Report No. 65909-ME Montenegro After the Crisis: Towards a Smaller and More Efficient Government Public Expenditure and Institutional Review January 2012 Poverty Reduction and Economic Management Unit Europe and Central Asia Region Document of the World Bank CURRENCY EQUIVALENTS Currency unit = euro (€) €1.00 = US$1.342 (As of December 4, 2011) FISCAL YEAR January 1 – December 31 ACRONYMS AND ABBREVIATIONS ALOS Average length of stay MIC Middle-income country BC Business Court MLSW Ministry of Labor and Social Welfare CALIMS Montenegro Drug Agency MoES Ministry of Education and Science (to 2010) CBCG Central Bank of Montenegro MON- Statistical Institute of Montenegro STAT CPD Continuing professional development NGO Non-governmental organization CR Credit Registry NHS National health service DRG Diagnostic related groups NPA Non-performing asset DZ Primary health-care centers NPL Non-performing loan EC European Commission OECD Organisation for Economic Co-operation and Development ECA Europe and Central Asia (World Bank) OOP Out-of-pocket ECD Early childhood development PEIR Public Expenditure and Institutional Review EECA Eastern Europe and Central Asia PIB Tax identification number EU European Union PISA Programme on International Student Assessments FDI Foreign direct investment PPP Purchasing power parity FMS/ Family material support R2 Coefficient of determination MOP Fund Pension and Disability Insurance Fund ROSC Report on Standards and Codes PIO FYR Former Yugoslav Republic SDR Standardized death rates GDP Gross domestic product SGP Stability and Growth Pact HIC High-income country SME Small and medium-sized enterprises HIF Health Insurance Fund UCG Univerzitet Crne Gore HRMA Human Resource Management Agency UDG Univerzitet Donja Gorica ICT Information and communication technology UM Univerzitet Mediteran IFC International Finance Corporation UN United Nations IFRS International Financing Reporting Standards UNDP United Nations Development Programme IMF International Monetary Fund VAT Value-added tax JMBG Citizen’s identification number WEF World Economic Forum LAMP Land Administration and Management WEO World Economic Outlook Project LEP Law on Enforcement Procedures WTTC World Travel and Tourism Council Vice President Philippe Le Houérou, ECAVP Country Director Jane Armitage, ECCU4 Sector Director Yvonne Tsikata, ECSPE Sector Manager Satu Kahkonen, ECSP2 Task Team Leader Jan-Peter Olters, ECCKO CONTENTS ACKNOWLEDGEMENTS……………………………………………………………….………....…... i EXECUTIVE SUMMARY…………………………………………...…………………….…….…...…. ii I. ENSURING FISCAL SUSTAINABILITY 1. PROLOGUE: THE EFFICIENCY/SUSTAINABILITY NEXUS…..………………………….…….. 1 A. INTRODUCTION AND REPORT OBJECTIVES………………………………………………. 1 B. BACKGROUND AND CRISIS LESSONS…………………………………………………….. 4 C. FISCAL SUSTAINABILITY…………………………………………………………………. 9 D. COMPETITIVENESS AS CENTRAL PILLAR OF THE MIC-TO-HIC STRATEGY …...…...…... 18 E. CONCLUDING REMARKS……..…….…….….…….…….…….……….………….……. 23 2. PENSIONS: THE SINGLE BIGGEST EXPENDITURE ITEM…………………………………….. 25 A. INTRODUCTION………………………………………………………………………..… 25 B. KEY CHARACTERISTICS OF THE PENSION SYSTEM……………………………….…….. 27 C. PENSION SYSTEM’S PERFORMANCE……………………………………………….……. 30 D. ISSUES AND RECOMMENDATIONS………………………….…….…………….……..… 40 3. THE WAGE BILL: MODERNIZING PUBLIC ADMINISTRATION…..…….………………...….. 43 A. INCREASED DEMANDS ON PUBLIC ADMINISTRATION…………….……………….….... 43 B. THE GRADING SYSTEM……………………………………….….……….…….……...... 45 C. CONCLUSIONS………………………………………………………………………..….. 58 II. IMPROVING SOCIAL OUTCOMES 4. EDUCATION: FOCUSING ON OUTCOMES……………………………………………….….…. 61 A. INTRODUCTION…………………………………………………………………..……… 61 B. EDUCATION OUTCOMES……………………………………………………………..….. 63 C. EDUCATION SPENDING………………………………………………………………….. 68 D. IMPROVING QUALITY AND EFFICIENCY………………………………………….……... 72 E. REALIGNING NON-SALARY COSTS AND ENHANCING THE RELEVANCE OF TRAINING……………………..……………………………………………………..…... 77 F. CONCLUDING REMARKS……………………………………………………………..….. 79 5. HEALTH: SWITCHING TOWARDS AN OUTPUT-BASED SYSTEM….……….………………… 82 A. INTRODUCTION………………………………………………………………………….. 82 B. OVERVIEW OF HEALTH SPENDING…………………………………………..……..……. 84 C. HEALTH-CARE EFFICIENCY OR “VALUE FOR MONEY” ……………………………..….. 86 D. HEALTH SPENDING AND FINANCIAL PROTECTION……………………….……………... 91 E. CONCLUDING REMARKS…………………………….……….…………..…….….…….. 97 6. SOCIAL ASSISTANCE: FACILITATING LABOR-MARKET ACTIVATION…………….……… 100 A. INTRODUCTION…….……………………………………..……………………………. 100 B. SOCIAL ASSISTANCE PROGRAMS AND SPENDING…….…………………………..…… 106 C. COVERAGE, TARGETING ACCURACY AND GENEROSITY……..……………………....... 106 D. COST CONTAINMENT AND COST OPTIMIZATION…………………….………………… 111 REFERENCES……………………………………………………………..…………….………….… 114 TABLES Key Policy Recommendations…………………………………………………………... ix Table 1.1. The Structure of Fiscal Deficits…………………………………………………….…… 10 Table 1.2. Fiscal Performance Vis-à-Vis PEIR-1 Fiscal-Policy Benchmarks, 2007–10………….... 16 Table 1.3. Montenegro and Neighboring Countries: Selected International Indices On Factors Affecting the Business and Investment Climate…………………… 21 Table 2.1. Fund PIO Receipts and Expenditures and Government Subsidies, 2006–15……...……. 26 Table 2.2. Comparison of Various Indexation Methods………………………………………….… 28 Table 2.3. Old-Age Dependency Ratios, 1950–2010………………………………………………. 31 Table 3.1. Public-Sector Wage Bill, 2008–10 Averages…………………………………………… 44 Table 3.2. Titles, Grades, and Coefficients of the Salary System………………………………….. 46 Table 3.3. Average Public-Sector Salaries…………………………………………………………. 47 Table 3.4. Length of Service and Salary Increases……………………………………………...….. 48 Table 3.5. Example of the Influence of Seniority on Grade 17 Salary….…………………..….….. 49 Table 3.6. Germany: Salary Grading for Civil Servants with Horizontal Pay Progression….…….. 50 Table 3.7. Selected OECD Countries: Base Salary as Percent of Total Remuneration………….… 53 Table 3.8. Planned Phasing-In of 2011–12 Base Pay for the General Labor Force………...……… 55 Table 4.1. University Density per 100,000 Inhabitants……………...………….…………….….... 62 Table 4.2. Overall Net Enrollment Rates, 2008……………………………….………………….… 63 Table 4.3. School Characteristics and Performance in Mother Tongue, 2008……………….…….. 65 Table 4.4. Education Spending, 2006–10………………………..……………………………...….. 69 Table 4.5. Per-Pupil Spending in Montenegro and Selected Countries, 2009…….…..………….... 70 Table 4.6. Categories of Expenditure by Sub-Sector, 2009…………………………………….….. 71 Table 4.7. Student and Teacher Numbers, 2006–10……………….………………………………. 72 Table 4.8. Average Class Sizes, 2008…………………………………………………………….… 73 Table 4.9. School Units and Classes with Less Than 10 Students, 2009/2010…………………..… 74 Table 4.10. Minimum Annual Time (in Hours) for Teachers Spent on Continuing Professional Development ..………………………………………………….… 77 Table 4.11. Overview of the Montenegrin Education System, 2009/2010………………………..… 80 Table 4.12. Students in Tertiary Education by Discipline of Study….…….………………….…….. 81 Table 5.1. Social-Insurance Contribution (Percent of Gross Wage), 2005–11…………………...... 85 Table 5.2. Sources of Health Insurance Fund Revenues, 2008–09………………………………… 89 0 Table 5.3. Physicians per 100,000 Inhabitants, 2000–07………………………………….….….… 90 Table 5.4. (Non-)Medical Health Workers……………………………………………………….… 90 Table 5.5. Hospital Indicators…………………………………………………………………….… 86 Table 6.1. Public Spending on Social Protection, 2000–10………………………………………. 101 Table 6.2. FMS/MOP Beneficiaries and Benefits………………………………………………… 104 Table 6.3. Monthly Child Allowances…………………………………………………………….. 106 Table 6.4. Spending on Main Social-Assistance Benefits, 2000–10…………………………….... 114 Table 6.5. Targeting Accuracy, Coverage, and Generosity of Social Protection…………………. 114 FIGURES Figure 1.1. Total Public Expenditure, 2010…………………………………………………….…….. 4 Figure 1.2. Risk to Investments, 2006–10………………………………………………………..…... 4 Figure 1.3. Montenegro’s Boom-Bust Cycle and Global Financial Crisis………………….………... 6 Figure 1.4. Growth Rates and Relative Per-Capita Income……………………………………….….. 6 Figure 1.5. Banks’ Boom Behavior and Post-Crisis Portfolio Quality……………………………..… 7 Figure 1.6. Differential Growth and Catch-Up with Average EU Income………………………..….. 9 Figure 1.7. General Government, 2001–10……………………………………………………..…… 10 Figure 1.8. Economic Classification, 2006–10…………………………………………………....… 11 Figure 1.9. Import Reliance of VAT Receipts, 2004–10………………………………………….… 11 Figure 1.10. Financing the Budget Deficit, 2007–11……………………………………………...….. 13 Figure 1.11. Fiscal Deficits and the Corresponding Maastricht Criterion, 2007–10……………….… 14 Figure 1.12. Public Debt and the Corresponding Maastricht Criterion, 2007–10…………….….…... 15 Figure 1.13. Montenegro: Sectoral Growth Rates, 2007–10…………………………………………. 19 Figure 1.14. Montenegro: Selected Productivity Measures, 2005–10…………………………….…. 20 Figure 1.15. Montenegro: Selected Subcategories of Competitiveness Measures…………………… 22 Figure 2.1. Montenegro: Demographic Trends, 1950–2100………………………………………... 31 Figure 2.2. UN Projections of Old-Age Dependency Ratio in Montenegro, 2010–2100…………… 32 Figure 2.3. Montenegro: Number of Pensioners, 1999–2009………………………………………. 33 Figure 2.4. Distribution of Old-Age Benefits, 2009……………………………………….……..…. 34 Figure 2.5. Value of a Pension Point (in Euro), January 2004–January 2010……………………… 35 Figure 2.6. Wage, Inflation, and Pension Developments, 2004–10.………………………….…..… 35 Figure 2.7. Fund PIO Expenditures, 2004–09………………………………………………………. 38 Figure 2.8. Source of Fund PIO Revenues, 2006–09………………………………………………... 39 Figure 3.1. Illustrative Single-Pay Spine………………………………………………………….… 56 Figure 4.1. Average School Performance on Grade-3 Mother Tongue, 2008…………………….… 64 Figure 4.2. Changes in Learning Outcomes, PISA 2009 vs. PISA 2006………………………….… 65 Figure 4.3. Education Performance and Per-Capita Income………………………………………… 66 Figure 4.4. Budgetary Spending and Education Outcomes…………………………………………. 67 Figure 4.5. Total Expenditure on Educational Institutions, 2007…………………………….……... 67 Figure 5.1. Health-Care Spending on Outcomes………………………………………………….… 82 Figure 5.2. Life Expectancy at Birth, in Years……………………………………………………… 83 Figure 5.3. Death Rate Per 1,000 People….….….………………………………………………….. 83 Figure 5.4. Public Spending on Health (Percent of Total Expenditure), 2008……………………… 87 Figure 5.5. ECA and EU: Health-Care Spending and Per-Capita Income, 2008……………………. 87 Figure 5.6. Health Insurance Fund Revenues, 2002–09…………………………....……………….. 88 Figure 5.7. Primary Health-Care Center Productivity and Expenditures…………………………… 89 Figure 5.8. Average Length of Stay in Various Hospitals, 2006–09…………………………..……. 92 Figure 5.9. Bed Occupancy in Various Hospitals, 2006–09………………………………………… 92 Figure 5.10. Private Health Expenditures (Percent of Total Expenditure), 2008…………………..… 93 Figure 5.11. ECA and EU: Out-of-Pocket Spending and Per-Capita Income, 2008……………….… 93 Figure 5.12. Household Share of OOP Payments by Income Quintile……………………………….. 94 Figure 5.13. Household Share of OOP Payments by Region and Quintile…………………………… 95 Figure 5.14. Breakdown of Household OOP Payments……………………………………………… 96 Figure 5.15. Household Health Expenses, by Month, for Medicines Purchased at Pharmacy………. 96 Figure 5.16. Effect of Out-of-Pocket Payments on Household Consumption…………………….…. 97 Figure 6.1. Public Spending on Social Assistance………………………………………………… 101 Figure 6.2. Public Spending on Social Assistance by Programs……………………………….….. 102 Figure 6.3. Spending on Main Social Programs, 2000–10………………………………………… 104 Figure 6.4. Coverage of Social Protection, Overall and by Types of Benefits…………………….. 107 Figure 6.5. Coverage of the Poorest Quintile by Overall Social Assistance in ECA..…………….. 108 Figure 6.6. Targeting Accuracy of Social Protection, Overall and by Types of Programs……....... 108 Figure 6.7. Targeting Accuracy of Overall Social Assistance in ECA…………………………….. 109 Figure 6.8. Generosity of Social Protection, Overall and by Types of Programs………………….. 110 Figure 6.9. Generosity of Last-Resort Social-Assistance Programs in ECA………………………. 111 BOXES Box 1.1. Risks From Accumulating Payment Arrears…………………………………………….. 3 Box 1.2. PEIR-1: Fiscal-Policy Recommendations………………………………………………. 16 Box 2.1. Principles of a Value-Point System…………………………………………………….. 36 Box 3.1. Potential Timeline for Public-Sector Remuneration Reform Program……………….… 60 Box 4.1. Determining Teachers’ Salaries in Montenegro..….………………………………….… 70 Box 4.2. Description of Licensing System for Teachers in Montenegro………………….…….... 76 Box 4.3. Chile’s Teacher Evaluation System……………………………………………….…..… 77 Box 4.4. Life-Long Learning.………………………….…………………………..…….……….. 78 Box 6.1. Main Non-Contributory Benefits in Montenegro…………………………………….... 103 ACKNOWLEDGEMENTS The team comprised Jan-Peter Olters (ECCKO, task team leader), Snježana Plevko (ECSH3), Csaba Fehér (consultant), Svetlana Proskurovska (LCSPS), Nina Arnhold (ECSH2), Toby Linden (SASED), Roberta Malee Bassett (ECSH2), Kari Hurt (EASHH), Ethan Yeh (ECSH1), Boryana Gotcheva (ECSH3), Victoria Strokova (ECSHD), and Danijela Vukajlović-Grba (ECSPE). Very helpful comments, questions, and suggestions were received from Satu Kahkonen and Ronald Hood (both ECSP2), Željko Bogetić (ECSP3, peer reviewer), Ardo Hansson (EASPR, peer reviewer), as well as by H.E. Dr. Milorad Katnić (Minister of Finance), Radoje Žugić (Central Bank Governor), and Jane Armitage (ECCU4). These comments have contributed to a considerable extent to improve the text, sharpened the analysis, and enhanced the presentation, and are herewith gratefully acknowledged. Basically, the euro serves as vacuum cleaner: nothing remains under the carpet. With such a strong currency, one has to put every economically relevant issue openly on the table as one cannot simply reprint money. —Prime Minister Igor Lukšić1 EXECUTIVE SUMMARY 1. Recognizing the important progress with structural reforms so far, on October 12, 2011, the European Commission recommended the opening of formal membership negotiations for Montenegro. Further, in December 2011, Montenegro was granted the membership in the World Trade Organization (WTO), and in June 2012, the country is expected to open the first chapters in membership negotiations with the EU. These are major milestones on Montenegro’s road to the European Union provides a broad external and institutional context––and the destination––for its ongoing reform efforts in a complex international economic environment. In this context, Montenegro needs to continue to strive to improve its policy environment and manage its public expenditures efficiently and improve the quality of public services for all its citizens. Building on the past analytical reports and ongoing economic policy dialogue, this report aims to provide new analysis, ideas, and policy advice to Montenegro’s authorities on how to accelerate its public expenditure reforms with the view to reducing cost, improving quality, and increasing public service delivery. Report objectives 2. Specifically, the present report aims at deriving policy recommendations that would underpin a two-pronged strategy to contain public expenditure growth and increase the “value for money” in Montenegro’s public administration. These objectives reflect concerns of possible inconsistencies in tax and expenditure policies, which could, in the difficult international environment, result—over the medium-term horizon—in unsustainably high budgetary deficits and levels of public debt and conflict with the unilaterally euroized country’s commitment to shadow the obligations of eurozone membership. Moreover, the structure of public spending and the general quality of public services provided by public institutions could be improved substantially in order to foster longer-term growth, support private-sector development, and achieve the Government’s overarching policy goal of becoming a member of the European Union within the foreseeable future. 3. The use of the euro shielded Montenegro’s small, open—and, at the time, dangerously exposed—economy from the worst effects of the global financial crisis, but its use as legal tender requires a strong policy focus on fiscal consolidation and productivity increases. Recent developments in Southern Europe have drawn global attention to the economic risks and social repercussions resulting from the disregard of obligations from eurozone membership, particularly as they relate to the need for prudent fiscal and debt policies. In Montenegro, the authorities adopted an exchange-rate anchor more than a decade ago, after having experienced one of the worst hyperinflations during the mid-1990s. However, this decision has necessitated a commitment to sound fiscal policies and a legal and institutional framework that encourages private-sector activities. With the euro as the sole legal tender, policymakers are precluded from using inflation and currency depreciation as instruments to (i) reduce the real burden of (local currency-denominated) public debt; (ii) offset wage increases in excess 1 See Frankfurter Allgemeine Zeitung of March 31, 2011: “Im Grunde erfüllt der Euro die Funktion eines Staubsaugers: Nichts bleibt unter dem Teppich. Mit einer so starken Währung ist man gezwungen dazu, alle ökonomisch wichtigen Fragen offen auf den Tisch zu legen, schließlich ist es nicht möglich, einfach Geld nachzudrucken.” i of productivity growth; and/or (iii) boost international competitiveness. With euro as the domestic currency, policymakers will need to put focus squarely on structural reforms that would address the root causes of fiscal imbalances and macroeconomic misalignments. Otherwise, as any other dollarized or euroized economy, without strong and sustainable fiscal policy, Montenegro could risk suffering from economic stagnation, rising rates of unemployment, and increasing poverty, either gradually (because companies face increasing difficulty in competing successfully in domestic and foreign markets) and/or in the wake of an acute fiscal and/or balance-of-payment crisis. 4. While Montenegro’s economy has—in principle—enormous growth potential, significant structural, economic, and fiscal risks could, if left unaddressed, tip this favorable outlook. Through the combination of additional efficiency gains in key sectors of the economy and the realization of large- scale, productivity-enhancing foreign-direct investments, Montenegro should be able to set its economy onto a trajectory of sustained economic growth that would close the per-capita income gap with the European Union within a generation. However, large-scale projects with long-term amortization horizons can only be realized in a politico-economic environment that assures investors of a high degree of predictability and security, both in terms of macro-fiscal stability and the business climate. Related concerns include (i) the liquidity-constrained private sector; (ii) the potential inconsistency between the State’s prescribed role and actual tax policy; and (iii) the inability to foster increases in productivity and international competitiveness in a manner that avoids prolonged periods of economic stagnation. 5. With no planned changes in tax policies as key element to its fiscal and institutional adjustment strategy, the Government must rely on the containment of, or cuts in, several of the larger items of discretionary expenditure as principal policy instrument. In 2010, about half the budget was spent on the public-sector wage bill and pensions, exerting considerable pressure on other budgetary items. With few discretionary budget lines, it is clear that the loss of fiscal control over these two budget items alone threatens medium-term fiscal sustainability and macroeconomic stability as well as the Government’s ability to realize (productivity-raising) public investments. The overall economic environment necessitates restraint to avoid fiscally damaging increases in the wages and social-security payments that have been tolerated in recent years. In various strategic Government documents, policymakers have committed themselves to contain public expenditures in select areas. 6. In parallel, the Government needs to put an emphasis on microeconomic (or structural) reforms, especially as it entered, in October 2011, the important stage of negotiations on the accession to the EU. These are intended to help to (further) (i) increase the effectiveness of public institutions; (ii) improve the efficiency in the delivery of public services; (iii) streamline bureaucratic procedures; and (iv) strengthen the rule of law and business climate. To this end, the authorities have requested the World Bank input provided by the report into the ongoing debate on how to best create a more modern and more efficient public administration that can be financed sustainably from the country’s tax base. Economic context: The efficiency/sustainability nexus 7. The economic crisis during 2008–10 exposed pre-existing fissures and structural weaknesses in the foundation of Montenegro’s economy, underpinning the notion that the 1¾-year recession had been the starting-point of a structural regime shift rather than a temporary interruption in growth. The combination of several—a priori unconnected—developments have caused a-more-severe-than- anticipated recession, leaving Montenegro with four distinct episodes in its first five years of its post- independence history: (i) the early boom (2006-Q2 to 2007-Q2); (ii) real-estate-based speculation and the bursting bubble (2007-Q3 to 2008-Q3); (iii) the fallout from the global financial crisis (2008-Q4 to 2010- Q2); and (iv) the fragile recovery (since 2010-Q3). It was, in particular, the highly pro-cyclical lending ii patterns of commercial banks that had contributed to the lasting effects and the continued fragility of Montenegro’s post-crisis economy. 8. During the 2008–10 crisis, the Government sought to pursue, to the extent possible, moderately counter-cyclical fiscal policies. With growth slowly recovering, it now needs to tighten expenditures, consolidate the budget, and ensure long-term debt sustainability—not least because the tax base will not return to the levels seen during the previous economic boom. This projection follows the expectation of a smaller external capital-account surplus and a correspondingly narrower current-account deficit, with imports being the main element of adjustment. This means that trade-related tax revenues, especially the value-added tax, will decline accordingly as a share of GDP. In response, the Government has taken politically difficult measures (i) postponing capital expenditures; (ii) freezing public-sector wages; (iii) adjusting pension entitlements; and (iv) integrating extra-budgetary funds into the central Government’s budget. 9. With a view to supporting the Government’s twin economic objectives of maintaining macroeconomic stability and advancing socio-economic development, the previous PEIR published in late 2008 had advocated a more rules-based approach to fiscal policies. However, the severity of the crisis had made their implementation particularly difficult. As argued in the previous PEIR, to achieve the fiscal objectives spelt out in the Economic and Fiscal Programme, the Government might want to implement its fiscal-adjustment strategy within the context of explicitly defined “fiscal benchmarks” so to lengthen budgetary planning horizons and be able commit—as credibly as possible—to a pre-announced fiscal-policy path. The analysis contained in the current PEIR thus supports the set of fiscal-policy principles derived in the predecessor document: (i) Fiscal policy stays within the constraints of the Stability and Growth Pact, reflecting implicit requirements to shadow the obligations of eurozone membership; (ii) Recurrent expenditures are financed from the regular tax base, thereby precluding Government from borrowing for reasons other than public investments; (iii) Future budgets do not increase but aim at a gradually reducing the ratio of recurrent expenditure to GDP and improving the composition of expenditures towards productive public investments. The benefits of such a rules-based fiscal-policy approach, if adopted, would be the Government’s ability to combine features of short-term budgetary flexibility with medium-term fiscal discipline. A fiscal framework that comprises these benchmarks should result in a gradual decline of sovereign risk premia and, subsequently, lower interest rates for public debt and more fiscal space for productive public investments. 9. The rapid increase in Montenegro’s the public debt-to-GDP ratio during 2009–11 and the vulnerabilities in the world economy suggest the usefulness of complementing the formal “fiscal rules” with tighter “policy objectives”. The aforementioned fiscal rules will likely prove insufficient, given: (i) the fiscal weakening that had occurred after the previous crisis; (ii) the sensitivity of Montenegro’s small, open economy to external shocks in general; and (iii) the risk that a follow-up shock caused by the sovereign debt crisis in the Southern eurozone countries and/or the US would set off adverse debt dynamics. To recreate the fiscal buffer and protect the markets’ positive assessment of Montenegro’s creditworthiness, the authorities would benefit from a benchmark on public debt significantly below 60 percent. To allow for a comparable response to any future crisis, economic policymakers should outline a longer-term fiscal-adjustment program anchored on a debt-to-GDP ceiling of 35±5 percent, with a limited response window that would allow Government increase debt during iii extraordinary periods but commit it to a post-crisis adjustment to pre-crisis levels within a pre-specified period of time. Pensions: the single biggest budgetary expenditure item 10. Long-term fiscal consolidation can only succeed if the pension system expenditures—on which about one-quarter of the budget is spent—is brought under control. Amplified by adversary demographics and an expected rapid increase in the old-age dependency ratio, Government has to follow through on its envisaged path of gradually decreasing, as percent of GDP, the expenditures to the Pension and Disability Insurance Fund, making pension reform—like elsewhere in Europe—a pivotal long-term policy priority. Important reforms, including the gradual increase in the pension age to 67, have already been implemented in recent years, but the combination of a tight fiscal situation in the post-crisis era and the gradual aging of the society places related reforms at the very center of a structural reform agenda — notwithstanding the indications that Montenegro’s welfare system has been relatively successful in (i) protecting the elderly against poverty; and (ii) ensuring that the level of contributions paid into the system (while employed) is reflected by the ultimate amount of pension benefits received (once retired). As in many other countries with aging population, the pension system’s current financial position is unsustainable, with contribution revenues falling systematically below benefit expenditures. The budget cannot afford to subsidize the current pension system on the order of about 4 percent of GDP on a long- term basis (as it has been doing during 2009–11). 11. Fiscal problems of the pension system can be addressed, however, through constraining entry benefit levels, benefit indexation, and—to some extent—the inflow of new beneficiaries. (i) In terms of indexation, Government has decided to link benefits mostly to inflation and not to increases in wage rates. This policy, which will limit expenditure growth, is in line with (emerging) international practice. This measure will only have a noticeable impact over a multi- year period, but—increasingly important—compound effects will help the Government to achieve its fiscal consolidation objectives in this area. In terms of benefit adequacy, price indexation ensures that beneficiaries can continue to purchase the same basket of goods and services but are excluded from the benefits of productivity increases (or real wage growth rates) that occurred after their retirement. (ii) Entry pensions are assessed on the basis of pension points, which measure the relative contribution intensity and thus maintains horizontal equity while allowing room for longitudinal adjustments should financial sustainability so require. Point systems help to ensure that benefits paid to people of the same cohort reflect contribution histories fairly, while allowing policymakers to adjust point values, which provides the basis for assessing pension benefits of consequent cohorts in accordance with long-term fiscal objectives. By automatically adjusting point values to the same composite index as benefits, the Montenegrin pension system forgoes the longitudinal freedom inherent in the point system. It is important for pension adequacy in the long run that the adjustment of the point values is decoupled from the benefit indexation regime and is based on long-term equilibrium values based on actuarial projections. (iii) The established schedule for increasing the retirement age for both men and women to 67 years already places the country among the most stringent ones in Europe. Another measure that Government might consider is limiting the inflow of new beneficiaries through reducing the scope for occupational early retirement. It is also recommended that the pension law is amended in a way which introduces automatic retirement-age increases to ensure that life expectancy at the statutory retirement age remains constant. iv 12. In addition, the Government needs to decide whether it wishes to build a robust, domestic voluntary pension industry. If so, (i) supplementary pension savings will need to be tax-advantaged; (ii) domestic capital markets developed (especially in terms of tradable securities); and (iii) the regulator’s technical capacity improved. In the absence of these steps, the fragile—i.e., barely existent—voluntary pension industry will remain unviable. In such an environment, mandatory second-pillar schemes should not be introduced. In (re-) assessing its policy stance towards voluntary pension savings, the Government would need to revisit the rules under which Montenegrin citizens can join —either directly or through domestically licensed agents—open pension schemes that are registered in selected developed markets. This approach would necessitate a review of international regulations concerning financial-sector regulations and tax administration, a task that would need to be undertaken in the context of the EU accession process. The wage bill: modernizing public administration 13. The wage bill represents the principal risk to the overall fiscal sustainability, not only because about one-quarter of general Government expenditures is reserved for salaries but also because of political pressures to further increase staff levels. Equivalent to about 12 percent of GDP, Montenegro spends about 3 percentage points more on public-sector wages than the average of neighboring countries in the Western Balkans—reflecting, in part, the “fixed cost” element of running a small country and the requirement for the public administration to respond to an ever more complex EU integration-related structural-reform agenda. In such an environment, there are growing pressures to recruit additional public-sector staff (rather than reduce public-sector employment) and/or to provide monetary incentives to highly qualified staff with expert knowledge. A clearly defined reform agenda is even more important as recent international-competitiveness surveys have identified “inefficient Government bureaucracy”—despite noted progress—as one of the most problematic factors of doing business in Montenegro. 14. In its present form, the civil service encounters constraints in its efforts to provide public goods and services as responsively and accurately as would be expected from an upper middle- income country preparing for EU membership. In particular, the inherited system of remuneration of civil servants and state employees includes a number of poor incentives, which —if realigned appropriately—would give scope for considerable efficiency gains. In addition, capacity will need to be built up among currently employed civil servants and state employees to manage the process of EU accession. The authorities have taken important steps towards reforming and streamlining the public administration, including by adopting (i) the 2010 Personnel Policy Paper; and (ii) the so-called “two- for-one” principle (by which the Government limits its ability to recruit one additional staff by requiring the concomitant release or retirement of two civil servants or state employees). 15. Given the increased complexity of tasks to be carried out by the public sector over the medium term, it is critical that related reforms succeed in retaining and promoting qualified and high-performance staff. To be able to use existing resources more effectively, it is recommended to increase the transparency of the payment structure by merging the “variable” and “supplemental” pay elements with the base pay structure (based on recommendations derived by a recent EU report). In developing a new pay policy, the Government might want to consider setting up a working group charged with the management and coordination of reform efforts in this area. Key objectives would be to (i) ensure that the salary system is fiscally affordable and, at the same time, competitive in the labor market for well-qualified candidates; (ii) design measures to improve the transparency of the remuneration system through streamlining legislation, developing a concept of the base pay that consolidates current variable and supplemental pay elements; (iii) implement the principle of “equal pay for equal work”; (iv) recognize performance through the pay system in a transparent manner, based on, and subject to, regular v performance appraisals; (v) improve staff retention through attractive career-development prospects by introducing an adequately decompressed salary structure, objective criteria for pay progression, and support to skill development; (vi) facilitate decision-making capabilities in Government by ensuring the availability of accurate pay information; and (vii) develop and implement a time-bound program of measures (and responsible officials) to introduce a new remuneration system for the public sector that is based on the core principles of the pay system and assign resources to implement this program. 16. Finally, the Ministry of Finance leads the work on development of the legislation on public service remuneration. It would be important that a working group be established for the development of a draft law on public sector remuneration which would include representatives of the HRMA, ministry of labor, ministry of interior (local government department), and representative of the trade unions. This multi-skill group would be able to provide input on issues related to the job classification and job valuation for different categories of civil servants and public employees, as well as help build support for new pay policies, embedded in the law. Education: focusing on outcomes 15. The objective of modernizing education within well-defined budgetary constraints is particularly difficult as Government has spent as much as other countries in the region in this sector but without achieving comparable outcomes. Major reforms in the education sector, including those to increase transparency, still lie ahead. At present, particular obstacles comprise (i) the lack of easily accessible information on spending, staffing, graduation rates, graduate employment, and research output in the (higher) education sector (making very difficult an effective steering of reforms in this sector); (ii) the financing of public higher education (which is rather static and has no performance-related component); and (iii) the fragmentation and internal organization of the public university (faculties continue to act as quasi-autonomous entities). 16. Montenegro faces difficult choices in the twin challenge of improving the quality of education and curtailing fiscal expenditures in this sector. To this end, Government has given particular attention to education-sector reforms, as signaled by the splitting of the Ministry of Education and Science into two separate ministries that are charged with the implementation of a comprehensive reform strategy aimed at (i) improving education outcomes; (ii) capping per-student expenditures (which are roughly at regional—but still below OECD—averages); and (iii) rebalancing the composition of spending on public education (where an exceptionally high percentage of expenditures are used for teachers’ salaries). To be able to achieve all of these objectives concomitantly, Government will have to consider a gradual reduction in the number of (non-)teaching staff—taking advantage of a situation in which a non-negligible percentage of teachers has already passed retirement age, thereby opening the possibility of implementing a (relatively painless) policy of workforce attrition with relatively limited effects on overall teaching ratios and class sizes. However, bringing about these changes will probably require more extensive changes to the organization of schools and universities. Health: moving towards an output-based system 17. The situation in health care is slightly different as Montenegro, relative to its per-capita income, spends an above-average amount but also generates above-average outcomes. In fact, most health-related indicators are at a level close to or above those prevailing in the Central European EU member states, and Montenegro continues to build on the progress made during its post-independence years, which comprised reforms to (i) define the legal and strategic framework for the organization of the health system; and (ii) develop the required institutional and management capacities (supported by capital vi investments in tools such as an integrated health-information system). Montenegro has embarked on specific reforms to improve the efficiency and effectiveness of the health system, starting with primary health care. The country has changed its system by introducing chosen doctors for primary care, with primary health-care centers being reorganized to support their work and the health outcomes for their registered patients. The reforms to the primary health-care centers were implemented to motivate the registration of patients and the provision of priority services. 18. Health-sector reforms are, however, work in progress, with core challenges remaining in the areas of fiscal sustainability and the efficiency of service delivery. Health Insurance Fund has declined substantially after the economic crisis, partly due to reduced social insurance contribution rates, and financial sustainability of the HIF and health sector spending is a concern. Primary health-care centers remain inefficient at this early stage of reform, and hospital indicators show room for increasing productivity. Finally, out-of-pocket payments for health remain low overall, but have become more regressive in recent years with the poor paying more for health and medicines. Principal recommendations to achieve the policy objectives include the following: (i) Introducing a DRG payment system; (ii) Providing greater autonomy to health facilities while building capacity for effective management; (iii) Defining the basic benefit package for secondary and tertiary care; and (v) Taxing “unhealthy” foods. (vi) Closely monitoring out-of-pocket payments for health. Social assistance: facilitating labor-market activation 19. Government can increase the socio-economic impact of social assistance by linking it to the provision of social care and employment/activation services. Social-care services help to overcome multiple vulnerabilities faced by the families that receive social assistance. Labor market activation supports “graduation” from social assistance for those who are able to work. Activation requires an authorizing environment (i.e., laws that allow flexible transition from social assistance to work but also preserve the right to resorting to social assistance if the job proves to be temporary). It also requires tax and benefit structure that preserves the incentives of taking a job instead of staying on social assistance for long time. 20. At present, spending on social assistance does not pose a fiscal risk, partly because cash transfers are means tested. However, since the overall spending envelope for social assistance and services is lower than in other ECA countries (while needs might be increasing), it is important to adequately finance the existing programs and to allow them to expand if demand goes up —not least to be able to implement critically important reforms to the contributory programs. The options for reallocation of additional resources to the family material support (FMS/MOP) within the current social-assistance spending envelope are quite limited, given its small overall size (and that of other programs). Whereas keeping the spending on family material support at around 0.5 percent of GDP is desirable, increases in the coverage and spending on child protection (especially on the monthly child allowance) would strengthen social policy and have very positive welfare benefits. Given that children in families with more than two children are at a higher poverty risk, it would be reasonable to waive the limit of three children in one recipient family and extend the child allowance to all children in poor families, especially if they attend school. Some potential for savings and opportunities for cost containment could be found within the category of disability benefits. Their costs could be contained by tightening the procedures for disability certification and—possibly—by the introduction of means-testing for certain disability benefits vii that go to recipients who have low disability rates and for whom social inclusion through participation in employment would be more beneficial. 21. The family material support is a well-targeted program, but with a low coverage. It could be scaled up to capture a larger share of poor that are not covered by this program—e.g., by relaxing the income thresholds for access to the FMS/MOP (as it has already happened in January 2011) and indexing the benefits annually with the CPI. In addition, the limitation of having up to five beneficiaries in one family could be waived to have the program poor large families in their entirety. Changes in the eligibility rules, such as the removal of the exclusionary “filters” and their replacement with a scoring formula (which would take into account asset ownership and remove automatic exclusion) could significantly improve coverage. Finally, proactive outreach to families who do not apply because of lack of knowledge, information, and/or guidance through the application process can also help to enhance the take-up and coverage of the FMS/MOP. To make expansion of coverage possible without significant increase in the budget envelope for this program, it would be necessary to scrutinize rigorously and assess realistically the incomes and assets of applicants and to avoid further increases in program generosity for the existing recipients. It would also be necessary to apply a “work test”/workfare requirement to the recipients of FMS/MOP. By determining who would be able to work, such a requirement would exclude those who work informally. viii Key Policy Recommendations over the short-term horizon over the medium-term horizon Chapter 1. Prologue: The Efficiency/Sustainability Nexus • define fiscal policy objectives: • adopt fiscal rules: recreate fiscal buffers by aiming for (i) stay within Maastricht criteria; (ii) golden rule; and public debt-to-GDP ratio of 35±5 percent (iii) non-increasing current expentiture-to-GDP ratio Chapter 2. Pensions: The Single Biggest Expenditure Item • constrain entry benefits: • constrain inflow of new beneficiaries: adjust benefit indexation and valorization of reduce the scope for occupational early retirement, pension point benefits introduce automatic retirement-age increases Chaper 3. The Wage Bill: Modernizing Public Administration • reform and modernize remuneration system •Improve public administration effectiveness through reforming develop policy paper (with objectives and the public sector remuneration system to reflect job content, results indicators) and program of activities required competences and staff performance, while keeping the pay structure simple and transparent. Chaper 4. Education: Focusing on Outcomes • increase transparency and sector efficiency: • reform the financing of higher education, make available information on spending, staffing, while increasing proportion of expenditures directed graduation rates, and graduation employment towards quality-enhancing measures Chaper 5. Health: Moving Towards an Output-Based System • tax unhealthy foods; • provide for output-based system, • define basic benefit packages accompanied by greater management and establish criteria for inclusion/exclusion from benefits budgetary autonomy for hospitals and HIF Chaper 6. Social Assistance: Facilitating Labor-Market Activiation • strengthen family material support, • link social assistance to provision of social care a well-targeted program with low coverage: and employment/activation services: essential for reforms to contributory systems facilitate "graduation" from social assistance ix I. ENSURING FISCAL SUSTAINABILITY 1. PROLOGUE: THE EFFICIENCY/SUSTAINABILITY NEXUS2 Montenegro is in the process of implementing a comprehensive reform and modernization program that aims at ensuring fiscal sustainability and improving the efficiency of social spending. The following analysis is meant to provide input into the public debate on ways to achieve the country’s overarching economic-policy objectives within the context of binding post-crisis/pre-integration constraints and against the backdrop of increasingly evident fragilities in the external environment. The 2008–10 recession, a consequence of the concurrence of the domestic boom-bust cycle and the global financial crisis, exemplified the ease with which adverse debt dynamics could be set off, depriving Montenegro’s small, open, and euroized economy of fiscal buffers (potentially) needed for future crises. Considerable macro-fiscal benefits could be derived by defining fiscal rules and objectives that would lengthen budgetary planning horizons and commit the Government to a pre-announced fiscal-policy path. Particular attention will need to be paid in the short- to medium-term to reducing the problem of internal payment arrears. A. INTRODUCTION AND REPORT OBJECTIVE 1.1 Montenegro’s economy has an enormous growth potential. Through the combination of further efficiency gains in key sectors of the economy3 and the realization of large-scale, productivity- enhancing foreign-direct investments (FDI) in tourism, energy, and services, Montenegro can set its economy onto a trajectory of sustained economic growth and close the per-capita income gap with the European Union (EU) within a generation. The World Travel and Tourism Council (WTTC), for instance, has simulated that Montenegro would be the world’s fastest-growing travel and tourism economy and its fastest-growing destination for investments during 2011–21 (WTTC, 2011). A similarly important impetus could come from the substantial interest that currently exists in Government and among foreign investors to develop Montenegro’s untapped hydro potential and transform the import-dependent energy sector into a regional energy hub.4 1.2 However, there are significant structural, economic and fiscal risks that —if left unaddressed—could tip the favorable outlook and set off a prolonged period of socio-economic stagnation. Like elsewhere, large-scale projects with long-term amortization horizons will be realized only in a politico-economic environment that assures investors of a high degree of predictability and security. Highly publicized litigation cases in (international) courts act as much a deterrent to investors as the absence of a national consensus on the assessment of inherent costs and benefits of (irreversible) public-infrastructure investments.5 In general, the economic recovery that began to take hold in 2010-Q3 2 Prepared by Jan-Peter Olters (ECCKO). 