Report No. 66253-BA Bosnia and Herzegovina Challenges and Directions for Reform A Public Expenditure and Institutional Review February 2012 Europe and Central AsiaRegion Document of the World Bank BOSNIA AND HERZEGOVINA GOVERNMENT FISCAL YEAR January 1 – December 31 CURRENCY EQUIVALENTS (Exchange Rate as of March 1, 2012) Currency Unit = KM US$1.00 = KM 1.45 Weights and Measures Metric System ABBREVIATION AND ACRONYMS AAC Annual Allowable Cut HBS Household Budget Surveys ALOS Average Length of Stay HIF Health Insurance Fund BH Bosnia and Herzegovina IEA International Energy Agency BHAS Bosnia and Herzegovina Agency for IFI International Financial Institutions Statistics BPR Business Process Review IMF International Monetary Fund CAD Current Account Deficit IPH Institutes of Public Health CFO Cantonal Forest Office Ktoe Thousand tons oil equivalent CIT Corporate Income Tax MAFWM Ministry of Agriculture, Forestry and Water Management CFMC Cantonal Forest Management Companies MoCA Ministry of Civil Affairs CPI Consumer Price Index MoH Ministry of Health CSW Centers for Social Work Mtoe Million tons oil equivalent DERK State Electricity Regulatory Commission NCD Non-Communicable Diseases DPL Development Policy Loan NOE Non-Observed Economy DRG Diagnostic Related Group OECD Organization for Economic Co-operation and Development ECA Europe and Central Asia OKFÅ  Timber Tax EE Energy Efficiency OOP Out-of-Pocket EMIS Education Management Information PEIR Public Expenditure and Institutional Review System EPHZH Elektroprivreda Hrvatske Zajednice Herceg PHC Primary Health Care Bosne EPBH Elektroprivreda Bosne i Hercegovine REERS Regulatory Commission for Electricity EPRS Elektroprivreda Republike Srpske RMUs Regional Management Units EU European Union RS Republika Srpska FBH Federation of Bosnia and Herzegovina SME Small and Medium Enterprises FCS Forest Stewardship Council STRs Student-Teacher Ratios FDCP Forest Development and Conservation TFC Total Final Energy Consumption Project FDI Foreign Direct Investment TIMSS Trends in International Mathematics and Science Study FERK Regulatory Commission for Electricity TPES Total Primary Energy Supply FFO Federal Forest Office TPPS Thermal Power Project FFI Federal Forest Inspection TSO Transmission System Operator FTE Full-Time Equivalent VAT Value Added Tax FMPs Forest Management Plans WHO World Health Organization FMUs Forest Management Units FP Forest Program GDP Gross Domestic Product Vice President: Philippe Le Houerou, ECAVP Country Director: Jane Armitage, ECCU4 Sector Director: Yvonne Tsikata, ECSPE Sector Manager: Satu Kahkonen, ECSP2 Task Team Leaders: Ron Hood and Damir Cosic, ECSP2 ACKNOWLEDGEMENTS This report was prepared by a core team led by Ron Hood and Damir Cosic and including Boryana Gotcheva, Anita Schwarz, Clelia Rontoyanni, Ethan Yeh, Mohinder Gulati, Sachiko Kataoka, Igor Kheyfets and Ahmad Slaibi. Velma Pijalovic helped with data gathering. Tanja Bijev translated the document into local language. The peer reviewers were Costas Christou and Anton Gladnishki. This report also benefited from discussions and comments from colleagues, in particular: Anabela Abreu, Goran Tinjic, Jasmina Hadzic and Sanja Tanic. Desktop publication was done by Mismake Galatis (ECSPE). TABLE OF CONTENTS EXECUTIVE SUMMARY ........................................................................................................................ I CHAPTER 1: PUBLIC EXPENDITURE AND THE POLITICAL SETTING .................................. 1 CHAPTER 2: THE FISCAL PICTURE AND RECENT ECONOMIC DEVELOPMENTS ............ 5 I. INTRODUCTION – OVERALL FISCAL DEVELOPMENTS ..................................................................... 5 II. COMPOSITION AND TRENDS IN PUBLIC EXPENDITURE .................................................................... 7 III. TAX STRUCTURE AND PUBLIC REVENUES ..................................................................................... 10 IV. RECENT ECONOMIC DEVELOPMENTS AND PROSPECTS .................................................................. 14 V. RECOMMENDATIONS ...................................................................................................................... 15 CHAPTER 3: PENSIONS AND SOCIAL TRANSFERS .................................................................... 17 I. PENSIONS ....................................................................................................................................... 17 II. RECOMMENDATIONS FOR FBH PENSION REFORM ........................................................................ 25 III. RECOMMENDATIONS FOR RS PENSION REFORM ........................................................................... 34 IV. NON-CONTRIBUTORY CASH TRANSFERS ....................................................................................... 35 V. RECOMMENDATIONS FOR NON-CONTRIBUTORY CASH TRANSFERS ............................................. 48 CHAPTER 4: PUBLIC SECTOR WAGES .......................................................................................... 51 I. OVERALL WAGE PATTERNS ............................................................................................................ 51 II. RECOMMENDATIONS FOR WAGE BILL CONTAINMENT ................................................................... 54 CHAPTER 5: HEALTH ......................................................................................................................... 59 I. INTRODUCTION............................................................................................................................... 59 II. OVERVIEW OF HEALTH SPENDING .................................................................................................. 63 III. EFFICIENCY OF HEALTH SPENDING ................................................................................................ 65 IV. HEALTH SPENDING AND FINANCIAL PROTECTION.......................................................................... 68 V. AFFORDABILITY OF PHARMACEUTICALS ....................................................................................... 72 VI. RECOMMENDATIONS ...................................................................................................................... 75 CHAPTER 6: EDUCATION .................................................................................................................. 77 I. INTRODUCTION............................................................................................................................... 77 II. OVERVIEW OF THE EDUCATION SYSTEM – STATE LEVEL ............................................................. 78 III. OVERVIEW OF THE EDUCATION SYSTEM – ENTITY LEVEL............................................................ 83 IV. ANALYSIS OF THE EQUITY AND EFFICIENCY OF EDUCATION SPENDING IN RS AND FBH CANTONS ....................................................................................................................................................... 87 V. RECOMMENDATIONS ...................................................................................................................... 99 ANNEX 6.1: MAIN FINDINGS AND RECOMMENDATION ON HIGHER EDUCATION FINANCING ............... 102 ANNEX 6.2: RECOMMENDATIONS FOR IMPROVING DATA COLLECTION PRACTICES IN EDUCATION .... 103 CHAPTER 7: FORESTRY ................................................................................................................... 107 I. INTRODUCTION............................................................................................................................. 107 II. STRATEGIC, POLICY AND INSTITUTIONAL FRAMEWORK ............................................................. 109 III. FINANCIAL ARRANGEMENTS AND PERFORMANCE ...................................................................... 111 IV. RECOMMENDATIONS .................................................................................................................... 117 CHAPTER 8: ENERGY ....................................................................................................................... 119 I. SECTOR OVERVIEW ...................................................................................................................... 119 II. COAL SECTOR RESTRUCTURING AND INVESTMENT NEED IN THE POWER SECTOR ....................... 122 III. IMPROVING THE EFFICIENCY OF PUBLIC SPENDING BY TAPPING THE ENERGY SAVINGS POTENTIAL ..................................................................................................................................................... 128 IV. RECOMMENDATIONS .................................................................................................................... 136 ANNEX 8.1: PLANNED INVESTMENTS BY THE STATE UTILITIES TILL 2015 .......................................... 137 ANNEX 8.2: ENERGY SAVING POTENTIAL BY ENERGY EFFICIENCY IN COMMERCIAL AND PUBLIC BUILDINGS ................................................................................................................................... 138 ANNEX 8.3: EXAMPLES OF ENERGY EFFICIENCY PROJECTS IMPLEMENTED IN SOUTH EAST EUROPE 139 TABLES Table 1.1: Revenues, Expenditure and Balance by Level of Government ....................................................................3 Table 2.1: General Government Revenues and Expenditure (as percent of GDP) ........................................................6 Table 2.2: Consolidated General Government Expenditures as a share of GDP ...........................................................8 Table 2.3: Government Spending in Southeast Europe in 2010 ....................................................................................9 Table 2.4: Issuance of T-bills and Bonds in 2011 ....................................................................................................... 10 Table 2.5: Tax Rates in Southeast Europe ................................................................................................................... 11 Table 2.6: Labor Informality in BH ............................................................................................................................. 12 Table 2.7: Government Revenues in Southeast Europe in 2010 .................................................................................. 14 Table 3.1: Pension Indicators for FBH ........................................................................................................................ 18 Table 3.2: Pension Parameters in the FBH .................................................................................................................. 22 Table 3.3: Pension Indicators for Republika Srpska .................................................................................................... 28 Table 3.4: Parameters of New and Old RS Pension Laws ........................................................................................... 31 Table 3.5: Expenditure on Non-contributory Cash Transfers in BH, percent of GDP ................................................ 36 Table 3.6: Expenditure on Veteran-related Benefits in FBH 2005-2011, million BAM ............................................. 39 Table 3.7: Expenditure on Veteran-related Benefits in RS 2005-2011, million BAM ................................................ 39 Table 3.8: Expenditure on Non-veteran (Civilian) Benefits in FBH, 2005-2011, million BAM ................................. 40 Table 3.9: Last-resort Social Assistance, Child Benefits, Civilian Disability Benefits and Benefits for Civilian Victims of War – Number of Beneficiaries in FBH, 2011 .......................................................................................... 41 Table 3.10: Expenditure on Non-veteran (Civilian) Benefits in RS, 2005-2011, million BAM. ................................ 41 Table 3.11: Tax Wedge at Various Levels of the Average Wage in BH, FYR Macedonia, and Serbia (%, 2008) ..... 46 Table 4.1: Monthly salary levels prior to austerity measures in KM (without allowances), 2009 .............................. 54 Table 5.1: Investments in Capital Projects in the Health Sector .................................................................................. 61 Table 5.2: Total Health Expenditure as a Percentage of GDP, 2000-2009 .................................................................. 63 Table 5.3: Share of HIF Expenditures by Functional Categories by Entity, 2006-2009 ............................................. 64 Table 5.4: HIF Membership by Enrollment Category, 2009 (in thousands)................................................................ 65 Table 5.5: Public Health Expenditures, 2006-2010 (unadjusted for inflation) ............................................................ 66 Table 5.6: Per Capita Public Health Expenditures by Canton, 2009 (in 2006 KM) .................................................... 66 Table 5.7: Medical and Non-medical Health Workers, 2004 and 2010 ....................................................................... 67 Table 5.8: Average Length of Stay in Acute Care Hospitals, in Days 2006-2009 ...................................................... 67 Table 5.9: Bed Occupancy Rate in Acute Care Hospitals, 2006-2009 ........................................................................ 67 Table 5.10: Health Out-of-Pocket Payments in the FBH............................................................................................. 72 Table 6.1: Enrollment by Level of Education, 2009/10 ............................................................................................... 79 Table 6.2: Education Statistics for Primary and Secondary Schools by Entity, 2009 ................................................. 88 Table 6.3: Education Statistics for Primary and Secondary Schools by FBH Canton, 2009 ....................................... 89 Table 6.4: Average Wages for the Education Sector and Total Economy by Entity, 2006-2009 ................................ 91 Table 6.5: Ratio of Salary After 15 Years of Experience (Minimum Training) to Earnings for Full-Time Full-Year Workers with Tertiary Education Aged 25 To 64 ....................................................................................................... 92 Table 6.6: Average Wages for the Education Sector and Total Economy by Canton, 2009 ....................................... 93 Table 6.7: Salary Coefficients for Primary and Secondary Education Sector Employees, RS, Effective Since 2008 95 Table 6.8: Avg. Wage for Teachers, All Employees in School, and All Employees in Canton by Canton, 2009 ....... 96 Table 6.9: Factors Determining School Budget for Primary Schools .......................................................................... 98 Table 6.10: Standards for Class Sizes and Teaching Loads by Canton, 2011 ............................................................. 99 Table 7.1: Forest and Forest Land Area (ha) – (Dennison 2006 and Tosterud and Nordberg 2011) ......................... 107 Table 7.2: Value Added as percent GDP ................................................................................................................... 108 Table 7.3: Analysis of Unit Costs .............................................................................................................................. 112 Table 7.4: Employment in the Forestry Sector .......................................................................................................... 112 Table 7.5: Staffing in JPS Å ume RS .......................................................................................................................... 113 Table 7.6: Tab OKFÅ  Collection and Federal Expenditure FBH (‗000KM) ............................................................. 114 Table 7.7: Example of Cantonal Forest Office Expenditure (KM) ............................................................................ 115 Table 7.8: Financial Performance CFMCs 2009 - FMAWMF (2010) ...................................................................... 116 Table 7.9: Profit and Loss - JPS Å ume RS ('000KM) ................................................................................................ 117 Table 8.1: Western Balkans Select Energy Indicators ............................................................................................... 121 Table 8.2: Overview of Write-offs by the Government of FBH & Liabilities of the Coal Mines Transferred to EPBH ................................................................................................................................................................................... 123 Table 8.3: Overview of Government Liability Payment Schedule. ........................................................................... 123 Table 8.4: Investment Needs for Coal Mines in FBH by 2030. ................................................................................. 124 Table 8.5: Energy Savings Potential in South East Europe by Sector ....................................................................... 132 Table 8.6: Public Building Stock ............................................................................................................................... 132 Table 8.7: EE Saving Potential in Commercial and Public Buildings ....................................................................... 133 FIGURES Figure 1.1: Administrative and Fiscal Structure in Bosnia and Herzegovina ................................................................ 2 Figure 2.1: Growth and Fiscal Balance before and after the Crisis. ..............................................................................5 Figure 2.2: Structure of Consolidated General Government Expenditure (in percent of GDP) ....................................8 Figure 2.3: Structure of Consolidated General Government‘s Current Spending (in percent of GDP) .........................8 Figure 2.4: Consolidated General Government‘s Revenue Structure (in percent of GDP) ......................................... 10 Figure 2.5: Tax Wedge for Singles at 33 Percent of average Wage in OECD and Western Balkans (2008/09) ......... 12 Figure 2.6: Tax Wedge for Singles at 100 percent of average Wage in OECD and Western Balkans (2008/09) ....... 13 Figure 3.1: Dependency Rates in FBH Compared to High Income Countries ............................................................ 19 Figure 3.2: Dependency Rates in FBH Compared to Transition Countries ................................................................. 19 Figure 3.3: Contributors as Percentage of Working Age Population in FBH Compared to High Income Countries .. 19 Figure 3.4: Contributors as Percentage of Working Age Population in FBH Compared to Transition Countries ...... 19 Figure 3.5: Pension Spending as Percentage of GDP in FBH Compared to High Income Countries ......................... 19 Figure 3.6: Pension Spending as Percentage of GDP in FBH Compared to Transition Countries .............................. 19 Figure 3.7: Percentage of Elderly Population in FBH Compared to other Transition Countries ................................. 20 Figure 3.8: Age Distribution of Old Age Pensioners in FBH ...................................................................................... 20 Figure 3.9: Types of Pensioners in Bosnia, Federation ............................................................................................... 21 Figure 3.10: Retirement Ages in Transition Countries ................................................................................................ 22 Figure 3.11: Dependency Rates in RS are High Compared to both High Income Countries and Transition Countries ..................................................................................................................................................................................... 29 Figure 3.12: Contributor Coverage in RS is Among the Lowest in Europe. ............................................................... 29 Figure 3.13: Pension Spending in RS is High Relative to Other Countries ................................................................. 29 Figure 3.14: A Large Percentage of Individuals below the Retirement Age are Collecting Pensions. ........................ 30 Figure 3.15: A Substantial Percentage of Pensioners are below the Age of 65, Particularly Widows ........................ 30 Figure 3.16: Social Assistance Spending in BH and ECA, % of GDP, 2008-10 ......................................................... 36 Figure 3.17: Dynamics (2002-2010) and Composition (2010) of Social Assistance Spending in BH ........................ 37 Figure 3.18: Dynamics of Veteran and Civilian Social Assistance Spending by Entity, million BAM, 2005-2010 ... 37 Figure 3.19: Trends in Spending on Veteran Benefits and Permanent Social Assistance in BH and Other ECA Countries (2000-2010, % of GDP) .............................................................................................................................. 38 Figure 3.20: Targeting Accuracy of Total Non-Contributory Benefits Received by the Poorest Quintile in BH and ECA Comparator Countries (%) .................................................................................................................................. 42 Figure 3.21: Last-resort Social Assistance – Share of Total Benefit Amount Received by the Poorest and Richest Quintiles in BH and ECA Comparator Countries (%) ................................................................................................. 43 Figure 3.22: Coverage of All Non-Contributory Benefits in BH and ECA Comparator Countries, Poorest and Richest Quintile (%) .................................................................................................................................................... 44 Figure 3.23: Generosity of All Social Assistance and Last-resort Social Assistance in BH, Poorest Quintile (%) ..... 45 Figure 4.1: Annual growth rates (nominal terms) of the wage bill, revenues, and public expenditure ....................... 51 Figure 4.2: Wage bill as % of GDP and total public expenditure ................................................................................ 51 Figure 4.3: Wage Bill as % of GPD............................................................................................................................. 51 Figure 4.4: Wage Bill as % of Total Outlays ............................................................................................................... 51 Figure 4.5: Wage Bill Trends by Entity (in KM million) ............................................................................................ 52 Figure 4.6: Wage Bill Trends by Entity (in KM million) ............................................................................................ 52 Figure 4.7: Public Employees (without health sector and SOEs) as % of Population and .......................................... 53 Figure 4.8: Total Public Employees ............................................................................................................................ 53 Figure 4.9: Public Sector Salaries ................................................................................................................................ 53 Figure 4.10: Average Gross Monthly Salaries (in KM)............................................................................................... 53 Figure 4.11: Functional Composition of Public Employment in RS ........................................................................... 56 Figure 4.12: Composition of State Wage Bill in 2009 ................................................................................................. 56 Figure 5.1:Infant Mortality Rate (per 1,000 live births) .............................................................................................. 59 Figure 5.2: Private Households' Out-of-pocket Health Payments as a Percentage of Total Health Expenditure, 2009 ..................................................................................................................................................................................... 63 Figure 5.3: Changes in Share of Out-of-pocket Health Care Payments, 1997-2010 ................................................... 68 Figure 5.4: Out-of-pocket Payments on Health as a Percentage of Total Household Expenditures, by Quintile and Equity .......................................................................................................................................................................... 69 Figure 5.5: Out-of-pocket Payments on Health as a Percentage of Total Household Expenditures, 2004 and 2007 .. 69 Figure 5.6: Out-of-pocket Payments on Health as a Percentage of Total Household Out-of-pocket Payments in the Previous Month, by Category ...................................................................................................................................... 70 Figure 5.7: Effect of Out-of-pocket Health Care Payments on Household Consumption ........................................... 71 Figure 5.8: Reference Prices of Generic Medicines in FBH and Croatia, 2011/12 ..................................................... 72 Figure 5.9a: Prices of Generic Drugs for Cardiovascular Disease, by Entity .............................................................. 74 Figure 5.10: Differences between the Highest and Lowest Prices of Generic Drugs in FBH, by Canton ................... 75 Figure 6.1: Gross Enrollment Rates, 2009 or latest ..................................................................................................... 80 Figure 6.2: BH‘s Public Spending on Education in Comparison to Other ECA Countries ......................................... 81 Figure 6.3: Public Sector Wages in BH, 2008-2010 .................................................................................................... 82 Figure 6.4: TIMSS 2007: BH 8th Grade Performances in Comparison to Other ECA Countries ............................... 82 Figure 6.5: Enrollments and Student-Teacher Ratios by Level of Education, 2001/02-2009/10 ................................ 84 Figure 6.6: Student-Teacher Ratios by Level of Education – International Comparison, 2008 or Latest ................... 84 Figure 6.7: Share of Secondary Enrollment by Type of Education, Academic Year 2001/02-2009/10 ...................... 85 Figure 6.8: Public Spending on Education by Entity, 2005-2009................................................................................ 86 Figure 6.9: Public Expenditure on Education as % of Total Government Expenditure, 2009 or Latest ...................... 86 Figure 6.10: Per Student Spending on Primary and Secondary Education by Entity, 2005-2009 ............................... 87 Figure 6.11: Per-Student Expenditure and Average Class Size in Primary and Secondary Schools by FBH Canton, 2009 ............................................................................................................................................................................. 90 Figure 6.12: Ratio of the Education Sector Average Wage to Total Economy Average Wage, 2006-2009 ............... 91 Figure 6.13: Ratio of the Education Sector Wage to GDP Per Capita, 2006-2009...................................................... 92 Figure 6.14: Ratio of the Economy-Wide Wage to GDP Per Capita, 2006-2009........................................................ 93 Figure 6.15: Unweighted Average Wage for Primary and Secondary Teachers Relative to Cantonal-Wide Average Wage by Canton, 2006-2009 ....................................................................................................................................... 97 Figure 7.1: Forest Cover ............................................................................................................................................ 107 Figure 7.2: Overview of Institutional Arrangements in BH ...................................................................................... 111 Figure 7.3: Timber Output per Employee .................................................................................................................. 113 Figure 7.4: Profitability of State Forest Companies .................................................................................................. 117 Figure 8.1: Primary Energy Supply ........................................................................................................................... 119 Figure 8.2: Total Final Energy Consumption ............................................................................................................ 120 Figure 8.3: Demand and Supply Balance in FBH ...................................................................................................... 125 Figure 8.4: Demand and Supply Balance in Mostar .................................................................................................. 125 Figure 8.5: Demand and Supply Balance in RS ........................................................................................................ 125 Figure 8.6: Self Financing Ratio of Power Utilities .................................................................................................. 126 Figure 8.7: Debt Service Coverage Ratio .................................................................................................................. 126 Figure 8.8: Domestic Tariffs and Year-End Cash Balance in EPBH......................................................................... 127 Figure 8.9: Domestic Tariff and YECB in EPHZHB ................................................................................................ 127 Figure 8.10: Domestic Tariffs and Year-End Cash Balance in EPRS ....................................................................... 127 Figure 8.11: Institutional Framework in countries in South East Europe .................................................................. 130 Figure 8.12: Electricity Prices in South East Europe in 2009 .................................................................................... 131 EXECUTIVE SUMMARY Profound political challenges pervade almost all aspects of life in Bosnia and Herzegovina (BH) including the management of public revenues and expenditures. This is obvious from the 15-month delay in forming the current government, from the recent suspension of payments on international debt, from the inability of the Fiscal Council to agree on the most basic parameters for allocating public resources and from the fragmented and heterogeneous nature of public institutions. While it is beyond the scope of a World Bank report such as this to make recommendations regarding a country‘s basic political architecture, a frank acknowledgement of the current state of affairs is necessary from the outset if one is to gain a sound understanding of the narrower challenges involved in managing public resources. Resolution of these political problems offers the greatest promise for achieving greater efficiency in the provision of public services, and it is within the authorities‟ grasp. The duplication and parallelism in public sector institutions especially in FBH of Bosnia and Herzegovina (FBH) is not primarily because of the Dayton Peace Accords. The Dayton Peace Accords are actually framed in quite general terms. They do not require the proliferation of multiple institutions at the canton level. The authorities have the power to adopt more centralized and efficient institutional arrangements. Similarly the authorities are free to determine the structure and functioning of central decision-making institutions such as the Fiscal Council. These could operate much more effectively, given political will. In essence the factionalism along ethnic lines that characterizes public debate and the allocation of public resources is a matter of behavior. It is not forced by constitutional constraints. With that as background, this Public Expenditure and Institutional Review (PEIR) presents the view that Bosnia and Herzegovina faces four main challenges in improving its public expenditure and revenue system, the first of which is to reshape public expenditures which at present are excessively concentrated on current spending including wages, pensions and social transfers. These types of expenditure are being made on a scale that displaces more productive spending on such things as infrastructure and therefore compromise future growth. They embody perverse incentives, and those that should form part of the social safety net system, are in fact strikingly ineffective at reaching the poor and vulnerable. The second challenge is to improve the structure of the revenue system. This system is heavily reliant on payroll taxes which provide a strong inducement to both employers and employees to retreat into the informal sector to avoid (or evade) taxes. It is difficult for foreign investors and potential exporters to do this and consequently they bear a disproportionate share of payroll taxes and of taxes in general. This thwarts FDI and undercuts competitiveness and growth. Consequently there is a need to reduce the size of the labor tax wedge, focusing particularly on the changes to social contribution rates. However, it is crucial that this be done in a manner that ensures fiscal sustainability. In particular, reductions in social contributions must be matched by reductions in the associated benefits or other expenditures, or by increases in revenues from other less distorting taxes. The third challenge is to reverse the excessive growth of the public sector. Pre-crisis tax reforms and a boom in growth gave the authorities the room to sharply increase spending. But they went even further and the expansion in spending actually outstripped both revenue and GDP growth. Not only did this represent an inappropriate pro-cyclical fiscal stimulus, it has resulted in BH having one of the largest public sectors in the region – largely to accommodate the growth of relatively unproductive expenditures. Moreover, within individual sectors such as health and education there is a need to undertake structural reforms that will improve the efficiency of spending. The fourth challenge is to complete the overall fiscal architecture in the country by strengthening the Fiscal Council. BH has to dedicate monetary policy to supporting the currency board arrangement. i Consequently fiscal policy is the only instrument left to manage aggregate demand. Moreover BH is also a decentralized country making it more critical to coordinate fiscal policy effectively. The Fiscal Council should therefore be a key institution for the management of fiscal policy. However, as evidenced by the protracted budgetary stalemate at the Fiscal Council since the October 2010 and from the recent suspension of payments on international debt, it is evident that the overall fiscal architecture is still from complete. The report is organized into eight chapters, the first of which briefly outlines the political setting. In keeping with the above, it identifies some of the features of the current system and recent history that affect the management of public resources, but it refrains from making recommendations of a political nature. The second chapter examines recent economic developments and the structure of revenues and expenditures. It outlines how the size of the public sector increased during the pre-2008 boom but failed to lay a solid foundation for economic growth. It traces how recurrent spending swelled and labor taxation increased. It highlights the risks inherent in the current situation where fiscal-financial demands have been allowed to grow against a backdrop of turbulent external conditions and sharply reduced access to international finance. It makes the following recommendations: ï‚· Strengthen the role of the Fiscal Council and improve fiscal management ï‚· Reduce the magnitude of the tax burden, especially on labor. ï‚· Reshape public expenditures which at present are excessively concentrated on current spending including wages, pensions and social transfers. ï‚· Reverse the excessive growth of the public sector by increasing efficiency of spending. The ensuing six chapters examine in more detail how to implement these recommendations in individual sectors. Chapter three deals with pensions and social transfers. The analysis of this chapter concludes that pensions and social transfers, as well as wages in the public sector, are the main culprits in the excessive level of recurrent spending. They embody poor incentive effects with respect to informality and labor force participation, and they are poorly targeted. The pension system is characterized by high benefits, high dependency ratios, low contributor numbers, large numbers of early retirees, inappropriate scaling and qualification criteria for disability benefits, and the prevalence of unfunded ―privileged‖ pensions. There are important differences between the systems in the two entities, however. The Republika Srpska (RS) has recently introduced a new law that provides a gradually increasing retirement age for women and a move to a point system that will generate a flatter accrual rate - changes which are all positive. However earlier versions of the law moved RS much closer to international best practice. The law which actually passed is unlikely to bring the pension system to full fiscal sustainability and will require further changes in the future. The following measures are recommended for RS: ï‚· Introduce strict curbs on early retirement. ï‚· Establish actuarial pension reductions for those who do retire early. ï‚· Eliminate double-dipping between war-related benefits and pension fund benefits. ï‚· Revise the disability benefit formula to provide higher benefits for the fewer people who will qualify as disabled. ï‚· Provide survivor pensions only at retirement age and withdraw them if remarriage occurs. ii In FBH a complex system of ―coefficients‖ introduced to contain costs, as well as numerous ad hoc adjustments and indexing arrangements, has rendered the system opaque and essentially unsalvageable. A complete makeover is needed in FBH shaped around the following recommendations: ï‚· Revamp the benefit structure, eliminating the coefficient as it now exists. ï‚· Introduce strict curbs on early retirement. ï‚· Establish actuarial pension reductions for those who do retire early. ï‚· Eliminate double-dipping between war-related benefits and pension fund benefits. ï‚· Tighten the eligibility conditions for disability and revamp the benefit structure to correspond with the new old age structure. ï‚· Provide survivor pensions only at retirement age and withdraw them if remarriage occurs. ï‚· Separate privileged pensions from the general pension pool, fund them separately and align them more closely with general pensions. ï‚· Limit benefits indexation to prices rather than wages. BH has a history of high and fiscally unsustainable spending on non-contributory cash transfers dominated by programs designed to protect veterans and/or their surviving dependents. The system is expensive and poorly targeted. In 2010, the Governments of FBH and RS started a move towards restructuring social assistance expenditures to address these matters by adopting Framework Laws on Targeting of Cash Benefits to Individuals, setting the stage for a shift from the predominantly rights- based approach to a needs-focused approach in social assistance. However there has been a complete failure to implement these laws and no significant progress has been made towards improving the equity and fiscal sustainability of non-contributory cash transfers. The following measures are recommended: ï‚· Avoid addition of new benefits, especially for all veterans and survivors. ï‚· With priority, in 2012, have FBH Ministry of Labor and Social Policy revise the current Law on Basic Elements of Social Protection in FBH, or replace it with new law(s).  