parastatals. (Table 3.1 provides details, while box 3.1 discusses the rationale for including the parastatal sector inthe discussion). Interest payments o f the central government accounts for 21percent o f its total revenue (4 percent o f GDP). l4 Judgedby its externaldebt only, Mauritiusis classifiedas a moderatelyindebtedcountry bythe World Bank's GlobalDevelopmentFinanceReport2002, as the presentvalue ofexternaldebt relativeto Gross NationalIncome(GNI) for 2000 stoodat 56 percent. This is a few points above the 48 percentcut-offbetween less indebtedandmoderatelyindebtedcountries. 21 Table 3.1 Public Sector Debt End-June 2002 (Percent of GDP)' Domestic External Total Guar. Non-G Total Guar. Non-G Total Guar. Non-G Total Total Debt 55.1 2.2 57.6 14.5 5.3 19.8 69.5 7.5 77.4 Long 14.6 0.9 15.5 14.2 4.6 18.8 28.5 5.4 34.3 Short 40.7 1.4 42.1 0.3 0.7 1.o 41.0 2.1 43.1 Central Gov. 49.2 49.2 6.2 6.2 55.4 55.4 Long 8.6 8.6 6.2 6.2 14.8 14.8 Short 40.6 40.6 40.6 40.6 Parastatals 4.7 2.2 6.9 8.3 5.3 13.6 13.0 7.5 20.5 Long 4.5 0.9 5.4 8.0 4.6 12.6 12.5 5.4 18.0 Short 0.1 1.4 1.5 0.3 0.7 1.o 0.4 2.1 2.5 Private sector 1.5 0.0 1.5 0.0 0.0 0.0 1.5 0.0 1.5 guarantees' Long 1.5 0.0 1.5 0.0 0.0 0.0 1.5 0.0 1.5 Short Notes. Long: Long-term debt; Short: short term debt; Guar. Government Guaranteeddebt; Non-G: Government Non-Guaranteeddebt. The table excludesparastatalarrearsto centralgovernment, pensionfunds, supplies, and others, as this information is not centrally available.The arrears ofthe largestparastatalamountedto 1.4percent of GDP at the end o f2001/2. Bank of Mauritius Lines of credit for sugar and textile sectors, guaranteedby government. Includes only 2001/2 disbursements. Source: Government of Mauritius. 3.5 Mauritius' totalpublic sector debt carries considerablerisk. First, at the endof June 2002, short-term debt accountedfor 56 percent of total consolidated public sector debt. Central government owes over 90 percent ofthe short-term debt (all to domestic creditors) androlls over an average of 3 percent of GDP inshort term debt per month. With Mauritius' open capital account, domestic interest rates are subject to domestic public sector borrowing needs andworld markettrends. Should world interest rates increase (see discussion further inthis section), there wouldbeanimmediate impact oninterestpayments ofthe central government. Second, two thirds of the parastataldebt (13 percent o f GDP) i s guaranteedby the central government. Although governmentrarely has hadto meet debt service obligations of parastatals, chapters 1and 2 showedthat subsidies, capital transfers and lending to parastatalscan reach2 percent of GDP annually; therefore, giventhe fungibility of money, it would be fair to claim that government has beensubsidizing the debt service of para~tata1s.l~ Itseems that any reductionincentral government outlays to parastatalscould raise the risk that the central government would needto meet debt service obligations ofthe parastatals -or risk reduction inparastatalservices. Finally, parastatalsface a foreign currency risk: two thirds of their debt (13.6 percent o f GDP) i s external, while their revenues l5The debt service ofguaranteedparastataldebt came to about 1.5 percent o f GDP in2001/2, roughly equivalentto the totality o f subsidies, capital transfers, and lending to parastatalsby central government inthe same year. 22 are indomestic currency indicating a currency mismatchof revenues and debt payment obligations. 3.6 Growthprospects are uncertain in the medium term. Despite the success of Mauritius' economy over the pasttwenty years, Mauritius may not be able to maintain the high average growthrate of the past 20 years, 5.5 percent per annum, into the mediumand longterm as its external environment i s poised to change rapidly with the eliminationof the MFA after December 2004 andthe reductionofthe preferential treatment for sugar in2008 (as discussed inchapter 1). Indeed, eventhough the ongoing efforts to restructure the sugar sector andthe textile efforts are expectedto pay off, as are the efforts to motivate the IT sector, the current international environment poses substantiverisks to those sectors as well as to tourism andto the financial sector. Hence, a growthrate o f 4.5 percentper annum is considered for the mediumterm sustainability analysis. (Sensitivity analysis is undertaken with economic growthrates of 5.5 percent growth per annum and 3.5 percent per annum). Box 3.1 Whyshouldparastataldebt be includedin an analysis of sustainabilityof central government expenditures? Parastatalsare enterprisesowned by the central government. The main reasonto include parastatal debt inthe sustainability analysisof centralgovernmentexpenditures is becausecreditors to parastatalsanticipate that the central government will honor their obligations should parastatalsbe unableto repay them. For Mauritius, there are two additional reasons. First, the financial performance is inpart determinedbyutilitypriceswhich are regulatedby government. Second, in the past, the central governmenthas honoredparastatals'debts, hence atrack recordexists. Not taking the existenceo fparastataldebt into account would be imprudent, as it would result inan understatementof existing potential central governmentobligations. This also recognizesthat there i s a problem of moral hazardwith parastatals (as well as large private companies) in that there is anticipation that the government will bail them out ifnecessary. 3.7 Interest rates are likely to increase in the medium term. Today, interest rates the world over are low owing to the relatively poor outlook for the global economy. Short-term rates, inUS.dollars, are 1.9 percent for 6 monthLibor while interest rates on 10-yearU.S. government bonds are 4.6 percent. The prospects are that interestrates will increaseby 2.6 and 1.7 points in 10years for short- and long-term rates, respectively; similartrends are forecastedfor interest rates inEurope andJapan.l6With Mauritius' open capital account, Mauritius will have to adjust interest rates upwards as this happens. Consequently, Mauritius' short-term public debt faces substantial interest rate risk, as it gets rolled-over. This will be explicitly consideredinthe sustainability analysis, as will the possible impact of gradual shifts inthe composition ofthe debt from short- to long-term maturities. How DEBTCANINCREASE 3.8 Factors determining debt developments. Public finance contains a very simple equation which is useful to focus our discussion o f debt sustainability.l7Itposits that debt, relative to GDP, (i) increaseswhenthe primary deficit (as % o f GDP) increases, (ii) increases l6 Source:World Bank, Development Economics ResearchGroup. I' For further discussionson governmentsolvency and fiscal policy sustainability refer to Buiter, William H."A Guide to Public Sector Debt and Deficits," Economic Policy, Vol.l(November 1985). 23 when the real interest rate i s greater than the rate o f growth o fthe economy, and (iii) decreases as seigniorage i s collected. More formally, the following equation applies: Change in Primary Real Growth Debtl = deficit, + interest * Debb - rate of * Debb - Seigniorage, ratel economyl The equation is expressedrelative to GDP. Naturally, the equation can be expressed inthe reverse, i.e. primary surpluses would leadto debt reduction, and GDP growth rates greater than real interest rates contribute to debt reduction. Seigniorage enters the equation because governments can print money to finance their debt and, through inflation, undervalue the existing debt. 3.9 I n Mauritius theprojected trends in key variables are rather unfavorable. All those factors determining the long-term evolution o fthe debt/GDP ratio are currently showing unfavorable developments: the primary deficit, real interest rates, and the growth rate o f the economy. First, we have analyzed earlier that primary deficits are currently rising. Through 2000/1 this hadhappened because tax revenues had eroded (exemptions, regional commitments); inthe future, the rise inexpenditures, partially stemming from pensions (due to the aging population) and partially from NEA expenditures (on the recurrent and capital side), are likely to put additional pressure on the primary balance. Second, interest rates globally are on the rise; withthe current tight monetary policy, real rates are likely to rise in Mauritius. With about one half o fMauritius' debt being short term, rising rates will likely leadto higher interest payments very soon. Third, prospects for the economy are more moderate than previously, with rising pressure on the EPZ and on the sugar sectors, and with uncertainty ininternational travel affecting tourism receipts. 3.10 The composition of debt contains substantialfiscal risks. Mauritius' debt stock can change (increase) quite abruptly. Table 3.1 indicated that the level o ftotal public sector debt i s 77 percent o f GDP. Certain categories o fthat debt are risky inthat they can leadto the accumulation o fmore debt: first, parastatals may not be able to service some o f their debt (which could be exacerbated inthe presence o f rising oil prices or exchange rate depreciation) and may needto borrow more to cover operational losses; and second, there is uncertainty inthe level o f central government debt as the Bank o f Mauritius i s able to issue such debt independently.l8Inaddition, the central government faces directly the interest rate riskassociatedwithits short-term debt andthe exchange rate risk (indirectly inmost cases) o f parastatal debt. 3.1 1 The present section demonstrated that the components inthe equation above are likely to change ina direction that will lead debt to increase (owing to the primary deficit and interest rates increasing, GDP growing less that historical average and, and the debt stock possibly increasing). The next section discusses what may a reasonable target level of debt for Mauritius giventhe current state o f its development o fthe economy. Since October 2003, after the drafting ofthis report, the Bank ofMauritius beganto issue its own bills for the conduct of monetarypolicy. 24 A LONG-TERM TARGET CONSOLIDATEDPUBLICSECTORDEBT FOR 3.12 Some benchmark on the level o f debt and on the composition o f the debt i s needed for the central governmentto manageandcoordinate total public sector debt. Thepresent section addresses the establishment o f a benchmark for the total public sector debt while takinginto consideration important aspects o fdebt composition. 3.13 Benchmarks are somewhat arbitrary. What is especially important indiscussing a benchmark for the level o ftotal public sector debt i s that the benchmark (i) i s sustainable, (ii) notconsumetoomanypublicresources(debtservicerelativetotaxrevenues), does (iii) notcontributetoescalatingpressureoninterestrates,and(iv)limitsapotential does negative impact on the private sector (crowding out). 3.14 The composition of thegovernmentdebtportfolio is as important as its size, and usingthe debuGDP ratio as the sole indicator offiscal sustainability is too simplistic. Fiscal risks stemming from debt maturity, currency composition, or interest rate composition, as well as large contingent liabilities can have destabilizing effects. Note that the size o f government debt itself, but difficulties inits refinancing(rolling over) have triggered recent fiscal crises, such as those o fEast Asia, Russia, Turkey, Brazil, and Argentina. Not fiscal fundamentals, but market sentiment, confidence, and liquidity may become the leading factors o f fiscal difficulties. Ifinvestors suddenly refuse to roll over debt, the government mustpay higher interest rates to attract lenders. And ifthe liquiditycrunch lasts long enough, higher interest rates can damage fiscal accounts and threatento become a solvency problem. 3.15 Implications of alternativedebt targetsfor thepublic sector. The impact on the debt ratios o f alternative debt targets for the public sector (central government and parastatal) on various debt ratios i s shown intable 3.2. The table shows average debt ratios for a 10-year period consistent with a ceiling revenue o f 22 percent o f GDP and growth rates o f 4.5 percent. The table suggests that Mauritius would not face substantial problems with greater indebtedness from the external sector particularly because the share o f external debt intotal debt is small. The debt service/exports ratio would increase to a relatively comfortable 13 percent even inthe case where the debt/GDP ratio i s 100percent. However, Mauritius could face problems from the fiscal side, where debt service could consume an increasingly larger share o f domestic resources. Inthe case o f a 100 percent debt/GDP ratio, interest payments would absorb more than one third o ftotal government revenue. 3.16 For Mauritius, a reasonablelong-term debt targetfor the consolidatedpublic sector may be around 60percent of GDP. Although there exists no specific cut-off for what would be a sustainable debt rate for a country's public sector, the European Unionchose a cutoff rate o f about 60 percent o f GDP for central government debt for its accession countries who are mostly middle income or ~ea1thier.l~The average tax base o f those countries was about 30 percent of GDP in 1995. GivenMauritius' low tax base (18 percent o f GDP in 2001/2), it may be prudentto consider applying the 60 percent debt target to total public The EUdebt target includes central government, local government, and outstanding pension debt, and applies to short and long-term obligations. It does not include unfunded pension obligations or parastataldebt. 25 sector debt. Then, as Table 3.2 shows, interest payments would be limitedto 20% of central government revenues. Inaddition, Mauritiushas no outstanding explicit pension debt, so a 60 percent o f GDP benchmark may provide reasonableleeway for the government to adopt it. We proceedwitharatio of60 percent ofGDPinthe present analysis as areasonable starting point for discussion.20 Table 3.2 TargetPublic Sector Debt Stock Scenarios - Simulation of Debt Service Ratios (4.5 percent economic growth assumed)' Target LongRun Debt / GDP (YO) Inpercentage 55 60 80 100 Interest/ Central Gov Revenue 18.3 20.0 26.9 33.7 External debt service/exports 0.9 0.9 0.9 0.9 NPV of interest / Cent Gov Rev. 163.5 178.8 240.1 301.3 NPV of external debt service / exports 7.5 7.5 7.5 7.5 NPV of Dejcit / GDP -27.0 -29.5 -39.7 -49.8 'The effective real interest rate is 3.5 percent on the projections while the real interest rate on new debt that has been used is 4.5 percent. The NPV is for 10years calculation, with a discount rate o f 8 percent. Debt service assumed average amortization rate o f 0.3 for the 10-year period. For improved comparability across the debt scenarios, the fiscal account gapfill has drawn on the domestic sector. Source: World Bank staffcalculations. 3.17 I n order to decrease totalpublic sector debtfrom 77to 60percent of GDP in 10 years, the public sector would need an average primary surplus of 0.9percent of GDPfor theperiod. Table 3.3 provides an illustrationofthe primary balance ofthe public sector needed to attain convergence to a debt rate of 60 percent o f GDP in 10years, dependingon the average growth rate of the economy andthe real interest rate. The table suggeststhat with a GDP growth rate of4.5 percent (compared to the historical 5.7 percent), the public sector will, on average, needa primary surplus 0.9 percent of GDP for the ten-year period. :.: Table 3.3 Average 10-Year Primary Balance Necessary to Reduce Total Public Sector Debt ::: from 77 percent o f GDPto 60 percent o f GDP in 10 years GDP Effective real interest rate Gr. 1.5 2.5 3.5 4.5 5.5 3.5 III 0.3 0.9 1.6 2.2 2.9 4.0 0.0 0.6 1.2 ~ 4.5 -0.4 0.3 1.5 2.2 5.0 -0.7 -0.1 0.6 1.8 5.5 -1.0 -0.4 0.3 0.9 1.5 6.0 -1.3 -0.7 -0.1 0.6 Note: Effective real interest rate is based on new debt and existing o fpipeline debt; real interest rate on new debt is 4.5 percent; seigniorage set at 0.1 percent o f GDP. Source: World Bank staff calculations. 2oThe analysis implies a tax elasticity o f one; iftax revenues increase by less that the percentage increase in GDP, then the proposed debt ceiling would be too low. 26 3.18 Note that higher interest rates and/or lower growth will require higher primary surpluses ,while lower interestrates andhigher GDP growth rates will permit a lower primary balance. For example, the requiredprimary surplus would be 1.9 percent o f GDP if the growth rate is 4.0 percentandthe real interest rate4.5 percent. Note alsothat each combination o fthe growth rate andthe real interest rate inTable 3.3 would define a different fiscal pathto reachthe targeted debt/GDP ratio -and a different expenditureenvelope with implicationsfor the affordability of the NEA. 3.19 Three important tradeoffs inherent inthe sustainability analysis presented above should be noted. First, because we are working with consolidatedpublic sector debt, a reduction inthe debt/GDP ratio canbe achievedby a combination of reductions inthe debts owed bythe central government and the parastatals. For example, the sustainable fiscal framework scenario presentedlater inthis chapter shows a decline inboth central government andparastataldebt (Table 3.4). 3.20 Second, the primary surplus of 0.9 percent of GDP is an average over 10 years, which i s consistent with aprimary deficit inthe initial one or two years and a primary surplus in later years. This flexibility is essentialto meetthe needfor substantial investmentin infrastructure inthe next few years to facilitate the economic transformation the economy is going through. Table 3. 4 IllustrativeAverage 10-Year Primary Balances o fthe Public Sector Necessary to Reduce Public Sector Debt from 77 percent o f GDP to 60 percent o f GDP in 10years Illustrative Primarybalances(% GDP) PublicSector primary bal. 0.9 - - 0.9 0.9 Parastataloperational bal. 2.1 0.0 -2.1 Central Gov primary bal. -1.2 0.9 +3 .O Note: GDP growth rate at 4.5 percent, effective real interest rate at 3.5 percent, seigniorage set at 0.1 percentof GDP. Source: World Bank staffcalculations. 3.21 Third, the primarydeficit of 0.9 percentof GDP refers to the consolidatedpublic sector. That is, it is an average of the central government andthe parastatals, and can be achievedby a combination of fiscal outcome o fthe central government andthe parastatals. Table 3.4 illustrates this trade off. For example, an average 0.9 percent primary surplus could be achievedby an average 0.9 percent primary surplus by the central government and zero operational balance by the parastatals. An operational surplus inthe parastatalssector would allow the central government to runhigher deficits. Similarly, a deficit inthe parastatalsector would necessitatethe central government to achieve higher primary surplus to meet the target debt/GDP ratio o f 60 percent. The strengtho f adjustment inboththe central governmentand the parastatals sector will dependonthe options the government selects with regardto these three trade offs. 27 A FISCAL FRAMEWORK FORTHE CENTRAL GOVERNMENT 3.22 Inthis sectionthe findings ofthe fiscal sustainability analysis, which are expressedin 10-year averages, are translated into a detailed illustrative fiscal framework for the central government using annual data (Table 3.5). The resultingfiscal scenario i s broadly consistent withthe NEA'stwo mainquantitative objectives: reducingthe central government deficit to about 3 percent of GDP by 2005/6 and room for reasonable investment spending through 2005/6, spendingabout Rs. 50 billion onthe capital projects over five years. Table 3.5 Indicative NEA FiscalFramework (average GDP growth 4.5 percent): A Sustainable ProgrammableExpenditureEnvelope Actual Est. Proj. Proj. Proj. Avg. 2003- Proj. Proj. Proj.Avg. 2006- 01/02 02/03 03/04 04/05 05/06 2005 08/09 10/11 12/13 2012 (InPercent ofGDP, unless otherwise stated) Revenue 18.5 20.5 20.7 21.1 21.7 21.2 22.1 22.1 22.1 22.1 Expenditures 24.4 26.4 26.2 25.6 25.2 25.7 25.2 25.1 25.1 25.1 O/w Interest 3.3 4.3 4.0 4.0 4.1 4.1 3.9 3.8 3.7 3.9 Overall deficit -5.9 -5.9 -5.5 -4.5 -3.5 -4.5 -3.0 -3.0 -3.0 -3.0 Primary balance -2.6 -1.6 -1.6 -0.5 0.6 -0.5 0.9 0.8 0.7 0.9 Memo Debt stock 55.3 56.5 57.5 57.4 56.3 57.1 52.6 50.5 48.7 51.6 Interest / Tax revenue(%) 21.1 21.1 19.3 19.3 19.2 19.3 17.9 17.3 16.6 17.6 External debt service/ Exports ("/.I 1.0 1.0 1.0 1.0 1.0 1.0 0.9 0.8 0.7 0.8 Real non-int expend(02103 Rs. Bill) 30.1 33.4 34.9 35.4 36.1 35.5 41.5 45.5 49.9 43.6 Note: (1) The real interest rate on new debt used is 4.5 percent. Simulations with a real interest rate o f6 percent yielded a primary balancethat was about 1percentof GDP worse than that presentedinthis table. (2) Real non- interest expenditures are in200213 Rupees. Source: Governmentof Mauritius and World Bank estimates and projections. 