3 In terms of basic productivity figures, Rojec et al. (2010) reported a considerable improvement of average GDP produced by employee, whether nominal in euro or in US dollar-based PPP terms, increasing by an average 8.4 and 12.0 percent annually during 2001–08. No data, however, were reported for (post-)crisis years. 4 Montenegro can only exploit its natural resources sustainably—i.e., become and remain a high-end tourist destination and a net exporter and regional hub for energy—if the economic development of these land-intensive sectors occurs in a complementary rather than predatory manner. Given the permanent nature of any large-scale investment in either sector, the sustainable development of Montenegro’s natural resources necessitates careful analyses of long-term economic and environmental impacts and the broadest possible participation and inclusion of stakeholders and citizens. 5 The progress made by Montenegro in its EU integration process—culminating in the European Council’s decision in December 2010 to grant Montenegro the official status of candidate country—ha s proven an important factor in 1 has remained fragile, comprising considerable macroeconomic and fiscal risks. In its latest Article IV Report, the International Monetary Fund (IMF) warned the authorities of a potential “relapse into recession” (IMF, 2011). Overall, fiscal (and environmental) sustainability concerns are the key risk factors to Montenegro’s socio-economic development potential, including those related to (i) the liquidity-constrained private sector (see Box 1.1); (ii) the State’s prescribed role and the actual tax policy, which puts extra burden on expenditure adjustment as an instrument of overall fiscal adjustment; and (iii) the inability to foster increases in productivity and international competitiveness in a manner sufficiently consistent to prevent prolonged periods of economic stagnation. 1.3 Within this delicate context, policymakers need to carefully balance economic opportunities and (fiscal, social, and environmental) sustainability risks. This task is made even more challenging by Montenegro’s small country size, and the use of the euro as legal tender makes fiscal policy the key policy instrument. While the latter feature precludes the authorities from relying on exchange-rate and monetary-policy instruments, the former constraint limits the effectiveness of fiscal stimuli, not least because the multiplier effect—in an economy as import-reliant as Montenegro’s—is (very) small. In its Economic and Fiscal Programme 2012–15 (which outlines a fiscal-adjustment path towards a budgetary surplus by 2013), the Ministry of Finance (2011) recognized the critical importance of devising prudent fiscal policies over a longer-term horizon so as to ensure their sustainability over different phases of the business cycle. In light of the limited policy effectiveness of standard macroeconomic policy instruments, Government will have to put a greater emphasis on microeconomic (or structural) reforms —for the aforementioned economic reasons as well as for EU integration-related ones. Central to corresponding policies are measures aimed at (i) increasing the effectiveness of public institutions; (ii) improving the efficiency in the delivery of public services; (iii) streamlining bureaucratic procedures; (iv) strengthening the rule of law and business climate; and (v) ensuring broad-based support within the country for critical infrastructure investments. 1.4 Against this background, the authorities strive for a more modern, more efficient Government that can be financed sustainably (and without fundamental changes in tax policies 6), calling for a “strategic” debate over medium-to-long-term reform options needs to be had on how best to combine the objectives of increased efficiency with those of sustainability. Thus far, the Government has resisted calls to consider changes in tax policies as (key) pillar in the adjustment program. In its 2010 Article IV Consultation, the IMF (2010), for instance, had warned of the risk of “an inconsistency between low tax rates and the large size of the public sector.” A year later, the IMF (2011) recommended that the authorities consider “small increases in income tax and VAT rates” as means of balancing otherwise difficult-to-implement cuts to current expenditures. Objective of the Report 1.5 In 2010, about one-half the budget was spent on the public-sector wage bill and pensions, thereby exerting considerable pressure on other budgetary items (Figure 1.1). With few “discretionary” budget lines, the Government has recognized that the loss of fiscal control over these two budget items alone risks (i) fiscal sustainability and macroeconomic stability; and (ii) the Government’s ability to realize public investments at a level that would be required to crowd in productivity-enhancing private-sector activities. Already, expenditures on salaries and pensions have increased at an helping to strengthen investor confidence and attract FDI that could bring in the funds and the know-how necessary to raise overall productivity. 6 While the 2010 budget broadened the tax base for the personal income tax and increased the rates for social contributions, business climate considerations have convinced the Government not to increase the rates of the income or value-added taxes. At the same time, core public infrastructure continues to suffer from the legacy of decades of neglect and requires investments and maintenance (World Bank, 2008a). 2 unsustainable rate during 2008–10. As summarized in Figure 1.2, the share of capital-to-total expenditures was a mirror image to the Government’s relative spending on the wage bill and pensions. Box 1.1. Risks From Accumulating Payment Arrears The resumption of economic growth in mid-2010 has not led to improved payment discipline among Payment Arrears, 2006 –10 households and firms. Partly a legacy from the boom- (In percent of GDP) bust cycle, accumulated payment arrears have reached macro-relevant levels, threatening economic recovery 2006 2007 2008 2009 2010 and banking-sector recovery (Vukajlovic-Grba, 2011). Already on the eve of the global financial crisis, the Total arrears (gross) 7.8 7.8 14.4 25.4 36.4 private sector—with overpriced (real estate) assets in their balance sheets—had been struggling to stay current of which: NPLs 1.3 2.8 6.9 11.6 16.8 with payment obligations. In its wake, firms and Source: Vukajlovic-Grba (2011). households lost access to bank (re-) financing and: (i) drew down available buffers (bank deposits from the private sector contracted by an average monthly rate of The Quality of Banks’ Assets, 2006‒10 6 percent during September 2008–April 2009); (ii) resorted to parallel financing mechanisms (trade credits and barter arrangements); and (iii) accumulated payment arrears towards employees, suppliers, and creditors. Because of the deteriorating quality of loan portfolios, banks raised their clients’ default risk, limited credit exposure, and demanded higher interest rates. This resulted in three years of negative private sector credit growth rates, reducing the stock of credit by about one- third. The economy, unable to roll over or refinance existing debt, entered a “vicious circle” in which the post-crisis credit crunch triggered the build-up of payment arrears and non-performing loans. An increasing stock of payment arrears adversely affects the balance sheets of otherwise healthy companies—thereby risking the Government’s ability to maintain macroeconomic stability and protect the proper functioning of markets. To ensure sustainability Source: CBCG; and World Bank staff estimates. of a successful inter-enterprise arrears clearance process, the strategy would need be complemented by multi-layered set of measures aimed at redressing existing institutional constraints related to credit registries, bankruptcy procedures, contract enforcement and judicial reform, as well as auditing, accounting, and corporate/public financial management. The problem of payment arrears will require concerted focus, monitoring and measures to Policy Recommendations stop the creation of new and reduce the stock of old arrears. The government has formed a over the short-term horizon over the medium-term horizon working group to deal with the problem but more aggressive actions might be needed. By • strengthen financial discipline • improve the quality and transparency addressing the private sector’s illiquidity, (to minimize delays in the payment of business and credit registries including through bilateral and multilateral of commercial transactions) (facilitate firms' "rational" decision-making) nettings of (overdue) debt, key obstacles to stimulating higher growth—the increased • initiate comprehensive inter-enterprise • improve the effective enforcement of availability of bank financing and the improved clearance procudure contracts (to help to restore firms' financial discipline of other market participants— (including multilateral compensations) creditworthiness and access to capital) would be addressed and provide businesses with • implement effective bankruptcy • enhance knowledge and skills an increased degree of planning certainty (and procedures (to improve quality/disclosure of balance Government with a steadier stream of fiscal (preclude balance-sheet contaminations) sheets and financial management) revenues). In a separate report now under preparation, the Bank is analyzing the problem of Source: Vukajlovic-Grba (2011). payment arrears with the view to assisting the authorities in gradually resolving this significant financial problem for Montenegro’s private sector. 3 Figure 1.1. Total Public Expenditure, 2010 Figure 1.2. Risk to Investments, 2006–10 Source: Ministry of Finance. Source: Ministry of Finance. 1.6 The Government is conscious of the need to stem against the politically tempting, but fiscally harmful trend of increases in the wage bill and social security. To this end, the Government has requested World Bank input into its efforts to develop a two-pronged strategy with which to (i) constrain the growth in expenditures; and (ii) increase the “value for money” in the public administration.7 In response, this follow-up Public Expenditure and Institutional Review (PEIR-2) seeks to build on the set of policy recommendations derived in the previous report (World Bank, 2008a) and, in so doing, contribute to the public debate on how to best modernize Montenegro’s public administration and create the legal, regulatory, and institutional foundation for a smaller and more efficient Government. B. BACKGROUND AND CRISIS LESSONS 1.7 The economic crisis has exposed pre-existing fissures and structural weaknesses in the foundation of Montenegro’s economy, underpinning the notion that the 1¾-year recession had been the starting-point of a structural regime shift rather than a temporary interruption in growth. Already during the period during which Montenegro had been “one of the world’s fastest growing non-oil economies” (World Bank 2008a), a number of domestic and international reports had warned the authorities of the inherent risk of overheating and of an increasingly likely hard landing.8 However, in the end, it was the combination of several, a priori unconnected developments that caused a-more-severe-than-anticipated recession. During the first five years of Montenegro’s independence, it is useful—to be able to draw the appropriate lessons from this crisis—to distinguish four distinct episodes of recent economic history. While the initial Public Expenditure and Institutional Review or PEIR-1 (World Bank, 2008a) looked at the first two periods, the present text will focus on the latter two: 7 In Montenegro’s context, this is a particularly challenging task. Given the small size of the economy and the “fixed cost” of running a country, Montenegro’s public administration is very small in absolute—but unsustainably large in relative—terms, with the EU expecting Montenegro to achieve the same standards in every sphere of public administration as any other member country of any size. 8 During this period, annual rates of real GDP growth averaged about 9 percent. For early warnings of a potential overheating and the risk of a hard landing, see, e.g., CBCG (2008), World Bank (2007), and IMF (2008). 4 x Independence and the early boom (2006-Q2 to 2007-Q2). The pent-up investment demand that had been released after the referendum on—and the international recognition of—Montenegro’s statehood in 2006-Q2 turned an anemic economy into a booming one, as reflected in the quadrupling of stock- market indices over the period of just one year. For about a year, up until mid-2007, growth was principally driven by domestic factors that can be subsumed under the heading of “post -independence euphoria”. Even though credits to the private sector increased by close to 90 percent during the eight- month period through February 2007 (i.e., at an annualized rate of about 130 percent), the banking sector’s credit-to-deposit ratio remained constant, at a (healthy) level slightly above 100 percent. x Real-estate speculation and the bursting bubble (2007-Q3 to 2008-Q3). The bullish phase gradually slid into a foreign-financed, real estate-based speculative bubble. During the period up until end- 2008-Q3, the credit-to-deposit ratio increased from about 100 to more than 170 percent. In an aggressive race for market share, and reflecting the excess liquidity in international capital markets, foreign (parent) banks provided substantial funds to their Montenegrin branches, amounting to about €1.2 billion (or about 40 percent of GDP). Concomitantly, the private and public sector propped up the domestic bank’s liability side at an even faster pace, leaving the entire banking sector dangerously exposed on the eve of the global financial crisis (Figure 1.3). The irrational exuberance9 prevailing during the boom’s mature phase masked the fact that Montenegro’s economy had become overly dependent on large (but impermanent) capital inflows and private-sector borrowing. During 2008-Q3, when real-estate prices in Podgorica exceeded those in Berlin and Beijing (World Bank, 2008a), it had become evident that the real-estate bubble was about to burst and set off a self-reinforcing cycle of declining asset prices, decelerating real growth rates, and tightening credit markets. x The fallout from the global financial crisis (2008-Q4 to 2010-Q2). Montenegro was just beginning to slide into recession when its—by that time, precariously vulnerable—economy was forced to absorb the storm surge emanating from the collapse of Lehmann Brothers.10 Against the background of the bursting real-estate bubble, the global financial crisis made the recession deeper and more prolonged than in neighboring countries (Figure 1.4). The subsequent deterioration in Montenegro’s fiscal position, the near collapse of the industrial sector, increasing levels of payment arrears, and the emergency rescue operation for Prva banka were some of the immediate consequences of the radically altered economic landscape. Adverse private-sector debt dynamics during the boom years thus precluded a simple turning back to the bullish market of the immediate post-independence years. x The fragile recovery (since 2010-Q3). The unexpectedly strong summer tourist season in 2010, paired with favorable weather conditions and metal-price developments, had the secondary and tertiary sector pull Montenegro’s economy out of the “twin crisis”-induced recession, tipping GDP growth from –5.7 percent in 2009 to modest growth in 2010 (Figure 1.4).11 However, the factors that fueled the modest growth rates in late 2010 might not prove as dynamic as hoped, with 2011-S1 data for the industrial sector already showing signs of reappearing negative growth trends. In addition, the accumulation in payment arrears within the private sector represents an important risk to the resumption of more dynamic rates of economic growth. 9 During the so-called “dot-com” bubble in the late 1990s, the former Chairman of the Federal Reserve Board, Alan Greenspan, coined this term to warn of overvalued asset prices and an impending bust. It was first used in a speech on “The Challenge of Central Banking in a Democratic Society” in Washington on December 5, 1996: “But how do we know when irrational exuberance has unduly escalated asset values, which then become subject to unexpected and prolonged contractions …?” (http://www.federalreserve.gov/boarddocs/speeches/1996/19961205.htm). 10 Inherent policy challenges have been analyzed by, inter alia, Bénassy-Quéré et al. (2009). 11 In late 2011-Q3, MONSTAT published the official 2010 GDP figures (€3.054 million), with a higher -than- expected growth rate of 2.5 percent (compared to initial projections of a +1.1 percent growth rate for 2010). With nominal GDP estimates being only €31 million (or 1 p ercent) higher than initial projections, GDP ratios in the following text are not affected. 5 Figure 1.3. Montenegro’s Boom-Bust Cycle and the Global Financial Crisis Source: Central Bank of Montenegro; Montenegroberza; and World Bank staff estimates. Figure 1.4: Growth Rates and Relative Per-Capita Income Real Growth Per Capita Income Sources: IMF (World Economic Outlook, April and September 2011); and World Bank staff estimates. 6 1.8 The highly pro-cyclical lending patterns of commercial banks have contributed to the inherent fragility of Montenegro’s post-crisis. The privatized banks’ aggressive race for market share culminated in private-sector credit growth rates of 140 and 180 percent during the boom years 2006 and 2007, respectively. These rates exceeded those of any other country. At this pace, banks were not able— partly as a result of the myriad of institutional constraints—to assess business plans and the creditworthiness of borrowers adequately, thereby sowing the seeds for their post-crisis balance-sheet problems. The banking sector’s own pre-crisis exposure was amplified by the fact that, on the banks’ balance sheets, more than 40 percent of bank assets were not covered by deposits. Funds had been made available by their respective parent organizations. From one day to another (after the collapse of Lehman Brothers), the credit expansion was not only not funded any longer, but Montenegrin branches were required to consolidate their balance sheets—setting off a ten-quarter episode of negative credit growth. The stock of loans extended to the private sector, which stood at €2.8 billion at 2008-Q3, declined by 27 percent to €2.0 billion at 2011-Q1; see Figure 1.3. The cash-constrained and income-deprived private sector, facing difficulty in repaying its debt, refinancing existing loans, or borrowing fresh capital, has increasingly resorted to alternative financing mechanisms. The arrears accumulation process started with the most highly indebted companies, gradually infecting the balance sheets of other firms and, increasingly, the banking sector. Non-performing loans, which represented about 16 percent of total loans at end-2010, increased to about 25 percent during 2011-Q2. The most affected banks are, of course, those that have lent most aggressively during the boom, and these are the four systemically important banks in the system; see Figure 1.5. Figure 1.5. Banks’ Boom Behavior and Post-Crisis Portfolio Quality Trendline: NPLt = 4.77 + 0.059 ∆ct‒1, R2 = 0.707. Sources: Central Bank of Montenegro; and World Bank staff estimates. 7 1.9 The euro (debt) crisis of 2010–11, which unfolded in direct succession to the global financial crisis, reinforced the need to adjust the underlying growth model and rely more on productivity- enhancing reforms. Having experienced one of the worst hyperinflations during the mid-1990s,12 the authorities adopted an exchange-rate anchor more than a decade ago,13 and the experiences of Greece, Ireland, Portugal, and other member countries of the eurozone’s Southern periphery demonstrate the risks of managing small, open economies without any recourse to exchange-rate policies. The combination of fiscal, debt, banking, and/or competitiveness vulnerabilities affecting these countries directly relate to the aforementioned risk factors of Montenegro’s economic recovery. To avoid being ca ught in a vicious circle of eroding competitiveness and rising levels of fiscal and/or external deficits, the focus of post- crisis reforms will have to be the adjustment to the underlying growth model so as to lay the foundation for—what the EU has termed—smart, sustainable, and inclusive growth,14 fueled by innovation and increased productivity (both endogenous and FDI-induced). 1.10 The focus on growth is crucial, with Montenegro’s per-capita income representing only about 35 percent of the EU average. Having taken office in the immediate aftermath of the European Council’s decision to confer to Montenegro the status of an official EU candidate country, the new Government defined as the overarching priority of economic policymaking the objective of increasing potential growth rates. Then-Prime Minister-designate Igor Lukšić (2010), in presenting his Government’s program to Parliament, outlined this as one of his principal policy objectives: Montenegro’s economy has the potential to generate long-term economic growth in the coming decade at an average rate of 5 percent annually. Such growth rates would mean accelerated convergence to the EU member states’ average growth level and would substantially improve Montenegrins’ living standards. Achieving continuously high economic growth will require significant economic reforms. Hence, this Government’s objective is to create the preconditions for turning high rates of potential growth into reality. After the income gap vis-à-vis the EU average had narrowed by more than 6 percentage points during 2006–08, Montenegro’s per capita GDP, as percent of the EU average, slipped from 35.9 percent in 2008 to 35.3 percent in 2010 (Figure 1.4). To avoid falling into the so-called middle-income trap, policymakers will have to anchor economic policies in a longer-term, “middle-to-high-income-country” (MIC-to-HIC) strategy in order to achieve sustainably higher-than-current potential growth rates. The aforementioned objective is ambitious but achievable. While the economic benefits for any given year might be relatively small (and difficult to measure), the compound effects of related reforms can be very considerable. The calculation in Figure 1.6 shows that, if Montenegro succeeded in raising its average growth rates from the current 1 to about 3.5 percent above the average rates prevailing in the EU—i.e., from about 2.5 to 5 percent15—, the country’s per capita income would be equal to the EU’s average not in 105 but already in 30 years (Figure 1.6). This also means that, to achieve the goal of a 5-percent potential growth rate, the 12 With a monthly inflation rate of 313 million percent, Hanke and Kwok (2009) included the Yugoslav hyperinflation as one of the highest in history. On the subsequent euroization decision, see also Fabris et al. (2004). 13 Six countries outside the 17-member eurozone have adopted the euro, three with a formal agreement with the ECB (Monaco, San Marino, and the Vatican) and three without (Andorra, Kosovo, and Montenegro). As a result, Montenegro has no recourse to exchange-rate and monetary policy instruments, and its fiscal space reflects implicit obligations to shadow the corresponding criteria spelt out in the Stability and Growth Pact. 14 In response to the global financial crisis, the European Commission (2010a) has sought to revive the European economy by devising a ten-year strategy that was to “turn the EU into a smart, sustainable and inclusive economy delivering high levels of employment, productivity and soci al cohesion.” The three pillars of this strategy aimed at (i) developing an economy based on knowledge and innovation (“smart”); (ii) promoting a more resource efficient, greener and more competitive economy (“sustainable”); and (iii) fostering a high-employment economy delivering social and territorial cohesion (“inclusive”). 15 During 2001–10, the EU economy grew by an average of less than 1½ percent. 8 structural reform agenda would need to increase potential growth—generally estimated at about 4 percent (IMF, 2010)—by “only” about 1 percentage point. Figure 1.6: Differential Growth and Catch-Up with Average EU Income The graph represents the following relationship: yn = (1 + [)n y0 or n = (ln(yn/y0)) (ln((1 + [))‒1, where yn = average per capita income in the EU (benchmark, set at 100 percent); y0 = Montenegrin per capita income in 2010 relative to the EU average (35.3 percent), [ = the difference between average growth rates in Montenegro and the EU, and n = years. Source: World Bank calculations. C. FISCAL SUSTAINABILITY 1.11 During the 2008–10 crisis, the Government sought to pursue, to the extent possible, moderately counter-cyclical fiscal policies. With growth rebounding, it now needs to tighten expenditures, consolidate the budget, and ensure debt sustainability. The Government increased total (current) expenditure from an average level of 44 (37) percent of GDP during the three boom years (2006–08) to 47 (41) percent during the two crisis years (2009–10); see Figure 1.7. Within this budgetary envelope, the Government boosted the share of expenditures on social services, health, education, and the environment from close to 57 percent of total expenditures during 2006–08 to more than 59 percent during 2009–10 (Figure 1.8). In view of the adverse socio-economic impact of the crisis and the decision to support the restructuring of the struggling industrial sector, the Government allowed —temporarily—a relaxation of its fiscal stance. As budgetary (tax) revenues decreased from an average 47 (39) percent of GDP during the boom to 43 (37) percent in the crisis years, the overall fiscal position deteriorated correspondingly. The budgetary balance tipped from an average surplus of 2.8 percent of GDP during 2006–08 to a 4.1 percent deficit during 2009–10. As shown in Table 1.1, the variations have largely been 9 due to the central Government, including budgetary funds, largely reflecting the counter-cyclicality of budgetary revenues. Having introduced a flat income-tax with a single rate of 9 percent for both households and corporations, the value-added tax (VAT)—especially those levied on imports—has become an increasingly critical source of budgetary revenue. Nominal VAT receipts thus decreased from €440 million in 2008 (14¼ percent of GDP) to €364 million in 2010 (12 percent of GDP). Table 1.1. The Structure of Fiscal Deficits, 2006–10 2006 2007 2008 2009 2010 (In percent of GDP) Overall fiscal balance 3.0 6.4 -0.3 -5.3 -3.9 Central government and funds 3.4 10.0 0.5 -4.1 -2.2 Municipal governments -0.4 -3.7 -0.8 -1.2 -1.7 Sources: Ministry of Finance; and World Bank staff estimates. Figure 1.7. General Government, 2001–10 Sources: Ministry of Finance; and World Bank staff estimates 10 Figure 1.8. Economic Classification, 2006–10 Sources: Ministry of Finance; and World Bank staff estimates. Figure 1.9. Import Reliance of VAT Receipts, 2004–10 Trendline: VAT = 57.3 + 0.16 IM, R2 = 0.908. Sources: Ministry of Finance; IMF; and World Bank staff estimates. 11 1.12 The tax base will not return to the levels seen during the economic boom. At the height of the boom in 2007, gross FDI had represented more than 40 percent of GDP, with domestic credit increasing by more than 50 percentage points of GDP. But these extraordinary demand stimuli will remain a unique episode of Montenegro’s economic history (see also World Bank, 2008a). At present, the private sector is over-indebted and has little access to domestic or international capital markets, while the levels of FDI are experiencing a trend decline. As a result, the external capital-account surplus will continue to shrink, resulting in a corresponding narrowing of the current-account and trade deficits. As a share of GDP, the latter two ratios declined from a peak of, respectively, 51 and 68 percent in 2008 to an estimated 26 and 44 percent in 2010—principally caused by lower imports. This trend adjustment is expected to continue further over the medium-term horizon. As shown in Figure 1.9, this means that trade-related tax revenues, especially the VAT, will decline correspondingly as a share of overall output and—in the absence of changes to the underlying tax policies—not revert back to pre-crisis levels. 1.13 The Government’s response to the twin crises caused public debt to increase by almost 17 percentage points of GDP during 2007–10. Instead of borrowing directly from international banks, as in mid-2009,16 the Finance Ministry placed its maiden 5-year, €200-million Eurobond at an interest rate of 7⅞ percent in September 2010 and a second €180-million Eurobond at 7¼ percent in April 2011.17 As a complementary source of financing, the Government chose to issue (and roll over) 6-month Treasury bills placed in the domestic market, at interest rates that have gradually declined from more than 4 percent in 2009-Q1 to a weighted average of about 2⅔ percent in 2011-Q1 (Figure 1.10). The moderately counter-cyclical fiscal policies, while cushioning the domestic impact of the collapse in global demand, resulted in a faster deterioration of the budgetary situation than in most neighboring countries and fiscal deficits in excess of the 3-percent Maastricht criterion during 2009–10 (Figure 1.11), an implicit policy objective that reflects Montenegro’s interest in formalizing the use of the euro in the context of EU membership negotiations. The inherent risk of such a policy stance is related to the speed with which the country’s public debt position worsened (Figure 1.12). In the wake of the crisis -response measures, Montenegro’s stock of public debt increased from 26 percent of GDP in 2007 to about 43 percent in 2010 (52 percent, if sovereign guarantees18 are included), demonstrating the ease with which the fiscal situation could spin out of control and the economy enter a vicious downward spiral akin to countries in the eurozone’s Southern flank. 16 Instead of committing to an IMF-supported program, Montenegro decided, at roughly 6 percentage points above Euribor, to borrow €30 million from Erste Bank and €90 million from Credit Suisse in 2009. 17 In a sign of increased capital-market confidence, Montenegro sold €180 million in 5 -year Eurobonds to 67 investors from 21 countries in April 2011, with an annual fixed coupon of 7.25 percent (i.e., 62.5 basis points lower than for its maiden €200-million issuance in September 2010). The sale occurred a few days after Moody’s had changed its outlook for the Ba3 sovereign bond ratings from “negative” to “stable”. The credit-rating agency explained its decision with measures taken to address the key risks that had triggered the negative outlook in April 2009. Moody’s noted that (i) the Government had started to reverse the fiscal deterioration and address structural weaknesses in its public accounts (including a lower-than-planned 2010 budget deficit, the pension reform, and the strict public-sector recruitment policy); (ii) the economy was showing signs of a gradual recovery; and (iii) the situation in the (still fragile) banking sector was beginning to stabilize (given strengthened levels of capitalization and parent banks’ willingness to provide support for their local subsidiaries). Moody’s expected that the EU candidate status would provide an anchor to Government policies. A further upgrade would be dependent on a continued reduction in the budget deficit and a reversal in the rising trajectory of public debt. The credit rating agency underlined that the lack of timely and high-frequency data continued to hamper its analysis. 18 Montenegro granted sovereign guarantees for debt amounting to about 9.4 percent of GDP in support of major enterprise restructurings in the industrial sector, in exchange for equity positions in those companies. This means that the Government—by reversing these transactions at some point over the medium term —could improve its debt situation tangibly. This would presuppose, however, that (i) the labor force and debt-restructuring operations have helped to ensure the firms’ economic viability over the longer-term horizon; and (ii) related receipts are indeed used for debt repayment purposes. 12 1.14 The Government is conscious of the fiscal challenges and has taken politically difficult measures aimed at consolidating expenditure growth. In reacting to the significantly worse-than- expected economic environment in the immediate post-crisis period, the Government responded by (i) postponing capital expenditures; (ii) a freeze in public-sector wages19; and (iii) adjusting pension entitlements. This fiscal-policy stance reflected the attempt to balance the need for crisis-related interventions, including those in the especially affected financial and industrial sectors, with the continued commitment to limit the deterioration in public finances.20 With a view to tightening control over expenditures, the Government has integrated extra-budgetary funds into the central Government’s budget, leaving only municipal balances outside the Treasury’s direct control. Figure 1.10. Financing the Budget Deficit, 2007–11 Sources: Ministry of Finance; and World Bank staff estimates. 1.15 The severity of the shocks and the depth of the ensuing recession caused a larger-than- foreseen deviation from fiscal-policy objectives, making more urgent the need for a medium-term fiscal-adjustment path. Apart from reasons related to the need to regain macroeconomic stability, the 19 Some of these envisaged effects were undone by the requirement for a gradual integration of meal and holiday bonuses into the basic wage, foreseen to take place during 2011–12. Given that these income items were not subject to taxes and contributions before, the adjustments will lead to a 3-percent rise in public-sector wage during the adjustment period. 20 A rise in tax arrears, the bankruptcy of the steel mill (in which context a sovereign guarantee over about 1 percent of GDP was called), and the need to fund otherwise uncovered police-force pensions can be expected to result in a fiscal deficit close to one percentage point above the budgeted level of 2.6 percent of GDP. 13 political progress in the EU integration progress21 and the serious deficit/debt/competitiveness difficulties faced by some eurozone member countries make the political focus on cementing fiscal sustainability an even more urgent task.22 While the strength of the economic shocks and the effects on the economy— including the accumulation of payment arrears in the private sector (discussed below)—resulted in the non-compliance of several of the (never formally adopted) PEIR1-inspired policy benchmarks (Box 1.2), the ongoing economic recovery presents policymakers in Montenegro with an opportunity not only to present to the public and the market a consistent, post-crisis fiscal-adjustment path, as proposed in Ministry of Finance (2011), but also to accompany the Economic and Fiscal Programme with an Action Plan entailing accompanying measures and reforms. Figure 1.11. Fiscal Deficits and the Corresponding Maastricht Criterion, 2007–10 Sources: Ministry of Finance; Eurostat; IMF; and World Bank staff estimates. 21 The European Council’s decision, in December 2010, to grant Montenegro the formal status of a candidate country has increased the likelihood—subject to the fulfilment of the seven “policy priorities” spelt out in the November 2010 avis (opinion)—of opening accession talks in the foreseeable future. It can be expected that, in this context, the formalization of a exchange-rate arrangement not foreseen in the Treaty on Economic and Monetary Union will be discussed, requiring Montenegro to demonstrate economic convergence and adherence to membership requirements, including with respect to the conduct of fiscal policies. 22 This point has become even more valid after the decision to borrow €380 million from the international capital market. In principle, Montenegro’s ability to access the international capital market at this very difficult period of time, at rates significantly below what commercial banks in Europe or elsewhere would have charged, has sent a positive signal of the capital market’s confidence in Montenegro’s conduct of fiscal policies. Given this exposure, the capital market—as well as international credit rating agencies—will monitor closely the implementation of economic policies in Montenegro, which should help to enshrine fiscal discipline over the medium to long term, not least because Montenegro borrowed these amounts with an explicit communication to the market of its medium- term Economic and Fiscal Programme, making this an even more binding policy document. 14 Figure 1.12. Public Debt and the Corresponding Maastricht Criterion, 2007–10 Sources: Ministry of Finance; Eurostat; IMF; and World Bank staff estimates. 1.16. While the Government succeeded in responding to the immediate socio-economic effects of the global financial crisis relatively effectively, the “cost” has been the near doubling of the public debt-to-GDP ratio in only 2½ years. This experience suggests that the formal adoption of the “fiscal rules” described in Box 1.2 might prove insufficient in the current economic situation characterized by the rapid deterioration in the eurozone and the U.S. It is thus recommended that the (broader) “fiscal rules”— ideally enshrined in legislation—are complemented with (tighter) “policy objectives” that would reflect: (i) the fiscal weakening that had occurred after the previous crisis; (ii) the sensitivity of Montenegro’s small, open economy to external shocks; and (iii) the risk that a follow-up shock caused by the sovereign debt crisis in the Southern eurozone countries and/or the US would set off uncontrollable debt dynamics. Thus, in an effort to recreate the fiscal buffer lost during the 2008 –10 crisis and protect the market’s positive assessment of Montenegro’s creditworthiness,23 the authorities would benefit from a public-debt ceiling significantly below 60 percent. To allow the Government to react to any future crisis in a manner comparable to the post-2008 response, economic policymakers would be well-advised to outline a fiscal- adjustment program aimed at achieving (and maintaining) a debt-to-GDP ratio of no more than 35±5 percent, together with the definition of a limited response window that would allow Government increase debt during extraordinary periods (comparable to the global financial crisis) but commit to an immediate policy reversal, a return to pre-crisis levels within a pre-specified period of time less than an electoral cycle. 23 As reflected in the trend decline in interest rates (Figure 1.10) and the fact that the last decision taken by international credit-rating agencies was an upgrade (Moody’s changed the outlook on Montenegro’s Ba3 rating from negative to stable in late March 2011). 15 Box 1.2. PEIR-1: Fiscal-Policy Recommendations To support the Government in achieving its twin economic objectives of maintaining macroeconomic stability and advancing socio-economic development, the previous PEIR had advocated a rules-based approach to fiscal policies. In particular, the report recommended that Government defines “fiscal benchmarks”, such as the ones proposed below, as a means to lengthening budgetary planning horizons and committing—as credibly as possible—to fiscal policies that aim at creating the fiscal space for discretionary priority expenditures: (i) Given the implicit requirement to shadow the obligations of eurozone membership, fiscal policy should aim at staying within the constraints of the Stability and Growth Pact (SGP). This policy recommendation is a direct consequence of the unilateral euroization decision in 1999/2000. As Montenegro seeks to join the EU and formalize the use of the euro as legal tender, policymakers are bound —implicitly—by the SGP fiscal- policy constraints (or Maastricht criteria). For any given fiscal year, these fiscal criteria limit (i) the overall budget deficit of general government to 3 percent of GDP; and (ii) public debt to 60 percent of GDP. (ii) Recurrent expenditures should be financed from the regular tax base. This implies that: (i) Government can only borrow to finance investments; and (ii) any windfall revenues (whether privatization or non-recurring tax revenues from specific events) will be earmarked for the repayment of public debt and/or public investments. This “modified golden rule” should help to prevent policymakers from committing to recurrent expenditures during boom periods cannot be maintained during economic downturns. (iii) Future budgets should aim at a gradually declining ratio of recurrent expenditure to GDP . The rates of nominal increases to current expenditures should lower than —at minimum, be capped at—(expected) nominal GDP growth rates. Given existing pressures for increased expenditure on, particularly, the large budget lines for the wage bill and pensions, this principle appears to be the most difficult to attain (and maintain) and requires particular attention (largely motivating the focus in the present PEIR). The benefits of such a fiscal-policy approach, if adopted, would be the Government’s ability to combine short- term budgetary flexibility with medium-term fiscal discipline—by allowing for overall fiscal deficits during economic downturns and improving the fiscal fundamentals of medium-term budgetary plans; see Chapter II in World Bank (2008a). A fiscal framework that comprises all three elements should result in (i) an improvement in the grades conferred to Montenegro by international credit rating agencies; (ii) the gradual decline of sovereign risk premia; and (iii) lower interest rates for (public) debt. Notwithstanding the Government’s principal concurrence with the aforementioned fiscal-policy benchmarks, the two shocks emanating from the domestic boom-bust cycle and the global financial crisis have forced the Ministry of Finance —during the severe recession—into a “crisis management” mode (Table 1.2). With growth rebounding and given existing fiscal and economic vulnerabilities, it has become critically important to take a more strategic view of fiscal-policy priorities. Table 1.2. Fiscal Performance Vis-à-Vis PEIR-1 Fiscal-Policy Benchmarks, 2007–10 2007 2008 2009 2010 Shadow obligations in SGP Fiscal balance (deficit ≤ 3 percent of GDP) 6.4 9 -0.3 9 -5.3 8 -3.9 8 Public debt (stock ≤ 60 percent of GDP) 27.5 9 31.9 9 40.7 9 44.1 9 Borrow only to invest ("golden rule") Public investments (in percent of GDP) 7.0 … 10.1 … 2.1 … 2.1 … Difference of investments over deficit > 0 13.4 9 9.8 9 -3.2 8 -1.8 8 Reduce ratio of total recurrent expenditure to GDP Ratio of total recurrent expenditure to GDP 36.4 … 40.4 … 42.7 … 41.0 … Change in percentage points of GDP ≤ 0 -1.8 9 4.0 8 2.3 8 -1.7 9 Sources: Ministry of Finance; and World Bank staff estimates. 16 Inherent Risks and Macroeconomic Outlook 1.17 Montenegro’s economic vulnerability is largely a function of its economic prospects being closely linked to, and dependent on, the summer tourist season24 and the inflow of FDI. Both components are traditionally highly variable and largely dependent on external factors —and, as a result, very difficult to predict. This reliance explains the high degree of external vulnerability and is amplified by the current volatility in global markets, commodity markets (including aluminum, Montenegro’s key commodity export), and the global economy teetering at the brink of a “double dip” recession. 