Undertake legal, administrative and fiscal measures to curtail the costs of the veterans‘ benefits by enforcing the legislation which was adopted on paper and scheduled for enactment as of January 2011.  Develop updated tools to improve the targeting of non-contributory benefits for civilians aimed at reaching the poor and most vulnerable.  Develop effective targeting mechanisms for veteran benefits recognizing their dual role (social protection and war service) and aim at eliminating the rich among benefit recipients.  Diversify the support for veterans who would lose cash benefits (due to reforms) by providing employment services, support for small business development, training and re-qualification.  Introduce design changes in the social assistance programs to eliminate potential work disincentives and enhance the incentive compatibility of the social assistance programs to facilitate the transition to employment.  Improve the cost-efficiency, transparency and accountability of benefits administration.  Step up eligibility audit efforts. Chapter four covers public sector wages The overall level of wages and non-wage benefits is high in BH compared to most other countries. Given this and the uncertain outlook for growth and fiscal resources going forward it is clear that measures are needed to reduce compensation costs. However, the approach taken needs to embody more than simple across-the-board cuts that do not address structural issues in the current patterns of compensation. Moreover, the changes need to be seen as fair and to be underpinned by new institutional arrangements in order for changes to have a lasting effect. With that in view the following measures are recommended: iii ï‚· Institute meaningful establishment controls (particularly in the state and FBH) and regularly monitor staffing to control staffing and wage bill expenditure. ï‚· Rationalize the large number of currently vacant posts and cut wage bill appropriations accordingly. ï‚· Systematically monitor staffing numbers to ensure the observance of establishment and hiring controls for all categories of public employees to facilitate wage bill management. ï‚· Regularly monitor employment and pay levels in sectors other than the public administration, as these sectors make up the bulk of public employment. ï‚· Reduce the cost of various salary increments with a combination of the following measures: o Abolish discretionary fees (e.g. for attending committees and working groups) by law, as they represent unnecessary salary top-ups for functions that already fall within the officials‘ duties. o Eliminate universal entitlements to allowances (e.g. meal allowance) by absorbing them into base pay, as was done in RS. o Restrict the eligibility to allowances for housing, family separation, and home visits to expatriated officials (e.g. diplomats posted abroad) and limit it in time (e.g. up to three months) to other officials relocating upon first taking up employment in the public administration). ï‚· Cut other categories of recurrent spending related to staff numbers, such as official travel, rental of buildings, and telecommunications. ï‚· Reduce the number of auxiliary staff by outsourcing these functions or transferring these employees to labor law contracts. ï‚· Increase flexibility in negotiations with public sector trade unions to avoid disorderly wage bill cutbacks in the future. ï‚· Improve coordination in wage bill planning in the Fiscal Council. Chapter five is on the health sector. Total health expenditures in BH have been growing at an unsustainable rate. Social health insurance contributions are high in both entities. Over 50 percent of those who receive health insurance coverage are exempt from paying contributions (unemployed, disabled, and war veterans), and their health insurance is financed via transfers from other extra budgetary funds and from general revenues. Substantial progress has been made in terms of investments in family medicine, changes to health financing, the introduction of quality standards, and more. However, work must continue on these reforms as many (such as the introduction of DRGs) have long implementation periods that will require 10 years or more before their positive effects on efficiency and quality are fully realized. Financial sustainability of the health system is a serious issue for both entities. Despite high and rapidly growing health care expenditures, health outcomes generally lag behind other countries in the region. One underlying cause is the highly fragmented nature of BH health system particularly in FBH where responsibility for health services is delegated to the cantons, so FBH health sector includes FBH Ministry of Health (MoH), the 10 cantonal MoHs, the Federal Solidarity Health Insurance Fund (HIF), the 10 cantonal HIFs, and 11 Institutes of Public Health (IPH). The RS health system is centralized at the entity level. Other efficiency issues include excessive numbers of non-medical staff in FBH, and continued reliance on financing by inputs rather than population needs, or health outcomes. Although private health expenditures have declined substantially in BH, private out-of-pocket payments remain at a high level, above 40 percent of total health spending. Drugs accounted for a substantial iv portion (30 percent) of household payments in the last HBS (2007). The pharmaceutical sector in BH is extremely fragmented. Each of the 13 HIFs has its own drug procurement rules, procedures, and positive lists. Unfortunately, this decentralized procurement results in higher prices and substantial price variation across cantons and entities. Regulation is non-transparent and a survey done for the PEIR revealed price differences of over 50 percent across cantons for half of the generic drugs sampled. The Federal Minister of Health has updated the guide prices for the drugs on the List of Essential Medicines, nevertheless only Una-Sana Canton, Zenica-Doboj Canton and Sarajevo Canton have fully complied with this new regulation. Recommendations for the health sector are as follows: ï‚· Expand the insurance risk pool related to hospital and pharmaceutical care. ï‚· Review exemptions from contributions and consider shifting the collection and control of health contributions to at least the entity level to help improve collection. ï‚· Reduce fragmentation and duplicate functions. In FBH, considerable duplication of roles and functions exist concerning administration of health care in the MoHs and IPHs. ï‚· Leverage family medicine reforms to increase Primary Health Care (PHC) productivity and increase preventive medicine. ï‚· Fully implement hospital and primary health care financing reforms. Efficiency in hospitals remains relatively low, and financing reforms such as case-based payment systems (Diagnostic Related Groups, DRGs) and capitation payments in PHC may provide incentives and management tools to increase efficiency. ï‚· Centralize procurement of drugs in FBH to at least the entity level. Chapter six is on education. In contrast to the situation in other sectors, and despite recent increases, overall spending on education does not appear to be excessive or out of line with levels in other countries. However the allocation of spending within the sector and various practices affecting the efficiency of spending give cause for concern. Students in BH have performed poorly in an international assessment. Even within the current level of public spending, BH‘s students could perform better given a better learning environment. Recent sharp increases in wages for primary and secondary education sector employees, particularly in RS have prohibited non-wage spending and capital investment from growing. Teachers are already slightly (FBH on average) or significantly (RS) overpaid in terms of the ratio of their wages to GDP per capita. Rebalancing of expenditures is also needed between levels of education. More investment is needed in pre-school and tertiary education which have been neglected, as is clearly indicated by the extremely low enrollment rates at those levels. While the primary and lower secondary enrollment rates are reasonably high, the rate for upper secondary is very low compared to neighboring countries. The government needs to investigate the reasons for such a low rate at the upper secondary level and take necessary actions to improve enrollments. The 2006 PEIR recommended introduction of per capita financing of schools to replace the existing input-based financing. School financing based on the number of students (output-based) is expected to help enhance efficiency, equity, and transparency in school financing if implemented rigorously. There are significant differences in the efficiency of education spending between the entities and between the cantons in FBH. Some cantons are more efficient (e.g., higher student-teacher ratios, higher average class sizes, higher teaching loads, slower wage increases) than others. However, in order to improve the efficiency and equity of school financing, more detailed school-level data are essential. For instance, without knowing geographic, socio-economic, and demographic conditions, it is difficult to judge what would be an adequate class size or student-teacher ratio for any particular school or canton. Also, v information on students‘ learning outcomes based on a national student assessment is essential to measure how efficient and effective sector spending is. The following are the recommendations for the education sector: ï‚· Limit further total outlays for wages. ï‚· Develop a more decompressed wage structure which takes into account teachers‘ qualifications, experience, performance and other criteria to better attract and retain highly qualified teachers. ï‚· Rebalance spending between levels to enhance pre-school and tertiary education and ï‚· Address the quality issues in upper secondary that are at the root of low enrollment rates. ï‚· Introduce per capita financing of schools to replace the existing input-based financing system. ï‚· Establish an Education Management Information System that includes: o school characteristics: in addition to the total numbers of students, classes, and teachers, more detailed information must be collected on school facilities and equipment, school- based staff (with details on staff positions, qualifications, pay grades, etc.), and student backgrounds (including ethnicity, language of instruction, demographics, etc.) o school level data on education financing o demographic and community data o national student assessments. The seventh chapter is on forestry. A recent survey (the 2010 State Forest Inventory) indicates that BH forest cover is much larger, and the forests are in much better condition, than previously thought. Consequently the sustainable yield is now reckoned to be about 11 million m3 compared with a current annual cut of only about 6–7 million m3. Hence the forestry sector‘s contribution to the economy could be increased by increasing timber harvesting. However, the government has a major presence in the sector and a number of changes in how it operates are needed to unlock this potential. The benefits of this should not be overstated. Forestry presently contributes about 2.5 percent of GDP, including the wood industry, so even a significant scaling up of harvesting would not add more than 1.5 percent to GDP. However this is not a trivial amount and moreover the problems that exist in the sector are representative of challenges of public expenditure and management more broadly. Only 20 percent of the forests are in private hands - largely in small isolated plots that are not effectively exploited. Ownership of the public forests rests with the two entities and there is no BH national forest policy. There is a very complex set of overlapping institutions that governs exploitation of forest resources and, as in other sectors, these are rather different in the two entities. In FBH responsibilities are transferred to the cantons which delegate them in turn to Cantonal Forest Management Companies. Arrangements in RS are somewhat more centralized, but in both entities there is a lack of overall policy direction, the operational units are over-staffed with people with outdated skills, and the emphasis appears to be on maintaining jobs (including for veterans) rather than exploiting the resource in an efficient manner. A significant issue is that of setting prices of the harvested timber. Earlier attempts at commercial pricing through auctions were frustrated by international price variations, and currently the majority of timber is sold at administratively set prices that do not maximize the value of the resource. Consideration has been given to various schemes that would enhance the role of the private sector. However there are some basic policing issues here. Despite the apparent lack of knowledge about the size and quality of the resource stock, forest monitoring and control functions have been performed reasonably well in BH and the illegal logging and gross over harvesting that have afflicted the forest sectors in many other countries hava not been a problem in BH. vi The following are the recommendations for the forest sector: ï‚· Improve road access and conditions to help increase production, improve harvesting efficiency and increase forestry contribution to GDP. ï‚· Identify and reduce unproductive labor in public forest-sector institutions to improve productivity and profitability. ï‚· Invest significantly in human resources through development of additional skills in the areas of business and organizational development, change and change management, investment appraisal, cost benefit analysis, robust budgeting processes, public awareness and communication. ï‚· Implement institutional changes in FBH: transfer the responsibility for security and protection of the forest resource to the Cantonal Forest Management Companies; have the Cantonal Forest Offices report formally to the Federal Forest Office and operate effectively as daughter organizations; give the Financial Police the sole responsibility for OKFÅ  (timber tax) compliance; increase the staff within the Forestry Department in FBH in line with its responsibilities. In RS review the staffing and roles of both Forestry Department and Forest Agency. ï‚· Have the larger Cantonal Forest Management Companies within FBH and JPS Å ume RS establish a comprehensive Business Process Review (BPR) of all the major processes within their organization to identify opportunities to improve compliance, improve quality and reduce costs. ï‚· Develop a transparent and market-based mechanism for timber sales in a manner that allows independent verification and has a definite timeframe and implementation indicators. ï‚· Set the overall level of the OKFÅ  tax based on an in-depth cost-benefit analysis that reflects the true value of public goods being provided. The system for allocation of the OKFÅ  collected and its subsequent disbursement needs to be made clear and transparent with complete traceability and accountability. ï‚· At this stage, privatization is not judged to be a sustainable option in BH due to: (a) the absence of any sound market price to determine the value of the harvesting rights / timber concession; and (b) the presence of weak institutions and lack of capacity to effectively implement the legal framework. There is also a risk that privatization could lead to over- exploitation resulting in forest degradation, soil erosion and environmental damage. Chapter eight is on the energy sector. There are three main issues in the energy sector. The first concerns fact that there are large coal-mining operations that are currently making significant losses. The second is the publically-owned power producers have been forced to absorb the coal mines into their own operations and that without tariff reform they will not be financially viable. This will preclude their making investments needed to meet future demand including that from potentially profitable regional exports. It will also preclude private participation in energy production. Finally there are efficiency problems throughout the sector: in coal and electricity production because of lack of investment and maintenance; and in consumption, particularly in the electrical heating of buildings. This sector has a significant potential for impact on government finances as well as being critical to overall economic development. Energy demand has been growing strongly, led by residential and transportation consumption, while that of industry has actually declined. Energy end use is highly inefficient. The buildings sector (private, public, and residential) accounts for one third of total final energy consumption. Buildings are generally poorly insulated, creating heat losses of more than 30 percent. In addition, since the gas distribution network is not fully connected, the majority of the vii households heating requirement is fulfilled by the inefficient use of electricity. As a result, about 65 percent of the total electricity consumed in 2009 went to the residential and public buildings sector. Significant investments are required to complete the rehabilitation of power infrastructure. Generation has not recovered from its pre-war level (14.6 TWh in 1990) and its efficiency is still low (about 30 percent) due to lack of maintenance and spare parts, and the slow replacement of obsolete units. Several expansion and greenfield projects will be needed to keep up with demand growth. Under current policies financing this investment will be challenging. BH‘s energy sector is characterized by a high degree of government ownership and large capital expenditure needs, especially for coal and power infrastructure. Coal accounts for 65 percent of electricity generation and energy supply, and 75 percent of coal production is sold to power generation companies. Hence the destinies of these two industries are closely linked. Investment needed in them to meet demand reliably is estimated at over €4 billion by 2020. Up to now, the power sector has been financially viable and creditworthy, and as such has been able to cover its costs and finance its capital expenditures. However, the coal mining companies in FBH, have suffered from severe lack of investments in modern technology, below-cost pricing (which is in effect a subsidy to power production), and high costs because of low productivity. They have racked up huge liabilities in the form of contributions owed to pension, health insurance and other funds, and by end-2008 their total liabilities amounted to over BAM 265 million (135 million euro). To address the long-standing economic difficulties in the coal sector, the Government of FBH decided to merge the coal mines with one of the government-owned electric utilities. This restructuring of the coal sector has, of course, had a significant fiscal impact on the Government of FBH. Following the merger, the Government of FBH took over the obligations for income taxes and pension contributions. These liabilities are being financed directly from the budget and amortized over a seven-year period. But this is not the end of the story. Other significant coal company liabilities to suppliers remain on the books of the inheriting power utility. Moreover, all the power utilities must make substantial investments to improve productivity and create additional capacity if they are to meet rising demand. However, the existing and expected future tariff levels are inadequate to sustain the needed investments, putting at risk the sector‘s financial and commercial viability. Compounding these problems is fragmented regulation of generation and distribution at the state and entity levels. There are three separate electricity regulators. First, the State Electricity Regulatory Commission based in Tuzla is responsible for electricity transmission tariffs. Then there are separate regulators at the entity level. The FBH Regulatory Commission for Electricity based in Mostar is responsible for electricity generation and distribution in FBH. The RS Regulatory Commission for Electricity based in Trebinje is responsible for electricity generation and distribution in RS. The entity regulators are responsible for licensing, proposing network access tariffs and end-user energy prices, and protecting customers. Without addressing the tariff issue which will enable productivity-enhancing investment, the sector will inevitably make further claims on the budget. Pricing adjustments themselves would help to improve energy efficiency. But further measures are needed to enhance awareness, introduce standards and regulations and provide public information. To reduce the potential burden of the power sector on public finances by improving its financial and commercial viability the following measures are recommended: ï‚· Continue the restructuring and modernization of coal mines within the new institutional setup. Efforts should focus on improving efficiency by adopting modern technologies and improving labor productivity. viii ï‚· Continue restructuring of the electricity sector towards EU standards for corporate governance and set tariffs at their full cost-recovery levels, while putting in place targeted social protection mechanisms to protect vulnerable groups. ï‚· Prepare and adopt an investment strategy for each of the power utilities based on a comparative least-cost investment plan which takes into account the power sector demand forecast at the entity, national, and regional levels. Options for private sector involvement should also be considered. To improve the efficiency of public spending in energy services the authorities should implement a set of Energy Efficiency (EE) measures including following: ï‚· Implement the EE program for public buildings, focusing for instance on schools and hospitals. ï‚· Develop a basic legal framework and set up institutional arrangements, including establishing a ―change agency‖ to champion the implementation of EE programs. ï‚· Set up a comprehensive and systematic energy data gathering and reporting system and launch targeted information campaigns. ï‚· Establish a financing support mechanism for EE investments, such as Energy Efficiency Funds. ix Highest Priority Recommendations Short term Medium term Fiscal ï‚· Strengthen the role of the Fiscal Council ï‚· Reduce the magnitude of the tax burden, especially ï‚· Reshape public expenditures which at present are on labor excessively concentrated on current spending ï‚· Ensure that any reductions in labor taxes are including wages, pensions and social transfers. balanced with benefit reductions or other less distorting revenue increases in order to preserve fiscal sustainability ï‚· Reverse the excessive growth of the public sector by increasing efficiency of spending, in particular, on health and education. Pension RS ï‚· Introduce strict curbs on early retirement. ï‚· Establish actuarial pension reductions for those who do retire early. Pension FBH ï‚· Revamp the benefit structure, eliminating the coefficient as it now exists. ï‚· Introduce strict curbs on early retirement. ï‚· Establish actuarial pension reductions for those who do retire early. ï‚· Tighten the eligibility conditions for disability and revamp the benefit structure to correspond with the new old age structure. ï‚· Separate privileged pensions from the general pension pool, fund them separately and align them more closely with general pensions. ï‚· Limit benefits indexation to prices rather than wages. Non-Contributory Cash Transfers ï‚· Avoid addition of new benefits, especially for all ï‚· Develop effective targeting mechanisms for veteran veterans and survivors. benefits recognizing their dual role (social ï‚· With priority, in 2012, have FBH Ministry of Labor protection and war service) and aim at eliminating and Social Policy revise the current Law on Basic the rich among benefit recipients. Elements of Social Protection in FBH, or replace it ï‚· Diversify the support for veterans who would lose with new law(s). cash benefits (due to reforms) by providing ï‚· Undertake legal, administrative and fiscal measures employment services, support for small business to curtail the costs of the veterans‘ benefits by development, training and re-qualification. enforcing the legislation which was adopted on ï‚· Improve the cost-efficiency, transparency and paper and scheduled for enactment as of January accountability of benefit administration. 2011. ï‚· Develop updated tools to improve the targeting of ï‚· Step up eligibility audit efforts. non-contributory benefits for civilians aimed at reaching the poor and most vulnerable. ï‚· Introduce design changes in the social assistance programs to eliminate potential work disincentives and enhance the incentive compatibility of the social assistance programs to facilitate the transition to employment. x Public Sector Wages ï‚· Institute meaningful establishment controls ï‚· Systematically monitor staffing numbers to ensure (particularly in the state and FBH) and regularly the observance of establishment and hiring controls monitor staffing to control staffing and wage bill for all categories of public employees to facilitate expenditure. wage bill management. ï‚· Rationalize the large number of currently vacant ï‚· Regularly monitor employment and pay levels in posts and cut wage bill appropriations accordingly. sectors other than the public administration, as these ï‚· Reduce the cost of various salary increments with a sectors make up the bulk of public employment. combination for the following measures: ï‚· Cut other categories of recurrent spending related to o Abolish discretionary fees (e.g. for attending staff numbers, such as official travel, rental of committees and working groups) by law, as buildings, and telecommunications. they represent unnecessary salary top-ups for ï‚· Reduce the number of auxiliary staff by outsourcing functions that already fall within the officials‘ these functions or transferring these employees to duties. labor law contracts. o Eliminate universal entitlements to allowances ï‚· Improve coordination in wage bill planning in the (e.g. meal allowance) by absorbing them into Fiscal Council. base pay, as was done in RS. ï‚· Increase flexibility in negotiations with public o Restrict the eligibility to allowances for sector trade unions to avoid disorderly wage bill housing, family separation, and home visits to cutbacks in the future. expatriated officials (e.g. diplomats posted abroad) and limit it in time (e.g. up to three months) to other officials relocating upon first taking up employment in the public administration). Health ï‚· Expand the insurance pool related to hospital and ï‚· Review exemptions from contributions and consider pharmaceutical care. shifting the collection and control of health ï‚· Centralize procurement of drugs to at least the entity contributions to at least the entity level to help level. improve collection. ï‚· Fully implement hospital and primary health care financing reforms such as case-based payment systems (Diagnostic Related Groups, DRGs) and capitation payments in Primary Health Care. ï‚· Reduce fragmentation and duplicate functions. In FBH, considerable duplication of roles and functions exist concerning administration of health care in the MoHs and IPHs. Education ï‚· Limit further wage increases. ï‚· Develop a more decompressed wage structure ï‚· Rebalance spending between levels to enhance pre- which takes into account teachers‘ qualifications, school and tertiary education. experience, performance and other criteria to better ï‚· Introduce per capita financing of schools to replace attract and retain highly qualified teachers. the existing input-based financing system. ï‚· Establish an Education Management Information ï‚· Address the quality issues in upper secondary that System. are at the root of low enrollment rates. Forestry ï‚· Identify and reduce unproductive labor in public ï‚· Improve road access and conditions to help increase forest-sector institutions to improve productivity and production, improve harvesting efficiency and profitability. increase forestry contribution to GDP. ï‚· Develop a transparent and market-based mechanism ï‚· Invest significantly in human resources through xi for timber sales in a manner that allows independent development of additional skills in the areas of verification and has a definite timeframe and business and organizational development, change implementation indicators. and change management, investment appraisal, cost ï‚· Set the overall level of the OKFÅ  tax based on an in- benefit analysis, robust budgeting processes, public depth cost-benefit analysis and reflecting the true awareness and communication. value of public goods being provided. The system ï‚· Have the larger Cantonal Forest Management for allocation of the OKFÅ  collected and its Companies within FBH and JPS Å ume RS establish subsequent disbursement needs to be made clear and a comprehensive Business Process Review (BPR) of transparent with complete traceability and all the major processes within their organization to accountability. identify opportunities to improve compliance, improve quality and reduce costs. ï‚· Implement institutional changes in FBH: transfer the responsibility for security and protection of the forest resource Cantonal Forest Management Companies; have the Cantonal Forest Offices report formally to the Federal Forest Office and operate effectively as daughter organizations; give the Financial Police the sole responsibility for OKFÅ  (timber tax) compliance; increase the staff within the Forestry Department in FBH in line with its responsibilities. In RS review the staffing and roles of both Forestry Department and Forest Agency. Energy ï‚· Continue restructuring of the electricity sector ï‚· Continue the restructuring and modernization of towards EU standards for corporate governance and coal mines within the new institutional setup. set tariffs at their full cost-recovery levels, while ï‚· Implement the EE program for public buildings, putting in place targeted social protection focusing for instance in schools and hospitals mechanisms to protect vulnerable groups. ï‚· Set up a comprehensive and systematic energy data ï‚· Prepare and adopt an investment strategy for each of gathering and reporting system and launch targeted the power utilities based on a comparative least-cost information campaigns. investment plan which takes into account the power sector demand forecast at the entity, national, and regional levels. Options for private sector involvement should also be considered. xii CHAPTER 1: PUBLIC EXPENDITURE AND THE POLITICAL SETTING 1.1. This report presents the view that Bosnia and Herzegovina (BH) faces four main challenges in improving the public revenue and expenditure system. The first of these is to improve the structure of public expenditures which is excessively concentrated on current spending including wages, pensions and social transfers. These types of expenditure are being made on a scale that displaces more productive spending on such things as infrastructure and therefore compromise future growth. They embody perverse incentives, and those that should form part of the social safety net system, are in fact strikingly ineffective at reaching the poor and vulnerable. Moreover, within individual sectors such as health and education, there is a need to undertake structural reforms that will improve the efficiency of spending. 