Report No. 42155-MK FYR Macedonia Public Expenditure Review February 4, 2008 Poverty Reduction and Economic Management Unit Europe and Central Asia Region Document of the World Bank ACKNOWLEDGEMENTS This Public Expenditure Review was sponsored by Orsalia Kalantzopoulos (Country Director), prepared under the direction of Bernard Funck (Sector Manager), and has benefited from comments from Cheryl Gray, Ardo Hansson, Sanjay Kathuria, Satu Kahkonen, Bojana Naceva, Camille Lampart Nuamah, Anita Schwarz, Richard Martin Humphreys, Andreas Schliessler, Ziad Nakat, Pablo Lopez Murphey, and Mark Griffiths. Erika Jorgensenand Willem van Eeghenwere the peer reviewers. The principal authors of the report are Bruce Courtney and Ivailo Izvorski, drawing from background papers and contributions from Evgenij Najdov (Macroeconomic and Fiscal Settings), Sue Ellen Berryman, Toby Linden, Borko Handziski, Evgenij Najdov and Marija Risteska (Education), Sarbani Chakraborty and Rajna Cemerska (Health), Snjezana Plevko, Patrick Weiss and Agnieszka Chlon-Dominczak (Pensions), and Gaetane Tracz (Transport). Nancy Davies- Cole and Ekaterina Stefanovaprocessedthe report. Detailed comments from the Government on the draft document are gratefully acknowledged. CONTENTS EXECUTIVE SUMMARY................................................................................................. v A PUBLIC EXPENDITURE . B EDUCATION............................................................................................................ VI VI C D. PENSIONS.............................................................................................................. .. DEVELOPMENTSCHALLENGES.................................... AND HEALTH ................................................................................................................. VII E. TRANSPORT............................................................................................................ VIII IX F. CONCLUSIONS ......................................................................................................... x 1 THE STRATEGIC SETTING........................................................................................ . 1 ......................................................................................................... 1 B MACROECONOMIC ...................................................................................... A. INTRODUCTION . Output, Inflation, Living Standards .................................................................... 2 SETTING 2 Monetary Developments..................................................................................... Balance of Paymentsand External Debt............................................................. 4 4 C. RECENT Structural Reforms.............................................................................................. 4 FISCAL DEVELOPMENTS .............................................................................. 5 D. GOVERNMENT E. THELEVEL GOVERNMENT REVENUES......................................................................................... 7 OF SPENDING Economic Classification ................................................................................... .................................................................. 9 Functional Classification .................................................................................. 11 12 F FISCALSUSTAINABILITY . ......................................................................................... 13 G EFFICIENCY FISCALPOLICY ............................................................................... H CONCLUSIONS ........................................................................................................ .. OF 14 16 2 EDUCATION ............................................................................................................... . 19 A. INTRODUCTION ....................................................................................................... 19 B. INSTITUTIONALFRAMEWORK EDUCATIONEXPENDITURES AND ............................. 20 C. EDUCATIONOUTCOMESAND EXPENDITUREEFFICIENCY ....................................... 24 24 D. Efficiency of Resource Use .............................................................................. Education Outcomes......................................................................................... 30 RECOMMENDATIONS .............................................................................................. 31 31 E. CONCLUSIONS ........................................................................................................ 35 ImprovingEducation Performance................................................................... ImprovingEquity and Efficiency of Spending................................................. 33 3 HEALTH ...................................................................................................................... . 37 A INTRODUCTION 37 B INSTITUTIONAL 37 C 38 D. HEALTHCARE ... ....................................................................................................... .............................................................................. HEALTH OUTCOMES............................................................................................... ENVIRONMENT REVENUES HEALTHCAREAND 39 E. HEALTHCARE FINANCING....................................... EXPENDITURES ................................................................................. 42 F. THEDELIVERY Preventive andPrimary Healthcare .................................................................. AND UTILIZATIONHEALTHCARE OF ............................................... 44 45 46 Benefit Package ................................................................................................ In-Patient Care.................................................................................................. 47 G. RECOMMENDATIONS PharmaceuticalExpenditures............................................................................ 47 .............................................................................................. 48 Health Financing............................................................................................... 48 H CONCLUSIONS ........................................................................................................ . Health Expenditures.......................................................................................... 48 50 4. PENSIONS ................................................................................................................... 51 A INTRODUCTION . ....................................................................................................... 51 B. STRUCTUREOFTHEPENSIONSYSTEMAND PENSIONEXPENDITURES Coverage........................................................................................................... .................... 52 52 The PAYG Pillar............................................................................................... 53 C. The FundedPillar.............................................................................................. 56 ASSESSING ADEQUACY FUTURE .................................. 59 The Likely PensionOutcomeUnder the New Multi-Pillar Scheme ................60 THE OF PENSION BENEFITS MinimumPensionGuarantee ........................................................................... 65 Guarantees inthe FundedPillar ........................................................................ 66 D. ................................................ 66 E CONCLUSIONS AND RECOMMENDATIONS ................................................................ . FINANCIALSUSTAINABILITY OF THE PAYGPILLAR 70 5 TRANSPORT ............................................................................................................... . 73 A INTRODUCTION . ....................................................................................................... 73 B CHARACTERISTICSOFTHEROADSECTOR.............................................................. 74 C.. PUBLICEXPENDITURE ........................................................................... 75 D. THEMAININSTITUTIONS INTHEROADSECTOR..................................................... ON ROADS 76 E FUTURE . 78 F. THERAILWAYSECTOR........................................................................................... FINANCING REQUIREMENTS...................................................................... 81 G RECOMMENDATIONS . .............................................................................................. 86 Mainrecommendationsfor the Railway Sector ............................................... Mainrecommendationsfor the Road Sector .................................................... 86 87 H. CONCLUSIONS ........................................................................................................ 87 Tables . Table 1.2: Fiscal Developments. 2001-2006....................................................................... Table 1 1:FYRMacedonia .Key Economic Indicators ..................................................... 3 6 Table 1.3: Government Revenues Structure -Comparison of Selected ECA Countries....8 Table 1.4: Macedonia and Selected Comparators: Government Spending, Fiscal Balance and Other Indicators (2006 or latest available year).......................................................... 10 Table 1.5: Economic Classification of Government Expenditures inSelected ECA Countries............................................................................................................................ 11 Table 1.6: Functional Classification of Government Expenditures inSelectedECA Table 1.7: Public Debt Dynamics and Fiscal Sustainability Projections .......................... Countries............................................................................................................................ 13 Table 1.8: Measures of Efficiency inGovernment Outcomes .......................................... 14 16 Table 2.2: Comparisons of Teacher Salaries, Macedonia, the EU15,and the NMS.........21 Table 2.1: Macedonia: Education Expenditures, 2002-2005 ............................................ 21 11 Table 2.3: Economic Classification of Education Spending. Macedonia and Selected Comparators....................................................................................................................... 22 Level of.............................................................................................................................. Table 2.4: Distributionof Goods and Services (non-staff recurrent) Expenditures by 22 NMS................................................................................................................................... Table 2.5: Expenditure Shares by Level ofEducation for Macedonia, the EU15, and the 23 Table 2.6: Total Education Expenditure per Capita Relative to per Capita Outlays for Primary Education, Macedonia,the EU15, and the NMS................................................. 23 Table 2.7: Average Expenditure per Primary Student, 2005............................................. 24 Table 2.8: Percent Scoring at DifferentCompetency Levels on PISA Reading Scale: Macedonia, EU 15, andNeighboring Countries................................................................ 25 Table 2.9: Annual Compulsory Instructional Hours by Age, Macedonia, EU15, andthe NMS................................................................................................................................... 26 Table 2.10: EnrollmentRatesfor Macedonia, the EU15, andthe NMS........................... 27 Table 2.12: Enrollment Ratesby Age and Household Consumption Quintile..................28 EUl5 andthe NMS............................................................................................................. Table 2.11:Educational Attainment of 25-64 Year Old Population for Macedonia, the 27 Table 2.13 : Enrollment Ratesby Age, UrbadRural Residence, and Household Table 2.14: Enrollment Ratesby Age, Gender, and Household Consumption Quintile...28 Consumption Quintile........................................................................................................ . . 28 Table 2.15: Percentof Total EnrollmentsinPrivate Schools at DifferentLevels of Table 2.16: Student-TeacherRatios, Macedonia, the EU15 andthe NMS....................... Education for Macedonia, the EU15, andthe NMS.......................................................... 29 30 Table 3.1: Morbidity and Mortality Indicators, 1991, 1995, 2000, 2004.......................... 39 Levels................................................................................................................................. Table 3.2: HealthOutcomes inMacedonia and Other Countries with Similar Income 39 Table 3-4:Macedonia: Revenuesand Spending of the Extrabudgetary Health Fund.......41 Table 3.3: Social InsuranceFinancing Mechanisms, EU8 and SEE Countries ................40 Table 3.6: HIF Expendituresby Level o f Care (2000-2004) ............................................ Table 3.5: Health Expenditures inEU8, SEE OECD and other MIC Countries...............43 44 Table 3.8: Health Seeking Behavior by Income Quintile.................................................. Table 3.7: Outpatient Contacts inSelectedCountries....................................................... 45 45 EuropeanRegion, 2004 or Latest Available Year ............................................................. Table 3.9: Inpatient Utilizationand PerformanceinAcute Hospitals inthe WHO Table 4.1: Revenuesand Expenditures of the Pensionand Disability Fund, 2006...........46 53 Table 4.2: Public and ImplicitDebt for Selected Countries inCentral and EasternEurope ........................................................................................................................................... 56 Table 4.3 : Approaches to PensionIndustry Structure of Charges..................................... 59 Table 4.4: Annuity Provisions inECA Countries ............................................................. 59 Table 4.5: Annuity Projections: Sensitivity to Fee Size, Number of Years of Saving and 61 Table 4.6: MinimumPensionsSize and Financing ........................................................... Gender (Inpercent o f averagewage) ................................................................................ 64 Table 4.7: Demographic Assumptions .............................................................................. Table 4.8: Economic Assumptions.................................................................................... 67 67 Table 5.1: Macedonia, Density of Road Infrastructure (latest observation available, 1997- 2003) .................................................................................................................................. 74 111 ... ........................................ Table 5.3: Annual Railway Expenditure (Inmillions of euro). 2002-2005 ...................... Table 5.2: EstimatedAnnual Maintenance Expenditure Needs 79 Table 5.4: Evolutionof M Z Financial Indicators .............................................................. 83 86 Figures Figure 1.1:Government Spending. Macedonia and Selected Comparators ..................... -11 Reading.............................................................................................................................. Figure 2.1:StudentPerformanceand GDP per Capita. Selected Economies. PISA 2000. 25 Figure 2.2: Enrollment Share Relative to Population Share by EthnicGroup and Level of Figure 2.3: Student-TeacherRatios at DifferentLevels of Education (1992-2005) .........29 Education (Inpercent. 2002) ............................................................................................. Figure 4.1: FYRMacedonia. Population over 65 years with PensionCoverage..............31 Figure 4.2: PensionExpenditure inMacedonia andEUCountries................................... 52 Figure 4.3: Contribution Levels andReplacement RatesinSelected ECA Countries......54 54 Figure 4.4: System and Demographic Dependency Rates in Selected Countries in2002 '' ........................................................................................................................................... 55 Figure 4.5: Foreign Investment Limits.............................................................................. 57 Figure 4.6: Investment Structure of PensionFunds (June 2006) ...................................... 58 Figure 4.7: Multi-Pillar vs .Mono-PillarPensions ............................................................ 63 Figure 4.8: Gross Replacement Rates for New Entrants.Mandatory PensionPrograms. Figure 4.9: Impactof Fees on Asset Level for Average Wage Earner.............................. MenAverage Wage Earners.............................................................................................. 64 Figure 4.11: Baseline Scenario: PDF Revenuesand Expenditures................................... Figure 4.10: PDF RevenuesMinusExpenditures (Excluding Budgetary Transfers)........65 68 69 Figure 4.12: Decomposition of the RequiredBudgetary Transfer .................................... 69 Figure 5.2: Percent of Firms Stating that Transport is a Problem for Doing Business .....75 Figure 5.1: Regional Comparison of RoadNetwork Quality............................................ 75 to 2005 ............................................................................................................................... Figure 5.3: Trends inRecurrent and Capital Expenditure inpercentageo f GDP from 1998 76 Figure 5.4: Car Ownership and per Capita Income (PPP) inSelectedEuropean Countries Figure 5.5: FuelPrices and Taxes ..................................................................................... ........................................................................................................................................... 78 80 81 Figure 5.7: M Z Freight Traffic. 1996-2005....................................................................... Figure 5.6: M Z Passenger and Freight Traffic. 1990-2005............................................... Figure 5.8: Evolutionof Staff Number and Productivity .................................................. 82 83 iv EXECUTIVESUMMARY 1. Progress in advancing political and economic reforms has been substantial this decade. The authorities have moved with determination in implementing the Framework Agreement for Peace (the Ohrid Agreement) that ended the 2001 hostilities, including in enhancing the representation o f minorities in governmental structures. This, together with economic reforms and the implementation o f the Stabilization and Association Agreement (SAA) with the EU, led the European Council to grant FYR Macedoniathe status o f a candidate country in December 2005. While no date to begin negotiations has been set yet, the goal o f EU integration i s shared by all key political parties, providing incentives for further reforms and cohesion in what is still a tense political dialogue within the country. 2. Macroeconomic stability was quickly restored and economic growth has gradually recovered in the years following the 2001 conflict. In 2003, the Government introduced a stabilization program focused on tighter fiscal policy and supported by the continuation o f the de facto peg o f the exchange rate against the euro. The budget deficit was virtually eliminated inthe first year o f the program and in the subsequent years spending has been sharply reduced as a share o f GDP. The fiscal adjustment helped improve financial stability and placed public debt ratios on a steadily declining path, with government debt amounting to about 30 percent o f GDP at the end o f 2006. The balance o f payments has improved since 2003, and especially since 2005, thanks to a surge in recorded private transfers. In fact, FYR Macedonia has the smallest current account deficit in the region equivalent to 0.4 percent o f GDP. 3. However, robust growth has remained elusive. Real GDP growth averaged about 3.5 percent a year during2003-2006, well below most countries in the region. Output remains driven by a few key sectors o fthe economy, and investment levels have remained relatively low at about 20 percent o f GDP. In contrast to the new member states o f the EU (NMS), FYR Macedonia has attracted only modest inflows of greenfield foreign direct investment (FDI), with only a small impact on the expansion o f output and exports. 4. Prospects for faster growth have improved following the implementation of key structural reforms in 2005 and 2006 and the new Government's commitment to further accelerate structural reforms focused on improving the investment climate and strengthening public sector governance. Real GDP growth picked to about 5.5 percent from a year earlier inthe first halfo f 2007, and formaljob creation, while still modest, has strengthened. 5. Still, a key challenge for the new Government is to adopt policies that would facilitate formalization of economic activity, including employment. In this context, the authorities are urgedto explore options to reduce social securitycontribution rates, starting with the elimination of the minimum contributionthreshold which represents an enormous disincentive for formal sector employment for low-wage and part-time workers. A cut in social security contribution rates should be carried out in a framework o f adjusting the method of financing o f health expenditures and perhaps the pay-as-you-go (PAYG) portion o f pensions to general revenues. Reductions in social security contribution rates would need to be accompanied by expenditure cuts (see below) andor increases in value added tax (VAT) rates or the personal income tax (PIT) rate. An increase in the latter could be accompanied by an increase in the tax- V exempt minimum to make the tax system more progressive. Concerted efforts could also be made to expand the tax base and further improve tax administration. As a comparison, Slovakia has successfully registered fast rates of real GDP growth and substantially reduced unemployment with a flat PIT, CIT andVAT rates harmonizedat 19 percent. The flat PIT rate in Estonia, another country that has experiencedrapidreal GDP growth, is set at 22 percent. A. PUBLIC EXPENDITUREDEVELOPMENTS AND CHALLENGES 6. FYR Macedonia's fiscal adjustment since the 2001 conflict has underpinned macroeconomic stability. Following a reduction in the fiscal deficit to nil in 2003, fiscal prudencehas beenmaintained, resultingin small surpluses in 2004 and 2005 and a modest deficit in 2006. With the adjustment program focused on sharp expenditure reductions, government spending fell from 44 percent of officially-estimatedGDP in 2001 to 34 percent of GDP in 2006, enabling a reductionin the tax burden. Relativeto GDP adjusted more fully for the size of the non-observedeconomy (NOE), spending is likely to amount to less than 30 percent of GDP, one of the lowest in Europe and Central Asia (ECA), but still higher than in some other emerging markets, such as Chile and Thailand. Reducing spending relative to GDP further is possible and should help ease further the fiscal burden on the economy, but the priorities for the Government should be in improving the efficiency of public spending, thereby raising the outcomes achieved. 7. Looking forward, the Government faces a number of challenges in improving the quality and effectivenessof publicspending. Thesechallenges needto be addressedtogether in the context of Governmentplansto reduce expenditures by an additional 2 percent of GDP over the next few years. Meetingthis goal, would present an opportunity to reduce social security contributionrates as discussed above. Stronger growth in real GDP, together with effortsto keep spending increases limited in many sectors should help achieve the goal. However, further reductions in spending will be difficult in the face of new spending pressures. Years of under-spending on road maintenance have left a large maintenance backlog that the authorities will be well advised to tackle within the available budget envelope. Moreover, new spending may be requiredto supportthe country's EUaccessionprocess. Reducingthe overall level of spending while meeting these spending pressures would require reductions in outlays elsewhere. One potential area for savings is defense, public order and security, where FYR Macedonia spends about 2 percent of GDP more than most countries in the region. Tackling this issue will be challenging, given the country's aspirations to join NATO, but exploring the options will be well worth the government effort. As discussed in the report, spendingon botheducation and health sectors in FYR Macedonia is modestly higherthan many countries in the region relativeto officially-estimatedGDP (and broadly similar relativeto GDP more fully adjusted for the size of the non-observed economy, NOE), but health and education outcomes are poor. All told, scope for spending reductions in these areas appears rather limited. In fact, the Government's key challenge in these sectors is to improve the quality and effectiveness of public spendingto ensurebetter outcomes. B. EDUCATION 8. Government spending on education as a share of officially-estimated GDP is modestly higher than among most countries in the immediate region. Student achievements are poorer relative to the same countries, however, and substantially more so comparedwith the countries of the EU. Enrollmentrates in secondary education are lowerthan in the EUand skills acquired in schools are not in line with market needs. Access to education seems equitable at the vi primary level, but gaps open up at the secondary and tertiary levels, with the level o f wealth, place o f residence and ethnicity all affecting enrollment rates. Moreover, substantial differences appear inthe amount o f spending on schools with different ethnic compositions. 9. Poor learning outcomes are one of the factors that have contributed to sluggish economic growth and high unemployment in FYR Macedonia. Boosting the productive capacity o f the economy should be accompanied by raising the quality and relevance o f education, key components to helping improve living standards and accelerate EU integration, both o f which are the top government priorities. The authorities are well advised to redouble their efforts to address these issues urgently. 10. The report concludes that the education sector at this time has adequate overall resources, but they are used unproductively,resulting in poor learning outcomes. It is not recommended that the authorities commit additional resources to the sector: the focus needs to be on improving the efficiency and equity o f spending. A move to capitation financing, as part o f the overall decentralization process, is recommended to help move toward a more equitable distribution o f resources across municipalities and create stronger incentives for improving the efficiency and outcomes o f spending. Increasing the amount o f instructional time that students receive in schools should also be a priority, including by increasing the amount o f time teachers spend on teaching rather than on non-teaching activities, increasing the number o f classes primary school students take, and revising the length o f these classes back to 45 minutes from 40 minutes at present. It is also recommended that savings from capitation financing be used to help stimulate poorer households' demand for education by reducing the cost burden o f education. Declining school-age cohorts over the next decades offers a fortuitous opportunity to reallocate expenditures to more efficient uses while raising enrollment rates. The authorities should make the best use o f this opportunity. 11. The government intends to spend substantial resources to equip every student with a computer in the next several years. The costs o f procuring computers will bejust a fraction o f the outlays on subsequent maintenance and operation, as well as training and preparing classrooms for installing computers. Curricula would also need to change to effectively use this new technology. Ifall o f these things are not done, then the potential benefits o f these computers may not be realized. Also, while the government should seriously consider taking steps to improve the use o f informationtechnology in schools, the priority should be improving education outcomes as identified in the report. The total resources needed for the "computer for every student" program could crowd out much needed resources for the reforms noted inthis report. Education: Summary ofRecommendations " 0 Move to allocate resources on a per-pupil basis to generate savings. 0 Slow the pace o f increase inteacher salaries. 0 Increase instructional time in schools. 0 Increase access to learningmaterials, especially textbooks. 0 Increase teacher training. I l'See Chapter 2, pp.3 1-35 for the full list o f recommendations. C. HEALTH 12. Health outcomes in FYR Macedonia are better than in countries with similar income levels, but these outcomes are achieved at high cost. Overall health spending levels is vii amongone ofthe highest inthe ECA regionand close to OECD levels. Despitehighgovernment spending, out-of-pocket payments, including informal payments are not small, thereby threateningfinancialprotectionamongvulnerable groups. Moreover, the bulk of health spending is allocated for relatively expensive inpatient and specialized interventions as compared with cost-effectivepreventive and primaryhealth care services. Productivity in health centers is low as compared with other countries in the ECA region and there is mal-distributionof key health sector inputs, includinghealthpersonnelandnumberof bedsfor acute care. 13. The report recommends that the authorities consider several priority avenues for reform. The priority areas for reforminthis report includea revisionof the benefits package for social health insurance, adoption of modern provider payment systems and tighter procurement rules for pharmaceuticals. Other recommended measures include the need to strengthen the regulatoryregime for healthcare institutionsand to improve the targeting of eligible persons for publicly-financed health insurance premiums, while enforcing contribution collection from nonexemptgroups. Reorganizingthe hospitalsector in line with the Medical Mapthe authorities are preparing should help streamline with the size of the sector and make it more attuned to the needsofthe citizens. Health: Summaryof Recommendations I' 0 Review andbeginreformingthe generous healthbenefitpackage. Consider shiftinga part ofthe financingofhealthcare from payrolltaxes to general revenues, startingwith analyzingthe fiscal impactof such a change. 0 Introducegeneric referencingfor outpatient drugs andexternal referencingfor outpatient drugs.Reviseco-paymentsfor pharmaceuticals. Develop andbegin implementinga plan for rightsizingthe healthsector, especially with regardsto hospitaland specialist outpatient healthservices. " See Chapter 3, pp.48-50for the full listofrecommendations. D. PENSIONS 14. The financial situation of the pension system worsened dramatically during the 1990s primarily due to the significant decline in the contribution base on the back of reduced formal employment. This, together with the build-up of substantial implicit pension obligations, led the authorities to begin implementingpension reforms, startingwith parametric changes of the PAYG pension pillar. A mandatory fully-funded second pillar was introduced in 2006. Pension reforms have been successful thus far, guided by the objectives of reducing the impact of demographic changes on the sustainabilityof the pension system and helping assure adequate pension incomes for bothcurrent and future generations of pensioners. More work will be neededto ensure that future implementationproceeds smoothly and the pension system copes with all challenges. 15. Under the report's realistic baseline scenario, the fiscal sustainability of the pension system looks likely to be attained (as measured by the deficit of the pension fund stabilizing at less than 1 percent of GDP a year), with pensions are likely to average 50 percent of wages (replacement rate), compared with 55 percent at present and 68 percent on average in the NMS. This result depends on the assumptions that the government will continue to implement parametric reforms already agreed, including the gradual increase in the retirement age. A recent reversalof a decision to give much larger weight to inflationthan wage growth in the pensionindexationformula will be counterproductive, however. viii 16. The likely outcome is reasonable and the decline in the replacement rate from current levels reflects the impact of the reforms the authorities are implementing. The difference with the EU should be interpreted with care, as many EU countries have yet to begin tackling the acute demographic problems they face. Moreover, stronger real GDP growth than recently observed should enable the decline in the replacement rate to be smaller than projected under the baseline scenario. Nonetheless, the authorities are well advised to begin work on managing public expectations that pension reforms in progress will likely result in a reduction in the replacement rate. These efforts should be coupled with efforts to strengthen growth in output and formal employment, and reduce substantially the fees charged on second pillar accounts to improve the return on fully-funded pensions. The authorities are also invited to consider ways to reduce further payroll contributionrates in a fiscally responsible way to help reduce informality. Pensions: Summary ofRecommendations '` Unifythe retirement age for menand women at the 64 years currentlyenvisaged for men and consider increasing the retirement age for both to 65 years. 