1.18 A renewed deceleration in economic growth worldwide would reverse recent terms-of- trade improvements and leave the still vulnerable sectors exposed to the vagaries of nervous markets. A renewed—exogenously caused—economic shock would affect, in an especially brutal manner, industry (currently undergoing an adjustment process to ensure its economic viability) and a banking sector (which is in the process of addressing its portfolio quality challenges, a legacy from the boom-bust cycle and the effects from the global financial crisis 2008 –10). At this stage Montenegro would have few buffers to protect its economy from a new wave of economic strains— with the risk that, ultimately, costs will have to be borne by the budget. This would jeopardize the realization of envisaged socio-economic development outcomes, a risk that is particularly pronounced in a country of Montenegro’s size where even relatively small shocks may have a sizeable impact on the economy. 1.19 Over the medium-term horizon, the current-account deficit—while remaining at a level considerably higher than elsewhere in Southeastern Europe—is expected to narrow considerably. Preliminary information seems to indicate that the political and economic instability around the Mediterranean has provided the tourism sector a (temporary) boost but dampened (further) FDI inflows. Many tourists visiting Montenegro during the 201 summer season appeared to have shied away from relatively riskier destinations and chosen destinations along the Montenegrin coast instead. The same developments have, however, had an inverse effect on FDI. Without special effects (such as privatizations in 2006, real-estate purchases in 2007, the partial privatization of the electricity company in 2009, and bank recapitalizations in 2010), Montenegro is projected to observe a gradual decline in the FDI-to-GDP ratio. Both higher inflows in the services account and a gradually declining capital account surplus would be consistent with a continued narrowing of the external current-account deficit, possibly beyond that what had been forecast by the Ministry of Finance (2011) and the IMF (2011).25 The adjustment in external imbalances is amplified by a strong increase in exports (especially aluminum and electricity) and transport services, albeit from a very low base and notwithstanding negative or anemic growth rates in Montenegro’s main trading partners (close to two-thirds of exports go to Serbia, Greece, Italy, or Hungary), and nominal growth rates in imports lower than those of projected GDP. Already, the adjustment the current-account deficit from 51.3 percent of GDP in 2008 to an estimated 25.6 percent in 2010 is entirely explained by the adjustment in imports, and the continuation of this trend to a projected 19 percent of GDP in 2013 is 24 During 2005–10, tourists spent an average of 66 (88) percent of total tourist overnights during July –August (June– September). 25 Both the IMF (2011) and Ministry of Finance (2011) expected the current-account deficit to decrease only marginally in 2011 to 24.5 percent and 25.4 percent, respectively, with the difference being explained by discrepancy in net FDI projections of 15.4 percent of GDP (IMF) and 16.2 percent (Ministry of Finance). Over the medium term, the IMF (2011)—similar to the World Bank—sees a faster adjustment in the current-account. 17 expected to follow a similar decline in imports,26 from about 83 percent of GDP in 2008 to a projected 40 percent in 2013. 1.20 With current rates, the tax revenue-to-GDP ratio is forecast to decline further, reflecting the expectation of lower imports (as share of overall output). To a large extent, this projection reflects Montenegro’s participation in regional tax competition, as manife sted in a flat— corporate and personal—income tax rate of 9 percent, a symbolic 1 percentage point lower than the corresponding ones in Albania, Bosnia-Herzegovina, Kosovo, FYR Macedonia, and Serbia. With given tax policies, income taxes can be expected to generate relatively constant tax revenues of about 5 percent of GDP. By contrast, the VAT-to-GDP ratio has gradually declined from close to 15 percent in 2007 to about 12 percent in 2010, with the post-crisis adjustment in imports explaining most of the trend decline. With the VAT representing almost one-half of tax revenues, a further narrowing of external imbalances (as forecast) would translate into a contraction of the tax base and—in the absence of a change in tax policies—require a further efforts in containing public expenditures and ensuring the highest possible “value for money” for resources made available to public administration. To be successful, the Government will have no option but to enforce pre- defined budgetary envelopes for line ministries and ensure the long-term sustainability of wage and pension commitments—as well as to increase the overall effectiveness of public expenditure. D. COMPETITIVENESS AS CENTRAL PILLAR OF THE MIC-TO-HIC STRATEGY 1.21 A word of caution upfront: placing “competitiveness” at the center of a reform agenda is fraught with complexities, not least because the concept is multi-faceted, ill-defined, and difficult to measure (and even more so in a country with traditionally limited data availability). Typically, “international competitiveness” is defined quite narrowly as the ability of enterprises in a given country to sell goods and provide services profitably in (unprotected) domestic and foreign markets. As such, the comparative concept relies on easy-to-measure elements of price and quantity (rather than more-difficult- to-capture aspects of quality). This “standard” definition of competitiveness was challenged by Krugman (1994), who described it as “a seductive ideal, promising easy answers to complex problems”—warning that “[t]hinking in terms of competitiveness leads, directly and indirectly, to bad economic policies on a wide range of issues.” According to him, domestic productivity growth, not some measure of competitiveness, determines a country’s welfare—which, in turn, is a function of capital, know-how, and technology that FDI can bring to a country. This “broader” competitiveness concept thus comprises elements of business climate, legal framework, contract enforcement, labor force qualifications, and the quality of public institutions and infrastructure. These elements, Olters (1999) sought to incorporate when defining as “competitive” an economy that managed to foster a high-quality export sector with a stable demand base over the entire business cycle and the capacity to easily adapt to changing (especially unfavorable) environments. If applied to Montenegro, this definition would make transport, communication, and tourism—as well as energy (if measured since 2008)—some of the most competitive sectors in the economy, showing both high average growth rates and low standard deviations. These sectors have proven capable of buffering the effects of adverse external developments and reacted only moderately to the effects of the crises (Figure 1.13). 26 As the over-indebted private sector has only very limited access to domestic and international capital markets, it will be the capital inflows from FDI and tourism receipts that will effectively cap imports. 18 Figure 1.13. Montenegro: Sectoral Growth Rates, 2007–10 Sources: MONSTAT; and World Bank staff estimates. 1.22 In both the narrower and broader definition of “competitiveness”, Montenegro’s post -crisis economy has shown some signs of gradual improvement. The IMF (2011) looked at a set of “narrow” competitiveness measures, including the wage and producer price indices, concluding that the “nascent adjustment in costs” has “improved competitiveness”. The—admittedly not overly robust—results of a trend reversal to the loss in international competitiveness seen over the course of the boom-bust cycle are broadly consistent with complementary productivity and competiveness measures summarized below in Figure 1.14. In terms of the broader competitiveness definition, there is no single figure capturing progress. Table 1.3 summarizes the most important competitiveness figures and compares Montenegro’s position with those of (i) neighboring EU countries (Bulgaria, Greece, Italy, Romania, and Slovenia); (ii) EU candidate countries (Croatia, FYR Macedonia, and Turkey); and (iii) and potential EU candidate countries (Albania, Bosnia-Herzegovina, Kosovo27, and Serbia). The unweighted average over all competitiveness indicators not only shows a gradual improvement from an implied 68th to the 64th rank, but also a relatively favorable position between the neighboring EU countries (53rd rank in 2010) and the EU candidate countries (67th). 27 Given the non-inclusion of Kosovo in many international rankings, it has not been included in the averages for potential candidate countries 19 Figure 1.14. Montenegro: Selected Productivity Measures, 2005–10 Sources: Central Bank of Montenegro; MONSTAT; and World Bank staff estimates. 1.23 Maybe more important than the overall position in international surveys might be the various rankings’ subcategories, which comprise critical information that could help to prioritize a structural reform agenda (Figure 1.15). In Montenegro’s case the banking sector’s fragility, the credit squeeze, and increasing non-performing loans are all reflected in (i) the relative decline of financial sector development; (ii) businesses’ increased concerns related to access to finance (both Global Competitiveness Index); or (iii) the World Bank’s getting-credit measurement (Doing Business); see Figure 1.5. By contrast, the surveys show consistently that institutions or policy, rules, and regulations have improved during 2008–10 (Global Competitiveness Index and Travel and Tourism Competitiveness Index), while inefficient Government bureaucracy has become somewhat less of a problematic factor of doing business (Global Competitiveness Index). Measured improvements in tourism infrastructure, environmental sustainability, and price competitiveness (Travel and Tourism Competitiveness Index) represent an important input to explaining both the average 7-percent growth rate for the tourism sector during 2007–10 and the low standard deviation over this four-year period. The gradually improved quality of Montenegro’s tourism product has been reflected in the stability of its demand base and is thus consistent with Montenegro’s 16-rank improvement to 36th (bringing Montenegro to Slovenia’s and Croatia’s immediate vicinity) in the Travel and Tourism Competitiveness Index 2011 (World Economic Forum, 2011), highlighting the fact that public institutions and good governance, macroeconomic and financial stability, as well as infrastructure improvements and environmental sustainability/protection form critical ingredients to Montenegro’s international competitiveness and its medium-term growth potential. 1.24 The business-climate-related aspects of the rule-of-law reforms demanded by the European Commission have given the (broader) competitiveness agenda a new impetus. The reconfirmed EU integration perspective in itself provides investors and businesses with assurances of strategic policy commitment to structural reforms. In addition, as the European Commission (2010) has made the opening of membership negotiations conditional on a list of seven policy priorities, viz., (i) a legislative framework for elections; (ii) public administration reforms; (iii) rule of law and reforms in the judiciary; (iv) anti-corruption; (v) fight against organized crime; (vi) media freedom; and (vii) anti-discrimination. The Government of Montenegro (2011) has agreed on—and harmonized with its European partners—a comprehensive, time-bound Action Plan for Monitoring the Implementation of Recommendations given in the EC Opinion. European Commission President Barroso (2011) encouraged Prime Minister Lukšić on his 100th day in office to implement the rule-of-law reforms, linking these to improvements in the business and investment climate: 20 Table 1.3: Montenegro and Neighboring Countries: Selected International Indices on Factors Affecting the Business and Investment Climate (Relative country ranks, with "1" being the highest-ranked country) Sample Unweighted Regional Averages Neighboring EU Member States (A) EU Candidate Countries (B) Potential EU Candidate Countries (C) Size MNE A B C1 BGR GRC ITA ROM SVN HRV MKD TUR ALB BIH RKS SRB Unweighted averages 2010 … 64 53 67 83 63 60 46 61 36 60 68 74 81 87 … 81 over all indices2 2009 … 66 51 66 86 63 50 44 62 34 62 66 72 88 90 … 80 2008 … 68 49 66 85 59 46 42 61 35 58 72 68 88 89 … 77 IMF WEO, October 2010, 2010 181 78 46 65 87 68 30 28 70 32 48 84 63 95 92 … 74 GDP (PPP), per capita 2009 181 75 44 65 86 67 26 28 68 32 47 82 65 94 92 … 73 2008 181 72 44 63 87 68 29 28 65 30 48 81 61 96 91 … 74 Reporters sans frontières, 2010 178 104 57 89 71 70 70 49 52 46 62 68 138 80 47 92 85 Press Freedom Index 2009 175 77 48 78 63 68 35 49 50 37 78 34 122 88 39 75 63 2008 173 53 42 63 60 59 31 44 47 30 45 42 102 79 36 58 64 Transparency International, 2010 178 69 63 60 85 73 78 67 69 27 62 62 56 87 91 110 78 Corruption Perception Index 2009 180 69 61 66 92 71 71 63 71 27 66 71 61 95 99 … 83 2008 180 85 56 64 87 72 57 55 70 26 62 72 58 85 92 … 85 UNDP, 2010 169 49 36 68 64 58 22 23 50 29 51 71 83 64 68 … 60 Human Development Index 2009 182 65 39 65 71 61 25 18 63 29 45 72 79 70 76 … 67 2008 179 64 36 63 70 56 18 19 62 26 45 68 76 69 75 … 65 World Bank, 2011 183 66 68 62 94 51 109 80 56 42 84 38 65 82 110 119 89 Doing Business 2010 183 65 64 62 94 51 97 76 54 43 89 36 60 81 110 118 90 2009 183 77 64 81 99 42 100 74 45 58 110 69 63 89 119 107 90 World Economic Forum, 2010-11 139 49 63 72 95 71 83 48 67 45 77 79 61 88 102 … 96 Global Competitiveness Index 2009-10 134 62 59 72 99 76 71 48 64 37 72 84 61 96 109 … 93 2008-09 135 65 60 71 100 76 67 49 68 42 61 89 63 108 107 … 85 World Economic Forum, 2010 125 43 55 54 69 78 55 51 54 35 45 56 62 59 80 … 67 Enabling Trade Index 2009 121 … 50 52 83 72 47 45 54 31 39 68 48 63 102 … 2008 118 … 43 54 81 60 36 33 57 31 42 81 38 73 89 … … World Economic Forum, 2011 139 36 40 53 83 48 29 27 63 33 34 76 50 71 97 … 82 Travel & Tourism Comp. Ind. 2009 133 52 41 57 95 50 24 28 66 35 34 80 56 90 107 … 88 2008 134 59 40 57 92 43 22 28 69 36 34 83 54 92 105 … 78 1 Excluding Kosovo. 2 Excluding WEF's Enabling Trade Index . 21 Figure 1.15. Montenegro: Selected Subcategories of Competitiveness Measures Sources: World Economic Forum; and World Bank. 22 It is important that the Government has elaborated a detailed Action Plan that focuses on the shortcomings identified in the Commission’s Opinion, which provides a clear road map for progress. The key now is implementation. Strong leadership and ownership are needed to address the priorities, in particular those related to the rule of law. We need to see concrete results in this crucial field —from fighting organized crime and corruption to strengthening judicial independence. This is also key to the business environment and thus in Montenegro’s own interest. While the EU’s policy priorities reflect, to a large extent, some of the key challenges identified in international competitiveness surveys, even the EU’s democracy agenda (as, e.g., its focus on the electoral framework and Parliament’s legislative and oversight role) comprises a clear economic component. Among other researchers in the currently very active field of institutional and political economics, Persson and Tabellini (2008) and Aghion, Alesina, and Trebbi (2008) have found (surprisingly) strong empirical growth effects from a democratic superstructure. E. CONCLUDING REMARKS 1.25 The sovereign debt crisis within the eurozone and in the US has added additional uncertainties to the world economy and, in turn, to Montenegro. Montenegro being a small, open, euroized economy that is dominated by a few, highly pro-cyclical sectors means that any deterioration in the global economic environment is posing considerable risks to the nascent economic recovery. These risks are exacerbated by Montenegro’s geographical proximity to Greece and Italy and the fact that market participants have to find a way to clear the very large (and still increasing) stock of inter- enterprise arrears. This assessment is unrelated to the fact that growth in 2010 has proven considerably more robust than initially projected. While the political response to the 2008–10 crisis was generally successful, stabilized the most vulnerable banking and industrial sectors, the Government has doubled the public debt-to-GDP ratio in its wake, leaving it largely without fiscal buffers to respond to any subsequent shocks. As a result, the challenges of ensuring fiscal sustainability and improving the efficiency of social spending define economic policymaking over the medium-term horizon. 1.26 In general, Montenegro’s economy has an enormous growth potential, but significant structural, economic, and fiscal risks—if left unaddressed—could tip the favorable outlook. Through the combination of additional efficiency gains in key sectors of the economy and the realization of large-scale, productivity-enhancing foreign-direct investments, Montenegro should be able to set its economy onto a trajectory of sustained economic growth so as to close the per-capita income gap with the European Union within a generation. However, large-scale projects with long-term amortization horizons will be realized only in a politico-economic environment that assures investors of a high degree of predictability and security, both in terms of macro-fiscal stability and the business climate. 1.27 Resisting suggestions to include (fundamental) changes in tax policies as one instrument in its fiscal and institutional adjustment strategy, Government will have to concentrate its efforts on expenditure cuts. In 2010, about one-half the budget was spent on the public-sector wage bill and pensions, exerting considerable pressure on other budgetary items. With few “discretionary” budget lines, it is clear that the loss of fiscal control over these two budget items alone threatens medium-term fiscal sustainability and macroeconomic stability as well as the Government’s ability to realize (productivity-raising) public investments. The overall economic environment necessitates a considerable degree of political maturity to be able to stem against the politically tempting, but fiscally harmful trend increases in the wage bill and social security similar to those that have been observed in recent years. 1.28 This PEIR (like its predecessor study) recommends implementing the fiscal-adjustment strategy by defining “fiscal benchmarks” so as to lengthen budgetary planning horizons. As such, 23 the Government should commit—ideally through legally binding constraints—to a pre-announced policy path. According to PEIR recommendations, (i) fiscal policy should aim at staying within the constraints of the Stability and Growth Pact, especially given implicit requirements to shadow the obligations of eurozone membership; (ii) recurrent expenditures should be financed from the regular tax base; and (iii) future budgets should aim at a gradually declining ratio of recurrent expenditure to GDP. The benefits of such a fiscal-policy approach, if adopted, would be the Government’s ability combine features of short - term budgetary flexibility with medium-term fiscal discipline. A fiscal framework that comprises the three aforementioned elements should result in: (i) an improvement in the grades conferred to Montenegro by international credit rating agencies; (ii) the gradual decline of sovereign risk premia; and (iii) lower interest rates for (public) debt. 1.29 The rapid increase in Montenegro’s the public debt-to-GDP ratio during 2009–11 and the vulnerabilities in the world economy suggest that the authorities would be well-advised to complement the formal adoption of (broader) “fiscal rules” with (tighter) “policy objectives”. The aforementioned fiscal rules will likely prove insufficient, given (i) the fiscal weakening that had occurred after the previous crisis; (ii) the sensitivity of Montenegro’s small, open economy to external shocks l; and (iii) the risk that a follow-up shock caused by the sovereign debt crisis in the Southern eurozone countries and/or the US could set off uncontrollable debt dynamics. To recreate the fiscal buffer and protect the market’s positive assessment of Montenegro’s creditworthiness, the authorities would benefit from a benchmark on public debt significantly below 60 percent. To allow for a comparable response to any future crisis, economic policymakers should outline a fiscal adjustment program anchored on a debt- to-GDP ceiling of 35±5 percent, with a limited response window that would allow Government increase debt during extraordinary periods but commit it to a post-crisis adjustment to pre-crisis levels within a pre-specified period of time. 1.30 The Government is conscious of the need to stem against the politically tempting, but fiscally harmful trend increases in the wage bill and social security. To this end, the Government has requested World Bank input into its efforts to develop a two-pronged strategy with which to (i) constrain the growth in expenditures; and (ii) increase the “value for money” in the public administration. In response, this follow-up Public Expenditure and Institutional Review (PEIR-2) seeks to contribute to the public debate on how to best modernize Montenegro’s public administration and create the legal, regulatory, and institutional foundation for a smaller and more efficient Government. 24 2. PENSIONS: THE SINGLE BIGGEST EXPENDITURE ITEM28 Long-term fiscal consolidation can only succeed if the pension system—on which about one-quarter of the budget is spent—is brought under control. Amplified by adversary demographics and an expected rapid increase in the old-age dependency ratio, the Government has to follow through on its envisaged path of gradually decreasing the expenditures of the Pension and Disability Insurance Fund (Fund PIO) as a share of GDP as well as budgetary expenditures and making pension reform a pivotal, long-term policy priority. Even if it might lead to “tactical” gains in the political debate or electoral campaigns, losing fiscal control over pension-related expenditures will crowd out the fiscal space for public investments and other budgetary spending of critical importance. The Government has implemented important reforms in recent years, including the gradual increase in the pension age to 67. However, the combination of a tight fiscal situation in the post-crisis era and the gradual aging of the society requires further pension reforms at the center of the broad structural reform agenda. A. INTRODUCTION 2.1. With total expenditures of €331 million (or 10.9 percent of GDP in 2010), the pension system is the single largest public spending program in Montenegro .29 This relative weight within the consolidated budget makes the financial sustainability of the pension system imperative to public finances in general. Losing control over pension-related expenditures would (gradually) crowd out the fiscal space for public investments and other budgetary spending of critical importance. As the main pillar of securing old-age income, adequate (relative and absolute) benefit levels are important from a welfare perspective as well as for ensuring political stability. However, to ensure long-term financial sustainability and benefit adequacy, Montenegro reformed its public pension system in 2004 and started to lay the foundations for a structural reform introducing privately managed —fully funded—pension schemes. The pension reform started a few years prior to independence, but the initial enthusiasm became engulfed by the economic boom: as the exceptionally favorable economic situation translated into fiscal surpluses, the budget could subsidize the pension system (where it should not have) and fill in—without too much difficulty—revenue gaps (which removed a considerable amount of reform pressure). The turning point came with crisis, which required a substantial tightening of fiscal policies, bringing questions of policy priorities to the fore. In this altered economic environment, pension reforms have been placed back at the very center of the political debate, and this has already led to the—very welcome and foresighted—enactment of further adjustments to the mandatory pension system in late 2010, including the gradual increase of the pension age to 67 years of age. 2.2. Whereas the public system’s basic characteristics are in line with current international practice, further parametric adjustments will be necessary to contain fiscal costs and react to demographic changes. This will include a policy revision to the privately managed, fully funded schemes to reflect domestic capital market developments as well as institutional and fiscal constraints. Central to the reform efforts will be the need to safeguard the consistency of the pension policy and not to sacrifice it for the sake of addressing temporary (short-term) labor market tensions or (even shorter 28 Prepared by Snježana Plevko (ECSH3) and Csaba Fehér (Consultant). 29 Over the medium to long term, the pressures from the pension system will increase. Chawla et al. (2007), e.g., point to the inherent policy challenges in societies with declining and aging populations, amplified by a combination of low labor force participation of the population aged 50 –64 and a very large increase in the old-age dependency ratios (partly reflecting the increase in life expectation). As these projected trends get confirmed over the medium term, still deeper and more comprehensive reforms will need to be gradually prepared and implemented to ensure the pension system’s sustained viability. 25 term) political objectives. Like elsewhere, all pension-related decisions will have to be made in the full knowledge of the likely fiscal consequences over periods far exceeding any medium-term budgetary framework. In this, it is pivotal that a mechanism is found to ensure that all decision-makers (parliamentarians in particular) and the public at large will be made fully aware of the long-term costs and consequences of policy proposals and their alternatives. Inherent “political economy” challenges stem from the ability to provide the electorate with generous promises, of which the fiscal effects will be borne in the long term by future Governments. All proposals to modify pension legislation should, therefore, be accompanied by long-term fiscal impact analyses and, possibly, be subjected to a legally defined set of fiscal criteria. 2.3. In previous years, Government has tended to budget pension-related receipts and expenditures carefully, resulting in lower-than-forecast budgetary subsidies (except for the crisis year 2009). Still, a trend increase in related expenditures by the Pension and Disability Insurance Fund (Fund PIO), both relative to total Government spending and GDP), has been evident in recent years (Table 2.1), which the Government seeks to reverse over the time span of the current Economic and Financial Programme 2012–15 by stabilizing contribution receipts at a level of about 7.1 percent of GDP and gradual reduction of Fund PIO expenditures from an estimated 11 percent of GDP in 2011 to 9.9 percent in 2015. This implies a gradual erosion of the real value in pension benefits and/or a stricter definition of eligibility criteria. Table 2.1. Fund PIO Receipts and Expenditures and Government Subsidies, 2006–15 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Ministry of Finance projections1 (In millions of euro) Receipts from contributions 140.4 176.7 213.9 199.5 233.5 218.4 230.9 244.9 257.0 269.7 Subsidies from the budget 54.3 58.8 37.1 124.0 97.5 124.7 130.8 122.8 116.8 110.0 Expenditure 194.7 235.4 250.9 323.5 331.0 343.0 361.7 367.7 373.7 379.7 (In percent of general government expenditures) Receipts from contributions 15.6 15.4 13.8 14.0 16.8 15.4 15.8 16.2 16.2 16.2 Subsidies from the budget 6.0 5.1 2.4 8.7 7.0 8.8 9.0 8.1 7.4 6.6 Expenditure 21.6 20.5 16.2 22.8 23.8 24.2 24.8 24.3 23.5 22.8 (In percent of GDP) Receipts from contributions 6.5 6.6 6.9 6.7 7.7 7.0 7.1 7.1 7.1 7.1 Subsidies from the budget 2.5 2.2 1.2 4.2 3.2 4.0 4.0 3.6 3.2 2.9 Expenditure 9.1 8.8 8.1 10.9 10.9 11.0 11.1 10.7 10.4 9.9 Memorandum item: Forecast error in budget subsidies 2 In millions of euro -14.8 -10.3 -6.2 9.6 -24.9 … … … … … In percent of total expenditure -1.6 -0.9 -0.4 0.7 -1.8 … … … … … In pecent of GDP -0.7 -0.4 -0.2 0.3 -0.8 … … … … … 1 Projections for pension contributions and expenditure are based on the 2011 budget and the Economic and Financial Programme 2012-15; otherwise World Bank staff estimates and projections. 2 Actual net of initially budgeted subidies (+ = higher-than-budgeted subsidies for old-age and invalidity pensions). Sources: Ministry of Finance; and World Bank staff estimates and projections. 26 B. KEY CHARACTERISTICS OF THE PENSION SYSTEM Mandatory Pension System 2.4. Montenegro’s pension system consists of mandatory social-insurance (first-pillar) and privately managed voluntary (third-pillar) pension schemes. The introduction of mandatory (second- pillar) private-pension schemes had been envisaged to commence in 2013, but fiscal considerations— together with discouraging lessons from other Eastern European countries —will likely delay the launch of second-pillar schemes. Social insurance is universal, meaning that every person entering the labor force is insured. Inversely, all labor income as well as certain transfers designed to partially replace labor income (such as unemployment or maternity benefits) are subject to contribution liability. The system is uniform in the sense that no special branches of social insurance are operated for certain occupation groups (e.g., farmers, civil servants, or the armed services). The social-insurance pension scheme is administered by the Fund PIO, although—as part of a large scale modernization project supported by the World Bank—registration of insured contributors and their employment status, as well as the collection and enforcement of contribution and ensuring reporting compliance, has been transferred to the Tax Authority in early 2011. The pension system, in line with traditional social security arrangements, pays old age pension, early retirement old age pension, various full and partial disability pensions, survivor benefits to orphans and widow(er)s as well as work injury compensations and funeral allowance. Following the country’s independence, pensions in service were continued and accrued pension rights were recognized with regard to other former Yugoslav republics. Montenegro has signed bilateral social security agreements with other states of the former Yugoslavia, providing for the orderly translation of multiple and extraterritorial service histories into pension entitlements received from different sources. 2.5. Since 2011, the Government has begun to gradually increase the retirement age, irrespective of gender, to 67 years of age. Eligibility for old age pensions is conditional on reaching retirement ages, and on accruing qualifying service period. In 2010, the Government introduced gradual change of the retirement age and qualifying period. The retirement age for men will increase two calendar months per year and reach 67 by 2025. For women, the retirement age will increase by three months per year and reach 67 by 2041. The qualifying period, at the same time, will remain at the current 15 years of work history. However, in case of work histories of at least 40 years, retirement remains available without early-retirement reductions; this effectively weakens the retirement-age increase. Hazardous and hard-work conditions result in accelerated accrual of work histories, i.e., one month of actual employment counts more than one month of work history. The rate of acceleration ranges between 14 months and 18 months of work history per 12 month of employment. Hazardous work periods also allow early retirement with 30 years of service, of which 20 years in hazardous occupation. Consequently, for an 18-year old employee in hazardous occupation, the retirement age can be reduced below 50 years of age. 2.6. Eligibility for disability pension depends on the extent of lost capacity and accumulated service history. While the service history criterion is absolute in the case of old age pensions, it is a function of age in for disability pensioners: younger contributors need shorter service years to qualify. As part of the 2004 reform, it is the remaining general capacity to work that is assessed as opposed to the former regime where capacity to perform the applicant’s previous job which determined eligibility. As a result of stricter assessment procedures, disability uptake has shrunk since 2004 and disability pensioners now make up a smaller portion of beneficiaries than before. Eligibility for survivor pension depends on the survivor’s age, educational status (for orphans), marital status, the deceased person’s benefit or work history and is in line with international best practice. 2.7. Entry pensions are assessed on the basis of pension points which measure the relative contribution intensity. Every contribution period results in a pension point less than, equal to, or greater 27 than unity, depending on the individual’s relative wage in the same contribution period (service year). At retirement, all pension points are averaged and multiplied by the number of service years. Average pension points for the entire calculation period cannot exceed 4 or fall below of 0.5. The resulting points are translated, using a legally set monetary value, into an entry pension. The period over which pension points are averaged is gradually increasing from the best consecutive 12 years (2004) to 40 years (2018). As the assessment period approximates a full career, the average points increasingly reflect lower earning years of a contributor’s career, thereby reducing the average point value per service year, which, over time, results in lower entry pensions. 2.8. Total pension points are translated into an entry pension using pension point values. In 2004, at the time of introducing the current system, the value of a pension point was defined as a percentage of average wages (1.42 percent). Later, it was redefined as a nominal euro amount, regularly indexed in the same manner as benefits in service. In 2010, the value of a pension point amounted to €6.08. In addition to minimum pension determined by the average minimum point value of 0.5, the pension system provides for a lump-sum, minimum benefit, set at €92 in 2010 and payable to all pensioners who meet the minimum qualifying conditions (retirement age and service years). The lump- sum minimum benefit is applicable to old age and survivor pensioners as well and is regularly indexed in the same manner as pension points. Since the lump-sum minimum benefit is not conditional on the level of contributions paid, its level is the same for farmers, too, despite farmers’ significantly lower contribution rates. There is no explicit benefit ceiling; however, the relative maximum is determined by the limit of 4 average pension points. Table 2.2. Comparison of Various Indexation Methods Between the minimum pension and Country Indexation Method the theoretically possible maximum, entry benefits are Austria De facto combination of prices and wages (price plus irregular real increases) linear both in service Belgium Price indexation years and relative Croatia Nominal wage indexation wages. This linearity Czech Republic Roughly equivalent to 70 percent prices and 30 percent wages is only manifesting Finland 80 percent prices and 20 percent wages since 1995 itself gradually, as France Price indexation retirees from the Germany Net wage indexation reformed system Hungary Minimum CPI, maximum 50 percent prices and 50 percent wages, depending on GDP growth replace the stock of Ireland Price indexation pensioners whose Italy GDP growth indexation benefits were set Latvia Price indexation under the pre-2004 Luxemburg 50 percent prices and 50 percent wages regulations. Poland 80 percent prices and 20 percent wages Portugal Price indexation (plus irregular increases) 2.9. Pension- Slovenia Net wage indexation indexation Spain 75 percent price and 25 percent wages (only price in 1996) provisions have been Switzerland 50 percent prices and 50 percent wages changed in 2010, United Kingdom Price indexation from adjustments twice a year by a retrospective30 composite index of 50 percent wage growth and 50 percent consumer price index (“Swiss indexation”) to a formula with a 25 percent weight on wages and 75 percent on inflation; see also Table 30 Indexation based on past observations—in this case, indices of the previous year—is retrospective, as opposed to indexation regimes based on expectations concerning future periods (typically, the budget year). Sources: National authorities. 28 2.2. In order to ensure that the same relative contribution performance results in the same relative pension compared both within and across cohorts (i.e., regardless of the year of retirement), the monetary value of pension points is valorized in the same manner as benefits already in service. While the decision to move more towards CPI indexation is welcome in terms of budgetary affordability, it would dampen initial replacement rates further and, consequently, reduce long-run pension adequacy. Private pension schemes 2.10. The institutional environment does not support the introduction of a second-pillar pension system (yet). Similarly to other countries and Central and Eastern Europe, the pension reform process (which begun in 2004) envisioned a three-pillar pension system consisting of a reformed PAYGO public pension scheme, a mandatory second pillar of privately managed defined contribution schemes operating in a competitive market and a voluntary pension pillar, operating along principles similar to those of the second pillar but collecting voluntarily paid, additional contributions. The process was to start with reforms to the public scheme and introduce third-pillar schemes subsequently. Mandatory second-pillar schemes were to be introduced only after the enabling environment was present, with established financial-service providers, well-functioning regulators and financially literate citizens already familiar with the principles and operations of private pension schemes as demonstrated by voluntary schemes. Such enabling conditions include a reasonable supply of securities—fixed income, equity, and pooled instruments—and a deep, liquid and well-regulated market where transactions can be safely executed, prices readily discovered, and asset values established accordingly. In case of a small, open, euro-based economy, controlled access to overseas markets is among the crucial preconditions. In terms of institutional development, the presence of investment professionals, banks offering custody services, scheme administrators, auditors is crucial as well as that of experienced regulators. Finally, introducing mandatory pension schemes involves redirecting some part of mandatory pension contributions away from the public schemes and into individual accounts. Since the public scheme’s obligations reflect their contribution revenues only as a general trend with multi-decade lag, introducing a mandatory private- pension pillar generates a financing gap which may be addressed by higher deficits and correspondingly higher debt, increased public revenues or reduced Government expenditures. Introducing a second pillar requires additional resources in the short and medium run and may only result in an overall reduction of the implicit pension debt in the long run. 2.11. The Law on Voluntary Pension Schemes was enacted in 2004, but the industry has not yet established itself. Currently, there are two voluntary pension schemes operating, with approximately 1,700 members and assets of about €150,000. Assets are kept mainly in bank deposits and cash (40 percent), and securitized Government debt (45 percent) while equity represents a small portion of the industry’s portfolio (11 percent). The development of the voluntary pension market is hindered by both demand and supply side constraints. In terms of demand, the main issue appears to be the lack of tax advantages. Unlike other forms of investments, pension savings are inaccessible until retirement and their use is tied entirely—or, at least, partly—to particular types of financial products such as annuities or other welfare-optimizing withdrawals. These constraints on liquidity are perceived as welfare-losses and need to be countered, the most common mode of which is favorable tax treatment of voluntary contributions. Currently, no tax deductibility or other favored tax treatment is available for voluntary pension savings in Montenegro. The situation is further aggravated by the lack of a domestic investment fund market which would foster financial literacy and nurture the financial services involved in providing private pension products. Given Montenegro’s use of the euro, savers’ easy access to foreign financial services (either directly or through agents operating in Montenegro), their preference for bank deposits and real estate investments, the limited domestic supply of securities, it is unlikely that institutional investors would be able to generate the depth, liquidity, diversity and safety needed for developing a pension market. 29 2.12. The introduction of mandatory second-pillar pension schemes appears premature, given the fiscal and financial-sector constraints. The absence of a robust voluntary private-pension pillar denies the Government, the contributors, and the industry the necessary experience. Thus, the Government may wish to formally revisit its strategy concerning a fully funded second pillar. Mandatory private-pension savings is directly linked with fiscal and social policies, for which reason the reliance on foreign schemes and regulators is less desirable than in the case of voluntary savings. At the same time, even under the best of conditions, the emergence of a competitive market of mandatory pension schemes and providers cannot be expected due to economies of scale considerations. Thus, a second pillar may only be a viable proposition in Montenegro if some level of centralization and unification is enforced. In this respect, the Swedish experience—where scale economies are realized at the scheme administration level while independent asset managers may compete for commercially viable portfolios—may offer a useful example. At the same time, the experience of Kosovo is also worth considering: the country, in light of its small contribution base, nonexistent capital markets and underdeveloped financial sector, decided to create a single pension scheme that is managed by an international board of trustees. It placed the management of the accrued assets with seasoned overseas asset managers. Singapore has chosen a similar route. 2.13. From a fiscal perspective, it is important to note that international best practice suggests that contribution rates to privately managed pension schemes of less than 5 percent are unsustainable. In those instances, the fixed and variable costs of scheme administration, asset management as well as the cost of annuitization at retirement reduce net returns realized on individual accounts significantly—below alternative saving instruments and even below inflation. Diverting 5 percent of gross wages to private schemes will deny Fund PIO 25 percent of the contribution revenues expected from members enlisting in the new, multi-pillar system. There is a trade-off between the size and dynamics of the financing gap—or additional deficit—attributable to the reform and corresponding to the share of contributors joining and contributions directed to the second pillar versus the speed of the reform: a faster shift to the new regime results in greater deficits. At the same time, in present value terms, the total cost of a prolonged reform (such as one only allowing new labor market entrants into the new system) is higher. C. PENSION SYSTEM’S PERFORMANCE 2.14. Montenegro’s population will continue to age, implying that every employed person will have to support an increasingly larger number of pensioners. Its population is about 0.63 million (2011), of whom 14½ percent are above the age of 65 years, against 25¼ between the ages of 0 and 19 and the remaining 60¼ percent at a working-age level (Figure 2.1). The country’s demographic situation is marginally better than other countries in the region, but the problems facing Montenegro are similar: population aging stemming from a below-replacement fertility rate of 1.7 and increasing age-specific life expectancies. According to UN population projections, the country’s old age dependency ratio—defined as the ratio of the population 65 years and older to those between 20 and 64 years of age—has increased from 13.7 percent in 1985 to 20.2 percent in 2010 (Table 2.1), and it is expected to deteriorate further to one-third by 2035 (Figure 2.2). 30 Figure 2.1. Montenegro: Demographic Trends, 1950–2100 Source: United Nations, World Population Prospects: 2010 Revision. Table 2.3. Old-Age Dependency Ratios, 1950–2010 Old-age dependency ratio (Ratio of population aged 65+ per 100 population 20-64) 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 Montenegro 16.4 15.1 14.6 14.9 15.5 15.6 15.2 13.7 14.4 15.1 17.2 21.3 20.2 Southern Europe 13.6 14.0 14.8 16.1 18.0 19.7 21.2 20.4 22.3 24.7 27.0 28.3 29.1 Albania 15.8 14.4 11.8 10.9 10.3 10.2 11.0 10.6 10.2 12.0 14.0 15.2 16.4 Bosnia and Herzegovina 8.7 7.6 7.0 8.2 9.5 10.4 10.8 9.3 10.0 13.1 18.0 21.6 21.6 Croatia 14.1 13.0 12.7 14.5 16.6 18.5 19.5 16.9 18.5 21.9 25.8 27.9 27.8 Greece 12.5 13.2 14.3 15.5 19.7 21.8 23.2 23.1 23.0 25.3 27.5 29.7 29.9 Italy 14.3 15.3 16.4 17.7 19.4 21.7 23.7 22.1 24.5 26.9 29.4 32.0 33.5 Macedonia, FYR 18.7 14.2 11.0 10.1 11.3 11.6 12.5 11.4 12.8 15.0 16.9 18.0 18.7 Malta 11.5 14.2 15.2 16.3 16.6 16.3 16.5 16.0 16.6 18.2 18.8 21.5 21.9 Portugal 13.0 13.7 14.7 15.9 18.2 19.6 21.2 21.6 23.7 25.2 26.6 27.7 29.1 Serbia 14.1 12.3 11.3 12.8 14.7 15.9 16.9 14.4 16.1 19.8 23.2 24.7 23.3 Slovenia 12.5 12.8 13.7 15.9 17.3 19.0 19.8 17.2 18.3 20.4 22.3 24.3 25.5 Spain 12.8 13.5 14.5 15.8 17.7 19.2 20.6 21.4 23.6 25.7 27.3 26.4 26.8 Source: United Nations, World Population Prospects: 2010 Revision. 