1.2. The second challenge is to improve the structure of the revenue system. This system is heavily reliant on payroll taxes which provide a strong inducement to both employers and employees to retreat to the informal sector and avoid (or evade) taxes. It is difficult for foreign investors and potential exporters to do this, and consequently they bear a disproportionate amount of the payroll taxes and of taxes in general. This thwarts FDI and undercuts competitiveness and growth. Consequently there is a need to reduce the size of the labor tax wedge focusing particularly on changes to social contribution rates. 1.3. The third challenge is to reverse the excessive growth of the public sector. Pre-crisis tax reforms and a boom in growth gave the authorities the room to sharply increase spending. But they went even further and the expansion in spending actually outstripped both revenue growth and GDP growth. Not only does this represent an inappropriate pro-cyclical fiscal stimulus, it has resulted in BH having one of the largest public sectors in the region – largely to accommodate the growth of relatively unproductive expenditures. This problem is exacerbated by the fact that BH has an aging population that will automatically give rise to fiscal pressures as social transfer demands for health and pensions rise and contribution bases dwindle. In addition there are significant fiscal challenges looming in the largely publicly-owned power sector where demand is rising and the sector is not well positioned to make the necessary investments. 1.4. The fourth challenge is to complete the overall fiscal architecture in the country by strengthening the Fiscal Council. BH has to dedicate monetary policy to supporting the currency board arrangement. Consequently fiscal policy is the only instrument left to manage aggregate demand. Moreover BH is also a decentralized country making it more critical to coordinate fiscal policy effectively. The Fiscal Council should therefore be a key institution for the management of fiscal policy. However, as evidenced by the protracted budgetary stalemate at the Fiscal Council since the October 2010 and from the recent suspension of payments on international debt, it is evident that the overall fiscal architecture is still from complete. 1.5. To some extent these challenges have roots in the political framework and the legacy of the war. The Dayton Peace Agreements, which ended the war in 1995, gave BH important distinctive features. BH was constituted as a country with a State Government holding relatively limited powers, and two largely autonomous sub-national Entities (and subsequently the small autonomous Brcko District). The larger of the two Entities, with two thirds of BH population, is called the Federation of Bosnia and Herzegovina (FBH). The smaller is called Republika Srpska (RS). 1 1.6. BH population is largely made up of three constituent peoples: Bosniaks, Serbs and Croats.1 Prior to the war the three groups were more evenly distributed throughout the territory of BH, but now FBH is predominantly populated by Bosniaks and Croats, and RS by Serbs. Since the war, there has been further polarization along ethnic lines within FBH, and the emergence of political parties affiliated with Bosniak and Croat communities, complicating consensus-building over important issues in that Entity. These tendencies are less marked within the more ethnically homogeneous RS. 1.7. The Government structures and budgetary arrangements of the two Entities are starkly asymmetric. FBH is subdivided into 10 cantons, each with its own executive, legislative, and judicial branches of Government. Each canton is subdivided into municipalities.2 In addition to a high degree of decentralization, the division of competencies between the cantons and FBH is unclear, rendering this Entity more difficult to govern. The RS is more centralized and is subdivided only into municipalities, and is thus easier to govern. Public budgetary arrangements reflect these features (Figure 1.1 and Table 1.1). While the direct taxes are in direct purview of the Entities, all indirect taxes are collected at a single location (BH Indirect Tax Authority). These revenues are used for debt service and funding of BH State budget, while the remainder is split according to the shares of final consumption among the two Entities and Brcko District. In FBH much of the spending is done at the cantonal level. These arrangements make it difficult to achieve consensus on the overall budget framework and tend to leave the State level with a residual share of resources. Figure 1.1: Administrative and Fiscal Structure in Bosnia and Herzegovina Bosnia and Herzegovina (State-level institutions) Federation of BH Republika Srpska Brcko Extra Entity Government Entity Government Extra District Budgetary Budgetary Funds (35) Funds (6) Pension (1), Cantons Pension, Extra Employment (10) Employment, Budgetary (11), Health, Funds (2) Health (11), Child, Employment, Road (11) Municipalities Municipalities Road, (79) (63) Health Motorway (1) Motorway Source: Constitutions of BH, FBH, RS. 1 The last census took place in 1991, prior to the outbreak of the war in 1992. At that time there were 43.5 percent Bosniaks, 31.2 percent Serbs, and 17.4 percent Croats. In 1991, BH had 4.3 million inhabitants. While there was no census since 1991, the World Bank estimates that BH had 3.8 million inhabitants in 2009. 2 The high degree of decentralization in FBH is the result of an agreement reached during the 1992-1995 war between the representatives of Bosniak and Croat ethnic groups to avoid dominance of the Bosniak majority over the Croat minority. 2 Table 1.1: Revenues, Expenditure and Balance by Level of Government 2008 2009 2010 2011 2008 2009 2010 2011 Share of GDP Share of consolidated GG Revenues State-level 3.5 3.9 3.6 3.4 7.6 8.9 7.7 7.0 FBH consolidated 27.9 26.1 27.9 28.6 60.1 59.4 60.2 59.4 FBH center 6.9 6.5 7.1 .. 14.8 14.8 15.4 .. RS consolidated 14.1 13.0 13.7 15.3 30.4 29.6 29.6 31.7 RS center 7.1 6.0 6.5 .. 15.2 13.7 14.1 .. Brcko district 0.9 0.9 1.1 0.9 1.9 2.1 2.4 1.9 2008 2009 2010 2011 2008 2009 2010 2011 Share of GDP Share of consolidated GG Expenditures State-level 3.4 3.9 4.0 3.6 6.9 7.9 7.9 7.0 FBH consolidated 29.5 28.6 29.2 30.3 59.0 57.8 57.9 59.0 FBH center 8.0 7.8 8.0 .. 16.0 15.7 15.9 .. RS consolidated 16.0 16.0 16.3 16.5 32.0 32.3 32.2 32.1 RS center 8.4 8.1 8.1 .. 16.8 16.3 16.0 .. Brcko district 1.1 1.0 1.0 0.9 2.1 2.0 1.9 1.8 2008 2009 2010 2011 2008 2009 2010 2011 Share of GDP Share of consolidated GG Overall Balance State-level 0.1 0.0 -0.4 -0.2 -2.9 -0.9 9.7 6.9 FBH consolidated -1.6 -2.5 -1.4 -1.7 45.0 45.5 32.4 53.0 FBH center -1.2 -1.3 -0.9 .. 32.7 23.4 20.6 .. RS consolidated -1.9 -3.0 -2.6 -1.2 52.6 54.6 61.2 38.8 RS center -1.4 -2.1 -1.5 .. 38.4 37.8 36.7 .. Brcko district -0.2 0.0 0.1 0.0 5.3 0.9 -3.3 1.3 Source: IMF and BH Authorities. Notes: Data for 2010 are preliminary while for 2011 are estimates and are not available for all levels. 1.8. The authorities have little choice but to confront these challenges directly. It is true that the transfers to certain powerful interest groups such as public sector workers, soldiers and veterans are politically difficult to contain. It is also true that consensus-building in a fractious multi-party system is hard. And it is clear that there are certain public sectors overheads associated with the complex administrative structures that reflect the ethnic geographic reality. But the current situation is not sustainable. For one thing the authorities are up against a financing constraint. Beyond the exhaustion of some short-term privatization receipts, they will find it difficult to finance fiscal deficits if they are not contained. They do not have access to international markets. There is some potential for borrowing from local banks and this has been done both in RS and FBH. But this is not a viable long term strategy. Moreover, there has been a protracted budgetary stalemate at the level of the Fiscal Council, with the State and Entity governments unable to come to an agreement on the fiscal framework. Consequently the State-level institutions were forced to operate without an adopted budget for 2011 and had to function on basis of decisions on temporary financing that can only fund operations on the same level as appropriated in the 2010 budget. This problem persisted into 2012 threatened to frustrate access to official financing from IFIs at time of international crisis when private financial flows are likely to be scarce, and closing imminent fiscal gaps may become very challenging. This underlines the fact that fundamental reforms urgently need to be implemented. 3 CHAPTER 2: THE FISCAL PICTURE AND RECENT ECONOMIC DEVELOPMENTS I. INTRODUCTION – OVERALL FISCAL DEVELOPMENTS 2.1. After a period of improvement between Figure 2.1: Growth and Fiscal Balance 2004 and 2006, public fiscal management before and after the Crisis. deteriorated sharply in 2007-2008 in the run up to the Global Financial and Economic Crisis. Following the successful introduction of indirect taxation in 2006 and the associated revenue windfall, fiscal performance deteriorated in 2007 and 2008 as public expenditure increased substantially, magnifying the impact of the rapid private sector demand growth on the current account deficit (CAD). Overall, in the period prior to the crisis, public expenditure grew faster than GDP, with spending on public sector wages and social benefits leading the way. 2.2. Declines in GDP (2009 real GDP contracted Source: IMF and BH Authorities. by 2.9 percent, which represents an almost 9 percentage point drop in the GDP growth rate compared to 2008) and the associated drop in revenues exacerbated fiscal outcomes (see Figure 2.1). Because of expansionary fiscal policies in 2007 and 2008, the deficit reached 3.5 percent of GDP in 2008 despite strong revenue growth. The budget deficit increased to 5.5 percent in 2009, as the country entered a recession. 2.3. The measures implemented in the context of the IMF program were critical to deficit reduction in 2010 and the budget deficit was brought down to 4.2 percent of GDP. In addition, as a part of the first Public Expenditure Development Policy Loan (DPL), the Entities took the first step toward a fundamental reform of the system of veteran and civilian cash-transfer benefits. This was to be accomplished by shifting to a needs-based, rather than a rights-based system of transfers. However, given the delay in formation of governments following the October 2010 general elections, progress in implementing these reforms has been slow. 2.4. The consolidated budget deficit continued to narrow in 2011 to an estimated 3 percent of GDP from 4.2 percent in 2010. At the time of writing this report (January 2012), BH Fiscal Council had not adopted the Global Fiscal Framework for 2011-2013 (due in May 2010) as well as one for 2012-2014 (due in May 2011). Consequently, the State-level institutions have been operating without a budget for 2011 and on basis of Decisions on Temporary Financing that can fund operations on the same level as appropriated in the 2010 Budget. This will likely continue for the first quarter of 2012. Meanwhile, the Entities have adopted their own budgets for 2011 and 2012 outside the coordinating mechanism of BH Fiscal Council which is a direct contradiction of the law.3 Consequently, the IMF has not been able to complete the fourth review of the SBA which was to have taken place in early spring 2011 and has since 3 Article 12 of the Law on Fiscal Council in Bosnia and Herzegovina requires that in absence of agreement on the Global Fiscal Framework, BH Council of Ministers, FBH Government and RS Government undertake to pass Decisions on Temporary Financing until agreement is reached. Under this scenario, governments can fund operations on the same level as appropriated in the last adopted budget. 5 been long delayed. In the meantime, the Entities have turned to issuance of domestic treasury bills and bonds for the first time in 2011 to finance their needs. Table 2.1: General Government Revenues and Expenditure (as percent of GDP) 2005 2006 2007 2008 2009 2010 2011 Nominal GDP 17,127 19,252 21,760 24,718 24,004 24,486 25,617 Revenues 46.7 48.2 47.0 46.0 44.9 46.3 46.0 Tax revenues 35.4 38.8 38.7 38.1 37.2 38.8 39.1 Indirect taxes 19.9 22.3 22.2 19.9 18.5 19.6 19.7 Direct taxes 3.2 3.3 3.1 3.5 3.0 3.0 3.2 Social s ecurity contributions 12.3 13.2 13.4 14.3 15.1 15.5 15.7 Other 0.3 0.6 0.6 0.5 Nontax revenues 5.5 5.5 6.2 5.5 5.6 5.6 5.3 Grants 4.6 3.7 2.1 2.4 2.1 1.9 1.7 Expenditures and net lending 45.9 46.1 46.8 49.5 50.5 50.5 49.0 Non-interes t current s pending 38.4 39.3 39.5 41.5 42.3 42.4 41.6 Capital expenditure and net lending 6.9 6.1 6.7 7.5 7.6 7.5 6.7 Foreign-funded capital projects 4.3 3.8 3.3 3.1 3.6 3.7 3.7 Other 2.6 2.3 3.4 3.6 3.1 3.0 2.9 Interes t 0.6 0.7 0.6 0.5 0.5 0.6 0.7 Balance on a commitment bas is 0.8 2.2 0.2 -3.5 -5.5 -4.2 -3.0 Primary balance 1.4 2.9 0.8 -3.0 -5.0 -3.6 -2.3 Source: IMF and BH Authorities. Notes: Data for 2010 are preliminary while for 2011 are estimates. Definition of General Government includes expenditures by State-level institutions, 2 entity governments, 10 cantonal governments, Brcko District government, all municipalities and 43 extra-budgetary funds (see Figure 1.1). 2.5. Establishing an effective mechanism for coordinating fiscal policy is critical in a country like Bosnia and Herzegovina. BH is extraordinarily decentralized country for its size, which is a direct legacy of the 1992-1995 war and Dayton Peace Agreements. Its many governments (14) and off-budget units (43) have a large degree of autonomy over expenditure planning, but their willingness to adopt budgets to support overall macroeconomic stability is undermined when others fail to do likewise. If left unchecked, the aggregate outcome could be excessive expenditure levels and macroeconomic instability. In short, a durable and effective fiscal coordination mechanism is needed to balance the objectives of individual administrative units against the broader ones of the country as a whole. As demonstrated by the current crisis in the euro zone, efforts in fiscal coordination are even more important within a single currency area or, as in the case of BH, with a currency board which offers no scope for independent monetary policy. 2.6. After years of delay, it was a major breakthrough when in 2008 the Fiscal Council (FC) was finally established. The FC is made up of three prime ministers and three ministers of finance and two non-voting members (the governor of the Central Bank and representative of Brcko District). The main responsibility of the FC, although not the only one, is to adopt the so-called Global Framework of Fiscal Balance and Policies (Fiscal Framework). This document is a key input for 3-year Framework Budget Documents of each of the state and entity governments which in turn forms the basis for adoption of their respective annual budgets. 2.7. However, several steps are needed to enable the FC to function effectively. As evidenced since 2010 and the protracted budgetary stalemate, optimal functioning of the FC depends on the political will of its members. Comprehensive and timely data, together with strengthened technical support and expert advice should help in reaching decisions. It is essential to strengthen the formal technical support structure for the FC in the medium term. In the short-term, the FC could rely and actively engage other domestic institutions that have such analytical capability, such as the Macroeconomic Unit of the Indirect Tax Authority, Department of Economic Planning, Sector for Economic Research as well as Department for Economic Research and Statistics of the Central Bank to provide analytical inputs for decision making. 6 2.8. In addition to its role in expenditure planning, the FC should take the lead in activities stipulated in Article 5, paragraphs 5 and 6 of the FC Law pertaining to the improvement of public finance management. This is particularly important in the following areas: i) public sector pay management; ii) public investment coordination and management; iii) social expenditure policies; and iv) public sector procurement. In order to do these things effectively the FC will need accurate and timely data for the entire general government (14 governments and 43 extra-budgetary funds) on budget execution, especially greater frequency of data on execution within a fiscal year. Most levels of government are using Treasury Single Account (TSA) which would enable obtain electronic records of expenditures, reconcile data and generate timely, accurate and comparable fiscal data that could improve frequency and timeliness to better inform policy making. The World Bank has completed several similar exercises through its BOOST4 initiative and would be able to offer assistance to this end. 2.9. Furthermore, there are several deficiencies in the current FC law that should be improved. The first shortcoming is the provision for breaking of the deadlock. At present, Article 12 of the Law on Fiscal Council in Bosnia and Herzegovina requires that in absence of agreement on the budget, BH Council of Ministers, FBH Government and RS Government undertake to pass Decisions on Temporary Financing until agreement is reached. In this situation, governments can fund operations on the same level as appropriated in the last adopted budget. However when this provision was last invoked both Entities violated its provisions by adopting their budgets irrespective of the lack of agreement on Fiscal Framework. Clearly, this provision could be made more workable by incorporating the same legal requirement in the Entities‘ Budget Laws. 2.10. The second shortcoming of the FC Law is that the penalty envisaged for noncompliance is especially weak. The violating party is fined 10 percent of any deviation from the agreed budget target and the fine can be used to pay off the violator‘s domestic debts. Strengthening the penalties for noncompliance is not straightforward. However, on-going discussions in the European Union aimed at redefining the Stability and Growth Pact and introducing a new fiscal compact could be of assistance to the authorities in incorporating penalties into the law that would effectively discourage deviations from the agreed targets. 2.11. Third, and perhaps easiest to overcome, is the fact that the definition of the consolidated budget balance excludes capital spending. However, it is the overall deficit that of most relevance for aggregate demand management. This can be addressed by simply defining expenditures under Article 1 of the law to include both current and capital spending regardless of the source of financing. II. COMPOSITION AND TRENDS IN PUBLIC EXPENDITURE 2.12. Given the complex political structure and the incomplete and stalled privatization process, the public sector plays a very significant role in the economy. During the reconstruction period (1996- 2002) public spending hovered around 50 percent of GDP. As shown in Table 2.1, the share of public spending in GDP fell to around 46 percent between 2004 and 2006. Enabled by the surge of revenues following the introduction of VAT, the governments in BH expanded spending from 2007. Rapid increases in public spending, in particular on public sector wages and cash transfers, resulted in a fiscally unsustainable position in the lead-up to the recession in 2009, and public spending again exceeded 50 percent of GDP. At the same time it is worrisome that there remain some 43 extra-budgetary funds that should be consolidated with the main budget to maintain a clearer picture of the overall fiscal position. 4 BOOST is not an acronym. It is the name of a new data tool developed at the World Bank to help enhance public sector performance. For more information go to http://go.worldbank.org/S14MC5Z9L0 7 Table 2.2: Consolidated General Government Expenditures as a share of GDP 2005 2006 2007 2008 2009 2010 2011 Consolidated expenditure on a commitment basis 45.9 46.1 46.8 49.5 50.5 50.5 49.0 Non-interest current expenditures 38.4 39.3 39.5 41.5 42.3 42.4 41.6 Wage bill 11.4 11.1 11.5 12.0 13.1 12.9 13.1 Goods and services 9.7 10.0 9.5 9.2 9.9 10.2 9.6 Transfers and Subsidies 13.2 13.7 15.9 16.9 17.0 16.8 16.5 Pensions 7.4 7.8 8.0 9.1 10.1 10.1 10.2 Social benefits 0.4 0.4 1.1 1.5 1.1 0.8 0.8 Other transfers (veterans, unemployment) 4.1 4.0 5.2 4.9 4.3 4.1 4.0 Subsidies to industry and agriculture 1.3 1.5 1.6 1.5 1.5 1.7 1.6 Other current spending 4.1 4.4 2.6 3.4 2.3 2.5 2.3 Interest payments 0.6 0.7 0.6 0.5 0.5 0.6 0.7 Capital expenditure and net lending 6.9 6.1 6.7 7.5 7.6 7.5 6.7 Source: IMF and BH Authorities. Notes: Data for 2010 are preliminary while for 2011 are estimates. Definition of General Government includes expenditures by State-level institutions, 2 entity governments, 10 cantonal governments, Brcko District government, all municipalities and 43 extra-budgetary funds (see Figure 1.1). 2.13. Consolidated general government Figure 2.2: Structure of Consolidated General expenditures are not laying down a Government Expenditure (in percent of GDP) foundation for future growth (Table 2.2 and Figures 2.2 and 2.3). Current expenditures rose sharply to 42 percent of GDP in 2011 (up from 38 percent in 2005). This represents fully 86 percent of total spending. The trend towards a greater share of recurrent expenditure was given further impetus by the crisis because capital expenditures cuts were found more politically palatable than cuts to recurrent spending. 2.14. Large recurrent expenditure stems partly from the complex constitutional make up of the country, but also from past unsustainable increases in cash transfers. Source: IMF and BH Authorities. Recurrent spending is crowding out investments needed to support longer term growth and is not Figure 2.3: Structure of Consolidated General sustainable. The largest share of this spending Government‟s Current Spending (in percent of goes on three categories: (i) public sector wages GDP) (13 percent of GDP in 2011, up from 11.4 in 2005); (ii) transfers and subsidies (16.5 percent of GDP in 2011, up from 13.2 in 2005); and (iii) procurement of goods and services (10.2 percent of GDP in 2011, up slightly from 9.7 percent in 2005). 2.15. Government size in BH is large, even when compared with the other countries in the Western Balkans. Using expenditures as a measure, Bosnia and Herzegovina, Montenegro and Serbia can be classified as having large governments, even when compared to the new EU member countries (EU10). Albania and Source: IMF and BH Authorities. 8 Kosovo can be classified as small government economies while FYR Macedonia has a moderate size of government. The size of the government sector is largely driven by social transfers and public sector wages. The large government countries (Bosnia and Herzegovina, Serbia and Montenegro) have public sector wages bill in excess of 10 percent of GDP, reaching almost 13 percent of GDP in Bosnia and Herzegovina. While Kosovo and FYR Macedonia spend less as percentage of GDP on wages, measured as percent of total expenditures, the wage bill still represents a considerable burden on the budget. Bosnia and Herzegovina, Montenegro and Serbia also spend significantly more funds on social transfers, driven largely by pensions. With the exception of Kosovo, the allocation for capital expenditures in SEE countries is relatively small. Table 2.3: Government Spending in Southeast Europe in 2010 ALB BIH KOS MKD* MNE SRB as percentage of GDP Wages and salaries 5.3 12.9 7.4 8 11 10 Social transfers 8.6 15.4 3.9 10.4 13.7 19.3 Other current spending 10.4 15.0 7 12.6 17.1 12.4 Capital spending 5.0 3.0 11.9 4.5 5.3 3.5 Total spending 29.3 46.3 30.2 35.6 47.1 45.6 Source: World Bank staff calculations based on data from MoF of SEE countries. IMF data for BH, excluding foreign financed capital expenditures in order to make the data comparable across countries, and MKD definition of general government were used. 2.16. The public sector wage bill as a share of GDP is not only one of the highest in the Western Balkans it is also one of the highest in the entire Europe and Central Asia region and is both a reflection of constitutional complexity as well as unsuitable hiring practices. This issue will be dealt in greater detail in Chapter 4. As shown in Table 2.3 the largest component of social transfers is pensions (10 percent of GDP) and results both from the aging population as well as from practice of retiring former soldiers early under so-called ―privileged‖ conditions. The rest of social transfers consist of non- contributory cash benefits for unemployment benefits, child protection and civilian and veteran disabled (nearly 5 percent of GDP). The latter benefits are particularly poorly targeted and they reach only 20 percent of the poorest quintile of the population. These issues will be dealt in greater detail in Chapter 3. Goods and services are sort of ―catch-all‖ category that includes some spending on education and health together with procurement of materials and services needed for daily operations of government. Issues relating to the expenditure on health and education will be dealt in chapters 5 and 6 respectively. 2.17. While public debt appears sustainable in the base case scenario, the scope for further deficit finance is limited. The government has no access to foreign finance (other than the IFIs). Its ability to raise domestic debt is limited. Short term t-bills were issued for the first time in 2011. Towards the end of 2011, limitations of domestic borrowing are becoming evident with interest rates rising and bid/offer ratios falling (Table 2.4). Commercial banks, the majority buyers of these bills and bonds, will likely face their own funding restrictions going forward as their foreign parents seek to limit exposures in Central and Eastern Europe. 9 Table 2.4: Issuance of T-bills and Bonds in 2011 Average Amount at Amount of Bid/Offer Date Issuer Maturity Amount sold Interest Rate, offer bids Ratio in percent 11-May RS MoF 6 months 30,000,000 83,918,686 2.80 36,000,000 2.48 20-Jun RS MoF 9 months 28,300,000 65,113,114 2.30 28,300,000 3.20 27-Sep FBH MoF 6 months 65,000,000 104,510,000 1.61 65,000,000 2.32 8-Nov FBH MoF 6 months 25,000,000 35,750,000 1.43 25,000,000 2.31 16-Nov RS MoF 7 years 120,000,000 120,000,000 1.00 120,000,000 6.83 29-Nov RS MoF 8 months 36,500,000 24,989,195 0.68 15,283,950 4.10 15-Dec RS MoF 6 months 21,000,000 12,950,801 0.62 9,835,080 3.36 Source: Banja Luka and Sarajevo Stock Exchanges. III. TAX STRUCTURE AND PUBLIC REVENUES 2.18. The main sources of tax revenue in BH Figure 2.4: Consolidated General Government‟s are indirect taxes (Value Added Tax (VAT), Revenue Structure (in percent of GDP) Customs Tariffs, Excise Duties). Following a reform of the indirect taxation system and successful introduction of a VAT in 2006, indirect taxes continue to be the main source of revenue for BH representing some 45 percent of total revenues (tax and non-tax) or 85 percent of tax revenues collected (Figure 2.4). Furthermore, BH is efficient in collecting VAT as there is a single rate and no exemptions (Table 2.7). 2.19. After the indirect tax reform, the two Entities undertook substantial reform of their Personal Income Tax (PIT) and Corporate Income Tax (CIT) during 2006-2009. Reform Source: IMF and BH Authorities. of PIT and CIT represents the only major tax reform since 2006 and these taxes have been simplified and largely harmonized across two Entities. However, given low rates and relatively recent introduction (of the PIT in particular), direct taxes still represent a very small share of public revenues (3.3 percent of GDP or 7.2 percent of overall revenues). 2.20. The key feature of the current tax structure is the very heavily reliance on labor taxation. Social contribution rates are much higher than all other tax rates in BH and are high compared to other countries. The combined social contributions rate in FBH amounts to 41.5 percent of gross salary. This is higher than in RS (33 percent) or the OECD average (29.5 percent) or the EU-85 average (38.1 percent). The consumption tax rate (VAT) in BH is 17 percent while income tax (both personal and corporate) rates are at 10 percent. 5 The new member states of the European Union. 10 Table 2.5: Tax Rates in Southeast Europe Bosnia and Herzegovina Albania Kosovo Macedonia Montenegro Serbia FBH RS BD Employee 17 18 17 8.8 10 18 15 11 Pension Employer 6 0 6 or 0 12.8 5 0 5.5 11 Health Employee 12.5 12.5 12 1.7 0 7.8 8.5 6.15 Insurance Employer 4 0 0 1.7 0 0 3.8 6.15 Social Unemploy- Employee 1.5 1 0 0.0 0 1.2 0.5 0.75 Contributions/ ment Employer 0.5 0 0 0.9 0 0 0.5 0.75 Payroll tax Other Social Employee 0 1.5 0 0.7 0 0 0 0 rates Contributio Employer 0 0 0 1.3 0 0 0 0 ns/Taxes Employee 31 33 29 11.2 10 27 24 17.9 Total Employer 10.5 0 6 or 0 16.7 5 0 9.8 17.9 Total 41.5 33 29 or 34 27.9 15.0 27.0 33.8 35.8 VAT 17 17 17 20 16 5 or 18 17 8 or 18 Personal Income Tax Rate 10 10 10 10 0 / 4 / 8 / 10 10 12 Corporate Income Tax Rate 10 10 10 10 10 10 9 10 Notes: Bosnia and Herzegovina - In the FBH contributions are levied on the wage only, while hot meal and transportation allowance are not included in the tax base. - In the RS, total income is subject to taxation for social contributions and there are no exclusions. - Brcko District does not have its own pension fund. Employees choose which Pension Fund they wish to contribute and employers pay the corresponding tax. - In the FBH, there is a personal exemption in the amount of KM 300/month (KM 3600/year) from the Income Tax. - In the RS, there is no personal exemption from the Income Tax. Kosovo - PIT in Kosovo is based in four brackets on the monthly income: EUR 0-80 0 percent, EUR 80-250 4 percent, EUR 250-450 8 percent and above EUR 450 10 percent. Macedonia - Corporate and Personal Income are taxed at flat rates of 10 percent. Effective rates are considerably lower. Personal income tax is levied on income after contributions and after tax credit. Generous CIT exemptions also exist; CIT is paid only on distributed profits; reinvested profits are exempt from CIT. - Social contribution rates are applied on gross income. Health insurance rate may be higher for certain occupations than the rate indicated in the table. Health insurance rate includes a statutory rate of 7.3 percent of gross income and additional 0.5 percent of gross income for occupational injuries. - Food, agriculture inputs, IT products, communal services and other necessities are taxed at the preferential VAT rate of 5 percent. Exports are exempt from VAT taxation. Serbia - there is a non-taxable exemption - first 7300 dinars (73 EUR) of the wage are not taxed at all. As a result, effective PIT rate is approximately 10.3 percent - there is an additional PIT which is paid annually by people who annually earn more than five times of the average wage. Marginal tax for this income is 10 percent. However, less than 1 percent of employed are subject to this tax - Social contribution tax is applied this up to five times average wage. If a persons‘ wage is higher than five times the average, no social contributions are paid for this additional amount. - 18 percent is the main VAT rate. There is also 8 percent VAT rate applied to some products, mostly food staples, but also some stuff that the Government thinks are "strategically important", e.g. IT equipment 2.21. One of the consequences of high taxes on labor is extensive informal employment. Income taxes and social contributions increase the difference between labor costs and take-home pay, which is defined as the tax wedge. In other words, the tax wedge is the share of taxes in total labor costs. Increasing the tax wedge, therefore, increases incentives for informality. 11 Table 2.6: Labor Informality in BH 2008 2009 2010 Total number of employed, LFS definiton 890,239 859,218 842,831 Total public employed, LFS definiton 263,897 249,427 255,535 Total non-public employed, LFS definiton 626,343 609,790 587,296 (percentage of total emplyoment) Informally employed (no pension) 30.3 28.3 26.1 Informally employed (no healthcare) 26.9 26.6 23.9 Informally employed (no pension, excluding public employment) 41.3 39.0 36.3 Informally employed (no healthcare, excluding public employment) 36.7 36.7 33.4 Source: WB Staff calculations based on LFS data. 2.22. The high level of informality (as well as high unemployment) is evident from World Bank analysis of the 2008-2010 Labor Force Survey data produced by BH Agency for Statistics (BHAS), (see table 2.6). Some 26 percent of employees are not covered by mandatory contributions to a pension fund (the first definition of informally employed), while some 24 percent are not covered by mandatory contributions to a health fund (the second definition of informally employed). If the employees from the state-owned companies are excluded from the total, the share of the informally employed is even higher -- 36 percent under first definition and 33 percent under the second definition. Informality appears to have been falling recently but the reason for this is that the non-public employment is decreasing much more rapidly than the public employment. Most informally employed are in the non-public sector. 2.23. The labor tax wedge is the Figure 2.5: Tax Wedge for Singles at 33 Percent of highest for the groups experiencing the average Wage in OECD and Western Balkans (2008/09) highest rates of informal employment. Using OECD Tax and Benefit Model to analyze labor taxation and benefits provided we find that agriculture workers and low-wage earners have the highest rates of informal employment, as the labor tax wedge on low-wage earners is one of the highest in Europe (Figure 2.5). These two categories of the population have the potential to absorb a significant number of unemployed people. The labor tax wedge drops off for higher earning individuals (Figure 2.6) implying very low progressivity of labor taxation in BH. This is particularly true in RS since February of Source: OECD Tax and Benefit Model 2011 when RS Government eliminated the personal exemption for the income tax, so the tax wedge for a single earner is the same irrespective of the wage. 2.24. Some countries in Europe can sustain higher labor tax wedges than BH can. These are generally higher income countries with a long history of respect for the rule of law and high rates of tax compliance. These behavioral patterns are bolstered by a sustained experience that pension contributors have received benefits as promised, and that high quality health services have been readily accessible 12 without the need for additional informal payments. Such confidence in public systems has not become deeply rooted yet in BH which means that evasion and informality are triggered at much lower contribution rate thresholds. Figure 2.6: Tax Wedge for Singles at 100 percent of 2.25. Certain design features of social average Wage in OECD and Western Balkans benefit systems also play a role in tilting (2008/09) incentives in favor of informality. An employer has a disincentive to register employees in order to save on high social contribution rates, and employees are willing to agree to such an arrangement as they can register as unemployed in order to claim a free health insurance benefit. In addition, provision of many social benefits is eliminated as soon as a recipient reports formal income, inducing those in the informal sector to stay there. 