0 Reduce the fees on fully-funded accounts, including to MAPAS. 0 Consider consolidating the supervision o fthe PAYG and the fully-funded pillars. " See Chapter 4, pp.70-71 for the full list o f recommendations. E. TRANSPORT 17. Years of under-spending on maintenance in the transport sector has led to a significant deterioration of the quality of public assets. The deterioration o f the quality o f roads, especially at the local level, has created road conditions worse than in any other country in Southeast Europe except Albania. Besides limiting scope for growth and exports, increasing vehicle amortization and resulting in a waste o f time and fuel, the deterioration o f the country's roads represents a serious fiscal risk which will grow further if not addressed quickly. The financial condition o f the FYR Macedonian Railways (MZ) has been poor and the company has been unable to maintain its existing network. The MZ has been implementing reforms to increase labor productivity through a program o f staff retrenchment and the company will be broken up in the middle of 2007 into a company in charge o f the infrastructure and another in charge o f the transport. 18. The report recommendsthat the Government address the maintenance backlog and increase spending on regular maintenance. Fully addressing the backlog could increase road expenditure by as much as 2 percent o f GDP a year for five years, but the Government will be hard pressed to afford such an increase within the available fiscal envelope and given ambitions to further reduce the overall level o f spending. This cost o f maintenance appears overstated by the outdated classification o f roads; prioritizing maintenance according to the level o f traffic should enable the government to focus on the most needed maintenance. The report also proposes a number o f institutional changes to improve the efficiency o f spending. Capacities to identify, appraise, prioritize, and effectively monitor the execution o f maintenance and investment spending should be strengthened. ix Transport: Summary of RecommendationsI' 0 Increase maintenanceoutlayswithin the available fiscal envelope to reducethe maintenancebacklog. Introduceopencompetitivebiddingfor maintenanceactivities. 0 Strengthenthe technical capacity ofMoTCandFNRRstaffto improvethe identification andprioritizationof roadsmaintenanceand investment expenditures. ''SeeChapter 0 StrengthenMZ-INationalProgramandMZ-T businessplans. 5, pp.86-87 for the full list of recommendations. F. CONCLUSIONS 19. Reducing spending by 2 percent of GDP over the next several years while tacking fiscal pressures will require a combination of cuts in some categories of existing spending, spending restraint in others and stronger real GDP growth. Spendingpressures includecosts relatedto advancingEUintegrationand on reducingthe substantial backlogof road maintenance. This report suggests that measurable spending cuts are not likely to be found in health and education where the challenge will be to improve the efficiency of spending and reallocate outlays to significantly improveoutcomes and support growth. The demand on the budget from furthering pension reform will increase in the following years, but should ease subsequently. Together with sustained progress in advancing structural reforms and revamping government spending, increasing the formalization of economic activity is the key to stronger growth in output. The authorities are urged to move to reduce social security contributionrates, starting with unifying and then abolishingminimum contribution thresholds. Reductions in contributionrates should be carried out inthe context ofreformingthe financingof social transfers, mainly health and further on pensions, toward general revenues. To ensure that the resultingreductionin revenues does not endangerthe authorities'targets for the fiscal deficit, a cut in contributionrates should be offset by more ambitious spending reductions or increasesin the VAT andPITrates. X 1. THE STRATEGIC SETTING A. INTRODUCTION 1.1 FYR Macedonia has made measurable progress since the Framework Agreement for Peace (the Ohrid Agreement) ended the 2001 hostilities. Progress in implementing the Agreement has been considerable, including in enhancing the representation of minorities in governmental structures. Two consecutive regular parliamentary elections in 2002 and 2006 were held peacefully and largely democratically. On December 16, 2005 the European Council granted candidate country status to FYR Macedonia, in recognition of the country's substantial progress in completing the legislative framework related to the Ohrid Agreement and in implementingthe StabilizationandAssociationAgreement (SAA). While no startingdate for the negotiations has been set, the shared goal of EU accession providesintegrative incentivesfor all political parties in what is still a tense political and ethnic dialogue within the country. As pointed out in the latest EC Progress Report (November 2006), the limited political dialogue seriously impedesthe country's ambitionto accede to the EU. 1.2 Macroeconomic stability was quickly restored and economic growth gradually recovered following the 2001 conflict. In 2003, the Government instituted a successful stabilization program focused on tightened fiscal policies and supported by the continuation of the defacto peg of the exchange rate to the euro. The budget deficit was virtually eliminated in the first year of the program and expenditures were sharply reduced as a share of GDP. The fiscal adjustment quickly placed public debt ratios on a declining path. The current account deficit and the overall balance of payments have improved since 2003, especially following a surge in recorded private transfers in 2005 and 2006. Progress on structural reforms has also resumed, buildingon the strong stabilizationefforts. 1.3 However, robust growth has remained elusive and living standards have not improved. Growth averaged about 3.5 percent a year during 2003-2006, well below most countries in the region. Output expansion remains narrowly based on few key sectors and investment levels have remained relatively low, at about 20 percent of GDP. Until 2006, few new jobs were created, keeping unemployment oversized. Unemployment, in turn, is strongly correlatedwith poverty, with the country's povertyrate andprofile remaininglittle changed. 1.4 There are signs that economic performance is improving. While preliminary official data indicatethat real GDP growth slowed to 3.1 percent in 2006, a number of indicatorssuggest that growth exceeded 4 percent.' Building on strong implementation of structural reforms, it appears that the economy is well placedto grow faster in2007 and beyond. Credit to the private sector has been growing rapidly, the employment rate has now begun to improve, and the investment climate is improvingwith the continued implementationof critical structural reforms. ' The State Statistical Office is receiving technical assistance from the IMF to improve their national accounts statistics. IMF staff estimate in a forthcoming review o f the Stand-By Arrangement (April 27, 2007) that real GDP growth was more than 4 percent in2006. A number o f key indicators were stronger in 2006 than in 2005. These include a shift to a 17 percent increase in the imports o f investment and intermediate goods in 2006 compared with a decline o f about 3.6 percent in 2005. Sales of passenger cars surged 24.9 percent in 2006 after remaining little changed in 2005. Real wages and employment also grew in2006 at faster ratesthan in2005. FYR Macedonia's aspirations to join the EU have further strengthened investor sentiment. The experience ofthe New Member States ofthe EU(NMS) illustrateshow Europeanintegrationcan foster the overall political, economic and administrative reforms in prospective new member countries. While an acceleration of growthand improvedliving standards ultimatelywill depend on the private sector responseto ongoing structuralreforms, there is a role for fiscal policy and public expenditure to support both growthandEUintegration. 1.5 This Public Expenditure Review analyzes a number of issues that bear on the effectiveness, level,and composition of publicspending. The overridingobjective ofthe report i s to better understand the nature and composition of public expenditure to support growth and EU integration. The report reviews the challenges and options in the areas of education, health, pensions, andtransport. B. MACROECONOMIC SETTING Output, Inflation, Living Standards 1.6 Economic activity picked up in 2007, with real GDP rising about 5.5 percent from a year earlier in the first half of the year, following a period of only modest expansion after the conflict of 2001. While direct conflict damagewas limited, output contractedby 4.5 percent in 2001, investment markedly dropped and defense-related expenditures pushed both the fiscal and current account deficitsto 7 percent of GDP. The effects of the conflictwere still clearly felt in2002, as investment and export growthremainedsluggishwhile the fiscal and current account deficits remained high. The current account deficit reached 9.4 percent of GDP in 2002 and internationalreserves fell to 3.5 monthsof imports (Table 1.1). 1.7 Growth began to recover after 2002, but the performanceof the economy has lagged most countries in the region. Recorded growth averaged only 3.5 percent during 2003-2006. Such growth rates place FYR Macedonia amongthe slowest growing economies in ECA in this period. Also, the recovery remains narrowly based on a few key sectors and unemployment remains high. While official data have consistently overstated unemployment, at about 36 percent FYR Macedonia's official unemployment rate is among the highest in the region. High and persistent unemployment statistics reflect low newjob creation in the formal sector. Inpart this has been due to an overly restrictivelabor market (until mid-2005)and a high tax wedge. It has also beendue to the poor corporate governancewhich emergedfollowingthe mainly insider- oriented privatizationprocess in the mid-1990s, which brought the newly-privatizedcompanies neither adequate knowledge nor access to markets and finance. On the contrary, post- privatization enterprise restructuring has been limited and unfair competition practices have discouraged potential new firms from entering the market. An overly burdensome business regulatory environment until recently further hampered new business startups and job creation. Collectively, these forces have contributedto the high unemployment rate and have encouraged informality. 1.8 While recent structural reforms aimed at improvingthe investment climate should begin to bear fruit, domestic and foreign investmentlevels have been low and are likely to remain a constraint on growth. At about 20 percent of GDP, investment in FYR Macedonia has lagged significantly the faster growing economies in the region. Inflows of foreign direct investment(FDI) havealso beendisappointingover the past decade. Apart from the spike inFDI in 2000 and 2001 when the largest bank and the telecom company were privatizedand in 2006 when the electricity distribution company was sold, inflows of FDI has averaged 1.5 percent of 2 GDP, one o f the lowest rates among transition countries. Inflows of greenfield FDI have been negligible. Table 1.1: FYR Macedonia Key EconomicIndicators - 2002 2003 2004 2005 2006 2007 Actual Est. National Accounts RealGDP growth (Inpercent) 0.9 2.8 4.1 4.1 3.7 5.1 Gross Investment (Inpercent of GDP) 20.6 20.0 21.4 20.7 21.8 23.1 Gross National Saving(Inpercentof GDP) 10.6 17.1 12.9 18.0 20.9 20.8 External Accounts (Inpercentof GDP) Exports of Goods and Services 36.0 37.6 39.6 44.0 47.4 53.5 Imports of Goods and Services 57.8 55.1 61.9 62.8 67.2 72.1 Current Account Balance, Including Transfers -10.0 -2.9 -8.4 -2.7 -0.9 -2.3 ExternalDebt 44.1 39.5 37.9 38.8 37.7 35.1 Foreign ExchangeReserves (Inmonths ofnextyear's imports) 3.4 3.2 3.2 3.7 4.1 4.6 Fiscal Accounts (Inpercentof GDP) Revenues 35.2 38.4 36.5 35.2 33.6 35.4 Expenditures 40.5 38.5 36.1 35.0 34.1 34.6 Overall Balance -5.3 -0.1 0.4 0.2 -0.6 0.8 Public Debt 43.O 39.0 36.7 39.5 33.2 28.8 Inflation (Inpercent, period average) 1.8 1.2 -0.4 0.5 3.2 2.3 Sources: State Statistical Ofice, Ministry of Finance, National Bank of the Republic of Macedonia. 1.9 FYR Macedonia's failure in attracting FDI has hindered a stronger growth in output and exports. FDI has played a critical role in the N M S in transferring technology, market access, know-how, and crucial intra-industry trade linkages. FYR Macedonian exports have grown on average at a modest rate o f 4 percent a year duringthe last decade. The country's share in world exports recovered slightly in 2005 and 2006, but remains below the levels of the mid-1990s. FYRMacedonian exporters have been able to regain market share in some traditional trading partners, in particular Serbia, Greece, and Germany, but have lost shares in eastern markets such as Ukraine and Russia, and have been unsuccessful in penetrating new markets. FYR Macedonia still specializes in low-value added goods (iron and steel, textiles, tobacco, agriculture produce). 1.10 The country's modest economic growth has not been sufficient to improve living standards, especially given the modest job creation. The employment rate remains among the lowest in the region. The employment rate fell from about 40 percent in the years preceding the 2001 conflict to about 37 percent in 2004. Since then the employment rate has begun to recover and stood at 39.6 percent in 2006. This largely "jobless" growth has not led to a reduction in poverty. An estimated 21 percent o f the population lives below the absolute poverty line. 1.11 The de facto exchange rate peg of the Denar to the euro has helped ensure price stability for more than a decade. Consumer price inflation has remained in single digits with the exception o f a brief period o f deflation in 2004 when a sharp decline in international reserve cover to 3 months o f imports forced a tightening o f monetary policy in defense o f the exchange 3 rate peg. Lack o f administered price adjustments (including o f prices o f electricity) has also contributed to keeping inflation low. Balance of Payments and ExternalDebt 1.12 Stronger export growth and, more importantly,a surge in privatetransfers, largely eliminated the current account deficit in 2005 and 2006. The current account deficit fell to 1.4 percent o f GDP in2005 following a 4 percent o f GDP increase inrecorded private transfers to 17.5 percent o f GDP, the highest in Southeast Europe. Recorded transfers rose by an additional one percent o f GDP in 2006 and, together with an unanticipated delay in making a large dividend payment by the telecom company to foreign shareholders, led to a further reduction inthe current account deficit to 0.4 percent o f GDP in 2006. The overall balance o f payments has strengthened considerably, with foreign exchange reserves increasingto about 4.5 months o f imports. External debt has gradually declined to less than 40 percent o f GDP, helping improve the country's access to international capital markets. Monetary Developments 1.13 With the Denar fixed to the euro, the strong balance of payment position has led to an easing of monetary conditions since 2004. Robust international reserves, the prepayment o f external debt, and continued fiscal prudence have helped improve market,confidence, resulting in lower market interest rates. Credit to the private sector has grown rapidly, increasing by 20 percent in 2005 and 30 percent in 2006, albeit from a relatively low base o f 22 percent o f GDP in 2004. Deposits have also grown rapidly from a low base, led by a rapid build up in Denar- denominated deposits. 1.14 Price competitivenessappears broadly appropriate. The CPI- and the PPI-based real effective exchange rates (REER) have steadily and modestly depreciated since the 2001 conflict by a cumulative 7 percent and 13 percent, respectively. The REER deflated by unit labor costs has fallen more, depreciating by about 30 percent over the same period. Growth in average real wages has averaged 3.5 percent a year in recent years and has been driven mainly by the nontradable sectors. Real wages in manufacturing and agriculture have grown at a much slower pace o f less than 1percent, slower than productivity growth. Structural Reforms 1.15 Structural impediments have been the largest constraints to a stronger improvement of competitiveness. The concentration o f the economy and exports in a few key sectors with low value added and declining world trade shares suggests that structural reforms are needed to encourage a reallocation o f resources towards more dynamic and productive sectors. With the implementation o f key structural reforms in 2005 and 2006 and with the new Government's commitment to further accelerate such reforms, prospects for faster growth have improved. 1.16 A number of reforms to improvethe investment climate are under implementation. A comprehensivejudicial reform was launched in2005 which, ifvigorously implemented further, will firmly establish the rule o f law includingcreditor, contract and property rights. A new Labor Law was enacted in July 2005 to introduce more flexibility into the country's labor markets. In the banking sector, supervision, regulation, and governance are being strengthened. Foreign banks have recently showed increasing interest in entering the market. A number o f reforms have 4 easedthe entry, exit andregulationon ongoing businesses. The launch of a new "one-stop shop" business registration system in January 2006 significantly eased the time, cost and burden of registeringnew business. The new system provides a transparent registrationfor new businesses and has reduced registration time from 48 days to 5 days. A revision to the bankruptcy framework strengthens creditor rights and focuses on the institutional development of the bankruptcy trustee profession. Reforms are underway to reduce the burden of unnecessary regulationon businesses. Around 2,000 regulatoryacts which constrain business activitieshave been identifiedand will be rationalized. Inaddition, a compulsory regulatoryimpact assessment of all new legislation will become mandatory. The latest Progress Report of the European Commission notes that FYR Macedonia is well advanced in establishing a functioning market economy, having maintained a broad consensus on the essentials of economic policies, though sustained efforts will remain critical to enable the country to cope with competitive pressure and market forces withinthe Unioninthe rnedium-tenn2Further,the authorities have make progress in reformingthe tax administrationand social security. 1.17 Reformsto improvethe effectiveness of publicsector governance and public service delivery have also been underway. The treasury reforms have resulted in sound and effective controls over spending. The budgetpreparationprocesshasbeenrevampedwith the incorporation of the extra-budgetary funds in the process and the introduction of ceilings for budget user expenditures. In addition, the budget is prepared based on a medium-term fiscal and macroeconomic outlook and also has greater alignment with the strategic priorities of the authorities. Internalaudit is rapidly improvingand reforms are ongoing in public procurement. The introduction of the second pension pillar in 2006 was an important step in improving the sustainabilityof the pension system. The government is advanced in introducinga third pillar of the pension system. Most recently, reforms in health financing are being introduced to reduce fiscal risksto the government and improvethe qualityof services. C. RECENT FISCALDEVELOPMENTS 1.18 FYR Macedonia's fiscal adjustment since the 2001 conflict has underpinned macroeconomic stability. In 2003, the Government introduced a fiscal adjustment program supported by a new IMF Stand-by Arrangement. The Government did manage to reduce the budget deficit in 2002 but expenditures remained high and revenues were boosted by a tax on financial transactions which had been introduced in mid-2001. In 2003, the Government's new fiscal adjustmentprogramquicklycut the budget deficit to 0.1 percentof GDP (Table 1.2). Fiscal policywas tightened further in subsequentyears andthe budget movedto small surpluses in 2004 and 2005. The fiscal balancereturnedto a small deficit in 2006, largelydue to the shift of part of pension contributionsto new privatepensionfunds in the wake of the launch of the funded pillar of the pensionsystemand also as a resultof emergency fiscal transfers to the energy sector. 1.19 The adjustment programfocused on sharp expenditure cuts. The public expenditure to GDP ratio declined in each year during the adjustment program, falling from 44 percent of GDP in 2001 to 34 percent in 2006. Demobilizationof army and policereservists and declining purchases of military equipment generated large savings in 2002 and 2003 (see goods and non- labor services line in Table 1.2). Thereafter, expenditure cuts were more broadly spread, with public consumption, capital investment and transfer payments falling by 1.5 percent of GDP, 1.2 percent, and 0.7 percent, respectively, between 2003 and 2006. The decline in public EuropeanCommission, 2006, ProgressReport. 5 consumption between 2003 and 2006 reflected reductions both o f the wage bill3and outlays on goods and non-labor services. Capital expenditure, already modest by regional standards at about 4 percent of GDP, was cut to about 3 percent o f GDP by 2006. Table 1.2: Fiscal Developments,2001-2006 (Inpercent ofGDP) 2001 2002 2003 2004 2005 2006 Total Revenue and Grants 37.0 38.2 38.4 36.5 35.5 33.7 Tax Revenue 31.5 33.1 30.7 30.8 29.8 29.9 Taxes on Incomeand Profit 4.4 4.2 4.3 3.8 3.8 4.3 Social Insurance Contributions 10.8 10.4 10.8 10.6 10.1 9.7 Domestic Taxes 12.2 13.1 12.9 13.9 13.7 13.0 Import Duties 2.6 2.6 2.4 2.2 1.9 1.8 Other Taxes 1.4 2.7 0.3 0.3 0.3 0.6 Non-tax and Capital Revenue 5.3 5.2 6.3 5.1 5.2 3.8 Non-tax Revenue 4.2 4.4 5.4 4.3 3.7 3.3 Capital Revenue 1.1 0.8 1.o 0.8 1.5 0.6 Grants 0.2 0.0 1.3 0.6 0.5 0.5 Total expenditures 43.6 42.2 38.5 36.1 35.3 34.3 Current Expenditure 39.7 37.4 34.0 32.7 31.7 31.7 Goods and services 18.0 15.9 13.5 12.9 12.6 12.0 Wages and Salaries 7.5 8.1 8.5 8.3 8.0 7.7 Goods andnon-labor services 10.5 7.9 5.0 4.6 4.6 4.2 Transfers 19.8 20.0 19.4 18.9 18.2 18.3 Interest 1.9 1.5 1.1 0.9 0.9 1.o Capital expenditure 3.8 4.6 4.1 3.1 3.6 2.9 Other 0.1 0.2 0.3 0.3 0.0 0.1 Fiscal Balance -6.6 -3.9 -0.1 0.4 0.2 -0.6 Note: 2001and 2002 data differ from previously published IMF data. Two adjustments were made. Discretionary spending of budget users financed from the "special revenue account" was incorporated into official data from 2003. Estimatesin this table includesuch revenueand expenditurein 2001 and2002. Also, convertiblenotes issuedin 2002 to compensatedepositorsof the failed "TAT pyramid" scheme in 1997 are excludedfrom this table. Sources: Ministry of Finance, IMF,and World Bank staffestimates. 1.20 Revenues remained high in 2003 as an influx of donor grants and non-tax revenue from the sale of commodities in the "strategic reserve" offset the elimination of the tax on financial transactions. Revenue subsequently declined by about 5 percent o f GDP by 2006. Non-tax revenue led the decline as the scope o f activity o f Commodity Reserve Bureau was curtailed. Also, trade liberalization led to a decline in customs duties and the sluggish labor market led to an erosion of the contribution base and a decline in payroll taxes. 1.21 The fiscal adjustmentof the past few years has improvedthe sustainabilityof public debt, the latter falling from 49 percent of GDP in 2001 to 34 percent at the end of 2006. While such levels o f indebtedness are moderate by international and regional standards, debt is subject to exchange rate risks as nearly all public debt i s denominated in, or indexed to, foreign currency. Efforts have been made to develop the government securities markets ahead o f the The wage bill rose in2003, following a 16percentincrease inpublic sector salaries inmid-2002. 6 launchingof the funded pension system in 2006, but the share of Denar-denominated debt will remainsmall for a number ofyears. The Government issuedan inaugural150 million Eurobond in December 2005 which was 4 times over-subscribed, usingthe proceeds to pre-paythe London Club inJanuary 2006. FYR Macedonia subsequentlyreached agreement to prepay Paris Club in early 2007 and plansto prepay some debt to internationalfinancial institutionslater in 2007. 1.22 While the size of the fiscal adjustment is reflected in the reduction of the fiscal deficit and the ratio of public expenditure to GDP, the overall quality of the fiscal adjustment is more difficult to measure. On balance the fiscal adjustment was accomplished through expenditure cuts rather than increases in revenue. In fact, the sound fiscal position has providedsome scope for the Government to cut corporate and personal income tax rates in 2007, with further cuts plannedfor 2008. 1.23 The reductionin the wage bill was accomplished in challengingcircumstances. The Government was able to improve the representation of ethnic minorities in the public sector, fulfilling its obligations under the Ohrid Agreement; while implementing a decompression of civil servant wages and simultaneously ensuringa gradualreductioninthe wage bill in percent of GDP. Also, a major pension reform was introduced including parametric changes and the introductionof a funded secondpillar. 1.24 The need to transfer the equivalentto about 0.3 percentof GDP to the energy sector late in 2006 indicates that some contingent liabilities had built up in the broader public sector. Reforms of the energy sector are needed, including improved payment discipline, to ensurecost recoveryandreducegovernment subsidies. 1.25 The Government has increased budgeted capital expenditures, but outcomes have consistently fallen short of plan. Under-fundingof operations and maintenance (O&M) has led to deterioration in important public assets such as the road network, health care facilities, and educationalfacilities. 1.26 While the level of fiscal deficit and public expenditures favorably positions fiscal policy in FYR Macedonia to support growth, a deeper look at the outcomes of public spending in key sectors indicates that further reforms are needed for public expenditure to better supporteconomic growth and EUaccession. D. GOVERNMENTREVENUES 1.27 As government revenues have fallen to 34 percent of GDP over the last several years, the fiscal burden on the economy is lower than many countries in South Eastern Europe and the lowestamong the countries of the former Yugoslavia. Only Latvia, Lithuania and Slovakia from the NMS and Romania and Albania from the sub-regionhave lower public revenuesto GDP ratios.Excludingcapitalrevenuesandgrants, the fiscalburdenwas 33.5 percent of GDP in 2005, inthe middleof all ECA countries(Table 1.3). 1.28 Revenues from income taxes are very low, averaging about 4 percent of GDP over the last several years, 8 percent lower than in the EU on average and 1.5 percent below the average for the region. This reflects both low rates (ahead of the introduction of the flat tax in 2007, 98 percent of personalincometax was paidat the lowest personal income tax rate) andnarrowbases. This is particularly so for the corporate incometax (CIT) where the allowances for fixed assets in FYR Macedoniaare probablythe most generous inEurope. 7 Table 1.3: GovernmentRevenuesStructure-Comparisonof Selected ECA Countries Total Revenuesand Grants Total Revenuesexcl. Grants Tax Revenues Corporate IncomeTax PersonalIncome Tax Social Contributions 12.1 15.9 11.9 12.0 10.7 11.2 10.7 14.1 8.9 11.4 9.2 9.4 4.7 Other Direct Taxes 0.0 1.4 0.0 0.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Taxes on Property 0.6 0.6 0.0 0.9 0.0 0.4 1.5 0.6 0.8 0.5 0.4 0.9 0.0 Domestic Taxes 20.3 13.8 17.7 14.6 15.2 12.0 10.9 11.0 10.0 11.3 10.5 11.3 10.2 VAT 14.9 8.7 12.2 8.4 10.2 8.2 6.7 6.6 6.7 7.5 6.4 6.9 7.6 Excises 5.1 3.5 5.6 3.5 5.0 3.8 4.1 3.5 3.3 3.3 3.0 3.3 2.1 Other Domestic Taxes 0.3 1.6 0.0 2.7 0.0 0.0 0.1 0.9 0.0 0.5 1.1 1.0 0.5 InternationalTaxes 0.8 0.3 2.6 0.2 0.8 0.2 0.2 0.1 0.2 0.1 0.2 0.7 1.8 Other Taxes 0.2 0.1 2.5 0.5 1.1 0.1 0.3 0.0 0.3 0.0 0.0 0.0 2.3 Non-taxRevenues 5.8 5.3 3.6 4.2 6.9 4.7 7.8 2.3 6.6 3.4 1.3 1.5 2.1 CapitalRevenues 0.5 0.0 0.5 0.8 0.0 0.0 0.4 0.4 0.0 0.3 0.7 0.0 0.0 Other Revenues 0.0 0.0 0.0 0.4 0.0 0.0 0.9 0.0 0.1 0.0 0.0 0.1 0.0 Grants Legend: CRO- Croatia; H - Hungary; SLO - Slovenia; SaM - Serbia and Montenegro; PL - Poland; CZ - Czech Republic; BG - Bulgaria; EST - Estonia; SVK - Slovak Republic; MKD- Macedonia; L T - Latvia; L V - Lithuania; ROM-Romania; ALB -Albania. Note: Datarefers to 2004. Source: ECA Regional FiscalDataset. 1.29 Despitelow personalincome (PIT) rates, relatively large social security contribution rates raise concerns that the labor tax wedge is contributing to the high unemploymentin the country. The labor tax wedge for the average worker in FYR Macedonia amounted to 67.2 percent o f net wages in 2006 compared to an OECD average o f 57.1 percent, but the wedge was much higher on the margin for lower-paid workers. (The introduction o f the 12 percent flat PIT in 2007 and the further cut to 10 percent in 2008 should reduce the wedge to 59.8 percent). Social security contributions accounted for 82.3 percent o f the labor tax wedge, with contribution rates totaling 32 percent of gross salaries (21.2 percent for pension and disability insurance, 9.2 percent for health insurance, and 1.6 percent for unemployment insurance). These contribution rates are somewhat higher than OECD averages and, in a broader international comparison, excessive relative to countries with similar levels o f income per capita. A unique feature o f FYR Macedonia's social security contributions is the minimumbase for contributions which was set as a percent o f the average salary. This feature is complicated by the imposition o f multiple minimum bases for different social contributions and for workers in different industries. Contributions on many low wages are calculated on the basis o f the minimumbase rather than on the actual wage. This practice significantly increased the labor tax wedge for lower paid and part time workers. In2007, as part o f the base harmonization reform, the Government is unifyingthe minimumbase for pensions and unemployment contributions at a high 65 percent of the average wage, The minimum base for health contributions will be increased from 50 percent o f the average wage to 65 percent in 2008. 1.30 However, given the exceptionally low rate of employment in FYR Macedonia: the Government should investigate options for reducing the labor tax wedge further. The employment rate (40 percent) is lower than all other countries in the region, and well behind the 68 percentaverage amongthe EUl5 countries. 8 Eliminatingthe minimumbase for all social contributionswould simplify the system and reduce the labor tax wedge especially for low wage and part-timeworkers. The Government couldalso consider reducing the contributionrates themselves. Reductions in contributionrates should be carried out inthe context of reformingthe financingof social transfers, mainly health and further on pensions,toward general revenues. So that the resultingreductioninrevenues does not result in an increase in the fiscal deficit, the drop in contributionrevenues should be offset by more ambitious spending reductions or increases in the VAT and PIT rates. An increase in the PIT could be accompaniedby an increase in the tax-exempt minimum to make the tax system more progressive. Some EU countries such as Sweden and the United Kingdom have strong tax financed systems and some countries such as France and Spain, with a long tradition of social insurance, have also been shifting to greater reliance on general revenues. The Government's plan to reduce overall expenditures by an additional 2 percent of GDP in the next few years presentsan opportunityto cut contributionrates. 