31 Figure 2.2. UN Projections of Old-Age Dependency Ratio in Montenegro, 2010–2100 Source: United Nations, World Population Prospects: 2010 Revision. Beneficiaries and benefits 2.15. A pension system’s beneficiary-to-contributor ratio tends to be considerably worse than the overall old-age dependency ratio. In the case of public pension schemes with universal coverage, a country’s old age dependency ratio as determined by its natural demographics presents a more favorable picture than the public pension scheme’s own system demographics, as described by the ratio of beneficiaries to contributors. The difference is determined by the actual age of entry into the labor force, economic activity ratios, the share of contributors within the economically active population, the effective retirement age, and the number of working age people economically inactive on account of retirement (typically, invalids). In theory and over the long run, the share of old age pensioners may also diverge as a consequence of insufficient contribution histories and a higher propensity after reaching retirement age (but this is not yet a consideration in the Montenegrin case). In 2009, the number of contributors to Fund PIO was approximately 160,000, representing 41 percent of the 20 –64 age groups. The system dependency ratio of the pension system was, in the same year 0.6631, meaning that there were almost 7 beneficiaries for every 10 contributors. The gap between the system dependency ratio and the demographic old-age dependency ratio is driven by low activity rates and the growth of beneficiary numbers over the past three years. 31 Including approximately 7,000 pensioners in other former Yugoslav republics. 32 2.16. In April 2010, there were 97,000 Montenegrin resident pensioners receiving old age, invalidity or survivor benefits from Fund PIO. Of this number, 46.4 percent received old-age pension and 24.8 percent was paid disability benefits, the remainder received young (orphan) or old (widow/widower) survivor benefits. The number of beneficiaries has grown substantially over the last 10 years: while the total number of beneficiaries was 86,000 in 2001, it grew by more than 19 percent to over 97,000 by early 2010. The total increase disguises an important structural shift, which renders the system demographic situation even more challenging. In 2001, the share of invalidity pensioners was 31 percent of the total, against a 41 percent share of old age beneficiaries, by the beginning of this year the proportion of old age pensioners exceeded 46 percent. This shift is the result of administrative efforts to curtail invalidity pension uptake and the pronounced increase of 40 percent within 10 years in the number of old age pensioners. 2.17. The growth of old-age pensioner population is one of the two primary reasons for the unsustainable trend in pension expenditures (Figure 2.3). Old-age pensioners cannot be “un-retired” (as opposed to invalidity pensioners who may be submitted to regular reviews and possibly reintegrated into the labor force). The number and proportion of pensioners within the population can therefore only be reduced through the slowing down of the inflow of pensioners. Since the stock of pensioners cannot be reduced (apart from the marginal effect of re-assessing young recipients of invalidity pensions), the main avenue of achieving this goal is increasing the effective retirement age through legislating a higher statutory retirement age and curtailing the proportion of retirees eligible for benefits earlier. When designing the new pension law, the Government took note of the fact that the retirement age of men was already 65 years (and that previous years’ beneficiary inflows had already r educed the potential impact of increasing the effective retirement age), the most obvious measures was the increase of women’s retirement age. For both men and women, the pension retirement age is now gradually increased to 67 years of age. Figure 2.3. Montenegro: Number of Pensioners, 1999–2009 100,000 90,000 80,000 70,000 60,000 50,000 40,000 30,000 20,000 Old age Invalidity Survivor 10,000 0 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 Source: Fund PIO. 33 2.18. The distribution of pensions remained broadly constant in recent years as pension benefits, including the minimum pension, were indexed in a uniform manner. The dispersion of entry pensions around their mean remained similar. The average net32 benefit in 2009 stood at €257, representing 55½ percent of the average wage. Average old-age benefits were 20 percent above this level at €306, while the average invalidity and survivor benefits were 10 percent and 18 percent lower, at €229 and €204, respectively. The average old-age benefit replacement rate is comparable to those observed in Serbia and above regional comparators. Principally, they do not stand out in comparison to other European countries. The distribution of benefits, shown in Figure 2.4, shows that benefits are skewed to the left, with invalidity and survivor pensions more heavily affected by the generally applicable minimum pension. Figure 2.4. Distribution of Old-Age Benefits, 2009 Source: Fund PIO. 2.19. New old-age benefits are calculated on the basis of relative contribution performance. The average annual pension point earned through contributions is equal to the ratio of the individual’s annual contribution liable earning to the nationwide average wage, up to a ceiling of 4 (i.e., four times the average wage) and a floor of 0.5. The value of one pension point was set at €2.97 in 2004 and, through composite indexation similar to that of pension benefits, has increased to €6.08 by 2010, representing a nominal increase of 100 percent in 6 years; see Figure 2.5. 32 Pensions are not liable to personal income tax. 34 2.20. This marked increase was driven by a high real-wage growth and two ad-hoc point value increases in 2008 and 2009. In addition to what was legally required, in 2008 and 2009 point value was raised by an additional 10 percent. At the end of 2009, when real wage growth turned slightly negative, the point value was also adjusted downward; see Figure 2.6. It must also be noted that Pension in service can be adjusted downward to respond to negative wage growth. Figure 2.5. Value of a Pension Point (in Euro), January 2004 – January 2010 Source: Fund PIO. Figure 2.6. Wage, Inflation, and Pension Developments, 2004–10. Sources: MONSTAT; and Fund PIO. 35 --------------------------------------------------------------------------------------------------------------------------- Box 2.1: An intuitive explanation of the “value point” system The „value point system“ determines the level of individual pensions and their indexation. Regarding the level of pensions, the value of pension is calculated as a product of personal points of the insurer (LB) and the value of pension per personal point on the day of retirement (VPLB): P = LB x VPLB Personal point (LB) is the product of personal coefficient of the insurer (LK) and her years of service (PS): LB = LK x PS Personal coefficient is the sum of annual personal coefficients divided by the period for which they are calculated. Annual personal coefficient is the ratio of the insurer’s salary (i.e., pension basis in a given year ) and the average annual salary in the country for the year. Regarding the changes in pensions, the value points are indexed to inflation (75%) and the average increase in salaries in the country (25%). This results in a gradual decline in the ratio of pensions and salaries and in medium term. As a result, younger generations of retirees will have lower benefits from older retires in relation to their contributions. Source: The Central Bank of Montenegro. 2.21. The dynamics observed with point values can also be seen with pension indices for the same period. The consequence of high wage growth is that pensions awarded in 2004 or before have a purchasing power 64 percent more than in 2004. (This statement is not equivalent to comparing average pensions, as the latter would also include the effect of attrition which pushes the index of average pensions even higher. This may change later on as workers with more fragmented contribution histories and consequently lower individual pensions start replacing people who retired after a life of full employment. This effect of shortening average contribution histories is not felt yet and may never present itself, depending on labor market conditions. 2.22. Disability benefits are calculated in a similar manner to old-age pensions. The major difference consists of a calculation, according to which missing service periods caused by the onset of invalidity are assumed to have been accrued at two-thirds until age 55 and one-half between 55 and retirement age. This approach is equal to the assumption that people under 55 work, on average, 2 years out of every 3 between entering the labor force and 55 years of age and 1 year out of every 2 thereafter. Pensions based on work-related disability receive more favorable treatment through the assumption of a 40 year work history. Due to the nature of benefit assessment rules which cannot take into account the difference between average wage growth and the individual's career wage profile, the distribution of young beneficiaries' invalidity pensions is more compressed around the average and is generally lower than that of people awarded an invalidity pension at a later age. 2.23. Survivor pensions are determined on the basis of the deceased persons benefit or accrued rights (if the death occurred before retirement), the age and lineal relationship of the survivor to the deceased and the number of eligible survivors. If a survivor is over 50 at the time of being awarded a survivor benefit or reaches that age while in receipt of the benefit, the person remains eligible for the rest of his/her life. One person can only receive one benefit with the choice of benefit resting with recipient. Current rules are in line with best practice while uptake and expenditure levels are determined partly by the dynamics of old age and disability pensions and partly by the process of women eligible for a sufficiently high old age pension gradually replacing women with more traditional life paths of fragmented or short contribution histories. 36 2.24. In addition to service-related minimum pension, the Montenegrin pension system includes a lump-sum statutory minimum pension which is conditional on meeting the minimum service history criteria but is unrelated to the level of contributions paid. The minimum lump-sum pension was €45 in 2004 and was indexed in the same manner as all other pension benefits, reaching €92 in 2009. The minimum pension provision applies equally to disability benefits. Expenditures and revenues 2.25. The Fund PIO’s total expenditures amounted to €364 million in 2009, of which about 90 percent (€329 million) were spent on benefits and health insurance contributions on behalf of pensioners. As shown in Figure 2.7, pension expenditure grew dynamically in euro terms, more than doubling in nominal terms. As a share of GDP, pension expenditures were lowest following the high- growth in 2006 and 2007 when high wage growth was not yet reflected incorporated into benefit levels. Since the growth rates observed in the middle of the decade, it is expected the GDP-proportional pension spending will stabilize around 11 percent of GDP, comparable to other middle income transition economies. 2.26. Pension expenditures comprise of benefits paid to pensioners and health contributions transferred to the HIF). Health contributions were paid at the rate of 19 percent on the basis of pension benefits. However, from January 2010, the contribution rate was reduced to 1 percent. This measure is expected to increase transparency in public finances through reducing cross-subsidies between budgetary entities. At the same time, this may lead to a revenue shortfall at the HIF in return for a nominal reduction of Fund PIO expenditures. Figure 2.7. Fund PIO Expenditures, 2004–09 Source: Fund PIO 37 2.27. Administrative expenses represented 2.2 percent of total expenditures in 2009 which is low compared to other public scheme operating in countries of similar populations. Indeed, operating expenses have been kept broadly constant at around €7–7.5 million. This was possible via strict controls over the volume of wages in a high-wage growth environment and reducing non-wage expenses. In this respect, Fund PIO made efforts to use the relatively expensive postal payment services and rely, instead, on commercial banks. 2.28. The pension system is universal and contributory, thus every employed person is required to pay contributions. Over the past years, the number of contributors closely mirrored employment, which, in return, followed the post-independence boom and, later, the economic crisis. The Montenegrin economy relies heavily on tourism and related services; this makes employment and contribution revenues volatile, not least because key industries are labor intensive. Consequently, with the economic crisis, pension finances had to rely more on direct budget subsidies than before. The current contribution rate is 20.5 percent, comprising a 5.5 percent employer and a 15 percent employee contribution. Higher employer contribution rates apply if and while a contributor is employed under conditions which count towards early retirement eligibility or accelerated accumulation of service periods. Expenditures are entirely pay-as-you-go financed without recourse to actuarial or liquidity reserves 33. Fund PIO revenues comprise contribution revenues (55 percent of total in 2009), budget transfers (44 percent in 2009) and other revenues. The crisis year 2009 was extraordinary in terms of the revenue shortfall and the resulting budget transfer; however, contribution revenues never covered more than 74 percent of Fund PIO expenditures (2008) and budget transfer never went below 22 percent. Whereas expenditures are slowly trending with little room for volatility—as they are determined by the stock of pensioners and the average benefit which are only marginally influenced by the net effect of attrition—, revenues reflect economic cyclicality, seasonality and shocks. It must be noted, however, that the continuous need for substantial budget support points at imbalances which go beyond mere cyclicality; see Figure 2.8. 33 PIO had reserves of approximately one month’s benefit payments but this reserve was consolidated into the budget from 2010 when PIO ceased to function as an extrabudgetary fund. 38 Figure 2.8. Source of Fund PIO Revenues 2006–09 Figure 5.8. Source of PIO Revenues 2006-2009 400,000,000 Contribution revenues 350,000,000 Budget transfers 300,000,000 Other income 250,000,000 200,000,000 150,000,000 100,000,000 50,000,000 0 2006 2007 2008 2009 Source: Fund PIO. 2.29. These imbalances are partly due to conscious policy choices, including the lowering of contribution rates (done to improve competitiveness). Reducing the cost of labor by closing the labor tax wedge—thus contributing to the country’s competitiveness and reducing the informal economy—has been a policy pursued by Government since 2006. Pension contributions were cut from 24 percent to 20.5 percent, while the relative share of employer vs. employee contributions have also been revisited, changing from a 12–12 split of the respective employer-employee rates to one of 5–15.5. 2.30. The contribution base is the same as the base for the personal income tax, making collection and the reconciliation of data and money flows more effective. The contribution base is broad, including all the typical categories of remuneration and allowances and is defined on a gross basis. Further extension of the contribution base through revisiting its definition has little scope. At the same time, gradually increasing the contribution base of farmers to the statutory minimum wage is an option which may be explored. Given the small number of contributors registered as farmers, the immediate revenue impact would be small. However, further uniformity in the contribution base could help to avoid a potential wage-tax arbitrage observed in other transition economies that operate different contribution bases for farmers and agricultural workers. In terms of establishing the contribution base, Montenegro relies on self-declarations, although traces of presumptive taxation exist in the form of minimum contribution bases for the self-employed and for farmers (including people working in the fishing and forestry subsectors): the minimum contribution base for those self-employed subject to lump- sum taxation; the base is 50 percent of the economy-wide average wage while the minimum is set at 20 percent of the average wage for farmers. 39 2.31. The types of income and remunerations in the contribution base are harmonized across health insurance, pension, and disability insurance as well as the personal-income tax base, leaving little room for further expansion34. Contributions for pension and disability contributions are levied at a uniform rate of 20.5 percent. In the case of employees and income earned from employment (as opposed to entrepreneurial income), contributions are split between the employer and the employee, at rates of 5.5 percent and 15 percent, respectively. In addition to the standard rate, further contributions are to be paid by the employer if the nature of the work qualifies the employee to accrue contribution histories at an accelerated rate. Additional contributions are payable at rates ranging between 6 percent and 18 percent. 2.32. From March 2010, annual and monthly returns are filed with and contributions are collected by the Tax Authority. Employment, wage and contribution data are relayed to Fund PIO for further processing, archiving, and benefit calculations, while contributions paid towards the single treasury account are credited onto institutional subaccounts on a daily basis. All employment and contribution records are kept electronically, going back to the early 1970s, making benefit-assessment response times short (typically 30 days or less). The system of registering contributors, contribution- liable earnings, and contributions is currently undergoing a major reform supported by the World Bank (PSAIP), completed in September 2010. The new system introduces a unified reporting system across the three branches of social security (pensions, health, and unemployment) and requires both the redesigning and unification of the current system of individual and business identifiers as well as ensuring that existing databases can communicate on the basis of the new identifiers and the unified reporting forms. 2.33. From 2010, Fund PIO ceased to function as an extra-budgetary fund and is now only responsible for administration of pension benefits. Under the new regime, Fund PIO will only be responsible for planning pension expenditures while revenue planning will rest with the Ministry of Finance. Beyond the annual budget, the Ministry also prepares a rolling, 3-year medium-term budget plan which includes Fund PIO. It is unclear, however, on what basis pension expenditures are planned and, especially, how the effect of new pensioners replacing existent ones is accounted for. Neither the Fund PIO nor other Government agencies prepare long-term pension projections which would help to establish the balance sheet effect of current and alternative policy scenarios. The Fund PIO lacks the necessary staff to produce long-term projections or to help legislative work through assessing the medium and long-term impact of proposed legislative changes. These capacities are crucial if the Government wishes to manage the pension system's parameters in a manner which improves sustainability, and improves contribution financing. Since January 2010 Fund PIO is financed through the single treasury account and has no role in revenue or liquidity management. D. ISSUES AND RECOMMENDATIONS 2.34. Fiscal consolidation can only succeed if the pension system—on which about one-quarter of the budget is spent—is brought under control. Amplified by adversary demographics and an expected rapid increase in the old-age dependency ratio, Government has to follow through on its envisaged path of gradually decreasing the expenditures by the Pension and Disability Insurance Fund (Fund PIO) as a share of GDP and budgetary expenditure, making pension reform—like elsewhere in Europe—a pivotal, long-term policy priority. Even if it might lead to “tactical” gains in the political debate or electoral campaigns, losing control over pension-related expenditures will crowd out the fiscal space for public investments and other budgetary spending of critical importance. Important reforms, 34 The only category left out of the income and wage tax base is the holiday supplement/support while other regular and discretionary earnings -from multiple and further jobs- are included. 40 including the gradual increase in the pension age to 67, have been implemented in recent years, but the combination of a tight fiscal situation in the post-crisis era and the gradual aging of the society places related reforms at the very center of a structural reform agenda —notwithstanding the indications that Montenegro’s welfare system has been relatively successful in protecting the elderly against poverty and to ensure that relative contribution performance is reflected by pension benefits.35 The pension system’s financial position is unsustainable, with contribution revenues falling systematically below benefit expenditures and Government being unable to keep on subsidizing the pension system in the order of about 4 percent of GDP (as it has been doing during 2009–11). 2.35. Fiscal problems can be addressed through constraining entry benefit levels, benefit indexation, and—to some extent—the inflow of new beneficiaries. In this respect, two areas merit particular attention, benefit indexation and the valorization of pension point values. In terms of indexation, Government has decided to link benefits mostly to inflation and not to increases in wage rates. This policy, which will limit expenditure growth, is in line with (emerging) international practice. It has to be understood that this measure will only have a noticeable impact over a multi-year period, but—increasingly important—compound effects will help the Government to achieve its fiscal consolidation objectives in this area. In terms of benefit adequacy, price indexation ensures that beneficiaries can continue to purchase the same basket of goods and services but are excluded from the benefits of real wage growth rates that occurred after their retirement. However, valorization with such a high weight of price inflation dampens initial replacement rates and, consequently, reduces long-run pension adequacy. 2.36. Indexation rules should not be interpreted as a minimum by which pension benefits increase. In case of periods with prolonged high and sustainable growth rates, when additional increases appear possible, Ad-hoc indexations and point value adjustments should be avoided as these introduce arbitrariness in pensions and contingent fiscal cost. 2.37. The main advantage of a point system is that it maintains horizontal equity while allowing room for longitudinal adjustments should financial sustainability so require. Point systems help to ensure that benefits paid to people of the same cohort reflect contribution histories fairly, while allowing policymakers to adjust point values, which provides the basis for assessing pension benefits of consequent cohorts in accordance with long-term fiscal objectives. By automatically adjusting point values to the same composite index as benefits, the Montenegrin pension system forgoes the longitudinal freedom inherent in the point system. In the medium run, the Government may need to decouple valorization from indexation and base a valorization rule on long-term equilibrium values based on actuarial projections. 2.38. The established schedule for the increase of retirement age for both men and women to 67 years already places the country among the most stringent ones in Europe. Another measure that Government might consider is limiting the inflow of new beneficiaries through reducing the scope for occupational early retirement. It is also recommended, once the retirement age reaches 67, that the pension law is amended in a way which introduces automatic retirement-age increases to ensure that life expectancy at the statutory retirement age remains constant. 35 According to the 2007 Montenegro poverty assessment, retirement did not increase the risk of poverty for those living in a household headed by a retiree and household demographics were not a good predictor of . Since 2007, pensions uniformly increased at a rate higher than inflation while newly assessed pension benefits did not noticeably fall. Thus, pensioners should be considered better protected against poverty than they were in 2007. 41 2.39. The Government needs to decide whether it wishes to build a robust, domestic voluntary pension industry. If so, (i) supplementary pension savings will need to be tax-advantaged; (ii) domestic capital markets developed (especially in terms of tradable securities); and (iii) the regulator’s technical capacity improved. In the absence of these steps, the fragile—i.e., barely existent—voluntary pension industry will remain unviable. In such an environment, mandatory second-pillar schemes should not be introduced. In (re-) assessing its policy stance towards voluntary pension savings, the Government would need to revisit the rules under which Montenegrin citizens can join—either directly or through domestically licensed agents—open pension schemes that are registered in selected developed markets. This approach would necessitate a review of international regulations concerning financial-sector regulations and tax administration, a task that would need to be undertaken in the context of the EU accession process. 2.40. Fiscal rules and impact assessment are instruments that are crucial in evaluating parametric and policy reform options. In the case of a point system, it is long-term projections through which the equilibrium point value can be determined, where the equilibrium is defined vis-à-vis the policymaker's long-term fiscal objectives. In this respect, it is recommended that actuarial modeling capacity is developed and a model is used for adjusting the point values and assessing the medium- and long-term fiscal impacts of legislative proposals and their affects on the public pension system as a whole. In the absence of a set of general fiscal rules aimed at debt sustainability and fiscal discipline, it would suffice if a partial fiscal rule concerning pensions—the largest public expenditure—was devised. Such rules may be expressed in terms of the Fund PIO’s unfunded liability or any other such indicator which can be expressed in net present value terms to accurately describe the balance-sheet impact of changes in legislation, macroeconomic and demographic circumstances, and changing assumptions. The most important information derived from such a system is the total impact and the time-path of the fiscal impact that result from new assumptions and changes to parametric or policy variables. 2.41. Coverage and system dependency are issues which impact both on the financial sustainability of the system and its capacity to provide adequate pensions. Unless today’s low activity rates start improving, the public pension scheme—providing pensions commensurate with contribution history and intensity—will be unable to pay pensions to a growing segment of the elderly population. This is not yet a problem, but in approximately 15 years’ time, contributors with insufficient contribution histories will reach retirement age and will require either increased social assistance or a amending the pension system’s rules pertaining to minimum benefits. Montenegro’s system of unemployment insurance caters to a large number of people close to retirement who, in addition to receiving unemployment benefits, are also covered by pension contributions paid by the Employment Bureau. In other words, actual contributor numbers are lower than in the official statistics as some contributions are actually paid by the budget. The average age of a registered unemployed is 44 years, while the average span of unemployment is three years. These figures are consistent with a relatively high structural unemployment rate which will continue to undermine the integrity of the pension system—and cannot be treated by pension policies. 42 3. THE WAGE BILL: MODERNIZING PUBLIC ADMINISTRATION36 The wage bill remains the Government’s key challenge in ensuring that the twin objective of developing a fiscally sustainable and operationally effective public administration can be met. One-quarter of general Government expenditures are reserved for salaries of civil servants and state employees, and pressures exist—partly for EU integration-related challenges—to increase staff levels further. For fiscal sustainability reasons, Government will have to exert very tight control over the wage bill and focus on the improvement of the public administration’s effectiveness and, within the available budgetary envelope, build up capacity needed to manage the process of EU accession. Given the increased complexity of tasks to be carried out by the public sector over the medium term, it is critical to retain and promote qualified and high-performing staff with salaries that are competitive with the private sector. With the private/public sector wage gap for qualified staff increasing, it is becoming increasingly important to reform and modernize the remuneration system for civil servants and state employees. A. INCREASED DEMANDS ON PUBLIC ADMINISTRATION 3.1. Montenegro spends about 12 percent of GDP on public-sector wages—or about 3 percentage points more than the average of neighboring countries in the Western Balkans. This level of spending on public-sector wages is only comparable to Bosnia-Herzegovina, generally an outlier because of its complex politico-institutional structure. To some extent, Montenegro’s elevated ratio reflects the “fixed cost” element of running a small country, a feature that is shared by other EU candidate and EU member countries with comparable population figures. Iceland, Malta, and Cyprus,37 the smallest countries in this group, have an even more elevated wage bill-to-GDP ratio, spending almost one-third of their budget on salaries for the public administration. In comparing Montenegro’s expenditures on the public wage bill with these two groups of comparator countries, Table 3.1 alludes to the inherent risk of having the wage bill crowd out other high-priority items in the budget, particularly relevant during a period with increasing demands on the public administration. Five years after independence, Montenegro’s public administration has to respond to ever more complex EU -integration challenges and the labor-intensive reform agenda aimed at strengthening the business and investment climate. As a result, there will be growing pressures to recruit additional public-sector staff (and fewer options to reduce public sector employment).38 With little room for savings on the staff-level side and constant pressures on salary levels for qualified staff, which continue to fall behind those paid in the private sector, the limited fiscal space makes it critically important to step back and (re-)assess options to increase public administration effectiveness and increase the “value-for-money” of salaries paid to civil servants and state employees. 3.2. This is even more important as inefficient Government bureaucracy—despite noted progress—remains one of the most problematic factors of doing business in Montenegro. Even though international competitiveness surveys show that the relative importance of perceived Government inefficiency has decreased gradually over the last three years (WEF 2011, 2010, 2009 and Figure 1.15), the high reliance of Montenegro’s economy on FDI inflows makes continued reforms in this area an automatic priority. Similarly, despite the fact that Montenegro’s business climate compares favorably with many countries in the region, the World Bank’s Doing Business surveys have identified key public- 36 Prepared by Svetlana Proskurovska (LCSPS), with input for Danijela Vukajlovic-Grba (ECSP2). 37 Luxembourg, the country with the highest per-capita GDP in the EU (about 18 times that of Montenegro), is a little bit of an outlier, reflecting a very different role that Government plays in that country. 38 The Government has sought to stem against these pressures by adopting a Personnel Policy Paper in 2010 that established a “two-for-one” principle—i.e., the Government can recruit one additional staff for every two that are leaving the public sector. 43 sector responsibilities as key obstacles to providing enterprises with an attractive business climate. In comparison to other countries, (i) dealing with construction permits; (ii) paying taxes; (iii) enforcing contracts; and (iv) registering property were areas with particularly pronounced performance weaknesses by Montenegro’s public administration. Contrary to neighboring countries (and certainly to EU member countries), the absolute staff size in the Ministries and public institutions is small, thus creating the principal problem of providing high-level services in the breadth and depth that would be required and expected. In general, civil servants and state employees are able to rise to the occasion, when required, and they have been able to address the myriad of post-transition, state-building, and EU-integration challenges39—efforts that were crowned by the EU’s decision in late 2010 to award Montenegro the status of candidate country. The European Commission (2010) has sought to support their efforts by defining two of seven policy priorities in the area of public administration reform. Thus, EU membership negotiations will only commence if Government can complete essential steps in public administration reform, including amendments to the law on general administrative procedure and the law on civil servants and state employees and the strengthening of the Human Resources Management Authority and the State Audit Institution, with a view to enhancing professionalism and de-politicization of public administration and to strengthening a transparent, merit- based approach to appointments and promotions; and strengthen the rule of law, in particular through de-politicized and merit-based appointments of members of the judicial and prosecutorial councils and of state prosecutors as well as through reinforcement of the independence, autonomy, efficiency and accountability of judges and prosecutors. Table 3.1. Public-Sector Wage Bill, 2008–10 Averages Countries in the Western Balkans MNE MKD RKS† ALB BIH HRV SRB average Population 0.6 2.1 2.2 3.2 3.9 4.4 7.4 … Wage bill, in percent of GDP 12.1 5.3 7.2 6.2 12.8 10.2 10.5 9.2 Wage bill, in percent of total expenditures 25.5 15.8 17.1 19.5 25.4 24.7 23.6 21.7 Other small EU member or candidate countries ISL MLT LUX CYP EST SVN LVA average Population 0.3 0.4 0.5 0.8 1.3 2.0 2.3 … Wage bill, in percent of GDP 14.7 14.1 7.6 15.2 7.8 10.5 9.7 11.4 Wage bill, in percent of total expenditures 31.2 32.4 18.8 33.3 17.4 23.6 23.4 25.7 † Most recent IMF staff reports with averages for 2008–10, except for Kosovo, for which only data for 2009–10 are available. Source: Latest IMF Staff Reports; and World Bank staff estimates. 39 Following the formal application for EU membership in December 2008, Montenegro’s public administration, during July–December 2009, succeeded in compiling comprehensive responses to the EU’s questionnaire containing 2,178 questions on the progress made in approximating its legal framework to the acquis communautaire, with 673 follow-up questions processed during early 2010, on which basis the European Commission and European Council awarded the Montenegro the formal status of candidate country. 44 3.3. In its present form and under existing budget constraints, the public sector will be hard pressed to provide services as responsively and accurately as would be expected and required in Montenegro’s current politico-economic environment. In particular, the inherited system of remuneration of civil servants and state employees includes a number of poor incentives, which—if addressed—would give scope for considerable efficiency gains. This chapter will highlight areas in which reforms to the underlying pay system would encourage high-quality and more efficient public- sector performance, including by (i) strengthening the merit-based elements in the determination of salary and promotion; (ii) streamlining the regulatory framework; (iii) tightening job descriptions and defining performance standards more objectively and in a measurable manner; (iv) encouraging learning and professional skills upgrading; and (v) strengthening Ministry of Finance control over the wage bill. B. THE GRADING SYSTEM 3.4. Civil servants’ pay consists of a fixed salary, a variable part, and an income supplement. The fixed salary component is tightly regulated with a relatively high compression rate, while the variable part is discretionary and—despite best intentions to the contrary—only loosely linked to performance. This could lead to civil servants’ (rational) career-planning considerations to include variables other than those linked to work performance, skills, experience, and qualifications, which —if left unaddressed—would contribute to the development of a grading system that de-motivates staff and contributes to a decrease in average work efficiency. The salary’s fixed components 3.5. The salary’s fixed part is a simple product of the legally prescribed base wage (the so- called “salary price”) and a wage-grade-specific multiplier. According to the Law on Salaries of Civil Servants and State Employees, amended in 2009, positions in the public sector are graded as summarized in Table 3.2. Civil servants are classified from the lowest wage grade 26 for entry level civil servants (advisor III) to the highest grade 1 (head of administrative authority). State employees are graded from the lowest grade 38 (state employee) to the highest grade 15 (senior state employee I). The grading systems for civil servants and state employees—and thus their salaries—overlap in the wage grade range 15–28. For each position, the fixed salary part is calculated by multiplying the salary price (determined by collective bargaining) with the multiplier (or coefficient, determined by law). At present, the salary price, fixed at €55, has been decreed in 2007.40 3.6. Given very narrow bands, professional growth and performance cannot be recognized within a given wage grade. The pecuniary performance recognition takes place through promotions; the present system does not provide for a sufficiently large fixed-pay differentiation within a given salary grade. A senior state employee I, for instance, has a salary range of only €14.30, from €314.60 to €328.90. For employees who remain in the same position for years, the fixed salary—except for the lowest wage grade—can increase, at most, by between 4 percent and 9 percent (Table 3.2). The largest within-grade salary range of 43 percent is for state employees in the lowest grade, following a promotion from salary grade 38 to 36.41 For higher-ranked state employees in grade categories IV–II, the inability to recognize professional growth within a salary grade creates incentives for public-sector employees to 40 The (reference) salary price is applied for the civil servants in the central public administration —but not for local Governments, which have their own values of salary price (leading to significant salary disproportions). In addition, judges have a higher salary price, currently set at €70. The law stipulates that the value of salary price is determined based on consumption needs of a family of four, labor productivity, and the economic situation in the country. 41 The highest grade for civil servants and state employees is 1, the lowest 38. 45 (i) keep their eyes open for higher-paid employment outside Government; (ii) accelerate a career through partisan activities and political loyalty, or (iii) improve personal qualifications and skills to be able to climb up the position ladder on professional grounds. Needless to say, the public administration derives the benefits only from the third form of self-motivation. Table 3.2. Titles, Grades, and Coefficients of the Salary System Civil State Job title Based on a salary price of €55 Width servants† employees‡ Wage grade Multiplier Fixed wage of range from to from to from to (percent) Head of authority n.a. Head of administrative authority† 1 9.90 544.50 or service President of misdemeanor council† 2 9.75 536.25 Managers and n.a. Secretary of a ministry† 3 9.23 507.65 appointed persons Deputy minister† Deputy head of admin. authority† 4 8.71 479.05 Judge of misdemeanor court† 5 8.19 450.45 Deputy head of service† 6 7.15 393.25 Pres. of regional misdemeanor board† Judge of regional misdemeanor board† I grade n.a. Advisor to the head of authority† 8 6 6.89 7.15 378.95 393.25 3.8 Chief inspector† Chief authorized officer† Independent advisor I† 10 8 6.63 6.89 364.65 378.95 3.9 Inspector I† Authorized officer I† Independent advisor II† 12 10 6.37 6.63 350.35 364.65 4.1 Inspector II† Authorized officer II† Independent advisor III† 14 12 6.11 6.37 336.05 350.35 4.3 Inspector III† Authorized officer III† II grade II grade Senior advisor I† 17 15 5.72 5.98 314.60 328.90 4.5 Senior state employee I‡ Senior advisor II† 19 17 5.46 5.72 300.30 314.60 4.8 Senior state employee II‡ Senior advisor III† 21 19 5.20 5.46 286.00 300.30 5.0 Senior state employee III‡ III grade III grade Advisor I† 24 22 4.68 5.01 257.40 275.55 7.1 State employee I‡ Advisor II† 26 24 4.42 4.68 243.10 257.40 5.9 State employee II‡ Advisor III† 28 26 4.10 4.42 225.50 243.10 7.8 State employee III‡ n.a. IV grade State employee IV‡ 31 29 3.45 3.77 189.75 207.35 9.3 State employee V‡ 33 31 3.19 3.45 175.45 189.75 8.2 State employee VI‡ 35 33 2.93 3.19 161.15 175.45 8.9 n.a. V grade State employee‡ 38 36 2.00 2.86 110.00 157.30 43.0 Source: Law on Salaries of Civil Servants and State Employees. 46 3.7. Considered low by international standards, the decompression of pay between the lowest specialist position and the most senior managerial post is 1:2.4. In the civil service, positions are graded into two groups—managerial and specialist. This would, however, abstract from the fact that the fixed part of the salary is only one component of total remuneration. The variable pay (discussed below) provides opportunities for increasing this ratio. It is a common phenomenon in many countries with low salary decompressions that lower-end Government jobs tend to be fairly attractive on the labor market, while the remuneration gap between high-level private- and public-sector Table 3.3. Average Public-Sector Salaries jobs makes the latter positions increasingly uncompetitive. Average Public Health Education Consequently, the key concern labor force administration sector sector relates to policymakers’ ability to ensure that (highly) qualified staff remains interested in applying for— (in euros) or staying in—public-sector positions. Over recent years, average Average 2007 497 506 427 433 salaries in Montenegro’s public January 2011 772 754 740 662 administration have eroded Nominal increase 55.3 49.0 73.3 52.9 somewhat vis-à-vis those paid in the (relative to average labor force) private sector. In early 2011, average salaries (gross) in the economy Average 2007 100.0 101.8 85.9 87.1 (€772) exceeded those paid in the January 2011 100.0 97.7 95.9 85.8 public sector (€754) by 2.4 percent; this relation has tipped since 2007; Source: MONSTAT; and World Bank staff estimates. see Table 3.3. While, on average, the wages rose by more than 55 percent in the economy as a whole42 (January 2011 relative to average of 2007), they rose by “only” 49 percent in the public sector.43 3.8. Anecdotal evidence suggests that skilled officials have left the public administration for the private sector, but it is not possible to confirm this assertion on the basis of the centrally consolidated personnel database (which is incomplete and covers only a minor share of the civil service). Statistics for the whole of the public administration are unavailable—largely because the Civil Service Register does not cover all statutory civil servants. In March 2010, only 30 percent of the estimated total civil servants were included in the central personnel registry, in spite of the mandatory legal requirements for a central database44. Staff attrition, observed at the level of higher-specialist positions, might be an indication of inefficient remuneration policy. 3.9. The narrow decompression ratio among civil servants is balanced by non-monetary aspects of a public-sector career. Incentives for staying in the civil service over years and pursuing professional growth and career progression depend on several variables, of which the pay and grading structure is but one aspect. Other factors include (i) pay benchmarking with the private sector for jobs comparable in complexity, independence, and responsibility; (ii) career prospects; and (iii) the adequacy of rewards for growing job complexity. Apart from the salary considerations, the most important reasons for choosing a 42 Policymakers are faced with the challenge to ensure that productivity increases, in both the private and public sectors, stay ahead of wage increases so as to prevent a (further) erosion of international competitiveness. 