2.26. The obsolete health financing system is one of the main reasons for high social contribution rates. The Source: OECD Tax and Benefit Model. foundations of BH system of social contributions were developed during socialist times when the economy was operating near full employment and the size of the gray economy was negligible. Currently, however, the ratio of those who work in the formal sector and pay social contributions to those who are entitled to healthcare is estimated at 1:5. A system of health insurance that is largely financed by social contributions, in current circumstances with huge informal employment, is not only unfair, it also undermines firm competitiveness and creates disincentives to formality. 2.27. In order to sustain international competitiveness, BH authorities face the challenge of reducing the relatively high social contribution rates which increase unit labor costs directly, but they have difficulty finding the fiscal space to do so given past surges in the cost of wages and transfers. Collection of social contributions has been streamlined and transferred to a single collection point in the Entities‘ tax authorities. This aids with tax compliance and fights evasion (i.e. an employer cannot pay one tax without paying all social contribution taxes). The authorities have been trying to reduce social contribution rates in recent years, and have managed to reduce the contribution rate for unemployment from 2.5 to 1.5 percent. However, in the face of increasing current expenditures on social transfers and the public sector wage bill, there has been no fiscal space for a more serious reduction. On the contrary, RS government increased social contribution rates for pension, health and unemployment as well as the personal income tax rate from February 2011 in order to finance growing deficits in pension and health funds. There is a further problem because as workers move to informal employment the only ones left in the formal system are those in foreign-owned enterprises and in the public sector. A higher tax wedge on public sector wages does little expand fiscal space. And a higher tax wedge on wages of workers in foreign-owned enterprises discourages foreign investment and the technology transfer that would otherwise sustain competitiveness and growth and expand fiscal space. 2.28. Progress has been uneven across the two Entities. The RS has moved away from a health insurance system financed entirely by contributions to one that is partially financed by general revenues. This system is more efficient than the one in FBH, which is almost entirely (96 percent) dependent on social contributions. The FBH Government, on the other hand, has had an ongoing dialogue with the 13 World Bank on reducing the social contribution rate for health insurance, with the idea of shifting some of the financing burden to general revenues. 2.29. Social insurance contributions appear to be the biggest factor contributing to such large revenue differences in Southeast European (SEE) countries. Bosnia and Herzegovina, Montenegro and Serbia derive a substantial part of government revenue from social insurance contributions, and social transfers account for a large part of spending (though these also include relatively generous non- contributory social transfers such as social assistance and other benefits). Albania on the other hand has a much smaller social insurance scheme which does not cover a considerable part of the population, while Kosovo does not have one at all (although a privately managed pension insurance scheme does exist). Poor labor market outcomes have strained the finances of the social insurance schemes in most of SEE countries requiring increasing transfers from the budget. Pressures on the social insurance schemes are likely to increase going forward and will warrant reforms to the systems. Growth in the near future may be jobless while population aging will put additional strain over the medium and long term. Table 2.7: Government Revenues in Southeast Europe in 2010 ALB BIH KOS MKD* MNE SRB as percentage of GDP Direct taxes 3.6 3.0 2.9 2.9 4.9 5.7 Indirect taxes 12.9 19.7 18.2 13.4 19.5 17.3 Contributions 4.3 15.5 0 9.1 12.2 10.8 Other revenues 5.4 5.6 6.4 7.7 5.9 7.1 Total revenues 26.2 43.8 27.5 33.1 42.6 41 Source: World Bank staff calculations based on data from MoFs of SEE countries. IMF data for BH, excluding foreign grants in order to make data comparable across countries, and MKD definition of general government were used. IV. RECENT ECONOMIC DEVELOPMENTS AND PROSPECTS 2.30. As with other countries in the Western Balkans, BH‟s overall economic performance has been affected by the financial crisis in 2008 and the unfolding Euro area crisis. GDP growth in BH in 2011 is estimated at 1.7 percent - significantly lower than in 2008. The 2010 recovery gained momentum towards the end of the year and the rebound was modest at 0.8 percent. Investments appear to have recovered as imports of capital goods grew by 13 percent in 2011 compared to 2010, while long-term bank credit to private enterprises declined by 2 percent yoy, which was especially evident towards the end of 2011. Consumption appears to have recovered as well, as imports of consumer goods grew by 6.3 percent through in 2011 compared to 2010, while bank credit to households was weak at 5 percent yoy 2011. However, it is likely that it will take some time before a more robust recovery emerges. 2.31. These events have left their mark on public sector finances. With the pro-cyclical fiscal policy and the slowing of revenues during the crises, public debt has increased rapidly, although not yet to an unsustainable level. Largely as a result of the increase in public sector borrowing (during the crisis and in the post-crisis period), BH external debt (both public and private) is expected to grow from 49 percent in 2008 to 59 percent in 2012, but to fall thereafter. External public debt will grow from 17 percent in 2008 to 27 percent in 2012, while total public debt grew from 31 percent in 2008 to 40 percent in 2011 and will fall thereafter. The debt appears sustainable under the base-case scenario of the IMF debt sustainability 14 analysis embedded in the overall macroeconomic framework which assumes sustained fiscal consolidation through the implementation of the planned reforms. The scenario would be vulnerable to (unlikely) currency depreciation, a non-interest current account deficit shock, an unexpected slowdown in GDP growth and inadequate fiscal consolidation. However, these projections do not include public guarantees and recognition of potential liabilities such as restitution. Clearly, these risks need to be managed very carefully as they could endanger the overall debt sustainability. 2.32. BH public debt consists largely of concessional external debt to IFIs in the amount of 26 percent of 2010 GDP and of domestic debt in form of long-dated bonds (14 percent of the 2010 GDP). The largest share of the public external debt (27 percent) is towards the International Development Association (IDA) acquired during the post-conflict reconstruction phase (1996-2002). The remainder of the external debt is to European IFIs and IMF as BH does not have access to international capital markets. Domestic debt is in form of bonds that were issued to settle pre-war frozen foreign currency savings and war damage claims by citizens. Given the limited development of local financial markets, governments in BH did not turn to them for regular issuance of bonds and bills for deficit financing until 2011. However, as mentioned above, limitations of domestic borrowing are becoming evident with interest rates raising and bid/offer ratios falling (Table 2.4). 2.33. Looking forward, BH faces a significant challenge in maintaining a prudent fiscal policy. With limited access to foreign financing and possibly week revenue performance, expenditure pressures must be kept in check, especially in the short term. This will require vigilance particularly with respect to the large transfer programs that currently account for such a disproportionate share of spending. Over the longer term the authorities need to take account of demographic realities. The BH has an aging population. Because of this there are going to be significant additional fiscal pressures from rising health and pension systems as the population ages and from a decline in the relative size of the working age population. The excessively generous and inefficient nature of these transfer programs makes the need for reforms all that much more urgent. Similarly, policies in the energy sector must be designed to avoid the buildup of arrears and contingent liabilities. Full advantage should be taken of potential growth areas such as forestry. Without such measures BH faces the risk of rapidly sliding into an unsustainable public financial situation. V. RECOMMENDATIONS 2.34. BH authorities face a stark choice regarding the management of public resources going forward. On the one hand the authorities can continue on the current path with increasing (and unproductive) recurrent expenditure, higher taxation, and most recently, a search for additional borrowing mechanisms. It may be possible to borrow more at least for while, but this is very risky. Direct borrowing abroad will be impossible for the foreseeable future. Borrowing from local banks the majority of which are foreign-owned may or may not be possible and such borrowing is, and will continue to be, both expensive and short term. There is a significant volume of repayments of domestic bonds issued to settle pre-war frozen foreign currency savings claims and IMF debt coming due in 2013 and 2014. The existing debt is largely with IFIs which are unlikely to provide more without the kind of structural reforms identified above. But, most importantly, further borrowing without the necessary structural reforms will raise debt and stunt growth thereby accelerating the country towards a position that is unsustainable. 2.35. The right course is to recognize the need to shrink the size of the public sector, reduce the magnitude of the tax burden, especially on labor, and make appropriate changes to the structure of expenditures and taxation that will enhance efficiency and growth. Unfortunately specific 15 recommendations for this which were laid out in the last PEIR have largely been ignored. In order to improve the broad management of fiscal policy the following recommendations are offered: Strengthen the role of the Fiscal Council (FC) and improve fiscal management ï‚· Improve the Law on Fiscal Council in the following manner: i) improve the provision on breaking the deadlock in case of non-adoption of Global Framework on Fiscal Balance and Policies by inserting similar provisions in the Entities‘ Budget Laws; ii) strengthen the penalty for non-compliance; iii) include capital expenditures in the definition of consolidated expenditures and fiscal targets. ï‚· Exploit data from electronic Treasury Single Accounts to obtain electronic records of expenditures, and generate more frequent fiscal data (World Bank‘s BOOST6 initiative). ï‚· Strengthen the analytical capacity of the Fiscal Council by forging stronger links to the Macroeconomic Unit of the Indirect Tax Authority, the Sector for Economic Research of the Directorate for Economic Planning as well as the Department for Economic Research and Statistics of the Central Bank. ï‚· The FC should take the lead in activities stipulated in Article 5, paragraphs 5 and 6 of the FC Law pertaining to the improvement of public finance management. This is particularly important in the following areas: i) public sector pay management; ii) public investment coordination and management; iii) social expenditure policies and iv) public sector procurement. Reduce the magnitude of the tax burden, especially on labor ï‚· Harmonize bases for social contributions and personal income tax across entities by eliminating non-taxable exemptions (hot meal allowance, transportation to work, etc.) by including all these non-wage benefits into the bases. ï‚· Reduce social contributions following the broadening of labor tax base to achieve revenue neutrality following the harmonization of bases for social contribution. ï‚· Consider further reductions in health contribution rates, especially in FBH, and replace forgone revenues with other revenues, namely excise and higher income taxes. ï‚· Increase excise taxes and harmonize related rates with those of the EU and neighboring countries, especially on tobacco and earmark these revenues towards financing of health sector. ï‚· Increase the progressivity of personal income tax with introduction of higher tax brackets for higher income earners. ï‚· Ensure that any reductions in labor taxes are balanced with other benefit reductions or other less distorting revenue increases in order to preserve fiscal sustainability. Reshape public expenditures which at present are excessively concentrated on current spending including wages, pensions and social transfers. See recommendations in subsequent chapters that deal with these individual sectors. Reverse the excessive growth of the public sector by increasing efficiency of spending. Further specific suggestions on this are offered in the subsequent chapters including those on health, education, forestry and energy. 6 BOOST is not an acronym. It is the name of a new data tool developed at the World Bank to help enhance public sector performance. For more information go to http://go.worldbank.org/S14MC5Z9L0 16 CHAPTER 3: PENSIONS AND SOCIAL TRANSFERS I. PENSIONS 3.1. Pension systems in Bosnia-Herzegovina are provided at the entity level. Since each of the systems is somewhat different, the pension analysis in this chapter is presented separately for each entity. Pensions in FBH 3.2. A key feature of FBH pension system is the very high dependency rate of 73.3 percent, i.e. each contributor needs to support 0.73 pensioners. Figure 3.1 compares the dependency rates in FBH with rates in high income countries and shows that FBH dependency rate exceeds that of countries like Greece, Spain, and Portugal, and is outmatched only by Italy and France. The FBH looks like less of an outlier in comparison to the transition countries of the Europe and Central Asia Region as shown in Figure 3.2, but even here FBH is in the top third of countries. The transition countries tend to have higher dependency ratios than non-transition countries for a number of reasons. Lower retirement ages tend to boost the number of beneficiaries. More generous definitions of disability tend to further boost the number of beneficiaries. In the case of BH, the wartime history has also increased the number of disabled. But the dependency rates are also driven by the number of contributors, and the high unemployment rates and the growth of informal labor markets in the transition countries. These factors tended to reduce the number of contributors at the same time that the number of beneficiaries was rising. This is particularly true in FBH and can be seen by looking at the number of contributors as a percentage of the working age population, defined internationally as the population aged 15-64 (see Table 3.1). The FBH‘s contributors, at 28.6 percent of the working age population, are among the lowest in Europe. Compared to high income countries as shown in Figure 3.3, FBH‘s percentage of contributors is clearly an outlier. Even compared to the other transition countries as shown in Figure 3.4, FBH‘s low number of contributors is matched only by Albania in Europe and by Armenia and Azerbaijan in the former Soviet Union. 3.3. The low contributor numbers and relatively high pension spending numbers lie at the core of the problems of FBH pension system. Figure 3.5 shows the pension spending in FBH at 9.4 percent of GDP as among the highest among high income countries, higher than in countries like Sweden and Japan which are both substantially older and wealthier than FBH. Figure 3.6 shows that even among the transition countries FBH is in the top third of pension spenders. While pension spending is clearly related to the percentage of elderly in the population (elderly defined as aged 65 and older), and BH is older than many of the other transition countries, Figure 3.7 shows that there are a number of older transition countries such as Lithuania, Latvia, and Bulgaria which have successfully managed to contain pension spending. The growth in spending in FBH is even more startling than the actual number of pensioners. Spending has risen from 7.6 percent of GDP in 2005 to 9.4 percent of GDP in 2009, in four short years. Part of this is attributable to a decline in GDP in 2009 due to the financial crisis which tended to boost the share of pension spending upward. But the pension spending share still represents the amount of resources that have to be freed up from elsewhere in the economy to support pensioners. 17 Table 3.1: Pension Indicators for FBH Indicators Total Men Women Number of 473,385 286,615 186,770 Contributors Number of Total Total Men Male Women Women Total Beneficiaries Regular Privileged Regular Privileged Regular Privileged Total 347,097 328,427 18,670 148,030 17,346 170,043 1,166 Old Age 154,106 137,451 16,655 95,351 16,158 42,100 497 Disabled 78,348 77,106 1,242 52,391 1,188 24,715 54 Survivors7 114,643 113,870 773 288 - 103,228 615 System Dependency 73.3% Ratio Expenditures (total in Total Total Men Male Women Women mil KM with % of Total Regular Privileged Regular Privileged Regular Privileged GDP in parentheses Total 1,424.63 (9.4% ) Old Age 715.53 617.42 98.11 442.15 94.92 175.27 3.18 (4.7%) (4.1%) (0.6%) (2.9%) (0.6%) (1.2%) (0.02%) Disabled 287.8 282.13 5.67 193.01 5.44 89.12 0.23 (1.9%) (1.9%) (0.04%) (1.3%) (0.0%) (0.6%) (0.0%) Survivors8 421.31 (2.8%) Average Monthly Pension Old Age 386.45 489.56 296.36 533.52 Disabled 307.00 381.39 300.49 359.54 Survivors 313.16 Contribution Revenues (in mil KM 1371.75 with % of GDP in (9.0%) parentheses) Budget Financing (in 83.4 mil KM with % of (0.5%) GDP in parentheses) Average Covered Wage (monthly in 790.075 thousand KM) Contributor Coverage 28.6% (contributors/pop 15- 64) Pensioner Coverage 61.8% (pensioners/pop 65+) All Beneficiaries/Pop 107.1% 65+ Source: Pension Fund Administrations. 7 Total number of survivors includes widows, widowers, and orphans, but the gender-specific breakdown does not include orphans as the gender for orphans was unavailable. 8 Survivor expenditures were not available by gender or by degree of privilege. 18 0 2 4 6 8 10 12 14 16 Iceland 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 0.0% 100.0% Australia 0.00% 100.00% 10.00% 20.00% 30.00% 40.00% 50.00% 60.00% 70.00% 80.00% 90.00% Ireland United States Countries Canada Bosnia Fed Spain Netherlands New Zealand Canada Ireland Netherlands Italy Norway Greece Canada Israel France Denmark United Kingdom Belgium New Zealand Denmark Ireland Norway High Income Countries United States Germany Finland Luxembourg Switzerland Luxembourg Austria Japan Australia Belgium Sweden Netherlands Spain Spain United States Finland United Kingdom Greece Japan Portugal Sweden Belgium Sweden Portugal Bosnia Fed New Zealand Japan Finland Germany Norway Austria Portugal Switzerland Germany Greece Iceland in FBH Compared to High Income Countries Austria Bosnia Fed Denmark France Israel Italy Italy Luxembourg France Age Population in FBH Compared to High Income Figure 3.3: Contributors as Percentage of Working Figure 3.1: Dependency Rates in FBH Compared to Figure 3.5: Pension Spending as Percentage of GDP Source: World Bank, Human Development Social Protection. Source: World Bank, Human Development Social Protection. Source: World Bank, Human Development Social Protection. 19 0 2 4 6 8 10 12 14 16 18 20 10% 20% 40% 50% 60% 70% 90% 30% 80% 0% 100% Kosovo 20.0% 40.0% 60.0% 80.0% 0.0% 100.0% 120.0% Kyrgyz Republic Armenia Georgia Azerbaijan Uzbekistan Armenia Albania Kazakhstan Kazakhstan Bosnia and … Latvia Azerbaijan Bosnia Republika … Kyrgyz Republic Russian Federation Kyrgyz Republic Poland Albania Turkey Belarus Turkey Moldova Turkey Uzbekistan Macedonia, FYR Czech Republic Romania Macedonia, FYR Transition Countries Montenegro Bulgaria Montenegro Serbia to Transition Countries Czech Republic Estonia Ukraine Latvia Serbia Romania Lithuania Lithuania Kazakhstan Slovenia Moldova Slovak Republic Russian Federation Romania Bosnia Fed Croatia Hungary Macedonia, FYR Poland Russian Federation Montenegro Bulgaria Bosnia Fed Poland Slovak Republic Bulgaria Belarus Hungary Azerbaijan Bosnia Republika … Lithuania Moldova Croatia Slovenia Slovak Republic Figure 3.4: Contributors as Percentage of Hungary Belarus Bosnia Republika … Estonia Czech Republic Croatia Slovenia Estonia Ukraine Figure 3.6: Pension Spending as Percentage of Working Age Population in FBH Compared to Serbia Latvia Albania Ukraine Uzbekistan Armenia GDP in FBH Compared to Transition Countries Figure 3.2: Dependency Rates in FBH Compared 3.4. One reason for the large number of pensioners and thus Figure 3.7: Percentage of Elderly Population in FBH the high level of pension Compared to other Transition Countries spending is the prevalence of young pensioners. Figure 3.8 below shows the percentage of old age pensioners in different age groups. A full third of the old age pensioners are younger than 65, while 59 percent of the disabled pensioners are below the age 65 and 38 percent of widows are below the age of 65. These younger individuals will be Source: World Bank, Human Development Social Protection. collecting pensions for a very long time. FBH pension system is not exclusively or even primarily about providing support to the elderly, but provides support to all age groups which creates incentives for younger individuals to withdraw from the labor force. Furthermore, based on the estimated population statistics, as few as 61 percent of the elderly (above age 65) are collecting any type of pension from the state. While the population may be overestimated, there is such a large gap between the 61 percent receiving some form of pension and the 100 percent that a high- coverage pension system would provide, that even if the elderly population is somewhat lower, it still suggests that a substantial percentage of the elderly are receiving little support from the Government. And, as the working-age cohorts of today, of whom less than a third are contributing and earning eligibility toward a pension, reach retirement, even fewer elderly will be covered under the publicly provided pension system. Figure 3.8: Age Distribution of Old Age Pensioners in FBH >85 2% <55 6% 75-84 19% 55-64 27% 65-74 46% Source: World Bank, Human Development Social Protection. 3.5. The number of disabled pensioners is almost 50 percent as large as the number of old age pensioners, as shown in Figure 3.9. Some of this might reflect FBH wartime history, but also reflects the relatively generous definition of disability employed in FBH as discussed below. Looking at women pensioners, the majority are collecting widow pensions. Again, this partly reflects the wartime history, but it also suggests that even prior to the war, relatively few women were working and earning pensions in their own right. 20 Figure 3.9: Types of Pensioners in Bosnia, Federation All Pensioners Male Pensioners Female Pensioners 3% 0% Old Age 25% Survivors Disability 30% Old Age 32% 44% Old Age Survivors 68% 61% Disability 14% Disability 23% Source: World Bank, Human Development Social Protection. 3.6. Another issue that is often cited for the increased pension expenditures is that of privileged pensions. These are pensions provided to those who were involved in the war. They are provided at much younger ages, with far fewer years of contribution, and with far higher benefit levels than normal pensions, hence the designation ―privileged‖. Currently these pensions comprise more than 7 percent of total pension expenditures. Benefit levels are anywhere from 20 percent to 80 percent higher than for comparable non-privileged pensions, and given the young age of many of these retirees, they will be collecting pensions for a very long time. The Constitutional Court of the FBH has ruled unconstitutional the Government Decrees that served as a legal basis for retiring of individuals under privileged conditions, even though the application of these Decrees has expired and could not be used as a basis for early retirement. The new law on privileged pensions is being proposed in order to legalize already retired individuals under privileged conditions. However, there is a risk that this process will further open up the possibility for additional retirement of beneficiaries under these privileged conditions. Parameters of the FBH Pension System 3.7. Another key feature of FBH pension system is the low average age of retirement. Table 3.2 summarizes the parameters of FBH pension system. While the pension law says that individuals, both men and women, can retire at age 65 with 20 years of service, exceptions were initially granted until 2005 that allowed men to retire at age 60 with 35 years of service, and women at age 55 with 30 years of service. This exception has been continually extended, most recently until 2015. In addition, individuals with 40 years of service can retire at any age and will be able to do so even after 2015. In the former Yugoslavia, many people earned additional months of service credit for each year actually worked, enabling individuals in certain occupations to reach 40 years of service credit as young as age 48. And since no additional pension is given beyond the 40th year of service credit, people want to retire immediately after completing their 40th year. The appropriate remedy for this issue is not to grant immediate retirement after the 40th year, but to allow people to continue accruing pension benefits and credits beyond the 40th year. In most European countries, 45 years is considered a full career with extra pension given to those that work even longer. Figure 3.10 provides a comparison of retirement ages in high income countries and among transition countries. Not one of the high income countries allows retirement as young as FBH and the only transition countries to do so are the Russian Federation, Belarus, Ukraine, and Uzbekistan. Given these low retirement ages, it is not surprising that the dependency rates are so high and that the percentage of young pensioners is so high. 21 Table 3.2: Pension Parameters in the FBH Retirement Age 65 M/F with 20 years of contribution as of 2003; Any age with 40 years of contribution Until 2015, can retire at 60M/55 F with 35M,30 F years of contribution, but subject to penalty of 1 percentM/0.5 percent for each year before the age of 65; Accrual rate 2.25 percent per year for first 20 years; 1.5 percent per year for next 20 years; 0 beyond 40 years Pensionable base Average net wages earned during the previous 40 years (increasing from 18 year average to 40 years between 2006 and 2015), revalued with historical average wage growth Indexation post- To average wage growth by law, but not implemented as such, subject to revenue retirement constraint Eligibility for Contributed for 1/3 of working life from age 20 to age of disability; disability pension No requirement if work injury; Defined as inability to work at previous job Level of Disability 50 percent for up to 20 years‟ contributions; Pension 1.67 percent more for additional years with a maximum of 75 percent reached at 35 years; Automatic maximum irrespective of years if work injury Eligibility for Age 45, incapacity, or caring for a child in case of widow; survivor‘s pension Age 60, incapacity, or caring for a child in case of widower; Children aged 15 or under and until age 25 if students or lifetime if incapacitated Level of survivor‘s 70 percent of contributor‘s entitlement if 1 person; 80 percent to be divided by 2; 90 pension percent to be divided by 3; and 100 percent to be divided if 4 or more Contribution Rates 23 percent of gross salary – 17 percent by employer and 6 percent by employee Source: FBH Pension Regulations. Figure 3.10: Retirement Ages in Transition Countries Retirement Ages in High Income Countries in Retirement Ages in Bosnia Federation Compared to Comparison to Bosnia Federation Other Transition Countries 70 70 65 65 60 60 55 55 Men Men 50 Bosnia and … Serbia Women Armenia Women Belarus Albania Georgia Montenegro Croatia Slovenia Estonia Russian Federation Ukraine Latvia Moldova Romania Bulgaria Azerbaijan Kyrgyz Republic Lithuania Slovak Republic Poland Hungary Kosovo Czech Republic Uzbekistan 50 Turkmenistan Macedonia, FYR Kazakhstan Greece Austria Luxembourg Ireland Netherlands Belgium Iceland Sweden Japan Australia France Israel United Kingdom Italy Canada Denmark Germany Portugal Spain Switzerland Finland Norway New Zealand United States Source: World Bank, Human Development Social Protection 3.8. Pension benefits are determined by the three parameters: the accrual rate; the pensionable wage base; and indexation rules. The accrual rate is the percentage of the pensionable base awarded to a newly retiring pensioner per year of service credit. Years of service credited can be substantially higher 22 than years of actual contribution as a large number of occupations are awarded more than one year of service credit per year of actual contribution. In addition, all workers have been provided double service credit for the war years. The accrued percentage is then multiplied by the pensionable base to obtain the initial pension. The initial accrual rate in FBH is both high and encourages short working careers. The accrual rate of 2.25 percent for the first 20 years is higher than the international best practice which is around 1.5 percent. As a result of the higher rate, workers can receive 45 percent of their pensionable base after only 20 years of service. In most countries, workers have to work 30 years to reach this level of pension. The faster accrual in FBH encourages workers who have reached a retirement goal to withdraw from the formal labor force early even if they have to wait until a later age to receive their pension. On the pensionable base, FBH is moving to using a career average by 2015 which is international best practice, rather than the average of the last 10 years which they had inherited. However, when calculating a career average wage, the nominal wages of each past year are typically revalued based on the cumulative economy wide average wage growth in the years between when the wage was earned and when the individual retires. FBH law follows this international best practice, although the actual implementation is quite different, as discussed below. The law also indicates that pensions in payment are to be indexed to the growth in average wages. International best practice suggests that allowing pensions to keep pace with wage growth is typically not affordable, particularly in aging countries like BH. The best practice is to index pensions to inflation so that they maintain their purchasing power. 3.9. Under BH law pension benefits are also conditioned by a unique feature that limits payments through imposition of a revenue constraint. Under this system a coefficient is determined by adding up all the pension amounts which need to be paid in a given year, adding up the revenue which is collected from contributions in that year, and dividing the pensions by the revenue. Multiplying the pension due by the coefficient to determine actual payments in principle limits payments to the available resources. The coefficient was put into place just after the war and was meant to avoid accumulating arrears from a potentially over-generous pension system which might not be payable in the foreseeable future. However, the long-run implementation of this revenue constraint coefficient has led to a number of problems. First, while the coefficient should legally be adjustable both up and down, politically, there has been a strong resistance to any downward movement of the coefficient. The coefficient rose fairly steadily until 2008 as revenues grew, but has been stuck at the level of 1.65 since August 2008. The FBH‘s implementation of this coefficient involves summing the pensions which were initially awarded to each pensioner, calculating the coefficient on that basis, and then applying this coefficient to the initial benefit award. But what this implies is that the person who retired ten years ago receives the same 65 percent above his initial award as the person who retires this year and whose pension benefits are already based on the nominal wage of today. The newly retiring person in 2011 does not need an additional 65 percent added to his newly calculated benefit while 65 percent additional may not be sufficient to cover cumulative inflation for an older retiree. Second, partially to correct for this and also to remain within the revenue constraint coefficient of 1.65, the pension fund has not been revaluing earnings earned since 2000 when creating the pensionable wage on which the pension is calculated. This does not fully correct the first inequity and causes yet another one where people and their earnings in different cohorts are not being treated equally. Third, the Government meddles in the pension system by creating special benefits for certain categories like the privileged pensioners or by dictating levels of minimum pensions. While FBH government is now transferring to the pension fund the money necessary to fund some of these benefits, in the past little or no money was provided to cover these largely politically-motivated benefit awards and increases. Moreover, when the system is revenue constrained, the coefficient system has been applied in a manner that protects the privileged pensions and squeezes the regular pensions. 3.10. The pension system is broken to a degree that requires a wholesale makeover. Given the various ad hoc measures added on the benefit side, there is almost no way to salvage the existing benefit structure. It is not surprising that the opacity of the pension system and the lack of clarity on what 23 benefits individuals will receive should they contribute results in people being unwilling to contribute. The FBH Government urgently needs to abandon the current structure and move to a new benefit system with everyone‘s benefit calculated under a consistent set of rules. Rather than reduce pensions for those who might have ended up with higher pensions than what they had fairly earned, the approach might be to maintain current pensions for those individuals without any additional increases until they reach the appropriately fair level of pension. 3.11. FBH law on disability is far more generous than in other countries. It considers individuals who cannot perform their previous jobs as qualifying for disability. This is a far less restrictive definition than common in other countries, where the definition of disability is the inability to perform any type of job and might be responsible for the current higher level of disability in FBH than elsewhere. Individuals who become disabled are frequently able to perform other jobs which require less physical demands or can be retrained for less physically demanding jobs. The law seems to preclude any such job change or retraining in the formal sector, but could encourage individuals to both collect a full disability pension and work in the informal sector. 