1.3 1 Aside from distortionary social security taxes, the bulk of revenues are generated from less distortionary indirect taxes. A VAT was introduced in 2000 with a standardrate of 18 percent and a lower 5 percent rate applying to food, water supply, publications, books and newspapers, and certain agricultural inputs. Over the past few years, the range of goods and services subject to the lower rate has beennarrowed. There is a small group of zero-ratedgoods and services such as exported goods, international air transport, and sale of goods to the free economic zones. Improved collection efforts and strengthened compliance should be the key steps to take for the Governmentto offsetcuts incontributionrates. E. THELEVEL GOVERNMENT OF SPENDING 1.32 Government efforts in reducing spending and the fiscal deficit have been commendable. Government spending fell to 34 percent of GDP by 2006, placing FYR Macedonia among the countries with the lowest government spending relative to GDP in Southeast Europe. Spending is roughly in line with the Europe and Central Asia (ECA) average (33 percent of GDP) but lowerthanthe average for the NMS(39 percent of GDP). 1.33 I n fact, since official data likely understate nominal GDP in FYR Macedonia, expendituresare probably even lower as a share of GDP. The State Statistical Office uses only 2 of the 8 categories of Eurostat's methodology for incorporating estimates of the non- observed economy (NOE) into official GDP statistics. The adjustments for NOE amount to 20 percent of the observed economy in FYR Macedonia (or 16.3 percent of overall GDP), compared with unofficial estimates that are twice as large. Still, using official GDP data, spending is only marginallyabove a trend line relatinggovernment expenditures and income per capita in PPP terms for 25 ECA countries and other countries with a similar level of income (Table 1.4 andFigure 1.1). 9 Table 1.4: Macedonia and Selected Comparators: Government Spending, Fiscal Balance and Other Indicators (2006 or latest available year) 2005 GNI per capita, FiscalBalance Expenditures Revenues at PPP (In percentof GDP) (InUS$) EuropeandCentralAsia -0.9 33.1 32.1 9,152 NMS11 -1.7 39.3 37.7 14,918 SEE /2 -0.8 38.7 37.8 8,435 High-growthcountries/3 -0.6 31.4 30.8 9,670 LAC41 -1.7 29.5 27.7 8,116 EastAsia andPacific -0.4 27.7 27.3 5,914 Ireland 1.1 33.3 34.3 34,720 Austria -1.6 49.9 48.3 33,140 Slovenia -1.1 42.9 41.8 22,160 Czech Republic -1.9 41.1 39.2 20,140 Portugal -5.7 47.3 41.6 19,730 Hungary -7.6 50.7 43.1 16,940 SlovakRepublic -3.1 39.2 36.1 15,760 Estonia 1.6 35.6 37.2 15,420 Lithuania -1.3 34.6 33.3 14,220 Poland -3.9 43.1 39.2 13,490 Latvia -1.2 37.0 35.8 13,480 Croatia -4.1 49.0 44.9 12,750 Chile 4.7 21.1 25.8 11,470 RussianFederation 8.1 31.6 39.7 10,640 Malaysia -2.9 27.9 25.0 10,320 Mexico -1.5 24.5 23.0 10,030 Romania -0.8 31.1 30.3 8,940 Bulgaria 2.3 38.3 40.6 8,630 Thailand 0.4 21.3 21.7 8,440 Turkey -3.7 38.7 35.0 8,420 Brazil -3.0 45.2 42.2 8,230 Tunisia -2.9 32.6 29.7 7,900 BH 0.8 49.4 50.1 7,790 FYRMMacedonia (2006) -0.6 34.3 33.7 7,080 Ukraine -2.3 42.0 39.7 6,720 Albania -3.6 28.0 24.4 5,420 Armenia -2.6 16.0 13.4 5,060 Georgia -2.4 25.4 23.O 3,270 1/ NMS, including Romaniaand Bulgaria. 2/ SEE includesCroatia, Romania, Bulgaria, Albania, BHand Macedonia. 3/ High-growth countries are those growingby above 6 percent on average during 2003-2005. 4/ Latin American and the Caribbean. Source: World DevelopmentIndicators and World bank staff estimates. 10 50 ................................................................................................................ + -. + c) 6 20 -.................................................................................................................. *+ 6 15 + Source: World Bank staff estimates. Economic Classification 1.34 The composition of expenditure, however, is not as supportive of growth as in other countries in the region. While total spending in percent of GDP is around 90 percent of the average of other NMS and SEE countries, FYR Macedonia's spending on capital expenditures and maintenance is among the lowest in the region. Outlays on wages and transfers are somewhat higher than regional comparators, while interest payments are much lower due to modest levelsof public debt (Table 1S). 1.35 The discussion below refers to ratios to officially-estimated GDP. An adjustment of officially-estimatedGDP to reflect more fully the NOEwill result in smaller ratios of spending to GDP, inter alia Total Expenditures 516 493 4 7 9 455 4 3 7 4 1 9 4 0 0 3 7 9 3 6 2 3 5 9 3 1 9 307 2 9 4 Current expenditures 427 4 3 6 4 3 4 4 1 6 4 1 0 3 2 1 3 5 8 3 4 4 3 2 8 325 2 8 8 2 7 6 2 4 0 Goods andServices 195 183 2 1 4 1 8 1 1 9 2 3 0 1 7 0 148 1 5 1 1 5 0 147 1 2 1 9 5 Wages and Salaries 108 120 1 1 9 103 6 7 3 0 6 1 7 8 5 9 7 3 8 9 4 9 6 4 Transfers Subsidies InterestPayments CapitalExpenditures Legend: CRO- Croatia; H-Hungary; SLO -Slovenia; SaM- Serbiaand Montenegro; PL-Poland, CZ -Czech Republic; BG-Bulgaria; EST -Estonia; SVK - Slovak Republic; MKD-Macedonia; L T- Latvia; L V Lithuania; - ROM-Romania; ALB -Albania. Note: Data refers to 2004. Source: ECA Regional Fiscal Dataset. 11 1.36 Current transfers account for more than half of all expenditures. The largest share o f transfers (1 1.3 percent o f GDP) is to households for pension and disability insurance, social assistance, and unemployment benefits. Transfers include healthcare spending. 1.37 Expenditures on goods and non-labor services are lower compared to most countries This has led to a deteriorating quality o f public assets such as roads, schools, and hospitals. It also led to a deteriorating quality o f public services provision. 1.38 Capital expendituresin FYR Macedonia are amongthe lowest in the region relative to GDP. In fact, actual capital expenditures are likely to be even lower than officially reported, as it is hard to determine if capital transfers (which account for more than 40 percent o f capital expenditures) are expended for the intended purpose or usedto supplement current spending. 1.39 I t is also unclear how productive public investments are in FYR Macedonia. The Public Investment Program lacks proper prioritization o f projects, monitoring i s weak and linkages among sector strategies are almost non-existent. As a result, execution o f projects is frequently delayed with completion ranging from 70-80 percent o fthe budgeted amounts. FunctionalClassification 1.40 A functional review of Government expenditures also reveals concerns about the structure of spending. Several features distinguish FYR Macedonia's expenditure pattern (Table 1.6). 0 FYR Macedonia has the highest security-related expenditures in percent of GDP. The allocation to public order and safety i s the largest in the regional comparators, while defense expenditures are higher than all countries except Bulgaria and Croatia. The high security-related expenditures in FYR Macedonia reflect large employment in these sectors. Nearly two-thirds o f all expenditures on defense and public order and security are wages and salaries and, compared to the sample o f countries presented below, only Bulgaria has a greater percentage o f its labor force engaged in these two ~ectors.~While this situation is a legacy o f regional and domestic conflicts o f the last 15 years, regional peace may provide the opportunity now to reallocate public resources to more productive areas. 0 Compared to FYR Macedonia, only Poland and Lithuania spend less on economic affairs and only Latvia and Lithuania spend less on housing and community amenities. 0 FYR Macedonia's spending on health as a share of GDP is higher than most countries in the region (see Chapter 3) and much higher than the average for lower middle income countries. The high spending on health is a legacy o f the past, as almost all former SFRY countries are outliers in this respect, and may explain the fact the health outcomes are good for the level o f development. However, as elaborated in Chapter 3, dissatisfaction in FYR Macedonia with the health system is high and health outcomes lag most EUcountries. 0 FYR Macedonia's public spending on education as a share of GDP is modestly higher than most countries in the region. (see Chapter 2) Despite this level o f spending, student achievements are poorer, hindering prospects for productivity World Bank, 2006 World DevelopmentIndicators. 12 improvements and representing a drag on potential growth. Declining school-age cohorts over the next few decades offers an opportunity to reallocate education expenditures to more efficient uses while raising enrollment rates. With outlays on social security and welfare amounting to about 11.4 percent o f GDP, FYR Macedonia ranks in the lower part o f the distribution among the countries o f the region. Pensions dominate this category, accounting for 8.8 percent o f GDP, broadly in line with regional comparators. Pension reform was advanced with the introduction o f a filly-funded second pillar in January 2006 and ongoing implementation o f structural changes to the first pillar (Chapter 4). Other forms o f social assistance are not excessive, which may explain the highGini coefficient exhibited by FYRMacedonia. Table 1.6: Functional Classification of Government Expenditures in Selected ECA Countries H POL CZ SLK LT LV BG ROM CRO MKD Total Outlays 49.3 45.2 42.2 39.2 35.9 33.3 40.0 32.1 49.6 35.3 GeneralPublic Services 4.2 2.5 2.8 5.2 3.0 3.9 2.9 1.7 2.6 2.3 Defense 1.3 1.2 1.8 1.8 1.2 1.4 2.3 1.4 2.3 2.0 PublicOrder and Safety 2.1 2.0 1.4 2.0 2.3 1.9 2.8 1.8 2.6 3.2 EconomicAffairs 5.3 3.7 6.8 5.1 4.8 3.8 5.0 5.1 6.2 4.4 Environmental 0.7 0.0 0.0 0.7 1.9 0.4 0.0 0.2 0.0 0.1 Protection Housingand 1.3 2.2 3.1 1.1 0.0 0.4 1.5 2.3 2.7 0.7 CommunityAmenities Health 4.9 4.4 6.4 2.3 3.4 4.6 4.7 3.8 6.8 5.6 Recreation, Culture and 1.6 0.8 1.1 1.o 1.3 1.o 0.8 0.6 1.3 0.6 Religion Education 6.3 6.2 4.2 4.3 5.9 5.9 4.3 3.0 4.4 4.8 Social Security and 16.5 19.5 12.9 15.7 10.1 9.9 13.8 9.5 18.3 11.4 Welfare Other Expenditure 5.3 2.6 2.1 0.0 2.0 nla 1.8 2.7 2.4 0.0 - Bulgaria; ROM-Romania; CRO -Croatia; MKD-Macedonia. Legend: H- Hungary; POL - Poland; CZ Czech Republic; SLK - Slovak Republic; LT - Latvia; LV - Lithuania; BG Note:Datafor CRO refersto 2002; ROM and SLK 2003, MKD2005. All others 2004. Source: World Bank ECA Regional Fiscal Dataset for all countries except MKD; MKD from Staff calculations from Ministry of Financedata. F. FISCALSUSTAINABILITY 1.41 Risks to public debt sustainabilityare moderate, reflecting sound fiscal policy in recent years and the prospects of continued fiscal prudence. Public debt amounted to about 34 percent o f GDP at the end o f 2006 following continuous decreases every year since 2000, when government debt jumped to 48 percent o f GDP as a result o f settlement o f frozen foreign currency deposits and restitution claimsS6 The large fiscal deficits in 2001 and 2002 were financed by draw-downs o f government deposits with the banking system and privatization receipts. 1.42 Gross debt issuance has been limited since 2001. Special government bonds with a face value equivalent to 6.5 percent o f GDP have been issued since 2001, part o f which were used to finance the 2001 clean-up o f Stopanska Banka's non-performing loan portfolio during its privatization and the rest to help settle restitution claims. Data for the end o f 2005 show an increase in public debt levels due to the issuance o f the country's first Eurobond in December of 2005. The increase was reversed as the proceeds were used in early 2006 to retire older more expensive debt towards the LondonClub of Creditors. 13 1.43 Interest rates on public debt are low, reflecting both the concessional nature o f some credits and the favorable conditions on the special domestic bonds issued to cover frozen foreign currency deposits, non-performing loans o fthe banking sector and other "legacies o f the past". 1.44 The debt sustainability analysis presented in this report projects that with a primary deficit amounting to about 0.6 percent of GDP a year and real GDP growing by 5 percent a year, public debt will remain on a declining path relative to GDP, decreasing to about 29 percent by 2012 (Table 1.7). Public debt could stabilize under a lower growth scenario with real GDP growth of 4.5 percent a year and primary deficit o f 0.6 percent o f GDP a year. However, the inflexibility o f the budget makes the public debt sustainability vulnerable to output shocks. A strong economic contraction in 2007-2008, in absence o f adjustments on the government expenditure side could pushpublic debt to higher levels. 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Actuals Est. Projected As percentageof GDP Public Sector Debt 48.8 43.0 39.0 36.7 39.8 33.8 31.7 31.0 30.5 30.0 29.6 29.2 ofw: External debt 31.1 27.3 24.5 23.0 26.8 20.7 20.1 19.8 19.4 19.0 18.4 17.6 Domestic debt 0.0 15.7 14.5 13.7 13.0 13.1 11.6 11.2 11.1 11.0 11.2 11.6 Change inpublic sector debt: -1.9 -5.9 -4.0 -2.3 3.1 -6.0 -2.0 -0.8 -0.5 -0.5 -0.4 -0.4 Primary deficit 4.5 4.1 -1.0 -1.3 -1.2 -0.5 0.1 0.3 0.6 0.6 0.6 0.6 Revenue and grants 34.0 34.9 38.4 36.5 35.5 34.2 32.7 31.1 30.7 30.5 30.5 30.5 Primary (noninterest) expenditure 38.5 39.0 37.4 35.2 34.3 33.7 32.7 31.4 31.3 31.1 31.1 31.1 Automatic debt dynamics 0.0 -4.6 -5.0 -2.4 2.8 -4.1 -0.4 -1.2 -1.1 -1.0 -0.9 -0.9 Other debt-creating flows -10.6 -0.6 0.6 0.9 4.0 -0.6 -1.7 0.2 0.0 -0.1 -0.1 -0.1 Residual, including asset changes 4.1 -4.8 1.4 0.5 -2.5 -0.9 0.0 0.0 0.0 0.0 0.0 0.0 Alternative scenarios: Key variables at historical averages 32.3 32.1 31.6 31.0 30.4 29.7 No policy-change scenario (2006) 31.7 29.6 28.1 26.6 25.1 23.6 Stress tests: GDP shock in 2007-08 34.4 37.9 40.1' 42.2 44.5 46.9 Fiscal shock in 2007-08 2 st. dev. 36.3 39.7 38.9 38.1 37.5 36.8 GDP +fiscal shocks in 2007-08 1 st. dev. --2 st.dev.- 35.3 38.1 37.4 36.7 36.1 35.5 30 percent depreciation in 2007 40.6 39.5 38.8 38.1 37.4 36.8 1.45 I n addition, the structure of the public debt indicatesvulnerabilities requiring close monitoring. Around 88 percent o f public debt is denominated in foreign currencies, although the recent sharp reduction in the current account deficit and continued government efforts to increase the share o f local-currency denominated debt should help ease somewhat exchange rate risks. An unanticipated depreciation o f the exchange rate by 30 percent would increase public debt by 9 percent o f GDP. The refinancing risk is moderate, with the average maturity o f government external and domestic debt above 15 and 3 years, respectively. This, together with the strong recovery in government deposits with the banking system has considerably reduced the liquidity risk. G. EFFICIENCYOFFISCALPOLICY 1.46 FYR Macedonia has a relatively lean public sector by regional standards, with mixed measures of the efficiency of government spending. Access to education and health 14 services is nearly universal, as is access to the electricity grid and water supply.' A ranking o f the quality o f public services, however, based on a composite index taking into account developments in education, health, judiciary and economic outcomes, ranks FYR Macedonia in the bottom half among 25 ECA countries (Table 1.8). 1.47 Educationoutcomes are low and overall perception of the education sector is poor. Primary education is compulsory and the network o f schools enables easy access even in distant rural communities. (The government agreed in April 2007 that secondary education will be compulsory from 2009.) But expenditures per student vary substantially across municipalities and the level o f skills and knowledge o f FYRMacedonian students lag those o f students in nearly all countries in the region. Moreover, FYR Macedonian students at the primary level receive significantly fewer hours o f instruction per year than the average student in the EU15 and N M S , an outcome reflecting the prevalence o f double-shift schools and curricula that needs to be revised. While primary enrollment i s in line with levels in the EU, secondary enrollment is well below the EUand N M S averages. 1.48 Health service coverage is also satisfactory but popular dissatisfaction with the quality of healthcareis high. Coverage with health insurance is almost universal and 90 percent o f the population is able to contact a health provider in less than 30 minutes. However, the recently produced Medical Map indicates wide regional discrepancies and inefficiencies in service provision. Health outcomes are better than countries with similar levels o f income but lag almost all N M S and the EU15. 1.49 Social policy outcomes are mixed. Pensions provide substantial support to the elderly and are an effective buffer against poverty. A recent Poverty Assessment for FYR Macedonia suggests that absolute poverty levels would double in the absence o f pensions. Recent parametric changes to the pay-as-you-go pension pillar and the launching o f a funded second pillar have improved the financial situation o f the pension system. However, fiscal sustainability o f the pension system will require full implementation o f parametric reforms currently in train, includingthe gradual increase in the retirement age and determined implementation o f the revised pension indexation formula. The projected average pension is likely to fall to 45-55 percent of average wages, well below the Government's target o f 70-80 percent. Also, the fees levied on second pillar accounts are high by international standards. The Government should consider decreasing these fees to boost net returns and the available stream o f income for future pensioners. 1.50 The Government has been less successful in performing its redistributionfunction. A Gini coefficient o f about 0.37 places the country among the more unequal countries in the region. Significant discrepancies also exist in access to public services; for example, children in families in the lower consumption quintiles have significantly lower enrollment rates in secondary and tertiary education and poor individuals are much more likely to suffer from chronic illness. According to the 2005 Household Budget Survey (HBS), 99 percent of the households have access to the electricity grid (up 13 percent since 1995) and 96.3 percent have access to the water system (up 24 percentagepoints since 1995). Sewage disposal systems are installed in 86.3 percent of households, while 88.8 percenthavephone connection. 15 Table 1.8: Measures of Efficiency in Government Outcomes (Relativeranks, lower is better) Education Health Enforcing Economic Public sector efficiency efficiency Contracts outcomes efficiency ranking 11 ranking 21 ranking 3/ ranking 4/ (weighted) Lithuania 7 7 1 9 1 Hungary 3 8 3 1 2 Slovenia 4 1 23 2 3 Czech 5 2 19 3 4 Estonia 2 3 5 7 5 Latvia 6 11 2 11 6 Poland 1 6 25 6 7 Croatia 10 4 8 5 8 Slovakia 15 5 20 4 9 Belarus 9 10 12 13 10 Armenia 13 15 4 18 11 Bulgaria 11 12 17 12 12 Georgia 8 9 10 24 13 Kazakhstan 12 16 7 16 14 Ukraine 21 18 6 10 15 Serbia 14 13 22 15 16 Azerbaijan 23 22 11 8 17 Romania 19 17 16 14 18 FYR Macedonia 20 14 21 17 19 Mongolia 16 20 15 21 20 Albania 18 19 24 19 21 Uzbekistan 17 23 9 25 22 Kyrgyz 24 24 13 20 23 Tajikistan 22 25 14 23 24 Moldova 25 21 18 22 25 1/ Measures public perceptions of education services (satisfaction with quality and perceptions of corruption) and education outcomes (enrollment rate in primary and secondary schools and primary completion rate). 2/ Measures publicperceptionsof health services (satisfaction with quality and perceptions of corruption) and achievements on health outcomes (infant mortality rates and life-expectancy). 3/ The indicator "enforcing contracts" from the Doing Businesssurvey. 4/ A summary indicator of GNI per capita at PPP, the Gini index and poverty levels at USD2 aday. Sources: World Bank staffcalculations based on EBRD-WorldBank Living inTransitionSurvey (LITS); World Bank World DevelopmentIndicators; World Bank DoingBusiness 2007 andWorld Bank "Growth, Poverty and Inequalityin EasternEuropeand Former Soviet Union." H. CONCLUSIONS 1.51 Reducing spending by 2 percent of GDP over the next several years while tacking fiscal pressures will require a combination of cuts in some categories of existing spending, spending restraint in others and stronger real GDP growth. Spending pressures include costs related to advancing EUintegration and on reducing the substantial backlog o f road maintenance. This report suggests that measurable spending cuts are not likely to be found in health and education where the challenge will be to improve the efficiency of spending and reallocate 16 outlays to significantly improve outcomes and support growth. The demand on the budget from furthering pension reform will increase in the following years, but should ease subsequently. Together with sustained progress in advancing structural reforms and revamping government spending, increasing the formalization of economic activity is the key to stronger growth in output. The authorities are urged to move to reduce social security contributionrates, starting with unifying and then abolishingminimum contributionthresholds. Reductions in contributionrates shouldbe carried out in the context of reformingthe financingof social transfers, mainly health and further on pensions, toward general revenues. To ensure that the resultingreductionin revenuesdoes not endangerthe authorities'targets for the fiscal deficit, a cut in contribution rates should be offset by more ambitious spending reductions, improved efficiency of spending or, in cases of larger reductions in contributionrates, by increases in the VAT and PITrates. 17 2. EDUCATION A. INTRODUCTION 2.1 Government spending on education as a share of officially-estimated GDP has been on a declining trend, but is still modestly higher than most countries in the immediate region. Relative to GDP adjusted for the size o f the NOE, public spending on education is not out o f line with comparator countries. Student achievements are poorer relative to the countries in the immediate region and substantially more so relative to both the EU15 and the new member states of the EU(NMS). In addition, enrollment rates in secondary education are lower than in the EU15 and the NMS, and skills acquired in schools have tended to be at odds with market needs. Access to education seems equitable at the primary level, but gaps open up at the secondary and tertiary levels, with the families' level o f wealth, place o f residence and ethnicity all affecting enrollment rates. Moreover, substantial differences appear inthe amount o f spending on schools with different ethnic compositions. 2.2 Poor learning outcomes are one of the factors that have contributed to sluggish economic growth and high unemployment in FYR Macedonia. The authorities understand well that boosting the productive capacity o f the economy should be accompanied by raising the quality o f education outcomes and making education more relevant for market needs. The authorities are well advised to redouble their already strong efforts to address these issues urgently. 2.3 The report concludes that the education sector at this time has adequate overall resources, but they are used unproductively, resulting in poor learning outcomes. Higher publicspending on education is not the answer to the current situation, however. The focus needs to be on resolving the outstanding equity and efficiency issues.' This chapter discusses these issues and makes recommendations to the authorities on how to tackle them. Institutional reforms already in progress, including decentralizationand changes in the way schools are financed, should help move toward a more equitable distribution of resources and create stronger incentives to improve the efficiency and outcomes of spending. Increasing the amount o f instructional time that students receive in schools should be considered, and public resources should be used to increase poorer households' demand for education by reducing the cost burden o f education. Declining school-age cohorts over the next decades offers a fortuitous opportunity to reallocate expenditures to more efficient uses while raising enrollment rates. 2.4 The rest of the chapter is organized as follows. Section B reviews the level and structure o f spending on education. Section C analyzes the efficiency o f using available resources and the education outcomes. Section D presents the chapter's recommendations and section E the conclusion. This chapter does not discuss the issues o f research, science or tertiary education. These domains absorb a relatively small proportion o f public spending at present and are beingaddressed in a separate study prepared by the World Bank. *Extra transitional resources may be needed in line with a plan to move to a more effective system, but these shouldalso be allocatedwithin the available overall fiscal envelope. B. INSTITUTIONALFRAMEWORKEDUCATION AND EXPENDITURES 2.5 The institutional setup of education is evolvingafter the start of reforms that have decentralizedfunctions and financing for primary and secondary education. In Phase 1 o f decentralization reforms that began in mid-2005, maintenance functions (such as the provision o f heating, energy, materials and services) and the transport o f students to school were decentralized to municipalities. Teacher salaries, however, are still paid by the central g~vernment.~ Municipalities became responsible for all primary and secondary schools except the schools for children with special needs and public tertiary institutions. Where clear cadastre documents exist, school facilities and their assets have been transferred to municipal governments, although the central government retains the titles to the lands on which these facilities were built. In 2005, the transfers to municipalities to cover the costs o f their responsibilities were based on historical costs. As o f 2006, the allocations are based on a formula that includes a fixed basic amount for each municipality and a variable amount based on enrollment. Such a formula should help reduce historical inequities, but data are still not available to evaluate the outcomes. The effect o f the allocations is likely to be relatively modest, however, since municipalities currently receive three separate grants - two for primary and secondary school operations and maintenance and one for primary school transportation - and these funds cannot be co-mingled. Phase Io f the decentralization is to be followed by a second phase, and the options facing the government are discussed below; by law, July 2007 is the earliest date Phase I1can start. 2.6 Government spending on education declinedfrom 4.5 percent of GDP in 2002 to 4 percent in 2005, while private outlays remained little changed (Table 2.1). Government education spending relative to GDP adjusted for the NOE i s broadly similar to the level in most countries in southeast Europe, but is lower than the average in the N M S (Relative to officially- estimated GDP, government spending is modestly higher than in the countries o f southeast Europe.) From a historical perspective, outlays on the order o f 4 percent o f GDP are similar to the amounts spent by Japan in the 1970s, as well as the Philippines, South Korea and Thailand in the 1990s, indicating that overall government expenditures in FYR Macedonia are not low. Private outlays on education in FYR Macedonia as a percent o f GDP are higher by about 50 percent than the average for the EU15 and by about 15 percent than the average for the N M S (reference year 2002). Given the experience o f countries outside Europe, private spending, especially at the tertiary level, has been helpful in improving education outcomes and should be encouraged. 2.7 Salaries for primary and secondary teachers seem high relative to officially estimated per capita GDP compared with the average teacher salary in the EU15 (Table 2.2). Relative to GDP adjusted for the NOE, however, the differences are smaller. After 15 years of experience, the salary per hour o f net teaching time at the primary level (in PPP) is slightly higher, while at the secondary level slightly lower than the average for the new EU members. The Ministry of Labor and Social Policy used a similar decentralization for the financing of preschools. Although most expenditure is still financed by the central government, maintenance costs are financed through an earmarked grant from the central government to the municipalities. The latter, in turn, transfer finds to the preschools. The above earmarkedgrants, together with the transfers for wages and remaining current outlays, will be convertedinto block grants starting during Phase I1ofthe decentralization. 20 Table 2.1: Macedonia: EducationExpenditures,2002-2005 (Inmillionsofdenarsunless indicatedotherwise) Public Expenditureson Education 11,029 10,941 11,186 11,477 InpercentofGDP 4.5 4.4 4.2 4.0 5.215.6 4.715.0 Inpercentoftotal public expenditures 11.2 11.3 11.7 11.5 11.7 9.4 PrivateExpenditureson Education2/ 3,111 3,126 3,431 3,962 Inpercentof GDP21 1.3 1.2 1.3 1.4 0.8 1.1 Total EducationExpenditures(Public andprivate)21 14,140 14,066 14,617 15,440 Inpercentof GDP21 5.8 5.6 5.5 5.4 6.016.4 5.816.1 I' For each pair of estimates separated by a slash, the first number refers to expenditures on educational institutions and ''theEstimatesof second on all education expenditures both within and outsideeducationalinstitutions. private expenditures include: a) the government's cost recovery from households for all levels of public education; b) estimated private costs of preschool; c) private expenditures on textbooks; d) tuition fees; and e) private lessons and school supplies. Costs of student transport are not included. Sources: Ministry of Finance; State Statistical Office and authors' estimates. Estimates for the EU15 and the NMS are from OECD's Education at a Glance, 2005, tables B2.la, B4.1, and B6.1. Table 2.2: Comparisonsof Teacher Salaries,Macedonia,the EU15, and the NMS Macedonia (2005) * EU15 (2003) NMS(2003) Lower and Upper Primary Education Ratio ofsalaryafter 15years of experience 2.00 1.29 0.86 to GDP per capita Salaryperhour ofnetteachingtime after 15years ofexperienceinUS dollars InUS dollars, at PPP $20.7 $54.3 $19.0 Upper Secondary Education Ratioof salary after 15years of experience 2.37 1.46 0.95 to GDP per capita Salaryper hour ofnetteachingtime after 15 years of experienceinU S dollars InUSdollars, at PPP $19.8 $69.8 $23.8 For Macedonia, the wage data are for 2005; the purchasingpower parity (PPP), for 2004. Source: Macedonia:Ministry of Finance, Ministry of Education and Science, and World Bank World Development Indicators 2006. Comparators:table D.3.1, OECD, 2005. 21 2.8 The wage bill is higher as a proportion of spending than in other countries. The further increases inthe share o f the wage bill at the primary level at a time when student numbers are decreasing is problematic (Table 2.3, Table 2.4 and Table 2.5). After wages, the largest share o f outlays is dedicated to utilities, transport and contractual services rather than on inputs which improve the quality o f education like books and materials. 2.9 Given these constraints, it is a cause for concern that the government has decided to spend 25m on the purchase of computers without any strategic plans for how these would actually be used in the classroom by students and teachers. This amount o f money exceeds what the country currently spends on all non-salary recurrent spending in primary and secondary education (about 18m). Moreover, the cost of purchasing the computers would represent only about 30 percent o f the ultimate cost that will need to include outlays for maintenance, upgrades, training and security. Much o f these latter burdens will fall on the municipalities, but the government has not prepared a financing plan to meet these costs. Table 2.3: Economic Classification of Education Spending, Macedonia and Selected Comparators (In percent of total government spendingon education) Basic and Secondary Education Countries YOrecurrent YOcapital Wages as YOof Non-wagerecurrent as recurrent % of total recurrent Macedonia(2005) 95.7 4.1 86.6 13.5 EU15 (2002) 92.7 7.3 81.5 18.5 NMS(2002) 92.5 7.6 72.9 27.1 Tertiary Education Countries YOrecurrent YOcapital Wages as % of Non-wagerecurrent as recurrent % oftotal recurrent Macedonia(2005) 93.O 5.9 51.6 48.4 EU15(2002) 88.6 11.4 68.2 31.8 NMS(2002) 89.0 11.0 56.9 43.1 Sources: Ministry of Finance; Table B6.3, OECD 2005. Table 2.4: Distribution of Goods and Services (non-staff recurrent) Expenditures by Level of Education (2005) Level of Utilities/ Contractual Maintenance TraveWer Other Total Education Transport Services and Supplies Diems Total 8.9 8.8 1.1 0.7 2.7 22.3 Preschool 26.7 2.8 2.1 0.2 0.9 32.8 Primary 6.2 2.6 0.5 0.0 0.4 9.8 Secondary 6.7 3.1 1.8 0.3 7.4 19.3 Tertiary 9.4 28.0 1.3 2.7 3.6 45.0 Source: Ministry of Finance. 