3.23 The mainbuildingblocks ofthis scenario are the following. The reduction o fthe debt/GDPratio from the present 77 percent to the target 60 percentwill be achieved by a reductiono f both central government andthe parastatals debts. As the Chart 3.1 suggests, the centralgovernment andparastatals debts will bereducedto about 49 percent and 11 percent o f GDP, respectively. This would be achieved by a combination o f a zero operational balance for parastatals and 0.9 percent o f GDP for the primary surplus for the central govemment, indicating substantial fiscal adjustment inboth sectors. Inthe central government, the primarybalance needs to be improved from a deficit o f 1.7 percent of GDP to a surplus o f 0.9 percent o f GDP as quickly as the current conditions allow. As discussed inchapter 2, mainlyenhancing revenuecollection withsome moderation andbetter phasing incapital expenditurecouldmakethis adjustment. 28 3.24 This scenario requiresthat the central government revenueis raisedto the historical level o f about 22 percent of GDP andtotal expenditure is kept around 25 percent of GDP. In other words. the burdenof fiscal adjustment will fall onrevenueenhancement rather than expenditurereduction.*l This wouid allow 1 real non-interest expendituresto increase Ch 3.1Public Sect Debt Stocks I by about 4.5 percent per annum, at the rate 80 of economic growth. The required 70 adjustment inthe parastatal sector would 60 i be equally challenging. Itwould involve 50 initially maintainingan operational --_ _ I I I balanceinthe sector by adjusting prices to ' - Total public sect I cover the costs. Inthe mediumterm, a 30 40 Centr Gov NEA I comprehensive program i s neededto restructure, commercialize, andprivatize the enterprises as appropriate. 0-0 2001/02 2005/06 2009/10 2012113 3.25 Itis possible for the central Source: Governmentof Mauritius and World Bank staff government to own debt higher than 49 projections percent o f GDP ifparastataldebt is further reduced. This couldbe achievedby selling off some of its equity inparastatals or by selling individualparastatalsthat are ingood standing, such as the State Investment Corporation, with its holdings ingambling andtourism sectors, andusing the revenue to retire the parastataldebt. Although such a strategy was followed in2000/1 whenMauritius Telecom shares were sold, equivalent to about 5.5 percent o f GDP, it would be currently advisable only as an integral part of the fiscal adjustment package.22 3.26 Note that fiscal adjustment of the central government follows a gradual path. The primary balancedoes not become positive before 2005/06. This meansthe debtlGDP for the central government would increasea couple of percentage points initially before it converges to 52 percent. This gradualism is necessaryto allow the government to undertakethe necessary investmentinhuman andphysical infrastructure inthe coming years to facilitate the ongoing transformation ofthe economytowards to a "high-tech andhighvalue added- services andknowledge" economy. 3.27 Thepotential impact of high-riskparastatals. Chapter 2 discussedthe significant role o fparastatalsinthe economy. The table 3.6 presentsthe projected debt service o fhigh riskparastatalsonthe stock ofdebt available as ofJune 2002. Shouldthe central government be requiredto pay this debt, that would reducethe expenditureenvelope available to implementtheNEA by about 1percent o f GDP a year for the mediumterm. The impact of suchadditional expenditureson the resources available for other programs forms the basisofthe expenditureanalysisofchapter 4. "Chapter4discussesoptionsforraisingrevenues. ''Mauritius has infact set up the privatization find inthe mid-1990s and borrowed to meet investment needs against future privatization revenues. The fund was closed and its activities were integrated into the budget in 2001/2. 29 Table 3.6 Projected Debt of Selected Parastatals -Highest Risk to Central Government; Stock of debt as of June 2002 200213 200314 200415 200516 2008109 2010111 2012113 Debt Stock 4.1 3.3 2.4 1.9 0.8 0.2 0.1 Interest 0.3 0.2 0.2 0.1 0.1 0.0 0.0 Principal 1.4 0.7 0.6 0.4 0.2 0.2 0.0 Debt service 1.7 0.9 0.8 0.5 0.3 0.2 0.1 Memo Primarybalance 1/ -3.3 -2.5 -1.3 -1.1 0.6 0.6 0.6 Expend.Envelope (real 02103 Rs Bill) 21 30.8 33.5 34.1 35.3 40.9 45.1 49.7 Notes: 11Primary balance o ftable 3.5 after transfers equivalent the debt service o f high riskparastatals has been accounted for as subsidies. 21Expenditureenvelope o f table 3.5 after transfers equivalent the debt service o fhighrisk parastatals has beenaccounted for as subsidies. Source: World Bank staff calculations. 3.28 Additional debt incurredbythe high-risk parastatalsfrom 2002/3 onwards i s excludedfrom the exercise; this meansthat the debt service associatedwiththat newdebt is paid out of operating surplusesthrough price adjustments or efficiency gains. The high-risk parastatalsinclude CEB, STC, DBM, MHC, NHDC; refer to Annex 3 for details. CoNcLusIoN 3.29 The analytical work inthe present chapter suggests that the originalNEA target fiscal deficits are sustainable but that Government will likely not be able to finance planned expendituresandmeet deficit targets at the same time. TheNEA framework formulated in June 2001 provided for an overall fiscal deficit o f about 3 percent of GDP by 2005/2006. Assuming that this is attained and such deficits maintained, the outlined target of a consolidated public sector debt of about 60 percent would be attained in 10years. A debt target of 60 percent would implythat about one fifthof revenueswould be usedfor interest payments. However, the originally planned expenditurelevels o fthe Agenda, especially the doubling of capital investmentsuntil2005/2006, are not consistent with attaining such deficit levels, Inpart, this stems from the expectedslower growth rate o fthe economy and the likely need for central Government to bail out severalparastatalcompanies inthe next years. 3.30 The fiscal adjustmentnecessary for gradual reduction of fiscal deficit to about 3 percent o f GDP should come largely from enhancing revenue collectionrather than expenditurereduction. The medium-termfiscal framework indicates that total central government expenditure can be maintained at the historical level o f 25 percent o f GDP, but the revenues have to be increasedfrom the present 20 percent to 22 percent of GDP. The government has already taken steps inthis direction. The rate o f VAT has beenraisedtwice inthe pasttwo years. The follow uprevenue measureswouldinclude streamliningtax exemptions and other tax incentives andimproving the tax and customs administration. 3.3 1 The next chapter looks at the nature of the tradeoffs facing the central government would have to face shouldit maintain the proposed fiscal framework. 30 4. POLICY OPTIONS 4.1 The previous chapter addressedthe questiono f "how much government spendingi s affordable to Mauritius?" and explored the implications o f different debt levels on public finances. Using a sustainable path ofindebtedness and an assumption of improved tax collection, the chapter found that a reduction o f the fiscal deficit to about 3 percent o f GDP by 2005/06, as was assumed inthe initialNEA baseline, would provide for an expenditure envelope amounting to approximately 25 percent o f GDP. 4.2 The objective ofthe present chapter is to examine the policy options available to Mauritius inexpenditure, revenue, and debt management that would enable Mauritius to implement the NEA withinthe framework explored inchapter 3. Onthe expenditure side, we employ a linear projectiontechnique and findthat (i) room for capital expenditures is the significantly smaller (by about one half) compared to programmed investment spending; and (ii) currentexpenditureenvelopeismoreinlinewithprojectedspendingbutthatsome the prioritization and savings would have to take place as well. Giventhese results, the chapter analyzes a number o f sectors that offer options for the Government to achieve some o fthe needed savings. Onthe revenue side, we explore the set o f revenue options that would enable Mauritius to meetthe revenue objectives inthe sustainable fiscal framework. The critical issue i s the extent and speedwith which Mauritius wishes to part with the past practice o f segmentingthe economy. Similarly, interms o f debt management, we look at potential institutional improvements that could mitigate risks to the sustainable expenditure envelope. 4.3 This chapter is written under the premisethat there is no single recommendationor solution to the public finance challenges that Mauritius faces. There are, however, many policy options that deserve to be explored closely for their (i) consistency with the NEA objectives o freorienting andmodernizing the Mauritian economy, (ii) efficiency inmeeting those objectives, and (iii) priority standing givenbudgetary constraints. The present chapter aims to illustrate some existing policy options and to point to potential tradeoffs. The chapter seeks to pose the type o f questions that the Government is likely to face inthe future and i s intendedto suggest areas for further analysis. EXPENDITURES Baseline projections 4.4 We use simple linear projections for the different expenditure categories to derive the overall likely expenditure level (given current trends) and its composition. We can then compare the aggregate spendingestimate to the sustainable aggregate derived inchapter 3. Further, we canthen also mapthe NEA plannedexpenditures (especially for capital goods) to the available resource envelopes. As such, the exercise does not represent the medium- term Budget o fMauritius. The exercise aims to examine potential expenditures that are likely to materialize and look at the type o f choices facing Government. For the baseline 31 expenditure projections, we use two alternative classifications o fthe budget-the economic and the functional classification. 4.5 Expenditure projections are based on historical trends andforecasted expenditures for sub-categories, butthey do not take into account the planned NEA investments. Table 4.1 provides the resulting projections o frecurrent and capital expenditures based on this "no policy change" scenario. The table incorporates the existing expenditure categories found in Mauritius' budgetary tables for 2001/2 and 2002/3 together with projections for sub- categories. Assumptions underlyingthe projections included: (i) paymentsare based interest on rising interest rates globally; (ii) wages assume a 15 percent wage increase for civil servants; (iii)"goods and services" expenditures are increased to meet the higher recurrent costs resulting from higher investment; (iv) pensionoutlays are calculated based onthe recipient population; and (v) subsidiesinclude government coverage o fthe debt service o f highriskparastatals (presented intable 3.6). The projections do not, though, include the specific capital investment plans ofthe NewEconomic Agenda. 4.6 The projections suggest a need for Table 4.1 restrainingtotal expenditures. The average Breakdown of the Indicative Sustainable annual expenditure (calculated on the basis of Expenditure Envelope (2003/4-5/6) past trends and likely future demands) for the (As % o fGDP) period 20031'4 - 2005/6 could reach 27.0 Ann. Actual Estimate Avg. percent o f GDP. Table 4.1 shows that this 03104- spending level (which does not yet take 01/02 02/03 05/06 account o fthe specific NewEconomic SUSTAINABLE Agenda investment plans) i s highcompared to EXPEND. ENVELOPE 24.4 26.4 25.7 the sustainable envelope of 25.7 percent o f GDP. The exercise assumes an average PROJECTED TOTAL EXPENDITURESONA annual capital expenditureo f 4.3 percent o f "NO POLICY CHANGE" GDP (less than the approximate 6 percent of BASIS 24.4 26.4 27.0 GDP pew annum assumed inthe fiscal PROJECTEDCURRENT - 20.4 20.8 21.3 framework o ftable 2.2) and requiredsavings Interest 3.3 4.3 4.1 of about 1.3 percent o f GDP from the Wages & salaries 6.4 6.1 6.7 recurrent side. Goodsand services 2.2 2.1 2.3 Subsidiesandtransfers 8.4 8.3 8.2 4.7 Table 4.2 extends this analysis. First, PROJECTEDCAPITAL I/ 3.7 - - 4.3 5.0 itindicates that anaverage capital investment Acq. o f fixed capital 2.8 4.0 3.3 of 4.3 percent o f GDP for the three year Purchase of land 0.1 0.1 0.1 period 2003/4-2005/6, i s equivalent to about Capitaltransfers 0.9 0.9 0.9 12.9 percent of GDP for the three year period; NET LENDING - 0.4 this is muchlessthanthe programmed(not - - 0.6 0.1 yet implemented) NEA capital expenditures o f SAVINGSNEEDED 0.0 0.0 -1.3 21.7 percent of GDP (discussed intable 2.1). This suggestthat the plannedcapital outlays 11Breakdownof capitalexpenditureswas basedon original budget shares. of the NEA would have to be considerably Source:Governmentof MauritiusandWorld Bank staff phased andprioritized. Obviously, it would estimatesand projections. be possible for Government to increase capital spending beyond the 4.3 percent o f GDP used 32 as a baseline for the projections. Then, however, equivalent savings would haveto be made on the current expenditure side sincethe overall expenditure envelope i s going to be exhausted. Table 4.2 Capital ExpenditureProjections for a Sustainable Fiscal Envelope (as a percent o f GDP) Remaining 3-year 3-year Project average' total' Balances Total Capital 4.3 12.9 21.7 Generalpub. services 0.4 1.3 1.3 Defense 0.0 0.0 0.8 Public order 0.2 0.6 1.4 bducation 0.8 2.3 2.3 I Health 0.2 0.6 1.4 SOCsecurity, welfare 0.1 0.2 1.o pousing& Comm. 1.7 5.0 6.2 I Recreational 0.1 0.5 0.9 Fueland energy 0.0 0.0 0.0 Agriculture 0.2 0.5 1.3 Miningetc. 0.0 0.1 0.0 /Transport. & comm. 0.4 1.3 4.2 I Other econ. services 0.1 0.4 0.8 Local government 0.0 0.1 0.0 Other 0.0 0.0 0.0 1 Linear Projection of Capital Expenditures200314- 200516(by composition of200213 Capital Budget). Basedon GDP growth of 4.5 percent. Remaining Project Balancesin2002/3 Capital Budget for 200314 Onwards(as appear in200213 Capital Budget). Source: Bank staff estimates. 4.8 Second, table 4.2 takes the assumption o fthe average capital investmentof 4.3 percent of GDP for the three year period2003/4-2005/6 and applies to it, again for illustrative purposes, a breakdown using the public investment composition ofthe 2002/03 budget. Table 4.2 includes a comparison betweenthe current composition o fpublic investment spending andthe remaining NEA costs. The table suggeststhat the resulting capital expenditure spendingpattern i s not inline withNEA flagship areas. The transport and communication sector, for example, would require an additional 4.2 percent of GDP in order to implementthe remaining investments, whereas the current resource distribution impliesanallocation amounting to only 1.3 percent of GDP (over three years). Similarly, additional NEA expenditures inhousing andcommunity amenities come to 6.2 percent of GDP, whereas the actual distributionwould imply a 5.0 percent o f GDP share. 33 ExpenditurePolicy Options 4.9 Taken together, the previous analysis suggests that Mauritiuswill face some important tradeoffs interms of choosing whichof the Agenda flagship capital investments will be financed over the coming years. These tradeoffs not only concernthe phasing and prioritization within capital expendituresbut also betweencurrent and capital outlays. This section will provide anumber of expenditurepolicy options for the Government inorder to helpincreasethe room for prioritizedcapital expenditures. Outlaystoparastatals (utilitiesand importingmonopoly) 4.10 Historically Mauritius has made outlays to parastatalsthrough avariety of channels, namely, direct subsidies, capital transfers, and lending. Simpleprojections of outlaysto parastatalsbasedonpast experience (relative to GDP), coupled with the potential impact of fiscal risks, suggest that Mauritius could spend as muchas 3 percent of GDP a year on parastatals(for the period2003/4 - 2005/6).23 This i s high given the fiscal consolidation that will haveto take place. The Government will needto examine the situation andeconomic rational for eachparastatalseparately, considering inparticular the rational for continued state support. Below are descriptions oftwo ofthe more prominent parastatals. Central Electricity Board 300 - Chart4.1Cost of EnergySourcesand AverageSalesPrice 4.11 The Central ElectricityBoard 250 - of Electricity(2992=D0) (CEB), responsible for generation(about 50 percent of total production), transmission, and distribution of electricity, built up a large debt mainly as .. a result of lossesincurredinthe 1990s, 50 whenelectricity tariffs were very 92 93 94 95 96 diesel -fuel 97 98 99 00 01 oil Chart 4.1 presents indexesof the average - - - ..-.coal -electr import price inRupees of energy sources (diesel, fuel oil, and coal make up about 70 Source: CSO-Energy Statistics percent o fthe energy sources that produce electricity) andthe average price o f electricity (for all uses). Since 2000, tariffs have beenrevised upward, allowing the CEB to generate small operating surpluses. It is, however, unlikely, that future electricity tariffs would be raisedto such a level as to enable the CEB to service its debt.25While structural reforms inthe power sector canreduce the risk of further accumulation of government contingent liabilities in CEB, success of these reforms may actually require the government to take over the CEB's presently outstanding liabilities.26 23Excludestax expendituresand receipts (equity paymentsand repaymentso f loans). 24The CentralElectricityBoard also serves as a guarantor inatake-or-pay scheme. This scheme provedto be relatively costly, having anegative impact onthe Central Electricity Board's operating balance. 25Inrecentyears, the CentralElectricity Boardhas defaulted on governmentloans. 26Once a decisionhas beenmade as to how to addressCEB's debt (including the pensiondebt), thenthe direct cost to the governmenthas to be factored into the sustainability analysis. 34 4.12 The Governmentmay wish to assess the overallpolicy of maintaining below market electricityprices. This can be done by (i) allowing CEB to adjust tariffs to meet costs, not and (ii) subsidizing CEB with guarantees, debt service, and capital transfers. InMauritius, about 25 percent of households usedabout one third o f electricity production and accounted for about 37 percent o f value sold in2001. The remainingtwo thirds o f electricity produced was usedalmost evenly by commercial and industrial sectors, paying an average price o f 20 percent above and 30 percent below the household price re~pectively.~~Giventhe losses at CEB,all three categories o f electricity users (households, commercial, and industrial) are beingsubsidized to some extent, butthe industrial sector has receivedthe largest subsidy. The Government may wish to examine the cost, employment, and incentive implications o f raising electricity prices to better match costs, as the fiscal sustainability analysis has suggested that long-term support for all users o f electricity i s not likely to be feasible within the current set o fNEA priorities. Consequently, the governmentmaywishto articulate a concrete planfor support to some sectors for a pre-determinedperiod o ftime andto transparently allow the budget to carry the subsidy. 4.13 Government may also wish to consider the equity impact of usingtax revenues to subsidize electricity for households. With 75 percent o f households usingheating sources other than electricity, it i s likely that the upper quintile o f households benefits from the subsidythe most. Ineither case, a more transparent approachto electricity policy wouldbe to quantify the entire impact o fthe policy on the budgetand the benefits it brings to specific categories o f stakeholders inorder to assess the alignment o fthe present subsidy system with the country's strategic priorities underthe NEA. Inessence, the existing outlays to CEB directly compete for fundingresources against top priority sectors. State Trading Corporation 4.14 TheState Trading Corporation (STC) has an importing monopoly over oil, rice, flour, cement, and, since 2002,LPgas. It also maintains price subsidies for these commodities. A widening mismatchbetween international and domestic prices, most notably petroleum andpetroleum by-products, inrecent years gave rise to large losses under STC's trading operations. The cost o f the subsidy on rice and flour has beentransparently covered from the government budget within each fiscal year.28 On cement and oil trade (which constitutes more than 90 percent o f STC's business), however, STC's losses have not beenrecovered from the budget. These losses reached Rs.1.1 billion, Rs.2.1 billion and Rsl.4 billion, respectively, infiscal years 1999/2000,2000/01,and 2001/02. As a result, as o f June 2002, STC had debt amounting to Rs.3.4 billion outstanding, of which Rs.1.2 billion i s guaranteed by the government (Rs900 million of foreign credit and Rs.180 million o f domestic bank overdrafts). 