3.12. Furthermore, disability benefits are more generous than old age benefits, resulting in an incentive for individuals to qualify under disability rather than under old age benefits. After 20 years of contributions, a disabled individual collects 50 percent of pensionable wage rather than the 45 percent that an old age pensioner would collect. Each additional year of service also generates higher pension increases for the disabled compared to old age. While it is true that individuals who become disabled at younger ages will have had fewer years to earn service credit and will therefore receive lower pensions for their entire life, providing more generous benefits per year of service is not necessarily the best solution to this problem since it creates an incentive to qualify as disabled rather than through old age, but may still provide too small a pension for the disabled. Most countries provide disabled individuals who have completed minimum years of service, full service credit for the years up until some age which might qualify them for the earliest old age retirement and then some fraction of the years between that age and the normal retirement age. Then the old age pension formula is applied using this higher number of service years. This provides the disabled roughly the same pension as they might have earned had they worked through to old age, and provides for a better living standard for the disabled, rather than giving them a slightly higher accrual rate and only recognizing the years that they actually worked. However, in this case, the individual should not be eligible for veterans‘ be nefits on top of a full earned pension. It is also true that the disabled might have greater needs than the elderly. But again, the better solution might be to provide some type of caregivers‘ allowance to those who are unable to care for themselves rather than to slightly increase benefits to all disabled. 3.13. Survivors‟ benefits are also provided on a more generous basis in FBH than is typical internationally. The eligibility for survivors‘ benefits stems from the old Yugoslav system, but is not found elsewhere. Widows are provided pensions for life if they are caring for a child who falls within the survivor age limits, is disabled herself, or is aged 45 or above. These pensions are provided irrespective of whether the widow remarries. Should a widow be childless and below the age of 45 at the time of her husband‘s death or have children who reach the age of ineligibility before she reaches the age of 45, she will receive nothing. Similar provisions apply to widowers, but at age 60. Internationally, widows‘ and widowers‘ pensions are provided only when the widow reaches retirement age with the assumption that the widow is expected to work when she is of working age. Some support is provided to young children even if the widow is young which might enable the widow to stay at home and care for the children. Widows‘ pensions are typically withdrawn if a widow remarries with the rationale that the widow is now being supported by the new spouse and no longer needs support from her previous spouse. Given the high proportion of survivor benefits being paid in FBH - almost a third moving toward international norms would save a considerable amount of money, encourage women to join the labor force and focus government resources on those who are in the most need. 24 3.14. Furthermore, the benefit provided to widows and widowers is a bit generous when compared to international norms. It is well recognized that if an elderly couple is supported by only one pension and one person of that couple passes on, the remaining member needs more than just half of the pension to live on, since two can live more cheaply than one. As a result, survivor pensions typically range between half and two-thirds of the individual pension. At 70 percent of the pension, the single survivor pension in FBH is on the high side. However, it should be noted that all of these benefits are subject to the same revenue constraint as the old age pension and might not in practice be considered a generous pension. II. RECOMMENDATIONS FOR FBH PENSION REFORM 3.15. FBH has set up an expert group for pension system reform and developed a strategy for pension reform. Many of these elements need to be further refined and elaborated, but even this strategy has yet to be adopted by FBH government. The strategy includes the following elements: (i) instituting a unified system of registration, control, and collection of contributions; (ii) possibility of increasing the retirement age by one or two years as well as stricter criteria for early retirement; and (iii) stricter linking of contributions and benefits, perhaps through adoption of the point system, with more appropriate indexation of all pensions above the lowest level. The strategy also suggests that privileged pensions be entirely separated from normal pensions, and while managed by the Pension Fund, be financed by resources from FBH Government. This strategy needs to be further refined and adopted by FBH Government and should include the following elements: ï‚· Strict curbs on early retirement. While raising the retirement age higher than 65, as the strategy proposes, is a good thing, the issue in FBH is that most people are retiring well below that age because of the early retirement provisions, allowing men to retire at 60 and women at 55. Although these early retirement provisions are only in place until 2015, they need to be eliminated right away and should under no circumstances be extended further. ï‚· Establishment of actuarial pension reductions for those who do retire early. For those who continue to be able to retire early, pensions should be reduced not by the current 1 percent per year for men and 0.5 percent for women, but by the actuarially fair reduction of 6 percent per year for every year before the age of 65. It is likely that far fewer individuals will choose to retire early under these circumstances. ï‚· Complete revamping of the benefit structure, eliminating the coefficient as it now exists. Adoption of a new benefit structure is absolutely critical, and consideration should be given to adopting it for existing pensions as well. The point system, as suggested in the strategy, is one possibility, but more important than the format of the pension system is that its parameters be designed to be fiscally sustainable. This means that the accrual rate in a conventional defined benefit system or the value of the general point in a point system need to be set at realistic and sustainable levels. The pensionable base needs to be revalued according to growth in average wages, and pensions once received need to be indexed to inflation. The current approach of setting unrealistic parameters and then whittling down the pensions by applying a coefficient in a somewhat ad hoc manner needs to be avoided at all costs. ï‚· Eliminate double-dipping between war-related benefits and pension fund benefits. Individuals should only be eligible for one publicly funded benefit, either a war-related benefit or a benefit coming from the pension fund. This could be implemented by cross- checking a database between the two types of benefits. 25 ï‚· Tightening eligibility conditions for disability and revamping the benefit structure to correspond with the new old age structure. As effective retirement ages begin to rise, the demand for disability pensions will increase further. The eligibility conditions need to be tightened to provide full pensions only to those who are incapable of any kind of work. The disability benefit structure also needs to be modified in line with the old age benefit design, making sure that disability benefits do not exceed old age benefits, while making sure that those who become disabled at young ages still receive adequate social protection throughout their lives. ï‚· Survivor pensions need to be provided only at retirement age and withdrawn if remarriage occurs. Survivor pensions, currently provided at the young ages of 45 for women and 60 for men, need to be provided only at the retirement age of 65. They also need to be provided only in the cases when the survivor has not remarried. Furthermore, the benefits could be reduced to 60 percent for a single beneficiary; 70 percent for two; 80 percent for 3; and a maximum of 90 percent for 4 or more. ï‚· Privileged pensions should be separated from the general pension pool, funded separately and aligned with general pensions. The redistribution that took place in the past from the general pension pool to those considered privileged needs to be avoided. The best way to avoid this is to keep these pensions separate and to finance them separately as proposed in the strategy. ï‚· Limit benefits indexation to prices rather than wages. The international best practice is to have increases in pensions linked to inflation to preserve purchasing power not to link to wages which in general adds to pension costs over time. Pensions in Republika Srpska 3.16. Like FBH, RS has a very high system dependency rate of 75.6 percent, indicating that each contributor needs to support 0.76 pensioners. Comparing the dependency rates in RS to those in high income countries, as shown in Figure 3.11, it becomes apparent that the dependency rate exceeds that of countries like Greece, Spain, and Portugal, and is outmatched only by Italy and France. Even when compared to transition countries, RS looks high, outmatched only by Croatia, Ukraine, Albania, and Armenia. The other countries from the former Yugoslavia, Slovenia, Serbia, Macedonia, and Montenegro, all have substantially lower system dependency ratios than RS. 3.17. A large explanatory factor for the high system dependency rate is the low contributor coverage rate, the ratio of contributors as a percentage of the working age population. This percentage, at 29.1 percent, is about the same as the very low value for FBH and is among the lowest in Europe. Compared to high income countries, as shown in Figure 3.12, RS‘s percentage of contributors is clearly an outlier. Even compared to the other transition countries as shown in Figure 3.12, RS‘s low number of contributors is matched only by Albania in Europe and by Armenia and Azerbaijan in the former Soviet Union. 3.18. High spending numbers also derive from the high system dependency ratio. RS‘s spending at 10.3 percent of GDP is surpassed only by Germany, Portugal, Greece, Austria, France, and Italy among high income countries, as shown in Figure 3.13. However, countries like Sweden and Japan spend considerably less relative to their GDP even though they are demographically older countries than RS. Among transition countries, RS‘s spending is also high. RS lies among the highest quartile of spenders among transition countries, being outspent by Croatia, Serbia, and Slovenia from the former Yugoslavia, and Hungary, Estonia, and Ukraine among the non-Yugoslav transition countries. 26 3.19. Despite the high spending numbers, only about two-thirds of those aged 65 and older actually collect a pension. While one might think that the low coverage among the elderly is partly because of lower labor force participation rates among women, who are supported by pensions from their still living husbands, even among men, only 78 percent of men receive a pension of any type. The high dependency ratios are largely driven by the exceptionally large percentage of pensioners below the age of 65. Figure 3.14 shows that a third of the population aged 55-64 is currently receiving a pension, while only 62 percent of the most elderly, those above the age of 85, are receiving pensions. 3.20. As Figure 3.15 shows, more than 40 percent of those receiving pensions in RS are below the age of 65. Almost 30 percent of old age pensioners are below the age of 65 and almost 60 percent of the disabled pensioners are below the age of 65. And as Figure 15 shows, almost half of the family pensioners are below the age of 65. While some of the family pensioners are children and will lose their eligibility for a pension beyond a certain age, more than 60 percent of the family pensioners below the age of 65 are widows or widowers who will receive these pensions for the rest of their lives. Some of the excess of early pensioners can be attributed to the war which led to a higher percentage of disabled and widows than might have otherwise been the case, but more than 15 years have passed since the war‘s end, the numbers continue to be high. And some of the war-related beneficiaries are collecting benefits from their war status in addition to the pension benefits they receive from the pension fund, making them relatively well off in comparison to those elderly who receive nothing. 27 Table 3.3: Pension Indicators for Republika Srpska Indicators Total Men Women Number of 296,326 171,976 124,350 Contributors Number of Beneficiaries Total 224,069 106,998 117,071 Old Age 103,466 70,246 33,220 Disabled 40,380 28,996 11,384 Survivors 80,223 7,756 72,467 System Dependency 75.6% Ratio Expenditures (total in mil KM Total Men Regular Women Regular with % of GDP in parentheses Total 845.8 (10.3%) Old Age 443.2 325.7 117.5 (5.4%) (4.0%) (1.4%) Disabled 155.4 120.5 34.9 (1.9%) (1.5%) (0.4%) Survivors9 247.2 (3.0%) Average Monthly Pension in KM 314.5 347.5 108.5 Old Age 356.9 386.4 294.7 Disabled 320.6 346.2 255.6 Survivors 256.8 218.14 257.1 Contribution Revenues (in mil 588.6 KM with % of GDP in (7.1%) parentheses) Budget Financing (in mil 160.0 KM with % of GDP in (1.9%) parentheses) Average Covered Wage (monthly 689.72 in KM) Contributor Coverage 29.1% (contributors/pop 15-64) Pensioner Coverage 66.1% (pensioners/pop 65+) All Beneficiaries/Pop 112.0% 65+ Source: World Bank, Human Development Social Protection 9 Survivor expenditures were not available by gender. 28 10 15 0 5 Iceland 0.0% 100.0% 20.0% 40.0% 60.0% 80.0% Australia 100.00% 20.00% 40.00% 60.00% 80.00% 0.00% Ireland Canada Bosnia RS United States New Zealand Spain Netherlands Netherlands Canada Ireland Norway Italy Canada Israel Greece United Kingdom France Denmark Transition Countries Denmark Belgium New Zealand United States Ireland Norway Switzerland Germany Finland Luxembourg Luxembourg Austria Japan Sweden Australia Belgium Countries Spain Netherlands Finland United States Spain Income Economies Japan United Kingdom Greece Belgium Portugal Sweden Bosnia RS Sweden Portugal Germany New Zealand Japan Finland Portugal System Dependency Rates in RS Norway Austria Greece Switzerland Pension Spending as % of GDP in High Compared to High Income Countries Contributor Coverage in High Income Austria Iceland Germany France Denmark Bosnia RS Italy Israel Italy Luxembourg France 29 Source: World Bank, Human Development Social Protection. Source: World Bank, Human Development Social Protection. Source: World Bank, Human Development Social Protection. 10.00 15.00 20.00 0.00 5.00 100% 0% 20% 40% 60% 80% Kosovo 0.0% 100.0% 120.0% 20.0% 40.0% 60.0% 80.0% Kyrgyz Republic Armenia Georgia Azerbaijan Armenia Albania Uzbekistan Kazakhstan Bosnia Republika… Kazakhstan Azerbaijan Kyrgyz Republic Latvia Russian… Turkey Kyrgyz Republic Albania Poland Moldova Belarus Turkey Macedonia, FYR Turkey Uzbekistan Montenegro Czech Republic Figure 3.13: Pension Spending in RS is High Relative to Other Countries Romania Serbia Macedonia, FYR Figure 3.12: Contributor Coverage in RS is Among the Lowest in Europe. Bulgaria Ukraine Montenegro Czech Republic Romania Estonia Latvia Kazakhstan Serbia Lithuania Lithuania Countries Russian… Moldova Croatia Slovenia Slovak Republic Poland Romania Macedonia, FYR Hungary Transition Countries Montenegro Bulgaria Slovak Republic Russian… Poland Hungary Bulgaria Belarus Lithuania Azerbaijan Bosnia… Moldova System Dependency Rate in RS Pension Spending as % of GDP in Croatia Slovenia Slovak Republic Compared to Transition Countries Contributor Coverage in Transition Hungary Belarus Bosnia… Estonia Czech Republic Croatia Figure 3.11: Dependency Rates in RS are High Compared to both High Income Countries and Slovenia Estonia Ukraine Serbia Latvia Albania Ukraine Uzbekistan Armenia Figure 3.14: A Large Percentage of Individuals below the Retirement Age are Collecting Pensions. Percentage of Population Receiving a Pension by Age Group 80.0% 70.0% 60.0% 50.0% 40.0% 30.0% 20.0% 10.0% 0.0% 45-54 55-64 65-74 75-84 85+ Source: World Bank, Human Development Social Protection. Figure 3.15: A Substantial Percentage of Pensioners are below the Age of 65, Particularly Widows Pensioners in RS by age Family Pensioners in RS by Age 3% >85 2% <55 <55 75-84 75-84 6% 16% 19% 19% 55-64 27% 55-64 26% 65-74 65-74 36% 46% Source: World Bank, Human Development Social Protection. Parameters of RS pension system and the new draft law 3.21. Given the problems in RS pension system, the Ministry of Labor prepared a new law which was passed by Parliament in November 2011. Table 3.4 summarizes the main changes between the old law and the new law. Since one of the critical problems in RS pension system is the prevalence of early retirement, it would be useful to look at how well the new law addresses that issue. The new law raises the retirement age for women gradually to age 65 from the current practice of age 60, but will not reach that age until 2019. However, it should be noted that less than a third of old age pensioners in RS are currently women, suggesting that this provision will not affect that many people. And of the women collecting pensions, only 28 percent of them collect old age pensions, again suggesting that the new provisions will affect only a limited number of women. The new law also lowers the number of years of service credit required to allow individuals to retire at age 65 from the current 20 to 15, potentially increasing the number of individuals who qualify for old age pensions. Recognizing that there are many elderly without access to pensions, this change was made to allow more of the elderly to access pensions 30 and to contribute to poverty alleviation among the elderly. Individuals with 40/35 years of service credit can retire at age 60 for men and age 58 for women, which is an improvement over the previous law which allowed individuals with this length of service to retire at any age, although these ages will not be fully effective until 2025. However, individuals with special service can retire as young as 55. The average worker who begins work at age 20, without special service, could not have reached 40 years of service before reaching age 60 anyway. For women, this might mean a slightly higher retirement age as they could complete 35 years of contributions by age 55 if starting at age 20 and now must wait until 58. Table 3.4: Parameters of New and Old RS Pension Laws Pension RS Law until 2011 Newly passed RS Law10 Parameters Retirement Age 65 M/60F with 20 years of contribution as of 65 M/65F with 15 years of contribution fully 2006;11 effective as of 2019; Any age with 40M/35F years of contribution 60 M/58 F with 40M/35F years of contribution Accrual rate 2.25 percent for first 20 years; 1.14 percent per year for all years until reaching 40 years of service; 1.5 percent for next 20 years; 0 beyond 40 years; 0 beyond 40 years Pensionable base Net wages earned since 1970, excluding 1992 Wages earned since 1970, excluding 1992 and 1993, revalued with historical average wage and 1993, revalued with historical average growth net wage growth; Indexation post- To average wage growth but subject to To 50 percent of growth in average net retirement revenue constraint salary and 50 percent of inflation, but not lower than inflation with no revenue constraint Eligibility for Contributed for 1/3 of working life from age 20 Contributed for 5 years or achieved 10 disability to age of disability; years of work credit; after age 55, need pension only be diagnosed with reduced work No requirement if work injury; capacity and 5 years of contribution or 10 Defined as inability to work at previous job years of work credit; No requirement if work injury; Defined as inability to work at all; if can work, receive a supplement covering the difference between previous salary and salary at new job. Level of 50 for up to 20 years; Same benefit structure as old age pension; Disability for individuals with less than 15 years of Pension 1.25 percent for subsequent years with pension history, benefits will be calculated maximum of 75 percent; with 15 years of contributions; for those Automatic maximum irrespective of years if with work injury, benefits will be calculated as if there were 40 years of 10 Bold provisions indicate changes from the previous law. 11 Although the law indicates that retirement provisions are equal for men and women, women have the option of retiring at age 60 and at any age with 35 years of contributions, and most do. The benefit formula is the same as for men so that women who retire earlier receive lower benefits on the basis of having accumulated fewer years of service, but there is no actuarial reduction for the earlier retirement provision. 31 work injury work history; Waiting benefit of average net salary for year Waiting benefit of average net salary for preceding disability paid from date of disability year preceding disability paid from date of to day of deployment to a new job; disability to day of deployment to a new job; Professional rehabilitation benefit paid at Professional rehabilitation benefit paid at average net salary for year preceding disability average net salary for year preceding during period of rehabilitation; disability during period of rehabilitation; Lower salary benefit defined as difference Lower salary benefit defined as difference between salary earned in year preceding between salary earned in year preceding disability and salary in new job paid from the disability and salary in new job paid from the day of deployment at new job; day of deployment at new job; Waiting benefit, professional rehabilitation Waiting benefit, professional rehabilitation benefit, and lower salary benefit are all indexed benefit, and lower salary benefit are all with wage growth. indexed with wage growth. Eligibility for Age 45, incapacity, or caring for child in case of Age 50, incapacity, or caring for child in survivor‘s widow; case of widow; should the widow reach age pension 45 at the time of the husband‟s death or Age 55, incapacity, or caring for a child in case while caring for a child subsequently, she of widower; will receive the pension beginning at age Children aged 15 or younger and until 26 if full 50; time students or lifetime if incapacitated Age 60, incapacity, or caring for a child in case of widower; if widower reaches age 55 at the time of wife‟s death or while caring for a child, the pension will be paid beginning at age 60; Divorced spouses are entitled to pensions if supported by court ruling; For those married after the age of 65, family pensions will only be paid if there is a child or if the marriage lasted at least 2 years; Children aged 15 or younger and until 26 if full time students or lifetime if incapacitated Level of 70 percent of contributor‘s entitlement if 1 70 percent of contributor‘s entitlement if 1 survivor‘s person; 80 percent to be divided by 2; 90 person; 80 percent to be divided by 2; 90 pension percent to be divided by 3; and 100 percent to percent to be divided by 3; and 100 percent be divided if 4 or more to be divided if 4 or more Contribution 24 percent of net wage 24 percent of net wage Rates 3.22. The accrual rate, the benefit per year of service, has been established at approximately 1.14 percent of average wage and this is a huge improvement over the previous law where the accrual rate was 2.25 percent for the first 20 years and 1.5 percent subsequently. Using the same accrual rate for most years of service also promotes longer working careers instead of the previous system where an individual could earn a reasonable pension with only 20 years of service. However, the maximum length of service for which points are awarded is 40, giving workers no incentive to work beyond 40 years. A full working career is considered to be 45 years in most high income countries. 32 3.23. Indexation of pensions after retirement has been moved to 50 percent of inflation and 50 percent of average wage growth from the previous 100 percent of average wage growth, but the coefficient which constrained pension expenditures has now been removed. In order to prevent violation of the revenue constraint, previously pensions were fully indexed to wage growth, but not necessarily to today‘s wage growth. For example, in 2011, pensions were indexed fully to 2006 wages. While the new indexation formula is more fiscally sustainable than the previous one and therefore an improvement, it is still not fiscally sustainable, and without the coefficient may in fact result in higher expenditure. It should be noted that the implementation of the coefficient in RS had been considerably different from FBH. In RS, all pensions with full indexation to wage growth until a particular year were calculated, and the sum of these was divided by the available revenue to determine the coefficient. As a result, the coefficient was much smaller in RS than in FBH. A larger portion of the pension expenditures are determined to be financed from the budget and budgetary finance is available to cover these expenditures. 3.24. On disability eligibility, the reforms appear to have made it slightly easier for an individual to be eligible for a pension. An individual need make only 5 years of contributions to be eligible. In the previous law, an individual had to contribute for 1/3 of the time between the age of 20 and his current age. For anyone over the age of 35, the new law makes eligibility easier. The move to define full disability as inability to work at any job and to provide job supplements to those who have to take lower-paying jobs is also an improvement over the previous law and should reduce the number of disabled, gradually over time. On benefits, applying the same benefit structure as for old age pensions will tend to lower the pension levels for those who qualify for disability in the future compared to the old law which paid higher benefits to the disabled, lowering the incentive to qualify for disability. However, a truly disabled person might not be able to live reasonably on the pension that comes from only 15 years of contributions. The smaller benefit was meant to be a disincentive toward fraudulent claims, but there is concern that the pensions might not be adequate for the truly disabled. 3.25. While eligibility conditions have been tightened for survivors, with the age for widows rising from 45 to 50 and for widowers from 55 to 60, survivor pensions are still being provided to people significantly below the retirement age, 15 years early in the case of widows. International best practice suggests that individuals are expected to work until they reach retirement age. It also suggests that survivor pensions should be withdrawn if there is remarriage of the remaining spouse, which is still not a component of RS law. 3.26. One more improvement in the law which does not appear in the table is the guaranteed pensions for war-related beneficiaries. Under the previous system, a minimum of KM 600 monthly salary was used to calculate the pensions. Now pensions will be calculated based on actual salaries earned and this will be applied retroactively, resulting in a reduction of pensions for many in this category. Three categories of war-related beneficiaries will not face reductions in their benefits: (i) participants in the Second World War and their children; (ii) deceased soldiers‘ families; and (iii) war invalids. 33 III. RECOMMENDATIONS FOR RS PENSION REFORM 3.27. While the new law provides a gradually increasing retirement age for women and the move to a point system generates a flatter accrual rate, changes which are all positive, earlier versions of the law moved RS pension law much closer to international best practice. The law which actually passed is unlikely to bring the pension system to full fiscal sustainability and adequacy and will require further changes in the future. Nevertheless, as earlier versions of the law demonstrate, the authorities had proposed stronger versions of the law, but in the political process, some of the proposed changes were removed. Future recommended changes include the following: ï‚· Stricter curbs on early retirement. While raising the retirement age for women to 65 is a good thing, the issue in RS, as in FBH, is that most people are retiring well below that age because of the early retirement provisions, allowing men to retire at age 60 and women at age 58 after 40 and 35 years of service, respectively, with even earlier retirement possible for those with extended service credit. Early retirement needs to be reduced further to improve fiscal sustainability. ï‚· Establishment of actuarial pension reductions for those who do retire early. For those who continue to be able to retire early, pensions should be reduced by the actuarially fair amount of 6 percent per year for every year before the age of 65. It is likely that far fewer individuals will choose to retire early under these circumstances which may automatically reduce early retirement. ï‚· Eliminate double-dipping between war-related benefits and pension fund benefits. Individuals should only be eligible for one publicly funded benefit, either a war-related benefit or a benefit coming from the pension fund. This could be implemented by cross- checking a database between the two types of benefits. ï‚· Potentially revising the disability benefit formula to provide higher benefits for the fewer people who will qualify as disabled. The disability benefit structure has been modified to correspond with the old age benefit structure, but there is a concern that those who become disabled at young ages or with fewer years of service might not receive adequate pensions. Many countries provide benefits to the disabled which cover the years they actually contributed plus some fraction of the years between their age at disability and the retirement age. Thus, the pensions of the disabled tend to be roughly the same as if the person had worked until retirement age and retired with an old age pension. The focus should be on providing the pensions to only the truly disabled, but providing them at a livable level. All of the additional benefits like the low salary benefit given to those who might take a lower paid job, need to be eliminated. People can become unemployed or disabled and be forced to take lower paid jobs, but the government cannot be in the business of compensating everyone. The government needs to focus on preventing poverty, not on this type of compensation. ï‚· Survivor pensions need to be provided only at retirement age and withdrawn if remarriage occurs. Survivor pensions, currently provided at the young ages of 50 for women and 60 for men in the new law, need to be provided only at the retirement age of 65. They also need to be provided only in the cases when the survivor has not remarried. Furthermore, the benefits could be reduced to 60 percent for a single beneficiary; 70 percent for two; 80 percent for 3; and a maximum of 90 percent for 4 or more. 34 IV. NON-CONTRIBUTORY CASH TRANSFERS 3.28. This section discusses the patterns of expenditure on different types of non-contributory cash transfers in BH, and provides a comparative perspective on program spending, design and performance across the six countries in the Western Balkan region12. It also benchmarks spending and BH program performance against the spending and performance of similar programs in other countries in Central and Eastern Europe. These are programs that should respond to critical social needs of the most vulnerable members of society. At present, in BH they fail to do so: their capacity for poverty alleviation or for prevention of chronic and transient poverty is negligible. Moreover, spending on such programs poses a significant fiscal risk—partly because a large share of the cash transfers are not means tested, and are not targeted to those who need them most. The benefit system in BH needs a complete overhaul to improve targeting accuracy, broaden the outreach to, and the coverage of the most poor and vulnerable and improve their access to last resort social assistance, and make spending more equitable and fiscally sustainable. The policy focus should therefore be on: (i) making social assistance equitable and fiscally sustainable by containing the costs of veterans‘ benefits and means testing some of them; (ii) modernizing the design and implementation of the last-resort social assistance to strengthen its crisis response capacity, improve its coverage and targeting; (iii) improving the governance of non-contributory disability benefits, and (iv) limiting the ‗lock in‘ effect of social assistance by removing the potential incentives for long - term stay on cash social assistance and facilitating the transition to employment for those who are able to work. 3.29. Bosnia and Herzegovina is striving to overcome a history of high and fiscally unsustainable spending on non-contributory cash transfers13. In the 2000s, the two entities of BH allocated between 3.4 and 4.1 percent of GDP to four main types of non-contributory cash benefits that are typical for the Western Balkan countries. The non-contributory benefits (social assistance) are those that are financed from public expenditures and not based on social insurance contributions. In BH, there exist the following types of non-contributory benefits: (i) last-resort / minimum-income support social assistance; (ii) disability benefits for civilian disabled and civilian victims of war; (iii) family and child protection benefits; and (iv) benefits for war veterans and for members of families of fallen soldiers. The peak level of spending on them - 4.1 percent of GDP - was reached in 2006, and was stabilized thereafter at a somewhat lower level of 3.9 percent of GDP (see Table 3.5). Nevertheless this spending remained the highest among the Western Balkan countries and the whole Europe and Central Asia (ECA) region. This spending level is more than twice the ECA average spending of 1.7 percent of GDP (see Figure 1). The above allocations refer only to the entity-level cash transfer programs. In addition to them, the cantons in FBH, as well as the municipalities in FBH and RS, are allowed to design, implement and finance own non-contributory benefit programs which complement or duplicate the entity-level ones. When this spending is added to the equation, overall social assistance expenditure in BH could be as high as 7 percent of GDP. 12 Albania, Bosnia and Herzegovina, FYR Macedonia, Kosovo, Montenegro, and Serbia 13 The non-contributory benefits (social assistance) are those that are financed from public expenditures and not based on social insurance contributions. 35 Table 3.5: Expenditure on Non-contributory Cash Transfers in BH, percent of GDP 2008 2011 2002 2003 2004 2005 2006 2007 (actual) 2009 2010 (plan) Civilian Benefits 0.9 1 1 0.8 1.5 1.4 1.58 1.52 1.46 Veterans‟ Benefits 3.1 3 2.8 2.6 2.6 2.6 2.23 2.38 2.44 2.05 Total as % of GDP 4 4 3.8 3.4 4.1 3.9 3.81 3.90 3.90 Memo items: Nominal BH GDP in million BAM 13,736 14,505 15,786 16,928 19,106 21,641 Source: Central Bank of BH. Figure 3.16: Social Assistance Spending in BH and ECA, % of GDP, 2008-10 BiH 10 *Hungary 08 Croatia 08 *EU 08 Romania 08 Russia 08 *Slovenia 08 Ukraine 08 Belarus 08 Albania 08 *Estonia 08 Serbia 08 *Slovakia 08 Lithuania 08 Kosovo 08 Moldova 08 Bulgaria 08 Georgia 08 Montenegro 08 Armenia 08 FYR Macedonia 08 Azerbaijan 08 *Poland 08 Turkey 08 Kyrgyz Republic 08 Latvia 08 Tajikistan 08 0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0% 3.5% 4.0% 4.5% Source: ECA Social Protection database, 2011. 