22 Table 2.5: Expenditure Shares by Level of Education for Macedonia, the EU15, and the NMS Countries Preschool Primary Secondary (General +VET) Tertiary Macedonia (average 2002- 9.3 47.9 21.6 21.3 05) EU15 (2002) 7.8 42.0 25.2 22.7 NMS(2002) 11.4 40.5 23.2 22.0 Source: Ministry o f Finance; Table B.2.1.c, OECD 2005 2.10 The share of spending dedicated to the secondary level is in line with comparator countries,and unit costs in secondary education have fallen relativeto unit costs in primary education(Table 2.6). Moreover, average expenditure per student inthe gymnasium schools and the `mixed' schools (the schools offering both general and vocational, or VET, education) is similar, while expenditures per student at the VET schools average 90 percent o f the outlays in the gymnasium schools." The variation in funding across schools is the largest among the `mixed' schools, varying with the degree o f integration o f a vocational program into the general education, and the lowest for the gymnasium schools. Table 2.6: Total Education Expenditure per Capita Relative to per Capita Outlays for Primary Education, Macedonia, the EU15, and the NMS Macedonia(average 2002-05) * Countries Preschool Secondary (general+VET) Tertiary 1.79 1.14 2.47 EU15 (2002) 0.65 1.17 1.73 NMS(2002) 1.10 1.25 2.41 *The average for preschool is 2002-04. Sources: Ministry of Financeand MES; table B1.1, OECD 2005. 2.11 Expendituresper student vary substantially across municipalities,with a largepart of the variation reflecting low spending per student in Albanian and dominant-Albanian schools. The average expenditure per student in Albanian schools is about 75 percent o f the average expenditure in FYR Macedonian schools (Table 2.7); average expenditures are lower in Albanian schools regardless o f whether the schools are rural or urban. Less than a third o f this difference is explained by different student-teacher ratios (STRs). For example, primary schools in FYR Macedonian (Macedonian-dominant) municipalities have average STRs o f 17.6 (16.9, with the figure for Albanian (Albanian-dominated) municipalities about 18.8 (17.5). The bulk o f the difference appears to reflect the continued practice to allocate resources to schools on the basis o f number o f classes rather than the number of students. Class sizes, moreover, can vary substantially across the municipalities, resulting in unequal outcomes. More generally, the highest-expenditure municipality spends 4.3 times more per primary student than the lowest- expenditure municipality. Outcomes are similar at the secondary level. 10There are 82 schools offering secondary education in the country. Of these, 14 are gymnasium schools, 40 VET and 28 are mixed. 23 Table 2.7: Average Expenditure per Primary Student, 2005 (Indenars peryear) School Ethnicity Type Municipality Type Large city Rural Skopje Small city Total Albanian 18,442 20,408 18,02 1 18,593 19,381 dominantAlbanian 21,044 24,116 16,825 25,945 21,294 dominantMacedonian 20,867 32,998 25,401 27,293 26,287 Macedonian 25,727 32,279 20,464 24,486 25,346 Other 26,467 32,095 20,903 27,886 Total 23,456 27,606 20,133 24,840 24,150 Source: Own calculations based on the MES database. 2.12 The disparities in spending across the municipalities are a corollary of budgetary allocations that have not kept pacewith the reductionin student populationand shifts in the location of students. Additional important factors include an inequitable distribution across ethnic groups and social status. As proposed in the recommendations, the authorities are well advised to begin urgently to tackle these disparities, including through the introduction o f capitation financing. c. EDUCATION OUTCOMES AND EXPENDITURE EFFICIENCY 2.13 Learning outcomes are poor and need to be urgently improved. Except for almost no investments in school maintenance, physical capital is efficiently used, measured in terms o f school occupancy and student-class ratios. However, the system produces exceptionally low learning outcomes, with students performing substantially worse than in other countries at similar income levels or countries that spend a similar share on education relativeto GDP. Education Outcomes 2.14 The levels of skills and knowledge of FYR Macedonian students lag those of students in other countries. Results from international assessments and a FYR Macedonian national assessment o f fourth grade achievements in language and mathematics tell a consistent story o f poor learning outcomes. For example, in the Programme for International Student Assessment (PISA), an assessment administered to 15 year olds, 87 percent o f FYR Macedonian students score below level 3 (1 being the lowest and 5 the highest); only individuals who attain a score equal or higher than 3 are considered able to function adequately in a modern workplace (Table 2.8). Of the countries in the region, only Albania performed worse, with 91 percent o f the students below level 3. In the EU15, by comparison, only 40 percent o f the students score as poorly.'' 2.15 Although learning outcomes are broadly correlated with the level of economic development, FYR Macedonia's learning outcomes are worse than among countries with similar levels of income per capita (Figure 2.1). Improving the efficiency o f education spending, as discussed in this report, should help lift education outcomes to or above levels in I'Macedonia's performanceis similar in other international assessments. For example, on PIRLS, which assesses children in the fourth year of formal schooling on a range of reading comprehension strategies, fewer children reachedthe lowest acceptable benchmark in Macedonia (55 percent) than all neighboring and EUcountriesthat participated (Slovenia with 83 percentis the next lowest). 24 countries with similar incomes in the short term, and contribute to boosting real GDP growth over the longer term. 2.16 The poor outcomes seem to result from insufficientinstructional time, low level of spending on key quality inputs, and the 'tracking' system under which students at the end of primary education need to decide whether to attend general or vocational secondary schools. (There are limited opportunities to move from four-year vocational schools to gymnasium schools. The move and the opportunity to attend university is impossible for three- year vocational schools.) These problems have existed for some time and will take concerted effort. In secondary education, the different vocational and general education tracks reinforce rather than ameliorate differences in performance between students after leaving primary education. Table 2.8: Percent Scoring at Different Competency Levels on PISA Reading Scale: Macedonia, EU 15, and Neighboring Countries Country Percent scoring Percent scoring at levels3-5 s: level 3 Total YOlevels 3-5 YOlevel 3 YOlevels 4-5 Albania 90.9 9.1 7.7 1.4 FYR Macedonia 87.0 13.0 11.1 1.9 Bulgaria 67.3 32.7 21.5 11.2 Latvia 56.9 43.1 25.2 17.9 RussianFederation 56.7 43.3 26.9 16.4 Poland 47.4 52.6 28.2 24.4 Hungary 47.7 52.3 28.8 23.5 Czech Republic 42.3 57.7 30.9 26.8 EU15 40.1 59.9 28.8 31.2 Sources: OECD, 2001 andown calculations. Figure 2.1: Student Performance and GDP per Capita, Selected Economies, PISA 2000, Reading 350 - OAlbanta Pen10 300 I I I I I I I I I Source: Figure3.9, p.56.ExpandingOpportunitiesand Building Competenciesfor Young People:A New Agenda for Secondary Education, World Bank, 2005. 2.17 FYR Macedonian students at the primary level receive significantly fewer compulsoryinstructional hoursper year than the averagestudent in the EU15and the NMS (Table 2.9). These differences are so large that it is highly likely that FYR Macedonia's low instructional time partly explains its students' poor performances in the international learning 25 assessments, all of which measure learning outcomes during or at the conclusion of primary education.'* At the primary level, the average EU15 (NMS) student is entitledto 69 percent (37 percent) more instructionaltime than the average FYR Macedonian student. At the secondary level, by contrast, instructional time is broadly similar in FYR Macedonia and comparator countries. 2.18 The low instructionaltime reflects a combinationof double-shiftschools, mandated few classes and short lesson lengths, and the thinking of local psychologiststhat helps limit instructional hours. About 40 percent of primary schools are double-shifts, with the length of each shift equal to five hours, substantially less thanthe length of a shift in single-shiftschools in more advancedcountries. Further, severalyears ago the government reducedthe lengthof lesson times to 40 minutes from 45 minutes, resultingin a further decrease in instructionaltime. There is, moreover, a body of opinionamong local psychologiststhat children in the 1-4 grades should have few instructionhours. All of these factors, combined with a rigid system of regimenting teachers' time into teaching and non-teaching hours, has resulted in low instructionaltimes for primary students in FYR Macedonia. There is clearly room to extend the amount of lessontime for primarystudents without the needto buildnew schools.Moves inthis directionwould require changes to the number of hours per day full-time teachers spend teaching without increasingthe wage bill. Table 2.9: Annual Compulsory Instructional Hours by Age, Macedonia, EU15, and the NMS Total Instructional Comparator 7-8 Years 9-11 Years 12-14 Years Hours Grades 1- Age 15 8*** Macedonia A 432 488 563 783/853**'* Macedonia B 444 520 640 4016 837/907**** EU15 781 836 902 6776 830 N M S 583 680 770 5516 831 *Annual compulsory instructionaltime is definedas the number of minutesper class hour times the numberof compulsoryclass hoursper week times the number of weeks of instructionper year, dividedby 60. .*MacedoniaA (B) assumesthat those 7-8 years ofage are in grades 1-2(2-3). .**For the EU15 and the NMS, the ages of 7-14 are usedfor estimatingtotal instructionaltime for the first 8 grades of formal education. ****High school students are expectedto spend 70 hoursper year (60 hours for grade 12) in projectactivitiesunderthe guidanceof ateacher. Although compulsory, these are not instructionalhours. The first figure in the pair excludesthe projectactivities; the secondfigure includesthem. Source: MES for Macedonia, Table D1.l (OECD, 2005) for the EU15 and the NMS. 2.19 In addition to the low compulsory instructional hours at the primary level, FYR Macedonian students complete fewer years on average than their counterparts in the EU. A hypothetical five-year old is expected to complete a total of 13.4 years in school in FYR Macedonia, 16.6years inthe NMSand 17.6 years inthe EU15. l2 The amount o f compulsory instructional time in school has been established as a key driver o f student performance, even though the relationship between instructional time and student learning outcomes is not straightforward. Factors such as the effectiveness with which the learning time is invested or instruction is delivered, the organization o f instruction, and the extent and type o f learning opportunities prior to schooling affect how instructional time translates into learning outcomes. 26 2.20 Primary enrollment is in line with the levelsin the EU,but secondary enrollment is well below the average in the EU average, including in the NMS (Table 2.10).13 Lower secondary enrollments result in a lower educational attainment in the working population compared with countries in the region. FYR Macedonia has significantly fewer people with upper secondary and tertiary qualifications and far more with only basic education (Table 2.1 1). Based on trends over the last ten years, it would take 33 years (24 years) for the share o f workers with secondary (tertiary) education in FYR Macedonia's labor force to match the shares in the EU. Table 2.10: Enrollment Rates for Macedonia,the EU15, and the NMS Country Enrollment Enrollment Enrollment Expected years of rates for 5-14 rates for 15- rates for 20-29 education for year olds 19 year olds year olds hypothetical 5 year old Macedonia 96.8 70.9 21.6 13.4 EU15 (2003) 99.6 82.4 24.5 17.6 *NMS (2003) 97.9 91.3 20.3 16.6 Inthis table only, this refers to the Czech Republic, Slovakia, Polandand Hungary. Source: Macedonia:State StatisticsOffice; OECD, Education at a Glance, 2005, tables C1.1 and C1.2. Table 2.11: Educational Attainment of 25-64 Year Old Population for Macedonia, the EU15 and the NMS (Inpercent) Education level Macedonia (2005) EU15 (2003) NMS (2003) >Basic 10.2 20.0 1.5 Basic 30.9 17.5 16.3 Upper secondary 45.5 38.7 68.0 Post-secondary, non-tertiary 4.1 6.3 2.0 Tertiary 9.4 23.2 13.0 Total upper secondary + 59.0 68.1 83.0 tertiary Average years o f education 10.3* 11.8 12.0 *The LaborForceSurvey reportsthe number of individualsby five-year age groups, Le.: with no education, incomplete primary education, VET secondary, gymnasiumsecondary, post-secondarystudies, and university. Incompleteprimary educationwas assumedto last 5 years. Post-secondarytertiary education is assumedto last two years, but this may be an over-estimate. University education is assumed to last four years, but sometertiary programsrequiremore time. Source; Macedonia: State Statistics Office, Labor Force Survey, 2005; EU15 and recent EU entrants: OECD, Education at a Glance,2005, tablesAI.1aand Al.4. 2.21 Access to education appears to be equitable at the primary level (ages 6-14) for all segments of the population(except certain ethnic minorities), but sizable disparitiesappear at the secondary and tertiary level (Table 2.12).14 The gap in enrollment rates between the poorest and wealthiest quintiles amounts to about 18-20 percentage points both at the secondary and tertiary levels. There is a similarly large gap with respect to whether students live in urban or rural areas (Table 2.13). l3 Primary (secondary) enrollment rates in 2004 are 97(74) percent in Albania and 93 (73) percent in BH (World Development Indicators). l4 The enrollment rates for the 6-10 year olds are lower than for the 11-14 year olds. This anomalous difference may reflect methodological problems or the fact that some children do not start first grade until the age o f seven. 27 Table 2.12: Enrollment Rates by Age and Household Consumption Quintile Consumption Quintile Gap between Age Poorest 2 3 4 5 *'I Q.l Q5Q. and A. 3-5 years 0.7 0.0 1.o 1.3 2.0 1.1 -1.3 B. 6-10 years 86.1 88.8 81.4 90.6 87.3 86.9 -1.2 C. 11-14years 97.0 99.0 98.7 98.5 99.6 98.7 -2.6 D. 15-18years 71.8 77.8 78.3 79.7 90.1 80.3 -18.3 E. 19-26years 19.0 22.7 26.5 36.6 39.3 29.9 -20.3 *Data for the 3-5 year old age group are not representativebecausethe coefficient of variation is higher than 50 percent. The group denoted with B corresponds to lower primary; with C to upper primary; with Dto secondary; and with Eto tertiary education. Source: MacedoniaState Statistics Office, 2005 Labor Force Survey. Table 2.13: Enrollment Rates by Age, Urban/Rural Residence,and Household Consumption Quintile Age group Consumption Quintile All Poorest 2 3 4 5 6-lo years 87.0 84.9 88.3 89.2 85.5 77.6 93.0 87.7 88.5 85.8 88.6 84.9 11-14years 94.9 100.0 99.2 98.8 98.6 98.8 99.4 97.1 100.0 99.2 98.6 98.8 15-18 years 72.4 71.1 77.2 78.6 88.1 69.5 89.9 66.1 94.6 84.4 85.6 74.2 19-26years 24.4 10.6 31.1 12.5 32.9 18.8 48.9 19.8 47.6 25.3 38.5 18.0 *Data for the 3-5 year olds is not representativeand not included in this Table. Source: Macedonia State Statistics Office, 2005 Labor Force Survey. 2.22 Enrollment rates for girls exceed those for boys in the early years, only to reverse by the latter years of secondary education (Table 2.14).15 This is reversed yet again at the tertiary level, however, where girls do substantially better in all but one quintile. Table 2.14: Enrollment Rates by Age, Gender, and Household Consumption Quintile Consumption quintile Age group Poorest 2 3 4 5 All M F M F M F M F M F M F 6-lo years 88.5 83.0 84.5 93.2 82.4 80.4 88.2 92.9 81.4 93.7 84.7 89.2 11-14years 97.4 96.6 100.0 98.2 98.3 99.2 99.3 97.6 99.3 100.0 98.9 98.4 15-18years 75.3 67.9 82.8 72.9 83.1 72.5 84.3 75.8 93.5 86.1 84.6 75.8 19-26years 14.9 23.6 21.0 24.4 28.2 24.8 35.7 37.5 33.6 45.4 27.8 32.1 *Data for the 3-5 year olds is not representative and not included inthis table. Key: M= male; F = female. Source: MacedoniaState Statistics Office, 2005 Labor Force Survey. 2.23 The level of enrollments in private institutions is still quite low relative to other countries, although private spending as a share of overall education outlays is high relative to comparators (Table 2.15). Anecdotal evidence suggests that private spending i s largely used by the richer segments of the population at the secondary and tertiary levels. Based on the experience of other countries, enrollment rates at the tertiary level can increase substantially l5The results from the 2005 LFSare similar. 28 supported by private spending. Given the limited fiscal envelope, this will be a welcome development. Table 2.15: Percent of Total Enrollments in Private Schools at Different Levels of Education for Macedonia, the EU15, and the NMS Comparator Primary Lower Upper Tertiary Tertiary secondary secondary type B* type A* Macedonia(2005) 0.0 0.0 1.1 8.1 8.1 EU15 (2003) 15.6 18.8 23.4 23.3 11.4 N M S (2003) 3.0 3.8 10.8 23.6 11.8 *Tertiary type A programs are university programs; tertiary type B, other types of post-secondary programs, such as vocational-technicalprograms. Sources: Macedonia:State Statistics Office; EU15 and recent EUentrants: OECD, Education ut a Glance, 2005, Tables D5.1and D5.2. 2.24 In terms of access to education by ethnicity, the results are clearer. The enrollment share of children of Macedonian ethnicity exceeds their populationshare at all education levels, especially after the primary level (Figure 2.2). Roma children are underrepresented at all levels of education. Although the Albanian and Turkish groups are slightly over-represented at the preschoollevel, bothgroups are under-representedstarting at the upper primarylevel. Figure 2.2: Enrollment Share Relative to Population Share by Ethnic Group and Level of Education (In percent, 2002) D 2 D.15 n.1 dE5 4.1 Source: MacedonianState Statistics Office, 2002 Census. 29 Efficiency of ResourceUse 2.25 Scarcity of capital resources has resulted in an intensive use of school facilities. At the primary level, 42 percent o f the schools use double shifts and one percent o f the schools use triple shifts.16 At the secondary level, 87 percent o f the schools are double shift, with little variation between the types o f schools. For secondary education, all but 9 of the schools in the country have 500 or more students, a reasonable use o f school buildings. 2.26 It is not recommended that the country use scarce resources to eliminate quickly double-shift schools, as other investments would have a higher impact on improving the quality of education, as discussed below. While doing so is desirable in the longer run - especially in helping increase low instructional time - the costs o f moving to a single shift system are very high. It is recommended that the small number o f triple-shift schools be converted at least into double-shift schools: triple-shift schools cannot provide sufficient instructional time for each shift. 2.27 Student-teacher ratios at both the primary and secondary levels are higher in FYR Macedonia than among comparators, and are similar at the tertiary level(Table 2.16). It is unlikely that the higher burden on teachers is a substantial part o f the explanation for poor student performance, however. South Korea, for example, is one o f the top performers in PISA and has an average primary-school class size o f 37. Table 2.16: Student-Teacher Ratios, Macedonia, the EU15 and the NMS Comparators Preschool Primary Secondary Tertiary Macedonia(2005) * 9.0 16.0 16.0 16.9** EUl5 (2003) 14.5 13.5 11.8 16.1 ..NMS (2003) 12.2 14 13.3 15.3 *Preschoolratiosare for 2004. This ratioassumesthat each part-timestudentis a0.5 fulltime equivalent student. Source: Macedonia: State Statistics Office; the EU15 andthe NMS:Table D2.2,OECD, 2005. 2.28 The student-teacher ratios at the preschool and secondary levels have remained stable over a numberof years, but the system has been adding teachers at the primary level even as the number of students has been declining (Figure 2.3). Ifthe primary level had the same STR in 2005 as in 1992, the number o f teachers would have been lower by about one-fifth, or almost 2,800 teachers. With the size o f the cohort in secondary education expected to fall by 15 percent (20,000 students) over the next 25 years, and a similar decline expected at the primary level, an important opportunity will be presented to make efficiency gains by reducing the number o f teachers and schools, and redirecting resources to improve the quality o f education and increasing enrollments in secondary education. This i s likely to be somewhat offset by the increase in enrollments that should follow the introduction o f compulsory secondary education from 2009. l6Source: MoES. 30 Figure2.3: Student-TeacherRatiosat Different Levels of Education(1992-2005) 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Source: Ministry of Education and Science. 2.29 Student-teacher ratios vary substantiallyacross municipalities, rangingfrom a low of an average of 8 students per full-time equivalent teacher to a high of 26 students. Secondary schools show a similar variation, with a range o f 3-23 students per teacher. STRs vary little across municipality types or types of schools, being only slightly higher in gymnasium schools (18.5) than in VET schools (16.0). The wide variations across municipalities, but not across types o f municipalities or schools, probably reflects the existence o f inefficient schools, where the number o f teachers has not been adjusted even as the number o f students has fallen more sharply than for the country on average. The ratio o f teachers to non-teaching staff is reasonable: about 3 teachers to one non-teaching staff at the primary level and 4 teachers to one non-teaching staff at the secondary level. D. RECOMMENDATIONS 2.30 The analysis demonstrates that the education system needs urgent improvement, a conclusion with which the authorities fully concur. It is recommended that the authorities seize the opportunity o f declining number o f students to free resources, including for more productive use within the sector. All in all, improving the quality o f education outcomes will require determined efforts over a longer period o f time; the time for action needs to start now. It may be that some transitional funding is needed to enable resources to be freed up and moved to a more effective system. Such transitional funding needs to be allocated strictly in line with a plan to accomplish such a move, however, within available fiscal resources. ImprovingEquity and Efficiencyof Spending 2.3 1 The most substantialprogress toward a more equitable distribution of resources is likely to result from the introduction of a formula-based allocation of staff wages (school directors, teachers and pedagogicalstaff). This is planned for Phase I1o f the decentralization, 31 though no decision has yet been made on the timing o f introducing such a formula. According to the Law, allocations must be based on a formula linked primarily to the number o f students in each municipality, though it would be possible (and expected) to include some modifications to reflect the higher cost o f education in ruraljurisdictions. Inthis way, municipalities with similar characteristics would receive the same level o f funding, helping address a key inequity in the current system. Given the current inequalities in funding, any formula will result in some shifts in fundingacross municipalities-the existing allocations cannot simply be replicated. Financing o f wages would be done through block grants earmarked for education. 2.32 However, the benefits of the allocation formula will only accrue if there are commensurate changes in the degree of control that municipalities have over how to spend the resources allocated to them and over the factors that affect the level of spending. The key issues to consider duringPhase I1o f the decentralization include: The abiliX to switch funds between wage and non-wage recurrent budget categories. The opportunity to switch resources between uses is a key tenet o f decentralization, as it provides local governments discretion over how to meet their responsibilities in the education sector. Inthe short term, it should provide an incentive to close small schools and reallocate the funds saved. However, there is risk that savings may be used to increase the number o f teachers and other staff rather than increase spending on other inputs vital for improving quality. The authorities are invited to consider setting up national standards to regulate these issues (see below). Employmentof staff, especially teachers. The decision about how many staff to hire is essential in determining the education budgets for municipalities.l8 If funding for teachers' salaries i s allocated to municipalities on a capitation basis, municipalities must have sufficient control over their wage bills to respond accordingly. The transfer o f this responsibility to the municipalities, in collaboration with their schools, is a key feature o f Phase 11. The setting and verification of minimum service standards for municipalities. The central government will need to promulgate a wide range o f regulations setting the framework for municipalities in areas including health and safety, organization o f schools and finance. In many areas, there are existing regulations but these are out o f date. Decisions will need to be made about what new standards are appropriate, who is responsible for upgrading facilities that currently do not meet these standards, and the sanctions for failure to observe standards. The distribution of responsibilities between municipalities and schools. The authorities need to clarify the distribution o f responsibilities, with schools, their boards and the municipalities in charge o f staffing decisions, decisions over spending, curriculum and textbook choices and setting class schedules. 2.33 One of the most complex choices the authorities will need to make concerns the management of the school network. Declining school-age cohorts will provide an opportunity, Similarly, even if salaries are not decentralized to municipalities, the current system of three separate earmarked grants for transport and for maintenanceand operationsshouldbe mergedinto one grant. A more radical step would be to give municipalities control over the terms and conditions of employment. Such a step is not recommended, however, given the significant disruption this will cause, coupledwith other significant changes that would be taking place. 32 especially at the primary level, to reduce the number o f teachers and downsize or eliminate schools (although internal migrationwill require additional classrooms in some locations): The authorities need to decide whether the central or the municipal government will close inefficient schools when needed. A solution to consider is a partnership between the central and local governments, with the central government providing guidance on closure decisions and some transitional funding and the municipal governments responsible for implementingplans. There i s a further issue with respect to the secondary education network, as each municipality is not able, nor should it try, to offer all types o f secondary programs. As a result, collaboration among the municipalities is essential for providing access to secondary schools. 2.34 Each of the issues discussed in this section has been resolved differently in other countries. FYR Macedonia will need to learn from the experience and find the set of solutionsthat will work most effectively toward increasing efficiency and improving quality. This observation argues for an approach which includes the use o f pilot schemes to evaluate differentoptions. 2.35 As of early May 2007, the authorities intend to begin a conservative and cautious transitionto Phase 11,commencingwith a `detour' inthe flow o f funds ftom the Treasury to the bank accounts o f individual school staff via the municipal Treasury subaccounts o f selected municipalities, rather than directly as at present. These funds will be earmarked for salaries. While municipalities will authorize the Treasury to transfer the funds to the accounts o f individual staff, they will not have the authority to add staffing positions (without the concurrence o f the Ministry o f Education and Science and the Ministry o f Finance) or to change the salary structure o f its teachers. At present, there are no additional changes planned for the start o f Phase I1that involve the existing degree o f local control over the determinants o f the education wage bill or the allocation o f funds destined for this purpose. Nor will there be any change in the distribution of salary funding among jurisdictions. The amount assigned to each municipality will be identical to the amount spent on teacher salaries in the preceding year-except for any adjustments due to nationwide changes in the salary structure or any centrally approved changes inthe number of authorized teaching positions. 2.36 The proposed approach would have some advantages in that it would enable the new system of financial flows to be tested or adjustments madeas needed. Ifthe Government intends to transfer significant management responsibilities to local governments and/or allocate block grants on a meaningful formula basis, additional and more substantive steps in decentralization will need to be taken in the future, however. This will require, as was done before the start o f Phase I,significant preparatory work by both ministries and municipalities. ImprovingEducationPerformance 2.37 Additional funding for non-salary inputs is needed and reducing the share of spending on salaries is the key option to consider, especially given the decline in school-age cohorts in the coming years. Together with rationalizing the school network, efforts to keep increases inteacher salaries moderate would result in significant savings over time. 33 2.38 Increased resources for education could come from private sources, especially at the tertiary level. Based on evidence from other countries, FYR Macedonia can expect to increase enrollments in tertiary education significantly only with additional private resources. Given the very low levels o f enrollments in private secondary schools and the government's desire to increase overall enrollment, expanding private schools should be encouraged. This will, however, require attention to the regulatory framework, especially the way in which private and public institutions are accredited and quality assurance is conducted. 2.39 There are several areas within the education sector that could benefit from a reallocation of spending away from current outlays, especially wages. The following recommendations take the analysis and these factors into consideration: Increase instructional time. Without devoting additional resources, this could start by reversing a decision o f several years ago to cut lesson duration for primary students (grade 1-4) to 40 minutes from 45 minutes and limited to 4 the instruction hours for the same students. An increase in the share o f time teachers devote to teaching should also help. Increase access to learning materials. The priority investments should be in ensuring all students in primary and secondary education from poor families have access to learning materials (especially textbooks). The most effective way o f doing this (and increasing school attendance) would be to link the provision o f materials to the provision o f social assistance benefits, though care would be needed to ensure that families do not sell the new books they receive and buy used ones at a lower price. In addition, there is a case for improving school libraries, perhaps through a set allocation o f books provided by the central government (or a block grant to schools). 0 Improve teacher training. This should be focused in the first instance on key competences in mathematics, science, and literacy by providing every teacher with 10 days training over a two-year period. Move towards full enrollment in primary and secondary education. The additional recurrent costs o f reaching full enrollment in both primary and secondary education are estimated to be 0.4 percent o f GDP. The authorities should consider moving to full enrollment over time in order to deal with capacity constraints and to take advantage o f declining cohorts. A period o f 3 years is to be considered. 0 Restructure secondary education. The authorities should examine carefully the future o f vocational education in the country. In line with European experience, there are two key recommendations. First,the authorities are advised to eliminate gradually three-year VET programs as the labor market outcomes from these programs are very poor and once enrolled inthem, students do not have access to tertiary education. Second, the authorities should consider moving away from the sharp distinction between the curricula offered in gymnasia and four-year VET schools. In particular, in all programs there should be a focus on the application o f literacy and mathematics skills and knowledge, analytic skills, team work, and creativity. The opportunity to shift from gymnasia to four-year VET schools should be reinforced. In particular, given the evidence from PISA that placing students permanently in either a general education or a VET track ("tracking o f students") produces high variance in learning outcomes, it is suggested that the authorities reconsider the advisability o f the German dual system. 