4.15 Thegovernment may wish to investigate the rationalefor subsidizingoilproducts.29 As with electricity, subsidies o foil products indirectly subsidize the manufacturing (36 27 The commercial sector obtains almost 90 percentof its energy from electricity and the remaining 10 percent from LP gas. 28 The rice and flour subsidy amountedto 0.3 percentof GDP in2001/2. 29 The present discussiondoes not addressthe rationale for the rice and flour subsidy (0.3 percent of GDP in 2001/2). Interestingly, the rice and flour subsidy is transparentlypresentedinthe budget, and fully covers costs 35 percent o f energy used comes from fuel oils) andtransport sectors (99 percent o f energy used comes from fuel oils, almost evenly from gasoline, diesel, andjet fuel), as well as other sectors indirectly because about 45 percent o f all electricity generation inMauritius comes from diesel and fuel oil. The government should investigate the rationale for subsidizing oil products by answering three questions: (i) Transparency -what are the total costs o fthe price controls on oil products? The costs would have to include quantification o f guarantees, debts, direct subsidies, transfers, and lendingto STC. (ii)Affordability, equity, efficiency, andprioritization-at a time when fiscal consolidation i s required, andthere are important expenditure priorities in education and other sectors, are the current beneficiary sector of the subsidies the best use o fthe limitedpublic resources? (iii)Incentives-giventheneedtorestructureMauritius'economyinthemedium term, are the incentives provided by the current policy the right ones and are they permanent or temporary? 4.16 The Government may also wishto reconsider the rationale for maintaining a parastatal holding an import monopoly. Alternative methods o f organization, such as private imports with a regulator or a consortium o fprivate importers, could perform the task under the watchful eye o fgovernment. Pensions 4.17 Today, Mauritius has a three-tiered pension system that helps the poor and provides moderate replacement income for retirees and no regulatory protection for voluntary retirement schemes. The unfunded nature o fthe universal scheme, together withthe income maintenance scheme o fthe civil service, are exerting serious fiscal pressure. Today, public sector outlays for pensions (the Basic Retirement Pension and the civil servants scheme) account for about 4.2 percent o f GDP, absorbing about a quarter o f recurrent expenditures and a fifth of recurrentrevenues. Without Government action, the forthcoming demographic transition will be very costly. Expenditures for the basic pension alone are expected to be close to 5.9 percent o f GDP by 2020. Civil service pension outlays have already increased by 50 percent inreal terms between 1992 and 1998 alone; the civil service pensionbill, as a share o fthe wage bill, can be expected to increase from the current 20 percent to 30 percent over a period o f 15 years -prior to the introduction o f the salary increase in2003/4. 4.18 Mauritius has several options for modernizing the universal pension scheme. The first option would be to rise the retirement age. This would respondto the higher life expectancy of the population and would be compatible with evolving employment patterns. The second option, albeit controversial, would be to convert the current universal scheme into a targeted scheme for the poor while at the same time undertaking a one-time upwards adjustment in the pension. This would (i) the fiscal risk stemming from universality o f the scheme reduce to STC. The Government may wish, however, to evaluate the efficiency o fthe rice and flour program and to assess ifthat approach, at a cost o f 0.3 percent o f GDP in200112, is the most efficient way to deliver assistance to Mauritius' poorest citizens. 36 and (ii) more effective protection from poverty to those inneed. The thirdoption provide would be to leave the universality o f the systeminplace but to reduce benefits. This would be the least efficient measure, butitwould serve to reduce expenditures while respecting possible constraints. 4.19 The civil service pensionwould also need to be adjusted. Based on the World Bank's projections, pensionexpenditure for the civil service scheme will rise steadily, from the current level of 20 percent o fthe coveredwage bill to around 30 percent in 15 years time, and will reach 50 percent by 2050. The total implicit pensiondebt (for liabilities accrued to date) exceeds one third o f Mauritius' GDP. Given their limitedcoverage (slightly over 10 percent o f the labor force), the public sector pensiondebt lays a sizeable claim on Mauritius' scarce public resources. 4.20 While these benefit levels are not excessive when compared with neighboring countries inAfrica and South Asia, they are considerably higher than civil service benefits in OECD countries. The highpensionbenefits may reflect the traditional back loading in public sector compensation, andthe not too competitive overall compensationpackage for civil servants. However, this type o f compensation structure is inefficient because it often fails to attract qualified young staff, has a questionable incentive structure, and impedes the mobility betweenthe public andprivate sector. This calls for a review of the overall pay structure and its benchmarking against the private sector as a precondition for a comprehensive reform o fthe pension scheme. Housing 4.21 The National HousingDevelopment Company (NHDC) andMauritius Housing Corporation (MHC) mainly serve the government inimplementing housing assistance programs. Assistance inthe form o f interest rates subsidies i s funded from the government budgetwithin eachfiscal year (fully, inthe case o fNHDC andpartially inthe case o f MHC).30 By June 2002, NHDC had liabilities outstanding inthe amount o f Rs2.1billion, or 1.5 percent o f 2001/2 GDP, out o f which is Rsl.3 billionwas guaranteed by the government -Rs900millionfromforeignsourcesandRs400millionfromtheNationalPensionFund. M H C hadRs4.9 billion o f outstanding liabilities, or 3.6 percent o f 2001/2 GDP, o f which Rs2.7 billion are guaranteed, including Rsl.7 billion borrowed from the National Pension Fund. MHC estimates that about Rs700million o fits debt, or 0.5 percent o f2001/2 GDP, is associated with the cost of past interest rate subsidies that hadnot been covered by the budget. Overall, with the default risk onmortgage financing low (around 6 percent) and good collateral enforcement, the companies appear financially sustainable. Risk may arise particularly for NHDC if it embarks on ambitious real estate development (for instance beyondthe present scope o fthe o f Cyber City and Cyber Village) and suffers from insufficient demand. 4.22 A large disequilibriumremains in the low-income housing market. Despite the significant amount o f resources that have been invested inthe sector, the diagnosis carried 30Infiscal year 200112, the two corporations received0.3 percent ofGDP indirect subsidies. 37 out by the Bank a few years ago31,suggests that there is a large disequilibriuminthe low- income housingmarket, withtens ofthousands of applicantswaiting to obtain a housing unit. At the same time, there is evidencethat householdswithvery low incomes (below Rs.3000) are not being reachedthrough these programs. 4.23 There is little rationale for having a low income-housing program ifthe intended beneficiaries cannot be reached. The medium-termpriority should be to eliminate public interventionon the supply side altogether-moving out ofpublic construction andprovision, This would call for dismantling NHDC as aninstitutionfocusedonpublic construction of housing. Instead, efforts should focus onthe demand side andon tailoring financing mechanismsto the specific needs of the very poor. 4.24 The Government may wish to rethink the rationale behind the broad-based housing policy. Mauritiusmay wish to undertakea detailed analysis o fthe beneficiaries ofthe current housingpolicy; Mauritius may also wishto assess the necessarychanges inoperation policies for NHDC and MDC with a view to obtaining positive operating balances. Itmay separatelywishto consider the solutions ofthe outstanding debt ofthose parastatals, and examine whether it is equitable to meet those debts from tax revenues or by passingthe costs to the beneficiaries of the housing loans. Operationalization o fthe government's newvery low income housing program (introduced in2001/2), could continue to be financed transparently, as is, underthe capital budget. Any costsresulting to the private sector from implementing that policy (including NHDC andMHC) would also have to be coveredby the budget. Education and health care 4.25 Theprovision of free services has enjoyed reasonable success. Mauritius has adoptedpolices of free education and free health care for its population for a longtime. In effect, those polices have resultedinannualpublic and private expenditure amounting to about 5-6 percent of GDP for the past 20 years.32 The expenditure has been spentto provide mostly primary and secondary level services inboth sectors. Results have beenreasonably good given Mauritius' past development accomplishments, inthat, ineducation, enrollments inprimary andsecondaryeducationare high; while inhealth, communicable diseasedhave beeneliminated. 4.26 Education and health demand is likely tofurther increase. Mauritius' development strategy clearly rests on buildingup its human capital base so the country can better compete internationally. Inthe short term, this means aggressively investing ineducation infrastructure andtraining/acquisition o fteachers andtrainers to buildup an internationally competitive delivery base, starting from primary and secondary, but also extending into tertiary education. Mauritius needs to improve quality andmarket relevance. Similarly, inthe mediumandlongterm, maintaining the education andtraining delivery system(inline with marketneeds) will be criticalto preserve aneffective life-long learning strategy for the labor force. At the same time, on the health side, Mauritius i s faced with an aging population (the 31Source: WorldBank Aide-Memoire andPolicyNote: Rethinkingthe Welfare State (February2000). 32Over the pastten years, this has beensplit with approximately4 percentpublic and2 percentprivate. 38 share o f the population over 60 will increase from 9 percent in2000 to about 16 percent in 2020), and an increasing burdeno fnon-communicable diseases, such as diabetes, cardiovascular diseases, alcohol-related diseases, and injury. Both factors can be considered consequences of the development success o fMauritius. 4.27 Expenditures can be expectedto rise. The New Economic Agenda formulates the formidable challenges inboth the education andhealth sectors as evidenced by the poor level o freadiness of Mauritius' labor force to move to higher value added (and more knowledge intensive sectors) and by the country's aging population and surge innon-communicable diseases. Both sectors will require improvement insecondary and especially tertiary level services. The per capita cost o f services inthese sectors i s usually muchhigher than primary level services. For education this involves laboratories, more specialized trainerdfaculty, more costly learningresources (books, CDs tapes, etc); for health care, this involves more sophisticated technology, more skilled personnel, andmore funding for recurrent expenditure such as maintenance and training. This would also require a different skills mix, meaning that Mauritius will eventually have to increase its doctor to populationratio which currently stands at about 1:1,138, which is lower than most industrialized countries. 4.28 Public sector interventions in education and health need to be rationalized. Iti s quite likely that the government will continue to heavily fundthe educationandhealth sectors inthe future. However, inthe presence o f budget constraints, the critical issue i s to use the public fundingfor maximum impact interms o f equity and efficiency considerations. (i) Despite the increasingimportance o f the health and education sectors for the country's development and the population's well being, funding for the two sectors i s not unlimitedinthe short or mediumterm. Consequently, aggressive development programs must be accommodated by precise plans for cost saving. (ii) The Government may ask its citizens to pay for an increasingshare o f tertiary level education and health services.33Fundingo f these sectors from general taxation i s an explicit policy choice with redistributive implications. Ineducation, itimpliesthat the general public is payingfor tertiary level services, which is possibly inequitable as it i s mostly the better-off that benefit from those services. Inhealthcare, the issue is abitdifferentas market failures cancausethe poorer high-riskmembers o fsociety to be excludedfrom group insurance. Therefore, government has to assure that the poor will have access to a given level o ftertiary services inhealth, while better-off groups could well be demanded to share in their individualtertiary level costs. (iii) Thepublic sector's role may bymost efficient as a regulator for assuring a good mix o f sector efficiency andquality o f services, andguarantor o f service availability for the poor. The Government may want to explore how the private sector can increasingly become providers o f education andhealth services as well. 33The MauritiusUniversityofTechnology is makingheadwayincostrecoveryas user fees reportedlymeet70 percentofrecurrentcosts. 39 Table 4.3 Illustrative Trade-offs Betweenand Within Sectors ILLUSTRATIVEBENEFITS POTENTIAL 2001/2 Number of % Increase % Increasein SAVERS Expenditures yearly inNon- road (% GDP) teachers' salary maintenance salaries education budget covered spend. Parastataltransfers (equivalent to halfof 3-year average) 1.o 5,732 400% 340% Increase OR Increase Debt service ofhigh riskparastatals Housing - subsidy on interest ratesto MHC andNDHC OR Approximate 0.3 1,673 120% 100% Increase pensions savings Increase (basic pension indexation to wages insteadof `past practice' Halfofpost- 0.25 1,433 100% 70% Increase secondaryspending Increase Bank o fMauritius lines o fcredit to I.1 6,305 440% 370% Increase textiles and sugar Increase Notes: ' Refer to table 1.3. Not in 2001/2 Budget; Refer to table 3.6. Background work to Mauritius: Modernizing and Advanced Pension System, World Bank. 2001. Government of Mauritius budget, 2002/03 Not in 2001/2 Budget. Capital costs of the LRT. Not in200112 Budget. Assuming an annual salary of239,000 Rupees(Education officer Grade B); excluding civil servant pension costs. Government o f Mauritius budget, 2002/03 Mauritius Transport Sector Action Plan, World Bank 2003 An explicit look at tradeoffs 4.29 Illustrative inter-sectoral tradeoffs. This section looks at what potential savings in one sector imply for program financing inanother. Table 4.3 presents illustrative trade-offs betweensome expenditure itemsandteachers salaries, non-salary education expenditures, and road maintenance. For instance, saving o f about 1percent of GDPwhich could come from a halving oftransfers to parastatals or reduced responsibility for the management of debt of highriskparastatals couldpay for the salariesof 5,732 teachersina singleyear, pay for a quadrupling ofthe non-salary non-contributive budget o fthe Ministry ofEducation and Scientific Research(teaching materials, labs, etc), or increase the roadmaintenance budget threefold (almost completing the 5 year objective inroad maintenance) for about 20 percent of Mauritius' roads which are inpoor quality and needmaintenance and drainage work to 40 prevent deterioration. The table suggeststhat small adjustments inexpenditures can have large payoffs inareas where there are large needs. It also suggests that it is not only inter- sectoral changes, but intra sectoral tradeoffs needto be explored. REVENUES 4.30 Mauritius' tax model is outdated, creating inefficiencies in the economy andposing fiscal risks. Mauritius' tax model aims to mitigatethe economy's natural anti-export bias (distance, lack o fraw materials) by having differentialtax rates (and thus subsidizing) export industries. Mauritius also protects import-competing industries. The tax system has multiple exemptions for indirect taxes, especially import taxes or the VAT, and tax holidays for corporate taxpayers. Ithas evolved into a`system o f a narro6 tax base but with highrates. At the same time, the revenues generated bythe tax system are low for anupper middle income country. The inefficiencies are generated because (i) tax system artificially biases the resource allocation within the economy, and (ii) itgrants unsustainable advantages to export industries, undermining Mauritius' capacity to develop long-term viable sectors. These two shortcomings leadto further inefficiencies because new, emerging sectors will also need exemptions, thereby further reducingthe tax base o fthe economy. The fiscal risks are posed bythe fact that (i) revenues are constantly under threat o f beingreduceddue to additional exemptions, and (ii) inthe absenceofknowing the true levelof tax expenditures, the effective prioritization o fpublic resources i s undermined. 4.3 1 I n thefirst instance, it is necessaryfor Mauritius to quantify tax expendituresto assess true costs of its support to theprivate sector. Tax expenditures refer to revenue losses from a "normal" tax base owing to deductions, exemptions, exclusions, preferential rates, deferrals, and credits. These losses are direct expenditures out o fthe budgetand should be subjected to the same scrutiny as the rest o fthe budget, namely estimation o f costs, scrutiny by the legislature, andbe subject to evaluation. 4.32 I t is also importantfor Mauritius to assess the appropriatenessof the current tax structure vis b vis its developmentobjectives. First, the multiplicity of exemptions accompanyingsuccessive investmentschemes (there are 42 such schemes presently in Mauritius) suggests that there i s room for re-aligning Mauritius' tax structure. On successive occasions, Bank and IMF reports have emphasized the needto (i) the correct incentives (by givingroom for new industries to grow that would be globally competitive) and (ii)raise revenues.34 4.33 Mauritius should be able to raise its tax base while maintaining the "low tax" character of its economy. How can this be accomplished? Mauritius has several options for raising revenues: (i) broadenthe tax base for import duties by reducing exemptions and improvingtax administration (strengthening collection, audit, and reporting to revenue departments);(ii) broadenthe tax base on corporate taxpayers who are paying an effective tax rate of 15 percent compared to the statutory rate o f 25 percent; (iii) incorporate more individuals into the tax scheme (in2001 it was estimated that 30,000 Mauritians were paying 34World Bank CountryEconomicMemorandum; 1995;WTO, TradePolicy Review: Mauritius; October 2001; IMFArticle IV Report, June 2003. 41 income taxes out o f a labor force o f 500,000); and (iv) increase tax rates o f indirect or direct taxes. DEBTMANAGEMENT 4.34 The central governmentcurrently lacks a comprehensiveapproach to managing public sector debt. Mauritius' debt management currently consists o fmanagingthe central government's cash flows and diversifying exchange rate risk on the limitedexternal borrowing. Parastatal borrowing, guaranteed by the central government, i s subject to review by the Ministry o fFinance. The present practice has resultedinexcessive short-term debt (which may be less costly than medium-termdebt but carries substantial roll-over risk, and hence more expensive inthe long run) and the absence o f risk as a quantifiable benchmark for decision-making. It exposes Mauritius excessively to the fluctuations infinancial markets. Mauritius does lack a comprehensive perspective o f acceptable borrowing levels andacceptable risks o fthe public sector debt portfolio. 