3.30. For historical reasons, the non-contributory cash transfers have been heavily dominated by programs designed to protect veterans and/or their surviving dependents. Meanwhile, the ‗civilian‘ benefits, especially the cash transfers to the poor (last resort social assistance) and the child protection programs, are small and underfinanced, and provide limited and uneven coverage across the two entities. Thus, veteran-related benefits14 in RS absorb about three-quarters of total spending on non-contributory cash transfers, and this share has been fairly steady over the period 2005-2010. In the same period, veteran-related benefits in FBH absorb on average two thirds of the expenditure on the above-mentioned transfers, though this share declined from 82 percent of the 2005 total to 57 percent by 2009. In 2010, the average share of veteran-related benefit expenditure accounted for 62.6 percent of all spending on social assistance in BH. As a share of GDP, veteran-related benefits absorbed around 2.4 percent of GDP in 2008-10 (see Figure 3.17). This is the highest level of spending on these kinds of benefits in ECA, followed by Croatia which spends 1.75 percent of GDP on the same category. At the same time, the expenditures on all kinds of civilian benefits account for around 1.5 percent of GDP, including spending on benefits for civilian disabled and civilian victims of war of around 0.8 percent of GDP, and spending 14 For the ease of reference, all benefit programs that belong to this type transfer will be referred to as ‗veteran - related‘ benefits. 36 on last-resort social assistance for the poor, family and child protection benefits, ad hoc one-time financial assistance altogether of around 0.6-0.7 percent of GDP15. This spending is significantly lower compared to veteran-related expenditure. The differences in spending levels and trends by entity are negligible with respect to veteran benefits but quite significant with respect to the social assistance for the civilian population (see Figure 3.18). Figure 3.17: Dynamics (2002-2010) and Composition (2010) of Social Assistance Spending in BH 9 8 7 6 Total as % of 5 GDP Civilian Veterans‘ Benefits 4 Benefits Veterans‘ 3 Civilian Benefits 2 Benefits 1 0 2003 2002 2004 2005 2006 2007 2008 2009 2010 Source: Administrative data. Figure 3.18: Dynamics of Veteran and Civilian Social Assistance Spending by Entity, million BAM, 2005-2010 Veteran FBH Civilian FBH Veteran RS Civilian RS 2005 2006 2007 2008 2009 2010 2005 2006 2007 2008 2009 2010 Source: Administrative data. 3.31. This very high level of spending on non-contributory benefits is not only fiscally unsustainable but is also economically inefficient and socially inequitable. The high overall cost of social assistance is largely due to heavy reliance on the categorical veteran-related benefits, as opposed to needs-based ones. The means-tested benefits account for around 0.5 percent of GDP, while the remaining expenditure is allocated on a categorical principle to individuals and families which are not necessarily in need. Over time (2000-2010), the composition of spending is changing: the public outlays for categorical 15 This includes 0.19 percent of GDP for last-resort social assistance (Permanent Social Assistance); 0.33 percent of GDP for child and family allowances and 0.07 percent of GDP for other cash and in-kind benefits. 37 veteran benefits are decreasing from a very high basis in 2000-02 (see Figure 3.2) while in many other ECA countries they were growing during the same period. Nevertheless, BH veteran-related benefit spending remains the highest when measured as a share of GDP: the peak level was reached in 2002 when this spending accounted for 3.1 percent of GDP, while the lowest level of 2.23 percent of GDP was registered in 2008. In 2009 and 2010 the spending on veteran-related benefits increased again to 2.44 percent of GDP. The spending on the means-tested last-resort social assistance (the Permanent Social Assistance program) is quite low by regional standards and decreasing though less markedly. In contrast, the spending on civilian disability benefits increased fourfold in nine years– from 0.20 to 0.80 percent of GDP despite the termination of a large part of them (all benefits for people with disability rate below 90 percent in FBH) in 2008, and leaving only the benefits for severe cases (90 and 100 percent disability in FBH). This move triggered fiscal savings in 2009 and later, and radically reduced the number of beneficiaries, however, it also exacerbated the inequality in treatment of beneficiaries based on the reason for disability, and undermined the role of these benefits as an instrument for poverty reduction and social inclusion in case of disability. Figure 3.19: Trends in Spending on Veteran Benefits and Permanent Social Assistance in BH and Other ECA Countries (2000-2010, % of GDP) 3.0% 2.0% 2000 1.8% 2.5% 2005 1.6% 1.4% 2.0% 2008 2009 1.2% 2004 1.5% 1.0% 2004 2009 2000 0.8% 2003 2009 1.0% 0.6% 2009 2000 0.4% 2009 2009 2009 2009 0.5% 2009 2004 2009 2005 2005 2003 2009 2000 0.2% 2000 2009 2009 2000 2009 2008 2004 2000 2009 2000 0.0% 2009 0.0% 2000 Albania Bosnia and FYR Macedonia Kosovo Montenegro Serbia Croatia Romania Albania Bosnia and FYR Kosovo Montenegro Serbia Armenia Bulgaria Croatia Latvia Herzegovina Herzegovina Macedonia Western Balkans Comparators Western Balkans Comparators Source: ECA Social Protection Database, 2011. 3.32. The veteran-related benefits remained largely untouched by reforms and spending cuts in the 2000s. Following the introduction of framework legislation for income testing of non-contributory budget transfers, both entities committed themselves to applying income tests for a wide range of veteran benefits (disability benefits when the disability rate is below 60 percent, benefits for family members of fallen soldiers and disabled veterans except for benefits for children while in full-time education) as of January 1, 2011. This has not happened yet, despite the provisions of the entity-level framework laws. 3.33. In FBH, the most significant downsizing was related to the elimination of the unemployment benefits for demobilized soldiers. These benefits were introduced in October 2006 with a maximum eligibility time frame of three years, and were phased out in 2010. Subsequently they were not replaced with successor benefits, despite pressure to do so. The full fiscal impact of the phasing out will be felt in 2012 since the accumulated arrears were cleared only in 2011. Also in FBH, income testing was introduced only for the monthly cash transfers for medal holders in mid-2010. The eligibility threshold was set at a high level that allowed excluding only a very limited number of recipients – only those with significant personal income from other sources, and incurred modest savings (see Table 3.6). The limited scope of income testing and the elimination of the demobilized soldiers‘ allowance did not affect the overall level of expenditure on veterans‘ benefits in FBH, on the contrary, total spending 38 increased significantly in 2008-2010 due to sustained high level of spending on the main benefit categories – the monthly veteran (military) disability allowances and survivors‘ benefits. 3.34. In RS, the spending on all types of veteran benefits virtually exploded (increased by close to 60 percent) in 2008, and declined somewhat16 thereafter, but still remains significantly higher compared to the pre-2008 period. Similarly to FBH, most of the spending is allocated for military disability benefits and survivor benefits however the share of survivors‘ benefits is substantially higher (see Table 3.7). In 2012, RS introduced a new cash benefit for all veterans who reach 60 years of age. It will put additional - significant and long-term – fiscal pressure. The current costs of this policy and their growth over time had not been estimated before the enforcement of this new veteran protection policy. Table 3.6: Expenditure on Veteran-related Benefits in FBH 2005-2011, million BAM 2005 2006 2007 2008 2009 2010 2011 Military disability allowance 151 151 172.3 137.2 150.1 146.2 134.8 Survivors‘ (family) benefits 150 150 152.8 185.0 184.7 183.0 181.4 Demobilized soldiers‘ allowance 0 0 60.7 0 0 45.4 na Medal holders‘ allowance 0 17 17 na 16.5 15.5 na Other benefits 0 4.5 0 Source: Administrative data Table 3.7: Expenditure on Veteran-related Benefits in RS 2005-2011, million BAM 2005 2006 2007 2008 2009 2010 2011 Military disability and survivor benefit including: 106.6 112.5 130 Military disability allowance 71.7 70.4 67.5 65.0 Survivors‘ (family) benefits 143.6 134.9 127.3 119.9 Special supplement (soldier's payment) 0 0 10 12.0 12.0 12,0 na Medal holders‘ allowance 0 0 0.4 2.0 2.6 na na Other benefits 5.5 5.3 4.8 Source: Administrative data. 3.35. The two entities operate separate veteran-related benefit systems, but the benefit types and spending structures are quite similar. Both entities perceive veteran-related benefits as different from main stream social assistance because of their dual objective: compensation for loss of capacity to work or of bread-winner in the family, and recognition for service. FBH and RS allocate over 90 percent of their entity budget outlays for two main categories of veteran benefits: benefits for disabled veterans and family / survivors‘ benefits. Each category consists of numerous sub-categories, e.g. basic disability benefit which is further differentiated based on the disability rate, and supplement for care to be provided by another person. In 2010, the basic disability benefit is extended to around 54,000 veterans in FBH, divided into 10 categories depending on the level of disability. Similarly, around 36,000 veterans in RS received ten different entitlements depending on acquired disability rate in 2010. The benefits for members of families of fallen soldiers (survivors‘ benefits) are also similar. In 2010, RS has delivered such benefits to around 30,000 people, down from 33,000 in 2008. In FBH, survivors‘ benefits are received by around 42,000 beneficiaries. 16 Data for 2009 and 2010 is not complete, to allow more exact estimate of expenditure downsizing. 39 3.36. The civilian benefits (cash transfers delivered by the Centers for Social Work) also remain largely unreformed, despite the fact that some efforts at reform had been undertaken initially in FBH, and more recently in RS. 3.37. In FBH, the most important reform step was the attempt at radical limitation of access to civilian disability benefits (see Table 3.8). In an effort to control the colossal expansion of spending in 2007 and 2008 (by 4.2 times in 2007 compared to 2006, and by 5.65 times in 2008 compared to 2006), FBH Ministry of Labor and Social Policy limited the eligibility for such benefits only to those who have grave permanent and irreversible disability (disability rate of 90 or 100 percent). However, this policy move increased the inequality of treatment of civilian disabled in relation to war-related disabled while at the same time its fiscal impact remained negligible: in 2010, budget outlays for this expenditure category were reduced by only 13 percent, and they continue to dominate the structure of spending on civilian cash transfers. The tighter eligibility for civilian disability benefits brought about significant savings: the annual cost of the program was KM 98 million in 2010 and KM 108 million in 2011 compared to KM 169.5 million in 2008. However, the overall fiscal impact remained less significant because of the large arrears that remained to be paid as liabilities under the ‗old rules‘. Table 3.8: Expenditure on Non-veteran (Civilian) Benefits in FBH, 2005-2011, million BAM 2005 2006 2007 2008 2009 2010 2011 Social assistance 33.8 31.9 24.7 13.5 11.9 13.5 na Child and family benefits 30.8 33.1 35.8 18.3 17.6 17.3 na Civilian disability benefits 0 30.0 124.8 169.5 152.7 150.1 149.0 Benefits for civilian victims of war 0 9.0 42.3 27.8 26.7 24.8 30.0 Miscellaneous 0 25.0 0 100.0 95.9 90.6 na Source: Administrative data. 3.38. The effect of the radical restriction of eligibility for, and access to civilian disability benefits is yet to be seen. Instead of decreasing, in 2009 the number of recipients of personal disability benefit went up to 92,936 cases from 54,286 and 59,516 in 2007 and 2008 respectively. The same trend was followed by the benefits allocated for care by other person and for orthopedic devices due to slow enforcement and also due to double counting of cases as the recipients have exercised their rights under the old and the new rules throughout the transition period. In 2010, the number of beneficiaries of allowance for care by other person and the personal disability allowance declined to below the pre-reform level. The most recent cantonal level data indicate that in 2011they have had 47,491 recipients of personal disability benefit, 17,776 recipients of orthopedic allowances and 25,434 recipients of allowance for care by other person (see Table 3.9). FBH Ministry of Labor and Social Policy‘s data which is based on the actual payments (transfers to bank accounts) indicate that 42,151 beneficiaries have received bank transfers in 2011 for civilian disability benefit payments. 3.39. In FBH, spending on last-resort social assistance (permanent social assistance) remains negligible, at 4.6 percent of the overall spending on benefits delivered by the Centers for Social Work (CSW) and close to 7 times lower than the spending on one-time social assistance and other ad hoc social assistance. The last-resort social assistance spending contracted significantly in 2008-2010, compared to the mid-2000s. The take up remains limited due to complicated eligibility criteria and low benefit amounts. The spending on child benefits is also very low – at 5.8 percent of total spending. Both programs – last-resort social assistance and child benefits – did not expand during the crisis. 40 3.40. In RS, overall spending on civilian non-contributory cash transfers increased by over 50 percent in 2008-2010, compared to the mid-2000s (see Table 3.9). The monthly child allowance is the main expenditure category accounting for close to 48 percent of all spending, followed by the non- contributory benefits for civilian disabled with 26 percent of overall spending on all benefits delivered by the Centers for Social Work. The child benefits in RS are better funded in relative terms, they account for close to 55 percent of total spending on civilian benefits in 2010. Table 3.9: Last-resort Social Assistance, Child Benefits, Civilian Disability Benefits and Benefits for Civilian Victims of War – Number of Beneficiaries in FBH, 2011 SOCIAL ASSISTANCE BENEFITS FOR CIVILIAN VICTIMS OF WAR CIVILIAN DISABILITY BENEFITS CHILD BENEFITS Monthly Perm anent Other Personal Orthopaed Fam ily Personal Orthopaed personal Care Care Financial person's disability ic disability disability ic Child allow ance cash allow ance allow ance Assistance care benefit allow ance benefits benefit allow ance CANTON assistance UNSKO-SANSKI 486 111 461 30 226 28 1409 3775 1601 2534 0 POSAVSKI 234 0 31 2 8 0 25 342 85 129 0 TUZLANSKI 3320 254 493 126 230 32 1108 11022 3832 4872 14109 ZENICKO-DOBOJSKI 1234 582 323 35 118 14 312 7997 3200 4637 16141 BOSANSKO-PODRINSKI 148 407 117 22 46 8 214 704 326 269 1760 SREDNJOBOSANSKI 811 1730 303 24 180 18 393 4793 1751 2606 1707 HERCEG-NERETVANSKI 828 0 517 57 235 37 460 4827 1640 3074 875 ZAPADNO-HERCEGOVACKI 832 5 56 0 24 8 27 1794 597 733 0 KANTON SARAJEVO 1137 5512 28 0 0 377 0 11398 4270 5841 0 KANTON Br. 10 741 0 67 0 31 4 28 1289 474 739 0 TOTAL 9771 8601 2396 296 1098 526 3976 47941 17776 25434 34592 Source: Cantonal level administrative data. 3.41. Similarly to FBH, the spending on last resort social assistance is very limited – at 7.8 percent of total spending, and decreasing to less than half compared to 2005. Both programs – last resort social assistance and child benefits – did not expand during the crisis and in the immediate post-crisis period. The main reason for this inadequate and declining spending is the low nominal level of the last-resort benefit (BAM 41 per person per month) which has not been updated for quite some time. The number of beneficiaries of permanent social assistance is significantly lower compared to the beneficiaries of one-time (ad hoc) social assistance. A new law will regulate the provision of last resort social assistance, if enacted in 2012. Its provisions can potentially lead to an increase in the nominal value of the permanent / monthly social assistance benefit due to linking it to the growth of the average wage in RS, as well as to a differentiation of benefit levels for different family composition following the introduction of equivalence scales. Table 3.10: Expenditure on Non-veteran (Civilian) Benefits in RS, 2005-2011, million BAM. 2005 2006 2007 2008 2009 2010 2011 Social assistance 10.4 10 7.5 5.1 5.0 4.8 na Assistance for newborn 3.2 4.1 4.0 na Child benefits 25.1 37.5 38.4 26.3 28.5 29.3 na Benefits for civilian disabled 19.2 15.7 16.0 na Benefits for civilian victims of war 4.3 4.4 7.8 6.8 7.2 7.1 6.8 Source: Administrative data. 3.42. Owing to the prevalence of categorical benefits, the overall targeting of the budget allocations on non-contributory cash transfers in BH is poor. The poorest 20 percent of the population receive only 36.8 percent of the total social assistance budget. Such targeting performance of overall 41 social assistance is weak by regional and ECA-wide standards. It is the weakest in the Western Balkan region where the targeting performance of social assistance is impressively good in most cases (Albania, Kosovo, Serbia and Montenegro) with over 60 percent and even close to 80 percent of total spending being transferred to the poorest quintile of the population of the respective country. The targeting performance in BH ranks among the poorest in ECA outperforming only the programs social assistance in Russia and Belarus (see Figure 3.20). Figure 3.20: Targeting Accuracy of Total Non-Contributory Benefits Received by the Poorest Quintile in BH and ECA Comparator Countries (%) Social Assistance Social Assistance Percent of Total Benefits Received by the Poorest Quintile (%) Percent of Total Benefits Received by the Richest Quintile (%) 90 30 80 25 70 60 20 50 15 40 30 10 20 5 10 0 0 Source: ECA Social Protection database, 2011. 3.43. Social assistance spending is characterized by significant elite capture and sizeable targeting errors. Leakage of expenditure on non-contributory cash transfers to the richest quintile or benefit capture by the elite is 16.6 percent in BH17, which is far higher than the average in ECA, and also higher compared to the rest of the Western Balkan countries. In the ECA countries, elite capture of benefits is higher only in Belarus. The elite capture is attributed to the high share of the non-targeted relative to to the poor veteran benefits. Overall the benefits are highly regressive with 26.7 percent going to those in the richest quintile of the population, and less than 15 percent going to the poorest quintile. Elite capture is also attributable , though to a lesser extent, to deficiencies in the targeting design (income and asset tests) and the implementation practices related to the Permanent Social Assistance and child benefits. Finally, the elite capture increases because the disability benefits for civilian disabled and the benefits extended to civilian victims of war are also categorical. 3.44. Even the last-resort social assistance program which is small in scope and means-tested, leaks resources to the non-poor. In relative terms, the best targeted social transfer is the means-tested Permanent Social Assistance program, and the rest of cash benefits delivered by the Centers for Social Work (CSW) at 46 percent of the poorest quintile meaning that even these programs are ‗wasting‘ over half of the budget allocations on families that are not among the poor. The means-tested benefits in BH are less accurately targeted compared to last-resort social assistance programs with similar design.18 Other 17 Most recent data is from the latest HBS, conducted in 2007. 18 The comparison of the permanent social assistance programs in FBH and RS only, with similar last-resort social assistance in ECA is not possible due to data limitations. The 2007 HBS in BH did not collect separately information on incomes from permanent social assistance; all cash transfers through the CSWs were lumped in one category / question. 42 European and Central Asia (ECA) countries, including Western Balkan countries, manage to achieve targeting accuracy of over 70 percent, or even 80 percent (see Figure 3.21). The gaps in targeting performance of the last-resort social assistance in FBH and RS are indicative of design deficiencies leading to inability of the currently applied income and asset test to capture hard to verify incomes, including informal incomes from work, asset use, household agriculture and remittances. These gaps are also indicative of implementation weaknesses arising from limited and delayed data exchange across agencies to verify employment status, family composition and asset ownership, and subsequent overlap of benefit receipt, weak oversight and control mechanisms and over-reliance on document-based (desk) eligibility verification methods. Therefore, there is significant scope for improving the social assistance effectiveness and equity by increasing the share of benefits that are targeted to needs while reducing categorical targeting. There is also room for improving the applied targeting methods so that a higher share of the social assistance budget is captured by the poor, and not by the elite. Figure 3.21: Last-resort Social Assistance – Share of Total Benefit Amount Received by the Poorest and Richest Quintiles in BH and ECA Comparator Countries (%) Last-Resort Social Assistance Percent of Total Benefits Received by the Poorest Quintile (%) 100 90 80 70 60 50 40 30 20 10 0 Source: ECA Social Protection Database, 2011. 3.45. Despite the large budget outlays for all social assistance, the coverage of the poor is very low. According to 2007 House hold Budget Survey (HBS) data, the coverage of the poorest quintile with non-contributory cash transfers was 21 percent - the second lowest in ECA and comparable only with the coverage of social assistance in Montenegro (20 percent). This means that only 4 percent of the poor population (defined as the lowest consumption quintile) receives any kind of non-contributory transfer, in sharp contrast to some other social assistance systems in ECA which manage to cover more than 70 percent of the poorest quintile (Poland, Latvia, Lithuania, Estonia, Bulgaria) or even more (90 percent in Romania and over 95 percent in Hungary as indicated in Figure 3.22). The coverage of the richest quintile is not high; where existent it reflects the categorical nature of the disability benefits. 43 Figure 3.22: Coverage of All Non-Contributory Benefits in BH and ECA Comparator Countries, Poorest and Richest Quintile (%) Social Assistance Social Assistance Coverage of the Poorest Quintile (%) Coverage of the Richest Quintile (%) 100 60 90 50 80 70 40 60 50 30 40 20 30 20 10 10 0 0 Source: ECA Social Protection Database, 2011. 3.46. Apart from the fiscal pressure, the prevalence of categorical social assistance programs in BH may also create labor supply disincentives as well as an increase in informal employment. This is particularly true for the already mentioned unemployment benefits for demobilized soldiers in FBH. In 2004-2006, prior to the introduction of this benefit, the number of unemployed demobilized soldiers registered with the Employment Bureau, was around 60,000. Between September 2006 and December 2007, when the benefit was introduced and taken up, the number of registered unemployed demobilized soldiers rose by 46 percent, and their share in the total number of unemployed grew from 17.6 percent in 2006, to 23.8 percent in 2007 and 24.5 percent in 2008. Equally importantly, the mandatory registration as unemployed which used to be a qualifying condition for the benefit, motivated taking jobs in the informal economy. 3.47. The last-resort social assistance program may also create work disincentives despite its low generosity and the absence of a sizeable poverty reduction impact. As mentioned, the poverty- targeted Permanent Social Assistance reaches a small fraction of the population – 0.37 percent of the poor and non-poor and together with the one-time social assistance covers around 1 percent of it. At the same time, the relative poverty line which is the yardstick usually used in the EU for defining the target groups of poor, is estimated at 11 percent. Along with the low coverage, the social assistance transfers are not generous enough to lift the recipients out of poverty. The average transfer value of overall social assistance represents only 19.5 percent of the minimum wage in BH. By the same measurement, the generosity of the Permanent Social Assistance is even lower – 14.7 percent of the minimum wage. When measured with administrative data, the Permanent Social Assistance transfer value varies between 12.3 and 21.3 percent of the minimum wage in FBH and between 12.8 and 18.8 percent of the minimum wage in RS. Generosity of overall social assistance measured as share of the transfer in post-transfer consumption of the poorest quintile is average for ECA and below the average for the Western Balkan countries (see Figure 3.23). 44 Figure 3.23: Generosity of All Social Assistance and Last-resort Social Assistance in BH, Poorest Quintile (%) Overall Social Assistance Last-Resort Social Assistance Benefits as Percent of Post-Transfer Consumption, Benefits as Percent of Post-Transfer Consumption, Beneficiary Household, Poorest Quintile Beneficiary Household, Poorest Quintile 60 60 50 50 40 40 30 30 20 20 10 10 0 0 Ukraine XP program Armenia FB Prog Latvia GMI + dwelling Serbia MOP Kosovo SA Poland SA benefits Montenegro FMS/MOP BiH CSW Bulgaria GMI Hungary Regular SA Romania GMI FYR Macedonia SFA Georgia TSA Croatia S. Allowance Albania NE Albania 2008 BiH 2007 Croatia 2008 FYR Macedonia 2006 Poland 2008 Hungary 2004 Bulgaria 2007 Armenia 2008 Kosovo 2006/07 Latvia 2008 Montenegro 2009 Moldova 2009 Romania 2009 Ukraine 2008 Serbia 2007 Georgia 2007 Source: ECA Social Protection Database, 2011. 3.48. The potential for work disincentives is mostly due to specific program design features. Permanent Social Assistance is an open-ended benefit with unlimited duration. It requires registration as unemployed with the Employment Bureau which is also taken as proof that the registered unemployed person has no income. In addition, the unemployment registration and subsequent eligibility for social assistance grant access to health insurance. Finally, it is a poverty gap program which pays the difference between the (maximum) income threshold for an applicant family of a particular size, and the family‘s actual income, meaning that any earned supplementary income will trigger proportionate reduction in the received benefit. Over 50 percent of the beneficiaries have stayed on the lists for over 5 years. This is inefficient given that over 50 percent of them are able-bodied and of working age. 3.49. There is also marginal taxation evidence that BH‟s social assistance could create disincentives to work on the formal labor market. The marginal tax rate on earnings has been researched19 in three Western Balkan countries ‒ Bosnia and Herzegovina, Serbia, and FYR Macedonia (see Table 3.11). This research reveals that people working in low-skilled, low-paying jobs have a fairly high tax wedge on their earned incomes, which is likely to give them a disincentive to graduate from social assistance into full-time work. Compared with the situation in the OECD and EU countries for which similar data is computed regularly, the tax wedge in FBH is considerably higher than the median. The gap between the tax wedge on low wages (33 percent of the average wage, which is used as an approximation of a minimum wage) and the full average wage in the Western Balkan countries is narrower than in the comparator countries, which implies that there are fewer incentives to move out of welfare and into work when the job is a low-paying one. 19 Calculated using the OECD Tax-Benefit model. 45 Table 3.11: Tax Wedge at Various Levels of the Average Wage in BH, FYR Macedonia, and Serbia (%, 2008) Bosnia and Herzegovina FYR Macedonia Serbia Tax wedge for single earner with no children 33% of average wage 37.8 (FBH) 31.1 (RS) 28.5 36.7 50% of average wage 39.5 (FBH) 32.8 (RS) 30.9 38.0 100% of average wage 41.8 (FBH) 34.5 (RS) 33.2 39.3 Tax wedge for one-earner couple with two children 33% of average wage 37.8 (FBH) 30.6 (RS) 28.5 36.7 50% of average wage 37.8 (FBH) 30.6 (RS) 30.9 38.0 100% of average wage 37.9 (FBH) 33.0 (RS) 33.2 39.3 Source: OECD Tax and Benefit model Notes: The tax wedge is defined as the share of income tax and social security contributions by employers and employees over total labor costs. The numbers presented in this table refer to a one-earner couple with two children who receives average wage and works 33 or 50 percent part-time or full-time. Alternatively, in most, though not all, countries this can be interpreted also as the tax wedge of a one-earner couple with two children, working full-time, but receiving 33, 50, or 100 percent of average wage. In the latter case, working full time at 33 percent of average wage might be below the legal minimum wage. Values refer to 2009. 3.50. In 2010, the Governments of FBH and RS started a move towards restructuring social assistance expenditures, along with improving the coverage and targeting of the main cash transfer programs. First and foremost, they adopted Framework Laws on Targeting of Cash Benefits to Individuals. They set the stage for a shift from the predominantly rights-based approach to a needs- focused approach in social assistance. The Framework Laws are supposed to strengthen the fiscal responsibility in social protection by: ï‚· Introducing mandatory needs-based targeting of all types of social benefit programs, including benefits for war veterans and for civilian victims of war, and benefits in the case of non-war related disability. This targeting goes beyond the existing income test applied to determine eligibility for last-resort social assistance and child benefits. The new eligibility determination arrangement will take into account the actual material needs of the applicant or beneficiary. The targeting criteria are expected to include, but not be limited to indicators as income, assets, health and disability status, social position, disability level and other characteristics that are of importance for assessing actual material needs; ï‚· Introducing mandatory ex ante assessment of the fiscal impact of any new legislative proposal that incurs social assistance commitment before putting the legislative procedures in motion, and inclusion in any such legislative proposal of provisions for adjusting the size of benefits. The adjustments are to be made in accordance with the budget fund planned in the entity budget for the respective year with special coefficients for reduction of the due benefit. Passing unfunded legislation that opens new commitments (benefits) should be avoided; ï‚· Introducing clear rules for benefit increase and a shift from indexation of the amounts of the non-contributory benefits with wage growth to indexation with the inflation rate, along with limiting the maximum size of the benefit increase to the official rate of inflation; ï‚· Improving the governance and accountability in social assistance through establishment of beneficiary registries, and reinforcing the requirements for reporting and data exchange. 3.51. The Framework Laws however have not been followed by specific laws that „operationalize‟ the targeting to needs. Even worse, RS has adopted a new law on new veteran benefits which are categorical, in breach of the requirements of the Framework Law. The Framework Laws mandate that all 46 types of cash benefits paid to individuals from the entities‘ budgets must be regulated with specific laws that clearly define the criteria for recognizing the rights of individuals to the respective benefits, the purposes of such benefits, as well as the criteria for determining the amount, duration and method of payment. Without specification of the principles of the Framework Laws in subsequent legal acts that govern the design and delivery of all types of non-contributory cash transfers, the objective of improving targeting and making the entity-level social assistance systems more equitable and effective in poverty reduction will not be achieved. The transition ‗from framework to subsequent specific laws‖ was supposed to take place in 2011, or at least to start in 2011, however none of them has been enacted, except for the already mentioned amendments to the law on cash benefits for holders of state decorations (medal holders) in FBH which were enforced in May 2010. The delays with the implementation of the targeting to needs are attributed to the lack of political will for unpopular reforms and the slow government formation after the national elections in early October 2010. Over one year after the elections, the lack of political will to advance the targeting reform persists. 3.52. No significant progress has been made towards improving the equity and fiscal sustainability of non-contributory cash transfers. The most important developments in 2010-2011 relate to the following: ï‚· The RS adopted a new law that regulates the provision of last-resort social assistance. This benefit has been means-tested in the past and in this context, the changes that the new law introduces are mostly parametric. On the positive side, the new law regulates with a greater level of precision the income test, and envisages inclusion of all kinds of incomes for its purposes, including incomes from veteran benefits. The eligibility determination process is left open for further amendments, including for the introduction of a proxy-means test or a hybrid test upon availability of data and capacity in the Centers for Social Work to implement a new targeting methodology. ï‚· In FBH, the post-election entity government was not successful in reviving the Inter- ministerial Task Force on Targeting which was convened by the previous cabinet with the objective of ‗operationalizing‘ the Framework Law. This Task Force‘s main assignment was to elaborate common definitions and concepts that are going to be used in the subsequent specific legislation to ensure consistency of concepts across different legislative acts. The previous cabinet had decided that the consistent approaches and definitions had to be regulated in a separate Law on Income Testing which will be an interim step between the Framework Law and the specific legislative acts. However, the need for such an interim law is questionable: all basic definitions and principles of targeting could be elaborated and agreed upon and ‗translated‘ into legal provisions directly in the specific legislation, which will be less costly and administratively burdensome. The avoidance of adoption of an interim legislative act will also save time. Instead of working on it, the parties could focus on accelerating the parallel and coordinated drafting the separate specific laws. 3.53. The Entity Parliaments enacted legislation mandating the carrying out of eligibility audits for the right-based veterans‘ benefits, and clean the beneficiary registry from those who abuse the system (receive benefits without legal proof of eligibility) and/or receive duplicative benefits. In FBH, under the previous cabinet, the audit procedures were initiated but in a costly, lengthy and bureaucratic manner, without profiling of the beneficiaries and concentrating on the most risky, error and fraud prone cases. As a result, the audits were progressing extremely slowly and have not resulted in savings, increased transparency or restraint of fraudulent behavior. The current cabinet is more committed to proceed with the audits but need to remove certain administrative and legal barriers and to improve the efficiency of the procedure by focusing with priority to sensitive and suspicious cases (profiling) instead of auditing all files. 