0 Increase the use of information technology in the learning process, ensuring that the purchase and use o f computers is seen as a part o f the fundamental reform o f teaching in 34 the classroom and incorporated into the ~urriculum.'~Related to the current plan to procure computers for each student, the authorities should be mindful that the costs o f maintenance and operation will ultimately far exceed the initial investment in hardware, an outcome with substantial budgetary implications. All told, the authorities should reconsider the advisability o f such a move. 2.40 Significantly improve the quality of data collection and databases in the education system with a view to improving the management of the sector (including monitoring and evaluation). It is recommended that the authorities focus on creating or improving the following databases: (i)an internationally interpretable set o f statistics on FYR Macedonia's education system; (ii)a school mapping database, building on the database now managed by the Policy Analysis Unit (PAU);20 (iii)an expanded education module within the household survey managed by the statistics office; and (iv) a database that supports analyses o f learning outcomes and the ancillary informationthat is routinely collected at the time o f the assessment. 2.41 A separate study by the World Bank will discuss the issue of research, science or tertiary education. Preliminary conclusions from the study indicate that in tertiary education, one o f the highest priorities has to be given to the restructuring o f university programs and degrees in line with the requirements o f the Bologna process. It i s important that this is more than just a paper exercise to divide existing curricula into credit units. Rather, it should be seen as an opportunity to revise the curricula and the approach to teaching and learning. In fact, adhering to the Bologna process will have significant implications for the way universities are structured and managed and it is with this aspect that reform o f higher education should start. These changes would require significant resources over an extended period o f time. Increasing enrollments in tertiary education should come primarily through increases in private institutions. However, a robust accreditation system for both public and private institutions i s needed to ensure public resources are directed towards good quality institutions. Research, meanwhile, is an area where FYR Macedonia faces major challenges and will require significant resources to address. E. CONCLUSIONS 2.42 Better education outcomes are central to the future of the country, as they clearly lead to faster growth in output and higher living standards. Increasing public spending on education relative to GDP is not the answer to the current poor education outcomes, however. Rather, the authorities should focus on improving the efficiency and equity o f government education spending and encouraging the productive use o f private spending, especially at the tertiary level. Fast growing emerging market countries, such as Chile and South Korea, offer valuable examples on how to achieve better education outcomes. l9See "Computers in Schools: Policy Note," The World Bank, November 2006 for more on this issue. 2o The PAU does not receive all necessary (and available) data because of inadequate lines of communication with other institutions (the Ministry of Finance) and other departments within the MES, such as the departments for primary and secondary education. The data that PAU receives from various sources within the same institution or department often have inconsistencies, the data collection method (paper entry) being the main reason for the discrepancies and errors in the data. PAU's data management software is not user friendly and does not allow for in-depthanalysis. 35 3. HEALTH A. INTRODUCTION 3.1 The health system has sustained better health outcomes than in countries with similar levels of income per capita, and has retained almost universalhealth insurance, but popular dissatisfactionwith the quality of healthcareis high. Such outcomes, however, have been achieved at a high cost, both direct and indirect. Public health outlays are one o f the highest inthe region relative to officially estimated GDP (or similar relative to GDP adjusted for the size o f the NOE), and higher than among the N M S on both measures. Moreover, private co- payments, both formal and informal, are large, resulting in overall healthcare spending that is quite higheven measured relative to GDP adjusted for the size o f the NOE. Further, the bulk o f public health financing is through distortionary social security contributions that, although rather low inthe region, are contributingto the informalization o f the economy. 3.2 Improving the performance of the health sector will require ambitious government actions to improve both the efficiency and equity of spending. Measures will be needed to strengthen both the allocative efficiency o f public finds (that is, using public resources for achievingthe government goals in healthcare) and the technical efficiency (maximizing the value for money allocated). The chapter recommendsthat the authorities consider several priority avenues for reform. These include the need to revise and substantially tighten the benefits package for social health insurance, tighten procurement practices, especially for pharmaceuticals, and adopt modern provider payment and performance management systems. Other priorities include the need to ,strengthen the regulatory regime for healthcare institutions and to improve the targeting o f eligible persons for publicly-financed health insurance premiums, while enforcing contribution collection from nonexempt groups. 3.3 The rest of the chapter is organized as follows. Section B reviews the institutional environment. Section C discussed healthcare revenues and financing. Section D analyzes healthcare expenditures and section E focuses on the delivery and utilization o f healthcare. Section F presents the chapters recommendations and section G offers concluding thoughts. B. INSTITUTIONAL ENVIRONMENT 3.4 The health services delivery system in FYRMof Macedonia consists of an extensive network of public health facilities, ranging from small primary health care facilities (ambulanta), to larger primary health care centers with diagnostic facilities and specialists (zdravstven dom) and to hospitals. Health reforms in recent years have focused on encouraging greater private sector participation in the provision o f health services, and in comparison with other SEE countries this process i s well advanced. Four private hospitals exist (all in Skopje), pharmacy privatization is complete and privatization o f dental practice and primary health care is advanced. 3.5 Similarly to the NMS and SEE countries, FYR Macedonia has a publicly financed and administered health insurance system. At present, a single Health Insurance Fund (HIF) pools health insurance contributions levied on wages, transfers from the budget and co-payments, ultimately managing about 95 percent o f health sector resources. The remainder is channeled 31 through the Ministry o f Health (MOH) for administrative costs and for health services provided by specialized hospitals (for example, the military hospital). Regulatory functions are carried out by the M O H and bodies under its jurisdiction, with a recently established Food Directorate responsible for all food safety issues. 3.6 There is a purchaser-provider split in the health system, with the single payer, the HIF,contractingwith public and private providers to deliver health services. Given the small population size, the risk pooling through a single HIF is a substantial strength o f the health care system. Reflecting the comprehensive nature o f the benefits package financed through social insurance, however, voluntary health insurance remains underdeveloped despite provisions in the Health Insurance Law making it possible. Similarly, the high share o f the population covered (more than 90 percent) has been touted as strength, but it has also come at a cost, ultimately to the budget, given pervasive evasion from payments o f contributions. 3.7 Much has been achieved in addressing the internal controls and governance of the Health InsuranceFund. Experience from the broader region o f Eastern Europe and central Asia and elsewhere demonstrates that implementation o f complex health financing reforms is not possible in an environment o f weak controls and governance. The government's interventions so far have resulted in a reduction of HIF arrears, improving the governance o f the HIF board and strengthening HIF capacity in financial management and performance monitoring o f health care institutions. 3.8 The Government has recently adopted a comprehensive health sector strategy that outlines reforms towards improving the efficiency and equity of health spending. The following reforms are envisaged: (i)a revision o f the health benefits package to help improve access, quality and fiscal sustainability (for example, expanding the negative list o f services, restructuring co-payments and improving targeting), and (ii)strengthening contracting o f health care providers (for example, launching competitive procurement for health services, using the medical map to identify service needs in each region). Accelerated implementation o f the health strategy combined with the ongoing interventions on governance and internal controls will reap important sustainability and quality gains in public spending on health. C. HEALTH OUTCOMES 3.9 Health outcomes are better than in other countries with similar levels of income per capita and are on an upward trend, but - not surprisingly lag the EUcountries, including - most of the new member states on most indicators (Table 3.l)and (Table 3.2). The adult mortality rates in FYR Macedonia are roughly half the rates in Thailand and El Salvador, for example. By contrast, FYR Macedonia's standard death rate (SDR) for circulatory diseases is more than double the EUaverage and 25 percent higher than the NMS average. The incidence o f tuberculosis, moreover, is three times higher than the EU15 average. 38 Table 3.1: Morbidity and Mortality Indicators, 1991, 1995,2000,2004 1991 1995 2000 2003/2004 EU e EU15 Average SDR, Diseaseso fthe Circulatory System 527.5 603.8 582.2 599.1 262.4 233.2 per 100000 (2003) SDR, Malignant Neoplasmsper 100 139.5 149.2 163.6 165.1 184.2 177.9 000 (2003) Tuberculosis Incidenceper 100000 35.2 40.0 31.6 31.7 11.9 9.7 SDR External Causes of Injury and 40.9 30.8 37.9 32.9 42.8 37.8 Poisoningper 100000 (2003) Clinically Diagnosed AIDS Incidenceper 0.0 0.3 0.2 0 1.6 1.8 100000 Cancer Incidence per 468.2 100000 142.5 276.4 (2000) New HIV Infections Reportedper 100000 0.1 0.3 0.4 0.3 5.3 6.8 Source: Health in Transition(HiT), Macedonia, 2006. Table 3.2: Health Outcomesin Macedonia and Other Countrieswith Similar Income Levels Health GNI Per Expenditures Adult Survival to Capita per Capita Life Mortality Age 65 (YOof Country (US$) (US$) Expectancy U5MR per 1000 Cohort) Macedonia 2,420 161 74 14 115 80 Thailand 2,490 76 71 21 193 73 Albania 2,120 118 74 19 78 81 Belarus 2,140 116 68 11 249 68 BH 2,040 168 74 15 121 81 Kazakhstan 2,250 73 65 73 255 60 Ecuador 2,100 109 75 26 149 76 Salvador 2,320 183 71 28 185 75 Peru 2,360 98 70 29 160 74 Algeria 2,270 89 71 40 130 77 Jordan 2,190 177 72 27 155 78 Source: World DevelopmentIndicators, 2006. D. HEALTHCARE REVENUESAND HEALTHCARE FINANCING 3.10 Payroll taxes (social security contributions) are the main source of HIF revenues. Highunemployment and evasion, however, together with relatively low contribution rates (one of the lowest in ECA, see Table 3.3), have made it necessary to utilize additional sources o f revenues. These include general budget transfers and out-of-pocket payments. Formal out-of- pocket payments have also been supplemented by informal payments. Table 3.4 presents the key sources o frevenue and spending o fthe HIF. 39 Table 3.3: Social Insurance Financing Mechanisms, EU8 and SEE Countries Share of Country Sources of Funds Contributions (in percent of earnings) Contributions Paid by Employee Bulgaria SHIpremiums 1/ 6.0 50 Czech Republic SHIpremiums 13.5 35 Estonia SHIpremiums 13.0 0 Hungary SHIpremiums 23.5 25 Latvia Income tax and other N/A NIA generalrevenues Lithuania SHIpremiumsand 6 N/A - 30% of employee incometax income tax (upto 33% of salary goes to SHI) Poland SHIpremiums 7.75 100 Slovakia SHIpremiumsand 14 28 incometax Slovenia SHI premiums 13.25 47.3 Croatia SHIpremiumsand 16 56 incometax Macedonia SHIpremiums 9 0 Serbia SHI premiums 11 0 "SHI=socialhealthinsurance. Source: CIS 2005, World Bank 2005. 3.1 1 Health insurance coverage is almost universal, but only a fraction of those covered actually pay contributions. There are 14 categories o f insured, with the employed and the self- employed (including in agriculture) contributing at the statutory rate o f 9.2 percent o f gross wages and allowances, and others at different rates.*' Although the statutory contribution rate is relatively low (with Albania, Bulgaria, Lithuania and Poland having still lower rates in ECA), a statutory minimum threshold in place renders the effective rate much higher on the margin in FYR Macedonia, especially for those with low wages, poor skills and the young.22 Such a threshold encourages informality, together with the relatively generous benefits offered the unemployed (including full health benefits, leading many informally employed to register as unemployed). 3.12 The authoritieshave embarked on reforms to harmonize collection of socialsecurity contributions to help reduce administrative complexity and burden of multiple agencies. The intention is for the Pension and Disability Fund (PDF) to begin collecting all contributions, with the Public Revenue Office (PRO) ultimately taking responsibility for collecting both contributions and taxes. As an added advantage o f these reforms, the HIF will be free to focus on strategic management o f health resources, a function that is crucial for improving the efficiency and cost-effectiveness o f health expenditures. Similar collection reforms were undertaken in Slovenia and Croatia, resulting in substantial improvement o f contribution collection. 21Unlikemost countries inthe region, all contributions are fully paidby the employers. 22 Payroll payments for most categories must be no lower than those due on a minimum threshold equivalentto 65 percentofthe average net wage. 40 Table 3.4: Macedonia: Revenuesand Spending of the Extrabudgetary HealthFund (Inmillions ofdenarsunless indicatedotherwise) 2005 2005 2005 2006 Budget Outturn Budget Outturn Revenues 15,724 15,083 15,820 15,697 (InpercentofGDP) 5.5 5.3 5.2 5.2 Contributions from employers 9,257 8,982 9,276 9,506 Contributions from the PensionFund 3,547 3,417 3,637 3,583 Contributions from the Employment Fund 2,080 2,062 2,143 2,145 Contribution from the Ministry of Labor 64 64 46 56 Transfersfrom the central budget 68 46 66 42 For programs 64 44 41 40 For healthinsurancefor disabledpersons 4 2 25 2 Revenues from co-payment 600 434 573 241 Other revenues 108 79 79 124 Expenditures 15,724 15,206 16,615 16,308 (InpercentofGDP) 5.5 5.4 5.5 5.4 Healthcareat home 13,764 13,440 14,732 14,387 Medicaltreatment abroad 130 151 155 164 Other treatment (compensation) 1,050 1,062 1,200 1,145 Administration 317 276 418 373 Wages and allowances 193 163 194 149 Goods and services 124 113 224 224 Capital expenditures 220 106 110 57 Other expenditures, including interest 63 0 0 0 Balance 0.0 -123 -795 -611 (InpercentofGDP) 0.0 0.0 -0.3 -0.2 Sources: Ministry of FinanceandHealthInsuranceFund;and staff calculations. 3.13 Very low levels of formal employment in FYRM of Macedonia and near universal health insurance coverage have brought substantial inefficiencies, raisingserious questions about the appropriatenessof financinghealthcarewith payrolltaxes. Contributions are often made on the basis o f the minimum threshold rather than actual wages, resulting in regressive effective tax rates. Thus, given high unemployment, even the moderate level o f the payroll contribution rate adds to the cost o f labor on the margin and creates an additional disincentive for job creation. Demographic changes under way are a second key factor to consider, even if those older than 65 years account for a smaller share o f the population in FYR Macedonia(10.6 percent in2003) compared to the EUon average (16 percent). It is likely that going forward, population ageing will result in a smaller workforce, putting downward pressure on contributions. Moreover, the older the population, the more intense the use o f higher-cost health services. 3.14 Given these considerations, one option available to health policy makersis to move to financingthrough general revenues. Such a reform may require an increase in VAT rates or the flat income tax rates. A hike in the latter could be accompanied by an increase in the tax- exempt minimum. Such financing would, inter alia, help eliminate or sharply reduce payroll 41 contribution rates (with all positive consequences listed above) and strengthen the link between policy priorities and budget formulation. The negatives, however, are equally powerful, with the most important concern reflecting the risk that the public may now perceive resources to be readily available, resulting in a tendency to boost further spending on health. The authorities should consider the trend toward financing through general revenues even in countries with a long tradition o f social insurance (for example, France and Spain; also note recent proposals in Germany). Countries that already had strong tax-financed systems (Sweden, United Kingdom and New Zealand) have further strengthened this model. 3.15 The low effective levelof official co-payments (averagingabout one-half of statutory requirements) has limited their effectiveness in curbing demand for health services and is bringinglittle revenuesto the HIF. By law, all insured except the exempt are required to make co-payments for the use o f health services and pharmaceuticals. Co-payments are set in fixed amounts, with co-payments for more expensive services fixed at a proportionally lower percentage. There are various exemptions from making co-payments, usually set by population group and incidence o f disease. For example, insured low-income individuals pay smaller amounts, as do children under 5 years o f age, youth under 18, and people older than 65 years. In addition, there are co-payment exemptions for mothers, for immunizations, and the treatment of some diseases. In addition to low effective co-payments, there is little incentive for the health care institutions to collect them since, by law, co-payments collected have to be returned to the HIF. All these factors combine to limit revenues from co-payments to 3-4 percent of total HIF revenues, compared with 7-10 percent in most west European countries. E. HEALTHCARE EXPENDITURES 3.16 Given the growing reliance in healthcare on advanced appliances and pharmaceuticalsproduced abroad, the absolute amount of spending on healthcare matters more than in other sectors. FYR Macedonia spends about $200 per capita on healthcare, less than one-halfthan the N M S on average, but more or about the same than in countries with similar levels o f income per capita. Relative to GDP, government healthcare expenditures amount to about 5.7 percent, compared with 5 percent in the N M S and 6.5-7.5 percent in Croatia and Bosnia and Herzegovina (Table 3-5). Out-of-pocket payments, both formalized and informal, are estimated to amount to 3 percent o f GDP (higher than the 2 percent average in the NMS), bringing total healthcare spending to about 8.7 percent o f GDP.23 OECD countries spend as much on healthcare on average, but these are far wealthier economies whose health outcomes are substantially better. ~~ ~ 23 Official estimates of private outlays amount to 1.1 percent of GDP, but these underestimate the true magnitude. The figure inthe text is estimatedfrom the HouseholdBudget Survey (HBS). 42 Table 3.5: Health Expenditures in EU8, SEE OECD and other MIC Countries YOof GDP Public, YOof GDP Private, YOof GDP Per Capita US$ SEE Countries Bosniaand Herzegovina 10.0 7.4 2.6 168 Croatia 7.6 6.3 1.3 494 FYRMof Macedonia '' 8.7 5.7 3.0 208 Serbiaand Montenegro 8.6 5.7 2.9 181 EU8 Czech Republic 7.5 6.8 0.7 667 Estonia 5.3 4.1 1.2 366 Hungary 8.4 6.1 2.3 684 Latvia 6.4 3.3 2.1 301 Lithuania 6.6 5.0 1.6 35 1 Poland 6.5 4.5 2.0 354 Slovak Republic 5.9 5.2 0.7 360 Slovenia 8.4 6.3 2.1 821 EU8 Average 6.9 5.1 1.8 488 Other countries with similar level of income per capita Thailand 3.3 2.0 1.3 76 Peru 4.4 2.1 2.3 98 Algeria 4.1 3.3 0.8 89 Jordan 9.4 4.2 5.2 177 El Salvador 8.1 3.7 4.4 183 OECDAverage 8.9 6.1 2.8 NA "ThedataforMacedoniaincludetheestimatesofprivateexpendituresin2004basedonthe,HBS. Source: World DevelopmentIndicators, 2002-04. 3.17 The poor still spend more on healthcare as a share of total household expenditures than those better-off, raising concerns about access to care. The authorities have moved to address this issue, with the amount paid by the poorest quintiles falling by 27 percent during 2002-04 and the amount paid by the richest quintile risingby 18 percent. 3.18 High healthcare outcomes reflect in part high costs, the latter reflecting a number of factors. Key among them is the overemphasis on specialist and hospital-care and the excess number o f medical personnel, with both o f these factors resulting in an oversized wage bill. For example, the proportion o f the HIF budget spent on hospital care i s about 40 percent (although that share has been decreasing over time) (Table 3.6). The comprehensive benefit package, the sizable outlays on pharmaceuticals and cost-pressures resulting from medical technology (typically acquired at world prices) are also important. These factors are considered in the following sections. 43 Table 3.6: HIFExpenditures by Level of Care (2000-2004) (Inpercentoftotal HIF expenditures) 2000 2001 2002 2003 2004 Primary healthcare 20 19 21 37 37 Dental care 6 4 4 4 4 Inpatient care 46 50 42 34 34 Outpatientdrugs 14 13 14 13 13 Medical devices 1 1 1 2 2 Specialprograms 1 3 0 0 0 Treatmentabroad 1 1 1 2 1 Other healthcare expenditures 1 0 0 0 0 Capital investment and O&M 1 0 6 0 1 Pecuniarycompensation 6 6 7 7 7 Expendituresfor fund operations 3 2 3 2 2 Credits repaymentand coverage of Fund'slosses 2 1 1 0 0 Source: The Health InsuranceFund. 3.19 Wages account for a full one-half of the HIF expenditures, but this is a familiar pattern across many countries at similar income levels. (For example, medical wages in Bosnia and Herzegovina account for roughly one-half o f healthcare outlays.) As in other areas, the large share o f wages crowds out much needed outlays on investment and maintenance. This outcome is especially problematic given the absolute amount spent on healthcare o f about $200 per capita in current dollars. 3.20 The size of arrears of the HIF and the healthcare institutions (HCI) has been reduced substantially of late. Arrears fell to about 1.2 percent o f GDP by the end o f 2006 (3.7 billion denars) from 1.8 percent in mid-2005. HIF arrears have been reduced to nil within the total amount. The gradual reduction in arrears has reflected the implementation o f hard budget ceilings for the HCIs, and was aided by the implementation o f monthly financial reportingby HIF and HCIs (that allowed better tracking o f liabilities) and the introduction o f competitive bidding for pharmaceuticals (which contributed to a decline in pharmaceutical expenditures by one-third). Nonetheless, H C I arrears remain an issue that needs to be tackled. F. THEDELIVERY AND UTILIZATIONOF HEALTHCARE 3.21 Despite high spendinglevels, utilizationof health services is below the levelsin other EUcountries. The latest available data on outpatient visits per capita inFYRMacedonia is from 2001 (and excludes the private sector), indicating a much lower number o f visits per capita (3 on average, 1 in rural areas) compared with 6.8 among the EU15 countries and 8.5 among the Nh4S (excluding Bulgaria and Romania) (Table 3.7). The number o f outpatient visits, especially visits to primary care doctors is a good indicator o f the efficiency o f the system. L o w outpatient visits may indicate a bypassing phenomenon, where instead o f going to a primary care doctor, patients self-refer to secondary or tertiary care institutions. The bed occupancy rate in FYR Macedonia is only about 54 percent, compared with 77 percent on average for the EU15 and OECD countries. There is little information on the factors for low utilization o f health services inFYRMacedonia, but experience from other countries in the region indicates that the cost o f health care is typically 44 a barrier. This is also likely the case in FYR Macedonia, given the large size o f out-of-pocket payments that supplement mandatory co-payments. Table 3.7: Outpatient Contactsin SelectedCountries (per capita, peryear) 1990 1995 2000 2001 2002 2003 2004 Albania 3.3 1.8 1.6 1.6 1.8 1.9 1.9 Bulgaria 6.6 5.5 Croatia 8.2 4.8 7.0 7.5 7.5 7.6 7.6 Serbia and 7.0 5.0 7.1 Montenegro Slovenia 6.5 7.2 6.8 6.7 6.4 6.9 7.0 Macedonia 5.2 3.1 3.2 3.0 N A N A NA EU15 average 6.5 6.3 6.6 6.6 6.7 6.8 Source: WHO RegionalOffice for Europe, Healthfor All Database,January 2006. Preventive and Primary Healthcare 32 2 Health-seeking behavior among the population indicates limited use of primary care. More than one-half o f the population select ambulance/emergency care as the first contact with the healthcare system, followed by a hospital visit (Table 3.8). This indicates a preference for treatment in a hospital setting, and bypassing o f primary health care. The poorest quintile tends to use ambulance services and hospitals more than the richest quintile, although the margin i s relatively modest. Table 3.8: HealthSeeking Behavior by Income Quintile (Inpercentofthe total) Total Q1 4 2 4 3 4 4 Q5 Self-Treatment 7.8 8.1 7.6 8.4 8.6 6.3 Private Medical Doctor 11.8 6.8 11.9 9.4 11.3 18.7 AmbulanceEmergency Care 55.4 55.1 54 59.6 57.5 49.7 Clinic 5.6 4.7 5.2 5.5 5 7.4 Hospital 19.3 25.1 20.7 16.4 17.3 17.7 Alternative Medicine 0.4 0.3 0.7 0.5 0.3 0.1 Other 0.1 0 0 0.3 0.1 0.2 Source: HBS. 2005. 3.23 The health care delivery system in FYR Macedonia is still largely dominated by specialist and hospital-based care with insufficient emphasis on preventive and primary care. Data on primary healthcare expenditures (PHC) in FYR Macedonia do not differentiate between outlays on bona fide primary expenditures and those going to specialist outpatient services, thus making it difficult to judge recent trends (data for other countries do not group these two together). The authorities would do well to establish a more comprehensive database that separates the two to enable better policy analysis and planning. All told, overall allocations for primary health care and specialist outpatient services have increased in recent years, reflecting HIFcontracts with privatePHC doctors. 45 In-Patient Care 3.24 Expenditures on in-patient care in FYR Macedonia amount to about 40 percent of the total, broadlysimilar to the average in the NMSand the OECD.24Hospital productivity, however, is very low, with the inpatient admission rate in FYR Macedonia equivalent to half the average for the EU15 and lower than in all countries in the region but Bosnia and Herzegovina (Table 3.9). Similarly, the hospital occupancy rate is only 54 percent, compared with averages o f 78 percent in the EU15, 74 percent in the N M S and 87 percent for the Commonwealth o f Independent States. Table 3.9: InpatientUtilization and Performance in Acute Hospitals in the WHO European Region, 2004 or Latest Available Year Hospital Beds Admissions per Average Length Occupancy per 1000 100 Population of Stay in Days Rate (YO) Population Albania 2.7 BH 3.3 7.2 9.8 62.6 Bulgaria 7.6 14.8 10.7 64.1 Croatia 3.6 14.6 8.2 89.9 Czech Republic 6.2 20.8 8.2 74.8 Estonia 4.3 17.2 6.2 68.4 Hungary 5.9 23.5 6.5 76.6 Latvia 5.4 18.8 Lithuania 6.1 21.9 7.9 77.4 Poland 4.7 Romania 4.4 SAM 9.7b 69.0 Slovakia 6.1 17.8 8.4 68.6 Slovenia 3.9 16.6 6.2 73.2 Macedonia 4.9 9.0 11.8 53.7 EU15 average 4.0 18.0 6.9 77.0 NMSaverage 5.2 20.6 7.4 73.8 CIS average 7.4 19.5 11.6 87.1 Source: WHO Regional Office for Europe, Health for All Database, January 2006. 3.25 Although at the aggregate level bed capacity is not excessive, the inefficient use of hospitalbeds combineswith their unequaland irrationaldistribution. Hospital beds amount to 4.9 beds per 1,000 persons, compared with 6.2 beds per 1,000 persons in the EU on average. More than one-half o f the beds are in specialized and tertiary care. Outside Skopje there are 15 general hospitals and the bed capacity varies from 5.2 per 1,000 in Shtip to 1.3 in Kochani. The variation in bed capacity is largely attributable to the fact that the 15 hospitals have different specialties. Given the size o f the country and the fact that population in some regions have access to multiple general hospitals, there is scope for rationalizing service delivery through a combination o f planning and use o f output-based payment mechanisms (case-based payments). Under case-based payments, hospitals without sufficient numbers o f cases are forced to either rationalize or merge with other hospitals. In the region, successful implementation o f case-based payments for hospitals has contributedto 30-40 percent reduction in hospital fixed costs. 3.26 There is a surplus of health personnel and non-medical staff, contributing to high fixed costs, especially in the hospital sector. Overstaffing reflects in large part the generous 24Outlays in in-patient healthcare amounted to 37 percentof the total inMacedonia, 38 percent or lower in the OECD countries, and about40 percentinHungary, Polandand Slovakia. 46 staffing norms under the previous system that targeted 1 physician per 1,000 inhabitants. In addition, the percentage o f full-time non-medical staff (cleaning personnel, cooks, accountants) in health facilities accounts for about one-third o f total staff. Health facilities do not have flexibility over hiring and firing and labor unions in FYR Macedonia are strong, contributing to high fixed costs and limited opportunity for hospital managers over the hospital input mix. Labor rigidities are one o f the key obstacles to introducingincentive-based payment systems at the hospital level. 3.27 The recent introductionof "global budgets" for hospitals (which provide hospitals with a certain overall amount without specifying in detail what the amount should be used for) has helped bring greater fiscal discipline in the Health Insurance Fund and to the healthcare providers. Prior to 2006, the HIF faced serious fiscal risks generated by the lack o f clear and enforceable criteria for paying hospitals. While hospitals were billing the HIF for health services on a fee-for-services basis, the HIF ignored these claims and was making payments on the basis o f historical, line-item budgets. This led to confusion in the health sector and the gap between hospital claims and the HIF payments resulted in arrears. To bring this problem under control, in 2006 the HIFadopted global budgets for all HCIs, with the objectives o f strengthening fiscal discipline. The global budgets for 2006 were defined according to the historical budgets for the previous three years. BenefitPackage 3-28 The comprehensive benefits package needs to be evaluated using cost-effectiveness and allocative efficiency criteria. The other countries o f the former Yugoslavia have similar benefits packages. Currently, the benefits package in FYR Macedonia provides the 90 percent o f the population enrolled in the HIF with comprehensive coverage for outpatient and inpatient health services. In contrast to the EU countries, non-medical benefits including sick leave and .maternity benefits are also included in the healthcare package. 