43 The health sector represented an exception, with an average salary growth rate of 73 percent. 44 So far the degree of compliance with the legislation has not been strong, which is reflected in the opinion of the European Commission on the country’s readines s for European integration. 47 career in the public sector include job security, job and job content satisfaction, pension provisions, as well as additional benefits and supplemental payments available in the public sector. Taking into account these additional aspects, the decompression ratio of 1:2.4 may appear not to be exceptionally narrow. 3.10. In the older EU member states, the decompression ratio in the public sector—if calculated correctly—is not that much higher. In Germany, an example often cited in the literature on remuneration levels, the decompression in the civil service is 1:6.8. Such decompression exists for the positions of a courier and the state secretary in a federal ministry. The situation in Montenegro is similar to other countries in transition, in which entry level positions in the civil service are on the specialist/advisor level. These positions require relatively higher levels of education and professional qualification. Relevant decompression comparisons with the German civil service would be the base salary of the state secretary with that of a desk officer in a federal ministry. Calculating decompression on the basis of an entry-level desk officer would result in a ratio of 1:3.3. If the comparison is made for the state secretary in Germany and a desk officer who has reached the out end of the salary range, his or her highest base salary would be imply a decompression ratio of 1:2.6. This is comparable to the current decompression for the fixed part of the civil service salary in Montenegro. 3.11. In conjunction with other non-salary features of public-sector employment, civil service jobs remain moderately attractive positions in the general labor market. It would be important to estimate the full benefit package provided by the salary system and social benefit system to make a more informed judgment on the attractiveness of the public sector work. Indicators that would allow determining whether—or to which degree—senior positions are attractive for (highly) qualified candidates include (i) the number of vacant positions; (ii) the average length of vacancies (especially for positions which are difficult to fill); and (iii) the average number of qualified candidates applying (considered) for a given position. However, as implicit in the European Commission’s (2010) policy priorities, senior positions tend to be subject to political nominations rather than competitive selection. As a result, the positions in the senior civil servants may be filled fairly quickly, but this would then be unrelated to the perception of the relative attractiveness of the posts or the underlying salary regime. It would be advisable that the Human Resource Management Agency (HRMA) monitor the above indicators over a prolonged period of time to determine whether or not salary levels are adequate to attract appropriately qualified candidates. In the absence of corresponding time series —i.e., in basing policy recommendations simply on the comparison of the salary level in the labor market with the average salaries in the public administration and taking into consideration features of job security and a typically more challenging job content—a preliminary assessment points to public administration positions remaining “moderately attractive” relative to those in the general labor market. The impact of the length of service on the fixed salary component 3.12. The fixed salary system takes into account the length of service (Table 3.4). For Table 3.4. Length of Service and Salary Increases this reason, the “statutory” and “personal” fixed salary levels have to be distinguished. Years of service Multiplier for years of service The increment of personal fixed salary for the length of service depends on the number of 1-10 0.5 percent years. A percentage of the statutory fixed 11-20 0.75 percent salary that is added for the length of service 20+ 1.0 percent depends on the range of the length of years, as presented in Table 3.5. To exemplify the Sources: Law on Salary of Civil Servants and State Employees. significance of the seniority rule, consider a Senior Advisor I at wage grade 17 with a statutory fixed salary of €3 14.60. With everything else equal, the length of service affects the individual fixed salary in the following way: 48 Table 3.5. Example of the Influence of Seniority on Grade 17 Salary Percent of fixed pay Years of service Fixed salary, € Seniority increment, € increase from previous period 0 314.60 0.00 - 5 322.47 7.87 2.5 10 330.33 15.73 2.4 15 349.99 35.39 6.0 21 380.67 66.07 8.8 Sources: Law on Salary of Civil Servants and State Employees; and World Bank staff calculations. Without a change in performance, just for the 21 years of loyalty to the civil service, the civil servant will be rewarded with a €66-increment in monthly salary. Rewarding employees for the length of service rather than performance or skills represents a disadvantageous feature for the young, well-educated, and high-performing civil servants with a short length of service. During interviews, officials in Montenegro have observed that many employees with outdated skills continue working in the public administration with lower productivity but higher salaries. 3.13. Relatively generous seniority increments do, of course, crowd out alternative methods of incentivizing performance. The asymmetric treatment could be justified if the skills that new entrants bring to the job may not be that advantageous and the lack of job-related knowledge and skills create inefficiencies. The resolution of this debate should be linked to an objective job evaluation and a system that links pay to skills and performance—not through an ad hoc application of bonuses but rather through a predictable pay progression based on satisfactory performance appraisal. 3.14. Other countries have opted to substitute the automatic link between length of service and fixed-pay increases with a pay progression structure aimed at rewarding satisfactory or excellent performance. If such experience—as, for instance, implemented in Germany—were accepted in Montenegro, it would allow the Government to reward skills enhancement and job performance. It would presuppose the establishment of a system of regular appraisals, typically through an annual cycle of performance evaluations. This report recommends considering the introduction of a pay-progression system linked to individual performance (rather than the length of service). This would presuppose the institutionalization of an objective performance appraisal system. Germany’s experience of recognizing performance for pay progression has proven beneficial in increasing the average performance of the public administration and, as a result, been widely replicated in other countries of Europe and Central Asia. Lessons from Germany’s remuneration reforms 3.15. In the German civil service pay system, the remuneration of specialists envisages incremental pay increases that are subject to satisfactory performance (as determined by formal performance appraisals).45 For each specialist grade, the pay scale envisages eight steps of progression 45 This system stands in contrast to that employed for senior positions in the civil service (on the level of a state secretary, director in a federal ministry, or head of division). Salaries of senior civil servants are fixed and defined by the level of the post only. 49 within the basic salary. Higher salary steps are awarded after the completion of the evaluation period, which lasts from two years in the beginning of the career up to four years in later stages of the career, reflecting the assumption of declining marginal on-the-job learning increments. If learning on the job occurs at a faster rate in first years on the job, before the growth in performance starts to slow down, it is consistent to have this reflected in the decreased percentages for pay increases. For example, for the head of division in a ministry, the first pay increment is 5 percent, the second 4 percent, and the subsequent increments are 3 percent. However, for the desk officer, the relative pay increments between steps do not reduce as quickly as for a head of division. During one’s career in the same p osition, a head of division can move up through all 8 pay steps, provided his performance was no less than satisfactory—equivalent to a total pay increase of 27 percent. The sequence of pay steps for the desk officer would be 5 percent for the first three steps and 3 percent for the rest steps, with a total pay increase of 29 percent from the lowest to highest step within a pay grade. Such an approach allows for a reduction of decompression of salaries between the desk officer and the head of division over their careers from initial ratio of 1.39 in salary step 1 to 1.37 in salary step 8. The corresponding salary grades with step for the German federal ministry are presented in Table 3.6 for information (from the pay scale valid as of July 2009). Table 3.6. Germany: Salary Grading for Civil Servants with Horizontal Pay Progression Basic salary (monthly salaries in euro) Pay Avg. step Job title Grade Step 1 Step 2 Step 3 Step 4 Step 5 Step 6 Step 7 Step 8 increase (percent) Desk A13 3,416 3,586 3,755 3,925 4,042 4,160 4,277 4,392 4.00 Officer Head of A16 4,737 4,967 5,141 5,315 5,488 5,663 5,837 6,009 3.66 Division Source: Federal Ministry of Interior, 2009, The Federal Public Service (www.bmi.bund.de). 3.16. This approach differs from Montenegro’s in two important aspects. First, in Montenegro, the increase of the fixed salary with the years of service is applied automatically, without consideration of employee performance. By contrast, in Germany, an employee can move to the next fixed salary step after being positively appraised over the assessment period. Second, the alignment of an employee’s productivity growth and his or her salary increases appear more closely aligned in Germany. In Montenegro, marginal salary increments increase with the length of service, at a time when learning ability and skill development can be expected to slow down, risking that pay grows at rates increasingly likely to exceed a long-serving employee’s productivity increases. In this situation, it is advisable to improve the public administration’s efficiency by performance- rather than seniority-linked remuneration formulae as closely aligned to productivity indicators as the institutionalization of an overall performance assessment is able to achieve. The salary’s variable components 3.17. The salary’s variable component is to reward performance. The Government Regulation on Criteria for Determining a Variable Part of a Salary (dated April 4, 2008) defines that the variable part of pay is determined on the basis of professional qualifications and exceptional performance (results and quality) of civil servants’ work. The variable pay is determined by the Head of Organization and is 50 subject to approval by the Minister of Finance. The variable pay is contained in the approved annual budget. However, this regulation does not provide any indications on the size of the variable pay or the number of (potentially) eligible employees. 3.18. Poorly regulated and unrestricted, the variable pay component encourages favoritism and risks undermining staff morale. The provisions for variable pay on professional qualifications refer to academic education and vocational training for performing particular tasks, specific specialist knowledge relevant to the position’s requirements and duties, working knowledge of a foreign language needed to perform specific tasks, and the special knowledge in the area of IT-system design. The regulation on variable pay was introduced with a view to giving managers a tool to attract young employees with a specific set of skills that, otherwise, could not be attracted into the public sector. However, such a system of poorly regulated and unconvincingly justified increases in variable pay risk (i) impacting negatively staff morale; (ii) undermining incentive systems for good performance, and (iii) creating grounds for favoritism. Therefore, consistent with the recommendations derived for the fixed salary elements, Government might want to consider abandoning variable-pay provisions, thereby opening the fiscal space for the introduction of adequate fixed-pay levels that properly reflect requirements of qualification, skill, and experience. For this to yield the expected efficiency gains, it is critical that a methodology for the description and assessment of posts is developed, clearly describing related qualification requirements and linking these to a fixed pay scale capable of accommodating the increasing differentiation of jobs in Montenegro’s public sector. 3.19. At present, job performance expectations are not sufficiently aligned—because of the vagueness in the definitions of performance and manager/employee communication channels. As a result, performance assessment mechanisms tend to become too ad hoc to represent a useful human resource management tool. In the aforementioned Regulation, work program results are defined in a manner that tends to be difficult to measure (reliably and accurately). In the absence of monitorable indicators, any job performance assessment is subject to judgment and the discretion of managers. At the moment, the list of criteria for performance evaluation includes (i) a special contribution to program tasks of state administration; (ii) a special contribution to the improvement of work processes relevant to the work of public administration and Government policy; (iii) a special contribution to the processes of European and Atlantic integration; (iv) outstanding performance and quality of completed tasks; (v) work efficiency and effectiveness; (vi) independence on the job; (vii) increased workload; and (viii) extra work due to a vacancy in the list of organization’s systematic positions. No formal performance appraisal is required for rewarding results. 3.20. The general guidance provided by the Regulation—leaving a lot of discretion to the heads of organizations—introduces opaqueness into the salary system. As for professional qualifications, the requirements listed above should be seen as regular requirements for each position, in which case there would not be any need to reward employee qualities additionally. In a predictable (and permanent) manner, these would be used to determine the appropriate coefficients used to calculate the fixed salary. Only in instances, in which job requirements are very specific, non –standard, exceptionally rare in the labor market and need to be remunerated additionally, it would be appropriate to reward them with a variable pay, while clearly indicating the main measurable parameters used to determine the variable pay. 3.21. Outstanding results can be rewarded through (permanent) increases to the base salary or (one-off) bonuses. Financial rewards for outstanding individual performances are positively related to average public-sector efficiency only if they are generally understood as such. This implies that, for the reward of outstanding results, explicit—and measurable—parameters for rewards should be established, together with a system for rigorous, rule-based performance appraisals that record outstanding performances and suggest the size of reward. In so doing, there are two methods of rewarding good 51 performance, either through performance-based pay progression in the fixed pay (which will have a lasting effect on professional development) and another through the payment of (one-off) bonuses. If Government were willing to abolish purely seniority-based pay increases, it would create the fiscal space to introduce permanent salary increments based on (satisfactory and) outstanding performance into a pay scale. 3.22. Attracting and retaining highly qualified and productive staff requires the interplay of salary and non-monetary incentives. In devising a reform strategy, the Government of Montenegro might want to consider international experience, showing that financial incentives alone are not an efficient instrument to improve average performance of the public sector. In a recent OECD study, Ketelaar et al. (2007) concluded that prerequisites for good public-sector performance comprised a set of elements, including (i) selecting the right staff; (ii) motivating staff through promotion and recognition rather than through financial incentives; (iii) retaining competent staff through adequate and predictable compensation (adequate base salary); and (iv) embedding targets for individual performance within the wider agency/Government performance targets. One important result included the observation that performance can be encouraged more sustainably by using dialogue for assessing and recognizing results rather than simply applying sanctions or monetary rewards46—one additional argument in favor of a functioning system of an annual performance review. These would align individual with public-sector values and, as such, complement intangible benefits such as job security, prestige, social privileges and reputation, contributing to an overarching incentive system aimed at encouraging good performance. 47 3.23. Given the perceived lack of objectivity and impartiality, most OECD countries have opted for a combination of adequate base pay and limited performance-based elements. The OECD has concluded that, in principle, there is not sufficiently satisfactory evidence that would link performance- based pay (financial incentives in return for high performance) to improved overall public-sector performance. In surveys, many interviewed employees have mentioned that performance-based pay was perceived lacking in objectivity and impartiality, with measurable performance targets being difficult to 46 Recent studies of work motivation conducted by Massachusetts Institute of Technology (Cambridge, Massachusetts) and London School of Economics assert that in tasks that require application of cognitive—rather than mechanical—skills financial incentives do not improve performance; see Dan Pink (http://www.ted.com/talks/dan_pink_on_motivation.html). Instead, provided that employees are paid adequately and fairly, they are performing better if they are given creative autonomy, a chance to improve their mastery, and if the result of their work has a purpose. These three conditions create intrinsic motivation, more powerful than performance bonuses. These ingredients of high performance appear to be adversely impacted by financial incentives as the latter focus by employees impair their creativity in problem solving. Therefore, the focus of new remuneration reform in the public administration should not be on bonuses and supplemental pay, but on adequate (base) pay level and conditions for creative autonomy, mastery, and purpose. 47 See also Perry (2006), who has reviewed performance pay in the public and private sectors. They found that (i) financial incentives can moderately improve task performance, mostly in the private sector, but their effectiveness depends on organizational conditions; and (ii) individual financial incentives are ineffective in traditional public- sector settings. Performance-related pay was observed more for employees in manufacturing jobs, i.e., in sector, in which there was a clear and measurable relation between the quantity of output (at a given level of quality) and performance pay. In public administration, quality rather than quantity is the principal measurement of employee output, together with complementary indicators of good governance, such as timeliness and legality of actions taken. While group incentive systems are consistently effective, these incentive systems are not well tested in public-sector settings, where measures of organizational performance are often uncertain. Consistent with other studies, Perry at al. (2006) emphasized the positive effects on motivation and performance from (i) job design; (ii) the participation of employees in decision making; (iii) challenging and specific goals set for employees by their managers; and (iv) constructive and timely feedback on performance provided by managers. According to the authors, the effective management of personnel by supervisors and managers was critical to improving organizational performance and, by far, more effective than simply relying on performance-related pay. 52 set in the public sector. As a result, performance payments have not adequately rewarded good performance. In consequence, most OECD countries have returned to emphasizing adequate base pay in the public administration, while rolling back performance-related pay to a very small percentage of the total remuneration. In fact, many OECD countries do not have special performance bonuses as a separate payment at all. Instead, they incorporate performance recognition into the salary scale; see Table 3.7. In the end, the overarching incentive structures are anchored in systems employed to decide on performance recognition through promotions and the ability to plan a professional career within the public sector. Table 3.7. Selected OECD Countries: Base Salary as Percent of Total Remuneration Base salary as percent of total Base salary as percent of total Country Country remuneration remuneration Austria 100 percent in most cases Netherlands 93–100 percent Canada 100 percent Portugal 100 percent, except for one group Denmark 99 percent Spain 30–100 percent Finland 65–90 percent Sweden 100 percent France 75–95 percent Switzerland 100 percent Germany 70–90 percent UK 100 percent, except for police Iceland 100 percent USA 100 percent Source: OECD statistics Supplements as a separate element of variable pay 3.24. The effective remuneration of public-sector employees in Montenegro includes an array of supplemental payments, frequently legacies of the Socialist-era remuneration system. The supplemental elements of remuneration have comprised allowances for hot meals, transportation, holidays, family separation, the compensation for use of one’s vehicle for business needs, supplements paid in cases of illness or death of a family member, and —for judges of the Supreme Court that resided separately from their family—a housing allowance. The 2009 amendments in the Law on Salaries of Civil Servants and State Employees reduced some types of supplemental pay while the others (hot meal and holiday allowance) have been included into the base pay. Most of the listed supplemental pay elements are social guarantees, protecting the social welfare of public employees. These payments reflect also the historic tradition of trade unions’ role in ensuring minimum social care for public servants. Since 2010, some of these social payments have become subject to taxation in 2010, which this report supports. 3.25. The supplemental pay has contributed to the negative image of the public employees. For comparison, in the United States, unionized public-sector employees also used to enjoy a number of benefits, more generous than the private sector employees. The economic crisis highlighted the asymmetrical treatment of the labor force in the public and the private sectors. The current media outcry supports the argument that all the “perks” that the public employees enjoy leave the private -sector 53 employees at a disadvantage, although the latter actually create the wealth (and pay taxes) that enables Government employment in the first place. Also the arguments for equalizing the remuneration conditions in the public and private sector, presented by the USA Center for Union Facts, a think tank, emphasize that “public sector workers enjoy substantially greater job security than their private sector counterparts; their layoff and discharge rates as a percent of total employment is over three times lower than the private sector as a whole”. In conclusion, this report recommends that the public sector salary system should be seen as transparent and related to job values, rather than built of multiple segments with supplemental payments that seem to be questionable for the wider public. 3.26. In recent years, public-sector remuneration in Montenegro has been subject to a number of changes. Following the average 30-percent increase in public salaries granted at the height of the boom,48 Government pursued a more careful policy subsequently, including the adoption of a cap on civil servants and state employees. In the first part of 2010, most public employees saw their salaries decrease, following the need to adjust expenditure to the reduced revenues. In many cases, variable pay was cut through reduced budget allocations. However in May 2010, the Government issued a regulation49 with which it established salary increases for selected groups of public employees. The average raises in salaries constituted 30 percent; however, individual increases ranged from 10 to 40 percent. These increases recognized special working conditions, complexity of work, and harsh working conditions. The following categories of public employees were singled out for increases: policy officers, IT system specialists, penitentiary managers and specialists, staff in the tax and custom administration, rescue and emergency service employees, judges and magistrates, inspectors in most of the public sectors, and defense employees. Policies for remuneration of contractual employees in Montenegro 3.27. Policies on the remuneration of contractual employees are determined through a process of negotiations between the Government and trade unions. In December 2010 the Government and Trade Unions agreed to change the methodology for determination of wages for contractual employees. Although the proposed new method of calculation salaries for contractual employees does not have direct relevance to the salaries of civil servants and public employees, given the tendencies to level salaries in the public with private sectors (not least because of trade unions’ influence), the salaries in the private sector would also lead to changes in the public-sector salaries. Moreover, public institutions continue hiring contractual employees to meet their organizational challenges, especially when a cap on the number of civil servants, introduced through the systematization act, has been reached. Therefore, this report explores the changes introduced. 3.28. Government and trade unions have recently agreed on a new Agreement. With the economic growth gradually picking up in the second half of 2010, the trade unions and the Government have negotiated and signed a new Agreement related to the new method of calculation of gross salaries for employees, which sets a progressively increasing new base pay and recognizes job complexity and length of service or performance. The Agreement (published in Official Gazette of Montenegro, No. 80/10) established the following key job-related parameters and coefficients: (i) the length of the full- time working month is 176 hours; (ii) the starting value for the gross salary calculation is established as a percentage of €90; (iii) the percentage used to determine the initial base salary (by multiplying with €90) will be gradually increased (every 6 months until 2012); and, (iv) for salaries, which would have fallen below the minimum wage (which, for 2011, is expected to be at the level of €145), the employer will 48 Percentage increases are, of course, the easiest form of salary increases. They should only be granted in exceptional circumstances—and then linked to the overall pay grading structure so that relative jobs values could be presented alongside with other jobs to assure internal consistency of values of posts. 49 See Official Gazette of Montenegro, No. 54/10, September 14, 2010. 54 compensate the difference up to that level. Following the signing of the above Agreement, the Ministry of Finance issued instructions for determining gross earnings on January 17, 2011 (Official Gazette of Montenegro, No 5/11) to implement the agreed provisions. The instruction provides guidance on gross salary calculations. For the regular hours the salary is calculated as a sum of the base salary and increment for complexity. The base salary is determined according to the phase in schedule as follows: Table 3.8. Planned Phasing-In of 2011–12 Base Pay for the General Labor Force Time period Percentage of the fixed Resulting base pay, in € value of €90 Jan. 1 – Jun. 30, 2011 55 49.50 Jul. 1 – Dec. 31, 2011 60 54.00 Jan. 1 – Jun. 30, 2012 65 58.50 Jul. 1 – Dec. 31, 2012 70 63.00 3.29. As for the part related to job complexity and the length of service, the following rule is introduced. A complexity coefficient for the job (contract) is determined in the sectoral collective agreement for each particular expertise and complexity. The Ministry of Finance, in its instruction, uses three levels of complexity with respective coefficients 3, 6, and 9. The adjustment coefficient for improved performance is based on either the length of service or on percentage increase of employee’s performance (Table 3.8). Such calculation for the first period (January 1 – June 30, 2011) would yield the total salary of €333 according to the following assumptions: complexity coefficient 3, adjustment coefficient for better performance 1.05, and the base salary €49.50. The salary then is calculated as €49.50 + 3 * 1.05* €90 = €333. Whereas in cases, in which the complexity factor is 9, the calculation would lead to a considerably higher salary of €900 (i.e., €49.50 + 9 * 1.05 * €90). If this will have impact on the salaries in the public administration and social sectors, then one would expect more emphasis on the complexity of jobs in the public sector as well. However, presently the differentiation of pay levels in the public sector is being introduced as percentage increases, as discussed in the following paragraphs. Principles of a good remuneration system 3.30. There are “strategic” and “technical” principles that are important for Government consideration. On a strategic level, the remuneration of the public sector employees must be affordable. Firstly, this requires that the total public-sector remuneration (or wage) bill does not exceed a certain threshold of public expenditure as a share of GDP so as not to crowd out other public expenditures, such as supplies that guarantee the effective quality of public services (e.g., text books for schools or medicines for hospitals), communal services, or investment in infrastructure. Secondly, the remuneration in the public sector should support effective and efficient delivery of Government programs by qualified and motivated personnel. On a technical level, these principles are related to the design of the remuneration system. A good pay system is simple. This is achieved by eliminating additional supplemental pay so that the base pay is the major component of the total pay. The values of the total pay are based on an analysis of the standard (typical) jobs in the public administration (see also Figure 3.1). The position evaluation uses a number of parameters determining the value of job. 55 Figure 3.1. Illustrative Single-Pay Spine Illustrative single-pay spine General grades Teaching grades Judiciary (pay points) 100 Director Sr. Judge Sr. Professional 90 Judge Top 80 Professional Sen. Head teachers 70 Supervisors 60 Executive Main teaching grade 50 40 Skilled 30 Probationary Semi-skilled 20 Trainee 10 Unskilled 0 Source: World Bank Global Expert Team. The public remuneration system is transparent and predictable. Not only does it mean that information of salary levels for posts in the public sector is accessible to any person or organization, the effective design of a pay system also requires that pay levels are determined without ambiguity, prejudice, or discretion that renders the public service pay opaque. If the system allows for unreasonable managerial discretion, e.g. determining pay components without clear criteria, the pay system is perceived unjust and this reduces employee trust and motivation. Regulation of public 56 sector salaries in a single law and single regulation, rather than being regulated in a number of normative acts, improves transparency of the system. Attempts to regulate different aspects of remuneration system in multiple legal acts is a bad practice which makes public sector remuneration less transparent and potentially leads to pay level discrepancies among various occupational groups. Analysis of remuneration regulation in Montenegro is difficult, as specific provisions are scattered among a multitude of legal acts of various hierarchy (laws, Government regulations, instructions, and sector regulations). Some consolidation would make the pay system more transparent and better manageable. It would also eliminate the grounds for unduly managerial discretion and liberal interpretation of the law. The public remuneration system provides equal pay for equal work. This is achieved through a good classification of posts that are evaluated based on a common methodology (common principles). However, multiple legal regulations tend to introduce discrepancies in the overall consistency of evaluation of job values. A good remuneration system provides incentives for senior jobs. This is achieved through an appropriate salary decompression. Compensation for senior jobs shall be comparable to the remuneration that people with such qualification can receive in the private sector. Moreover, the system encourages professional growth and career development, through pay differentials between jobs with lower and higher complexity and responsibility. A single, universal pay system for the public sector is preferable. In such systems salaries are presented against a common pay “spine”. This method is preferred to multiple isolated systems, each for a different occupational group, because it provides useful comparisons for remuneration of various groups of public servants. This in turn improves transparency and equity between the remuneration regimes. An approach to creating a single pay system in the public sector is demonstrated in Figure 3.1 below. The remuneration system provides incentives for improved performance. This means that annual performance appraisal is linked to the annual salary review for civil servants. Rules for rewarding good and excellent performance of civil servants and public employees stipulate how pay levels of employees should be revised (in a transparent and predictable manner) in an annual budget cycle. To enable pay progression to be based on performance, rather than relying on the length of service or bonuses, requires introducing certain remuneration system’s design features that currently are not present in the salary system in Montenegro. Such a predictable and rule based progression of pay (even without a change of position grade or title) allows for easier automation of pay and better flexibility of the Government to link resources to provision of good job motivation. Public sector pay is managed through an automated payroll system of the Treasury. Having an automated payroll provides transparency, eliminates potential abuse of pay levels through automatic controls, provides historical data for analysis, allows to calculate fiscal impact and assess the potential fiscal impact of any amendments of major pay parameters; reduces human error in wage calculation; is easily linked to the banking system for administration of payment to the individual employee accounts. Pay is competitive with the wider labor market. To prevent the outflow of the professional and skilled employees, the pay levels does not fall far below the pay levels for comparable jobs in the private sector that can be open for public servants with their current qualifications. It is useful to consider that job security provided in the public sector, as well as some other benefits should always be considered as additional value factors that may compensate for lower that in the private sector remuneration. 57 C. CONCLUSIONS 3.31. Government has to resist very strong pressures to increase both staff numbers and wage levels (exerted largely for EU-integration purposes) and focus on the improvement of the public administration’s effectiveness. Capacity will need to be built up among currently employed civil servants and state employees to manage the process of EU accession. The authorities have taken important steps towards reforming and streamlining the public administration, including by adopting the 2010 Personnel Policy Paper and the so-called “two-for-one” principle by which the Government limits its ability to recruit one additional staff by requiring the concomitant release or retirement of two civil servants or state employees. Together with pensions, the wage bill remains the Governments key challenge in ensuring that the twin objective of developing a fiscally sustainable and operationally effective public administration can be met. This requires a strong focus by Government on “value for money” from the public administration. 3.32. Currently, the legislation that regulates salaries of civil servants and public employees is very fragmented. This results in complexities of administration of the remuneration policy and significantly reduces transparency of the remuneration system. The World Bank supports the government intention to develop a single law regulating public sector remuneration regimes, applicable both to civil servants and public employees. The new law on public sector remuneration should be guided by the following principles: transparency, equal pay for equal work, simple and transparent salary structure, links or the grading structure to the classification of positions (that is built based on job descriptions), recognition of links between performance and pay progression, consolidation of pay elements into the base pay, ensuring adequate pay decompression between junior and senior posts. 3.33. The Ministry of Finance leads the work on development of the legislation on public service remuneration. It would be important that a working group be established for development of a draft law on public sector remuneration which would include representatives of the HRMA, ministry of labor, ministry of interior (local government department), and representative of the trade unions. This multi- skill group would be able to provide input on issues related to the job classification and job valuation for different categories of civil servants and public employees, as well as will help build support around new pay policies, embedded in the law. 3.34. Given the increased complexity of tasks to be carried out by the public sector over the medium term, it is critical to retain and promote qualified and high-performance staff, including by having salaries that are competitive with the private sector. With the private/public sector wage gap for qualified staff increasing, it is becoming increasingly important to reform and modernize the remuneration system for civil servants and state employees. At present, the current system is constrained by the following factors: x the pay system is not transparent and predictable as it depends on (i) the variable, poorly regulated part of pay; and (ii) multiple normative acts that regulate the remuneration regime of various occupational groups, which risk distorting the principle of equal pay for equal work; x the remuneration system includes a fixed part, which is loosely related to jobs in the civil service and modified by the variable pay administered by the discretion of management; x the remuneration system does not promote the quality of work as the progression for the fixed salary is based on seniority rather than performance; x performance criteria are defined in very broad—and thus non-measurable—terms and are therefore subjective; 58 x remuneration levels in the central public administration and municipalities are different because of the different value of the applied “salary price” (perceived unjust by public-sector employees); and x the Ministry of Finance cannot effectively control the total public sector wage bill aside from setting limits for budget allocations and enforcing the limits on the permanent employment numbers. 3.33. In 2009, the EU financed a project that analyzed the remuneration system in the public sector and formulated reform proposals, with results that are fully consistent with the recommendations derived in this chapter. With this project, the EU sought to use existing resources more effectively to reward civil servants and state employees and improve the public sector’s competitiveness in the labor market. Consistent with the present analysis, the EU proposed to increase the transparency of the payment structure by incorporating variable and supplemental pay in the base pay structure (e.g., by eliminating food allowance and seniority increments). In addition, the EU assistance recommended to (i) revise the pay grading structure and reduce the number of grades; (ii) enlarge the vertical pay differentials for higher-level positions; (iii) increase the number of horizontal pay steps to accommodate recognition of performance on the job; and (iv) improve job classification as the basis for the new pay system. 3.34. To systemize and structure the complex public-sector remuneration system, the following measures are recommended to at least “freeze” the w age bill-to-GDP ratio and increase the effectiveness of public administration (for a summary of a potential reform timeline, see also Box 3.1). In an initial step, Government might consider designating responsible institution(s) to develop a new pay policy and manage and coordinate reform efforts in this area. Such a working group would have to comprise representatives from the Ministry of Finance, the Human Resources Management Agency, the Ministry of Labor and Social Protection, and the public sector union and tasked with analyzing current drawbacks of the remuneration system, taking the issues identified in the EU report and this chapter as a starting point. The group’s members would need to concentrate on the following issues and identify the approaches to address them to (i) ensure that the salary system is fiscally affordable and, at the same time, competitive in the labor market for well-qualified candidates and fiscally affordable; (ii) design measures to improve the transparency of the remuneration system through streamlining legislation, developing a concept of the base pay that consolidates current variable and supplemental pay elements; (iii) implement the principle of “equal pay for equal work” through (a) the adequate assessment of job classification requirements; and (b) an improved classification of jobs in the public administration/sector; (iv) recognize performance through the pay system in a transparent manner, based on (and subject to) regular performance appraisals; (v) improve staff retention through attractive career development prospects by introducing an adequately decompressed salary structure, objective criteria for pay progression, and support to sill development; (vi) facilitate decision-making capabilities in Government by ensuring the availability of accurate pay information; and (vii) develop and implement a time-bound program of measures (and responsible officials) to introduce a new remuneration system for the public sector that is based on the core principles of the pay system and assign resources to implement this program. 59 Box 3.1. Potential Timeline for Public-Sector Remuneration Reform Program First 4 months: Analysis and development of a policy paper with objectives and results indicators; development of a program of activities; identification of needed resources and potential technical assistance. Subsequent 8–12 months: Implementation of the measures outlines in the Program aimed at developing a new pay system, including: x Revising job descriptions and developing a methodology for assessing jobs x Revising the job classification system based on the above methodology x Developing a new pay grading structure that incorporates various pay elements into the base pay structure x Piloting the reclassification of civil servants and adjusting the pay grading structure to avoid considerable reduction of proposed pay levels compared with the current x Conducting fiscal impact assessment and planning future budget provisions x Identifying the required improvements in the personnel management system x Revising the provisions for performance appraisal to link them with the pay progression system x Developing normative basis for introduction of the new pay grading structure Subsequent 3–6 months: Approval and implementation. x Parliamentary approval of the new legislation x Developing and implementing a training program for all public institutions on job descriptions, job reclassification, assignment of civil servants to adequate pay grades and steps in the pay grading structure, and performance appraisal linked to pay administration x Additional training x Preparation, review and approval of the new pay grading structure for each institution x Launching the new pay system x Implementing a related performance appraisal system x Improving personnel management practices in public institutions x Monitoring of the implementation of the new pay grading system Future 12 months: Automation of payroll management; introduction of the new payroll calculation module in the treasury (with web-based access in budget organizations) for effective pay administration. 