47 V. RECOMMENDATIONS FOR NON-CONTRIBUTORY CASH TRANSFERS 3.54. In summary, there is no significant progress on the equity and fiscal savings of non- contributory cash transfers. The new entity cabinets have not been able – despite certain efforts – to unblock the path towards implementation of means testing and speeding up of eligibility audits. The following measures are recommended: ï‚· Avoid addition of new benefits, especially for all veterans and survivors. Consolidate the existing benefits, make rules more transparent, and prioritize those which ensure equity and poverty reduction. ï‚· With priority, in 2012, FBH Ministry of Labor and Social Policy revise the current Law on Basic Elements of Social Protection in FBH, or to replace it with new law(s). The cabinet has decided to follow the approach adopted by the previous MLSP leadership which decided in favor of replacement of the old law with four new laws as follows: (i) a Law on Basic Social Welfare and Minimum Social Protection which will legally define the concepts of ‗last-resort social assistance‘ and ‗social services‘, determine the financing responsibilities for social protection at all levels of government, and establish the detailed legal framework for eligibility processes / means-testing of social benefits; (ii) Law on the Basic Rights of People with Disabilities (civilian) and the draft Law on Protection of Civilian Victims of War both stipulate the introduction of means testing for disability benefits, and (iii) Law on Protection of Families with Children that is going to define the design and implementation of the means test for child and maternity allowances. There is some progress with the draft laws, but they are at different stages of preparation, and – more importantly - their elaboration is not well coordinated. The laws regulating disability benefits for civilian disabled and civilian victims of war have been approved by the former FBH cabinet and submitted to the Parliament. They have also been endorsed by the present MLSP and are pending Parliamentary approval. The preparation of the other two laws is lagging behind. This is a highly unwelcome situation, given that the Law on Basic Social Welfare and Minimum Social Protection is the one that defines the overall system of principles, rules and institutions for the design and delivery of non-contributory cash transfers and services, and as such should be drafted first. ï‚· Undertake legal, administrative and fiscal measures to curtail the costs of the veterans‟ benefits by enforcing the legislation which was adopted on paper and scheduled for enactment as of January 2011. This concerns some of the war-related disability and family (survivors‘) benefits: FBH authorities amended the Law on Rights of Defenders and Their Families and, RS authorities amended the Law on the Rights of War Veterans and Families of Deceased Soldiers to apply an income test for the benefits that are received by veterans with disability rate between 20 percent and 60 percent and by family members except for children in full-time education as of January 1, 2011, Since the income test for the above mentioned veterans‘ disability and family benefits was not enforced in the timeframe stipulated in the Laws, this is preventing the saving of 20 percent of the budget allocations for these benefits which were set as the target of this policy change. ï‚· Develop updated tools to improve the targeting of non-contributory benefits for civilians aimed at reaching the poor and most vulnerable. An improved targeting mechanism would help address the issues raised above and support a shift from the politically expedient rights- based approach to a more neutral needs-based approach that enhances benefits for those who are socially and economically vulnerable. 48 ï‚· Develop effective targeting mechanisms for veteran benefits, recognizing their dual role (social protection and war service) and aim at eliminating the rich among benefit recipients. These include replacement of rights-based benefits programs for war veterans with better targeted, means-tested programs that are more cost effective and supportive of the vulnerable. This should be done in a phased manner in step with establishment of databases and information systems necessary for implementation. ï‚· Diversify the support for veterans who would lose cash benefits (due to reforms) by providing employment services, support for small business development, training and re- qualification. ï‚· Introduce design changes in the social assistance programs to eliminate potential work disincentives and enhance the incentive compatibility of the social assistance programs to facilitate the transition to employment. ï‚· Improve the cost-efficiency, transparency and accountability of benefit administration. Improved benefit administration could be achieved through an in-depth functional review of the current registry processes, information flows and data management, institutional roles and responsibilities at all levels of governments with the ultimate objective of establishing entity- level unified registries. Finally, benefit administration will benefit from eligibility audits which are especially critical for the good governance and cost containment of programs which have recently experienced rapid growth of beneficiary numbers.  Step up eligibility audit efforts. 49 CHAPTER 4: PUBLIC SECTOR WAGES I. OVERALL WAGE PATTERNS 4.1. Containing the wage bill will be a key challenge for fiscal management in the coming years. In the past five years, the wage bill rose sharply from 11.1 percent of GDP in 2006 to 13.1 percent in 2011 (after having dipped to 12.9 percent in 2010). For most of this period, the growth in the wage bill has outpaced the growth in revenues (Figure 4.1) and appears increasingly unsustainable in view of the country‘s growth outlook for the coming years. The growth in the wage bill has been due mainly to salary increases, but also due to increases in employment. Contrary to the recommendations of the previous PEIR (2006), the centralization of functions at the state level was not matched by a reduction of government employment at the entity level. Despite the crisis, overall public employment continued to increase in 2009 and 2010, while in RS staff increases continued in 2011 as well.20 As a result, the country‘s wage bill as a share of GDP is one of the highest in Europe (Figures 4.3 and 4.4). Figure 4.1: Annual growth rates (nominal terms) Figure 4.2: Wage bill as % of GDP and total of the wage bill, revenues, and public expenditure public expenditure 25% 35% 20% 30% 15% 25% 20% 10% 15% 5% 10% 0% 2004 2005 2006 2007 2008 2009 2010 5% -5% 0% -10% 2004 2005 2006 2007 2008 2009 2010 revenues total expenditure wage bill % of GDP % of public expenditure Source: BH Central Bank Figure 4.3: Wage Bill as % of GPD Figure 4.4: Wage Bill as % of Total Outlays 14 35 12 30 10 25 8 20 6 15 4 10 2 5 0 0 2007 2008 2009 2010 2007 2008 2009 2010 Source: World Bank staff calculations based on IMF Government Finance Statistics data 20 Very modest staff reductions (of 0.6% and 0.2%) seem to have occurred at FBH central level and the state level respectively only in the first half of 2010 (Figure 4). 51 Figure 4.5: Wage Bill Trends by Entity (in KM Figure 4.6: Wage Bill Trends by Entity (in million) KM million) 4000 1800 3500 1600 3000 1400 1200 2500 1000 2000 800 1500 600 1000 400 500 200 0 0 2004 2005 2006 2007 2008 2009 2010 2011 2004 2005 2006 2007 2008 2009 2010 2011 state FBH RS BD state FBH RS BD Source: IMF based on data from Ministries of Finance 4.2. During the period 2004-2011, the wage bill grew fastest at the state level and in RS, whose shares of the total wage bill increased. During this period, the state wage bill grew by 362 percent in nominal terms and from 8 to 19 percent of the total wage bill; that of the RS grew by 103 percent in nominal terms (including a 35 percent increase in 2007-2008) and its share of the total increased from 26 to 27 percent. During 2004-2011, FBH wage bill grew by 53 percent in nominal terms, while its share of the total wage bill decreased from 67 to 51 percent. Finally, during the period 2004-2011, the share of the Brcko District (BD) in the total wage bill dropped from 4 to 2 percent of the total. 4.3. In 2011 the wage bill grew in both RS and FBH. There was another marked increase in RS wage bill in 2011 as a result of both new hiring and a shift in the tax treatment of public employees‘ wages, whereby the Government guaranteed net public sector wage levels by effectively taking on responsibility for part of employees‘ tax liabilities at a cost of KM 50 for 2011 (RS center only). In 2011 FBH also saw a significant increase in the wage bill, primarily due to the payment of compensation (arrears and interest) to employees who won court cases against FBH Government and/or the cantons for pay cuts imposed in 2009-2010 as part of the austerity measures without observing due legal process. However, these compensation payments have so far been made on an ad hoc basis as court decisions become due for execution. So far there has been no budget appropriation for the payment of this compensation to all employees, for which the total cost is estimated at KM 55 million for FBH center alone. Sarajevo canton is also faced with a compensation bill of similar magnitude. 4.4. At 3.5 percent of the population, public sector employment in BH is not excessively high, but wages in the public sector are high compared to average pay levels. While their share of the population is not high compared to other small countries in Europe and central Asia, on account of low labor force participation and high unemployment in BH, public employees (excluding SOEs and health workers) make up relatively large shares of the labor force (11 percent) and total employment (19 percent) (Figure 4.7). FBH cantons account for the bulk of public employment, followed by RS government, and the state (Figure 4.8). At the same time, public sector salaries have been higher than average levels in the private sector, though private sector salary data may be distorted by the existence of a large informal sector in BH. In response to the crisis, pay levels in the public sector began to decline in the second half of 2009 and in 2010, but rose again in 2011 (Figures 4.9 - 4.10). 52 Figure 4.7: Public Employees (without health Figure 4.8: Total Public Employees sector and SOEs) as % of Population and Labor Force (Q3 2011) 16 140,000 14 120,000 12 state 100,000 FBH funds 10 FBH municipal 8 80,000 FBH cantons 6 60,000 FBH central 4 RS funds 40,000 2 RS municipal 0 20,000 RS central BiH Croatia Slovenia Lithuania Latvia Estonia 0 % of population % of labor force 2008 2009 2010 2011 Source: National statistical offices using Source: Staff estimates based on IMF based on data from LABORSTA classification MoFs (2008-2010); MoFs data for 2011 Figure 4.9: Public Sector Salaries Figure 4.10: Average Gross Monthly Salaries (in as % of average in the Economy KM) 2,000 138% 141% 140% 141% 1,800 115% 1,600 110% 107% 104% 1,400 1,200 1,000 800 600 400 200 0 2008 2009 2010 2011 (3Qs) 2008 2009 2010 2011 (3Q) public admin & police education average in economy public admin, police & social security education Source: BH State Statistical Office 4.5. Pay levels vary at different levels of government and between the civil service and other categories of public employees. Public sector pay policies are not coordinated among levels of government. Nor does any state-level institution collect official payroll data or monitor pay levels for all levels of government in the country. Broadly speaking, the state and FBH entity governments offer comparable salaries, though the state pays more for higher-grade staff. The FBH pays higher salaries than RS – a difference that is more significant than Table 4.1 indicates, given that – unlike the state and FBH – RS employees do not receive allowances. In RS, the separate salary law adopted in 2008 covers different categories of public employees (education, police, public administration), establishing broadly equivalent pay levels for positions that require similar qualifications and responsibility. The FBH Law on Salaries adopted in 2010 covers only the public administration and excludes police and teachers who are employed at the cantonal level. As a result, pay levels vary at the cantonal level. 53 4.6. Different levels of government have Table 4.1: Monthly salary levels prior to austerity not been inclined to coordinate pay policies. measures in KM (without allowances), 2009 While officials at the state level have been in FBH favor of coordinating public sector pay levels State (center) RS with a view to eventual harmonization, the Unqualified worker 548 696 360 overall tendency in salary setting has been Qualified auxiliary competition among different government (e.g. driver) 944 974 450 authorities. Pressures from influential trade Entry level unions, which operate at each level of professional 1,189 1,311 1,080 government, have contributed to this CS director 2,406 1,715 1,980 competition. In this context, comparisons of Top civil servant 2,797 2,500 1,980 salary levels for equivalent posts paid by Sources: RS Law on Salaries, MoF data for the state and different entities could risk fuelling wage bill FBH expansion, as trade unions may be expected to push for equalizing salaries and allowances upwards, putting fiscal pressure on less prosperous government authorities. 4.7. Salaries are generally adequate to attract and retain staff, but remain highly compressed for professional-grade staff. Civil service agencies, which are responsible for the recruitment of civil servants, do not report difficulties in filling vacancies through competitive procedures – with the exception of information and communications technology engineers. On the other hand, salaries and allowances appear excessive at the lowest levels of the pay scales in the state and FBH administrations, far exceeding pay levels in the private sector for equivalent positions. Pay scales remain very compressed – with decompression ratios between the salary of an entry level civil servant (professional associate) and the top civil service grade of 2.4 in the state, 1.9 in FBH, and 1.8 in RS. Much of the decompression that has occurred in the past five years has concerned the top salary grades, which are reserved for elected officials and political appointees. Therefore it cannot be said that decompression has contributed to its ostensible objective of incentivizing staff and career development. 4.8. The salary law of FBH contains some provisions which risk spurring potentially unaffordable increases in the wage bills of the cantons. The law (article 60), which covers the salaries and allowances of civil servants and other government employees (including elected officials, political appointees, auxiliary staff, prison guards, and judicial police, but excluding police, teachers, and health workers), requires that salaries paid by canton administrations be limited to 90 percent of the salaries paid by FBH administration for posts of the same grade. At the same time, the law sets a floor of 80 percent of FBH salaries for cantons in case they fail to reach agreement with the trade unions. The latter provision has provided an argument for unions in lower-income cantons to demand that salaries be increased to 80 percent of the level paid by FBH. It would therefore be advisable to revisit the 80 percent floor, though this is likely to encounter resistance from the unions. On the other hand, there are signs that canton administrations are already seeking to reach the 90 percent ceiling. This would set a dangerous precedent for poorer cantons, which might not be able afford to match these salary levels. II. RECOMMENDATIONS FOR WAGE BILL CONTAINMENT 4.9. The overall level of wages and non-wage benefits is high in BH compared to most other countries. Given this and the uncertain outlook for growth and fiscal resources going forward it is clear that measures are needed to reduce compensation costs. However, the approach taken needs to embody more that simple across-the-board cuts that do not address structural issues in the current patterns of compensation. Moreover, the changes need to be seen as fair and to be underpinned by new 54 institutional arrangements in order for changes to have a lasting effect. With that in view the following measures are recommended: ï‚· Institute meaningful establishment controls (particularly in the state and FBH) and regularly monitor staffing to control staffing and wage bill expenditure. So far neither entity has put effective central controls on the establishment of new posts and the hiring of staff in place allowing budget organizations to define their own staffing requirements with scant regard for fiscal constraints. For example, if all vacant posts in approved organograms were to be filled in FBH administration, the number of public employees would increase by some 2,600. ï‚· Rationalize the large number of currently vacant posts and cut wage bill appropriations accordingly. So far wage bill appropriations have been based on the number of approved (rather than actually filled) posts. On the positive side, budget organizations tend to observe their wage bill appropriations, so cuts in appropriations should succeed in reducing the wage bill. Treasury rigidities actually prevent budget organizations from reallocating surplus funds (e.g. resulting from unfilled vacancies), leading to significant under-execution of wage bill appropriations (e.g. 6.7 percent at the state level in 2009; 6 percent at FBH level). In 2010 and 2011 staffing at the state level and in FBH central government remained flat due to the lack of funds to fill vacancies (owing to FBH compensation payments and the non-adoption of a budget at the state level). In RS, a hiring freeze introduced in late 2010 has not been fully observed, leading to the hiring of 80 new staff in the public administration (in addition to 220 new teachers) in 2011. In December 2011, the Government decided to change the method of appropriating the wage bill by instructing all line ministries to revise their organograms to include only actually filled posts. This measure is expected to ensure that no funds would be released for vacant posts. ï‚· Systematically monitor staffing numbers to ensure the observance of establishment and hiring controls for all categories of public employees. This would facilitate wage bill management. Sound wage bill and staff planning are currently hampered by the lack of a unified register of public employees at the entity level. Not only does such a register not exist at for BH as a whole, but neither RS nor FBH governments have full information about staffing numbers and composition, particularly at the cantonal and municipal levels. Civil service agencies currently monitor only the number of public administration staff, excluding large categories of public employees such as police and teachers.21 Mandating a single government body (e.g. civil service agency, MoF) to regularly monitor staffing and verify the observance of establishment and hiring controls is necessary both for effectiveness of hiring freezes and for sound wage bill management more broadly. In this context, centralizing the payroll function through staff and payroll databases with a unique employee ID system (at least at the entity level) would be a cost- effective investment. 21 Health workers are not part of the public sector wage bill in the BH, as they are paid under separate health insurance funds and their wage costs appear as ‗goods and services‘ in Treasury reports. 55 ï‚· Regularly monitor employment and pay levels in sectors other than the public administration, as these sectors make up the bulk of public employment. At the state level, the military is the largest category of public employment accounting for 44 percent of all staff, followed by the Indirect Figure 4.11: Functional Composition of Public Employment in Taxation Authority (11 RS percent) and the border guard service (10.5 3% percent). Civil servants 12% 4% account for 9.5 percent of 17% total public employment in civil service FBH (entity and cantons). auxiliary staff In RS, civil servants education 10% account for 12 percent of police and judiciary public employment, with municipalities teachers being the major health and social funds category, followed by 54% municipal staff, police and judiciary (Figure 4.11). In contrast to FBH, RS also has a larger SOE sector, Sources: RS Statistical Office; IMF based on Ministries of Finance data which employs some 13,000 staff22 which are not reflected in the wage bill but still result in budget expenditure, as SOEs absorb significant budget subsidies (5 percent of public expenditure in 2009).23 So major cuts in the wage bill could not be achieved by reducing employment and/or pay levels in public administration alone. ï‚· Reduce or eliminate Figure 4.12: Composition of State Wage Bill in 2009 allowances in the state and FBH. This could 2% 2% produce significant 2% taxes and contributions savings. Despite the 2% 7% net salaries adoption of the law on 30% commuting allowance civil service salaries in 4% 2009, FBH continues to housing allowance maintain a large number of separate living allowance allowances (food allowance, commuting meal allowance allowance, family vacation allowance separation allowance, 49% other allowances housing allowance, holiday pay). This situation is replicated at the state level, where Source: BH state external audit report 2010. allowances and various fees (e.g. for attending working groups) made up 26 percent of the wage bill and 37.5 of staff 22 This excludes the 15,000 health workers who work in the public sector, but are financed outside the wage bill (from the health insurance fund under ‗goods and services‘). 23 BH Central Bank data for 2009. 56 compensation (without payroll taxes and social contributions) in 2009; taxes and social contributions made up 38 percent of the wage bill (Figure 4.12). In 2011 the state achieved some savings by cutting back on senior officials‘ discretionary fees for attending committees and working groups. While the monetary levels of allowances are higher in the state administration, their overall cost in FBH is estimated to be higher, as entitlement to some costly allowances (e.g. housing) extends to auxiliary staff (e.g. cleaners, drivers, security guards, catering staff) as well as civil servants.24 By contrast in RS – with the salary legislation adopted in 2008 – most allowances have been absorbed into the base pay. As a result the pay system is more transparent and easier to manage (though the wage bill has actually expanded due to increases in staffing and the preferential tax treatment of public employees introduced in 2011). ï‚· Reduce the cost of various salary increments with a combination for the following measures: o Abolishing discretionary fees (e.g. for attending committees and working groups) by law, as they represent unnecessary salary top-ups for functions that already fall within the officials‘ duties. o Eliminating universal entitlements to allowances (e.g. meal allowance) by absorbing them into base pay, as was done in RS. o Restricting the eligibility for allowances for housing, family separation, and home visits to expatriated officials (e.g. diplomats posted abroad) and limiting it in time (e.g. up to three months) to other officials relocating upon first taking up employment in the public administration). ï‚· Cut other categories of recurrent spending related to staff numbers, such as official travel, rental of buildings, and telecommunications. According to external auditors, office rental costs could reportedly be reduced by using vacant government-owned premises instead, particularly at the state and FBH entity administrations. Savings could also be achieved by reducing travel and telecommunications costs, especially by limiting some agencies‘ practice of covering the mobile phone charges of employees. ï‚· Reduce the number of auxiliary staff by outsourcing these functions or transferring these employees to labor law contracts. According to FBH Civil Service Agency‘s estimate, in FBH (including cantons) auxiliary staff outnumber civil servants by more than 2:1 (at FBH central level auxiliary staff exceed civil servants by 1.2:1). At the state level and in RS, excess employment of auxiliary staff seems less of a problem. However, the impact on the wage bill is significant at the state and FBH levels, as pay levels for these staff exceed average private sector levels for similar positions, especially as these staff benefit from seniority pay and very sizeable allowances, which are not paid in the private sector.25 In addition, these staff benefit from higher job security and a more generous entitlement to severance pay in case of redundancy. Significant efficiency gains without impact on core government functions could therefore be achieved by transferring these staff to ordinary labor law contracts and/or outsourcing some of the functions they are employed in (e.g. catering, cleaning) to the private sector. However, either of these approaches may be expected to encounter resistance from trade unions, where auxiliary staff are strongly represented. 24 Unlike the state, FBH chart of accounts does not include a separate line for allowances and fees, so their exact share of FBH wage bill cannot be identified. 25 At the state level, the monthly total of the allowances to which all staff were entitled (meal allowance, commuting allowance, vacation allowance) amounted to KM 432 (approx $ 293) or 79 percent of an unqualified worker‘s salary (2010 data). 57 ï‚· Increase flexibility in negotiations with public sector trade unions to avoid disorderly wage bill cutbacks in the future. The unions‘ role in negotiating salary levels is enshrined in the legislation of all levels of government. In the past governments‘ avoidance of conflict with the unions has led to wage hikes in times of increased budget revenues and difficulties in adjusting pay in a downturn. An attempt to use a mechanism (in the organic budget law) linking revenue shortfalls to cuts in public employees‘ salaries without prior agreement with the unions has backfired in FBH, as individual employees and unions have successfully taken the central and cantonal governments to court. As a result, FBH government is faced with the prospect of having to retroactively compensate all employees for the pay cuts introduced in 2009-2010. Some cantons, however, managed to reach agreement with unions on pay cuts following the shortfall in revenues in 2009, thereby avoiding such court cases. Their example should be followed by other governments. Negotiations between governments and unions need to be placed on a firm factual basis with a realistic identification of budget limitations and a realistic dialogue on the possible trade-offs between pay cuts and redundancies, which may be necessary during a crisis. Increasing flexibility to enable governments to adjust pay levels in times of crisis without legal action by employees on a mass scale would require changes to the current legal framework in FBH. Such changes would make the procedures for amending or temporarily suspending comprehensive collective agreements, which include the public sector employees, more flexible in times of fiscal pressure. ï‚· Improve coordination in wage bill planning in the Fiscal Council. Setting and observing numerical targets for wage bill expenditure has so far been complicated by the country‘s institutional complexity, which militates against fiscal coordination and discipline. Despite the establishment of the Fiscal Council, which provides a mechanism for fiscal coordination between the entities, the multi-layered structure of BH institutions makes it difficult to conduct meaningful fiscal planning and coordinate governments‘ approaches to spending cuts in times of crisis. The effectiveness of the Fiscal Council with regard to wage bill management has so far been rather limited, not least due to the limited authority of FBH Government to influence the wage bill policies of the cantons. A coordination mechanism at FBH level could be helpful in this regard. 58 CHAPTER 5: HEALTH I. INTRODUCTION 5.1. Bosnia and Herzegovina has made significant progress in many health indicators. For example, infant mortality rates have declined in both entities - Federation of Bosnia and Herzegovina (FBH) and the Republic of Srpska (RS) and the District of Brcko - in recent years (see Figure 1).26 The standardized death rate (SDR) for all causes and all ages has declined in FBH from 863 deaths per 100,000 in 2004/2007 to 795 per 100,000 in 2010.27 In RS, there were 13,082 deaths in 2004, 13,232 deaths in 2006, and 14,146 deaths in 2007 but this number declined to 13,517 in 2010.28 However, health outcomes in Bosnia and Herzegovina (BH) still have room for improvement. While infant mortality rates have declined in both entities, the number of infant deaths remains high in FBH. In 2010, reported infant and child mortality rates were almost twice as high in FBH (7.8 infant deaths per 1,000 live births) as they were in RS (4.3 infant deaths per 1,000 live births). Moreover, like many countries in the region, BH suffers from high rates of mortality and morbidity due to non-communicable diseases (NCDs). Figure 5.1:Infant Mortality Rate (per 1,000 live births) 12 10 FBiH RS 8 Croatia Estonia 6 Montenegro Serbia 4 Slovenia EU-15 2 EU-12 0 2006 2007 2008 2009 2010 Source: FBH Data from Public Health Institute, RS data from RS Institute of Statistics, other data from WHO/Europe Health for All Database 5.2. According to WHO-Euro estimates, cerebrovascular and cardiovascular diseases and lung-related cancers account for 50 percent of all deaths in BH. The leading risk factors associated with these diseases 26 The mortality rates for BH differ greatly depending on the source. The official figures are those reported by the Statistical Offices in each entity, which report only the infant mortality rate. The European Health for All Database at the WHO Regional Office for Europe has no reported infant and child mortality rates for BH for the period 2006 to 2010. The World Bank‘s World Development Indicators report an i nfant mortality rate of 7.5 and a child mortality rate of 8.4, both for 2010. UNDP reports a child mortality rate of 15 for 2008. 27 Institute for Statistics of FBH 28 Institute for Statistics of RS. Demo bulletin 12 and 14 59 (for example, tobacco use, high blood pressure, alcohol consumption, high cholesterol, and obesity) are very prevalent in BH according to population surveys. They are mostly lifestyle-related and will require a strong, multi-sectoral, and coordinated approach to reduce their incidence. A campaign aiming to reduce these risk factors is being implemented in FBH, financed by the Federal Institute of Health Insurance and Reinsurance, including screening for the most preventable types of carcinoma (breast cancer and colorectal cancer) as well as preventive programs provided by family medicine teams. While the RS is also implementing a program for prevention of risk factors and NCDs, significant work is still needed, for example, to enforce free tobacco environment regulation. 5.3. In the last two decades, BH has been passing through a social, demographic and economic transition and has faced the considerable challenge of reducing the gap between the population‘s health needs and financial capacities. Several policy and legal frameworks were adjusted to increase the efficacy and efficiency of the health sector, to strengthen financial sustainability, and/or to improve the quality of health care. RS adopted the Strategy for Primary Health Care in 2006,29 the Strategy for Secondary and Tertiary Health Care in 2007,30 and the Law on Health Care in 2009.31 It also invested about €137.5 million in capital projects between 2006 and 2010 (Table 1). In addition, the government of the RS plans to introduce an amendment to the Health Care law, a law on health insurance with associated bylaws, and is developing a Strategic Plan for Strengthening of the Health System until 2020. The FBH has also developed a new Strategy for Primary Health Care and a Strategy for Health System Reform and has amended the Health Care law. 5.4. Meanwhile, both entities are introducing a new payment system to create incentives for providers to improve sector-wide results. Under this diagnosis-related group system (DRG), hospitals are paid according to the "products" that they provide. With the support of the European Union, the Secondary Health Care Financing Reform in BH Project is training staff in all hospitals that treat patients (19 in FBH, 11 in RS, and one in the Brcko District) in the principles of coding, the development of models for contracting services, the determination of DRG weight coefficients, and DRG units (prices). In the case of FBH, both of these developments (the introduction of the new payment system and the training of hospital staff) are requiring more effort and coordination than in RS because of the entity‘s decentralized health system. FBH will introduce a new per capita payment system in which family medicine (FM) teams will be paid based on the number of patients registered with their teams rather than given a standard salary. Under this system, patients will have the right to enroll with a different team if they are not satisfied with the care that they receive. A similar system is implemented in RS. In addition FBH is designing a payment mechanism to FM teams to promote the provision of a standardized set of prevention and promotion services. 