4.35 ..and it does not control the issuance of all central government T-bills. Chapters 2 and3 illustratedthe Banko fMauritius practice o f issuinggovernment T-bills which amounted to almost 4 percent o f GDP inJune 2002 sitting as idle cash. This practice represents a fiscal risk even ifthe central government i s compensated for the interest it pays on this debt. The risk stems from the fact that the central government lacks control over the issuance o f debt for which it i s accountable. The practice i s also not advisable because the central government is bearingthe cost o fmonetary policy. 4.36 Mauritius has severaloptionsto reduce the cost of borrowing. The options require institutional improvements indebt management, which would be part o f designing and implementinga new debt management strategy to meet more complex borrowing needs, adopt practice to international standards, and further develop its economy. Mauritius could: 0 Develop a strategic approach to debt management. A strategic approach to debt management i s neededfor Mauritius to ensure that public sector debt (even at the 60 percent o f GDP or lower) i s sustainable andcan be serviced under a wide range o f circumstances. To guide the future composition o f the government debt portfolio, it may be beneficial for the Government o f Mauritius to establish portfolio benchmarks related to the desired maturity structure, duration, and currency composition o f the debt. To assess risk, Mauritius' debt managers could regularly conduct stress tests o f the debtportfolio on the basis o f the economic and financial shocks to which the government, andthe country more generally, are potentially exposed. Insetting its objectives for public sector debt management (and government debt management specifically), the government will have to decide where to draw the line inthe tradeoff between risk and cost. For example, although lengthening the maturities o f borrowings may imply higher expected debt servicing cost (assuming an upward sloping yield curve), the risks associated with short-term debt andthe possible disruptions relatedto their realizationmake longer maturities muchmore attractive. Since risk management capacity inthe Government Debt Management Office as well as inparastatals i s only emerging, it would be wise to emphasize the objective o f reducing risk rather than minimizingexpected cost. (Most OECD countries pursue 42 specific risk and cost objectives and emphasize that the degree o f risk must be prudent.) 0 Take control o f the issuance o f all central government debt. The Government o f Mauritius needs to take control o f all debt issued on its behalf inorder to implement a debt management strategy. This means that the Bank o f Mauritius would needto issue debt it own debt to conduct monetary olicy. This would make the conduct and costs o fmonetary policy more transparent. R 0 Lengthenmaturities and deependomestic bond markets. This could reduce the government's overall borrowing cost as well as reduce the risk o f sudden increases in the cost o f debt service. Moreover, this wouldmake government access to debt financing more reliable, ensuring o f a muchwider pool o f investors. International experience suggests that the development o f a government bond market i s instrumental for the development o fthe overall domestic bond markets. Deeper domestic bond markets would then make it easier for parastatals to manage their debt and reduce volatility intheir future debt service levels (and intheir access to debt financing).36 0 Government bond markets can reduce government exposure to refinancing and other financial risks ifthey are deep and liquidby providing readily accessible domestic financing. Moreover, well-developed government bond markets also allow governments to reduce their debt service costs by lowering liquidity premiums embedded inthe government bonds' yields.37 4.37 Reducingthe long-term cost o f borrowing would free up resources underthe sustainable expenditure envelope to implement priority NEA tasks. Adoption o f risk parameters would helpto stabilize the cost o fborrowing and ultimately make debt services more predicable. CONCLUSION 4.38 Mauritius has many options for implementing a sustainable expenditure envelope. It can prioritize expenditures, raise revenues, or manage its debt better. It can also pursue a combination o f any or all o fthese options. 4.39 Onthe expenditure side, Mauritius is facing budgetary constraints that require expenditure prioritization. Using expenditure projections, this chapter has shown that the sustainable expenditure envelope will likely only be able to finance halfo f plannedNEA capital outlays andthat some cuts will also have to take place for current expenditures, albeit 35See footnote 18. 36Since July 2003, the government has increased the share o f issuance o f long term securities. A draft debt management strategy has been worked out with the assistance o fan external consultant, and implementation has been initiated. 37Since September 2003, the Bank o f Mauritius has sold government securities on the stock exchange o f Mauritius. 43 to a lesser extent. Interms of expenditure policy options, the chapter outlined potential savings that could materialize from a change inthe wage policy, lower parastatal subsidies, better targeted low-income housing support, a reformeduniversal and civil service pension, and cost recovery intertiary education andhealth care. 4.40 As a next step to identifying expenditure priorities, the Government may wishto assess the extent to which the current set o f expenditure programs correctly reflects its strategic priorities. This would include (i) identifyingthe policy objectives currently clearly supported by the budget and by quasi-fiscal activities; (ii)undertaking a transparent and comprehensive costing o f existing policies; and (iii) identifying the beneficiaries o fthe support programs. Such a general, backward looking program reviewwould also be a key ingredient for developing a MediumTerm ExpenditureFramework. It would also allow for costs to government (subsidies) to be more transparently reflected inthe budget. 4.41 Onthe revenue side, Mauritius may wishto broadenthe tax base for direct and indirect taxes. Itmay wish to quantify tax expendituresand to develop a more even-handed tax policy betweenexporting and import-competing sectors. 4.42 On debt management, Mauritius needsto develop a debt management strategy that takes into account central government and parastatal debt, and also establishes debt level and riskbenchmarks. The government may wishto try to reduce risk rather thanminimize expected costs. It would need to buildup its capacity to implement this new strategy. Mauritius may also wish to keep central government debt issuance as a prerogative to the central government and allow the central bank to benefit from the issuance o f its own debt. It may also wishto benefitfrom the deepening o fthe secondary market for government paper. 44 5. REFORMINGECONOMIC MANAGEMENT INMAURITIUS 5.1 A primarygoal of the Government ofMauritiusis to achievethe development objectives set out inthe NEA, by realigning resources with priorities and at the same time reducing the size o fthe fiscal deficit. The Government i s also focusing on improvingthe efficiency and effectiveness ofpublic resources. To achieve these goals, the Government has begunthe process ofintroducing a MediumTerm ExpenditureFramework (MTEF) as the key tool to plan and managepublic resources. At this moment there exists sufficient political commitment to launch anMTEF inMauritius, and substantial desire at the senior civil servant level to increase the efficiency and effectiveness o fthe government. Accordingly, the MTEFis intendedto enablethe government to strengthenthe focus onresults andto improve service delivery, while increasing the efficient use of public resources. This will be achieved by moving from a focus on inputsto assessingoutcomes inline with the NEA's strategic priorities, following the principles of results-basedmanagement (RBM). 5.2 From a cross-sectoralpoint ofview, sound economic managementhas three dimensions. First, itrequires the existenceof an adequate framework to plan, manage, and monitor the implementation o f Governmentpolicies and the allocation o f resources. This framework should support strategic decision-making andthe prioritizationof activities inline with overall Governmentpolicies. These Government priorities mustbe adequatelyreflected inthe Government budget, andbudgetimplementation shouldbemonitoredagainstthese targets. Second, sound economic managementrequires an adequateinstitutional set-upto deliverthe implementation ofGovernment policies andthe allocationofresources. Strong coordination and collaboration as well as the existence of consultative mechanisms are important elements for ensuringthat Government policies are plannedand implementedin accordance with key priorities. Third, economic management requires adequatecapacity in the public administration. This capacity must exist at a technical level to translate Government priorities into specific, time-bound activities and monitorable outcomesto enable Government to assess the implementation o f its policies. Technical capacity i s also important to evaluate implementationprogress inpriority areas. Management capacity i s requiredto make necessaryadjustmentsinthe course ofthe implementation of Government policies. Inaddition, compensation packages and incentives for civil servants needto match capacity. 5.3 This chapter analysesthe current status of economic managementinMauritiusby identifyingthe strengths and weaknesses ofthe institutional andprocedural set-up. Itputs forward a number o frecommendations for improvingthe economic management framework inorder to facilitate the movetowards results-basedmanagement withinamediumterm expenditure framework. 45 ECONOMIC MANAGEMENT INSTITUTIONSINMAURITIUS 5.4 Ingeneral, the institutional framework inMauritius provides a goodbasis for effective planning and management o fpublic resources. The current institutional framework consists of: The Ministry of Finance is responsible for overall budget and expenditure management 8 The Public Finance section plans andmonitors the budget; The Finance sections at the Ministryo fFinance (i) request for the process reallocation o f funds from line ministries for the recurrent and capital budget; 8 Staff o fthe Finance cadre are allocated to the line ministriesto administer the line ministry budget (keep accounts, process payments, andmake entries). 8 The Office o fthe Accountant General acts as the Government's cashier and prepares the Government accounts; e The Internal Control cadre undertakes internal audit functions; and 8 The ManagementAudit Bureauprovides internal management consultancy services to the Government by assisting inorganizational reviews, management audits, and strategic planning (mainly to parastatals). The Ministry of Economic Development contributes to the appraisal o f government projects, the formulation o f sector polices, and formulates Mauritius' overall development strategy. Inthat context, the ministry prepares and monitors the Public Sector InvestmentProgramme (PSIP) and also reviews sector policies and macroeconomic planning.3 The Ministry of Civil ServiceAffairs andAdministrative Reforms is responsible for the designand implementation o f civil service reforms, the review o f conditions o f services, andthe formulation o f human resource management policies. Sector ministries are responsible for planning, implementation, monitoring, and reporting their sector projects and activities. Seconded finance and internal control staff inthe sector ministries have a key role inthis context. The NationalAssembly debates and approves the budget estimates presentedbythe Minister o fFinance. TheAudit Of#ce is responsible for the external audit o f Government accounts. The PublicAccounts Committee o fthe National Assembly reviews Government accounts and produces an annual report setting out issues andrecommendations. 38InDecember2003, Mauritius mergedthe functions ofthe MinistryofEconomic Development with those of the MinistryofFinanceas part of its efforts to better coordinate policy formulation, budgeting, andmonitoring and evaluation. 46 ANALYSISOF THE MAURITIAN ECONOMIC MANAGEMENT FRAMEWORK 5.5 This section analyses Mauritius' economic management framework witha view to assess those components which are essential for an efficient MTEF andRBMprocess. The mainprinciples behindthe MTEFframework are: (a) the budgetis comprehensive, (b) the mediumterm fiscal framework precedes andunderliesall budgetdiscussions, (c) Government (Cabinet) buildsconsensus onthe macro framework and the expenditure priorities, (d) monitorable progress on program performance i s a pre-requisite exercise to new budget formulation discussions, (e) contestability inthe allocation process duringbudget formulation is necessary, and (f) transparency inall o fthe above steps is necessary. Figure 5.1 illustrates a `skeleton' process inwhich the macro framework i s initially prepared (box A), sectoral ceilings are set bythe Cabinet basedon governmentpriorities (box C), butalso based on analysis o f how well a particulate sector's budget was implemented inthe previous year and whether progress inmeeting program objectives was met (boxes B, G and H), rolling three-year sectoral ceiling are set by Cabinet (box D), andthe national assembly discusses and approves the next year's budget, and discusses the indicative medium-term allocations and programs (box E). The reader may notice that monitoring and evaluation enter every aspect o f the budget cycle, and are made use o f by all participants inthe budget process (boxes G,HandB). Figure 5.1: Features of a GeneralizedMTEF r It. TOP DOWN: MINISTRYFINANCE, I 1 OF CABINET, AND NATIONAL ASSEMBLY .t I IiI Macroeconomic Ceilings Cabinet NationalAssembly Review of ~ ~~ Framework Ministry ceilings Debatesceilings Approvesfirst year progress by developedbasedon basedon estimates,two forward centralministries bottomup Govemment years indicative to feed into requirements priorities settingo fceilings t I t F for next year Sector Review Managing and Reporting on Ministriesdefinetheir Ministries finalize Results objectives, results indicators, estimatesbasedon Ministriesuse information on activities,andcosts for three- approved ceilings for resultsto improve year periodbasedon policies three-year period managementandreport andpriorities progress 47 Macroeconomic andfiscalframework 5.6 One o f the prerequisites of a goodmediumterm planning and expenditure managementprocess is the capacity of Government to develop a sustainable mediumterm macroeconomic andfiscal framework whichtakes into account likely trends inthe global economy, forecasts of the national economy, current levels of debt, and the important fiscal risks. The assumptions behindthe macroeconomic and fiscal framework, as well as the explicit choices (i.e. revenue andexpenditurepolicies, borrowing andfinancing polices) of the framework itselfcan be subject to a healthy debate, internally andoutside government. This is particularlycrucial for small openeconomies suchas Mauritius, which contain inherent risks due to the amplifiedimpact of small changes inglobal factors, such as volatility inenergy prices, climatic conditions, or shocks to the world economy (exchange rate movements, wars, etc.). The Government's role on the macro and fiscal framework i s two-fold: first to develop consensus around a sustainable rollingthree-year framework as part of budget preparation, and second, to reassess and updatethe framework during the year under apredictable schedule. 5.7 A criticalfeature ofthe fiscal framework i s comprehensiveness:the framework needs to incorporate all of government's actual expendituresandtax expenditures. The needfor a comprehensive budget is two-fold. First, inincorporating all expenditures, the policymakers assure that all conceivable expendituresand risks of unscheduled expenditures or leakages are taken into account, and the government's sustainability analysis i s on stable footing. Second, inincorporating all expenditures,the integrity ofthe prioritizationexercise is established, as at the time o fprioritization, all trade-offs for the use o f limitedfiscal resources (revenues and grants) will be subject are beingexplicitly considered. For Mauritius, the necessityof the fiscal framework to be comprehensive was demonstrated in chapters 3 and 4: the existence of substantial quasi-fiscal activities has the capacity to derail macro-economic stability andthe implementation o f sectoralprograms. Consequently, the fiscal framework mustinclude a comprehensiveaccount o f quasi-fiscal activities andmust also budget for them. (Box 5.1 and annex 4 provide information). 5.8 The Government ofMauritius preparesathree-year macroeconomic andfiscal framework and, since 2002/3, publishesthe broad aggregates inan appendix to the budget. The technical work underlying the framework is done bythe Macroeconomic Coordinating Committee (MCC), which is a technical group that meets at once or twice a year to support the MinistryofFinance's preparationof the annual budget. The MCC consists of representatives ofthe Ministryof Finance, the Ministryo fEconomic Development, Financial Services, and Corporate Affairs, the Central Statistics Office, and the Bank of Mauritius. The MCC's composition andmodus operandi,of iterative meetings inwhichall sectors participate, are a good example of intra-government collaboration. However, the MCC suffers from several shortcomings. First, its does not always meet duringthe year to update the macro framework. Second, its analytical methods are not very robust and uniform across sectors. Finally, the MCC itself i s not responsible for the development or the updating of the macro framework; it isjust responsible for gathering anddiscussing inputs to the framework -Finance has the ultimate responsibility for putting the macro framework together. The latter issue points to abroader problem for Mauritius, that the there i s no high-level Government entity (such as anInter-MinisterialCommittee or anEconomic Committee of Cabinet) that 48 systematically assessesmacroeconomic trends or develops a coherentpolicy response to macroeconomic events onbehalfof Government. The presenceof such a commission would be im ortant inview of the macro and fiscal challengesthat Mauritius faces inthe short term. 8 Box 5.1 Incorporating fiscal risks into the MTEF Inits standardform, the MTEF tacklestheproblemoffiscal opportunism acrossbudgetaryprograms,by requiring policy makersto analyze and disclosethe assumptionsand expectedmedium-term (3-5 years) implications of their budget proposal and to be accountablefor any departures from their medium-term targetedlevels inyear-by-year decision-making. The standardMTEF aims at promoting government performance at three levels -fiscal discipline, equity, and efficiency - under government budgets. With respectto fiscal risk, MTEF should aim to strengthen government fiscal performance across non-budgetaryas well as budgetary activities of fiscal nature and with respect to both contingent and direct sources of fiscal risk. An expandedMTEF would promote a risk-awareness culture in government and would integratefiscal risk deliberationswith govemment policy-makingand with fiscal anddebt managementprocesses. Reporting. To ensure proper attention to their possibleimplications, fiscal risks needto be made known, Since no cash is spent from the budgetwhen governmentsassumes a contingent obligationor provides a tax exemption, cash- based accounting systems fail to detect suchfiscal risks. Transparency o f risks needsto be requiredinthe markets as well as in govemment. Inaddition, govemmentneeds capacity to monitor and analyze its risk exposurevis-his implicit sources o f fiscal risk. As apart ofmonitoringits risk exposures, government needs to monitor behavioral determinantso f fiscal risks. These include performance(moral hazard) under its guarantee contracts, legality of claims againsttax duties, andperformanceo f sub-nationalgovernmentsor other entities that may affect the government's future fiscal position Comprehensivecosting of programs: including budgetary and non-budgetary activities. Mauritius uses a broad array o f instruments to achieve programobjectives which range from budgetary expendituresto a variety of quasi- fiscal activities price ceilings (coupled with subsidiesto parastatals)andtax exemptions. Inorder to be able to effectively prioritize its policy choices andto managerisks, the government needs to have a comprehensivecosting of programs in operation (especially a costing -and risk assessment-of quasi-fiscal activities). Development of baseline cashprojections and stress testing. Government medium-term fiscal planning is hardly viable if it does not reflect implications of fiscal risk exposures. The best way to show medium-term implications of fiscal risks for the overall fiscal position is to develop baselinemedium-termcash projections and stress-test these projections with respectto specific risks(taking into accountpossible correlations). Similarly, equity and efficiency considerationsare relevant only ifthey addressthe whole range o f fiscal activities andtheir expectedand maximum likely costs. Stress testing, that is estimating sensitivity o f expectedfiscal cost to normal and abnormal changes inthe underlying assumptions, is critical for developing a sound medium-termfiscal strategy. Budgetingforfiscal risk. Contemporaryapproachesto budgeting for fiscal risk are basedonthe following principles: Recognizethe risksand full potential fiscal cost of guaranteesand other promises of contingent government support prior to their approval, as part o f the budget process. Apply ajoint ceiling for the cost of budgetaryand off-budget support for each sector in a fiscal year. Off-budget support is a form of subsidy, which should be consideredalong with other forms of govemment support for each sector. Calculate the cost o f off-budget support as the present value ofthe future expectedfiscal cost. Reflect the cost of off-budget programsin full within the fiscal year o f their launch. Transfer an amountequal to the off-budget support from the budgetaryenvelope o fthe related sectorto a central contingency reserve fund. Source:Annex 4 ~~ 39 InNovember 2003, ahigh levelinter-ministerial committee, the Economic Coordination Committee (ECC), was established to decide on the policy agenda, review progress, and identify constraints to delivering the Economic Agenda. The issues on the ECC agenda are: unemployment, the budget deficit and public sector debt, investment, and inflation. The ECC also discusses and reviews the macroeconomic and fiscal framework prepared by the Macroeconomic Coordination Committee. 