60 II. IMPROVING THE EFFICIENCY OF SOCIAL SPENDING 4. EDUCATION: FOCUSING ON OUTCOMES50 Montenegro has spent as much on public education as other countries in the region, but this has not yet resulted in comparable education outcomes. Major reforms in the higher education sector still lie ahead, with the key challenge consisting of improving the quality of education while curtailing fiscal expenditures in this sector. In the ongoing education-sector reform process, Montenegro faces difficult choices in an attempt to carve out fiscal resources to improve the quality of education. Given that a non- negligible percentage of teachers has already passed retirement age, one potential intervention for Montenegro could be implementing a policy of workforce attrition with relatively limited effects on overall teaching ratios and class sizes. This would, however, require more extensive research into the impact on local communities and the likely changes to the organization of schools and universities. A. INTRODUCTION 4.1. Comparable to other countries in the region, Montenegro has spent about 4½ percent of GDP on public education, but outcomes are not yet commensurate with this level of expenditure. Montenegrin student performance Programme for International Student Assessment (PISA, 2009) has revealed inadequate levels in reading, mathematics, and science among 15-year-olds51 and, as such, pointed to the need to continue (and deepen) the ongoing reform of teaching methods, curricula, and education-sector financing. On the whole, however, the country has succeeded in ensuring that almost all (non-Roma) children have been enrolled into primary education. Likewise, in secondary education, the overall net enrollment rate of 88 percent is close to the OECD average. Thus, beyond enrollment rates, the school reform process is geared towards ensuring improved quality of learning, linking reforms to measuring students’ academic outcomes on all levels—from elementary schools to universities. Over the past five years, Montenegro has invested considerable resources in primary education, including on upgraded infrastructure, new equipment, revised textbooks, and teacher training, with first indications of early success. The national assessment for 2008 found a positive correlation between the length of a given school’s participation in the school reform process and the overall average score; see Ispiti centar (2009). 4.2. Montenegro faces difficult choices in the twin challenge of improving the quality of education and curtailing fiscal expenditures in this sector. With the restructuring of Government in December 2010, particular attention was given to education-sector reforms, as signaled by the splitting of the Ministry of Education and Science into two separate ministries,52 charged with the implementation of a comprehensive reform strategy aimed at (i) improving education outcomes (which lag behind those of comparator countries); (ii) capping per-student expenditures (which are roughly at regional—but still below OECD—averages); and (iii) rebalancing the composition of spending on public education (where an exceptionally high percentage of expenditures are used for teachers’ salaries). To be able to achieve all of these objectives concomitantly, Government will have to consider a (gradual) reduction in the 50 Prepared by Nina Arnhold (ECSH2), Toby Linden (SASED), and Roberta Malee Bassett (ECSH2), with input from Danijela Vukajlovic-Grba (ECSP2). 51 The student assessments were undertaken prior to the World Bank-supported curriculum reform and, consequently, can be expected to not fully represent the current situation. The PISA 2012 survey will give an indication of the success of related reforms as, by that time, the majority of students from reform schools will have been included in the test cohort. 52 The Ministry of Education and Sports and the Ministry of Science. 61 number of (non-)teaching staff—taking advantage of a situation in which a non-negligible percentage of teachers has already passed retirement age, thereby opening the possibility of implementing a (relatively painless) policy of workforce attrition with relatively limited effects on overall teaching ratios and class sizes. However, bringing about these changes will probably require more extensive changes to the organization of schools and universities. 4.3. More than one-fifth of Montenegrin citizen attend schools or universities. Elementary education is compulsory and was recently extended from eight to nine years. As in other countries of former Yugoslavia, there are three types of secondary education, viz., general secondary education, 4- year vocational training, and 3-year vocational training. The first two types of secondary education allow students to access the Matura exam, which serves simultaneously to certify learning at the upper secondary education level and provides the basis for access to tertiary education. In 2009/10, 74,539 pupils were enrolled in primary education and 31,758 students in secondary education, about one-third of the latter in general secondary schools (gymnasia); see Annex A4.1. There are three accredited universities in Montenegro, of which the University of Montenegro (UCG, Univerzitet Crne Gore) is the only public university. The Mediterranean University (UM, Univerzitet Mediteran) and the University Donja Gorica (UDG) are the two (accredited) private universities.53 Collectively, these institutions provide Montenegro with a university density that is comparable to—or rather, as comparably high as in—most other countries in the region (Table 4.1). At the time of its establishment in 1974, the UCG was the only institution of higher education in pre-independence Montenegro, comprising three faculties (economics, engineering, and law) as well as various smaller units. Today, the state university consists of 17 faculties, four research institutes, and three independent study programs, which are located in five different towns throughout Montenegro. The university currently enrolls about twenty-thousand students. Table 4.1. University Density per 100,000 Inhabitants ALB BIH BGR HRV EST LVA LTU MKD MNE ROU SRB SVN Number of universities 45 25 42 19 20 25 22 19 3 84 25 29 Public 13 9 … 8 12 6 … 5 1 26 8 4 Private 32 16 … 11 8 19 … 14 2 58 17 25 Universities per 100,000 citizens 0.7 1.6 1.8 2.3 0.7 0.9 1.5 1.1 2.1 2.6 3.0 0.7 4.4. Looking forward, the key challenge for the Montenegrin education sector consists of balancing fiscal-sustainability considerations with increasing the quality of outcomes. Within these constraints, and given the particular from the EU integration process, this chapter will recommend that the education-reform agenda be linked to the overarching competitiveness and skills agenda. Already in the review of the education sector contained in the World Bank’s (2006) Public Expenditure and Institutional Review (the “2006 Report”), it was argued that the basic policy choice—on the basis of projected budget plans—consisted of measures to contain sharp increases in salaries and staff costs or to have resources for non-salary expenditures be used to further the school reform and quality enhancement agenda (as these two expenditure items represented significant budget commitments, especially in 2007 and 2008). The conclusions were drawn on the assumption that wages would rise by around 8–9 percent. The 2006 report identified—still relevant—areas in which costs could be reduced so as to create head room for these two expenditure items within the expected overall education budget, including (i) 53 For basic data on the Montenegrin education system, see Table A1 and A2 in Annex. 62 reducing the number of very small schools and school units; (ii) reforming secondary vocational education by shifting towards more general education programs; and (iii) eliminating three-year programs. 4.5. During the boom years, Government Table 4.2. Overall Net Enrollment Rates, 2008 increased significantly the overall education budget beyond original projections. However, against the Primary Secondary backdrop of a considerably tighter fiscal environment, the World Bank has been asked to reflect again on reforms that could both improve outcomes and contain Japan 100.0 Sweden 99.2 UK 99.6 France 98.5 expenditures. This chapter will therefore not only New Zealand 99.5 Japan 98.3 discuss the quantitative dimension of public Poland 99.5 Finland 96.5 expenditures in education but also look at quality Greece 99.4 Norway 95.8 aspects of education-sector expenditures. This chapter is Sweden 98.9 Korea 95.5 organized as follows. The next section will review Korea 98.8 Spain 95.0 progress in the education sector since the 2006 report. Netherlands 98.7 Poland 93.8 Access and quality of education—as well as outcomes Norway 98.7 UK 93.2 documented by PISA54 results—are described, as are Portugal 98.7 Greece 91.0 Belgium 98.4 Hungary 90.9 items of budgetary spending be reviewed. The final France 98.4 Iceland 90.2 section argues for refocusing education on skills Germany 98.2 Denmark 89.6 provision and contribution to Montenegro’s Montenegro 98.2 Ireland 88.4 competitiveness, especially in the tertiary sector. It also Mexico 98.1 Italy 88.4 reviews options for the Government in improving the Iceland 97.6 Netherlands 88.3 quality and efficiency of education spending—within Ireland 97.1 USA 88.3 the fiscal constraints that the sector faces. Italy 97.1 Australia 88.0 Australia 96.9 Portugal 87.9 B. EDUCATION OUTCOMES Finland 96.0 Montenegro 87.6 Luxembourg 95.7 Belgium 86.9 Denmark 95.6 Switzerland 84.7 4.6. Overall enrollment rates in primary and Spain 95.6 Luxembourg 83.7 secondary education are comparable to other Turkey 94.7 Turkey 73.9 European and OECD countries. Almost all children Switzerland 94.2 Mexico 42.4 are, according to official statistics, enrolled in primary USA 92.0 education at the appropriate age, which places Hungary 89.7 Montenegro favorably in the midst of higher-income Czech Republic 89.6 countries (Table 4.2). In secondary education, the overall net enrollment rate of 87.6 percent is creditable Sources: UNDP (2009) for Montenegro; and www.worldbank.org/education/stats for other countries. but somewhat lower than the OECD average. 4.7. The situation of the Roma minority, however, is quite different. Similar to other countries, accurate statistics on Roma are very difficult to obtain,55 but unofficial estimates put the primary enrollment rate at 25.2 percent, the completion rate of the first cycle of compulsory education at 32 percent (compared to 98 percent for the general population) and the corresponding rate for the second cycle at 7 percent (compared to 86 percent of the general population); see Roma Education Fund (2009). To be able to make noticeable progress in getting the remaining pupils into school will depend on 54 OECD’s Programme on International Student Assessment. For details see http://www.pisa.oecd.org/pages/ 0,2987, en_32252351_32235731_1_1_1_1_1,00.html. 55 Figures for the total number of Roma in Montenegro vary between 2,601 from the official census in 2003 and over 20,000 from NGO sources (which would be approximately 3 percent of the total population). In Montenegro, official documents treat Roma, Askhalia, and Egyptians as one group. Many of these individuals are not registered, which in part explains the very high variance in estimates of population numbers. 63 ensuring that those Roma children already enrolled will attend school regularly and with a view to graduating.56 4.8. The ongoing school reform process has led to better learning outcomes for students in primary schools. Over the past 5 years, Montenegro has invested considerable resources in primary education to improve the quality of learning.57 These changes, which included upgraded infrastructure, new equipment and textbooks, and teacher training, seem to have had some positive effect. The analysis of the national assessment results for 200858 indicates that the longer a school had participated in the school reform process the better was the overall average score—even though the improvements have been modest. Notwithstanding the large range of average school scores, the data show a less pronounced tail of poor performance in the schools that have participated in the reform process for a longer period of time (Figure 4.1). Regardless, these findings should be treated with caution since schools were not randomly assigned to the reform process and it could be that better schools were included in the early years of the reform. Another important caveat to the findings is that the schools which entered the reform process at the end are much smaller schools than the national average (Table 4.3). In general, small schools have much broader ranges of performance (at both the top and bottom ends). Curiously, there does not appear to be a consistent correlation between average school size and performance: for the grade 3 mother tongue assessment, the larger the school the better the performance, but for other assessments and other grades, this correlation disappears and in some cases reverses (Ispitni centar, 2009). Figure 4.1. Average School Performance on Grade 3 Mother Tongue, 2008 4 Year of entry into school reform 3 process 2 1 0 0 20 40 60 80 100 Average school score Source: Assessment Center (Ispitni centar, 2009). 56 Data from the Employers Survey 2009/10 (Employment Agency for Montenegro, 2010: 66) show that lack of education is a major obstacle for labor market integration of Roma. 57 The education reform was supported by the Education Reform Project and an associated US$5 million World Bank loan. 58 Montenegro conducts national assessments each year and the results for 2009 are available. However, only every four years are the assessment conducted through a completely external process; in the other years, teachers mark the test papers of their own students. The first assessment in 2008 was wholly external which is why these results are reported here. 64 Table 4.3. School Characteristics and Performance in Mother Tongue, 2008 (In Grade 3, by Year of Entering the School-Reform Process) Year Average Standard 5 percent Number of Average no. entering score deviation significance schools of pupils reform 1 67.105 7.599 3.330 20 86.8 2 65.082 7.113 2.846 24 89.8 3 62.436 8.129 3.187 25 80.3 4 59.989 12.557 3.669 45 29.6 Source: Assessment center (Ispitni centar, 2009). 4.9. Learning outcomes towards the end of secondary education show some signs of improvement, but overall results still lag behind those in other countries. Since the 2006 report, important new information on outcomes of Montenegro’s education system has become available. Montenegro participated in the 2006 and 2009 PISA international assessments, which measured the performance of 15-year-olds in reading, mathematics, and science. The two PISA studies’ results have shown that, between 2006 and 2009, (i) the overall performance remained low relative to key comparator countries (Figure 4.2); (ii) Montenegro was able to improve its average scores, significantly in reading and modestly in mathematics, while performance in science worsened (Figure 4.3); (ii) outcomes were equitable59 and socio-economic status of pupils had one of the lowest effects on performance of any participating country.60 As Montenegro strives to improve the quality of its education system, retaining this equitable outcome distribution will be important. Figure 4.2. Education Performance and Per Capita Income Trendline: PISA = 308.2 + 54.4 ln(gdp); R2 = 0.7225. Sources: OECD; IMF (April 2011 WEO data); and World Bank staff calculations. 59 This assessment excludes the Roma population, of whom only about 7 percent finish compulsory schooling. 60 The equity of outcomes in mathematics was not reported in OECD 2007. 65 Figure 4.3. Changes in Learning Outcomes, PISA 2009 vs. PISA 2006 Sources: OECD; and World Bank staff calculations. 4.10. The 2006 and 2009 PISA results imply that Montenegro will have to improve considerably on its overall performance in the education sector, with Montenegro’s average performance in each subject area having been significantly below relevant comparators in the region. As shown in Figure 4.3, Montenegro’s education system has performed poorly in relation to its income level, with its relative position being significantly below the trend line. This discrepancy is, of course, even more pronounced vis-à-vis higher-income comparators that are generally viewed as “benchmark” countries.61 The measured pace of overall improvement in performance (which do not reflect the effects from the latest education-sector reforms) remains to slow and uneven to be able to translate into productivity improvements necessary to narrow the per-capita income gap with the EU average (Figures 4.4 and 4.5). 4.11. Montenegro had a large proportion of students performing at the lowest levels. Level 2 is considered the baseline performance—i.e., the level at which pupils have a “basic grasp” of key concepts and skills. Across the OECD as a whole, 81.2 percent of pupils achieved this level in reading in 2009, compared to only 60.5 percent of young people in Montenegro. For mathematics, the corresponding figures are, respectively, 78.0 percent and 41.6 percent. Moreover, in each subject area, Montenegro had virtually no students performing at the highest level (level 6). In Montenegro, only 0.4 percent of pupils performed at level 6 in reading and 0.1 percent in mathematics, compared to OECD averages of 8.6 percent and 3.3 percent, respectively. Montenegro will need to closely monitor the outcomes of future PISA rounds. The 2012 results, in particular, will help to prioritize future reform initiatives as the majority of participating students will have been exposed to the aforementioned education reforms —as such, the results should serve as an important assessment tool of the reform impact and effectiveness. 61 Montenegro’s average reading score in 2009 was 408, as compared to the OECD mean of 493. 66 In percent of GDP 0 1 2 3 4 5 6 7 8 Iceland United States Denmark Korea Source: OECD (2010a) Chile Sweden Belgium Canada France New Zealand United Kingdom Mexico Finland Portugal Netherlands Sources: OECD, IMF, and World Bank staff estimates. Switzerland Norway 67 Austria Poland Australia Hungary Japan Spain Germany Montenegro Ireland Czech Republic Italy Slovak Republic Figure 4.4. Budgetary Spending and Education Outcomes OECD average EU19 average Figure 4.5. Total Expenditure on Educational Institutions, 2007 Israel Russian Federation Slovenia Brazil Estonia C. EDUCATION SPENDING 4.12. Between 2006 and 2010, total education expenditures increased by 50 percent, from €89.4 to €134.8 million. The largest nominal increases (year-on-year) occurring in 2007 (+30 percent) and 2008 (+24 percent). Altogether, the increase in the education budget between 2006 and 2009 exceeded the increase in average living costs (+15 percent), reflecting sizable real increases. Consolidation of education-sector expenditures started in 2009 and continued into 2010, with both years showing a nominal tightening in spending; see Table 4.4. The largest percentage increases between 2006 and 2009 were recorded in funding for pre-school education (84 percent) and in science (by a staggering 163 percent), although science spending has fluctuated dramatically during this period—with spending going up by 80 percent in 2008 compared to 2007, before falling by more than 60 percent in 2008, and returning back to the 2008 levels in 2010. Despite the recent increases, however, the pre-school and science sub-sectors remain relatively small parts of the overall budget (at 8.5 percent and 1.3 percent respectively). 4.13. While this report does not contain a detailed discussion of pre-school provision, this is a sub-sector which deserves to be analyzed in more detail in the future . A range of World Bank and other studies have highlighted the significant returns on investment in early childhood development (ECD), which show that, at the time when formal schooling starts, there are already important developmental differences amongst children.62 Even though ECD is resource-intensive and needs political commitment, related investments pay off in terms of social inclusion and competitiveness. Montenegro’s decision to invest in pre-school education is, therefore, in line with global best practices and shows commendable awareness of the full spectrum of education investments necessary to support comprehensive improvements in student outcomes. The largest increases, in money terms, were in primary education (in 2009, €21.3 million was spent beyond that what was made available to this sub- sector in 2006) and secondary education (with a difference of €10.6 million).63 4.14. As a share of the education budget, Montenegro spends 17 percent on primary education, as compared to the OECD average of 20 percent and the EU19 average of 19 percent64. Poland, a country which has significantly improved its PISA reading scores since 2000, ranks at the upper end with 25 percent; the Czech Republic, a country which has faced a steady decline in PISA reading scores since 2000, is at the lower end, with 14 percent per capita PPP spending for primary education. Thus, the major challenge facing Montenegro’s policymakers—especially during this time of crisis—is ensuring that (i) there will be sufficient funds for education; and (ii) these funds will be directed towards quality- enhancing measures. 4.15. The most important driver of the large increases in primary and secondary education spending has been the very sharp rise in teachers’ salaries and, to a lesser extent, in spending on utilities. The Government funded significant increases in teacher and other staff salaries, following the conclusion of the Collective Agreement with the teachers’ unions in 2006. A number of changes were introduced to wage rates, contributions, and taxes. Overall, primary education expenditure on gross salaries increased by 10.6 percent in 2007, 28.3 percent in 2008, and 7.9 percent in 2009; similar increases were recorded in secondary education (Box 4.1). Utilities as well represented a principal source of overall spending increases: between 2006 and 2008, they increased by more than 3.5 times in 62 For details, see, e.g., World Bank (2010). 63 The categories of primary and secondary education spending followed the data provided by MoES, though the ‘other’ category in Table 3 includes significant spending on dormitories and scholarships which it has not been possible to attribute to the different levels of education. 64 Per-student primary education expenditure as percentage of per capita GDP; see OECD (2010a). 68 primary education and more than 6 times in secondary education, largely caused by a country-wide increase in electricity prices.65 Table 4.4. Education Spending, 2006–10 2006 2007 2008 2009 2010 (In millions of euro; unless otherwise indicated) Total 89.4 116.3 144.0 134.9 134.8 percent increase year-on-year 30.2 23.8 -6.3 -0.1 in percent of GDP 4.2 4.3 4.7 4.5 4.5 in percent of budgetary expenditures 9.2 10.0 9.3 8.8 9.1 Pre-school 6.3 9.2 11.6 11.4 11.6 percent increase year-on-year … 44.8 27.1 -2.2 2.1 in percent of GDP 0.3 0.3 0.4 0.4 0.4 in percent of budgetary expenditures 0.7 0.8 0.7 0.7 0.8 Primary 39.7 54.7 66.8 61.0 62.5 percent increase year-on-year … 37.7 22.2 -8.7 2.4 in percent of GDP 1.8 2.0 2.2 2.0 2.1 in percent of budgetary expenditures 4.1 4.7 4.3 4.0 4.2 Secondary 17.7 25.1 31.0 28.3 28.2 percent increase year-on-year … 41.5 23.8 -8.6 -0.6 in percent of GDP 0.8 0.9 1.0 1.0 0.9 in percent of budgetary expenditures 1.8 2.2 2.0 1.9 1.9 Tertiary 11.5 13.0 16.3 17.4 14.8 percent increase year-on-year … 13.4 25.0 6.9 -15.2 in percent of GDP 0.5 0.5 0.5 0.6 0.5 in percent of budgetary expenditures 1.2 1.1 1.0 1.1 1.0 Science 0.7 1.0 1.8 0.6 1.7 percent increase year-on-year … 47.1 79.6 -66.2 194.3 in percent of GDP 0.0 0.0 0.1 0.0 0.1 in percent of budgetary expenditures 0.1 0.1 0.1 0.0 0.1 Other 13.4 13.4 16.4 16.1 16.0 percent increase year-on-year … -0.3 22.6 -1.9 -0.8 in percent of GDP 0.6 0.5 0.5 0.5 0.5 in percent of budgetary expenditures 1.4 1.2 1.1 1.1 1.1 Notes: 1. Tertiary is grant to University of Montenegro. 2. Figures are expenditures, except 2010. Sources: Government of Montenegro; and World Bank staff calculations and estimates. 65 According to figures provided by MoES, there was no spending on utilities in primary or secondary education in 2009. The budget figures for 2010 are still very high compared to those in 2006 —about 2.5 times in primary education and more than 3.5 times more in secondary education. 69 Box 4.1. Determining Teachers’ Salaries in Montenegro Teacher salaries in Montenegro are set under the branch (collective) agreement. As described in the previous Chapter, the salary base is a minimum rate and monthly salaries are a multiple of this base and the coefficients set by the branch agreement. Multiples are determined by qualifications, particular roles (e.g., being a mentor or an organizer of practical education), and years of experience. The Government provided a 10 percent increase in the coefficients on October 1, 2007 and an additional 20 percent increase on January 1, 2008. Source: Ministry of Education and Science; and World Bank (2006). 4.16. After increases in 2007 and 2008, Montenegro’s spending on quality-enhancing, non-salary items has remained low. In 2009, the proportion of the recurrent education budget used for gross salary costs was 93 percent in primary education and 92 percent in secondary education (Table 4.4).66 These are the same figures as in 2006, but in 2007, 11 percent of recurrent spending in both primary and secondary education (including vocational education and training) was on non-salary costs. In 2008, the figure was 9 percent (again for both primary and secondary education); however, almost half of these non-salary expenditures were for utilities, rather than quality-promoting items like textbooks or other learning materials. In 2010, the proportion of non-salary expenditure is expected to be even lower, at 6 percent. Capital expenditures as well were very low, at about one percent in 2009 and projected to fall, as a percentage, in both 2010 and 2011. As already noted in the 2006 report, Montenegro’s spending on non- salary items compares unfavorably to other countries. Within the OECD, spending on teachers’ compensation (i.e., gross salaries) accounts for 59.0 percent of total spending in primary, secondary, and non-tertiary higher education, with another 13.8 percent spent on compensation for other employees. On average, OECD countries spend 19.2 percent on non-salary recurrent items and 7.6 percent on capital expenditures.67 It is unfortunate that the large increases granted to education-sector spending have, for the most part, not been directed towards quality-enhancing investments (see also Tables 4.5 and 4.6). Table 4.5. Per-Pupil Spending in Montenegro and Selected Countries, 2009 Post-transition EU member countries Emerging economies MNE OECD EST HUN POL SVK SVN BRA CHN MEX (In US dollars) Total 2,095 8,216 4,783 4,811 4,134 3,694 7,560 2,080 1,404 2,598 Pre-school 1,664 5,447 2,232 4,304 4,658 3,419 8,464 1,599 882 1,979 Primary 1,523 6,741 4,058 4,656 4,063 3,499 7,981 1,862 778 2,111 Secondary 1,742 8,267 4,869 4,225 3,590 3,219 7,267 1,750 1,153 2,236 Tertiary 2,594 8,970 5,214 5,365 4,637 4,922 7,037 10,326 1,404 5,862 (In percent of 2007 per capita GDP) Total 21.2 23.8 29.6 35.1 37.1 26.5 32.1 28.3 53.1 26.5 Pre-school 16.8 15.8 13.8 31.4 41.7 24.5 35.9 21.8 33.4 20.2 Primary 15.4 19.5 25.1 34.0 36.4 25.1 33.8 25.4 29.4 21.6 Secondary 17.6 23.9 30.1 30.8 32.2 23.1 30.8 23.8 43.6 22.8 Tertiary 26.2 26.0 32.3 39.2 41.6 35.3 29.8 140.6 53.1 59.9 Source: Montenegro, authors' calculations; otherwise, OECD 2010a (Table B1.1.a) and IMF (April 2011 WEO data). Notes: n.a. = data not available; Slovenia, primary includes lower secondary. 66 Because of data limitations, it has only been possible to analyze the primary and secondary education budget in detail. However, these sub-sectors represent about two-thirds of total education spending. 67 Authors’ calculations based on data from OE CD 2010, Table B6.2b. 70 Table 4.6. Categories of Expenditure by Sub-Sector, 2009 Primary Secondary euro percent euro percent Total 61,012,852 28,095,516 Recurrent 60,523,681 99 27,707,426 99 Capital 489,171 1 388,090 1 Recurrent, of which Gross staff costs 56,562,882 93 25,459,327 92 Recurrent w/o gross staff costs 3,960,799 7 2,248,099 8 Total w/o staff costs, capital 3,960,799 6 2,248,099 8 Source: World Bank staff calculations based on MoES data. 4.17. In addition, due to volatility of the tertiary education budget, the structure of expenditure in this sub-sector is equally problematic. While the authors did not have access to the executed budget, budget plans for 2011 and recent years show the same patterns of misdirected resource allocation in tertiary education as in the other sub-sectors. Out of the €13.5 million foreseen in 2011 for the University of Montenegro, €12.5 million (or 92.6 percent) have been earmarked for salaries and €0.367 million (2.7 percent) for other compensation. The UCG budget, as far as public funds are concerned, is thus a mere salary budget, not leaving any space for training, academic exchange, or other quality-enhancing measures. The aim, of course, should be a more dynamic model that rewards performance, allows for broad investment in sustainable improvements across the sector, and ensures that activities of the university are aligned with overall public economic and social priorities. 4.18. Government has decreased spending on tertiary education but increased the budget for science. Across the sub-sectors, there has been little change in primary and secondary education but a significant reduction in the central UCG budget (about 15 percent), while the budget allocated for science and innovation activities has increased.68 In science and innovation, as in higher education, it will be crucial to agree on public priorities for the sector, foster closer connections to the private sector, and develop a more dynamic model of funding. There are no expected increases in teacher salaries for higher education and, in the 2011 budget, the most noticeable change is another reduction in tertiary level spending (by about 9 percent). Staffing levels 4.19. There has been virtually no change in the number of pupils and teachers since 2006. The pattern of pupil enrollment and staffing levels in both primary and secondary education has shown persistent fluctuations since 2006, but within very small bands (Table 4.5). The numbers of pupils in both primary and secondary education have risen by 0.3 percent, while the number of teachers in primary education has shown a marginal trend decline of 0.3 percent, resulting in a slight increase in the pupil-teacher ratio from 15.26 to 15.36. In secondary education, the number of teachers has increased over the period as the number of pupils has declined, resulting in a reduction of the pupil-teacher ratio from 13.94 to 13.31. 68 The increase in 2011 is almost 300 percent, but this amount only takes it back to the level in 2009, compensating for the very large cut in 2010. 71 4.20. Montenegro’s student-teacher ratios are slightly below those in comparator countries, at 16.4 in primary education (compared to 14.6 in the EU) and 13.7 in lower secondary education (11.5 in the EU); see OECD (2010). Similarly, class sizes in Montenegro are close to the international norms. Within the OECD, class size in primary education is 22 students per class (with a large standard deviation around this average, largely reflecting cultural factors). In lower secondary education, class sizes tend to be larger (by about two students per class) than in primary education. Table 4.7. Student and Teacher Numbers, 2006–10 Primary Education Secondary Education 2006/07 2007/08 2008/09 2009/10 2006/07 2007/08 2008/09 2009/10 Number of students enrolled (total) 74,339 74,539 74,220 74,596 31,647 31,381 31,333 31,758 Number of classes 3,451 3,411 3,374 3,390 1,164 1,175 1,175 1,150 Number of teachers 4,871 4,692 4,974 4,855 2,271 2,345 2,297 2,386 Pupil-teacher ratio 15.26 15.89 14.92 15.36 13.94 13.38 13.64 13.31 Pupil-class ratio 21.54 21.85 22 22 27.19 26.71 26.67 27.62 Expenditure per student (in euro) 534.40 733.80 900.70 818.00 560.00 798.50 989.90 892.20 Expenditure per student (in percent of GDP) 12.5 15 18.9 16.4 13.1 16.3 20.8 17.9 Sources: Ministry of Education and Science; and authors’ calculations. D. IMPROVING QUALITY AND EFFICIENCY 4.21. Montenegro will face very difficult choices to ensure that there are sufficient resources for improvements in education quality. The reform agenda will have to include the closing of small schools, increasing pupil-teacher ratios, making savings on energy efficiency, and/or reducing the costs of administration. 4.22. There is some scope for increasing class sizes overall, even though Montenegro’s current averages are close to international means. Class size varies considerably across countries. In Montenegro, the average class size in primary (secondary) education is 22.0 (27.6).69 As shown in Table 4.8, these figures are above OECD averages (marginally in the case of primary education). However, European countries tend to have lower class sizes than countries in other regions. It is important to keep in mind that these averages vary across a wide range and that there is no strong correlation with academic outcomes. 4.23. Teachers are a critical factor in improving the quality of education. It is important to ensure that both the number of teachers is appropriate and there are opportunities for professional development. By international standards, the proportion of the Montenegrin education budget for salary expenditures is very high—which means, inversely, that the level of expenditure on non-salary items is very low. This problem is aggravated by the relatively low per-student spending. There are different options for the Government to manage the number of staff employed in the education sector. Staff numbers can be allowed to decrease as a result of exits from the system, either on retirement or other voluntary mechanisms. However, in the case of Montenegro, the overall number of staff has not fallen as new staff has been recruited at levels higher than departures. The Government might want to (i) assess more closely staffing needs; and (ii) manage more actively the number of staff (by imposing retirement rules). 69 Authors’ calculations based on MoES data. 72 Table 4.8. Average Class Sizes, 2008 Lower secondary education Primary education (general programmes) Public Private Public Private institutions institutions institutions institutions OECD countries Australia 23.2 24.9 23.0 24.7 Austria 19.3 20.6 23.3 24.0 Belgium (Fr.) 19.7 20.8 n.a. n.a. Chile 28.8 31.6 29.5 31.6 Czech Republic 20.0 16.2 22.5 20.3 Denmark 20.0 16.8 20.4 18.1 Finland 19.8 18.4 20.0 21.7 France 22.7 23.1 24.1 25.1 Germany 21.9 22.4 24.7 25.5 Greece 16.6 20.4 21.7 25.6 Hungary 21.4 19.3 22.8 21.3 Iceland 18.0 14.4 19.8 13.0 Ireland 24.3 n.a. n.a. n.a. Italy 18.6 20.1 20.9 22.0 Japan 28.0 32.8 33.0 35.5 Korea 29.9 30.9 35.5 34.4 Luxembourg 15.6 18.1 19.5 21.2 Mexico 19.7 20.7 29.1 24.9 Netherlands 22.4 n.a. n.a. n.a. Poland 19.3 12.7 24.0 18.1 Portugal 18.6 20.1 22.2 23.4 Slovak Republic 19.4 18.3 22.0 21.1 Spain 19.7 24.4 23.6 26.2 Turkey 27.3 18.0 a. a. United Kingdom 25.7 13.6 21.3 12.8 United States 23.8 19.3 23.2 19.1 OECD average 21.6 20.8 23.7 23.2 EU19 average 20.3 19.1 22.2 21.8 OECD partner countries Brazil 27.1 17.8 30.5 25.0 China 36.6 41.8 55.5 51.6 Estonia 18.4 15.1 21.6 15.6 Indonesia 27.4 22.8 37.7 33.1 Israel 27.6 a. 32.5 a. Russian Federation 15.8 10.4 17.9 9.8 Slovenia 18.5 17.0 20.4 23.5 Note: Netherlands year of reference is 2006. n.a. = data unavailable; a. = data inapplicable because category does not apply. 73 4.24. At present, the system employs a significant number of education staff above the official retirement age; if they were to retire, the resultant savings could total up to €4 million per annum. The official retirement age has already been passed by 343 primary school teachers (or 7.1 percent of primary teachers) and 74 secondary school teachers (3.1 percent). In addition, among non-teaching staff, there are 93 employees beyond the official retirement age. An obvious route to reducing staff numbers would the enforcement of the retirement age, which would increase the pupil-teacher ratio in primary education from 15.36 to 16.60 and in secondary education from 12.89 to 13.31. These averages would still be well within the international norms. According to Ministry figures, the total average salary cost of a primary school teacher, including social contributions, is €8,828 per year, while for a secondary school teacher the amount is €9,585.70 Thus, if the 343 aforementioned primary school teachers were to retire, the total savings would be €3.03 million per year. The retirement of secondary scho ol teachers would save another €0.71 million, bringing the total savings—together with non-teaching staff and on the assumption that no new recruitment occurs—to close to €4 million.71 A further option would be to look at potential savings from those who are close to retirement age, resulting in comparable wage savings from those already over the retirement age. Given the total gross annual cost of a teacher, offering early-retirement benefits could yield additional savings. There are 402 teachers (343 primary and 69 secondary) and 82 non-teaching staff (57 Table 4.9. School Units and Classes With Less Than 10 primary and 25 secondary) who are within 5 Students, 2009/2010 years of retirement, which would yield total savings of roughly the same amount as those over retirement age. However, these steps No. of students School units Classes Avg. class size would need to be considered very carefully (as will be discussed further below). 1 22 22 1 4.25. The fiscal costs and socio-economic 2 42 42 2 benefits of maintaining the 155 small 3 65 67 2.9 schools (and school units) with less than 10 4 80 91 3.5 pupils need to be weighed carefully. As 5 106 121 4.4 discussed in the 2006 report, there are a 6 113 128 5.3 number of very small schools and school units 7 126 143 6.2 in Montenegro. National assessment data 8 138 156 7.1 shows that the quality of schools in rural areas is lower than in other schools; for example, in 9 146 164 8 reading in grade 3, students in the northern part 10 155 180 8.6 of Montenegro scored on average 239.991 compared to 252.728 for students in the central Source: MoES; and World Bank staff calculations. region (which includes Podgorica) (Ispitni centar, 2009). There are currently 155 school units in primary education with 10 or fewer pupils, the same number of such schools in 2005/2006 (Table 4.9) and 3 secondary school units with 50 or fewer pupils. Assuming there were 155 teachers in the 155 primary school units with 10 and fewer pupils (there may be more teachers because there are 180 classes), the savings would be €1.37 million per year. This is likely a conservative estimate since it does not include the cost of utilities and non-teaching staff costs. 70 These figures are calculated using the average figure for teaching staff with the title of “teacher”. There are, though, a number of different staff positions in schools which are considered teaching positions (school directors and deputy directors, specialist teachers and so on) and it is not known what proportion of those over retirement age are from these different categories. However, if anything, these average teacher cost figures are likely to be low since the older teachers have higher salaries and so those over retirement age can be expected on average to have the highest salaries. 71 For reasons given in the previous footnote, this is likely to be a conservative figure. 74 4.26. Closing small schools would need to be considered very carefully, however, and should not be applied in isolation. There is anecdotic evidence that it is difficult to fill positions at smaller schools in rural areas (particularly in the North). Any planned attrition in the teacher corps must, therefore, take into account the localized impact of losing teachers on student learning and success potential. It may well be the case that many of the teachers above retirement age are located in areas where teacher shortages may result from retirements in the existing pool of teachers. Teachers in small schools serving the country’s most remote areas might not have sufficient professional contact and access to training, but their impact on the students should not be underestimated. Though these schools seem to be disadvantaged on a variety of dimensions, but they do serve an important social (and demographic72) role, and it is imperative to open up learning possibilities for teacher and students and ensure access to good quality schooling for the children in these areas. The value of high-quality education, however, needs to be communicated to parents and communities and the safe access of children to remoter schools needs to be ensured. 4.27. Closing school units will need to be put in the context of the Government’s plans to promote economic development in the rural areas of Montenegro. For these economic development plans to be realistic, they must include enhancing the human capital of these regions. This places a premium on ensuring that young people get a good quality education in these regions, so that they can take advantage of and even generate economic opportunities. 4.28. To ensure pupils’ continued access to schooling, transportation arrangements would need to be carefully considered when closing very small school units. At a minimum, it would be necessary to enhance the transportation infrastructure and road safety so that children can safely—and throughout the entire school year—commute to their respective schools. This would imply that, as necessary complement to the consolidation of the rural school network, Government would have to accelerate the rehabilitation of the rural roads and integrate the transport infrastructure as one central pillar into a broader regional development strategy; see Chapter 3-B in World Bank (2008). In addition, it may also be necessary to consider more heterodox solutions, such as changing the academic schedule by lengthening the winter and reducing the summer break, to accommodate local constraints, while weighing the corresponding investments with the opportunity costs of not providing good quality education to the majority of the population in rural areas. 4.29. While other mechanisms, such as an appraisal or licensing system, are sometimes discussed in the context of savings, these are not likely to be effective approaches to reducing teacher numbers. The identification of teachers and other staff over (and close to) the official retirement age was discussed as a potential first option, largely because it is an instrument that the Government already has at its disposal. It also offers the advantage of being objective and transparent, with the numbers of staff involved and the costs easily identifiable. However, this report cautions against employing the licensing system as a cost-cutting instrument as this would undermine efforts aimed at improving teacher performance and implementing other quality-enhancing measures. 4.30. Broadly there are two ways in which teachers are evaluated in Europe, through individual assessments and through evaluations of the performance of the school as a unit. After a period in the 1990s, when individual assessments of teachers were on the decline, there has been an increase, since the middle of the 2000s, in the number of European countries that have (re-)introduced such schemes (Eurydice, 2008). Thirteen out of 26 OECD countries had some sort of periodic evaluation. In most of the remaining countries, no evaluation is done once a teacher acquires tenure (OECD, 2005). Where 72 Closing small schools in the mountainous North would accelerate the already considerable migration trend from the rural areas to the urban centers in central and coastal Montenegro. 75 assessment is done of individual teachers, in the vast majority of cases, this assessment is carried out by the head teacher (in, for example, the Czech Republic, Lithuania, or Slovenia). In only a few (Central European) countries, teacher-evaluation mechanisms have been linked to salaries. Evaluations of individual teachers are an opportunity to provide professional development (or, in the case of poor performance, remediation). Very few teachers are made redundant as a result of assessments of performance (OECD, 2005). Establishing a system for appraisal will take several years (Box 4.2). Box 4.2. Description of Licensing System for Teachers in Montenegro In order to be licensed, teachers—after one year of service—have to pass the professional exam. There are no procedures or requirements for re-certification or re-licensing. The Government has recently started to think of re-licensing in the context of professional development of teachers. Professional development strategies for each sub-sector of education are being developed but all have yet to be completed. The Ministry of Education and Science has established a committee for preparation of the general framework for licensing and re- licensing, comprised of five persons, from the Ministry, the Education Bureau and the University. No conclusions have been reached and no decisions have been made as of February 2011. At present, however, re- licensing is being considered within a five year framework during which teachers would collect proof of their professional achievements – such as published papers, studies, achievements of their students, etc. Source: MoES/Education Bureau. 4.31. Chile is an interesting example of a country which has instituted a comprehensive system of teacher evaluation. This was part of a comprehensive approach to improving the quality of teaching and learning in Chile, including school inspections and national assessments. The teacher evaluation system was introduced because of concerns about the quality of teaching. However, following the first evaluations in 2003, only 4 percent of teachers were considered unsatisfactory, though a further 30 percent were assessed as having only “basic” competence and also required to follow an individualized professional development program and be evaluated after one more year. Teachers who remained unsatisfactory performers were taken out of a teaching position but given another year to improve their skills. Only after failing a second re-evaluation, they could be made redundant (Box 4.3). 