3.29 There is an overlap and duplication of benefits. In addition to the coverage under the compulsory health insurance, certain categories o f citizens (children, students, pregnant women, persons suffering from renal failure, diabetes) are eligible for free care under "special programs" financed through general budget revenues. It is difficult to evaluate expenditures on such programs since some o f the costs are hidden under "expenditures on primary health care" and "expenditures under special programs." Special programs seem to be routinely under-funded relative to statutory needs, leading in the past to healthcare arrears. For example, the 2000-2004 annual approved budget allocations for special programs amounted to less than 40 percent o f statutorily-defined needs. The "special programs" also cover some o f the most expensive drugs and medical devices such as renal dialysis, cytostatic medicines (for the treatment o f cancer). PharmaceuticalExpenditures 3.30 Public pharmaceuticalexpenditures are a key cost driver of healthcare, although at 15 percent of HIF expenditures they are only half as large as the average for the NMS. In Macedonia, as in other EU countries, outpatient pharmaceuticals consume the bulk o f resources, while inpatient drugs only consume 10-15 percent o f expenditures. FYR Macedonia is also likely to face continued growth in pharmaceutical expenditures inthe short to medium-term. While this is inevitable, by strengthening system efficiency and tackling abuse, FYR Macedonia can at least guarantee that some o fthe cost increases can be neutralized and absorbed. 47 G. RECOMMENDATIONS 3.31 While healthcare reforms have been advanced, more remains to be done to ensure that the government's health policy goals are met, includingthe emphasis on improving health outcomes and setting the system on a financially sustainable footing. The following recommendations are based on the analysis in this chapter and are split into suggestions for immediate action (over the next one or two years) and a more medium-term policy agenda. 3.32 The authorities are invited to consider the following immediate policy actions: HealthFinancing Complete the implementation of reforms to harmonize benefits and bases for collection of benefits and improve administrativeefficiency in benefits collection. A key factor for success is promoting transparent and accountable transfer o f funds from PRO to HIF and data sharing between the agencies (e.g. beneficiary database). This will enable HIF to track contributions and disallow the use o f health benefits for non- contributors. Specify explicitly the groups whose health financing is provided from the budget, include the full cost of their premiums in the budget as special line items and monitorbudgetexecution. Consider moving healthcare financing, in part or in whole, from payroll taxes to general revenues, starting with analyzing the fiscal impact of such a potential change. Health Expenditures Strengthen the implementationof global budgets for hospitals and implement more relevant key performance indicators differentiated by the type of care (primary, secondary and tertiary). For hospitals that have adhered to hard budget ceilings, have developed business plans and begun implementation o f efficiency measures, some flexibility over line item re-allocations could be allowed. 0 Begin the (technically and politically challenging) reforms of the health benefit package. Key areas for reform include: (i)expanding the negative list o f services based on specific criteria (for example, cost effectiveness, public and merit goods); (ii)reducing or eliminating entirely non-health benefits (as Serbia did in 2006); and (iii)integrating special programs inthe benefits package. Advance reforms of the pharmaceuticalsector by: (i)introducing generic referencing for outpatient drugs and external referencing for patented drugs; (ii)improving information systems linking HIF with pharmacies and allowing HIF to monitor and address prescription fraud, (iii)strengthening mechanisms for determining the inclusion o f new (expensive) drugs on the HIF positive list, (iv) evaluating the effectiveness o f physician drugbudgets on prescription behavior o f physicians. 0 Evaluate the role of publicly provided, fully paid and co-paid services in encouragingmore appropriatepatientbehavior. The differentiationo fhealth services 48 into these categories should be based on an understating of the set of incentives and patient reactionto these incentives. Strengthen the capacity of the HIF,MOH and the HCIs to help limit cost pressures and eliminate wasteful spending. For the HIF, begin with (i)improvingprocessingof claims, and detecting and resolving fraudulent billing; (ii)strengthening contracts with provides and ensuring compliance; and (iii)collecting and analyzing provider data on a regular basis, and using this information to determine the price and volume of health services. 0 Implement the National Health Accounts to ensure provision of standardized and timely data for policy development andmonitoring. 3.33 The immediate policy actions are recommended to be supplemented with the following efforts over the mediumterm: 0 Develop and begin implementing a plan for reducing overcapacity in the health sector, especially in the hospital and specialist outpatient health services. One approach is to develop a regionalrationalizationplanthat can be usedto planthe delivery of services at the regional level and link the implementation of these plans with appropriate prospective paymentmethods. Strengthen institutionaland organization reforms to allow providers to respond to new incentive regime. For example, the implementation of prospective provider payment systems (e.g, case-based payments) needs to be combined with granting hospitals autonomy over budget allocation, allowing hospitals to retain profits, and determine staffingaccordingto needs. 0 Strengthen mechanisms for paying providers at other levels of the system. For example, FYR Macedonia needs to strengthen payment mechanisms for primary care aimed at enhancing the gate keeping function of primary care and encouraging group practice. 0 Rationalize further the health benefit package taking into account international experience and the fiscal constraints, and identify scope for rationalizing health benefits. This should include a re-evaluationand reforms of the system of co-payments with the objective of improving targeting (for example based on a revisedmeans-testing system) and introductionof co-payments for specific categories of non-essentialmedical services and products. Concurrently, support the development of voluntary health insurance(VHI) includingthroughan appropriateregulatory framework. 0 Continue pharmaceutical reforms by implementing provider prescribing guidelines and strengthening monitoring of out-of-pocket payments and drug use by population groups to helpcalibrateco-payments. 0 Strengthen the public health and health promotion system, given the burden of smoking, alcoholconsumptionand other preventable diseases. 49 H. CONCLUSIONS 3.34 Further ambitious health reform is needed to improve both the efficiency of spending on healthcare and the health outcomes. These top priorities include the need to revise the benefits package for social health insurance, tighten procurement rules for pharmaceuticals and adopt modern provider payment systems. Reorganizing the hospital sector in line with the Medical Mapthe authorities are preparing should help streamline with the size of the sector and make it more attuned to the needs o f the citizens. 50 4. PENSIONS A. INTRODUCTION 4.1 The financial situation of the FYR Macedonian pension system worsened dramatically during the 1990s due primarily to the significant decline in the contribution base on the back of reduced formal employment. These developments, and the build-up of substantial implicit pension obligations, led the authorities to begin implementing pension reforms, starting with parametric changes of the pay-as-you-go (PAYG) pension pillar and introducinga mandatoryfully-fundedsecondpillar in2006. 4.2 The implementationof the pension reforms has been successful thus far, guided by the objectives of reducing the impact of demographic changes on the pension system and helping assure adequate pension incomes for both current and future generations of pensioners. More work will be neededto ensure that future implementation proceeds smoothly and the pension system copes with all challenges. Using a realistic scenario, with real GDP growth and other key parameters below historicalaverages, this chapter concludes that the fiscal sustainabilityof the pension system is likely to be attained, as measured by a declining and ultimately stabilizing deficit of the pension fund (financed from general government revenues). This result depends on the realistic assumption that the authorities will continue to implementparametric reforms currentlyinprogress, includingthe gradualplannedincrease in the retirement age and maintainingthe recently agreedpensionindexationformula that gives a much larger weight to inflation than wage growth. The conclusion assumes only moderate growth in formal employment, increasingits share in the labor force from one of the lowest in the region, thus increasingthe numbersof contributors. 4.3 The sustainability of the pension system is likely to be assured. The outcome projected by this chapter is for the replacement rate (the ratio of average pensions to average wages) to average 50 percent,comparedwith 55 percent in 2006,68 percentamong the NMSand 78 percent in the EUon average." The projected outcome is reasonableand the modest decline from current levels reflects the impact of the reforms the authorities are implementing. The difference with the EUshouldbe interpretedwith care, as many EUcountries have yet to face the acute demographic problems they face. Faster growth in output and formal employment should result in a much smaller reductioninthe replacement rate, enabling an easier transitionto a more sustainable pension system. Further, reducingthe fees charged on second pillar accounts should also help, and the government is invited to consider the chapter's recommendations in this direction. The authorities are also invitedto consider ways to reduce further payrollcontributionrates in a fiscally responsible way to help reduce informality. 4.4 The rest of the chapter is organized as follows. Section B outlines the overall structure of the FYR Macedonian pension system and discusses its current performance. Section C analyzes the adequacy of future pension benefits, while Section D discusses the long-term 25The replacementrate is definedas the ratio of the average net pensionto the average net wages, unless notedotherwise. financial sustainability o f the system. Section E provides the chapter's recommendations and section F offers a conclusion. B. STRUCTUREOFTHE PENSIONSYSTEMAND PENSIONEXPENDITURES 4.5 The authorities initiated pension reforms earlier this decade in response to the deteriorationof the financial situation of the FYR Macedonianpension system during the 1990s. The deterioration reflected a significant decline in the contribution base, as formal employment fell markedly. A bias toward early retirement became entrenched, combining with a large number o f beneficiaries relative to contributors and generous benefits. The authorities have begun addressing some o f these issues, but further determined progress is needed to help ensure the long-term sustainability of the pension system. Pervasive evasion and a large informalization o f the economy have limited the inflow o f payroll contributions, making it necessary for the authorities to rely increasingly on budgetary transfers to fund promisedpension payments. 4.6 The FYR Macedonianpensionsystem consists of two mandatorycomponents. The first, a PAYG pillar, offers defined-benefit pensions based on lifetime earnings history with pre- defined accrual ratios. The second, a fully-funded pillar based on the principle o f defined contributions, was introduced from the start o f 2006. The funded pillar is mandatory for new entrants into the labor market after January 1, 2003 and optional for the rest o f the working population. Coverage 4.7 Low formal employment and high unemployment in FYR Macedonia has resulted in very low pension coverage. The number o f contributors in 2006 amounted to 425,263, or about one-half o f the officially estimated formal labor force, and a much smaller share o f the economically active population. This creates the risk that a significant part o f the population will reach retirement age with a short pension insurance record and will, thus, be entitled to very low pension benefits. Improving incentives for voluntary compliance has been one o f the main drivers o f reforms. L o w coverage is projected to continue (see below), underpinningthe urgency o f further reforms (Figure 4.1). Figure 4.1: FYR Macedonia, Population over 65 years with Pension Coverage (Inpercent) 90 0% 80.0% 70 0% 60 0% 50 0% 40 0% 30 0% 20 0% 100% 0 0% Source: MAPAS. 52 The PAYGPillar 4.8 Pension expenditures amounted to about 8.5 percent of GDP in 2006, down from about 8.8 percent in 2005. The level of expenditures in 2006 was lower than among the NMS and the EU15, with the outlays in the EU averaging 12.6 percent of GDP (Table 4.1 and Figure 4.2). Within the Western Balkans, however, FYR Macedonia's pension expenditures are similar to the levels in Bosnia and Herzegovina and Montenegro, but higher than in Albania (4.8 percent) which is a much younger country demographically. PAYG pensions are administered by the Pension and Disability Fund (PDF) and are funded through a combination of social security contributions and budgetary transfers. The contribution rate is set at 21.2 percent of gross wages, lower than in all N M S (average about 28 percent), with an annual contribution ceiling of 300 percent of average wages.26 The PDF retains the full contribution for individuals who have not joined the second pillar. For those who have joined a second-pillar fund, the PDF retains the equivalent of 13.78 percent of gross wages and transfers the equivalent of 7.4 percent to the funded pillar. Table 4.1: Revenuesand Expendituresof the Pension and Disability Fund, 2006 In Millions of In Percent REVENUES MKD of GDP Contributionrevenues 19,948 6.7 Budgetary transfers Regularbudgetary obligations 4,670 1.6 Pensiondebt repayments 85 1 0.3 Deficit financing 2,585 0.9 Transitional deficit 1,264 0.4 Employmentagency transfers 1,220 0.4 Other revenues 1,870 0.6 Transferred revenue fkom 2005 233 0.1 Total Revenues 32,642 11.0 EXPENDITURES Pensionbenefit payments 25411 8.5 Disability benefits 1,550 0.5 Healthcontributions 3,582 1.2 Administrativecosts 664 0.2 Total Expenditures 31,207 10.5 Source: Macedonian authorities and staff calculations. 4.9 Contributionsaccountfor about 60 percentof pensionfund revenues,with transfers from the budgetary and the extrabudgetary employment fund accounting for almost all of the remainder. In2006, the transfer from the budget amounted to 3.6 percent of GDP, of which 1.9 percent of GDP was to cover expenditure mandates for pensions to exempt individuals, 0.4 percent for the transitional cost of the introduction of the mandatory pillar (this is the first year in which such a cost has been incurred) and the remainder for the Fund's structural deficit. 26This (annual)contributionceiling compares with 250 percent set in Hungary and Polandand 800 percent inLatvia. 53 Figure 4.2: PensionExpenditure in Macedonia and EUCountries (InpercentofGDP), 2005 16 14 12 10 a 6 4 2 0 "EstimateddatafortheEUaverages,GreeceandBelgium;provisionaldatafortheothercountries. Source: World Bank pensiondatabase. 4.10 The structural deficit of the pension system reflects generous pensions relative to wages and the large number of beneficiaries relative to contributors. At 55 percent in 2006, FYRMacedonia has one of the highest replacement rates in the region, as measured by the ratio of average pensions to average wages (Figure 4.3). Nonetheless, the replacement rate has fallen from 62 percent in 2002, thanks to reform efforts. The system dependency rate with 77 beneficiaries per 100 contributors is also one o f the highest inthe region (Figure 4.4).27 Figure 4.3: Contribution Levels and Replacement Rates in Selected ECA Countries (Inpercent of average wage) 1 IPensioncontribution(%ofwage) Average net pensionlaverage net wage ~ Source: Chlon-Dominczak (2003). ''The ratio ranges fiom 43 pensionersper 100workers inMaltato 76 inItaly. 54 Figure 4.4: System and Demographic DependencyRates in Selected Countries in 2002 " I g m systemdependency rate (left axis) Idemographicdependencyrate(leftaxis) "Systemdependencyisthenumberofpensionersper 100workersanddemographicdependencyisthenumberof peopleolder thanthe statutory retirement age to 100 peoplein working age. Source: Chlon-Dominczak (2003). 4.11 The large number of beneficiaries reflects the relatively young age of exit from the labor marketdespite increases under the parametric reforms introduced earlier inthe decade. At 61.7 years for men and 57 years for women (rising to 64 years and 62 years, respectively), the retirement age at present is lower than in most countries in the broader region. Furthermore, the ratio o f the employed (in the formal sector) to those in working age i s lower than all other countries in the region, almost one-half the 70 percent target under the EU Lisbon strategy and the 68 percent average among the EU15 countries. Formal employment slumped from about 38 percent o fthe labor force in2001 to about 33 percent in 2004 before modestly picking up in 2005 to about 35 percent. Unemployment has remained broadly unchanged in recent years at about 37 percent, however, as rising employment was accompanied by modest increases in labor force participation.*' 4.12 Fallingfertility rates and population ageing, the relatively young current structure of the population notwithstanding, are signals to the authorities to implement further pensionreforms to assure the viability of the pension system. Indeed, the ratio o f those aged 60 years or older per 100 individuals aged 18-59 is projected to double from 24.5 in 2005 to about 50 in 2050. This upcoming wave o f retirees will further burden the pension system in FYR Macedonia. 4.13 The need for change is underpinned by the need to tackle the large implicit pension debt (JPD) in FYR Macedonia, a legacy of existing labor market conditions and policies. The IPD in FYR Macedonia is one o f the largest in the region as a share o f GDP, representing a substantial burden on future generations (Table 4.2). Facing high IPDs, most countries in the region have introduced parametric changes such as gradual increase of retirement ages or reduction o f pension indexation level. In some cases, however, these changes were not sufficient to regain the stability o f pension systems and further reforms were needed. The implementation o f multi-pillar reforms was aimed to further diversify the demographic risk related to ageing 28 Rutkowski, Jan and Stefan0 Scarpetta, 2005, Enhancing Job Opportunities, The World Bank, Washington, D.C. 55 imposed on the pension systems. So far, nine countries in the region of Central and Eastern Europeand CentralAsia havefollowedsuchpath. Table 4.2: Public and Implicit Debt for Selected Countries in Central and Eastern Europe Public Debt IPD by Discount Rate Country 1999/2000 2 percent 4 percent 5 percent As share of GDP Macedonia 41 441 291 244 Poland 43 379 261 220 Hungary 59 300 203 171 Croatia 33 274 201 175 Estonia 7 268 189 163 "ImplicitPensionDebt(IPD)isthepresentdiscountedvalueofallpensionsowedbythegovernment. Source: Holzmannet al. (2004). 4.14 Although FYR Macedonia also started with the implementation of parametric changes within the existing system, the projections available in the following sections indicated the need for a more systemic approach. The Law on Pension and Disability Insuranceenacted in 2000 introduced a set of parametric changes includinga gradual increase in retirement age and terminationof early retirement provisions,a decrease in the replacement rate through change in the pensionformula and a change in pensionindexationmethod, and provided a framework for the introductionofthe multi-pillarpensionsystem. The FundedPillar 4.15 The funded pillarwas launched from the start of2006. Beforethe launch, the Agency for Supervision of Fully Funded PensionFunds (MAPAS) selected pension fund managers. On the basis of an internationaltender carried out in July 2004, the MAPAS granted licenses to two pension companies.29(FYR Macedonia is the only ECA country to limit the number of pension fund managers, motivatingthis restrictionby the small populationsize.) The licenseswere issued in April 2005 for 10 years. Contributionsbegan to be transferred to the funds from the start of 2006. 4.16 The majority of those enrolled in the second-pillar funds joined by the end of 2005, with the pace of inflow slowing in 200630By the end of 2005, about 87,000 people hadjoined (comparedwith an expected numberof about 82,400), risingto about 128,000 by the endof2006. The rapid pace in 2005 could have reflected an aggressive marketing campaign by the fund managers (not accompaniedby a government campaign), and a broad lack oftrust in the existing pension system. Slightly more than one-half of those enrolledjoined voluntarily. In terms of age, the majority of those enrolled (108,626) are younger than 36 years, most likely having workingcareers on average shorter than 15years. 29 There are two authorized administrators: The Nov Penziski Fond-OtvorenPenziski Fond(with founders NLB from Slovenia with 51 percent of the capital and Tutunska Banka from Macedonia with 49 percent) and the KB Prvo drustvo za upravuvanje so penziski fondovi AD Skopje (founders Prva Pokojninska Druzba from Slovenia with 51 percent and Komercijalna Banka from Macedonia with 49 percent.) The founding capitalofbothadministratorswas 1.5 million. 30 Those that should be mandatory members of the second pillar are temporarily allocated to a random pensionfund by MAPAS immediatelyafter their employment is registered. These personshave the right to switchto another pensionfund within 3 months. 56 4.17 The second pillar is funded with contributions equivalent to 7.4 percent of gross wages. Although there is no age limit for the' individuals who join the second pillar, there are disincentives for participation to limit fiscal costs, similarly to Hungary. In the case of FYR Macedonia,the penalty is quite large, reflectingthe specificationthat for those who decidetojoin the multi-pillarscheme, only 5 years of contributionhistorypriorto 2006 will count for the future pension. As a result, under this chapter's baseline scenario (see below), only people with less than 15years ofwork experiencewill benefitfrom switching. 4.18 The law establishes limits on the types of investment and ceilings on investments in particular asset classes to ensure the security of pension fund assets, the diversification of investment risk and the maintenance of adequate liquidity. Under the limits, no more than 30 percent of pension fund assets can be invested in shares issued by FYR Macedoniancompanies, and no more than 40 percent can be invested in bonds and fixed-return instruments issued by FYRMacedoniancompanies. There is a limit of 80 percent set onpublicdebt instrumentsand 60 percent on bank deposits and similar instruments. Pension funds can invest up to 20 percent in the investment funds unitsS3' 4.19 The authorities have limited investments abroad to 20 percent of total assets. This compares with no restrictionsin Estonia (on investments in foreign securities issued in foreign investment by EU and IOSCO member countries), and 30 percent in Hungary and Latvia. The limit is higher than in Bulgaria, Croatia and Poland, however (Figure 4.5). The limit is not bindingat present(see below). Nonetheless, given the small size of the FYRMacedonianequity market, it will be advisable to monitorthe limit inthe future andconsider increasingit. Figure 4.5: Foreign Investment Limits 3 5 % 1 Source; Chlon-Dominczak(2003) 4.20 Reflecting an underdeveloped capital market and the usual home bias for pension funds, especially at the start of reforms, the investment structure currently is rather conservative. About three-fourths of pension fund assets are held in FYR Macedonian government securities, reflecting a shift from implicit to explicit government obligations to contributors, and almost all of the remainder in domestic bank deposits. Only about 7 percent of pension assets are invested in domestic equities, and there are no foreign investments (Figure 4.6). 3'See the Law on MandatoryFully-FundedPensionInsurance for details. 57 NPF KB PN sharesof domertic cash 0,03%<::y~ eivables bonds of domestic bondsof issures domeStic 65 36% lssures 70 37% short term securities ofI --.I domedic domestic issuers issuers 7.43% 13.57% 1 I I Source: MAPAS, Statistical reportno.-2. 4.21 Recognizingthat fees were very high relative to other countries and would erode return on investment, the authorities have moved to cut fees in 2007. There is scope for further prudent reduction in fees over time. Regulations are similar to those in Poland, and only certain fees can be levied on fund participants. The law sets the structure o f fees that funds can charge. The pension funds charge a contribution fee (the fee levied on inflows into individual accounts) o f 7.4 percent starting in July 2007, down from 8.5 percent previously, compared with a limit o f 7 percent in Poland to be reached by 2010.32 The annual asset fee (the fee levied on the total pension assets held in each account) is capped at 0.5 percent of assets, similar to that in Poland(0.54 percent, plus a 0.06 percent success fee). 4.22 The pension law in FYR Macedonia allows for pensions to be paid as either annuities or through programmed withdrawals (Table 4.4). Most countries inthe region with a mandatory funded pillar allow only annuity payments, however, as these provide better security for the system participants. Further, the FYR Macedonian law does not specify whether annuities will be based on unisex or gender-specific life tables. The authorities are advised to address this issue with view to help ultimately improve the predictability' o f instruments to the financial system. 32The maximum fee on contributions in Polandis set to decline from 7 percent (in force through the end of 2010), to 6.125 percent in2011, 5.25 percent in2012,4.37 percentin2013, and 3.5 percentin2014. 58 Table 4.3: Approachesto Pension Industry Structure of Charges Types of charges Limits on Admission Contribution- Asset Performance Country Charge Fee Based Fee Management Fee Fee from Structure Investment Return Hungary J a El J El Kazakhstan J El El El a Poland J El a la El Latvia J El la a' El Croatia J El a J El Bulgaria J El a El El Estonia J El a a El Macedonia J El J El El Note:types of charges: J allowed without limits; 0 allowed with limits; El not allowed. "Untii2003, the fee was paid from the state budget. Source: Chlon-Dominczak (2003). Table 4.4: Annuity Provisions in ECA Countries Products Providers Country Limitations Annuities Other Products Pension Funds Specialised Insurance anies on Life Tables Hungary J J J J Kazakhstan J J J Poland' J J J Latvia J J2 J Croatia J J J Bulgaria J J J Estonia J J J3 J Macedonia4 J J "No regulation exists as yet, basedon the currentgovernmentconcept. */Social security administration. 3/Additional capital requirements. 4'No regulation yet for providers and life tables. Source: Chlon-Dominczak (2003). c. ASSESSINGTHE ADEQUACY OF FUTURE PENSION BENEFITS 4.23 Providing retirees with adequate income in old age is one of the most important goals of the pension systems. Most countries design their pension systems to assure that both the elderly (particularly those with long working careers) and the disabled receive a minimum level of income. This section evaluates the extent to which the FYR Macedonianpensionsystem is set to provide such adequate pension benefits under the parameters currently in place and taking into account ongoing parametric changes. As customary when evaluating multi-pillar systems, the analysis focuses on the minimum income provision, the current and future level of pensionsrelatedto earnings fromthe mandatory scheme, the security of funded-pillarsavings and the system coverage. The discussion in this section is inseparable from the analysis of the sustainabilityofthe PAYGpensionsystem, as presentedinthe followingsection. 59 The Likely PensionOutcome Underthe New Multi-Pillar Scheme 4.24 The analysis in this chapter projects that for participantsin the multi-pillar system the replacement rate is likelyto amount to about 50 percent, as measured by the ratio of the average net pensionto the average net salary. The is compares with a replacement rate o f 62 percent at present and maximum o f 71 percent under the Law on Pension and Disability Insurance for a man with 35 years o f work history. The net replacement rate for the EU on average was 78 percent in 2005 and for the NMS amounted to 68 percent but the highrates reflect to a large extent lack o f efforts by many EU governments to face the acute pressure adverse demographics will have on government budgets. Reforms under way in Poland, one o f the leading pensions reformers in the region, are projected to reduce the net replacement rate to 44 percent by 2050 from 78 percent at present.33 4.25 The projected outcome for the replacement rate is reasonable, given the ongoing parametric reforms o f the PAYG pillar, the efforts to slow the pace o f pension increase by giving more weight to inflation rather than wage increases, and the ongoing rise in the retirement age. The outcome, moreover, reflects the rather cautious assumptions under the chapter's baseline scenario, with real GDP growth projected at 3 percent. A stronger real GDP growthof 4 percent ayear will boost the replacement rateto 63 percent. 4.26 The value of an annuity provided under the new system will depend on several factors other than the fee structure. Firstly, the authorities have not yet decided whether life tables used for annuity calculation will be gender-neutral on gender-specific. The latter would mean relatively lower annuities for men and higher for women. Secondly, the ratio o f fully- funded pensions to the average wage will depend on the difference between the rate o f return and wage growth. The higher the difference, the higher will be the expected level o f annuity. To help understand the interaction o f factors, Table 4.5 presents annuity projections by gender and years o f savings under three different fee regimes and under three scenarios: one based on unisex life expectancies, another under different calculations o f benefits for men and women, and the third with higher rates o f returnon investment. Analyzing theSwitch to Multi-Pillar Pension System 4.27 Forthose switchingto the multi-pillarsystem,the followingassumptions are used: e The accrual rate for years o f work prior to 2003 is 2.33 percent for each year for men (up to 11.65 percent cumulative) and 2.6 percent for women (up to 13 percent cumulative). This assumption is an extrapolation o fthe one inthe PDF law. e The accrual rate after and including 2003 is 0.75 percent for men and 0.86 percent for women. The assumption is extrapolated from the PDF law. 4.28 Pension levels for those who did not enroll in the fully-funded pillar and had more than 15 years o f history are set according to the PDF law. Accrual rates for those only in the PAYG pillar with fewer than 15 years o f contribution history is set at 2.33 percent for men and 33EuropeanCommission, 2006, "Current and ProspectiveTheoretical PensionReplacementRates," Report by the Indicators Sub-Groupofthe SocialProtection Committee, Brussels. 60 2.6 percent for women for every year before 2003 and 1.8 percent for men and 2.05 percent for women for 2003 and thereafter. It should be noted that these transition rules not only limit pension rights accrued before 2003 for those who switch, but also reduce pension rights accrued for 2003-2005, as these are credited with lower accrual levels even if no contributions were paid inthat periodto the funded pillar. Table 4.5: Annuity Projections: Sensitivity to Fee Size, Number of Years of Saving and Gender (In percent of average wage) Standard Assumptions 11 Return on Investment = 8% Unisex Life Expectancies Gender Specific Life Expectancies Unisex Life Expectancies No of Years of Savings Men, Retirement Age 64 years Fees Fees Fees Current None Reduced Current None Reduced Current None Reduced 10 5.0 5.6 5.3 5.37 6.0 5.7 5.5 6.1 5.8 15 7.8 8.9 8.3 8.34 9.5 8.9 9.0 10.2 9.5 20 10.7 12.4 11.5 11.52 13.4 12.3 13.1 15.2 14.0 25 13.9 16.4 14.9 14.93 17.6 16.1 17.9 21.1 19.3 30 17.3 20.7 18.7 18.57 22.2 20.1 23.6 28.4 25.6 35 20.9 25.5 22.7 22.48 27.4 24.4 39.3 37.2 33.0 40 24.8 30.7 27.1 26.66 33.0 29.2 38.2 47.9 41.9 Women, Retirement Age 62 years Fees Fees Fees ~ ~~ ~~ ~~~ ~ Current None Reduced Current None Reduced Current None Reduced 10 4.5 5.1 4.8 4.3 4.8 4.5 5.0 5.6 5.3 15 7.1 8.0 7.5 6.6 7.5 7.0 8.1 9.3 8.7 20 9.7 11.3 10.4 9.11 10.6 9.8 11.9 13.8 12.7 25 12.6 14.9 13.6 11.8 13.9 12.7 16.3 19.2 17.5 30 15.7 18.8 17.0 14.7 17.6 15.9 21.4 25.8 23.2 35 19.0 23.1 20.7 17.8 21.6 19.3 27.5 33.8 30.0 40 22.5 27.9 24.6 21.1 26.1 23.1 34.7 43.5 38.1 "Thestandardassumptionsusedinthe first sixcolumnsare:agrossrateofreturnonassetsof4 percent, annualwage growth of 2 percent and current mortality tables. The last three columns assume a return on assets of 8 percent. Annuity values are calculatedby dividing pensionfunds savings and life expectancies, excludingfees. The current fees are as described in the text. The reduced fees are: 0.4 percenton assets and 4 percenton contributions. Source: Authors' calculations. 