49 5.9 Regardingthe comprehensiveness o fthe fiscal framework, Mauritius has also made headway. There i s a good degree o ftransparency with respect to government explicit contingent liabilities, helping to ensure the comprehensiveness o f the fiscal framework and transparency of fiscal risks. The list o f government guarantees i s publicly available, with a full list o fbeneficiaries andlenders under government guarantees andthe amounts outstanding denominated inthe domestic currency as well as the currency o f origin. However, degree of government risk exposure emergingfrom guarantees, and the factors that affect whether or not a guarantee will be called are, however, not discussed. As for other risks, their transparency i s uneven. The Bank o f Mauritius' annual report provides detailed information o n the portfolio of government debt. The report, however, does not discuss the associated government risk exposure and its possible impact on f h r e debt service cost. In addition, quasi-fiscal operations conducted through public corporations and their potential future fiscal cost are only discussed occasionally. For example, the government has reflected on the potential fiscal risk emergingfrom the domestic price caps on electricity, water, and oil products inits 2000 report, "The Present State of the Economy." Inaddition, annual reports o f public corporations provide some insights into their fiscal function and finances. Finally, tax expenditures are not consistently reported. There i s a good example, though, on which the government can buildinthis regard-"The Present State o fthe Economy" report presents the estimated budgetary cost o f new tax concessions geared toward the electricity, water, andtransport sectors to alleviate the accumulation o f liabilities o f related public corporations. Linking ceilingstopriorities 5.10 Sound economicmanagement requires comprehensivecosting of programs and priorities ... Chapters 2 and4 have demonstrated that Mauritius uses a broad array o f instrumentsto achieve program objectives. The instrumentsrange from budgetary expenditures, a variety o f quasi-fiscal activities price ceilings (coupled with subsidies to parastatals), andtax exemptions. Inorder to be able to effectively prioritize its policy choices, the Government needs to have a comprehensive costing o f programs inoperation (especially a costing - and risk assessment -o f quasi-fiscal activities). This will reveal the "true" costs o f a program, and to juxtapose it against the "true" costs and risks of other programs inorder to prioritize activities. 5.11 ..and budgetary ceilings closely linked to clearly defined nationalpriorities. The Government must be able to make and enforce trade-offs between sectors that are consistent with spendingpriorities. This shouldalso include mechanisms to determine expenditure requirements duringthe execution phase o f the budget. 5.12 I n Mauritius, national objectives werepreviously defined in national plans, which have been replaced by theNEA.The current process o f setting budgetary ceilings involves an assessmento f the macroeconomic forecasts, agreement onoverall recurrent and capital expenditures,and allocation o fresources betweenministries based on priorities. The capital budgetceilings are based on mid-year implementation assessments o fcapital projects. As part o f this process, lower priority projects are dropped from the budget. The ceilings are consolidated ina Cabinet report, which outlines the overall macroeconomic situation, targets, and broad issues for reallocations o f resources between sectors. The ceilings are thenissued 50 to all ministries inthe BudgetCircular that sets out overall Government objectives and macroeconomic targets as well as detailed instructions for the Recurrent and Capital Budgets. The Budget Circular for the 2002/3 budgetincluded a section on MTEF's introductionand required ministries to submit three-year expenditure estimates. Sectoral planning andfocus on results 5.13 Effective sectoralplanning is keyfor the developmentof a realistic budget. The central feature of an effective planning process includes a detailed analysis of the existing policies and their costs, as well as a review o f other policy options that may enable Government to achieve policy objectives more effectively. It i s also important to plan holistically for the entire sector or for all o fthe line ministry's activities inorder to evaluate ifactionsareconsistentwithGovernmentpriorities. Box 5.2 Public-PrivateInitiativesand Public ExpenditureReformin Mauritiusand SouthAfrica The Government of Mauritius is currently designing two potential Private Provision o f Infrastructure (PPI) projects. One, which is in the early design phase, will focus on the delivery of potable water. The second, which is in a more advanceddesign stage, is an energy project that will finance the construction and operation of a power station to supply electricity to the CEB Grid. The water project may be able to have complete cost recovery from users, although there may be some level of Government subsidy to increase access, make services more affordable, pay for positive externalities, or finance public support during a transition period to cost-covering tariffs. Similarly, the electricity project will commit the Government to long-term payment obligations to the private power producer, some or all of which may be recoveredthrough end-user tariffs. The possibility o f long term fiscal exposures and the creation of potential contingent liabilities in these two projects alone suggests that there is an urgent need to make explicit the link between the PPI program and public expenditure reforms. Some countries have recognized this link, making PPI programs explicit and important components of their broader public expenditure reform programs. The South African Government, for example, has established a specialized PPIUnit in its National Treasury, thereby ensuringthat the ministry with overall fiscal oversight is also responsible for driving implementation of PPI. Infact, PPIhas been brought into the domain o f the country's medium term expenditure framework (MTEF), and sector ministries are expectedto demonstratehow risks are to be managed, performance measured, and value for money maximized in their large infrastructure projects and programs if they are to be approved in their MTEF budgets. Increasingly, this has meant that sector ministries have had to incorporate PPIoptions for infrastructure projects at the planning stage, many of which have moved into procurement and implementation phases as PPI models. Interestingly, as the sector ministries have got more comfortable with the approach, private participation has expanded from traditional infrastructure projects to include social programs inthe health and education sectors. InMauritius, the establishment of the Public-Private Partnership (PPP) Unit, the formulation of a PPP Policy, and the drafting of new PPP legislation are important steps in managing these risks and potential exposures. However, the Government still faces severalchallenges on this front. Key amongthese are ensuringthat: (i) all PPIprojects, regardlessofthe sector/ministry or Government agency from which they emanate, adhere to commonprinciples of sound fiscal management, inthe context ofproper long-term planning; (ii) theMinistryofFinance,asthe ministrywith overallfiscal oversight, plays aprimary, early role in assessingthe affordability and value for money ofPPIprojects; (iii) allpotentialPPIprojectsareconsideredandplannedinthecontextofoverallbudgetplanningcycles;and (iv) sector ministries embark on PPI projects not as a means to circumvent existing fiscal discipline and constraints on borrowing, but rather to improve the reach, efficiency, and cost-effectivenesso f public servicedelivew. 5.14 Theplanningprocess at the sector level in Mauritius differs in scope and coverage between sectors.A number o f sector plans exist (i.e. education); some ministries have 51 producedmaster plans for specific sub-sectors; others, such as the Ministryof Health, have produced draft action plans or developedpolicy documents on specific issues; and many ministrieshave defined their vision andmission statements.Thus, while some sectoral planning processes are inplace, the approachis not systematic andvaries interms of quality. Some of the main shortcomingsof the sectoralplans include: Itappearsthat individualsectorplansdo notcoverthe entire scope ofaministry's activities. For example, anumber of different ministries are involvedinthe cross- sectoral Environmental Strategic Plan. However, individual plans o f line ministries seldom take this engagementinto account. Similarly, the medium- and long-term costing implications of strategic partnerships with the private sector (inwater and electricity provision) are not fully assessed at the sectoral level (Box 5.2). Line ministriesare able to develop policy proposals andprojects for approval by Cabinet outside the budget process. This poses the risk that the financial implications of these policies and projects are not sufficiently analyzed. Costing of the plans is incomplete -without fully costing the proposals, ministries may implement activities without a clear understanding of the plan's affordability and effectiveness to achievethe stated objectives. Focus on results is weak. There are only rudimentary feedback loops to enable policy decisions basedon implementation progress. Inaddition, there is a common focus on inputsandtracking ofactivity completion for investment projects, butoutcomes and intermediate outcomes are not systematically identified, nor relevant tracking mechanismsestablished.However, an important pilot exercise is taking place within the Mauritius Civil Service (see Box 5.3). Box 5.3: Pilot Public Sector Reformsinthe Mauritius Civil Service The Government o f Mauritius is piloting arangeof public sector reforms, which are set out inthe 2001-2003 action plan entitled"Towards the Modernization o f the Public Service." The reforms focus on five areas: (i)performancemanagement:introduceanewperformanceappraisalframework; (ii)human resource management/development: review the existing human resource management framework, develop appropriate HR systems, review service schemes, and identify areas for capacity building; (iii)reengineeringandrestructuring thecivilservice: review the processofissuinglicenses, carryout organizational reviews of ministries, identify opportunities for alternative methods o f service delivery, and proposeareas for corporatization where appropriate; (iv) quality management: develop approaches for customer-focused public service, design and implement Total Quality Management framework, implement Citizen's Charter, and develop framework for Counter ServiceAwards and Individual Excellence ServiceAwards; and (v) financial management: review the current Financial Management manual; identify gaps in store management, asset management, and procurement systems; identify the need for technical improvements to financial management framework; formulate guidelines for management of inventory of assets; recommend improvements to the budgetary process; ministries/agencies produce annual reports covering use o f funds and outputs, gradually gearing towards performance measurement; and identify areas for training o fpublic officers in the field of financial management. These initiatives must be streamlined inorder to achieve consistentoutcomes and maximum efficiency. The Government needs to capitalize on the MTEFprocessas an entry point that will provide an opportunity to align and appropriatelysequencethese reforms. 52 Expenditureplanning and approval 5.15 I n Mauritius, the recurrent and capital budgets areplanned separately.The proposals for the capital budget are basedon the current status of capital projects' implementation andprojected requirementsfor the forthcoming year. Thus, the focus of expenditure planning is predominantly on individualprojects, rather than on abroader context aimed at achieving Government objectives. 5.16 There is also littleplanningfor the recurrent budget, despite thefact that its value is more thanfive timeshigher than the capital budget. Within the recurrent budget, budget planning i s primarily focused onthe estimation of personnelcosts, with ministries having to present information such as newposts, retirements, andpromotions. The personnel budget i s preparedon the basis of requests for additional personnel, whichhave to be approvedbythe Ministry of Civil ServiceAffairs prior to inclusioninthe budget.However, thisplanningis not undertaken ina systematic manner to assess the numbersand skills requiredinorder for the ministry to achieve its statedobjectives. 5.17 Once ministries receive the Budget Circular in early February, they have about three weeks toprepare and submit their estimates - which does not leave sufficient timefor detailed analysis or the developmentof policy options. Some ministriesstart budget preparationprior to receiving the budget circular. The budget estimates are reviewedinthe Ministryof Finance. A teamofofficers attheMinistryofFinanceis responsible for reviewing all o fthe Government's capital projects. This team has information on the current implementation status of eachproject and is able to makerecommendations to the Director of Public Finance on the amounts that should be allocated to eachproject. Hearings are heldin April andMay. The recurrent estimates are discussedwith the participation ofthe Ministry of Civil Service Affairs at the level ofthe Estimates committee. This i s followed by discussions o fthe capital estimates. The representatives from the Ministryof Economic Development attendthe hearings on capital estimates .40 5.18 There is no standardformat for preparation orpresentation of projects, nor do projects undergodetailedpre-evaluation such as a costhenefit analysis. Ministries develop newprojects, andproposalsare submittedto the Ministry ofFinance for approval andthento Cabinet for further approval. As noted earlier, these projects can be submittedfor approval during the entire fiscal year; consideration o f newprojects i s not part o fthe review ofthe overall expenditure framework and allocation priorities. 5.19 Thepresentation of the annual budget in the budget documents, inparticular the separation between recurrent and capital budgets, does notfacilitate external analysis. The budget is presentedby the Ministerof Finance inthe Budget Speechand i s debatedinthe NationalAssembly. After two-three weeks o fpublic debate of the budget document ina plenary at the NationalAssembly, the Committee of Supply (from the NationalAssembly) discusses the provisions ofthe budget for each individualministry. However, the national budget currently does not include any performance information, so discussions ofministry allocations are made inthe absence ofpastperformance data. 40See footnote 38. 53 Public expenditureexecution 5.20 An effective budget implementationsystem wouldguarantee that the Governmentis able to control overall expenditurelevels, and also ensure thatfunds are allocated to priority areas during implementation of the budget. Thus, rather than funds being allocated to projects on an individualbasis, ministrieswould submit a quarterly work planfor all projects inorder of priority, and funds would be allocated andusedinthis order. Inaddition, an effective cashmanagementsystemwould enable Government to manage its cashposition andto assure a close link betweenfiscal andmonetary policy. Thus, Government needs to reviewcurrent cashbalancesagainst expenditure andrevenuetrendsto ensure that there is no need to resort to unnecessary domestic borrowing ifthere are either sufficient idlebalances or ifGovernment expenditurescouldbe rephasedso as to coincide with revenuecollections. 5.21 Once the National.Assemblypasses the budget, a warrant i s approved for the recurrent budget's implementation for the entire year. Ministries have to preparetheir requirementson a quarterly basis, whichare updatedevery week andusedas expenditures control by the Office ofthe Accountant General. The Accountant General's office acts as the Government cashier, managing its bank accounts and cash flow; it also accommodatesthe line ministryrequest for funds by requestingthe Bank of Mauritius to issue government bills. Some ministries preparevouchers and sendthem to the Accountant General for processing andpayment, while others are self-accounting which meansthat they are able to issue checks andfunds are transferred into their accounts bythe Accountant General to cover these payments. 5.22 TheAccountant General's office recently introduced a state-of-the art computerized accounting system using Oracle Financials, which hasforced the Governmentto update expenditureinformation. The Public Finance Section of the Ministry of Finance can now access informationinthe accounting systemat any time. During the year, Ministriesare able to reallocate funds betweenrecurrent items of expenditure and capital projects as long as the total budget i s not exceeded. Overall compliance with public finance rules andregulations is strong. Managing, monitoring, and evaluatingprogress 5.23 I n Mauritius, reporting and monitoring systemsfocus predominately on expendituresandphysical implementationof capitalprojects. The Ministryo fFinance obtains expenditure (andrevenue) data on a daily basis through the treasury accounting system. Regarding the physical monitoringofproject implementation, ministries are requiredto submit quarterly reports to the MinistryofFinance for the purpose of monitoring of the budget. No system exists for reporting on outcomes or intermediate outputs. Some ministriesholdregular meetings to reviewproject implementationprogress. However, this type of monitoring i s not developed systematically (with standardizedinformation requirements),whichwould allow for identificationof bottlenecks or readjustment of expenditureplans. 54 5.24 Internal audit i s undertaken by the Internal Control Cadre (ICC) o f the Ministry o f Finance, which since 2000 has posted officers to the sector mini~tries.