4.32. Making continuing professional development of teachers (CPD) a requirement for maintaining licensing or teacher re-certification is another possibility to improve teacher performance, one that is not generally used in Europe. CPD is a professional duty in most European countries, but it is an effectively optional practice (Eurydice, 2008). The numbers of hours required as minimum vary considerably—even where they are specified (Table 4.10). If significant resources were found to enable all teachers to access training, then this could be introduced as a means to re-certify teachers. The major challenge for this option is the actual provision of high-quality CPD for teachers and the creation of mechanisms to support teachers’ access these opportunities. 4.33. CPD investments are critical to improve the quality of teaching and compensate for the lack of attention paid to this feature of teacher performance. It would, in principle, be simple to introduce a re-certification system by tracking the amount of CPD undertaken by teachers. Key issues to address would be to define (i) the amount of required CPD; (ii) the types of training that would be available; (iii) CPD training quality standards (including certification of providers, perhaps along the lines of the approach implemented in FYR Macedonia). It would thus be necessary to establish guidelines (and possibly regulations) about the process of deciding what CPD an individual teacher undertakes, since the priorities of an individual may not coincide with the priorities for the school to which that teacher belongs. 76 Box 4.3. Chile’s Teacher Evaluation System Chile introduced a national system—the Evaluación del Desempeño Profesional Docente—through which teachers are evaluated every four years. This process was introduced after extensive consultations between and among the Ministry of Education, the Teachers’ Association, the Chilean Association of Municipalities, and the public at large. x The evaluation is based on four domains: (i) the preparation of teaching; (ii) the creation of a setting which promotes learning; (iii) teaching for the learning of all students; and (iv) professional responsibilities. Within each domain, there are between four and six criteria. The documentation for the evaluation is from four sources, viz., (i) a portfolio of the teacher’s work, including a video of them teaching; (ii) the teacher’s self-evaluation; (iii) a peer interview; and (iv) a report from the teacher’s manager. The actual evaluation is done by a trained and accredited person from outside the school. x The evaluation gives the teacher an overall rating of “excellent”, “competent”, “basic”, or “unsatisfactory”. Teachers in the first two categories are recognized by having preferential access to various benefits, such as training, study tours, mentorships, or conferences. Teachers in the bottom two categories have to undergo compulsory professional development —based on the identified weaknesses— and are re-evaluated a year later. If someone in the “unsatisfactory” category does not improve after this one year, they are removed from the teaching position and given one more year to reach professional standards. Only if no improvement is made after another year can they be fired as teachers. Source: Organization for Economic Cooperation and Development, 2005 Table 4.10. Minimum Annual Time (in Hours) for Teachers Spent on Continuing Professional Development, ISCED 1 and 2, 2006/2007 BEL EST CYP LVA LTU HUN MLT AUT ROU FIN GBR NOR Hours 21 32 57 36 35 17 12 15 19 18 35 35 Source: Eurydice (2008). 4.34. The approaches discussed earlier in this section look at the overall number of teachers in Montenegro. Decisions would still need to be made about which teaching positions at which schools are not filled or where redeployment of a teacher from another school would be needed to fill the position previously occupied by someone over or close to retirement age. All of these decisions must be considered within a context of cost-savings and efficiencies as well as equity and access for all students. The quality of teaching and learning could be further enhanced by paying more attention to CPD of teachers and connect professional development to the licensing system. E. REALIGNING NON-SALARY COSTS AND ENHANCING THE RELEVANCE OF TRAINING 4.35. Enrollment in vocational education courses at the secondary level, which have higher unit costs, remains high. According to Ministry data, more than two-thirds of students in secondary education are in vocational education, although the number of students enrolled in three-year programs has decreased over the past five years. The latter is in accordance with the findings of the 2006 report which noted that students in three-year program often do not—contrary to the intention of these courses—have good success when entering the labor market after graduation. 77 4.36. Montenegro suffers from a mismatch between acquired skills and knowledge and the needs of the labor market. For example, the Employment Agency of Montenegro (2010) found that the “lack of trained labor in the market” represented the single most important reason that employers had given for not being able to fill positions, and 61 percent of employees who received training in the previous year were trained in basic skills and knowledge (basic computer literacy and other courses, and knowledge of English language. Little attention is paid to upgrading labor market skills (Box 4.4). Montenegro is not the only country that has encountered problems in this area. Jakubowski et al. (2010:26), in an analysis of recent Polish reforms, caution “policymakers about the effectiveness of vocational schooling—when that schooling is not designed to improve math and reading skills … which have become the real vocational skills in the world of work today.” Indeed, the issue of the jobs/skills mismatch is a global challenge being addressed in every region of the world. Box 4.4. Life-Long Learning Montenegro will need to move beyond isolated reforms for different sub-sectors and develop a comprehensive life- long learning system, including an expansion of provisions at the pre-school and adult education level. First steps in this direction have been undertaken by the development of a Qualification Framework for Life-Long Learning, which will allow for more flexible learning paths and will make the recognition of degrees and prior learning easier. However, the situation in adult education gives reason for concern: in 2009/10, 1,363 employees (or 1.84 percent of the labor force) underwent training in Montenegro, a significant decrease compared to 8.46 percent in the previous year (Employment Agency of Montenegro, 2010:27) 4.37. In response, the Government has introduced reforms to the vocational education and training (VET) system in secondary education. The reforms included (i) modernized curricula and a move towards a modular system whereby students can move through a hierarchy of qualifications or pick up qualifications as they go through working life; and (ii) changes to entry for higher education to give greater access to this level of education for graduates from secondary VET programs. 73 Typically, however, VET courses have lower student-teacher ratios and hence higher recurrent costs (also because of the need for more expensive materials). In addition, reducing the proportion of time students spend in vocational programs has the potential to lead to overall improvements in learning outcomes. For example, Poland was able to improve its performance on the PISA international assessment after carrying out an extensive reform of secondary education, with one of the main changes being delayed entry into vocational specialization; see Jakubowski et al. (2010) and Murthi and Sondergaard (2010). 4.38. There is considerable opportunity in reducing energy costs through investments in more efficient school heating systems. As noted above, there has been a very large rise in the amount of money spent on utilities. At the moment, there is no incentive for schools to economize on these expenditures since the relevant line ministry automatically pays the bills submitted to it by the schools. Moreover, a study of energy use and systems in five primary and secondary schools found potential savings of between 55 and 80 percent in the amount of energy used.74 Clearly, some investments are required to garner these savings, but this would be expected to take place within a few years (and possibly less if these very high levels of savings are achieved). 4.39. While the Ministry of Education and Science was observed as having good capital expenditures planning, there might still be room for further improvements in this area. A World Bank (2008b) report on public envestment in the Western Balkans suggests that the project preparation and the application of procurement techniques should be improved in order to avoid unnecessary costs 73 Ministry of Education and Science. 74 See Project Appraisal Document for the Montenegro Energy Efficiency Project, Report No. 44745-ME. 78 for public investments in Montenegro. If the costs encountered are unnecessarily high or are increasing during the project, they might further put pressure on the budget envelope for quality enhancing measures (because these are discretionary expenditures most easy to cut). 4.40. There may be some scope of modest savings in reducing the number of separate agencies under the Ministry of Education. At present, there are five agencies which support the Ministry of Education in implementing Government policy.75 These agencies perform important functions within the system. However, these agencies are required to have their own administrative staff (such as accountants, drivers, etc.) and there may be some scope for reducing costs if these centers were combined or would closer connect their administration. This need not reduce the professionalism or independence of these agencies’ work. F. CONCLUDING REMARKS 4.41. Montenegro has spent as much on public education as other countries in the region, but it has not yet resulted in comparable outcomes. As a result, the key challenge consists of improving the quality of education while curtailing fiscal expenditures in this sector. Government has paid particular attention to education-sector reforms, which seeks to (i) improve education outcomes (which lag behind those of comparator countries); (ii) cap per-student expenditures (which are roughly at regional averages); and (iii) rebalance the composition of spending on public education (where an exceptionally high percentage of expenditures are used for teachers’ salaries). 4.42. Major reforms in the higher education sector still lie ahead. As the chapter has indicated, financing of public higher education is rather static and has no performance-related component. Further, with 17 faculties, the University of Montenegro is highly fragmented and organizationally inefficient. Despite the change of the legal situation, it seems that faculties still act as quasi-autonomous entities, and there is little steering through financing not only in the Government –university relationship but also within the university. Based on the experience with fragmented university structures in the region, the high number of faculties and their current status suggests that there is scope for making significant efficiency savings; see Linden and Arnhold (2008). A strategy for higher education is currently under preparation and will need to take up these challenges. 4.43. The education sector will need to increase significantly its transparency. Currently there is little (easily accessible) information regarding spending and staffing, as well as graduation rates and graduate employment. This information, however, would provide the basis for efficient sector steering. Elements of such a move towards more transparency would be an accessible Higher Education Management and Information System, annual reports of higher education institutions, and possibly also graduate tracer studies. In particular, collection and dissemination of output data, like graduation and graduate employment rates, as well as research outputs and technology transfer, could be effectively utilized in support of continued and potentially expanded support for the sector. 4.44. Finally, the Strategy for Scientific Research Activity of Montenegro 2008 –16 has been an important step towards more focused research activities, a principle which should also be guiding for future higher education reforms. In order to become attractive for students and academics, not only within the region but also within a wider Europe, the sector will need to focus its efforts on areas of national economic, cultural and social priority. 75 Center for Vocational Education, Institute for Education, Institute for Text Books and Teaching Materials, Education Center, Institute for International Cooperation in Science, Education/Culture and Technical Disciplines. 79 4.45. The country faces difficult choices in order to identify resources which can be used directly to improve the quality of education in the country. The most obvious way to find these resources is to reduce the levels of teaching and non-teaching staff. If (non-)teaching staff above the official retirement left the active service, this alone would contribute to annual savings of about €4 million (it needs to be ensured, however, that such a decision would not affect disproportionately teaching ratios and class sizes in, particularly, the smaller, rural schools). But, bringing about these changes will probably require more extensive changes to the organization of schools, appropriate transportation and would ideally go hand- in-hand with a strategy for the development of the North, where some of these teachers are likely to be employed. 4.46. In addition to cost considerations, Montenegro will need to change the structure of its education expenditures and significantly increase the proportion directed towards quality- enhancing measures. While discussion so far has focused on the reform of primary and secondary education; early childhood development as well as higher education and lifelong learning will play an important role for Montenegro’s future competitiveness and further reform efforts will need to focus on these sub-sectors. Annex Table 4.11. Overview of the Montenegro Education System, 2009/2010 No. of institutions No. of pupils Regular 183 Preschool 21 Primary 162 74,539 Satelite schools 281 Secondary 49 General (gymnasium) 12 10,063 Vocational education 37 3-year VET 11 2,516 4-year VET 26 19,179 Education of adults 2 Special Education 26 Special primary education 11 Special Secondary educati 3 Music and ballet schools 12 Source: MONSTAT and MoES 80 Annex Table 4.12. Students in Tertiary Education by Discipline of Study 2006/07 2007/08 2008/09 University of Montenegro 14,737 18,761 21,131 University Mediteran 749 1,474 1,892 University Donja Gorica 0 205 795 Other tertiary institutions 538 1,034 1,489 Business Management Faculty 506 407 1,191 Sorority College 32 37 36 Faculty for State and European Studies … 35 104 Faculty for Management in Transport and Communications … … 122 Algonquin Faculty for Int'l Hotel and Tourist Management … … 36 Source: MONSTAT and MoES 81 5. HEALTH: MOVING TOWARDS AN OUTPUT-BASED SYSTEM76 Relative to its per-capita income, Montenegro spends an above-average amount on (public) health, while achieving health outcomes that tend to be above those of comparator countries. The health system has broadly adequate resources and is gradually improving health outcomes. In this vein, Montenegro has begun to focus its efforts on increasing the efficiency and sustainability of health-care spending, with health financing starting to shift from an input- to a more output-based system at levels of primary, secondary, and tertiary care. To ensure the largest possible impact on efficiency and productivity, it is recommended that these financing reforms be accompanied by greater management and budgetary autonomy for hospitals and the Health Insurance Fund, accompanied by strengthened capacity for effective management of health facilities. A more flexible financing system, no longer based on strict input-based norms, would re-align provider capacities and improve efficiency. A. INTRODUCTION 5.1. With health-sector expenditures of around 6 percent of GDP, post-independence Montenegro has managed to steadily improve its health indicators. Most indicators are at a level close to or above those prevailing in the new EU member states (Figure 5.1). This is a considerable achievement, keeping in mind that health outcomes had diverged from those in EU member states during the early 1990s, reflecting effects of the regional conflict and the sanctions. Compared to the pre-conflict (pre-independence) situation in 1990 (2005), Montenegro reduced the mortality rate of children under five from 11 (17) per 1,000 live births to 9 in 2009. Montenegro’s child mortality rate is now lower than Bulgaria (10), Romania (12), and close to the United States (8)77. However, further progress is still needed to reach the EU average (5). The maternal mortality ratio stood at 13 per 100,000 live births in 2007 and, as such, was marginally better than Romania but worse than Bulgaria, Croatia, or Slovenia. As summarized in Figure 5.2, life expectancy in Montenegro (which had gradually, but steadily declined during the period between the mid-1990s and independence) has slightly recovered and stayed above comparable figures for Bulgaria, Figure 5.1. Health-Care Spending and Outcomes Romania, or Serbia. Death rates tell a similar story, stabilizing during the post-independence years at a level comparable to the EU’s (Figure 5.3). The World Health Organization’s Europe Health for All database shows an even more significant improvement in recent years— with the standardized death rate (SDR) declining from 956 deaths per 100,000 inhabitants in 2005 to 846 in 2009. Other health indicators support this trend, and the general and broad-based improvement has been reflected in the relative increase by Montenegro to the 49th position in the 2010 Human Development 76 Sources: IMF and World Bank databases. Prepared by Kari Hurt (EASHH) and Ethan Yeh (ECSH1). 77 Data from WHO (Europe Health for All) 2011. U.S. data from World Development Indicators. 82 Index, up 16 ranks from the year before (UNDP, 2010). Figure 5.2. Life Expectancy at Birth, in Years Source: World Bank. Figure 5.3. Death Rate Per 1,000 People Source: World Bank. 83 5.2. In recent years, Montenegro has made significant progress in (i) defining its legal and strategic framework for the health system; and (ii) developing the required institutional and management capacities. The legislative framework is anchored in the Law on Health Care (Official Gazette of Montenegro, No. 39/04) and the Law on Health Insurance (Official Gazette of Montenegro, No. 39/04). In addition, a number of other specific acts have been enacted, such as the Patients Rights Act (Official Gazette of Montenegro, No. 40/2010), together with a significant body of secondary legislation, such as the rulebook on the right to primary health care (PHC) and the provision on the scope of rights and standards of health care from obligatory health insurance. Key strategic documents approved by the Government include the Master Plans for the Development of the Healthcare System in Montenegro 2005–10 and 2010–13, the Strategic Development Plan of the Health Insurance Fund up until 2011, and the Strategy for the Optimization of Secondary and Tertiary Health-Care (June 2011). In addition, the institutional capacity for management of the health system is being enhanced through investments in tools such as an integrated health information system. The information system has been completed for all PHC centers and pharmacies and is being scaled up for hospitals and in the clinical center. There is also support to develop the technical capacity of agencies such as the Montenegro Drugs Agency (CALIMS). 5.3. Montenegro has also embarked on specific reforms to improve the efficiency and effectiveness of the health system, starting with primary health care. Most countries in the region have started to reform their PHC systems in order to improve efficiency and deliver better quality care. The specifics of reforms in each country differ in terms of focus, sequence, and details. Montenegro has reformed its primary care system by introducing a model of chosen doctors. Patients can choose to register with one of three types of primary physicians: general practitioner (for adults), gynecologist, pediatrician (for under age 15). These chosen doctors become a “gate keeper” for the health system, and PHC centers (called dom zdravlja, DZ) were reorganized to support the work of chosen doctor teams (i.e., doctor and nurse teams). The DZ financing was changed to incorporate a per capita payment mechanism, called capitation, for physicians. In a per capita PHC payment system, all providers are paid a predetermined fixed rate to provide a defined set of services for each individual enrolled with the provider for a fixed period. This type of financing can facilitate many changes, including greater efficiency in resource use as doctors are paid for keeping people well (and more people well), rather than just for the services when treating sick patients. The PHC financing reform in Montenegro introduced a payment system where provider salaries were based 50% on capitation (based on the number of registered patients per chosen doctor) and 50% on fee for service (for the provision of a package of primary services). These reforms were supported by the development of the integrated information system’s PHC component. 5.4. Health sector reforms remain, however, work in progress, with the key challenge being the full implementation of financing reforms so as to improve efficiency and ensure fiscal sustainability. Related reforms, such as the planned introduction of diagnostic related groups (DRGs) in secondary and tertiary care, follow similar reforms (and good practice) in OECD and other Eastern European countries. However, these reforms—to align incentives and have an impact on the efficiency and sustainability of the overall system—need to be accompanied by greater budgetary and management autonomy. Health service managers, in particular, need to have the flexibility and autonomy to make organizational and hiring decisions within their facilities, based on the new data. Only then can the health system flexibly respond to the health needs of the population. 84 B. OVERVIEW OF HEALTH SPENDING 5.5. Montenegro’s health system remains characterized by a single public -health insurance fund (HIF), largely financed by a social health-insurance tax on labor paid by employees and employers. The coverage of the population by the social health insurance is near universal.78 The HIF negotiates standard contracts with public providers for the provision of broadly defined basic benefit packages. The public health system has 18 PHC centers (DZ), 7 general hospitals, 3 special hospitals, and 1 clinical center. 5.6. The health system is financed largely through social-insurance contributions, the rates of which have varied in recent years. In principle, Montenegro has sought to reduce the tax wedge on labor, but this proved impossible in the aftermath of the global financial crisis. In response, the Government increased the rate from 10.5 percent in 2009 to 12.3 percent in 2010–11 (Table 5.1).79 The HIF also receives revenue transferred from the state budget, which has increased during recent years, from 5.8 percent in 2006 to 20 percent in 2009. A portion of this transfer covers health insurance for certain population groups (long-term unemployed or refugees), but the transfer has been used also to finance health sector wages and severance payments. Unfortunately, most of the budget transfers to the HIF are not based on a clear formula or on defined principles to fill the gap between HIF expenditures and revenues. Table 5.1. Social-Insurance Contributions, 2005–11 (In percent of gross wage) 2005 2006 2007 2008 2009 2010 2011 Total Contribution 15.0 13.5 13.5 12.0 10.5 12.3 12.3 Employee 7.5 7.5 7.5 6.5 5.0 8.5 8.5 Employer 7.5 6.0 6.0 5.5 5.5 3.8 3.8 Source: Law on Contributions for Compulsory Social Insurance. 5.7. To further address fiscal sustainability, a number of revenue-raising options can be considered that capture the externalities of an unhealthy lifestyle. In particular, the Government might want to consider the introduction or increase of “sin taxes” on the consumption of alcohol, tobacco, and pre-packed food with high sugar, high salt, and/or fat content, i.e., on food that can be expected to cause health problems. As these would need to be treated subsequently, there is an economic rationale to have affected consumers (co-)finance corresponding expenditures. A part of these funds could be used to support health promotion and prevention programs aimed at addressing non- communicable diseases (NCDs) and reduce the demand on health services in subsequent years. 78 The Health Insurance Fund estimates universal coverage. More than 96 percent of respondents to a 2008 household survey by UNDP (which had focused on issues of social inclusion) indicated that they had insurance coverage directly or through the coverage of a family member. 79 A similar contribution rate is effective to farmers. The old age and disability pensioners had to contribute 19 percent of their benefits in 2009, but this was decreased to 1 percent in 2010. The Labor Office pays 5 percent contribution rate on behalf of unemployed who receive unemployment benefits. 85 5.8. Government spends more on health than other countries with similar income levels . In 2008, Montenegro’s public-sector health expenditures were 14.7 percent of total public spending, above average when compared to ECA and EU countries. Similar results are found when relating health-sector spending to per-capita income, see Figures 5.4 and 5.5. In general, health systems with higher levels of public funding tend to do perform better in attaining improved health outcomes and financial protection. So Montenegro’s health resources are broadly adequate to achieve key health-sector outcomes and objectives. 5.9. With the economic crisis, however, it is important that health spending remains adequate and financially sustainable. HIF revenues declined substantially after the economic crisis ( (Table 5.2 and Figure 5.6). During the period 2001–09, HIF revenues increased by a (nominal) factor of 2.3. The annual increase has been around 10 percent on average, with the exception of 2008 when revenues increased by 33 percent. This trend reversed in 2009, largely due to: i) the global financial crisis, and ii) reduced contribution rates. Depending on future growth scenarios and contribution rates, the financial sustainability of the HIF could easily become a key concern, representing yet another contingent liability for the budget. C. HEALTH-CARE EFFICIENCY OR “VALUE FOR MONEY” 5.10. Montenegro has already introduced substantial PHC reforms to improve efficiency. One central reform, beginning in January 2009, allowed the HIF to make payments to PHC facilities, based 50 percent on capitation and 50 percent on a fee-for-service basis. Capitation can have several benefits. Among those, particularly important in the one that provides primary physicians with greater incentive to treat more patients and to keep them healthy (as they would be paid the same per-capita amount for having a sick registered patient, which presumably would take more time to treat). Fee-for-service financing provides an incentive for more productivity and provision of services, which is a shift from the previous system, in which a physician’s income was based exclusively on salary. Over time, this health - financing reform has the potential to greatly improve efficiency and service delivery. 5.11. At this early stage, however, many primary health-care centers remain inefficient. Data from the first year of implementation have shown that DZ productivity continues to vary substantially, with a given physician receiving an average 404 to 784 visits per month. Figure 5.7 shows that there are many facilities with scope to increase their productivity. At the same time, there is no apparent relationship between DZ expenditures and productivity, suggesting that DZs with greater resources/expenditures are not necessarily more productive. Hence, financing reforms should address this situation by better aligning incentives and—to ensure maximum effectiveness—transfer to DZ directors greater autonomy over facility and budget decisions. In addition, individual doctor earnings should be based directly on the new payment model, as currently, only the overall budget for salaries distributed to DZs is based on capitation and fee-for-service. 5.12. The number of non-medical staff can still be reduced. Medical staff levels in Montenegro are lower than the EU-15 average (see Table 5.3), but there are some inefficiencies in staffing, with the number of non-medical staff representing 26.6 percent of total public health sector staff in 2008 (Table 5.4). While this share of non-medical staffing is similar to those in neighboring countries (e.g., the share of non-medical staffing in Serbia was 26 percent in 2007), the share does not appear to be declining over time. The Government is undertaking measures to reduce the number of non-medical staff, but it is also important that some staffing or optimization decisions be made by management of health institutions to identify the appropriate mix of medical and non-medical health workers at the facility level. 5.13. 86 Figure 5.4. Public Spending on Health, 2008 (In percent of total expenditure) Source: WHO (Europe Health for All). Figure 5.5. ECA and EU: Health-Care Spending and Per-Capita Income, 2008 Trendline: h = –9.52 + 2.313 ln(y); R2 = 0.529. Sources: WHO (Europe Health for All) and IMF (April 2011 WEO); and World Bank staff calculations and estimates. 87 Table 5.2. Sources of Health Insurance Fund Revenues, 2008–09 2008 2009 Share of total revenues 2009 Change in 2009/08 (in thousands of euro) (in percent) Total 183,417 162,157 100.0 -11.6 Contributions 144,805 129,768 80.0 -10.4 employees in productive branches of the economy 68,271 58,665 36.2 -14.1 self-employed 3,008 2,532 1.6 -15.8 farmers 362 349 0.2 -3.6 employees in public sector 43,765 35,565 21.9 -18.7 retirees 28,947 32,188 19.8 11.2 unemployed 452 470 0.3 4.0 State budget allocations 38,086 31,809 19.6 -16.5 Other revenues 526 579 0.4 10.2 Source: Health Insurance for You and with You (2009), Report on HIF Work and Performance 2009. Figure 5.6. Health Insurance Fund Revenues, 2002–09 Source: Health Insurance for You and with You (2009), Report on HIF Work and Performance 2009. 88 Figure 5.7. Primary Health-Care Center Productivity and Expenditures Source: MOH PHC Survey 2010; and World Bank staff estimates and calculations. Table 5.3. Physicians per 100,000 Inhabitants, 2000–07 2000 2001 2002 2003 2004 2005 2006 2007 2000-07 Chg Montenegro 186.7 189.4 180.5 192.8 191.4 193.0 197.5 196.9 5.5 Old EU members (EU-15) 308.0 309.9 313.5 321.3 328.2 337.5 340.1 343.5 11.5 New EU members (EU-12) 245.3 249.8 256.9 265.1 260.3 248.7 252.7 250.9 2.3 Slovenia 215.0 217.1 222.5 224.6 229.8 234.5 236.0 238.4 10.9 Serbia … … 259.2 265.2 265.3 262.0 264.2 271.1 5.0* Croatia 238.3 237.8 238.3 243.6 249.9 249.9 253.4 266.0 11.6 Estonia 326.4 317.4 314.1 316.0 321.3 317.7 320.6 326.5 0.0 CIS 372.7 370.2 373.6 373.3 371.3 372.0 377.2 376.8 1.1 * estimated. Source: WHO (Europe Health for All database); 2007 is latest year available. 89 Table 5.4. (Non-)Medical Health Workers 2007 2008 Health workers, total 8,807 8,295 Non-medical 2,212 2,207 Medical 6,595 6,088 Share of non-medical, in percent 25.1 26.6 Source: Health Statistics Yearbook, Institute for Public Health, 2007 –08. Table 5.5. Hospital Indicators 2006 2007 2008 2009 Number of discharges 67,209 71,504 75,033 76,329 Average length of stay, in days 9.64 8.98 8.58 8.59 Bed occupancy rate, in percent 69.36 71.26 72.62 73.74 Source: Institute for Public Health, 2006-09. 5.14. At the hospital level, productivity indicators in hospitals have been increasing over time, but more needs to be done. Table 5.5 shows that the numbers of discharges, the average length of stay (ALOS), and bed occupancy rates have improved slightly over the last 4 years. However, when benchmarked against comparator countries, gaps still remain. For example, average length of stay in 2008 was 8.6 days, higher than comparable levels for Estonia (7.9) or Slovenia (6.9). The bed occupancy rate of 73.7 percent is better than the level of new EU members (71.5 percent), but still below the average for EU-15 members (77.2 percent). 5.15. Variations in the average length of stay and bed occupancy rate among hospitals suggest room for improvement. Figure 5.8 shows ALOS for 7 general hospitals and one clinical center in Montenegro (special hospitals are excluded because their cases and treatments are somewhat different). The figure provides two key messages. First, there is great variation in ALOS across hospitals, ranging in 2009 from 5.6 to 8.9 days. The case mix across these hospitals varies, so they are not directly comparable, but the range in ALOS can be reduced. Second, some hospitals have improved their efficiency over time, while others have remained stagnant. Ideally, efficiency improvements should be taking place in all hospitals. The evidence is similar for bed occupancy rates (Figure 5.9). The variation in bed occupancy rates (63.3 percent to 77.7 percent in 2009) suggests that improvements can still be made by reducing the number of beds or by utilizing the existing resources better. 5.16. To address these inefficiencies and to improve the quality of care, the Government has recently adopted a hospital optimization strategy. As with most health systems, hospitals in Montenegro account for the largest share of health expenditures to providers. In 2009, 48 percent of 90 Montenegro’s expenditures were to hospitals, compared to 33 percent to DZs, 7 percent abroad, and 11 percent to other health-care providers. As such, the hospital strategy represents a substantial restructuring, moving the health system away from the current network of DZ, hospitals, and clinical center to eight vertically integrated health-care centers. These centers would include both primary and secondary care and, over the medium- to long-term horizon, transform into three regional reference centers in Northern, Central, and Southern Montenegro. This ambitious optimization strategy will require a transition from the financing of inputs to a system that pays for outputs and outcomes. Key success factors for optimizing the hospital network for greater efficiency include (i) the full implementation of a DRG payment system for hospitals; and (ii) the introduction of greater management autonomy. D. HEALTH SPENDING AND FINANCIAL PROTECTION 5.17. One of the primary objectives of a public health system is financial protection of households from high out-of-pocket (OOP) payments for health care.80 In addition to improving health outcomes, a health system should also ensure protection from financial risks, such as falling into poverty or inability to afford care. In terms of overall OOP payments, Montenegro compares well to other countries, with private expenditures representing 22.3 percent of total health expenditures. This ratio is lower than the regional average and lower than what would be expected when linking OOP payments to per-capita income (Figures 5.10 and 5.11). 5.18. At the household level, formal OOP payments have remained relatively low, but they have become more regressive. Using the annual Montenegro Household Budget Survey, the average share of household expenditures (per adult equivalent) for health has decreased slightly from 3.5 percent in 2006 to 3.3 percent in 2008.81 This share of household expenditures is low in comparison to other countries; however, the trend across quintiles provides some cause for concern. Figure 5.12 shows that the share for OOP payments has increased substantially among the poorest quintile, from 2.4 percent to 4.6 percent. In fact, the pattern across quintiles appears to have shifted from a progressive one (where the rich pay a higher share OOP) to a regressive one (where the poor pay a higher share OOP). The reasons for this shift in OOP payments are unclear, but there may have been level increases in OOP for drug expenditures or copayments. As a result, the share for health out of total expenditures may have increased more among the poor. 5.19. OOP payments appear to be highest in the Southern region. Since the South is the richest region, this result may partly be driven by households’ willingness to pay more, perhaps for supplements or other self-prescribed drugs. In both the North and the South, overall OOP payment shares increased from 2006 to 2008, while they decreased in Central Montenegro. However, Figure 5.13 shows that OOP payment shares increased for the poor across all the regions. Interestingly, OOP payments increased the most for the poor in the South, with a 3.9 percentage point increase in OOP payments among the poorest quintile. In Northern Montenegro, OOP payments increased among all quintiles. It will be important to continue to monitor OOP payments among the poor and in the North to ensure that poor households are protected from payments they cannot afford. 80 This is to ensure that people do not face medical expenditures that are either impoverishing (driving them below the poverty line) or catastrophic (exceeding some threshold of household resources). 81 The household expenditure methodology here differs very slightly from that used by MONSTAT for two reasons. First, we present here expenditures per adult equivalent. Second, the consumption categories included in total consumption vary slightly due to non-inclusion of some durable consumption goods. The amount of health expenditures reported by households is exactly the same, however. 91 Figure 5.9. Average Length of Stay in Various Hospitals, 2006–09 10.0 Pljevlja 9.5 Nikšić 9.0 Average length of stay, in days 8.5 Bijelo Polje 8.0 Berane 7.5 7.0 Bar Clinical Center 6.5 Kotor 6.0 5.5 Cetinje 5.0 2006 2007 2008 2009 Source: Institute for Public Health, 2006-2009 Figure 5.9. Bed Occupancy in Various Hospitals, 2006–09 Source: Institute for Public Health, 2006–09. 92 Figure 5.10. Private Health Expenditures, 2008 (In percent of total expenditure) Source: WHO (Europe Health for All) Figure 5.11. ECA and EU: Out-of-Pocket Spending and Per-Capita Income, 2008 Trendline: ph = 129.3 – 10.6 ln(y); R2 = 0.618. Sources: WHO (Europe Health for All) and IMF (April 2011 WEO); and World Bank staff calculations and estimates. 93 Figure 5.12. Household Share of OOP Payments by Income Quintile 5.0% 2006 4.5% 2008 4.0% 3.5% 3.0% 2.5% 2.0% 1.5% 1.0% 0.5% 0.0% 1 (Poorest) 2 3 4 5 (Richest) Sources: Montenegro HBS 2006 and 2008; and World Bank estimates. 5.20. Drugs purchased at pharmacies account for the majority of household OOP payments. Medicine-related payments average 41.1 percent, incomparably higher than for any other category of health-care expenditure (see Figure 5.14).82 Dentist services (17.9 percent) were the next highest category for OOP payments, followed by outpatient services (16.3 percent). Drug supplements, which include vitamins, minerals, and herbal drugs, account for a considerable share (7.8 percent) of household OOP payments. 5.21. The share of OOP payments for medicines purchased at pharmacies is highest among the poor. Figure 5.15 shows that the poor spend approximately 50 percent of their health expenditures on drugs, compared to only 31 percent among the rich. The actual amounts spent on drugs, shown in yellow, are similar across quintiles, suggesting that both poor and rich are buying similar amounts of medicines. Medicine costs should therefore be monitored to ensure that OOP payments do not increase further for the poor. 5.22. These health expenditures can have an impoverishing effect on households as they lower their consumption of other necessary goods in order to cover health costs such as drugs. Fortunately, households in Montenegro are relatively well protected from large health payments that lead them to fall below the poverty line. Figure 5.16 shows graphically that, in Montenegro, health expenditures in 2008 had a relatively minor impoverishing effect, primarily because of the low amount of (reported) OOP payments. The figure shows the amount of OOP payments (in vertical red lines) across the household distribution ranked by total expenditures (the curved line represents total household expenditure). The poverty line is the dashed black line, and only a few households fall below the poverty line due to health payments. The poverty rate of 4.9 percent in 2008 increases only to 0.8 percentage points when including households that are impoverished due to health expenditures. In contrast, 82 Unfortunately, the NHS 2008 does not have information on total expenditures, so the OOP payment share of total expenditures cannot be calculated. Thus, the analysis of the impoverishing effect of health payments relies on HBS data. Total health expenditures reported are much higher in the NHS data with a per-adult equivalent average of €63.90. The HBS 2008 per adult equivalent average is €10.30. This substantial difference suggests that HBS health data may be underestimated and that OOP payments should be monitored. 94 countries with high OOP payments such as Armenia show much greater impoverishment due to health, with an increase in the poverty rate of 6 percentage points as Armenian households fall below the poverty line due to OOP payments. Figure 5.13. Household Share of OOP Payments by Region and Quintile 6.0% 2006 5.0% 2008 4.0% 3.0% 2.0% 1.0% 0.0% North Central South 4.0% North 3.0% 2.0% 1.0% 0.0% 1 (Poorest) 2 3 4 5 (Richest) 5.0% Central 4.0% 3.0% 2.0% 1.0% 0.0% 1 (Poorest) 2 3 4 5 (Richest) 8.0% South 6.0% 4.0% 2.0% 0.0% 1 (Poorest) 2 3 4 5 (Richest) Sources: Montenegro HBS 2006 and 2008; and World Bank estimates. 95 Figure 5.14. Breakdown of Household OOP Payments Medicines Dentistry Outpatient (Basic/Specialist) Lab tests Drug supplements Other Hospital Medical aids 0% 10% 20% 30% 40% 50% Percent of Household OOP Payment Source: Montenegro NHS 2008; and World Bank estimates. Figure 5.15. Household Health Expenses, by Month, for Medicines Purchased at Pharmacy 60% 60 Monthly expenses, right-hand scale 50% 50 Share of OOP payments Expenditures in euro 40% 40 30% 30 20% 20 10% 10 0% - 1 (Poorest) 2 3 4 5 (Richest) Source: Montenegro NHS 2008; and World Bank estimates. 96 Figure 5.16. Effect of Out-of-Pocket Payments on Household Consumption Source: Montenegro HBS 2008; and World Bank estimates. 