4.29 In addition to these assumptions, the simulations assume: (i)that the retirement age for men (women) i s increased as planned under the law to 64 (62) years and unisex life tables are used for annuity calculations; (ii)pension fund fees remain unchanged and (iii)the annual gross return on investment i s either 4 percent or 6 percent (alternative calculations) and average wages grow at 2 percent a year. The results o f projections are presented in Figure 4.7. 61 4.30 These simulations suggest the following policy implications: Efforts should be made to reduce the costs o f the funded pillar to help increase pension levels. Limits imposed on the credit for the work history before 2003 have very strong impact on the expected pension level. For those with 10-15 years of contributions before 2003, the total replacement rate falls to below 40 percent o f wage. This can lead to higher poverty among this part o f the population and increase pressure on the government to boost the minimumpension promise. 4.31 The simulations in this chapter suggest that under a realistic scenario, average pensions will average about 50 percent of wages. Such a likely outcome is reasonable and a testament to the effect o f the reforms the authorities are implementing. The difference in replacement rates with the EU should be interpreted with care, as many EU countries have yet to begin tackling the acute demographic problems they face. Moreover, stronger real GDP growth than recently observed should enable the decline in the replacement rate to be smaller than projected under the baseline scenario. 4.32 The projected pension levels for FYR Macedonia are broadly similar to those likely to prevailin most EUand OECD member ~tates.3~ demonstrated inFigure 4.8, differences As inthe design o f the pension systems and promises offered to individuals differ quite significantly. Further, the evolution and level o f replacement rates depend on the design of the pension system and the entire social protection system in each o f countries. Additionally, depending on the design o fthe pension system, replacement rates reflect assumptions about future economic, social and demographic development. These considerations apply to FYR Macedoniaas well. 4.33 There are also significantvariations between countries in terms of the redistribution effect of pensions. While some countries have relatively flat pension schemes, others offer benefits that are closely related to individuals wage histories or contribution payments. The Macedonian pension system is less redistributive, although some redistribution between men and women i s included through differentiationo f accrual rates and retirement ages. Impact of Fees on Pension Savings and ProjectedAnnuity Values 4.34 As discussed earlier, the level of fees in FYR Macedonia is higher than in most countries in the region and above levels expected before the start of the pension reform. High fees will tend to discourage participation for people to join, and will erode the return on assets for those who participate. The latter effect for those who earn the average wage is presented in Figure4.9. As an illustrativecalculation, the fees charged in FYR Macedonia would reduce the size o f the pension assets o f the average earner by 24 percent relative to a scenario with no fees charged. This compares with 14.4 percent inPoland, 16.5 percent inKazakhstan and to 26.4 percent in Croatia. The FYR Macedonian pension scheme, as a result, i s relatively expensive, even more so given the investment restrictions and the current practice by the pension funds to keep most assets in government bonds and bank deposits. 34See OECD (2005) for the OECDcountriesand SPC (2006) for EUmember states. 62 Figure 4.7: Multi-Pillar vs. Mono-Pillar Pensions a. Men, 35 years of work history 80% m e5 60%. .. r 40% 0 20% 5 7 9 11 13 15 17 19 21 23 25 years of work prior to 2003 -Multi-pillarpension(6% return) b. Women, 35 years of work history 6 Whi I , , I , , , I I I I I I I , I I I I , I 5 7 9 11 13 15 17 19 21 23 25 - years of work prior to 2003 - Multi-pillar pension (4% return) -Multiipillar pension (6% return) Mono-pillarpension c. Men, 40 years of work history 80% ---"e 3 0 60% 1 c 40% 0 O % i I , , I 1 , I , I 8 I 8 8 8 I 8 I 8 5 7 9 11 13 15 17 19 21 23 25 - years of work prior to 2003 -mitipillar -Wltiipillarpensionpension 4% return Mono-pillar pension 16% return] Source: Authors' calculations. 63 Figure 4.8: Gross Replacement Rates for New Entrants, Mandatory PensionPrograms, Men Average Wage Earners (Inpercent) Source: OECD (2005). Table 4.6: Minimum Pensions Size and Financing Country Minimum Pension Qualifying Structure of the Guarantee Level Period(Women/men) (In percent of average wage in 2002) Hungary 17 20120 Financed from the state budget Kazakhstan 20 20125 Financed from the state budget, topping-up pensions receivedfrom the funded pillar Poland 30 ' 20/25 Financed from the state budget, topping-up pensions receivedfrom the first and second pillars. Latvia 17 10/10 Inthe first tier, financed from social security contributions. Croatia 10.52 15115 Inthe first tier, financed from social security contributions. Bulgaria3 19 15115 Inthe first tier, financed from social security contributions. Estonia 13 515 In the first tier, financed from the state budget - Macedonia 41 Over 30135 years 38 Over 20125 years Financed from social security 35 Over 15 years contributions, on top of pensions received from the first and second pillars. I/Relative to the average wage net of social security taxes. For the minimum qualifying period, with increases in case of longer service. 3'People who do not fill the eligibility conditions receive social pensions from the state budget (paid for those older than 70 years of age, basedon meanstest). Source: Chlon-Dominczak (2003). 64 Figure 4.9: Impact of Fees on Asset Level for Average Wage Earner 3,000,000 2,500,000 2,000,000 1,000,000 500,000 0 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 Years of savings I'Under the assumptions o f 8.5 percent fee on contributions, 0.6 percent fee on assets, average annual wage growth of 2 percent and gross annual return on investments o f 4 percent. Source: Authors' calculations. 4.35 The key conclusion from the simulations is that the authorities' targeted replacement rate o f 40-50 percent from the second pillar i s unlikely under the current fee structure and reasonable assumptions on rates o f return and wage growth. The replacement rate is reached only under the assumption o f rate o f return on investment substantially higher than wage growth and a substantial reduction in fees for persons with long participation periods. Such higher returns are unlikely under the current investment practice o f the fund managers, however, with the bulk o f the assets placed in (safer and lower-yielding) government bonds and deposits. The analysis demonstrates that a reduction in the fees can increase the replacement rate by 2-3 percentage points. Secondly, the annuity values for women are likely to be lower than those for men due to the lower retirement ages if gender-specific tables are used, with the gap doubling compared to results based on usingthe unisex life tables. MinimumPensionGuarantee 4.36 Countries with multi-pillar systems structure the first pillar to provide a minimum income to pensioners, typically by establishing an explicit minimum pension guarantee (a role that is separate from the system's role o f income replacement) or through a state guarantee for both mandatory pension pillars. The latter i s also the path chosen by FYR Macedonia, Hungary and Kazakhstan. 4.37 The level o f the minimum pension provision relative to the average wage is not only higher in FYR Macedonia than in Hungary and Kazakhstan (to compare similar systems), but also exceeds levels observed in all other countries in the region that have enacted a guarantee (Table 4.6). The guarantee in FYR Macedonia, unlike the other countries, i s differentiated by length o f contribution period, which can be seen as an incentive for longer working lives. On average, the qualifying period is similar to other countries. 65 Guarantees in the FundedPillar 4.38 The guarantee in the funded pillar in FYR Macedoniais relatively simple. Chapter 15 o f the Law on Mandatory Fully Funded Pension Insurance specifies that the state budget is responsible for covering 80 percent o f the depletion o f a pension fund caused by act that appears to be criminal, involves fraud or theft. Depletion due to breaches o f the investment principles is not covered. 4.39 The law does not providea guarantee of a minimum rate of return, similarly to the practice in Hungary,Poland or Croatia. Similarly to Croatia, the law in FYRMacedonia does not require the pension funds to keep any reserves on a company or industry level. In other countries such reserve funds are kept to finance the guarantees related to minimum rate o f return or covering fraud or misuse o f funds. Clearly, lack o f reserves puts the budget at a greater risk, even if this risk is relatively small at present due to the conservative investment mix o f the pension funds and the small size o f the assets. Nonetheless, the risk is likely to increase as the funds move into riskierinstruments and their size increases. D. FINANCIALSUSTAINABILITY OFTHEPAYGPILLAR 4.40 This section analyzes the PAYG pillar and concludes that under realistic assumptions its fiscal sustainability is likelyto be assured, provided the authoritiescontinue implementing the reforms already in progress and continue advancing the structural reforms to generate moderate growth in employment and output. The key underlying assumptions include the continued gradual increase in the retirement age and the implementation o f the indexation formula that increases substantially the share o f inflation and reduces the share o f wage growth. While sustainability is likely, the authorities are advised to consider it in the context o f the results o f the previous section, with the most likely pension benefit set to average 45-55 percent o fthe average wage. 4.41 Population ageing, a common concern in most European countries, has put the financialsustainability of pensionsystems at the heart of the debate about pension policies. As discussed above, the PAYG pension scheme in FYR Macedonia already requires substantial budgetary transfers, with the transfer from the budget to the PDF amounting to about 3.6 percent o f GDP in 2006 (with 2.2 percent covering pensions for exempt individuals, 0.8 percent covering the structural deficit o f the PAYG fund, and 0.6 percent to offset PDF's transfer to the second- pillar funds). 4.42 Going forward, during 2006-2015, the PAYG deficit that needs to be covered from general revenueswill decline despite the likely pickup in transfers to cover the transitional costs for the introductionof the second pillar. The decline is projected to reflect the absence starting in 2006 o f payments o f pension arrears; the increase in the retirement age as mandated under the law; and the decline in average pensions relative to wages because o f the change in the pension indexation formula. 4.43 The projections developed in this section are based on the following demographic assumptions. The total fertility rate is quite low at present and is assumed that it will rise only modestly to 2.1, at which level the population should be stabilize (Table 4.7). In addition, the expected lifetimes at birthare projected to rise at a rate o f about 1year per decade. 66 Table 4.7: Demographic Assumptions 2005 2010 2020 2030 2040 2050 Lifetime at birth: men 69.8 70.8 71.9 72.4 74.0 75.0 Lifetime at birth: women 74.1 75.3 76.5 77.1 78.8 80.0 Total fertility rate 1.50 1.60 1.75 1.90 2.10 2.10 Source: Wiese (2006). 4.44 Simulations are carried under three economic scenarios (baseline, optimistic and pessimistic) but the assumptions for all scenarios, including the optimistic, are rather cautious. The key difference among the scenarios is the pace o f real GDP growth, with the assumed real GDP growth under the optimistic scenario set at 4 percent, compared with an outcome o f about 3.7 percent during 2003-2006 (Table 4.8). For each o f the scenarios, real GDP growth is assumed equal to the growth in the real wage bill (employment growth plus growth in average real wages). Further, the rate o f contributor growth i s assumed equal to the rate o f (formal) employment growth (1 percent a year). A realistic, albeit restrictive assumption, is the proposition that unemployment cannot fall below 10 percent. These assumptions together imply that when the unemployment rate falls to 10 percent, real GDP growth slows along with employment growth. Although a more ambitious structural reform agenda than the one underpinning the results o f this chapter cannot be excluded (and, along with it, a lower deficit o f the pension fund over the longer term and higher average pensions), the conservative assumptions inthis chapter result inuseful scenarios for understandingthe financial sustainability o fthe PDF. 4.45 The parametric assumptions used in the chapter are also conservative. They are: (i) the transfer from the PDF to the Health Fundis held constant at its current level relative to GDP; (ii)thetransferfromtheEmploymentFundtothePDFisheldconstantatitscurrentlevelrelative to GDP; (iii)disability prevalence rates are held constant at their current levels; (iv) through 2008, pensions are assumed to be indexed to inflation (40 percent) and average nominal wages (20 percent),while the indexationafterwardswould be 80 percent to inflation and 20 percent to wages; and (v) the average retirement age is assumed to increase from its current level o f 62 years to 64 years for men by 2015, and from 57 years to 62 years for women over the same period.35 Table 4.8: Economic Assumptions (Inpercent) Pessimistic Base Case Optimistic Employment Growth 0.7 1.o 1.4 RealWage Growth 1.3 2.0 2.6 RealGDP Growth 2.0 3.0 4.0 Source: Wiese (2006). 4.46 Under all scenarios, the deficit of the PDF is expected to decline by 2015, reflecting the increased retirement age, revised pension indexation and other changes to the PAYG scheme (Figure 4.10). After 2015, the deficit is projected to increase before beginningto decline again after 2025. The scenarios are similar in the direction o f change o f revenues, expenditures and required budgetary transfers, but differ inthe magnitudes o f these variables. ''It i s possible that the effective average retirement ages will rise faster, provided no new exemptions are introduced. 67 Figure 4.10: PDFRevenuesMinus Expenditures(Excluding BudgetaryTransfers) (Inpercent ofGDP, underthe three different sets ofeconomic assumptions) 0.0% - - - - _____-- - -.---- - -.- 2006 2016 2026 2036 2046 2066 2066 1 *pessimistic --i)--.base case +optimistic Source: Wiese (2006). 4.47 The impact of changes in the labor market and pension system changes on the balance of PDF is presented in Figure 4.11. PDF revenues are projected to decline by about 2 percent o f GDP during 2005-2055, largely because o f the enrollment o f new labor force entrants inthe fully-funded pillar. Under the baseline scenario, PDF expenditures are projected to decline by 4 percent o f GDP during 2005-2050, falling to levels well-below current and future projected levels in the EU. Expenditures by the FYR Macedonian PAYG are projected to undergo three distinct periods. These are: (i)a decline through 2015 because o f PAYG parametric reforms and the end o f pension arrear payments in 2005; (ii)little change during 2015-2025; and (iii)a further drop after 2025, as the second pillar reaches maturity and thus reduces the financing burden placed upon the PDF. These projections assume a determined implementation o f all parametric changes. 4.48 The deficit of the PDF is assumed to be covered fully by transfers from the budget. Transfers from the budget are projected to ease by 2015 before rising steadily again to 3 percent of GDP by 2025. Transfers thereafter are projected to decline to about 1percent of GDP a year by 2055 (Figure 4.12). These results are encouraging given the demographic projections. The dynamics reflects three components. Firstly, the transitional costs related to the introduction o f the fully-funded pillar are set to rise through 2025, but fall to nil by 2055, as the second pillar matures. Secondly, as the number o f exempt individuals declines, as new entry to some exempt groups i s terminated, the budgetary obligation is projected to fall steadily to about 1 percent o f GDP by 2055. Thirdly, the underlying structural deficit o f the PDF is projected to disappear by 2015, thanks to the assumed implementation o f P A Y G parametric reforms, including the increase in the retirement age, but is projected to reappear during 2040-2055, as population ageing worsens the dependency ratio. 68 Figure 4.11: Baseline Scenario: PDF Revenuesand Expenditures (Inpercent ofGDP) 12.0% 10.0% 8.0% 6.0% 4.0% I 0.0% 1 1 2005 2016 2025 2035 2045 2055 2065 *expenditures u- revenues -0-expenditures -revenues Source: Wiese (2006). Figure 4.12: Decompositionof the Required Budgetary Transfer (Inpercent ofGDP) 3.5% 3.0% 2.5% __ _ _ _ _ ~ I 2.0% -~ -~ I.5% I.O% 0.5% 0.0% 2005 2015 2025 2035 2045 2055 2065 IB Budget Obligations o PDF Deflcit ITransitionCosts Source; Wiese (2006). 4.49 To summarize, the pension reform implemented by FYR Macedonia is projectedto lead to a gradual improvementof the financial outlookfor the pension system. Nonetheless, budget transfers will be necessary throughout the entire projection period. Under the baseline scenario, budgetary transfers are set to fall by 2 percent of GDP from 2005 to 2055, an encouragingresultgiventhe demographic projections. 4.50 These results can be compared with projections for the EU countries prepared in 2006 by the Working Group on Ageing Populations (within the Economic Policy Committee of the European Council). According to the projectionsabout the EU, populationageing would boost pension outlays by 2.2 percent of GDP on average during 2004-2050, with Cyprus experiencingthe largest increase (more than 10 percent of GDP) and Poland the largest decline (about 6 percent of GDP). Such a diversity of outcomes reflects several factors. Firstly, the design of the pension system matters. Some of the countries projectthat PAYG expenditure will decrease due to the introductionof mandatory funded components (Hungary, Poland, Latvia and 69 Estonia). Additionally, several countries have systems that link future pension benefits to changes in life expectancy (this applies most notably to the countries that have notional defined contribution systems, such as Italy, Sweden, Latvia and Poland). As a result, the pension projections automatically cover the pension benefit adjustment, which in other countries could take a form o f legal change that is difficult to predict. Secondly, the underlying assumptions also matter. The projections for the EU assume a relatively high wage growth. As a result, countries with pensions indexed to inflation (such as Poland, Estonia or Latvia) show a relatively large drop inpension expenditure through 2015. Such an outcome is also likely inFYRMacedonia. 4.51 Long-term projections are a useful tool for monitoringtrends in pension spending and the possible magnitude of changes. Because parametric changes to the system take hold only with a long lag, while macroeconomic developments are more dynamic, such projections needto be prepared regularly and used to improve public policy. E. CONCLUSIONSANDRECOMMENDATIONS 4.52 The FYR Macedonianpensionsystem is broadlywell suited to handle demographic and structural changes, but additional steps in reforming the PAYG pillar and strengthening the funded pillar should be consideredto ensure the long-termsustainability of the system and the adequacy of pension benefits. The analysis in this chapter assumes that pension contribution rates will not be increased, in line with government intentions. This is particularly important given the current labor market situation in FYR Macedonia- any increase o f the tax wedge is likely to have a negative impact on already very low (formal) labor market participation. Reducing social security contributions should be considered to help ease the burden, buildingon recent progress.36 4.53 To help create room for a reduction in the contribution rate and help ensure the long term soundness of the pension system, the authorities are invited to consider the followingmeasures: 0 Gradually unify the retirement age for men and women at 64 years, the targeted retirement age for men, and consider an increase to 65 years for both men and women. 0 Improve the disability assessment system, revise the eligibility criteria for disability pensionsand tighten, overall,disabilitypensionpolicies. Additional measures to be considered include the implementation of stricter assessment procedures for those who apply for disability pensions and the payment o f disability pensions only until retirement age, substitute thereafter with an old-age pension. 0 Implement efforts to scale back popular expectations that pension benefits will average 75-85 percent of wages. The likely outcome for the average pension as a share o f the average wage projected in this chapter is similar to the outcome in most EU and OECD countries and should form a reasonable benchmark for the authorities. 4.54 The pensionsystem, especially the funded component, needs to be transparent and the participantswell informed. Although a short time has elapsed since the funded pillar was introduced, several issues that require the authorities' attention were flagged in the text. The 36IMF, 2006, FYR of Macedonia: SelectedIssues, WashingtonDC. 70 legislative base for the fully funded pillar needs to be completed to set clear rules for the payment o f benefits. Additional recommendations related to the discussion in this chapter are listed below: Consider ways to reduce the charges on fully-funded accounts and fees paid by the second pillar funds. One method to consider would be to introduce a cap on contribution fees in line with regional best practices. 0 Consider consolidatingthe supervisionof the non-bankingfinancialsector. 0 Engage in a continuous government information campaign to help people understandtheir choicesand consequences inthe form of lost pension rights. 0 Consider allowing those who have enrolled in the fully funded pillar, but have 10 years or more of work experience prior to 2003, to exit back in the mono-pillar scheme. 4.55 The low coverage at present will also have important implications for the future. Because today's contributors have, in general, higher incomes than non-contributors, the transfers from the budget to fund the pension deficit in the future will be more regressive, Le., benefit disproportionately the better-off. Moreover, as the number of people receiving pensions decreases in response to today's low coverage, the government may come under pressure to provide additional social assistance to those that do not receive pensions to ensure that they do not fall below the poverty line. These implications are important for the government to ponder as they advance reform implementation. 71 5. TRANSPORT A. INTRODUCTION 5.1 FYR Macedonia has under-spent on maintenance of assets in the transport sector for many years. In the road sector the deterioration in the quality o f assets, especially local roads, has created road conditions worse than all other countries in Southeast Europe except Albania. The deterioration o f the country's roads is an obstacle for growth and exports, and increases vehicle amortization while causing wasting time and fuel. In addition these deteriorating road conditions represent a serious fiscal risk which will grow further if not addressed quickly. However, significant increases in expenditure in the next few years would be required to address the maintenance backlog. A number o f institutional reforms could ensure that through a new Road Law presently under discussion. . resources are spent more efficiently. In the road sector, these reforms will need to be defined 5.2 The financial condition of FYR Macedonian Railways (MZ) has been abysmal. While direct state aid has been very low, the Government has subsidized investments in the unfinished Corridor VIII, and whilst the completion o f this corridor is recognized as a strategic priority, current and projected demand suggests that it should be a medium to longer term objective. The Government has also absorbed large accumulated losses and financed MZ debt service to external creditors. MZ has been implementing reforms to increase labor productivity through a program o f staff retrenchment and the company was broken up in mid-2007 into two state-owned companies; one focused on infrastructure and the other focused on transport. Nevertheless, continued reforms are urgently needed to improve operating efficiency and avoid future losses. 5.3 The chapter recommends that the Government address the maintenance backlog and increase spending on normal maintenance to the extent possible within the fiscal envelop. Addressing the full backlog while boosting standard maintenance outlays is estimated to require an increase in road expenditure by as much as 2 percent o f GDP a year for five years. The estimated cost o f addressing the backlog could be reduced if the classification o f roads were updated to reflect low traffic volumes on some roads, as lower maintenance expenditures are required on lower volume roads. The chapter also recommends a number o f institutionalchanges to improve the efficiency o f spending. Capacities to identify, appraise, prioritize, and effectively monitor the execution o f maintenance and investment spending should be strengthened. In railways, the chapter calls for continued reform of the MZ to improve operating efficiency and avoid future losses. 5.4 The rest of the chapter is organized as follows. Section B outlines the characteristics o f the road sector. Section C reviews public expenditure on roads. Section D examines the main institutions in the road sector. Section E reviews the future financing requirement for roads. Section F examines the railway sector and section G presents the chapters recommendations. B. CHARACTEFUSTICS OFTHE ROAD SECTOR 5.5 FYR Macedonia has a fairly well developed road network but it suffers from Table 5.1: Macedonia, Density of Road heterogeneous quality. In general, the road Infrastructure (latest observation network in FYR Macedonia appears adequate in available, 1997-2003) terms o f scope, as it compares favorably with (km/1000 sq (km/1,000 respect to selected regional peers both in terms o f km) People) kilometers per 1000 square km and road kilometers Hungary 1,733 15.7 per 1000 people. Nonetheless, FYR Macedonia falls Czech well short o f the levels o f coverage in N M S and Republic 1,646 12.5 other upper middle income and OECD countries. (Table 5.1) Croatia 506 6.4 Bosniaand 5.6 The lengtho f the road network totals 13,124 Herzegovina 427 5.6 km. Roads are classified as arterial, regional and Serbia & local. The national road network comprises 4,707 Montenegro 494 4.8 km, 906 km of which are arterial road and 3,801 km Kosovo 783 4.2 are regional roads. The local roads network totals Albania 657 3.5 8,4 17 km, about 48 percent o f which are paved. The Highincome: OECD 1,340 17.3 arterial network, together with one third o f the Uppermiddle regional road network, carries about 80 percent o f income 1,076 9.2 the traffic. Other regional roads often carry less than Europe& 2,000 vehicles per day, and a review o f the CentralAsia 580 8.6 functional categorization o f some roads would Source: WDI and IEF databases. appear warranted. 5.7 FYR Macedonia has good north-south connections with Serbia and Greece, but poor east-west connectionswith Albania and Bulgaria. The 176 km Pan European corridor X (E-75) is the spine o f the system. Its development to four lanes is two-third complete. The Pan European corridor VI11 goes East-West linking the Bulgarian border to the Albanian border via Skopje with a total distance o f 304 km including 91 km at motonvay standards and a 25 km bypass around Skopje under construction. 5.8 While the proportion of arterial and regional roads in good condition compares favorably with countriesin the region, the conditionof the local road network is very poor. About 70 percent o f FYR Macedonia's local roads are estimated to be in poor condition. Arterial and regional roads have been better maintained than local roads but still, 33 percent o f arterial and regional roads are in poor condition. As a result, the quality o f FYR Macedonia's entire network lags well behind regional comparators, with the exception o f Albania (Figure 5.1). Only 16 percent o f the entire network is in good condition, a further 27 percent i s in fair condition, whilst the remaining 57 percent o fthe network is inpoor or very poor condition. This assessment i s corroborated by firms' deteriorating perceptions about the quality o f transport infrastructure (Figure 5-2). 5.9 Both road and rail traffic growth has been considerablerecently.Between 2002 and 2005, international railway freight traffic increased by an average o f 17 percent per year, or 6 times the average real GDP growth over the same periods3' Total road traffic has grown even 37 Internationaltraffic represents 99 percent ofthe total traffic carried by the railways. 74 faster at an average o f 27 percent per year, nearly 10 times the average o f GDP growth. A similar trend can be observed for passenger transport. As a result, roads handle now about 91.3 percent o f freight traffic and 92 percent o f passenger traffic, compared to 9.6 percent and 8 percent for the railways, respectively. Road transport demand tends to concentrate on several main links, notably around Skopje and Ohrid, while low traffic flows were found on most o f the network.38 Despite the recent increase in traffic, only a few sections around major cities presented ADT reaching 10,000 vehicles per day. In2002, traffic on the main corridors remained below 8,000 ADT.39Unfortunately, traffic trends by road category are difficult to assess since such traffic counts were discontinued after 2002. Figure 5.1: Regional Comparison of Road Network Figure 5.2: Percent of Firms Stating that Quality Tranmort is a Problem for Doing Business 80% 18% 70% 16% 60% 14% 50% oCroatia 12% 40% 10% 30% o Bosnia 8% 20% oAlbank 6% 10% 4% 0% 2% 0% Good Fair Poor Mac SEE ECA Source: BCEOM, 2005; FNRR,2005, World Bank PEIRs Source: BEEPS. 2005. for Kosovo and Bosnia and Herzegovina. c. PUBLIC EXPENDITURE ROADS ON 5.10 Since 1998,the primary focus of public expenditures hasbeenon capacity expansion and upgradingof regional and arterial roadswhile maintenanceneeds have been neglected. Since 1998, the authorities have concentrated more than 40 percent o f road expenditures on new investments and reconstruction, whereas operating and rehabilitation maintenance averaged around 35 percent o f total expenditures (Figure 5.3). Over the same period, FYR Macedonia allocated 85 percent o f road expenditures, or 1.3 percent o f GDP, on regional and arterial roads, half o f which was used for new investments and reconstruction. This concentration on regional and arterial roads is high compared to other countries in the region. Such expenditures have averaged about 1 percent o f GDP in each o f Albania, Romania and Bulgaria in recent years. In terms o f expenditures on maintenance, the amount spent on arterial and regional roads (routine, scheduled periodic and winter maintenance) in 2005 was 14.3 million, compared to an estimated need o f20.6 million (Error! Reference source not found.), a shortfall o f about 30 percent. On local roads, actual expenditures on maintenance in 2005 were 8.3 million, compared to an estimated need o f15 million, a massive shortfall o f nearly 45 percent. 3sSource:BCEOM2006 FYRMroadreport. 39JBIC study, 2005. 