~' These officers prepare reports for their accounting officers on specific issues such as levels o f overtime expenditures. These assessmentsand reports are not produced on a regular basis. The Ministry o f Finance's central Internal Control Unitmonitors their work and i s planning to produce an annual report summarizing the issues arising fiom their work. 5.25 The Government's Central StatisticalOffice (CSO) couldplay an important role in monitoring outcomes. CSO complements monitoring efforts from sector ministries through its ongoing data collection efforts (specialized surveys and censuses), data analysis, and publication o f results. Often, sector ministries pass their own monitoring informationto CSO for inclusion inthe national statistical database. However statistical data are not always effectively presented and some ministries feel that the data collected do not sufficiently meet their needs. To respond to these demands, CSO has developed a customer service model to improve supply andpresentation o f i n f ~ r m a t i o n . ~ ~ 5.26 Overall, evaluation of projects andprograms is weak in Mauritius today.Evaluation generally takes the form o f financial audits. Few examples o f engineering and quality control assessments for major capital projects exist. Similarly, there are rare examples o f cost effectiveness studies, although one has beenconducted inthe Ministry o f Health. Budget Oversight 5.27 The Office of theAuditor General is responsiblefor undertaking ex-post audits of Governmentaccounts and isplanning to introduceperformance audits. Government accounts are also scrutinized by the National Assembly's Public Accounts Committee, although the Assembly does not have a role inassessing budget proposals and therefore can only comment on actual expenditures rather than on expenditure plans. 5.28 Generally there appearsto be limited systematic review orpressurefrom civil society on budget and service delivery matters. The reviews o fthe overall budgetfrom civil society include (i) the Joint Economic Council's assessment o fthe budget focusedprimarily on macroeconomic targets andbroad expenditure issues; (ii) the Mauritius Commercial Bank's more detailed assessments o fthe Government's macroeconomic performance and targets; (iii) the Mauritius Business Magazine's publication o f extensive features on the budget; and (iv) reviewsby some consultancy and accounting firms Nevertheless, the lack of clear links between Government objectives (performance indicators at the sectoral levels) and expenditure allocations inthe annual budget document makes it difficult for civil society andjournalists to track whether policy objectives are beingmet andwhether expenditures are being allocated inline with the stated objectives. NGO attentionto budget relatedissues appears to be therefore limitedto specific interest groups. 41The ICC reportsto the supervisingofficersofthe line ministries. 42The Governmento fMauritiushasadheredto the GDDS ofthe IMF and is implementing improvementplans. The CSOperiodicallysurvey's lineministriesfor dataneeds. 55 Capacityfor EconomicManagement-Stayingthe Courseon theBig Picture 5.29 Capacityfor economicmanagementhas shown signs of stress. Mauritius' past success ineconomic managementhas restedontwo factors (i) government -private strong sector coordination primarily through the JEC but also through other private sector institutions, and (ii) coordination onmacro polices betweenthe Ministryof Finance (for fiscal stability and annual budgeting ), the Ministry ofEconomic Development (for the formulationand monitoring of five year plans), andthe Bank o f Mauritius (for inflation control and exchange rate management). However, the challenges that Mauritius is facing today, notably (i) the mismatch betweenlabor skills and development needs and (ii) the persistently highdeficits -albeit necessaryto overcomepoor investment decisions of the past, both suggest that the economic management ap aratushas not been functioning as desiredperhaps as far back as the midto late 1990s:' At the core was the absence of a `big picture' assessment ofthe progress of Mauritius, despite the fact that at a `micro' level, most projects that were incorporated inthe annual budgetswe progressing according to plans. 5.30 I t therefore seems that a reassessment of policymaking institutionsneeds to be undertaken. This would involve reassessinghow inputs from the private, civil society, and government analysts, are assessed, the means by which those inputs are developed into policies andprograms by government, andthe meansby whichperformance against benchmarks i s assessed. Although inpractice, the country's well establishedpolitical processmakes these determinations through elections andpolls, inreality, a muchsmoother, more predictable, andmore institutionalized processi s neededfor assessing economic management. 5.3 1 TheMinistry of Economic Development has developed capacityfor sectoral analysis, but remains underutilized. The Ministry of Economic Development i s the agency that has the capacity for sectoral analysis. However, inpractice, the ministry does not assume a commensurateresponsibility, either at the time of budget formulation or at any reviews o fpolicies or programs. Although 5-year plans had been produced by the ministry in the past, those plans, and the accompanying public investment programs have fallen by the wayside. The ministry does maintain sector economistshpecialists, who are responsible for sector policy analysis and review, have little input inthe policy development that occurs at the line ministry level. At the same time, the majority ofthe sector ministries lack skills and capacity inthe area of policy analysis, monitoringand evaluation, and results based management. Inthe past, economists were secondedfrom the Ministryo f Economic Development to sector ministries, however this practice has stoppedinrecent years. 43Two separate developmentoutcomeshelporient the discussion: first, the prolongedunder investment in educationhas beenoccurring sincethe early 1990s; and second, the fact that tax revenueshave beeneroded sincethe mid 1990s (under the pretextofregional integration, but without complementarymeasuresto raise revenues). Inlightof the needto maintainhighexpendituresto meet developmentobjectives, lower revenues resultedin underminingthe fiscal positionofthe government andperhapsthe economic stabilityofthe country. 56 Box 5.4: Finance and Economic Development Ministries -Together or Apart? 4n often debated topic of government machinery indeveloping countries is whether planning and finance ;hould be separate orjoined. Planning ministries were once amongthe most powerful ofthe central nanagementagencies, determining governmentpolicies and budgets. Now, with more reliance placed on narkets rather than direct intervention, the number of countriesproducing multi-year developmentplans has Iwindled. Consequently, the continuedexistenceof a separateeconomic planning ministry(agency or :ommission) has come into question. Shouldplanning ministries continue their separate existence, or should heybe wound up, given adifferent role, or incorporatedinto acombined financeand economic development ninistry? 4t independence, it was widely believedthat laissez-faire policies would not achievethe economic and social irogress people wanted. Governmentshadthereforeto take the lead inraising resources and channelingthem :o sectors ofthe economy that would grow as well as to neglectedsocial sectors and high return infrastructure. Ministries of finance were seen as fiscally too conservative for this role. Thus economic planning ministries (or :ommissions) were createdacross the developing world to pursuethe new developmentparadigm, many :onsciously modeled on the Indianplanning experience. Typically, the planners would produce a five-year plan :ontaining macroeconomic projections, sector policies and programs, and a public investment program (PIP). Zountriesadopted a dual budget structure, with the current or ordinary budget preparedbythe finance ministry, mdthe capital or developmentbudget assembledbythe planning ministry. But with differentorganizational cultures, it was always difficult to integratethe two parts ofthe budget. Cautious finance ministriesdistrustedthe ambitious investment schemes of planners. The latter felt economic growth could be acceleratedonly by aggressive public programstargetedto social and physical infrastructure investment,and, perhaps, direct investmentinstate owned enterprises. Dualbudgets, furthermore, tendedto be fiscally expansionary inthe long run, and systemically under-funded recurrent operating costs. Early on, some countries soughtto resolvethe tension by combining finance and planning ina single super ministry, while retaining a planning process. Insub-Saharan Africa, Botswana did this in 1970 and has retained this structureever since. Nigeria and Ghanahavekeptthem separate-untilvery recently for the latter-though inboththe planning function is weak. Malawi, Tanzania, Uganda, andZambia fusedtheirs inthe wake of structural adjustment, and strongMTEF-led budgeting prevails intwo o fthem. Kenya alternated, partly for political reasons. SouthAfrica experimentedwith the idea of a separate economic planning body, before opting for a strongNationalTreasury and a best practice MTEF process. InSouthAsia, separateplanning commissions have continued without interruption, though their influence has declined. Malaysia andThailand continue multi-year planning, though with increasedemphasis on the budget and operational performance in service delivery. InSingaporeand Hong Kong, a medium term budget process dominates. InLatin America, traditional planning bodiesretain influence, though some adaptation is taking place. Brazilmaintains a strong planning framework, with separate planning and finance institutions. For Colombia, the organizational issue lies aheadas the finance ministry movesto create a MTEF alongside atraditional planning commission. In transition countries economic planning has been either downgraded or abolished. Some more advanced countries like Germany (except informer GDR) retain an economy ministryfocusing on economicpolicy, principally for the industrial sector. Many others have never had one, or ifthey did, the experience is long forgotten. As longas roles and responsibilities are clear, governments can work well with diverse organizational forms and there is no single correct formulation. Inpractical terms, much will dependonthe relative power ofthe planning and finance bodies, who leads them, and which one is driving reform. However, ina growing number of countries, governmentsare focusing on their core business. MTEFs, which integratebudgets and seek to cost policies and programs rather than projects alone, are replacing national development plans as the guiding framework for policymaking and annualbudgeting. As a result, more and more resourcedecisions are shifted to line ministries within ahard budgetconstraint. This is likely to leave atraditional planning ministry with a top down culture built upon on centrally evaluatedpublic investmentswith a muchreducedrole. An often consideredoption is to terminate organizational duality by forming ajoint finance and economy ministry. In this way, planning staffcontribute skills in economic forecasting, review MDA strategicplans, advise on sectoral budgetallocations, and help monitor the efficiency and effectivenessof departmentalspending. The challenge,then, is to integratethe two cultures ina single organization. 57 5.32 TheMinistry of Finance lacks resources. Capacity o fthe Ministryo f Finance to undertake sectoral analysis i s limited. Concurrently, the Ministry o f Finance, which has among its core responsibilities, the development o fthe medium term macro and fiscal frameworks, to support the mediumterm vision o fthe economy, i s also short on capacity. First, capacity for macroeconomic forecasting andmodeling needs to be improved (as discussed earlier inthis chapter). Second, the capacity to monitor overruns incapital projects also needs to be developed. Other than the obvious financial implications, cost overruns also affect the government's capacity to implement its longterm vision ina timely manner because either the overrunprojects take more time to implement ina tight budget situation, or, other projects, awaiting insequence, get delayed and this contributes to slippages o fthe NEA. Either way, more accurate resource envelopes andbetter cost management both contribute to greater efficiency o fthe public sector. 5.33 The Government has recognized lack of sufficient capacity as a key issue, and is tackling it on severalfronts. Onthe `big picture' issue, certainly, the decision to develop a Medium Term Expenditure Framework is very much a response to the need to confront the strategic challenges Mauritius is facing ina systematic way. Mauritius will have to find the bestway to obtain the critical features ofthe MTEF(fiscal discipline, strategic prioritization, and efficiency) given its own country-specific circumstances. Perhaps this can come through existingconstellation o finstitutions (Finance, Economic Development), perhaps through a merger o f Finance with Economic Development, or perhaps with a secretariat to the Cabinet or a subcommittee o fthe cabinet for economic matters. The key issue for Mauritius will be to introduce consistency inits decision making among its responses to short-term crises and its medium term development challenges -and to come up with sustainable polices. At different level, another response to the lack o f capacity is made through the civil service reform program, under which training needs assessmentsare presently undertakeninseveral ministries. However, while ministries cansend their staffto courses at the Universityo f Technology, places are limitedand most ministries do not runtheir own in-house capacity buildingprograms. Officials interviewed commentedonthe fact that they hadnotbeen provided training sincejoining the civil service. Government may wish to review the issue of increasing capacity o f the civil service, especially inlight o f as part o f the needs placed by the MTEF as part o f its civil service reform program.44 5.34 The importance of goodpolicies should not be understated, and cannot be made up with good institutions. This chapter has focused on the needs for improvement o fthe institutional framework for economic management, and has discussed the importance o f closer collaboration among the Mauritius' economic agencies. However, improvement in the institutional aspects o f economic management will not make up for outdated policies, which were briefly discussed inchapter 4. nevertheless, good public expenditure management institutions create a challenging environment for assessingpolicies under implementation. This naturally requires strong monitoring and evaluation systems, open discussion o fresults withingovernment, and outside government, and capacity o f analysts and policymakers to respond andpropose new policies. This report does not address civil servicereform issues as additional analytical work would be requiredto cover the topic. 58 5.35 Overall, the foundation for soundeconomic management inMauritiusexists today. The Ministry ofFinance andCabinet discuss Government priorities andtranslate these priorities into sectoralallocations. Some sector ministrieshave developed sub-sector master plans and/or reform programs, andmost have developedmission statements andobjectives. Policy analysis andreforms are underway ina numberof ministries.Financialmonitoring systems and compliance with existing rules are strong. A broad range o f output statistics are collected and usedto some extent intaking managementdecisions. The Government has recognizedthe needto improve the delivery of services by the public sector andhas therefore initiatedfar-reaching civil service reform. Overall, support for the introductionof a more advanced system of economic managementbasedon MTEFandRBMappearsto be strong. 5.36 But asignificant problem that remains is that the planning andmanagementprocesses are not systematically linkedto the assessment of the effectiveness of Government programs and therefore effective prioritizationcannottake place. Inaddition, the current processesdo not encourage organizations to work together effectively; inparticular, the separate preparationo fthe capital budget at the Ministry of Finance andthe Public Investment Program at the Ministryof Economic Development does not facilitate cooperation between the two mini~tries.~~ Planning processes inthese two ministriesseem to overlap and are not fully integrated. There also appears to beno formal or informalmechanismto coordinate economic managementbelow the Cabinet level-as mentioned earlier, the MCC does not have adequate capacity to fulfill this role.47 5.37 To improve the efficiency andeffectiveness of economic managementinMauritius thisreport has eight recommendations: (i) Institutionalframework for economic management: The institutionalframework for economic management should be strengthenedby introducing systematic, regular coordination andcollaboration below the Cabinet level on economic management issues. This should include the establishment of formal or informal mechanisms for coordination andcooperation within Government. With progressive introduction of the MTEF, the separate public planning investment exercise is likely to become redundant.The Government may want to consider integrating the planning functions ofthe Ministries ofFinance andEconomic Development to improve operational efficiency and consistency (Box 5.4 discusses the international experience o fhaving Finance and Planning Ministries together or apart). (ii) Macroeconomicforecasting and analysis: While the Government collects and analyses macroeconomic data, this informationis not brought together ina systematic manner on a regular basis. Inaddition, there are no formal or informal institutional 45This section summarizes recommendations made throughout chapter 5, However, throughout chapter 5, the authors have noted occasions when government has already taken action, since the time the report was originally drafted. 46As occurred prior to the merger o fthe ministries o fFinance and Economic Development. 47See footnote 39 on the creation o fthe Economic Coordination Committee, inpart as a response to this problem. 59 arrangements for regular discussion o fmacroeconomic issues.48Setting up mechanisms for macroeconomic forecasting and analysis at boththe technical andthe policy levels will be important to ensure a more detailed, systematic, andregular analysis of macroeconomic data, forecasts, trends, andpolicy issues. Furthermore, quasi-fiscal activities should explicitly be included inthe fiscal fiamework. (iii) Settingbudgetceilings: ThecurrentprocessofdeterminingGovernmentpriorities and translating these priorities into budget ceilings couldbe further strengthened in three main areas. First, sector ministries' activities should be analyzed inmore detail and a clear link to Government priorities should be established. Second, resource allocation decisions of Cabinet should clearly reflect the trade-offs betweensectors. Third, the budgetcircular shouldbemore detailed andtransparent. (iv) Sector planning: Sectoralplanning processesand policy analysis should be more systematic. Planning documents should cover the entire scope o fthe ministry's functions and should be drawn relative to government's program objectives ineach sector. Policy proposals andprojects should be fully costed, clearly linkedto Government priorities, and approved as part ofthe annual budget process (andnot outside the budget cycle) and should be discussedinthe context o f trade-offs inview o fthe country's budget constraints. To monitor the implementation of Government policies, systematic feedback loops should be established to enable the progress assessments andthe adjustment ofpolicy decisions. (VI Expenditure planning: The budget process should be systematically aligned with the development of an overall planwithina ministry that i s basedon achieving an agreed set o f objectives. This should include planningo f recurrent expenditureneeds - currently, these expendituresare planned incrementally rather than strategically. Similarly, the present dual budget system, with recurrent andcapital budgetsplanned separately, should be replacedby an integrated budget approach, (4 Results based M&E framework: Systematic feedback loops should be establishedto monitor the effectiveness of programs inachieving Government objectives. Datathat i s collected ina number of ministries should be systematically used to inform managersandpolicy makes about progress and performance. Civil service reforms, expendituremanagement, andRBMreforms should be integrated into a consistent framework that aims at holding organizations and individuals accountable for results. Therole of civil society inassessing and monitoring Government performance should be strengthened andmechanismsfor enabling citizen feedback on Government services should be systematically established. (vii) Budget oversight: The Auditor General should introduce financial or performance audits to assess the efficient use ofpublic resources. The Government should report regularly on policy implementation to encourage debate. (viii) Capacityfor economic management: The Government should develop a comprehensive training planto improvethe capacity o f existing staff inthe area of economic management, focusing inparticular on MTEF,monitoring and evaluation, andresults basedmanagement. 