5.23. Similar to other countries in the Western Balkans, informal payments are a critical issue facing Montenegro. Anecdotal evidence suggests that these payments are common. A recent Health System Integrity Assessment conducted by the Ministry of Health, World Health Organization, and UNDP found that 56 percent of hospital patients had given to medical staff cash or in-kind gifts not related to official health fees. The Assessment found that informal payments were sizeable, at average of about €60 (and €112 for surgeons). The majority of informal payments were given as gifts after health services were provided to express gratitude for care; however, a substantial proportion of patients did think that informal payments were expected or related to the quality of care. In particular, of the patients visiting a surgeon, an anesthesiologist, or a lab technician/radiologist, 28 percent, 44 percent, and 50 percent, respectively, thought that informal payments would improve the quality of care. While informal payments are entrenched and part of the Montenegrin tradition, the health system should continue to monitor informal payments to ensure that they do not affect the equity of service delivery and financial protection in the system. The Ministry of Health can start by targeting both providers and physicians in cases where informal payments are more of an issue, such as in hospital delivery, anesthesiology and radiology. E. CONCLUDING REMARKS 5.24. Montenegro has made considerable progress in reforming its health system, putting it in a position to improve overall health outcomes and ensure financial protection. The current health system needs to maintain adequate public funding to achieve these objectives, with a continued focus on improving efficiency, particularly in secondary and tertiary care. Only by improving efficiency and 97 productivity can health costs be stabilized in the long-run. There are a number of actions that can be taken to improve efficiency and fiscal sustainability. (i) Introducing a DRG payment system. To accompany the financing reforms in primary care, the Ministry of Health and the Health Insurance Fund are in the process of introducing a case-based payment system, called diagnostic related groups (DRGs), which will shift the financing of hospitals away from line-item budgeting based on beds and staff to payments based on cases treated. The payment is based on the average cost of treating a patient for a particular medical case. An integrated hospital information system is being developed to support this new payment system, and it will be important to monitor case coding, readmission, and cost data to ensure that there are no unintended consequences such as reduction in quality. A DRG system will have to be introduced over time, however, with a phase focusing on hospital reporting and cost analysis, and perhaps a phase with global budgets per hospital. If successfully implemented, a rough idea of the potential impact of introducing DRG’s for hospital payments is provided b y Hungary, where the introduction of DRGs reduced spending on acute care by 14 percent. (ii) Providing greater autonomy to health facilities while building capacity for effective management. Introducing capitation, DRGs and plans to rationalize the hospital network will have limited effect if DZs and hospitals have limited management autonomy matched with management capacity. Hospitals, in particular, are effectively budgetary units, with budgets based partly on staff positions rather than on services. The legal, regulatory, and decision making environment for hospitals makes it difficult to improve efficiency by adjusting staff positions, optimizing work processes, or using remuneration to incentivize productivity. For example, hospitals are not allowed to keep savings from budgetary funds. . Under the new strategy for optimizing the health network, primary, secondary, and tertiary care facilities would be merged into health centers. Increasing the level of autonomy for these health centers, including greater decision rights over budgets and staffing, would increase incentives for efficiency. This type of autonomy would also entail greater budgetary autonomy for the HIF to act as an active purchaser of health care, contracting for services with hospitals rather than financing inputs. Currently, the budget classification rules make it difficult for the HIF to manage health expenditures in the most efficient manner. These reforms would not take place over night, but in the short- to medium-term greater autonomy could be given to health centers as capacity for management is developed. It will be necessary to change decision rights, budgetary authority, representation of their board of directors, as well as provide management training, and methods for cost containment. (iii) Defining the basic benefit package for secondary and tertiary care. To address fiscal sustainability of the system, the basic benefit package should be more clearly defined, as the current benefit package is rather generous. Work on the basic benefit package is ongoing, and it is critical that criteria be established for inclusion or exclusion from the benefit package in order to be flexible as new interventions and technologies are developed. In addition, costs of particular services and technologies must be considered when revising the benefit package. (iv) Taxing “unhealthy” foods. To address fiscal sustainability challenges, a number of revenue- raising options can be considered, such as the introduction or increase of “sin taxes” on the consumption of alcohol, tobacco, and pre-packed food with high sugar, high salt, and/or fat content, i.e., on food that can be expected to cause health problems (“externalities”), which would need to be treated subsequently. In recent years for example, most U.S. states and several EU countries have introduced taxes on high-sugar carbonated beverages. If designed and implemented effectively, these taxes can help change population behavior. A part of these funds, 98 especially if earmarked, could also support health promotion and prevention programs aimed at reducing future demand on health services. (v) Closely monitoring out-of-pocket payments for health. The analysis in Section D shows that OOP payments may be increasing, particularly for the poor. While most citizens remain well- protected from health shocks, the trends are concerning, especially since OOP payments are now highest among the poorest quintile in all regions. Moreover, medicines are the largest category for OOP payments, suggesting that affordability of essential medicines may increasingly become an issue. The analysis here relies on data before the crisis, so the situation is likely to have worsened. In fact, a rapid survey conducted in Montenegro in 2009 suggests that many households affected by the crisis reduced preventive care visits (World Bank, 2011). The Government can take several actions, beginning with closely monitoring the share of OOP payments for the poor and the impoverishing effect. In addition, the can continue to reform drug procurement and price lists to ensure that generic medicines remain affordable. Finally, the Government should continue to address informal payments by increasing awareness of the issue for both patients and providers, and also changing staff remuneration and incentives. 99 83 6. SOCIAL ASSISTANCE: FACILITATING LABOR-MARKET ACTIVATION Non-contributory programs respond to critical social needs of the most vulnerable members of society. At present, spending on social assistance does not currently pose a fiscal risk —partly because of cash transfers are means tested. The accurate targeting, broad coverage, and effective implementation of these programs are, however, essential to accompany reforms to the contributory elements of Montenegro’s system of social protection. The policy focus would, therefore, be on: (i) strengthening the family material support, which is a well-targeted program with a very low coverage; and (ii) linking cash social assistance to labor market activation to support the “graduation” from social assistance for those who are able to work. A. INTRODUCTION 6.1. Social protection in Montenegro is financed largely through its contributory programs, including pension, while the non-contributory programs—averaging 1.1 percent of GDP during the post-independence years—respond to critical social needs of the most vulnerable members of society. As such, the accurate targeting, broad coverage, and effective implementation of these programs are essential to accompany reforms to the contributory elements of Montenegro’s system of social protection. There are nine non-contributory benefit programs which can be broadly grouped into three categories based on their objectives, viz., (i) last-resort social assistance such as monthly Family Material Support (FMS/MOP) and ad hoc (emergency) cash assistance; (ii) benefits for families with children; and (iii) veterans and disability benefits. The safety net includes also social care (placement in institutions for residential care, accommodation in foster care or guardianship) and employment services (public works, assistance for professional rehabilitation and vocational training). 6.2 Spending on “social assistance” (defined as non-contributory benefits) is a relative small share of the overall spending on social protection in Montenegro and, as such, not a fiscal risk per se. While “social protection”, including contributory benefits (such as pensions, maternity and unemployment benefits), has average more than 13 percent of GDP during the post-independence years (Table 6.1), more than 90 percent of these expenditures have been allocated for contributory benefits (pensions, maternity and unemployment benefits). In 2000, spending on non-contributory benefits (social assistance) was 1.9 percent of GDP. It declined to 1.0–1.3 percent of GDP during 2001–10, following legislative changes that strengthened benefits’ means testing. In addition to social assistance, Montenegro spends at average 0.5 percent of GDP per year on social care and employment services. B. SOCIAL ASSISTANCE PROGRAMS AND SPENDING 6.3. Montenegro is one of the few ECA countries which contained social-assistance spending in the 2000s, and the country kept it below the ECA average throughout this decade. Spending tended to decline in the 2000s, reflecting the income benefits generated from the economic boom. It reached its lowest level of 1.05 percent of GDP in 2008. In 2009 and 2010, reflecting the crisis effects, it increased marginally to 1.21 and 1.12 percent of GDP, respectively, but continued to be lower (when measured as share of GDP) than the average spending in ECA (1.7 percent of GDP); see Figure 6.1. 83 Prepared by Boryana Gotcheva (ECSH3) and Victoria Strokova (ECSHD). 100 Table 6.1. Public Spending on Social Protection, 2000–10 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011* percent of GDP Overall spending on social protection 17.28 14.35 16.26 15.44 14.08 13.87 12.40 11.50 11.98 15.19 14.20 13.81 Social protection total 13.17 11.62 12.63 12.82 11.60 11.07 10.48 10.09 10.55 13.94 13.05 12.92 Contributory benefits 11.28 10.34 11.44 11.46 10.22 9.77 9.33 8.90 9.50 12.73 11.93 11.72 Non-contribiturary benefits 1.89 1.28 1.19 1.36 1.38 1.30 1.15 1.19 1.05 1.21 1.12 1.19 Child allowances 1.05 0.49 0.12 0.17 0.16 0.15 0.15 0.15 0.13 0.11 0.15 0.15 Family material support -FMS/MOP 0.43 0.44 0.77 0.64 0.50 0.50 0.48 0.49 0.42 0.45 0.46 0.48 Other persons care 0.12 0.11 0.07 0.19 0.14 0.14 0.15 0.18 0.15 0.12 0.17 0.23 Veterans and disability protection 0.23 0.17 0.17 0.26 0.50 0.42 0.35 0.31 0.29 0.45 0.31 0.30 Social care and employment services 0.62 0.60 0.62 0.20 0.39 0.66 0.87 0.36 0.58 0.51 0.46 0.33 Residential institutions 0.15 0.14 0.16 0.18 0.15 0.14 0.11 0. 09 0.09 0.06 0.08 0.08 Active labor market measures 0.44 0.43 0.44 n.a. 0.22 0.51 0.76 0.27 0.48 0.45 0.38 0.25 Other social protection expenditures 3.49 2.13 3.01 2.42 2.09 2.14 1.05 1.05 0.85 0.74 0.69 0.56 * Budget. Sources: Ministry of Labor and Social Welfare. Other social protection expenditure include staff and maintenance costs for the Ministry of Labor and Social Welfare, Centers for Social Work, Commissariat for Displaced Persons, Montenegro Red Cross, Employment Agency and Pension and Disability Fund. Figure 6.1. Public Spending on Social Assistance Sources: Europe and Central Asia Social Protection Database, World Bank. An asterisk (*) before a country indicates that data from Eurostat were used (includes spending on Family/Children, Housing and Social exclusion). Note: Montenegro data does not include subsidies for electricity 101 6.4. Social assistance spending is almost equally split between means-tested and rights-based categorical programs. Two types of benefits are targeted to the poor, requiring a means-test, viz., the FMS/MOP and the monthly child allowance. In 2008–09, they amounted to around 0.5 percent of GDP, which is comparable percentage to other ECA countries with similar levels of economic development, such as Serbia, Bosnia and Herzegovina, and Romania (Figure 6.2). However, Montenegro spends less on programs that are categorical and not well targeted. In the Western Balkans, only FYR Macedonia allocates a higher share of its social spending to means-tested benefits. Other countries, like Croatia and Bosnia and Herzegovina, spend up to three-quarters of their social assistance budgets on categorical benefits, which are often regressively targeted to the upper middle-income and rich quintiles. When compared to similar programs in ECA middle-income countries, the means-tested FMS/MOP is relatively large and more adequately financed. Conversely, due to the means testing, the monthly child allowance is much less generously financed. With average annual allocations of about 0.15 percent of GDP, it remains among the smallest programs of this type in ECA. Figure 6.2. Public Spending on Social Assistance by Programs Source: Europe and Central Asia Social Protection Database, World Bank (2011) 6.5. Since 2001/02, the spending on the main social assistance programs was increasing in nominal terms. This trend continued until 2009. In 2010, despite the need for counter-cyclical increase in spending on transfers for the poor, the allocations for FMS/MOP were cut. The same happened with the spending on caregivers’ allowance (allowance for care by other persons), which is the largest disability benefit in the country. Preliminary data on budget execution for 2010 indicate that the spending on the caregivers’ allowance, which was expanding most notably since 2002, contracted abruptly by close to 20 percent compared to the previous year. The spending on FMS/MOP also declined notably, by around 10 percent on a year-to-year basis. Only the budget for the monthly child allowance 102 continued increasing, although quite insignificantly (Figure 6.3)84. Spending on social assistance is expected to resume increasing in nominal terms and—possibly—as a share of GDP, due to updating the benefit levels for the main social assistance programs (such as FMS/MOP, child allowances, individual disability benefit and caregivers’ allowance, funeral grant), which had been “frozen” and not updated for several years. The number of beneficiaries is expected to increase. As the ceilings of benefit levels increase, there will be a corresponding rise in the income thresholds that determine the eligibility for the respective benefits. 6.6. The FMS/MOP is the biggest non-contributory program in terms of number of beneficiaries and budget outlays. The FMS/MOP beneficiaries increased throughout the 2000s (Table 6.2), reaching 13,746 families with 41,816 members in December 2010. Compared to 2000, the average benefit per family almost doubled, while the average amount per family member increased less, reflecting the changed demographic structure of recipient families/households. Their average size grew from 2 members in 2000 to 3.1 in 2006 and thereafter. Box 6.1. Main Non-Contributory Benefits in Montenegro The provision of non-contributory benefits is regulated by the Law on Social and Child Protection (2005) and the Law on Veterans and Disability Protection (applied as of 2004). These benefits are financed by the state budget, designed by the Ministry of Labor and Social Welfare (MLSW), and administered by the Centers for Social Work (CSW). Family Material Support (FMS/MOP). The FMS/MOP is a last-resort social assistance program. It targets poor households/families identified with an income (means) and asset test, and exclusionary “filters” for land and housing ownership. The maximum FMS/MOP amount, which is also the income threshold for eligibility, is determined in nominal terms and varies by household/family size. There is a “ceiling” of five eligible recipients in one household or family. Based on the same means test, the FMS/MOP beneficiaries are also eligible for health insurance paid by the state, monthly child allowance (if they have children), funeral grants and caregivers’ allowance (in case of disability). They receive a “package of benefit and services” which is more generous than the FMS/MOP itself, and addresses multiple social needs and vulnerabilities of the beneficiary family/household. Benefits for families with children. The benefits for families with children include (i) birth grant – universal benefit providing a flat amount for each newborn; (ii) monthly child allowance – a means-tested transfer for the first three children in FMS/MOP families, and categorical (non-means-tested) child allowance for children without parents and children with disability; (iii) compensation for unemployed women and full-time students when they are on maternity leave; (iv) salary compensations for parents who work part-time and take care of a sick child; (v) grants to cover holiday and recreation costs for children living in FMS/MOP and foster families. Along with these non-contributory benefits, Montenegro operates a contributory scheme of paid maternity leave, for which women with social insurance contributions are eligible. Veterans and disability benefits. They include 11 different benefits, among them (i) a personal disability benefit which is categorical and received by persons with disability acquired before reaching 18 years of age; (ii) a caregivers’ allowance (benefit to solicit other persons’ care) which is given to FMS/MOP and personal disability benefit recipients; (iii) orthopedic supplements – separate for equipment and consumables; (iv) benefits for members of families of diseased veterans; (v) health insurance, etc. Certain benefits are provided outside MLSW like the benefit for elderly farmers provided by the Ministry of Agriculture, Forestry and Water Resources. 84 See also Annex 1 for more detailed data on spending. 103 Figure 6.3. Spending on Main Social Programs, 2000–10 Source: Europe and Central Asia Social Protection Database, World Bank (2011) Table 6.2. FMS/MOP Beneficiaries and Benefits (In December of the respective year) Number of Number of family Average per Avg. per family Average families members family (in euro) member (in euro) household size 2000 7,936 16,022 45.67 22.62 2.0 2004 10,733 31,721 64.61 21.86 2.9 2005 11,889 36,340 68.26 22.33 3.0 2006 12,695 39,403 69.14 22.27 3.1 2007 12,741 39,281 76.74 24.89 3.1 2008 12,830 39,383 77.32 25.20 3.1 2009 13,130 40,222 85.93 28.05 3.1 2010 13,746 41,816 86.05 28.29 3.1 Source: Ministry of Labor and Social Welfare 104 6.7. The Montenegro FMS/MOP is adequately financed compared to similar programs in ECA. It absorbs more resources as share of GDP compared to the similar Family Material Support program in Serbia. Also, it is better financed than the minimum income programs in Poland and the Baltic states, and is comparable in terms of spending and share of GDP with comparable programs in Hungary and Slovakia. At the same time, some other EU new member states, like the Czech Republic, Romania and Slovenia, spend even more (up to 0.7 percent of GDP) on last-resort social assistance85. The budget plan for 2011 envisages increase in spending, following the aforementioned increases in eligibility thresholds and benefit levels.86 The number of FMS/MOP claimants is expected to increase, despite the resumed economic growth as, generally, the response to last-resort programs to crises comprises a one-year lag. It is advisable to keep the FMS/MOP’s spending at around 0.5 percent of GDP to accommodate the increased demand. It is equally critical that the program is able to identify accurately the poor so as to be able to channel resources to them (targeting accuracy will be discussed later). Disability benefits 6.8. The overall spending on veterans’ and civilian disability benefits is as large as the last- resort social assistance, and it is increasing. In 2010, it represented around 0.48 percent of GDP (in some years, the figure exceeded 0.6 percent of GDP). Compared to the beginning of the 2000s, spending on disability protection is increasing, and this trend is expected to continue in 2011, mostly due to the continuing expansion of the caregivers’ allowance (the largest disability program). As a share of GDP, the budget allocations rose 0.11 percent in 2001 to 0.18 percent in 2009. This share is expected to increase further to 0.23 percent of GDP, due to an increase in the size of the individual entitlements as of January 1, 2011. In nominal terms, the budget for the caregivers’ allowance has seen a six -fold increase during 2002–09. The eligibility criteria for access to caregivers’ allowance need to be scrutinized carefully. This applies to the disability assessment process as well to ensure that leakages to non-eligible households are precluded. As foreseen by design, disability benefits need to be restricted to people with severe disabilities who cannot take care of themselves alone. The personal disability benefit is much smaller in terms of spending but is increasing notably in nominal terms. Related expenditures need to be monitored closely for strict application of the eligibility rules—especially since the eligibility for personal disability benefits is the pre-condition for applying for the more generous caregivers’ allowance. The incentives for qualifying for disability benefits are usually high when the access to other cash transfers is restricted. The—relatively more generous—disability benefits are often regarded as an alternative to last-resort social assistance, also because of the relatively less cumbersome qualification procedures. The incentives for qualifying for disability benefits could escalate, with a 45-percent increase of individual benefits in 2011, form €75.00 to €108.80. Child and family benefits 6.9. Compared to most middle-income countries in ECA, Montenegro has a smaller child and family protection system, both in terms of spending and number of beneficiaries. While the mix of child and family benefits is comparable, the main child benefit (the monthly child allowance) is much smaller: in recent years, Montenegro allocates around 0.15 percent of GDP for monthly child allowances, i.e., less than most other ECA countries. The monthly child allowance program is smaller because it has eligibility restrictions; it is subject to the means test that is applied to determine eligibility 85 Spending on MOP in Serbia varies between 0.14 and 0.17 percent of GDP. Latvia, Poland and Estonia spend 0.2 percent of GDP, Lithuania – 0.3 percent. On the other hand, Slovakia and Hungary spend 0.5 percent, the Czech Republic and Romania – 0.6 percent, Slovenia – 0.7 percent (ESSPROS database and Ministry of Labor and Social Policy of Serbia). 86 As of the beginning of 2011, the eligibility thresholds and benefit ceilings were increased by €3.00for single member households to up to €5.70 for households of 5 and more members. 105 for FMS/MOP. Only FMS/MOP beneficiary families with children are entitled to child and family benefits. As a result, the child allowance spending is lower compared to other similar benefits, even when compared with the similarly rigorously means-tested child allowance program in Serbia.87 In 2011, the spending on the child allowance is planned to remain unchanged as share of GDP, while the benefit amounts paid to different categories of recipients (poor children, children with disability and orphans) have been increased insignificantly. The number of recipients will also increase, depending on the take up of the FMS/MOP under the higher eligibility thresholds. This will be consistent with the existing recipients’ growth trend: the number of families receiving monthly child allowance increased by over 30 percent in 2009 compared to 2004 (Table 6.3). Table 6.3. Monthly Child Allowance (In December of the respective year) Avg. amount per child Avg amount per family Number of families Number of children (in euro monthly) (in euro monthly) 2004 7,436 14,241 20 38.3 2005 8,600 16,505 16.4 31.5 2006 9,600 18,741 15.8 30.9 2007 9,475 18,524 17.2 33.7 2008 9,430 18,370 17.3 33.6 2009 9,734 18,949 19.1 37.1 2010 9,922 19,131 19.6 36.8 Source: Ministry of Labor and Social Welfare. 6.10. Following eligibility restrictions introduced in 2002, the spending on the monthly child allowance measured as share of GDP decreased almost 10 times. Between 1994 and 2001 the child allowance was universal. In 2002, means testing was introduced using the same income thresholds as for the FMS/MOP, and only children living in families receiving FMS/MOP remained eligible, with a ‘ceiling’ of maximum three eligible children in one family. The child allowance remained categorical for children without parents and children with disability who continued receiving it without means testing. Following these program design changes, the spending on child allowances dropped down to 0.12 percent of GDP in 2008-2009, compared to 1.05 percent of GDP before the reform. C. COVERAGE, TARGETING ACCURACY AND GENEROSITY OF SOCIAL ASSISTANCE Coverage of social assistance 6.11. Social assistance in Montenegro is well targeted and relatively generous compared to other countries in the region, but it reaches a small share of the deserving poor and vulnerable. The coverage of all social protection (contributory and non-contributory benefits) is quite high, due to high coverage with contributory pensions. During 2006–09, between 92 and 94.4 percent of the poorest 87 The monthly child allowance program in Serbia amounts to up to 0.4 percent of GDP. The main difference from Montenegro is that in Serbia the means tests for the last resort social assistance and the child allowance are independent from one another, and the take up of the child allowance is higher. Other reason for the higher size of the spending in Serbia is that the child allowance is extended to up to four children in a family against three in Montenegro, and that the benefit size increases with birth order. 106 population quintile was receiving at least one type of contributory or non-contributory social benefit (Annex 2). However, this broad picture masks big differences in coverage with contributory and non- contributory benefits. While the coverage with contributory benefits is quite high and improving, the coverage with social assistance is much lower and declining (Figure 6.4). Figure 6.4. Coverage of Social Protection, Overall and by Types of Benefits Source: Europe and Central Asia Social Protection Database, World Bank 6.12. The overall coverage with social assistance went down to 20 percent of the poorest quintile in 2009, and is one of the lowest in the ECA countries (along with Bosnia and Herzegovina); see Figure 6.5. An even smaller share of the poorest quintile (around 16 percent) receives FMS/MOP and related benefits and health insurance. The combined coverage with disability benefits, child allowances, and scholarships is far lower. In 2009, only 4.5 percent of the bottom quintile received these benefits. The reasons for the low coverage of the FMS/MOP are manifold. Some are related to the design of the program, which has strict rules for access (eligibility criteria) that include income test and additional “filters” related to availability of assets and employment status. The presence of any “filter” overrules eligibility that is based on low income and excludes applicants who would be eligible if only the income test were applied. Other reasons relate to the rigorous enforcement of the eligibility rules, which involve verification of the presented documents by the social workers, annual recertification (re-application), home visits with discretionary power, oversight and control mechanisms. The low coverage combined with the relatively high spending on the FMS/MOP renders good program generosity. Targeting accuracy 6.13. The targeting accuracy of social assistance is one of the highest in ECA. The distribution of social benefits is progressive: the lowest consumption quintile captures between 63.5 and 73 percent of all social assistance and between 58 and 84.4 percent of the FMS/MOP and related benefit in 2006 –09. 107 The targeting accuracy of all contributory and non-contributory programs is increasing except for child and disability benefits (Figure 6.6). Figure 6.5. Coverage of the Poorest Quintile by Overall Social Assistance in ECA Sources: Europe and Central Asia Social Protection Database; and World Bank staff estimates.’ Notes: Performance indicators for Europe and Central Asian countries are calculated using a standardized methodology so as to facilitate comparison across countries. This methodology ranks individuals into quintiles based on harmonized consumption aggregates (Alam et al, 2005) and pre-transfer consumption per capita. A * before a country indicates that the standardized methodology has not been applied in this case. Figure 6.6. Targeting Accuracy of Social Protection, Overall and by Types of Programs Source: Europe and Central Asia Social Protection Database; and World Bank staff estimates. 108 6.14. The targeting accuracy of all social assistance compares well with other ECA middle- income countries (Figure 6.7). The targeting accuracy of 84.4 percent for the FMS/MOP is exceptionally high (the highest in ECA in 2009). An increasing share of the social-assistance budget is going to the rural areas. This is consistent with the poverty analysis, which finds rural residents are at a higher risk of being poor and vulnerable. Between 2006 and 2009, the share of social assistance for urban areas declined while the share of rural areas increased. FMS/MOP targeting follows similar pattern, and as a result around two thirds of its allocation goes to beneficiaries living in the rural (and poorer) parts of Montenegro. Figure 6.7. Targeting Accuracy of Overall Social Assistance in ECA Source: Europe and Central Asia Social Protection Database; and World Bank staff estimates. Notes: Performance indicators for Europe and Central Asian countries are calculated using a standardized methodology so as to facilitate comparison across countries. This methodology ranks individuals into quintiles based on harmonized consumption aggregates (Alam et al, 2005) and pre-transfer consumption per capita. An asterisk (*) before a country indicates that the standardized methodology has not been applied in this case. Generosity 6.15. In terms of generosity, social protection in Montenegro delivers benefits which are higher measured as a share of post-transfer consumption compared to social protection in other ECA countries. This is particularly valid for the generosity of the benefits that are social insurance contribution-based (Figure 6.8), but social assistance is also relatively more generous. During 2006–09, social assistance accounted for 35-40 percent of the post-transfer consumption of the beneficiaries from the poorest quintile. In 2009, the generosity of the FMS/MOP was 30 percent of post-transfer consumption of the poorest quintile which is high according to regional standards and among the highest in the Western Balkans region. The generosity of the child allowance is even higher —close to 55 percent of the post-transfer consumption of the poorest population quintile in 2009. 109 Figure 6.8. Generosity of Social Protection, Overall and by Types of Programs Source: Europe and Central Asia Social Protection Database, World Bank 6.16. FMS/MOP not only ranks among the most generous last-resort programs in programs in ECA; it is also a qualifying condition for other benefits (Figure 6.9). Due to the FMS/MOP and the child allowance specific design characteristics where eligibility for FMS/MOP is also a qualifying condition for receiving the child allowance, if the family has child(ren), caregivers’ allowance in case of disability, funeral grants, free health insurance and other benefits and services, the degree of overlap of benefits is quite high, and those who are already ‘captured’ by the safety net system receive a ‘package’ of transfers and services that enhances the overall generosity. At the same time, the “package” covers only a narrow group of all who are eligible. The generous package of benefits is arguably creating disincentives for earning income because all earned income is taxed away (the actual amount received is the difference between the maximum benefit for which those with no income are eligible and the actual income of the family). The generous benefits also arguably create disincentives for acquiring certain assets that act as exclusionary “filters”. 110 Figure 6.9. Generosity of Last-Resort Social-Assistance Programs in ECA Source: Europe and Central Asia Social Protection Database, World Bank Notes: Performance indicators for Europe and Central Asian countries are calculated using a standardized methodology so as to facilitate comparison across countries. This methodology ranks individuals into quintiles based on harmonized consumption aggregates (Alam et al, 2005) and pre- transfer consumption per capita. A * before a country indicates that the standardized methodology has not been applied in this case. D. COST CONTAINMENT AND COST OPTIMIZATION 6.17. Spending on social assistance does not currently pose a fiscal risk—partly because of cash transfers are means tested. Since the overall spending envelope for social assistance and services is lower than in other ECA countries, and the needs might be increasing (given the lagging crisis response and the new risks for the poor associated with the current food and fuel price increases), it is important to adequately finance the existing programs and to allow them to expand if demand goes up —not least to be able to implement critically important reforms to the contributory programs. The options for reallocation of additional resources to the FMS/MOP within the current social assistance spending envelope are quite limited given its small overall size (and that of other programs). Keeping the spending on FMS/MOP at around 0.5 percent of GDP is desirable. In terms of possible scaling up, it would be desirable to increase the coverage and spending on child protection, especially on the monthly child allowance. Given that children in families with more than two children are at a higher poverty risk, it would be reasonable to waive the limit of three children in one recipient family and to extend the child allowance to all children in poor families, especially if they attend school. Some potential for savings and opportunities for cost containment could be found within the category of disability benefits. Their costs could be contained by tightening the procedures for disability certification and —possibly—by the introduction of means-testing for certain disability benefits that go to recipients who have low disability rates and for whom social inclusion through participation in employment would be more beneficial. 6.18. The family material support (FMS/MOP) is a well-targeted program, but with too low coverage. It could be scaled up to capture bigger share of poor that are not covered by this program— e.g., by relaxing the income thresholds for access to the FMS/MOP (as it has already happened as of January 2011) and indexing the benefits annually with the CPI. In addition, the limitation of having up to 111 five beneficiaries in one family could be waived and thus the program will cover more people living in poor large families. Changes in the eligibility rules, such as the removal of the exclusionary “filters” and their replacement with a scoring formula that takes account of asset ownership without automatic exclusion, could significantly improve coverage88. Finally, proactive outreach to families who do not apply because of lack of knowledge or information, and guidance through the application process can also help to enhance the take up and coverage of the FMS/MOP. To make expansion of coverage possible without significant increase in the budget envelope for this program, it would be necessary to scrutinize rigorously and assess realistically the incomes and assets of applicants and to avoid further increase in program generosity for the existing recipients. It would be also necessary to apply a “work test”/workfare requirement to the recipients of FMS/MOP. By determining who would be able to work, such a requirement would exclude those who work informally. 6.19. Priority spending on FMS/MOP makes sense—if the program’s feature of accurate targeting is preserved. Given that FMS/MOP beneficiaries are also eligible for a number of additional benefits, it is extremely important to correctly identify them and reduce leakage to middle- and high- income recipients. In an economy of high informality, the income test is usually not sufficient to correctly determine eligibility for last-resort social assistance. For this reason, some ECA countries supplement the income test with alternative indicators of living standards to allow capturing multi- dimensional aspects of poverty and to limit inaccuracies arising from high degrees of informality. Targeting is also a function of good implementation which requires monitoring of program outcomes and compliance with program criteria through electronic beneficiary registries which are also linked with the registries of other Government agencies (tax authorities, pension fund, employment services, etc.). 6.20. Government can increase the impact of cash social assistance by linking it to the provision of social care and employment/activation services. Social care services help overcome multiple vulnerabilities faced by the families that receive social assistance. Labor market activation supports “graduation” from social assistance for those who are able to work. Activation requires an authorizing environment (i.e., laws that allow flexible transition from social assistance to work but also preserve the right to resorting to social assistance if the job proves to be temporary). It also requires tax and benefit structure that preserves the incentives of taking a job instead of staying on social assistance for long time. Activation also requires joint efforts of the CSWs and the public employment services to profile the beneficiaries who could be subject to activation and design suitable for these profiles training and employment opportunities. Montenegro’s public employment services are well positioned to provide activation services to social assistance beneficiaries and facilitate their return to employment. The budget of active labor market services is increasing: from 0.19 percent of GDP in 2008 to 0.31percent of GDP in 2009 and 0.44 percent of GDP in 2010, and it is possible within this budget to design specific interventions for hard-to-serve clients, as the FMS/MOP beneficiaries. 88 Simulation of relaxation of the land ownership filter or removal of certain asset filters in the last-resort social assistance program of Serbia, which is very similar to the FMS/MOP in Montenegro, indicate that coverage could be increased by up to 50 percent, if take-up is full. 112 Annex Table 6.10. Spending on Main Social-Assistance Benefits, 2000–10 Caregivers’ allowance FMS/MOP Child allowance Expenditure in millions of euro 2000 1.27 4.40 10.76 2001 1.38 5.51 6.12 2002 0.91 10.04 1.55 2003 2.60 8.97 2.34 2004 2.28 8.28 2.64 2005 2.56 8.86 2.70 2006 3.16 10.30 3.31 2007 4.70 13.25 3.95 2008 4.60 13.00 4.05 2009 6.33 14.45 4.30 2010 5.15* 14.00* 4.42* Source: Ministry of Labor and Social Welfare * Preliminary data Annex Table 6.11. Targeting Accuracy, Coverage, and Generosity of Social Protection (In percent within poorest population quintile) 2006 2007 2008 2009 Targeting accuracy Total social protection programs 46.7 45.5 45.5 53.7 Social insurance programs 47.4 46 48.1 56.1 Social assistance programs 65.2 64.8 63.5 72.9 FMS/MOP and related programs 73.9 57.9 60.2 84.4 Child allowance, caregivers’ allowance and scholarships 44.8 77.3 70.2 62.3 Coverage Total social protection programs 90.7 92.9 93.8 94.4 Social insurance programs 83 85.5 86.9 88 Social assistance programs 23.8 21.1 22.4 20 FMS/MOP and related programs 18.8 16.9 15.7 15.9 Child allowance, caregivers’ allowance and scholarships 8.4 6.9 10.4 4.5 Generosity Total social protection programs 76.8 78.3 84.4 105.7 Social insurance programs 75.1 76.8 83.2 105.2 Social assistance programs 34.7 35.7 39.6 38.4 FMS/MOP and related programs 32.9 30 29.4 30 Child allowance, caregivers’ allowance and scholarships 24.3 39.2 42.9 54.7 Source: HBS 2006, 2007, 2008 and 2009 data processed with SP ADePT by Kiso and Sundaram 113 REFERENCES Aghion, Philippe, Alberto Alesina, and Francesco Trebbi, 2008, “Democracy, Technology, and Growth,” in: Elhanan Helpman (ed.), 2008, Institutions and Economic Performance (Cambridge, MA: Harvard University Press). Barroso, José Manuel Durão Barroso, 2011, “Statement by President Barroso Following His Meeting with Mr Igor Lukšić, Prime Minister of Montenegro,” Podgorica, April 8, 2011. Bénassy-Quéré, Agnès, Benoît Cœuré, Pierre Jacquet, and Jean Pisani -Ferry, 2009, “The Crisis: Policy Lessons and Policy Challenges,” Bruegel Working Paper 2009/06. Central Bank of Montenegro, 2008, Izvještaj Glavnog ekonomiste: 2007.godina (Chief Economist’s Annual Report for 2007), (Podgorica: Central Bank of Montenegro). Chawla, Mukesh, Gordon Betcherman, and Arup Banerji, 2007, From Red to Gray—The “Third Transition” of Aging Populations in Eastern Europe and the Former Soviet Union (Washington: World Bank). Employment Agency of Montenegro, 2010, Employers Survey 2009/2010—Results of the Research. (Podgorica: Employment Agency of Montenegro). European Commission, 2010, “Communication from the Commission to the European Parliament and the Council—Commission Opinion on Montenegro’s Application for Membership of the European Union,” Report No. COM(2010) 670 (Brussels: European Commission). Eurydice, 2008, Levels of Autonomy and Responsibilties of Teachers in Europe (Brussels: Eurydice). Fabris, Nikola, Danijela Vukajlović-Grba, Tanja Radunović, and Jelena Janković, 2004, “Economic Policy in Dollarized Economies with a Special Review of Montenegro,” Central Bank of Montenegro Working Paper No. 1. Hanke, Steve, and Alex Kwok, 2009, “On the Measurement of Zimbabwe’s Hyperinflation,” Cato Journal, Vol. 29, No. 2, pp. 353–64. International Monetary Fund, 2011, “Montenegro: 2011 Article IV Consultation—Staff Report; Public Information Notice on the Executive Board Discussion; and Statement by the Executive Director for Montenegro,” Country Report No. 11/100 (Washington: International Monetary Fund). International Monetary Fund, 2010, “Montenegro: 2010 Article IV Consultation—Staff Report; Public Information Notice on the Executive Board Discussion; and Statement by the Executive Director for Montenegro,” Country Report No. 10/155 (Washington: International Monetary Fund). International Monetary Fund, 2008, “Republic of Montenegro: Article IV Consultation—Staff Report; Staff Statement; Public Information Notices on the Executive Board Discussion; and Statement by the Executive Director for the Republic of Montenegro,” Country Report No. 08/48 (Washington: International Monetary Fund). Ispitni centar, 2009, Nacionalno testiranje učenika III i IV razreda devetogodišnje ssnovne škole maj 2008 (National Testing of the Third and Fourth Grade Primary School May 2008) (Podgorica: Ispitni centar). 114 Jakubowski, Maceij, Harry Anthony Patrinos, Emilio Ernesto Porta, and Jerzy Wisniewski, 2010, “The Impact of the 1999 Education Reform in Poland,” Policy Research Working Paper No. 5263 (Washington: World Bank). Ketelaar, A., N. Manning, and E. Turkisch, 2007, “Performance-Based Arrangements for Senior Civil Servants: OECD and Other Country Experiences,” OECD Working Papers on Public Governance, No. 5. Linden, Toby, and Nina Arnhold with Kirill Vasiliev, 2008, From Fragmentation to Cooperation: Tertiary Education, Research and Development in South Eastern Europe (Washington: World Bank). Lukšić, Igor, 2010, “Keynote address of Montenegro’s Prime Minister -Designate to Parliament,” (December 28, 2010); (http://www.predsjednik.gov.me/en/news/102241/KEYNOTE-SPEECH- OF-MONTENEGRIN-PRIME-MINISTER-DESIGNATE.html.) Ministry of Finance (2011), Smjernice makroekonomske i fiskalne politike 2012. – 2015. godine (Economic and Fiscal Program 2012–15), (Podgorica: Government of Montenegro). Murthi, Mamta, and Lars Sondergaard, 2010. “Skills, Not Just Diplomas: The Path for Education Reforms in ECA,” ECA Knowledge Brief (Washington: World Bank). OECD, 2005, Teachers Matter: Attracting, Developing and Retaining Effective Teachers (Paris: OECD). OECD, 2010a, Education at a Glance 2010 (Paris: OECD). Olters, Jan-Peter, 1999, “The Unfriendly Competitiveness Whip,” World Economic Affairs, Vol. 2, No. 3, pp. 78–79. Perry, James L., 2006, “Motivating Employees in a New Governance Era: The Performance Paradigm Revisited,” Public Administration Review, Vol. 66, No. 4 (August 2006), pp. 505–14. Persson, Torsten, and Guido Tabellini, 2008, “The Growth Effect of Democracy: Is it Heterogeneous and How Can It Be Measured,” in: Elhanan Helpman (ed.), 2008, Institutions and Economic Performance (Cambridge, MA: Harvard University Press). Roma Education Fund, 2009, Advancing Education of Roma: 2009 Country Assessment and the Roma Education Fund’s Strategic Directions (Budapest: Roma Education Fund). Rojec, Matija, Verica Hadži Vasileva-Markovska, Nikola Durović, and Milijana Komar, 2010, Competitiveness of Montenegrin Economy (Podgorica: Ministry of Finance). Vukajlovic-Grba, Danijela, 2011, “Risks From Accumulating Payment Arrears,” Public Expenditure and Institutional Review, Complementary Report (Washington: The World Bank), forthcoming. World Bank, 2011, The Jobs Crisis: Household and Government Responses to the Great Recession in Eastern Europe and Central Asia (Washington: The World Bank). 115 World Bank, 2010, Investing in Young Children: An Early Childhood Development Guide for Policy Dialogue and Project Preparation (Washington: The World Bank). World Bank, 2008a, Montenegro—Beyond the Peak: Growth Policies and Fiscal Constraints, Report No. 46660–ME (Washington: The World Bank). World Bank, 2008b, Balkans Investment Management Study—Country Report for Montenegro (Washington: World Bank). World Bank, 2007, Country Partnership Strategy for the Republic of Montenegro for the Period FY07 – FY10, Report No. 39800–ME (Washington: The World Bank). World Bank, 2006, Republic of Montenegro: Public Expenditure and Institutional Review (Washington: The World Bank). World Travel and Tourism Council, 2011, Montenegro: Travel & Tourism Economic Impact 2011 (London: WTTC); http://www.wttc.org/bin/file/original_file/montenegro_report_2011-pdf.pdf. 116 IBRD 34825R 19°E BOSNIA AND M O N TE N E GRO HERZEGOVINA SELECTED CITIES AND TOWNS NATIONAL CAPITAL RIVERS a in MAIN ROADS Dr To RAILROADS Priboj Cehotin OPSTINA (MUNICIPALITY) BOUNDARIES To a To Priboj Foca Gradac INTERNATIONAL BOUNDARIES Pljevlja Ljubisnja 20°E (2238 m) Du Bioc rm v Zabljak (2396 m) ito r ´ Durdevica SERBIA Tara San Rudinice dzak To Goransko Kom v Foca Tomasevo ari Ta Bijelo Polje Bajovo a Sin ra nc Polje jaj 43°N evi ˘ Savnik na Mojkovac To Vucitrn To Mostar Zeta v Rozaje Gvozd v Ivangrad Kapa Kolasin Velimlje Moracka v Petrovi ´ Petrovici ‘ Niksic (2227 m) v Matesevo Vilusi v Andrijevica Morakovo Medurijecje To Komovi To Dubrovnik (2656 m) Dakovica Grahovo Lijeva Rijeka Murino ca Ze M o ra ta Pelev Plav Crkvice Danilovgrad To Dubrovnik v Gusinje KOSOVO Risan Spuz v Perast Bioce Herceg- Novi Tivat Kotor PODGORICA Radovi´ci vn a ije Cetinje Tuzi C Plavnica ALBANIA Sveti Budva Stefan Virpazar Petrovac Lake A dr ia t ic Scutari Stari Bar Sea Bar Shkodër 42°N a un To -B Kukes a 0 5 10 15 20 Kilometers MONTENEGRO Bojan Ulcinj 0 5 10 15 20 Miles Sveti Nikola To This map was produced by the Map Design Unit of The World Bank. Tirane The boundaries, colors, denominations and any other information shown on this map do not imply, on the part of The World Bank Group, any judgment on the legal status of any territory, or any endorsement or acceptance of such boundaries. 19°E JULY 2009