75 5.11 This large under-fundingof maintenance has contributed to a deterioration of the local road network at a high cost for road users. The deterioration o f these assets represents a significant contingent liability. Paved local roads currently in a poor and fair state represent about half o f the total network maintenance backlog. Low-trafficked regional roads and main regional roads account for one-sixth o f the backlog each. Without adequate maintenance, roads deteriorate at an increasing rate until reconstruction is necessary. The cost o f reconstruction is considerably greater than any short term saving in maintenance expenditure which led to such imbalances. Heggie and Vickers (1998) report that rehabilitating a paved road is 3 times more expensive than maintaining it, in current terms, and around 35 percent more in net present value terms. In addition, failing to maintain a paved road is estimated to increase user costs by a factor o f three, interms o f additionaltime, fuel, and vehicular wear and tear. Figure 5.3: Recurrent and Capital Expenditure, 1998-2005 (In percent of GDP) 2.0% , 1.8% - 1.6% 7 1.4% - 1.0% - 0.8% - 0.6% - -. - _ # - - - - - - I - - 0.4% - 0.2% 0.0% 1998 1999 -- 2000 2001 2002 2003 2004 2005 Operating and rehabilitationmaintenance -Capital and financing expenditures e A d m i n i s t m t i o n and tolling costs -Total Notes: Feasibility study costs are countedin "Administration"in 2002 and 2003. Maintenance includesoperating, rehabilitationworks and localroadmaintenance. Source: FNRR (2002,2006), Country ReportFYROMacedonia(NE1Transport, 1999),WE3 assumptions. D. THEMAININSTITUTIONSINTHEROAD SECTOR 5.12 There are four main institutions in the road sector, The Public Road Act defines the institutional responsibilities for road management in FYR Macedonia. The four main institutions involved are the Ministry of Transport and Communication (MOTC), the Fund for National and Regional Roads (FNRR), Makedonija Pat (the public enterprise) and the municipalities. The Ministry o f Transport and Communication (MOTC) is responsible for: (i)developing and executing FYR Macedonia's transport policy; (ii)setting technical regulations and standards for roads, road traffic and vehicles; (iii)authorizing reconstruction and construction projects o f arterial and regional roads; (iv) inspecting roads; (v) approving the annual road budget; and (vi) supervising FNRR. FNRR, a semi-autonomous road fund, is in charge o f planning, financing, constructing and maintaining arterial and regional roads, while local and urban roads are the 76 responsibility o f municipalities. Finally, Makedonija Pat is a public sector enterprise which holds a legal monopoly on routine and winter maintenance for arterial and regional roads as well as toll operation. It receives 20 percent o f the total budget o f FNRR via direct contracting through annual plans. 5.13 The planning and budgetary process in the road sector remains weak, with a number of deficiencies that undermine the efficiency of expenditures within the sector. These deficiencies include: (i)the absence o f a formal sector policy and strategy, consistent with the needs and the fiscal envelope available to the Government; (ii)the absence o f a process that contributes to the formulation and revision o f a sound medium-term sector strategy by the MOTC; (iii) weaknesses in project identification and assessment both within MOTC and FNRR, (iv) poor budgetary control; and (v) limitations in the management o f the assets in the sector by FNRRand Makedonija Pat. 5.14 The delineation of responsibilities in the road sector leads to suboptimal road maintenance planning and sector monitoring. Firstly, the monopoly position o f Makedonija Pat in maintenance provision is likely to have a significant negative impact on allocative efficiency in the sector and should be ended. Secondly, although Makedonija Pat is the maintenance provider, it also holds (although it is not fully operational) a Road Financial Management Program, which would be more appropriately located at the network manager, in this case FNRR. Thirdly, the collection, recordingand analysis o f data on the condition o f roads are marred by a lack o f coordination and duplication o f efforts in MOTC, FNRR and Makedonija Pat. Fourthly, Makedonija Pat is responsible for enforcing axle load legislation on behalf o f the Ministry o f the Interior. This responsibility should rest with the network manager, FNRR, rather than the maintenance contractor. Finally, FNRR i s responsible for revenue allocation to the municipalities to maintain local roads, but the allocation has not met the legal funding ratio from excise taxes and vehicle registration taxes (23 percent in 2005 versus a legal requirement o f 35 percent). 5.15 The limited use of formal techniques of economic appraisalin project identification and prioritizationis a key weakness in the budgetary process. Emphasis has been placed on the development o f the Helsinki corridors VI11 and X without a clear action plan for phasing o f investments. Similarly, programming o f periodic maintenance works lack systematic and objective criteria in the absence o f updated data on road traffic. Not only should the PMS be managed by FNRR but Makedonija Pat does not keep it updated, nor use it to prepare a formal schedule o f maintenance works. The actual maintenance program is prepared by regional branches, and consecutively reviewed by Makedonija Pat, and then submitted for approval to FNRR.FNRRis itself understaffed and lacks training inmaintenance managementactivities. 5.16 Makedonija Pat should be more effectively regulated. There are concerns that Makedonija Pat overcharges tolling services and that more than 20 percent o f toll revenue is unaccounted for. Tolls are not related to the costs o f the infrastructure as required by the acquis. The government has enacted a decision according to which the collected toll shall reflect the utilization o f the road infrastructure rather than the price o f hela4' Makedonija Pat appears to be overstaffed. Over the last few years, more than a third o f Makadonija Pat's revenue (40 percent o f operating costs) was allocated to staff expenses. The high labor-intensity o f Makedonija Pat is also explained by a low level o f mechanization o f its maintenance practices and the poor condition o f its equipment. About one forth is life-expired and presents low level o f productivity. 40Official Gazette o fthe Republic of Macedonia No. OY2007. 77 E. FUTURE FINANCINGREQUIREMENTS 5.17 Over time, demand for road transport will continue to grow at a fairly rapid pace. Total road traffic has grown by an average o f 27 percent per year in recent years, about 10 times the average o f GDP growth. The projected increase in GDP per capita over the medium-term i s likely to continue to translate in increased mobility, while integration into the European Union is likely to foster international traffic. Figure 5.4 illustrates the positive relationship witnessed internationally between per capita income and car ownership. Figure5.4: Car Ownership and Per Capita Income(PPP) in SelectedEuropeanCountries 5 5 0 500 Germany 450 y = 1 . 7 8 1 7 ~ ~ ' ~ ` ~ ' Austria g 400 R ' = 0.8491 Spain 3 5 0 *Netherlands Qa 300 z:: 250 -L 200 1 B u Igaria Greece U 150 Macedonia l;: , , , , , , , , 0 $0 $3,000 $6,000 $9,000 $12,000 $15,000 $18,000 $21,000 $24,00 PPP Per capita GDP Source: Scetauroute International. E75 RoadSectionDemirKapija GevgelijaDetailedDesign of the - Upgrading.FeasibilityStudy / PreliminaryDesign FinalReport, July 1999. 5.18 The key challenge for public expenditure now is to provide sufficient funding to address the maintenance backlogand sufficientfunding on routine and winter maintenance sufficientto ensure assets are kept in operationalcondition. New investmentswould need to be weighed against these critical maintenanceneeds and the constraints of the overall fiscal envelop. 5.19 The level of expenditure required to address backlog maintenanceis significant and will increase if remedial action is not taken quickly, The annual expenditures necessary to address the backlog o f maintenance expenditures over the next few years is estimated at 84 million, or about 1.6 percent o f GDP. For arterial and regional roads, expenditure to clear the maintenance backlog is estimated at 46.4 million per year. For local roads, expenditure to clear the maintenance backlog is estimated at 37.8 million per year (Error! Reference source not found.). The estimates for the maintenance and rehabilitationneeds are based on the current road classification. Were this classification to be updated to reflect the actual low traffic volumes on some road sections, the actual maintenance needs would be reduced somewhat as lower maintenance standards can be applied to lower volume roads. 5.20 The level of expenditure needed for recurrent maintenance of the road network is higher than the amount historically allocated for this purpose. The level o f estimated recurrent expenditure required to ensure that the entire road network assets remain in a "steady 78 state" condition amounts to 35.4 million per year, or about 0.7 percent o f GDP. This estimate includes 21 million for routine and winter maintenance and 14.4 million for periodic maintenance. These sums include 15 million for the local road network, assuming the authorities take a number o f measures to improve the efficiency o f maintenance practices. Efficiencies can be gained through the use o f better prioritization o f periodic maintenance and rehabilitationrather than following a predetermined schedule. Prioritization should be based on systematic economic evaluation using an appropriate Road Financial Management Program. This would require the establishment o f a database providing up-to-date road condition data and traffic counts. A review o f road classifications as noted above could reduce expenditures as roads in lower categories require lower maintenance standards. There is also scope to increase private sector involvement in the maintenance o f roads by contracting out routine and winter maintenance to private contractors. Table 5.2: Estimated Annual Maintenance Expenditure Needs on the Road Network 2008-2012 (In euro million) 2008 2009 2010 2011 2012 Addressing backlog Maintenance National Roads 46.4 46.4 46.4 46.4 46.4 Local Roads 37.8 37.8 37.8 37.8 37.8 Normal Maintenance National Roads Routinelwinter Maintenance 10.6 10.6 10.6 10.6 10.6 Periodic Maintenance 8 8 8 8 8 Bridges and tunnels 1.8 1.8 1.8 1.8 1.8 Local Roads RoutinelWinter Maintenance 8.5 8.5 8.5 8.5 8.5 Periodic Maintenance 6.5 6.5 6.5 6.5 6.5 Total 119.6 119.6 119.6 119.6 119.6 Source:World Bank staff estimates. The following assumptions were made for backlog maintenance estimation: Arterial and high traffic regional roads are reconstructedwhen the condition is poor, and rehabilitated (overlay or mill and replace) when the condition is fair. For regionalroads with low volume, as well as for unpaved regional roads, only roads inpoor condition are rehabilitated. For local roads, only roads inpoor condition are rehabilitated. All unit cost are based on actual empirical data from Serbia, Bosnia, Albania, and Bulgaria. 5.21 Restructuring MakedonijaPat to ensure modern managerial practices and ensuring cost-effective performance is recommended and privatization could be considered. Non- core activities should be spun-off. The toll collection unit could be converted in a service company contracting its service with FNRR. The Road Management services unit could become an engineeringhonsulting company assisting FNRR in managing the network, preparing an annual maintenance program based on economic merits. The restructuring strategy should include a reduction o f excess labor through voluntary and involuntary retrenchment and revision the legislative framework including the Road Act. Makedonija Pat should compete with other private companies for medium-term performance-basedmaintenance contracts granted by FNRR. 5.22 While the level of capital expenditure has been broadly appropriate a national strategic plan is needed to guide such investment. Such a plan should be founded on clear economic justifications, consistent with the financing envelope available to the Government, and 79 should reflect, where considered appropriate, the Southeast Europe Core Regional Transport Network Development Plan for the period2006-2010.The planningand budgetary process inthe sector should be strengthened, with particular emphasis on the economic appraisal of projectsand the monitoring of their execution. FYR Macedonia's capital spending on arterial and regional roads has been concentratedon the constructionof the Skopje and Tetovo bypassesas well as the upgradingto motonvay standards of a number of sections of Corridor X focused on bringingthe entireFYRMacedoniansection of CorridorX to Pan-Europeanmotonvaystandards. 5.23 Addressing theses challenges would require an increase in resources for the road sector. FYR has spent about 1.4 percent of GDP per year on roads allocated to roughly 0.8 percent of GDP for capital expenditure, 0.5 percent of GDP for maintenance, and 0.1 percent of GDP for administration. Addressingthe backlogwould add 1.6 percent of GDP per year, for five years. Increasingannual maintenance efforts to adequate levels to avoid the creation of a future backlog would add 0.2 percent of GDP per year. Assuming capital expenditures remain at historic levels, overall spending on the road sector would rise to about 3.6 percent per year, for five years. Again, estimates of the maintenance backlogwere made on the basis of the current road classification. Were this classification to be updatedto reflect the actual low traffic volumes on some road sections, the actual maintenance needs would be slightly reduced as lower maintenance standards can be applied to lower volume roads. Also, the Government could consider addressing the backlog over a longer time period to reduce the annual cost. Such a strategy would leadto worsening conditionson some roads and so prioritizationof projectswould be very important. 5.24 However, the scope for funding additional road expenditures is constrained. The structure of road user charges follows internationalpractice through a combination of fuel taxes, vehicle ownership charges and tolls. The retail price of diesel and petrol, and the levels of fuel duty and tax on each, are towards the upper end of the levels in neighboring and comparator countries (Figure 5S).The length oftolled roads is unlikely to rise beyond 160km4'over the next 4 years. Figure 5.5: Fuel Prices and Taxes (November 2004) Blrgana S m a BiH Rcinarra MROM Cmaha Albania maDiaseI p c a US$ CentJlltre IpnceUS$llme Gasdire +Tas as a % of lhesel Pnce +Tap as a % of Gasoline Pnce Source: GTZ (2005) InternationalFuelPrices. 5.25 There are a number of steps the Government could undertake to more efficiently raise revenue. Toll collection could be significantly improved. The current open system is 41 Itstands at 130 kmtoday. 80 costly, time consuming, andpresents a highproportionof `leakage'. A Louis Berger study (2002) estimated the leakage to reach some 25 to 30 percent, while the BCEOM report (2006) estimated the leakage during a single week of observation at 15.5 percent. FNRR is aware of the deficienciesand has proposeda new advanced system of payment by paper-based debit (`smart') cards bought inadvance. This solutionwouldbe an improvement, and couldease the transitionto the creation of a closed system of toll collection. Higher revenue should more than offset the higher costs and time savings for the road users would be significant. Such a system could be completed once the gaps in the network are closed (that is, the Tetovo and Skopje by-passes are completed). 5.26 The use of PublicPrivatePartnerships (PPP) arrangementsappears limited to road maintenance and tollingmedium-termcontracts.Beforeentering the longand difficult pathof a large scale PPP transaction, the Ministry of Transport should develop a PPP strategy, and ensure that the institutional framework is revised for this type of transaction. It would set out the issues and sector priorities that could be addressed through PPPs as well as identify the institutional requirements related to the planningand implementationof such projects. Giventhe foreseen sector spending priorities and the current limited institutional capacity, priority should be given to 3 to 5-year performance-based contracts for road maintenance and tolling facilities operation or longer concession-type agreement pooling tolling, maintenance and eventually upgradingactivitieson the motonvaynetwork. 5.27 However,there is a needto ensure that the prerequisitesfor PPPs are in place first. An enablingenvironment for PPPs requires political stability, good macroeconomic policies, the rule of law, and a highdegree of transparency under a clearly defined institutionalframework. A recent survey bythe EuropeanInvestment Bank foundthat the majority of PPPs undertaken inthe absence of such circumstances in the regionhave been unsuccessful. Elsewhere in the region, the World Bank, with the support of the Public PrivateInfrastructureAdvisory Facility, has provided technical assistance to review the institutional framework, and identify the feasibility of PPPs. Suchwork would appear to be timely here. F. THERAILWAY SECTOR Corridor X which links Greece to * P..senge,r -Fr.lphl ..rmm 42This trend bringsMacedoniacloser to the EUaverage of6.6 percent for passenger services and 8 percent for freight services. 81 5.29 MZ has been undergoing a restructuring program. MZ will be divided into two separate state-owned companies in mid-2007. One company will focus on infrastructure (MZ-I) and the other on transport (MZ-T). The latter will be created as a shareholder company and is slated for privatization in future. The accounting o f the two companies was separated in2006. country's size, MZ freight services W - competitiveness with road transport ...Domestic +Import suffers from short average haul o f )50- _ _EXpofi about 128 km. Most local traffic is _-Transl 103. moved by truck and this is not likely - to change. As a result, 99 percent o f 1" the traffic carried by MZ is international and 95 percent o f j r n . freight revenue is generated on L a . y Corridor X. In recent years, MZ has 1w tried to carve out a growing market share in container and transit traffic. 5 0 ' In 2005, with an aggressive 0 7 ~ ~ ~ . ~ ~ . . . _ _ . . . . _ . _ _ _ . . . 5.31 Passenger services are loss making and suffer from low service quality. Tariffs are low and focused on providing a social service focused a limited number o f routes with the budget support o f the Government. Rail passenger services represent only 15 percent o f the traffic volume and 7 percent o f MZ total revenue. Until 2005, only two lines, Tabanovce-Gevgelija and Veles-Bitola, along Corridor X and Xd were operated. In spite o f the absence o f passenger market study and profitability data by passenger routes, the FYR Macedonian Government new line between Skopje and the border with Kosovo in February 2006 and another line from Skopje to Kicevo, on the western part o f Corridor VI11 in June 2006. Both lines are expected to bring 2 percent and 20 percent additional passenger-km compared to 2005 and generate additional losses o f about 200,000 and 500,000 respectively per year to the already loss-making passenger services. Current railway tariff, about 0.18 cents is estimated to be half o f that o f bus services. Interms of the quality of service, the main bottlenecks are the limited service frequency, low commercial speed, around 60 km/h, as well as numerous stops and long waiting time at border crossings which hinder the development o f international traffic. Passenger services should be rationalized based on survey and analysis o f market demand with a view towards ending uneconomical routes. 5.32 MZ needs to further improve its operational performance by lowering costs. MZ infrastructure, locomotives, wagons, coaches and staff were originally designed to handle many times the level o f traffic they currently carry. At 0.8 million traffic units per track-km (2006), the level o f utilization o f MZ infrastructure is low compared to Serbia, 1.1 million traffic unitsper 82 track-km, and the EU15 average of about 3.7 million traffic units per track-km. MZ must adjust its commercial practices, increasing productivity and shedding unproductive assets. Growth prospects for MZ seem to lay in freight transit traffic but operating efficiency would need to improve to effectivelycompetewith other regionalroutesandwith roads. - Government's restructuring m - B o x l action plan includingnatural attrition and retrenchment m . concurrent with an increase in traffic (see Figure 5.8). MZ staff levels have been p 1 5 0 - cut from 9,200 in 1990 to cB 2,879 in 2006. At 250,000 I 1 0 0 . traffic units per employee I 2 o w (2006), MZ labor productivity compares 5 0 - .-1 mo favorably with neighboring countries but remains lower 0 lss5 1896 1887 1988 ISSS Xm 2M1 Xm 2a33 XXU XO5 5.34 Annual railway expenditures have been modest. This is particularly true for capital expenditures. Although cross-countries comparisons on infrastructurespending should be taken with caution, MZ capital spending appears significantly lower than other countries in the region and in the EU. As showed in Table 5.3, from 2002 to 2005, MZ capital expenditures averaged about of 0.06 percent of GDP. This compares to the EU15 average of 0.37 percent of GDP and regionalpeersRomaniaof 0.5 percentofGDP andBulgariaof 0.6 percentof GDP. Table 5.3: Annual Railway Expenditure, 2002-2005 (Inmillionsof euro) 2002 2003 2004 2005 Current Expenditures, including 29.4 27.1 25.8* 29.1** (In percent of GDP) 0.7% 0.7% 0.6% 0.7% Staff 18.5 17.4 16.9 13.8 Energy 3.0 2.9 3.0 3.7 Maintenance 4.5 3.0 3.8 4.0 Capital Expenditures 12.4 5.5 0.1 2.5 Debt Service 2.8 0.1 0.1 0.8 Repayment of Principal 2.2 0 0.1 0.7 Interestsand Fees 0.6 0.07 0 0.03 Total Expenditures on Railways 44.5 32.7 27.0 31.0 ***Significantdecreaserelatedto drop in other operating costs. Includingseverancepayments. Source: MZ. 83 5.35 High staff costs and pressing fiscal constraints have led to low levels of expenditure on maintenance programs. The implementation o f MZ's staff reduction plan has led to a lower wage bill, decreasing from 63 percent o f operating expenditures in 2003 to 40 percent in 2006.43 Overall, staff expenses as a share o f operating costs remain higher than the EU15 average, which i s about 40 percent. Under-funding o f maintenance risks the build up o f a contingent liability as temporary speed restrictions on the network and Corridor X are developed. Unless adequate maintenance programs are implemented, this may lead to serious lack o f competitiveness, commercial damage and loss o f revenue to MZ. 5.36 I n addition to low level of spending, past maintenance and capital expenditure programs have suffered from excess capacity, poor prioritization and planning. MZ maintenance practices have been labor-intensive and machines, equipment and tools used to undertake maintenance and renewal works are to a large extent life-expired and technologically obsolete. As long as funds were available, MZ has used time-based overhaul cycles for complete replacement o f materials regardless o f their condition. Over-maintenance o f tracks carrying very low gross tonnage was therefore not uncommon. In case o f funding shortage, punctual works based on urgency were carried out, leading to the use o f spare-parts o f poor quality. In addition, failure to plan for and implement certain maintenance programs, such as periodic rail-grinding, has accelerated maintenance needs in terms o f sleeper and track replacement as well as ballast tamping. 5.37 Capital expenditure has focused on the new line Beljakovce - Deve Bair at the Bulgarian border which forms part of Corridor VIII. While this project may have been justified politically in 1994 at the time o f the Greek blockade, it was then and remains now an unviable investment from an economic and financial perspective. In the face o f the closure o f Corridor X Greek border, the oil transit point from Thessaloniki port to Skopje, opening a new railway link to Bulgaria would have helped restore an oil supply to FYR Macedonia. However, this railway line has not beencompleted and represents a sunk cost for the railway company and the government. Based on assumptions from the Ministry of Transport, total benefits of the construction o f this line would only start covering total costs after 90 years o f operation. Capital expenditure would be better placed on modernizing Corridor X, increasing the inter-operability and the quality o f operation o f international traffic 5.38 The financial position of MZ has been dismal. The lack o f a policy and legal framework for public contributions to the railway sector led to an inadequate subsidy transfer process and to the extensive use o f indirect public support in the form o f arrears and losses (Table 5.4). Widespread indirect forms o f public support lack transparency as ex post contributions cannot be linked to specific projects and costs. Indirect support also does not contribute to economic efficiency, since public contributions were not allocated based on minimum economic returns. Also, such support undermines incentives for MZ to adopt a more market-based business model. 5.39 Direct government transfers44to the sector have shrunk from 0.26 percent of GDP in 2002 to 0.03 percent of GDP in 2005 and have not been allocated according to sound 43 In2005, the 3.1 million of staff cost reduction was accompaniedby severance payments amountingto 3.2 million. 44 This does not account for severance paymentspaid by the State which represented 0.02 and 0.08 percent of GDP in2004 and 2005 respectively. 84 economic principles. The average state contributionto railways in the EU is 0.3 to 0.4 percent of GDP. Over the period 1991-2005, State Aid to railway operating amountedto 12.7 million. The Government's contribution to capital expenditures was concentrated on a single new investment, two third of which still remainsto be completed. Since the beginningof the project's construction in 1995, the new line Beljakovce - Deve Bair at the Bulgarian border has been funded by a subsidy of 120 million. Also, in the face of competing demands on the Government's budget, the Government's contributionto maintenance expenditures has declined by 13 percent in 2004 and 50 percent in 2005, In 2005, the Government provided a subsidy of 1.5 million covering only 21 percent of maintenance expenditures and 6 percent of total revenues. 5.40 MZ has funded its expenditures mostly out of own revenues. The growth in MZ freight services remains constrained by a lack of resources for investment in new systems and marketingcapabilities. Also, MZ revenues have been burdened by large cross subsidies to loss- makingpassenger services. In 2005, passengerservices revenues covered less than 15 percent of passenger services current costs. The remainder was funded by freight revenues. International experience clearlyshows that cross subsidiesfrom freightto passenger servicesproducethe worst of both worlds. They weaken the competitiveness of freight services and do not adequately finance the passengerside. 5.41 MZ's losses amounted to 10.8 million in 2006, following cumulative losses of 127 million until 2005, in spite of the recent implementationof revenue-increasing and cost- cutting measures by M Z (Table 5.4). The latter have resulted in an improvement in MZ working ratios without subsidy from 145 percent to 101 percent. It should be noted that MZ working ratio progression is significant when compared to other EU railways.45MZ built up arrears with suppliers, social and pensionfunds, andto the publicrevenueoffice. TheMinistry of Finance took over debt service payments for loans from internationalfinancial institutionsand commercial banks. Given its inabilityto fulfill this obligationso far, the Ministry of Financehas been charging penalties to MZ on overdue interest payments which further increased MZ short term financing payables. Arrears to one domestic construction company generated penalty interest payments which were also not paid. A successful law suit against MZ resulted, in the beginningof 2006. MZ's bank accounts were blocked until an agreement was reached in which MZ must pay off5 million in 31 installments. This monthlypayment represents about 8 percent of MZ monthlyoperational revenue. 45From 1990 to 2001, the average EU railway working ratio without subsidies only decreased from 160 percentto about 140percent NERA, 2005. 85 Table5.4: Evolutionof MZFinancialIndicators 2001 2002 2003 2004 2005 EBITDA46(In millions of euros, excluding subsidies) (1.1) (10.4) (7.4) 1.9 2.7 Subsidy- includingseverance payments 2.0 4.6 3.7 3.9 4.4 EBITDA (million ) (3.7) 5.8 7.4 Net Income(million e)includingsubsidy? 0.9 (5.8) (6.0) (15.5) (14.7) (7.1) (2.3) Accumulated losses (million ) (75.5) (89.9) (116.7) (123.8) (127.2) Operating cash-flow (million e) 1.7 0.5 (2.2) 3.O Net change in cash (million ) (2.2) (0.4) 0.8 1.15 Working ratio (%) 97% 121% 114% 82% 78% Working ratio (%, excludingsubsidies) 103% 145% 134% 93% 91% Current ratio: current assets/currentliabilities 0.40 0.27 0.26 0.32 0.37 Indebtedness:Total Liabilities/ (Total liabilities +Equity) 39% 41% 43% 38% 38% Debt structure: Short term liabilities/Total liabilities 82% 87% 93% 88% 93% Source: MZ financial statements G. RECOMMENDATIONS Main recommendationsfor the Road Sector 5.42 The composition of future road transport expenditures will need to be carefully planned and reviewed in order to address fiscal risks stemming from years of under- spending on maintenance. The planning and budgetary processes in the road sector need to be improved with particular emphasis on the appraisal of investment projects and maintenance programs, and effective monitoring of their execution. Improving the efficiency of current spending on the road network should also be a top priority. Towards that end, the authorities are advised to consider the following measures: Increasespendingon maintenanceexpenditures on roads to reducethe maintenance backlog; Review the existing categorization of roads in order to rationalize public spending by linking resourceallocation to road classification and demand; Introduce open competitive tendering for maintenance activities to improve the efficiency of spendingactivities in the sector; Strengthen the technical capacity of MOTC and FNRR staff to improve the identificationand prioritizationof roads maintenanceand investment expenditure; Establishment of an Asset Management System linked to an internationally recognizedeconomicappraisaltool; 46 Earnings Before InterestTaxation DepreciationandAmortization. 86 Develop a sector policy and strategy, consistentwith the fiscal envelope availableto the authorities, and establish a process to regularly revise and update the plan to informthe PIP and the annualbudget cycle; 0 Strengthen budgetary controls and use of competitive bidding in procurement of transportprojects. Main recommendations for the Railway Sector 5.43 The railwaysector medium-termexpenditure and financing plan needsto be revised to make it more financially sustainable. The railway expenditure planning framework could also be strengthened to improve the sector's governance and economic efficiency of Government's budget support. The full implementation of the ongoing sector restructuring program geared toward adopting the EU acquis offers significant opportunities to improve the level, quality and monitoringprocess of public expenditures in the railway sector. Towards that end, the following measuresare needed: Whilst, there may be a strategic, and future economic, case for the development of corridor VI11as part of the TEN and SEETO Core Network, current and projected trafficvolumes suggest that it should be a mediumto longterm objective; 0 Strengthen MZ-I National Program and MZ-T business plans so that they (i) carefully target maintenance and investment expenditures based on their economic returns and on real business needs, (ii) provide a sound basis for national investment decision-making and performance monitoring, and (iii) feed in the annualbudgetand PublicInvestment Programpreparationprocess; 0 Provide well-defined and economically justified infrastructure maintenance and social passengerservices subsidies based on the above plans thereby improvingthe predictabilityand transparency of public support to the sector and increasing the revenuesand competitivenessof freight services; 0 Harmonize the MZ-I National Program and MZ-T business plans with similar programs in neighboring countries to improve interoperability and to increase international traffic through the development of a regional railway transport market. H. CONCLUSIONS 5.44 Years of under-spending on maintenance in the transport sector has led to a significant deterioration of the quality of public assets. The deterioration of the quality of roads, especially at the local level, has created road conditionsworse than in any other country in Southeast Europe except Albania. Fully addressingthe backlog could increase road expenditure by as much as 2 percent of GDP a year for five years, but the Governmentwill be hardpressedto afford such an increase within the availablefiscal envelope and given ambitionsto further reduce the overall level of spending. This cost of maintenance appears overstated by the outdated classificationof roads; prioritizing maintenanceaccordingto the levelof traffic should enable the government to focus onthe most neededmaintenance. Regardingrailways, scarce resources have been devoted to the still unfinishedCorridor VIII, and whilst the completion of this corridor is 87 recognized as a strategic priority, current and projected demand suggests that it should be a medium- to longer-term objective. The MZ has been implementing reforms to increase labor productivity through a programof staff retrenchment and the company will be broken up in the middle of 2007 into a company in charge of the infrastructure and another in charge of the transport. 88