48Ibid. 60 ANNEX 1 GOVERNMENTEXPENDITURESAND PARASTATALS: SELECTEDDATA The present Annex presents selecteddataon expenditures and onparastatals. The source of the data i s the Government of Mauritius, MinistryofFinance. The tables present information available to the Ministry ofFinance. This data is considered confidential. Table A.l Budgetary Central Government: Composition of Total Expenditure (percent) 1986-88 1989-91 1992-94 1995-97 1997198 1998/99 1999/002000/01 Average Average Average Average General Government Services 17.1 20.4 20.4 20.4 18.8 17.9 17.5 17.0 General public services 10.4 10.9 10.5 10.3 10.3 9.2 9.3 9.0 Defense 0.8 1.3 1.5 1.3 0.9 0.8 0.9 0.8 Public order and safety 5.9 8.2 8.4 8.9 7.6 7.8 7.4 7.1 Com. and Social Services 38.4 41.7 46.9 49.5 50.7 50.7 51.3 50.2 Education 13.9 14.5 14.6 16.5 16.0 15.0 14.7 14.0 Health 8.6 8.8 8.9 8.1 8.1 8.3 8.3 8.2 Social sec. and welfare 13.2 13.3 15.7 17.6 19.4 19.9 20.5 19.9 Housing and com. amen 1.8 3.1 5.7 5.6 5.7 5.8 6.3 6.7 Recr., cult., religious act. 1.o 2.1 2.0 1.7 1.5 1.6 1.6 1.3 Economic Services 19.8 15.6 15.2 13.0 11.4 13.5 13.5 13.6 Fueland energy 0.1 0.1 0.1 0.0 0.0 0.1 0.1 0.1 Agri., fore., fishing 7.6 6.9 5.8 5.2 4.8 5.4 4.6 4.3 Mining, manuf., const 1.1 0.6 0.5 0.7 0.6 0.6 0.6 0.6 Transport, telecom. 7.5 4.2 6.1 3.8 1.7 4.1 4.1 4.4 Other econ. services 3.6 3.9 2.9 3.2 4.2 3.4 4.1 4.2 Other Expenditure 24.6 22.3 17.5 17.1 19.1 18.0 17.8 19.2 Public debt interest 19.9 17.5 13.4 13.6 16.0 14.8 14.5 16.4 Transfer to local govern 3.6 3.7 3.6 3.5 3.O 3.2 3.2 2.8 Other 1.1 1.1 0.6 0.1 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: Government of Mauritius. 61 Table A.2 Budgetary CentralGovernment: Composition of Capital Expenditure(percent) 1986-88 1989-91 1992-94 1995-97 1997/98 1998/99 1999/002000/01 Average Average Average Average GeneralGovernmentServices 12.9 20.5 19.1 18.1 12.1 9.9 9.1 8.7 Generalpublic services 9.8 9.1 8.7 7.4 8.9 4.5 5.1 4.7 Defense 0.0 0.1 0.0 0.1 0.0 0.0 0.0 0.0 Publicorder and safety 3.1 11.2 10.4 10.6 3.2 5.3 3.9 4.0 Com. and Social Services 22.4 39.0 41.3 48.4 60.4 48.5 48.7 48.7 Education 7.3 8.7 8.5 17.4 20.1 9.7 9.7 9.6 Health 4.5 5.9 5.7 3.2 4.4 3.3 2.5 2.3 Social sec. andwelfare 0.0 0.3 0.2 0.8 0.3 0.9 0.5 1.4 Housingand com. amen 7.7 15.4 20.4 23.1 32.2 30.3 32.1 33.4 Recr., cult., religiousact. 2.9 8.7 6.4 3.9 3.5 4.2 3.9 1.9 Economic Services 63.2 40.1 39.4 31.7 27.4 39.5 39.1 41.0 Fuelandenergy 0.2 0.3 0.2 0.2 0.2 0.3 0.3 0.2 Agri., fore., fishing 12.6 11.4 6.2 7.2 7.3 10.5 7.5 5.3 Mining, manuf., const 4.7 2.2 0.3 0.6 0.8 1.1 0.8 0.6 Transport,telecom. 45.7 24.2 30.7 20.7 9.2 26.5 26.0 22.9 Other econ.services 0.1 2.0 2.0 3.O 9.9 1.1 4.6 12.1 Other Expenditure 1.5 0.4 0.6 2.6 0.0 2.1 4.0 1.5 Publicdebt interest Transfer to local govern 0.6 0.3 0.4 2.6 0.0 2.1 3.1 1.5 Other 0.9 0.3 0.2 0.8 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: Governmentof Mauritius. 62 1 Date of Government Permanent Temporary Establishment Share Employment Employment Agricultural Marketing 1 I 1 1 Board 1964 PC* 184 Airports ofMauritiusLtd 1 1999 100 400 1 (AML) Cargo Handling 1 /6%,MPA-40%,1 955 Corooration Ltd 1983 SIC-54% 11 Development Works 1 Corporation 1971 PC* 700 1000 Mauritius Broadcasting 1 Corporation 1960 PC* 489 Mauritius MeatAuthority I11 1974 111 100 111 146 111 I 11 1 Mauritius Port Authoritv I 1976 I Rs48 m 397 I I Mauritius Shipping 16493 shares out 1I Corporation 1986 o f 6520 National Housing Development Corporation I 1991 1 SICOM-40% NationalTransport Corporation 1979 PC* Rose Belle Sugar Estate 1973 100 635 90 State Trading Corporation 1983 PC* Sugar Planters Mechanical Pool 1974 PC* 275 Corporation Source: Governmento fMauritius. 64 h 2 3 Table A.5 Financial Performance o fthe Largest Public Enterprises for 2001/2 and estimates for 2002/3 (million Rs.) Operating Balance Operating Subsidies 2001102 2002103 2001/02 2002103 Est 1.AgriculturalMarketing Board(AMB) 36.9 27.6 2. Airports of Mauritius Ltd(AML) 61.6 58.3 3. Cargo HandlingCorporationLtd(CHCL) 12.2 38.4 4. Central Electricity Board(CEB) 862.0 728.0 5. Central Water Authority(CWA) 38.0 35.0 134.0 21.0 6. Development Works Corporation(DWC) (169.1) (21.0) 7. Mauritius Broadcasting Authority(MBA) (2 1.4) 14.4 10.0 0.0 8. Mauritius Meat Authority (MMA) (7.8) (7.4) 5.0 5.0 9. Mauritius Port Authority (MPA) 191.0 227.0 10. Mauritius ShippingCorporation(MSC) (53.4) (48.1) 38.8 54.4 11. Mauritius Sugar Terminal Corporation(MSTC) 0.0 0.0 12.National HousingDevelopment Co.(NHDCL) (5.6) (3.2) 13.National Transport Corporation(NTC) (3 1.O) (50.0) 36.0 36.0 14. Rose Belle Sugar Estate(RBSE) (32.8) 8.7 15. Sugar Planters Mech. Pool Corp.(SPMPC) (60.5) (63.4) 16. State Trading Corporation(STC) 690.0 595.0 400.0 420.0 TOTAL 1,510.1 1,539.3 Percentof GDP 1.1 Memo Item: GDP(nomina1, million Rs) 137,375 /a Excludinggovernmentgranthubsidies Source: Governmento fMauritius. 66 3 0 0 P4 h 5 a 8 h 0 d 9 h u i n E u0 I m 0 a G 0 4i 0 12 m 3 cd I ANNEX 2 A HIGHERDEBTSCENARIOFORTHE CENTRAL GOVERNMENT AND THE COSTSOFADJUSTMENT The present annex looks at the possibility of adopting ahigher deficit pathfor the central government. The annex suggest that the costs of higher spending (higher interest rates, policy risk, signaling risk) outweighthe benefits of higher spendinginthe outer years ofthe horizon. The annex additionally looks at the option ofrunning amuchhigher deficit inthe short run, but suggest that the adjustment costs inthe event of an adverse shock would be too hightojustify such action. A higher debtalternative for the longterm Mauritius could alternatively make a strong case for ahigher debt and deficit target for the central government, as part of a set o fpolices that would stabilize the parastatal sector and reduce its debt inthe mediumterm. Although this approachwould give more spendingroom to government, with average annual growth o f expenditures of about 5.3 percent for the ten year horizon, the higher spendingbenefits (relative to the NEA framework) would materialize from 2005/6 onwards as that is when debt is scheduledto drop severalpercentagepoints below 60 percent of GDP under the NEA framework. However, ifMauritiusdidfollow the higher debt alternative, under the assumption that parastatal debt was being reduced, the country would have to consider three implications. 1.The impact of continued high domestic central government financing needs. An increase inthe central governmentdebtwouldplace increasingstrainondomestic credit andcontinue the crowding out ofthe private sector shouldthe higher sustaineddeficits be financed by recourse to domestic credit. Historical experience suggests that for every 1percent increase ingovernmentcredit(relative to GDP) domestic realinterest rates increaseby abouta 1/2 percent point. Alternatively, recourseto monetizing the deficit couldleadto inflation which i s also potentially destabilizing for the economy. Alternatively Mauritius could borrow from abroadto finance the higher deficits. Criticalthere would be the potential impact on the exchange rate, from the higher demand for the Rupeeinthe highdeficit years. 70 2. Policv risk. It i s entirely possible that government Chart A-2.1 could set course to attain a macro framework represented intable 3.5, but could runinto problems 10 ?of GDP % of GDP down the road. For instance, the deficit may be fall T 801 out o f control infor a few years due to political developments, andthe economy could be hit with a major cyclone or a global downturn. Chart A-2.1 presents such a scenario. It illustrates the difficulty o f returning to a l o w debt scenario fkom a highdebt one. Box 3.1 illustrates that it becomes more difficult to take corrective policy actionwhen ina highdebt situation. -Fiscal balancelGDP CG DebtlGDP 3. Signaling risk. Since 2000/1, Mauritius has been Source: World Bank staffprojections runningfiscal deficits o fclose to 6 percent of GDP every year. Continued fiscal deficits close to 6 percent o f GDP inthe short term without very strong signals on fiscal restructuring may be misinterpretedby the markets that the public finances of Mauritius are not controllable, despite the strongjustification for public investment. This couldbe damagingto the economy andcouldprompt interest rates to raise further reflecting creditors' demands for an additional risk premium. Table A 2.1 Higher DebtAlternative (average GDP gr. 4.5 percent) 01/02 02/03 03/04 04/05 05/06 08/09 10111 12/13 As YOof GDP ,unlessotherwise stated Revenue 18.5 20.5 20.7 21.1 21.7 22.1 22.1 22.1 Expenditure 24.4 26.4 26.2 26.1 26.7 27.0 26.8 26.8 O/w interest 3.3 4.3 4.0 4.0 4.2 4.3 4.4 4.4 Overall Deficit -5.9 -5.9 -5.5 -5.0 -5.0 -4.9 -4.7 -4.7 Primarybalance -2.6 -1.6 -1.5 -1.0 -0.8 -0.5 -0.3 -0.3 Debtstock 55.3 56.4 57.4 57.9 58.2 59.2 59.6 59.7 Memo: Interest/ Tax revenue("h) 21.1 21.1 19.3 19.3 19.4 19.7 20.1 20.1 Externaldebt service/ Exports(%) 1.0 1.0 1.1 1.0 1.o 0.8 0.8 0.8 Realnon-interestexpend. (2002/03 Rs.Bill) 30.1 33.4 34.9 36.2 38.6 44.3 47.8 52.1 Source:GovernmentofMauritiusand WorldBank estimatesandprojections. Underthe aforementioned circumstances, the government could embark on a strategy of (a) immediately selling off existing equity insome parastatals and/or (b) gradually preparing parastatals for strategic partnerships or privatization and usingsome o f the revenues to pay off existing debts. Either option would be healthy for the economy, for the companies, and for the public finances. 71 Box 1: Thedijj?cultyoffiscal adjustment The chart below illustrates that higher debtpolicypositions(lower bolder lines) needto make larger fiscal adjustments (because ofsteeper slopes) to maintaintheir debt levelinthe event of a GDP shock. The "iso- debt" (ID) lines representa levelofdebt relativeto GDP. Movementalongthe IDlinekeepsthe debt ratio to GDP the same, implyingthat highergrowthrates enableworse primarybalances withoutchangingthe debt level. Lower lines (bolder) representhigher levels of debt; butthese lines alsohave steeper slopes. This meansthat althoughwhenthe country is at a highlevelofdebt only reallybaddeficits will makethe debt situationworse (because it is starting from a very highdebt base). The steeper slopes suggestthat should GDP growth be lowerthanexpectedina givenyear (thena country with a larger debt needsto make a much larger adjustment(smaller primary balance)thana country with a smaller debt. Hence, a growing economy supportshigher primaryimbalanceswith a highdebt base, butneedsto make very painfuladjustmentsshould growth slow down. For instance, inchart 1, the ID linedhavethe followingslopes: ID(45)=-0.5; ID(55)=-0.6; ID(65)=-0.8; ID(75)=-0.8. An shortfall ingrowth from 5.5 percentof GDP to 4.5 percento fGDP would require a 0.5 percentofGDP improvementonthe primary balanceto avert deteriorationofthe debt at a debt levelo f45 percento fGDP comparedto a 0.8 percentofGDP improvement inthe primarybalancefroma starting debt levelof75 percentof GDP. Althoughthe difference seems modest, suchreductionin expendituresare difficult to implement. Chart 1. Is0 -Debt Lines Trade-offsbetweenGrowth Ratesand Prim Balance (higher growth rates afford higherprimary balances) 0.0 0c -1.0 a -2.0 2" 2 -3.0 B on 0 " -4.0 2 -5.0 1.5 2.5 3.5 4.5 5.5 6.5 GDPgrowth rates -Debt 75% GDP-Debt 65% GDP -Debt 55% GDP Debt45% GDP The case for yet higher short-term deficits A case canbe made that Mauritius' GDP growth is ChartA-2.3: GDP andFiscalAggregates now below its longterm average (or full t3, , 25 employment average) and hence there isjustification for a higher deficit, i.e. counter cyclical fiscal policy to bringthe economy to "full employment". The rationale is typically that tax revenues reflect economic performance and below average GDP '1 t1 o growth implies merelya revenue drop which i s . 2 1 , , , , , , , , , , , , -5 cyclical. Indeed, chart 3.3 illustrates, as has been 90 92 94 96 98 00 2f discussed inchapter 2, that GDP growth i s below its -GDPmp gr. - . - .- MinusDef ~ ~Rw/GDP (Rt axis) long-term average. The chart also illustrates that duringthe 1990sthe higherdeficits havenot Source: Govemment of Mauritius and World Bank staff calculations. necessarily come at times o f lower growth rates, as in 72 the 1994-96 periodandthe 1999-00period, so the cyclically hypothesis requires some investigation. Infact, for 2001 and 2002, revenues fell primarily due to the reduction intrade tariffs due to regional trade commitments. That beingthe case, coupled with the structural imbalances existing inthe Mauritianbudget, andunclear signaling effects of deficits more than6 percent o f GDP at this time, make this a difficult propositionto accept. 73 ANNEX 3 INCORPORATING FISCAL RISKSINTO THE FISCAL FRAMEWORK The present annex attempts to quantifythe risks onthe central government ofthe public sector debt andto examine the impact o fthose risks onthe programmable or feasible mediumterm expenditureenvelope for Mauritius. The annex looks at only the stock of existing debt at end June 2002, and does not consider the acquisition o fnewdebt by the parastatal sector. The analysis inchapters 1and2 indicated that the debt servicing o f parastatals can pose substantial fiscal risks to the central government as it has been consistently subsidizing some aspect o ftheir operation. The likelihood that these risks materialize for government inthe future depends onthe international environment, inparticular the world price o f oil. It also depends on the capacity o f the governmentto adjust domestic prices without adversely affecting the competitiveness o f the economy. Not planningfor these fiscal risks, and incurringthem, i s detrimental to (i) macroeconomic stability and (i) predictability o f budget envelopes. the Methodoloay. There are two possible approaches to evaluating the potential impact o f parastatal debt onthe central government. The first approach is to consider individual loans o fparastatals and assess which loanpayments are not likely to bemet by eachparastatal, andconsider whether the central government will step into meet the payment obligation. This however would require intimate knowledge o f each loan and this information i s not readily available and i s beyondthe scope o fthe present report. The second approach, the one selected for this analysis, i s to group parastatal debt into that which has: (i) hiah likelihood of being serviced by the central government, hence that which belongs to highrisk parastatals (CEB, STC, DBM, MHC, NHDC). (ii) moderate likelihood o f being serviced by the central government, andthis i s all government guaranteed debt (including the highlikelihood debt), and (iii) low likelihood of beingserviced by government, and this is total parastatal debt, including non-guaranteed and guaranteed debt. This classification allows usto break the central government repayment risk for parastatal debt into three cumulative categories: high, medium, and low. The reason for constructing cumulative categories i s that an implicit simplifyingassumption is made that all parastatals face risks from economic slowdown but that some parastatals -the relatively "higher" risk ones- are less poorly insulated inthe event o f an economic downturn. So the category (ii) "moderate likelihood o f debt being serviced by the central government" includes the parastatals o f category (9 Table A 3.1 presents the debt stocks and the debt service payments o fthe three categories of parastatal debt till June 2002. It should be noted that Category (i) service i s purely debt indicative. Since Droiections o fparastatal non-guaranteed debt do not exist ingovernment, we have assumedthat the stock o f debt relative to GDP i s reduced inlinear fashion; to obtain an 74 indicative quantification of debt service paymentsthat would be necessaryto bring the stock of 22 percent of GDP down to zero in 10years.49 The exercise is highly sensitive to the structure of the debt. For Categories (ii) (iii), draw onprojections ofexisting parastataldebtwhich and we the team compiled as part of this report preparati~n.~' Two simulations were runwith this datato test, first the impact of exchange rate movements, and second, the impact on interest rate movements. Although movements ininterest rates hadsmall impact on contracted debt because only a small share of it is invariable rate loans, exchangerate movements had a large impact of debt service. Table A3.1 Parastatal Debt Service (as percent of GDP, averageGDP gr. 4.5 percent) 01/02 02/03 03104 04/05 05/06 08/09 10111 12/13 Category I:Total Parastatal Debt -Lowest Likelihood of central government repayment Debt Stock 22.0 20.0 15.1.0 11.0 6.8 2.3 1.4 0.0 Debt service .. 4.3 6.7 6.5 6.3 1.5 0.8 1.1 Category 2: Guaranteed Parastatal Debt -Moderate Likelihood of central government repayment Debt Stock 13.0 8.3 8.3 6.3 4.8 1.9 0.9 0.4 Interest 0.5 0.5 0.5 0.4 0.2 0.1 0.1 Principal [1.9] 1.3 1.3 1.1 0.5 0.3 0.2 Debt service ..,... 2.4 1.8 1.8 1.5 0.7 0.5 0.2 Category 3: Debt of Selected Parastatals -Highest Likelihood of central government repayment Debt Stock 5.4 4.1 3.3 2.4 1.9 0.8 0.2 0.1 Interest .. 0.3 0.2 0.2 0.1 0.1 0.0 0.0 Principal 1.4 0.7 0.6 0.4 0.2 0.2 0.0 Debt service .... 1.7 0.9 0.8 0.5 0.3 0.2 0.1 Source: Government o f Mauritius and World Bank staff estimates and projections. A reasonableprogrammableexpenditureenvelope Table A 3.2 presents a sustainablefiscal framework, same as table 3.4, but explicitly showing the debt service of parastatalsas an expenditure item. The International Monetary Fund's Government Financial Statistics specifies any expenditureto parastatalsto be classified as subsidies. This would apply to funds usedfor paymentsofinterest or principalbypara~tatals.~~Table 3.6 indicates that over the three-year period2003/4 -2005/6 the averagedebt service ofthe high-risk parastatals will come to about 0.7 percent o f GDP. The year 2002/3 already reflects an amount of Rs.1.4 billion inSTC debt 49With central government debt reaching about 52 percent o f GDP ina 10year horizon, this implies that under the present scenario, parastatal debt would remain at about 8 percent o f GDP. The present exercise does suggest a "regime change", i.e. parastatals service all new debt acquired in200213 and beyond without government assistance. Inthe event that the central government takes over the title o fthe debt, Le. becomes the debt holder, thenthe debt service payments are presented as any other debt service o fthe central government, i.e. interest above the line and principal below the line. 75 whichis being covered by the central government. According to the methodology establishedin section 3.3, the effective average primary balance for the central government that will assure convergenceto the a lower debt level o f 60 percent of GDP, -0.5 percent of GDP, will be increasedby an average of 0.7 percent for the period2003/4 -2005/6, to 0.2 percent of GDP, meaning the government will have about 0.7 percent of GDP less to spend on non-interest non- parastataldebt expenditures. Those would come to about 20.9 percentper annum, or Rs. 34.6 billion. Inthe extremecasethat the central governmenthadto subsidizethe debt service ofall parastatals, the primary balanceplus the subsidywould come to 4.5 percent per annum for the period2003/4 -2005/6. For instance, relative to the no subsidy case, the central government would have 4.9 percent of GDP less to spendper annum for the period2003/4 -2005/6. Table A 3.2 A Sustainable Programmable ExpenditureEnvelope A v ~2003- . Avg. 2006- I 01/02 02/03 03/04 04/05 05/06 2005 08/09 10111 12/13 2012 (As % ofGDP, unless otherwise stated) Revenue 18.5 20.5 20.7 21.1 21.7 212 22.1 22.1 22.1 22.1 Expenditures 24.4 26.4 26.2 25.6 25.2 25.7 25.2 25.1 25.1 25.1 Interest 3.3 4.3 4.0 4.0 4.1 4.1 3.9 3.8 3.7 3.9 Subsidy for debt service of high risk parastatals .. 1.7 0.9 0.8 0.5 0.7 0.3 0.2 0.1 0.3 ProgrammableEnvelope (resid.) 21.1 20.4 21.4 20.8 20.6 20.9 20.9 21.1 21.3 21.0 Overall deficit -5.9 -5.9 -5.5 -4.5 -3.5 -4.5 -3.0 -3.0' -3.0 -3.0 Primary balance -2.6 -1.6 -1.6 -0.5 0.6 -0.5 0.9 0.8 0.7 0.9 Primary balanceplus subsidy for debt serv. ofhigh risk parastatals .. 0.1 -0.7 0.3 1.1 0.3 1.2 1.0 0.8 1.1 Memo Other Expenditures (Real. 2002/3 Rs.billion) 30.1 33.4 34.9 35.4 36.1 35.5 41.5 45.5 49.9 43.6 Primary balanceplus Subsidy for debt service of all parastatals (%GDP) .. 2.2 4.1 5.1 4.1 4.5 2.1 1.4 0.7 2.0 Source: GovernmentofMauritius and World Bank staff estimatesand projections. 76 ANNEX 4 AN INSTITUTIONAL FRAMEWORKFORMANAGING FISCAL RISK Introduction Starting in2000/01, the government introduced a more comprehensive definition o f its budget deficit: The real budget deficit was definedto include revenues and expenditures o f Privatization Fundand cost o fnew tax concession. Consolidated RealBudgetDeficit estimate evenrightly included deficits STC andCEB. These initiatives make Mauritius one o fthe countries approaching best practice indeficit analysis. However, the deficits o f STC and CEB, which had been coveredby their own borrowing, are yet to affect the reported government debt inthe future. Inthis aspect, the deficits generated by off-budget entities, such as public corporations, are differentfrom the real budget deficit, which affects the level o f government debt immediately. Transparencyoffiscal risk The current framework o fpublic finance institutions gives a good degree o ftransparency with respect to government explicit contingent liabilities. The list o f government guarantees i s publicly available. Documents o f the Ministryo f Finance, such as the annual Report o f the Accountant General, provide the full list o fbeneficiaries and lenders under government guarantees andthe amounts outstanding denominated inthe domestic currency as well as the currency o f origin. The degree o f government risk exposure emerging from guarantees, and the factors that affect whether or not a guarantee will be called are, however, not discussed. As for other risks, their transparency is uneven. The annual report o fthe Bank o fMauritius provides detailed information on the portfolio o f government debt. The report, however, does not discuss the associated government risk exposure and its possible impact on future debt service cost. Quasi-fiscal operations conductedthrough public corporations and their potential future fiscal cost are discussed occasionally. For example, the government has reflected onthe potential fiscal risk emerging from the domestic price caps on electricity, water and oil products inits 2000 report The Present State o fthe Economy. Inaddition, annual reports o f public corporations provide some insights into their fiscal function and finances.52 52 Pastreportso f STC, for instance,explicitly show the unmetrequests for reimbursementof losses incurredonthe domestic sales of oil imports. 77 Tax expendituresand divestment of public assets are not consistently reported. There is a good example, though, on whichthe government canbuild inthis regard: The Present State ofthe Economy report presents the estimatedbudgetary cost of newtax concessionsgearedtowardthe electricity, water and transport sectorsto alleviate the accumulation of liabilities of related public corporations. Ceilings and incentivesin governmentrisk management Presently, the legal andregulatory framework for public finance inMauritius does not set any explicit limits or safeguardson the amount of government guaranteesand other contingent liabilities andrisk. Guaranteesand other contingent liabilities are not subject to any ceilings (either on the total levels outstanding nor on the amounts issued ina fiscal year). The issuanceof government guarantees i s well centralized at the Ministry ofFinance, withinthe powers of the Minister. A copy of every guarantee agreement is submittedto the Assembly, but only after the agreementwas concluded. Itappearsthat even guaranteesthat, for any reason, are not submittedto the Assembly and are not reported inany government documents are legally valid. Guarantees and other promises of contingent government support are subjectedtoo mucha lower scrutinythan government cashoutlays. This is true throughout the budget process, includingthe budget preparation, evaluation and audit conducted by the State Auditor. For example, guarantees are issuedon an ad hoc basiswithout any particular relationship to budgetary priorities and constraints. The government does not charge any fee for a guarantee and does not make any budgetary provisions for future possible claims. Every government guarantee covers all risks that may be at the source o f a default on the guaranteeddebt. As for managing government exposureto risk, the government is yet to specify its objectives and strategy. The absence of clear risk management objectives and strategy i s most urgently felt in the area o f government debt management.53 But it is also making difficult to ensure proper incentives andprudence inthe issuance of government guarantees and inthe overall management of government contingent liabilities. Similarly, without a clear government risk management strategy, it is difficult to establish a systemfor prudentand efficient managementof risks inthe public sector. Monitoringfiscal risk Monitoring of contingent liabilities i s centralized at the Ministryof Finance. It is basedon collection of quarterlyreports from beneficiaries o f government guarantees and other quasi-fiscal operations. Monitoringfocuses on disbursements, repayments and amounts outstanding. Itdoes not aim at examining actual performance underprograms supportedby guarantees. Recent initiatives at the Bank of Mauritius and Federal Services Commission promise to improve the ~ 53Reports prepared for the government by Paul Sullivan (2002) and Ian Storkey (2001) elaborate on this issue. 78 monitoring of risks inthe banking and non-banking financial sector. The mechanism for monitoring fiscal risks of non-financial public corporations is less clear. Thecapacityto managefiscal risk Presently there is no obvious capacity to analyze and manage fiscal risk inthe government. Risksinvolvedinthe portfolio of government debt are expectedto bemanagedbythe newly establishedDebtManagement Office (DMO) o fthe Ministryof Finance. Relatedto the managementof risks under the government debt portfolio, the DMO i s also expected to play an important role inreshapingthe legal and regulatory framework and inbuilding adequate infrastructure for the developmentof domestic government bond market. Inthis respect, as well as inestablishing reliable debt managementfunctions, the task to strengthenthe DMO's is urgent. To complement the advice already receivedby the government from independentdebt management experts, this report would only emphasizethe needto bring government contingent liabilities withinthe scope of DMO's work. Along with the risks involved inthe government debt portfolio, the DMO should at least monitor and analyze risks arising from government . guarantees and from the public sector andthe economy more broadly. Eventually, rather than building risk management capacity ineach individual public company, suchas CEB and STC, the DMO could centralize the managementof the main common risk exposures, suchas those to the exchangerate and interest rate movements, inthe public sector. Broadening the scope of risk management across the public sector would allow DMO to immunize the government risk exposure by, for instance, structuring government liabilities as to develop natural hedges. Alternatively, the function of riskmanagement across the public sector couldbe contractedout to the private sector. If well specified, such a contractual arrangement can generatethe efficiencies of risk pooling while not relying on the DMO's scarce resources. Reducingthescopefor fiscal opportunismin thepublic sector On a broader basis, reducing the government exposureto fiscal risk will also require improvements ingovernance, specifically, inclosing the loops through whichpoliticians are able to affect investment, credit andor spending decisions of bothfinancial and non-financial, public andquasi-private corporations. This task is enormous and, obviously, its objectives would be muchbroader thanjust ensuring overall fiscal discipline of the government. Inthis regard, the Financial Services Commission currently draft laws for non-banking financial institutions i s considering anumber of useful measures. 79 Possible agenda for the future Motivation There are severalreasonswhy itmaybe wise for the present Government to improve the institutional framework for dealing with government fiscal risk. First, already inits 2000 report on the PresentState ofthe Economy, the current'Government recognizedthe dangers of fiscal opportunism and the negative fiscal effects of quasi-fiscal operations. Second, the current Government appears committed to promoting reforms, transparency andaccountability inthe public sector. And, third, pressure to reduce the budgetary deficits facing the future as well as the current Government may generate incentivesto move some budgetary programs offbudget, that is to transform budgetary expenditureinto government guarantees or into quasi-fiscal operationsto be (temporarily) fundedby non-government entities. International experience gives many examples of governments expandingtheir off-budget operations of fiscal nature and thus amassinghuge contingent liabilitiesas they focus onthe reductionoftheir budgetdeficits (Box 1).Fromthe macroeconomic viewpoint, however, tight control over all potential government liabilities is more important than budget balance accompaniedby increasing contingent liabilities. Box I Fiscal Risks as a By-product of Reducing BudgetDeficit When looserules for fiscal managementare accompaniedby pressure for fiscal adjustment, vote-seeking politicians and budget-seekingbureaucratshavecause for opportunism. Paradoxically, the incentive to mask the true cost of risks often rises when governmentcomes under pressureto tighten itsbudget constraints. When pressure for adjustment is slack, the governmentmay have little incentive to hide the costs of its financial commitments. Butwhen stringent fiscal targets are imposed, wily spenders have incentive to substitute illusory adjustments for actual ones. And when proper accounting rules and strong enforcement mechanismsdo not accompanythe targets, the spenders have ample opportunity to evade the controls. Various studies have shown that weak enforcementproducesbudgetary opportunism. In 1985, the United States enactedthe Gramm-Rudman-Hollings law which promised to progressively reducethe size ofthe deficit and to produce a balancedbudget by 1991. But insteadof genuine austerity, the law spawned budgetarylegerdemainthat increasedthe government's exposure to fiscal risk.The volume o f loan guarantees escalated, the governmentwas slow inrespondingto a costly crisis inthe banking sector, and it adoptedpolicies (suchas asset sales) that weakened its long-term fiscal posture. Ineffect, faced with the Gramm-Rudmanconstraint on fiscal deficits, the U.S. Congresshas reduceddirect lending by $50 billion and increasedloan guarantees by $178 billion, replacing budgetary outlays by explicit contingent liabilities. Source:Brixi and Schick (2002): Government at Risk: Contingent Liabilities and Fiscal Risk Possible measuresfor consideration There are severalmeasuresfor the government to consider indealing with its existing risk exposures and intrying to constrainthe scope for further accumulation of risks. Dividedinto two groups according to the level of urgency, these measures can be summarized as follows: 80 In the short term: (i) Revise the legal andregulatory framework and strengthen the infrastructure needed for the development of domestic government bondmarket. (ii) Incorporate informationonthe risks emerging inthe portfolio o f government direct and contingent liabilities inthe budgetary framework (iii) Discuss the expected future fiscal cost o f government off-budget operations of fiscal nature as part ofthe budget process (iv) Evaluate alternative forms of financing the cost o f off-budget operations as part of the budget process (v) Establishceilings onthe amounts of guaranteesissued and outstanding In the medium term: Formulate Government DebtManagement Strategy, emphasizing the needto prudently managethe risks anddebt service cost (with a broader debt management strategy inplace, it will be easier for the DebtManagement Office to implementthe objective ofreducing the share of short-term government debt ina deliberate manner, for instance, with consideration for providing a benchmark yield curve to support the development of domestic bond market) Buildthe capacity ofthe Government DebtManagement Office to monitor performance under government guarantees (as to ensure guaranteeddebt repayment and reduce the risk of default on guaranteeddebt) andto analyze risks arising from the guaranteedandnon-guaranteeddebt of the public sector (on how to build government riskmanagement system see Annex) Assess the possibility of divesting some of the holdings of public assets (for example, inthe StateInvestment Corporation) withthe objective ofreducing the stock of government's short-term debt Improvereporting of performance under quasi-fiscal operations and demand accountability similar as for the state budget. Inthis respect, appropriate laws should strengthenthe role ofthe MinistryofFinance and State Audit Office to build accountability for risk management andperformance oversight across quasi-fiscal operations Design and implement a comprehensive medium-termexpenditureframework that would bring fiscal risks fully into the context of government fiscal management(see below) BringingJiscal risks into the medium-term expenditureframework This section summarizes recommendationsfor government consideration infinalizing the Fiscal Responsibility Act and indesigning the Medium-TermExpenditure Framework (MTEF). Inits standard form, the MTEFtackles the problem o f fiscal opportunism across budgetary programs, by requiringpolicy makersto analyze anddisclose the assumptions and expectedmedium-term 81 (3-5 years) implications of their budgetproposal, andto be accountable for any departures from their medium-term targeted levels inyear-by-year decision-making. The standard MTEFaims at promoting government performance at three levels, fiscal discipline, equity and efficiency, under government budgets. With respect to fiscal risk, MTEF should aim at strengthening government fiscal performance across non-budgetary as well as budgetary activities o f fiscal nature and with respect to both contingent and direct sources of fiscal risk. An expanded MTEF would promote risk-awareness culture ingovernment and would integrate fiscal risk deliberations with government policy- making and with fiscal and debt management processes. The key features o f the expanded MTEFcould be summarized as follows: Reporting: Statement of contingent liabilities, tax expenditures and otherfiscal risks To ensure proper attentionto their possible implications, fiscal risks needto be made known. Since no cash i s spent from the budget when governments assumes a contingent obligation or provides a tax exemption, cash-based accounting systems fail to detect such fiscal risks.54 Therefore, the critical step for governments i s to regularly issue a separate statement o f contingent liabilities, tax expenditures and other fiscal risks. Good practice is offered by Australia, Canada, Netherlands, New Zealand, and the UnitedStates. Their statements list the various sources o f fiscal risk, discuss their nature and sensitivities, implications on future fiscal position and equity, and where applicable provide their face andlor estimated value. Budget documents o f these countries stress test the fiscal baseline with respect to major macroeconomic, policy and demographic risks. As Mauritius' own experience has confirmed, risk outside the government occasionally falls on government. Therefore, transparency o f risks needs to be required inthe markets as well as in government. Inaddition, government needs capacity to monitor and analyze its risk exposure vis-&vis these implicit sources of fiscal risk. As a part o f monitoring its risk exposures, government needs to monitor behavioral determinants o f fiscal risks. These include performance (moral hazard) under its guarantee contracts, legality o f claims against tax duties, and performance of sub national governments other entities that may affect future government fiscal position. Medium-termfiscal strategy to reflect risks Government medium-term fiscal planning i s hardly viable ifit does not reflect implications o f fiscal risk exposures. The best way to show medium-term implications o f fiscal risks for the overall fiscal position i s to develop baseline medium-termcash projections and stress-test these projections with respect to specific risks (taking into account possible correlations). Similarly, 54 Accrual-based accounting system is neither necessary nor sufficient as a remedy. International accounting standards, for instance, require onlyprobable contingent liabilities (contingencies with relatively high probability o f realization) to be included inthe balance sheet, leaving the others ina separate statement o f contingent liabilities. 82 equity and efficiency considerations are relevant only ifthey address the whole range of fiscal activities andtheir expectedandmaximum likely costs. Stress testing, that i s estimating sensitivity of expectedfiscal cost to normal and abnormal changes inthe underlyingassumptions, is critical for developing a soundmedium-termfiscal strategy. For instance, related to a specific program (e.g., price ceilings), stress testing will show how normal (5%-10%) and extreme (40%-50%) changes inthe underlying factors (e.g., commodity price) over the next 3,6, and 12 months affect its fiscal cost. More importantly, stress testing detects key medium-termfiscal vulnerabilities o f government. The portfolio approachto stress-testingaddresses the whole range of fiscal risks identifiedinthe fiscal risk matrix and shows their overall possible fiscal impact. Results o f stress testing will be critical for government risk managementto re-estimate periodically the sensitivities of its risk exposurewith respectto changes inthe underlying assumptionsandto decide on a hedging strategy accordingly. To ensure that policy makerscorrectly reflect fiscal risks intheir fiscal planning, they must become accountable for adequacy oftheir analysis and assumptions. Therefore, the role o f state audit extendsfar beyond its conventional role and covers all aspects o f government risk analysis andrisk management. With respectto implicitrisks, whether or not government decides to disclose its exposure, medium-termfiscal strategy will be at riskifnot accompaniedby confidential contingency plans on how to proceedwhen implicit and indirect risks materialize. Budgeting: Makeprovisions and accountfor fiscal risk Budgeting andprovisioningfor fiscal risk has three main objectives: First,to draw more attention to risk analysis, andthus buildrisk awarenessofpolicy makers andthe public. Inthis context, the process of risk analysis i s more important than the actual figure, that is the value assignedto government risk exposure. Second, to makepolicy-makers cashneutral, that i s neutral betweenalternative forms ofproviding government support from the viewpoint of deficit measurementand sectoral budgetary envelopes. Third, particularly for governments with a limitedaccess to borrowing(limitedmarket access), budgeting inthe sense of early provisioning for risk creates abuffer for the event that risks materialize. Contemporary approachesto budgeting for fiscal risk are basedon the following principles: 0 Recognizethe risks and full potential fiscal cost of guarantees and other promises of contingent government support prior to their approval, as part o fthe budget process 0 Applyjoint ceilingfor the cost ofbudgetary and off-budget support for eachsector ina fiscal year. Off-budget support is a form of subsidy, which should be considered along other forms of government support for each sector. 0 Calculate the cost o f off-budget support as the present value o fthe future expectedfiscal cost. 83 0 Reflect the cost o f off-budget programs infull withinthe fiscal year o ftheir launch. 0 Transfer an amount equal to the off-budget support from the budgetary envelope o f the related sector to a central contingency reserve fund. These principles pullthe expected fiscal cost o f off-budget support into the calculation o f government fiscal deficit. The amount transferred to the reserve fund i s accounted for like an expense. Onthe other hand, when contingent claims are paid from the reserve fund, fiscal deficit remains unaffected. Implementationo fthe Federal Credit Reform Act inthe U S has confirmed that this approach is plausible even whenaccounting and budgetingsystems are not fully on an accrual basis. 84