29 Official Gazette of RS, no. 72/06 30 Official Gazette of RS, no. 72/07 31 Official Gazette of RS, no. 106/09 60 Table 5.1: Investments in Capital Projects in the Health Sector Republika Srpska, 2006-2010 Government of RS BAM 24,739,242 RS Development Program BAM 45,914,173 Ministry of Health and Social Welfare of RS BAM 9,766,385 RS Health Insurance Fund BAM 22,640,905 Credit assets (Projects HSEP, SITAP , Koreja I, and Koreja II) BAM 71,563,999 Private partners (Public-private partnership model) BAM 55,992,616 Donations BAM 35,349,529 Other sources BAM 9,045,448 TOTAL BAM 275,012,297 Federation of Bosnia and Herzegovina, 2010* Government of FBH BAM 27,977,409 Other own resources BAM 29,124,660 Credit assets BAM 4,371,529 Donations BAM 22,023,240 TOTAL* BAM 83,496,838 Note: Investments in FBH include only 2010 figures 5.5. In recent years, this investment in the family medicine model has been the single largest reform implemented by the government in the health sector. In FBH, more than 75 percent of service users expressed their satisfaction with the approach of family medicine teams, as well as with the improvement in health care quality following the introduction of family medicine.32 Some of the most significant results have been the education of medical doctors and nurses through the additional training program and the FM specialization program and the founding of family medicine faculty departments and educational centers that are now operating independently and are implementing training programs. Through this successful program, one-half of the family medicine teams have been established so far. Nevertheless, there are still evident differences in ratio of teams per number of inhabitants between urban and rural areas as well as among cantons in FBH. Therefore, continued training in family medicine shall be necessary in the forthcoming period as well as improving the planning of resources in the PHC. Another positive outcome has been the up-to-date equipment installed in family medicine clinics. A register of patients who have chosen to use family medicine doctors has been initiated but is taking more time than planned in some cantons. Many clinical guidelines and pathways aimed at improving the quality of health services have been developed and are being applied. Nevertheless, the implementation of the PHC reform is hampered by a lack of information management and connectivity in the health system. In response to this problem, FBH is launching an IT program to define standards for different IT applications used within the health sector and has passed legislation in the area of health records to standardize data reporting. Regardless of the fact that a new payment mechanism has been implemented in FBH (capitation for family medicine teams), the process of reforming health financing is still slow, and this may have a negative effect on the success of the implementation of family medicine. Therefore, it is necessary to start implementing the unified methodology for contracting health services as soon as possible, primarily with family medicine teams. Preparations are ongoing under the HSEP Additional Financing to test incentive payments for family medicine teams for providing the standardized set of preventive and promotional services. The intention is to standardize the preventive and promotional work of family medicine teams while at the same time improving the quality of those services. RS has also been upgrading the clinical skills of its family medicine teams. It has completed the retraining program of its existing family medicine doctors, family medicine specializations are now regularly provided, and it is focusing on enhancing the role played by nurses as a vital component of the primary health care delivery system. The infrastructure and medical equipment available in family medicine clinics has significantly 32 Evaluation of PHC Reform in FBH 61 improved and patients are now registered with family medicine teams. In addition in RS an IT application is being used to support the activities of the FM teams and the management of the FM clinics. RS has completed the first phase of a scheme to accredit FM teams to improve the quality of health care and is ready to launch the second and final phase. 5.6. Health professionals are now being given health management training in both entities at the primary, secondary, and tertiary levels. As a contribution to the definition of strategies and polities in the health sector, a system of monitoring and evaluation is being developed by the relevant institutions in FBH with the objective of improving data collection, analysis, and reporting. This has been reinforced by the recently adopted law on health records in FBH, the implementation of which needs to begin as soon as the necessary procedural regulations have been developed. However, additional reforms are necessary in the areas of health financing, quality improvements, service delivery, and governance to improve health outcomes and increase financial protection. 5.7. One underlying cause of the high cost and poor performance is the highly fragmented nature of the health systems in BH. The Dayton Peace Accords stipulated that the entities should be responsible for health regulation, organization, financing, and service delivery. In FBH, the responsibility for health services has been further delegated to the cantons, so the FBH health sector includes the FBH Ministry of Health (MoH), the 10 cantonal MoHs, the Federal Solidarity Health Insurance Fund (HIF), the 10 cantonal HIFs, and 11 Institutes of Public Health (IPH). The RS health system is centralized at the entity level, so responsibility is shared only between the Ministry of Health and Social Welfare of the Republic of Srpska (RS MoH), a single health insurance fund (the Health Insurance Fund of the Republic of Srpska - HIFRS), and a single Public Health Institute (RS PHI). The District of Brcko also has a Department of Health and a separate HIF. As a result, BH has 13 MoHs, HIFs, and IPHs, which has led to substantial duplication and inefficiency. While it will be politically challenging to reduce this fragmentation and diffusion of authority within the health systems within BH, it would increase efficiency and improve both service delivery and health outcomes. 5.8. The World Bank Public Expenditure and Institutional Review in 2006 provided several recommendations for reducing fragmentation and increasing efficiency, and many of the recommendations are still pertinent. The PEIR noted that access to health care was inequitable and involved high out-of-pocket payments. Pharmaceutical spending was high, primarily out-of-pocket, and there was little transparency in drug pricing and procurement. The PEIR made specific recommendations for reforming health insurance, improving governance, rationalizing hospital services and pharmaceutical procurement, increasing the efficiency of care, and making access to health care more equitable. Unfortunately, these reforms have not been fully implemented. 5.9. This chapter will review and analyze health expenditure patterns in BH and then make some specific recommendations. The chapter is structured as follows. Section 2 provides an overview of health expenditures in BH, including the major sources of funds and the main spending categories. Section 3 focuses on the efficiency or ―value for money‖ of these health expenditures. Section 4 examines the impact of government health spending on the degree to which households are protected against high out-of-pocket spending. Based on this analysis, Section 5 presents recommendations on how best to reform the health systems in BH. 62 II. OVERVIEW OF HEALTH SPENDING 5.10. Total health expenditures in BH have continued a growing trend. Health expenditures, both public and private, as a share of GDP in neighboring countries such as Croatia and Serbia have remained relatively flat since 2000, but in BH expenditures have grown substantially during this period (Table 2). Total health expenditures are now estimated to be about 10.9 percent of GDP (having risen from 7.1 percent in 2000 and 9.4 percent in 2006), a level similar to EU-15 countries but higher than many of BH‘s neighboring countries. The growth in spending in the early 2000s was necessary since health spending had been under-funded for much of the previous period. This spending growth has slowed in recent years, especially in the case of public spending. Table 5.2: Total Health Expenditure as a Percentage of GDP, 2000-2009 Years 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 BH 7.1 7.22 7.08 8.04 9.16 9.12 9.36 9.78 10.32 10.94 Croatia 7.82 7.3 6.3 6.5 6.7 7.06 7.16 7.64 7.84 7.84 Estonia 5.3 4.86 4.86 5 5 5 5.02 5.24 6.12 7 Montenegro 7.9 8.86 8.98 9.56 9.08 8.74 8.58 8.12 8.38 9.26 Serbia 7.4 7.58 8.86 8.8 8.66 9 9.32 10.12 10.04 9.92 Slovenia 8.32 8.62 8.64 8.64 8.4 8.42 8.24 7.8 8.3 9.06 EU-15 8.67 8.86 9.08 9.41 9.48 9.57 9.57 9.53 9.78 10.5 EU-12 5.89 6.17 6.53 6.57 6.48 6.53 6.38 6.4 6.76 6.94 CIS 5.4 5.54 5.8 5.77 5.51 5.58 5.58 5.55 5.22 5.76 Source: WHO Euro estimates, 2010 data not available, WHO/Europe Health for All Database 5.11. Both public and private health expenditures have grown. About 60 percent of total health expenditures are paid for publicly. The other portion (41.8 percent) is financed by households through out-of-pocket payments.33 The proportion of private health expenditures remains extremely high, even in comparison with former CIS countries (Figure 2). Public Health spending as a share of GDP in the Republic of Srpska is below the EU average, and health-related public expenses as a percentage of GDP in RS amounted to 3.4 percent in 2007, while in 2010 they amounted to 4.5 percent. Figure 5.2: Private Households' Out-of-pocket Health Payments as a Percentage of Total Health Expenditure, 2009 CIS 37.5 EU-12 22.17 EU-15 13.91 Slovenia 12.24 Serbia 34.54 Montenegro 25.02 Estonia 20.32 Croatia 14.52 BH 38.66 Source: WHO Euro estimates, 2010 data not available, WHO/Europe Health for All Database 33 RS has enacted several reforms recently to reduce formal out-of-pocket payments including introducing contracts with private institutions for consultative-specialist care, reducing wait times and payments for diagnostics, and reducing co-payments for medicines. Unfortunately, there is no breakdown of private health expenditures as a percentage of total health expenditures by entity. 63 5.12. Hospital care continues to consume the largest portion of HIF spending. Despite various reforms including the introduction of family medicine, the spending allocations for different health functions have not changed much in the last five years (Table 3). Including drug costs, primary health care (PHC) spending accounted for approximately 38 percent of spending in RS and 46 percent in FBH (PHC, specialist and drugs spending combined) in 2009. Inpatient spending, including dialysis, chemotherapy, and treatment abroad, has remained high in both entities, accounting for 43 percent of HIF expenditures in FBH and 36 percent of HIFRS expenditures in RS. Officials in RS should continue to pay attention to health care costs abroad, which have declined from 11 percent to 6 percent of HIFRS spending but remain high.34 Table 5.3: Share of HIF Expenditures by Functional Categories by Entity, 2006-2009 2006 2007 2008 2009 FBH PHC 22% 22% 22% 22% Drugs 15% 14% 13% 15% Specialist 9% 9% 9% 9% Hospital 44% 44% 43% 43% Treatment abroad 1% 1% 1% 1% Sick leave 3% 3% 3% 3% Other 6% 7% 8% 6% RS PHC 18% 24% 23% 24% Drugs 9% 12% 12% 13% Hospital 26% 36% 36% 37% Dialysis 7% 7% 6% 6% Chemotherapy 2% 3% 3% 4% Treatment abroad 8% 9% 7% 6% Sick leave 1% 1% 2% 2% Other 29% 8% 10% 8% Source: FBH and RS HIFs 5.13. Sick leave spending is an area that still needs reform. While the share of sick leave expenditures is not large in comparison to other countries and EU average (for example, Serbia spends 4 percent of its health expenditure on sick leave), the amount spent on sick leave has been increasing in both entities over the last five years. Sick leave spending increased in RS from KM 2.9 million in 2005 to KM 10.2 million KM in 2009. This sudden increase of sick leave expenditures took place in 2009 when the law on health insurance was changed so that the amount of sick leave paid for by employers was reduced from 120 to 30 days, with the HIF then paying the additional expenses. While the HIFRS has taken positive steps to reduce sick leave expenses by reducing the percentage of salary paid by the HIF from 90 percent to 70 percent, thus preventing an even higher increase in that type of costs, the government should continue to monitor sick leave expenditures closely. 5.14. Social health insurance contributions are high in both entities. As is the case in most other countries in the region, in BH public health insurance is almost entirely funded by contributions collected from employers and from the salaries of employees. Over 50 percent of those who receive health insurance coverage (pensioners, the unemployed, the disabled, and war veterans) are exempt from paying personal contributions, and their health insurance is financed by transfers from other extra budgetary funds and from general revenues (Table 4). As a result, contribution rates remain high. RS has recently 34 The level of spending for treatment abroad has remained constant, from 25.1 million KM in 2006 to 23.5 million KM in 2010. 64 increased its contribution rate from 11.5 percent to 12.5 percent whereas the FBH contribution rate is 16.5 percent. The FBH contribution rate is higher than other European countries such as Serbia (12.3 percent of gross wages), France (13.6 percent of gross wages), and Bulgaria (8 percent of gross wages). The contribution rates for pensioners in both entities are extremely low (2 percent in the RS and 1.2 percent in the FBH), which further exacerbates the problem as this means that the employed have to shoulder the overwhelming burden of contributing to HIF income. The last PEIR by the World Bank recommended reviewing the contribution exemptions with a view towards increasing collections and broadening the risk pool. The high tax rates on labor are barriers to the creation of more formal sector jobs, which would then increase the pool of contributors. Table 5.4: HIF Membership by Enrollment Category, 2009 (in thousands) FBH RS Category Number % Number % Total members insured 1,980 1,026 Contributions 100% 100% Employees 870 44% 260 39% Farmers 22 1% 13 2% Self-employed 11 1% Pensioners 501 25% 221 33% Unemployed 396 20% 133 20% Disabled, veterans 77 4% 6 1% Foreign insured 65 3% 31 5% Other (refugees, displaced) 39 2% 1 0% Source: FBH and RS HIF Annual Reports Note: RS is for active insured only, does not include family members III. EFFICIENCY OF HEALTH SPENDING 5.15. As already mentioned, both entities are introducing performance-oriented payment systems. The FBH is also moving to a results-oriented approach at the primary care level. While work is still being done to fully implement the DRG system (staff are still being trained, the IT system is being developed, and data being collected to update and cost the epidemiological profile), both entities are well on track in this process. 5.16. However, because of the fragmentation of responsibility for health care in FBH, the sector does not operate efficiently. Table 5 shows that public per capita health expenditures vary greatly between the entities. The HIFs in FBH spend approximately double the amount per capita than the HIFRS even though there has recently been a substantial increase in public health spending in RS (Table 4). Although the health sectors in the two entities differ, the spending difference suggests that significant inefficiency must exist in FBH since health outcomes in the two entities are comparable. Health spending in the RS appears to be more efficient, probably as a result of the introduction of the service delivery and health financing reforms. 65 Table 5.5: Public Health Expenditures, 2006-2010 (unadjusted for inflation) 2006 2007 2008 2009 2010 FBH Total Health Expenditures (in million KM) 1004.40 1183.00 1449.42 1531.24 1484.06 Total Population 2325018 2328359 2327195 2327318 2338000 Health expenditures per capita 432.0 508.1 622.8 657.9 634.8 RS Total Health Expenditures (in million KM) 334.60 345.70 418.09 498.21 491.75 Total Population 1443709 1439673 1437477 1435179 1433038 Health expenditures per capita 231.8 240.1 290.9 347.1 343.02 Source: FBH and RS HIFs, RS population from Institute of Statistics 5.17. Health spending across cantons varies greatly, resulting in less than optimal spending on primary care. Table 6 shows per capita expenditures for select functional categories.35 PHC expenditures are approximately twice as high in Herzegovina-Neretva and West Herzegovina as they are in Una-Sana, Canton 10, and Posavina. Drug expenditures also vary widely, with Tuzla canton having the highest per capita spending (excluding Sarajevo) and Herzegovina-Neretva the lowest. There are also wide variations in sick leave and in the cost of treatment abroad, variations that reflect differences in usage, enforcement, and benefits between the HIFs. Table 5.6: Per Capita Public Health Expenditures by Canton, 2009 (in 2006 KM) Total Sick Treatment Cantons PHC Specialist Hospital Drugs health leave abroad Sarajevo 645.59 95.91 68.76 271.48 140.12 15.22 3.19 Herzegovina-Neretva 377.91 121.35 52.71 131.76 24.95 4.41 5.27 West Herzegovina 368.97 163.02 73.79 22.37 45.15 3.64 26.01 Bosnian Podrinje 353.64 84.47 16.24 165.17 53.13 10.92 - Tuzla 329.04 94.81 42.69 96.62 67.69 13.28 0.36 Zenica-Doboj 299.75 81.36 23.67 116.64 39.14 22.74 0.64 Posavina 291.59 76.19 - 126.26 22.63 3.16 37.32 Central Bosnia 281.64 95.37 - 125.98 34.75 8.68 0.61 Canton 10 256.47 65.67 16.85 102.26 28.82 2.66 15.33 Una-Sana 241.53 61.34 27.69 70.87 38.08 5.56 0.90 Source: Cantonal HIFs 5.18. There continue to be too many non-medical staff in FBH. Wages account for a large proportion of all health expenditures (40 percent in FBH and about 48 percent in RS), and a significant proportion of this expenditure seems to be on non-medical staff in health facilities (Table 7). FBH has been more successful in reducing its non-medical staff, as their numbers declined by 30 percent between 2004 and 2010, compared to only a 3 percent reduction in RS in the same period. The share of non- medical staff in FBH remains higher than in Serbia (26 percent) and the UK (13 percent). The numbers of medical staff, on the other hand, remain low in comparison to other European countries. There are 145 physicians per 100,000 population in RS compared with 183 in FBH, both lower than in Croatia (266), Serbia (271), and Slovenia (238) and considerably lower than in the older EU countries (an average of 344). Another staffing issue worth mentioning is that health care institutions are often subject to political pressure to hire new staff regardless of whether they are needed or not. 35 Sarajevo canton is somewhat of an outlier because it often provides inpatient services for patients from many other cantons. 66 Table 5.7: Medical and Non-medical Health Workers, 2004 and 2010 RS FBH 2004 2010 2004 2010 Total Medical 7,889 9,026 15,727 19,362 Total Non-medical 3,578 3,816 7,973 6,243 Total Personnel 11,467 12,842 23,700 25,606 Share of Non-medical 31% 30% 34% 24.3% Source: IPHs of FBH and RS 5.19. The recent PHC reforms have Table 5.8: Average Length of Stay in Acute Care increased efficiency at the primary care level. Hospitals, in Days 2006-2009 In RS, a variety of PHC reforms including the 2006 2007 2008 2009 introduction of capitation and of family RS 8.0 7.7 7.7 7.0 medicine has increased the number of FBH 9.3 9.2 9.0 8.7 examinations done every day by PHC Croatia 7.6 7.5 7.3 7.4 physicians from 14.3 in 2008 to 24.6 in 2011. In Estonia 5.9 6.0 5.8 5.6 FBH, examinations also increased from 28 per Montenegro 7.8 7.3 6.7 6.9 day in 2008 to 30 per day in 2010.36 Although Serbia 7.8 7.7 7.4 these figures only measure output and do not Slovenia 5.8 5.5 5.7 5.6 reflect quality, this increase in productivity is a EU-15 6.9 6.8 6.7 6.6 EU-12 6.9 6.7 6.5 6.4 positive development. The rates are now similar CIS 11.0 11.0 10.8 10.8 to the level in Serbia, where PHC physicians Sources: The IPHs of FBH and RS, WHO/Europe Health perform 26.7 examinations per day.37 for All Database for other countries 5.20. In FBH, improvements are still needed at the hospital level to reduce patients‟ average length of stay (ALOS) and to increase the Table 5.9: Bed Occupancy Rate in Acute bed occupancy rate (see Table 8). According to data Care Hospitals, 2006-2009 from the IPHs, the ALOS has stayed constant in FBH 2006 2007 2008 2009 at around [9 days] but has declined substantially in the RS 90 87 91 86 FBH 69 68 70 67 RS from 8.0 days in 2006 to 5.9 days in 2011. The Croatia 87 86 85 83 DRG payment system is being introduced in RS since Estonia 71 72 70 68 2011 at acute care hospitals has given hospitals a Montenegro 66 68 67 70 disincentive to extend patient stays. Similarly, there is Serbia 67 70 70 room to improve bed occupancy rates in FBH, which Slovenia 72 70 72 71 have stayed more or less constant between 2006 and EU-15 76 76 76 2009 (69 percent to 67 percent) (see Table 9). Bed EU-12 73 72 74 73 occupancy rates are much higher in RS (86 percent). CIS 84 85 85 85 Sources: The IPHs of FBH and RS, WHO/Europe Health for All Database for other countries 36 Data from the IPHs of RS and FBH 37 World Bank (2009), Baseline Survey on Cost and Efficiency in Primary Health Care Centers before Provider Payment Reforms, Serbia, Jan 26 67 IV. HEALTH SPENDING AND FINANCIAL PROTECTION 5.21. One of the primary objectives of a health system should be to protect households from having to make high out-of-pocket payments for health care. Patients should not have to face medical expenditures that are either impoverishing (driving them below the poverty line) or catastrophic (exceeding some threshold of household resources). Data on out-of-pocket payments are typically collected in household surveys, which reveal the actual amounts that households spend on medical care, drugs, dental care, and other health services. Unfortunately, the most recent household survey data on health expenditures were collected in the Household Budget Survey (HBS) that was conducted in 2007.38 The Ministries of Health in both entities are concerned that the data from the 2007 HBS do not accurately reflect actual out-of-pocket spending. Nevertheless, the statistics on out-of-pocket spending from the 2007 HBS survey are similar to the WHO Euro estimations for 2009, in other words, private out- of-pocket payments are equivalent to about 40 percent of total health spending (see Figure 2). While out- of-pocket payments are still a big proportion of all health expenditures, BH has achieved the greatest reduction in these payments in the region (Figure 3) Figure 5.3: Changes in Share of Out-of-pocket Health Care Payments, 1997-2010 BH Source: WHO 5.22. As stated in the HBS report, in 2007, out-of-pocket payments accounted for 4.4 percent of total household expenditures (per adult equivalent) in BH. The poor did not appear to pay disproportionately more for health care as the share of health expenditures was fairly consistent across quintiles (see Figure 4). In RS, the lowest two quintiles appeared to pay slightly more out of pocket than households in the higher quintiles, but the difference was small. Comparing the data from the 2007 HBS with those from the previous HBS in 2004, household out-of-pocket payments have not changed significantly, though there was a slight increase in the share of household expenditures allocated to health 38 The 2011 HBS report will be available in the fall of 2012. 68 care in BH overall (from 3.7 percent to 4.0 percent).39 Figure 5 shows that this small increase was consistent in both FBH and RS. Figure 5.4: Out-of-pocket Payments on Health as a Percentage of Total Household Expenditures, by Quintile and Equity 6.0 5.6 5.2 4.9 4.9 4.9 5.0 4.1 4.1 4.2 4.1 3.9 4.0 3.0 FBiH RS 2.0 1.0 - 1 (Poorest) 2 3 4 5 (Richest) Source: Bosnia HBS 2007, authors‘ calculations Note: Expenditures per adult equivalent Figure 5.5: Out-of-pocket Payments on Health as a Percentage of Total Household Expenditures, 2004 and 2007 6.0 4.8 5.0 4.4 4.0 4.0 3.6 3.7 3.4 3.0 2004 2007 2.0 1.0 - Federation of BH Republika Srpska BiH Source: Bosnia HBS 2004 and 2007, BHAS 5.23. Although the share of household expenditures on health was similar in poor and rich households, the amounts varied greatly. In FBH, total adult health expenditures in the poorest quintile amounted to only KM 12.0 whereas the richest households spent KM 66.2. Similarly in RS, poor households spent KM 14.0 on health care, whereas the richest households spent KM 68.0. Therefore, although the impact of health payments on household income was similar in both poor and rich households, rich households seem to be accessing and using more and better quality health care. 39 Following the methodology of BHAS, the expenditure categories are consistent for 2004 and 2007, so the 2007 consumption expenditure for health excludes informal health payments. 69 5.24. Drugs accounted for a substantial portion (30 percent) of household out-of-pocket payments in BH (see Figure 6). In FBH, drug spending accounted for a higher share of these expenses (32.8 percent) than in RS (27.1 percent) and was almost quadruple the amount spent on inpatient expenses and about six times more than was spent on outpatient expenses. On the other hand, out-of-pocket payment shares in RS for inpatient and outpatient expenses were higher (20.4 percent and 9.5 percent, respectively) than in FBH. The average adult in RS spent KM 13.2 on hospital expenditures and KM 6.2 on outpatient expenditures compared to KM 5.2 and KM 3.3 spent by the average adult in FBH. Surprisingly, expenditures on medical products were the largest portion of all health expenses in both entities (42.9 percent in FBH and 34.3 percent in RS), but it is likely that these expenses were slightly overestimated in the HBS questionnaire.40 Figure 5.6: Out-of-pocket Payments on Health as a Percentage of Total Household Out-of-pocket Payments in the Previous Month, by Category Medical products Medicines Hospital Outpatient RS FBiH Lab tests Dentistry Other 0% 10% 20% 30% 40% 50% Source: Bosnia HBS 2007, authors‘ calculations 5.25. Health expenditures are impoverishing a sizeable number of households in BH. Figure 7 shows the households that fall below the poverty line because of their health expenditures in FBH and RS. Households were ranked by total consumption (along the curved line), while the amount of out-of- pocket health payments is represented in the vertical red lines. The poverty line is the dashed black line (KM 385.7),41 and the figure shows that many households had large out-of-pocket payments that resulted in them falling below the poverty line. In RS, out-of-pocket payments appear to be higher for households between 0.2 and 0.7 of the distribution (along the horizontal axis), suggesting that out-of-pocket payments on health care may make households in RS more vulnerable to poverty than those in FBH. In fact, the poverty rate in RS increases from 21.7 percent to 25.2 percent when it includes households impoverished by health expenditures. In FBH, the poverty rate of 16.3 percent increases to 18.4 percent when it includes households impoverished by health expenditures. 40 The diary consumption module contained questions about only two categories of health expenditures (medicines and medical aids/products), and these categories accounted for the highest proportions of total health expenditures. Medical products mentioned in the diary module included bandages, syringes, thermometers, adhesive bandages, elastic stockings, condoms, and mechanical contraceptive devices. 41 The poverty line used is the same as by BHAS —60 percent of median monthly consumption expenditure in BH, 385.71 KM. 70 Figure 5.7: Effect of Out-of-pocket Health Care Payments on Household Consumption FBH RS 5.26. Due to the fragmented health system in FBH, households in certain cantons may be less affected by out-of-pocket expenditures on health care than others. The 2007 HBS showed the wide variation in the share of health spending in total expenditures ranged from 1.9 percent in Bosnian Podrinje to 5.4 percent in Central Bosnia (Table 10). The amounts spent on health and specifically on drugs varied widely as well, providing further evidence of inequity in the health system. In 2008 the FBH adopted a List of Essential Drugs with fixed referential prices, which before the list had generally been higher than the prices of the same medicines in the other countries of the region. However, many cantons were unable to ensure the accessibility of all the medicines from this essential list in line with the defined referential prices of medicines, which resulted in unequal availability of medicines across FBH. A study conducted in 2011 to compare the prices of the medicines under the Decision on the List of Essential Medicines of FBH compared the most frequently prescribed medicines in the Sarajevo Canton with the positive list of drugs list in Croatia and found that the reference price is now lower in the FBH (Figure 8). 71 Table 5.10: Health Out-of-Pocket Payments in the FBH Out-of-pocket Out-of-pocket Cantons share (%) amount (KM) Drug amount (KM) Central Bosnia 5.40% 43.11 13.84 Herzegovina- Neretva 4.90% 48.43 14.11 Zenica-Doboj 4.40% 29.68 8.63 Una-Sana 4.30% 31.62 10.24 Tuzla 4.10% 27.67 11.05 Posavina 3.50% 28.49 7.37 Sarajevo 3.40% 36.31 12.16 West Herzegovina 2.50% 19.4 8.15 Canton 10 2.20% 14.52 5.92 Bosnian Podrinje 1.90% 11.84 5.23 Source: Bosnia HBS 2007. Expenditures per adult equivalent, authors‘ calculations. Note: OOP = share of total household expenditures Figure 5.8: Reference Prices of Generic Medicines in FBH and Croatia, 2011/12 100 FBiH Croatia 90 80 70 Prices in KM 60 50 40 30 20 10 0 V. AFFORDABILITY OF PHARMACEUTICALS 5.27. The pharmaceutical sector in BH is extremely fragmented. Eighty-three percent of drugs were imported through at least 40 wholesalers in 2009. Moreover, each of the 13 HIFs has its own drug procurement rules, procedures, and positive lists. Unfortunately, this decentralized procurement results in higher prices and substantial price variation across cantons and between entities. The 2006 PEIR noted that this inefficient system for pricing and procurement increases prices for patients, reduces transparency, and increases the risk for corruption. Since then, RS has centralized the procurement of 90 percent of medicines, medical supplies, and other medical products, while FBH has centralized the procurement by hospitals of medicines from the List of Medicines of the Solidarity Fund of FBH. This 72 reform has not only reduced the price of medicines but has also increased the availability of drugs in hospitals and health centers. Greater centralization of procurement in FBH could greatly reduce the costs of medicines there as well. 5.28. To increase the transparency of drug prices in BH, a survey of the retail pharmacy prices of 36 common drugs was conducted in 2011 with World Bank support.42 These drugs included essential medicines and those used to treat common illnesses such as cardiovascular disease, diseases of the nervous system, diabetes, and respiratory diseases. Data were collected from 82 pharmacies that were selected randomly from the eight cantons in BH. Brand-name drugs are rare in BH, being available only 27 percent of the time, so these results focus on generic drugs. 5.29. Partly due to the non-centralized procurement in FBH, drug prices vary substantially between entities and across regions. Because RS has a centralized procurement system with reference prices based on generic drugs, prices in RS tend to be lower and more consistent across different RS areas. In FBH, however, the way in which the prices of most drugs were set within cantons was not transparent, leading to substantial variation. Figure 9 shows that prices were higher in FBH than in RS for 10 out of the 13 cardiovascular drugs and in some cases by 50 percent or more. The prices of the whole sample of drugs tended to be lower in RS than in FBH. Within FBH, prices also varied substantially, which had serious implications for the affordability of medicines for many residents (see Figure 10). For five drugs in the sample, the difference between the highest- and lowest-cantonal prices was over 100 percent. In fact, over half of the sample of generic drugs had price differentials of 50 percent or more. 5.30. To ensure that medicine is equally accessible throughout the entire FBH, the Federal Minister of Health has updated the guide prices for the drugs on the List of Essential Medicines.43 Also, a federal decree adopted in November 10, 2011 gave cantons 60 days to harmonize their lists of medicines with the Federal Essential List of Medicines. However, that timeframe was extended, and so far only Una-Sana Canton, Zenica-Doboj Canton and Sarajevo Canton have fully complied. The implementation of the Federal Essential List of Medicines under this Decision has resulted in lower prices of medicines for cardiovascular diseases, even compared to RS (Figure 9b). The Federal Minister of Health has also published a rulebook establishing more precise criteria for selecting and using medicines as well as procedures for drafting the list of medicines included in the FBH Solidarity Fund.44 The rulebook has also increased the number of medicines that can be used to treat certain diseases. A Law on Records is currently being debated in Federal Parliament, which, if passed, will require pharmacies to send regular reports to the FBH Ministry of Health on their drug sales and prices. This will make it possible to monitor the extent to which the population of FBH has equal access to pharmaceuticals. 42 The survey followed the WHO/HAI standard methodology for measuring medicine prices and availability (World Health Organization and Health Action International, 2008). 43 Official Gazette of the Federation of BH, issue No. 75/11 44 Official Gazette of the Federation of BH, issue No. 36/11 73 Figure 5.9a: Prices of Generic Drugs for Cardiovascular Disease, by Entity 0.4 0.35 RS FBiH Drug price in Euro 0.3 0.25 0.2 0.15 0.1 0.05 0 Source: World Bank Survey Figure 5.9b: Updated Prices of Generic Medicines from the New Essential List in KM 18 RS FBiH 16 14 Drug price in KM 12 10 8 6 4 2 0 74 Figure 5.10: Differences between the Highest and Lowest Prices of Generic Drugs in FBH, by Canton 140.0% Percent difference (between max and min 120.0% 100.0% 80.0% 60.0% canton price) 40.0% 20.0% 0.0% Amoxicillin_susp_clav_a… Lisinopril Enalapril_and_diuretics Loratidine Lisinopril_and_diuretics Lacipidine Amoxicillin Paroxetine Diclofenac Insulin Zolpidem Atenolol Carvedilol Alprazolam Ciprofloxacin_injections Bisoprolol Diazepam Amoxicillin_susp Glibenclamide Enalapril Metformin Furosemide Omeprazole Amlodipine Atorvastatin Ranitidine Co_trimaxazole_susp Isosorbide_mononitrate Cotrimaxazole Simvastatin Source: World Bank Survey Note: These prices prevailed before the implementation of the new Federal Essential List of Medicines in 2011. VI. RECOMMENDATIONS 5.31. Health system reforms must continue in BH. Substantial progress has been made in terms of investments in family medicine, changes to health financing, the introduction of quality standards, and more. However, work must continue on these reforms as many (such as the introduction of DRGs) have long implementation periods that will require 10 years or more before their positive effects on efficiency and quality are fully realized. We make the following recommendations to strengthen and enhance the reform process. Reform Health Insurance and Financing ï‚· Expand the insurance pool related to hospital and pharmaceutical care. Currently in FBH, cantonal HIFs have the legal authority to contract with public and private health providers, but this could be done more efficiently at the entity level. Since hospital and pharmaceutical care account for over 50 percent of health expenditures, expanding contracting and purchasing to the entity level would lead to greater efficiency and to a more appropriate allocation of resources. ï‚· Review exemptions from contributions and consider centralizing the collection and control of health contributions to the entity level to improve collection. In both entities, substantial legal exemptions have reduced the contributory risk pool to less than 50 percent of the population. Coordination with the Tax Authority is necessary to increase contributions, and all exemptions should be reviewed. In FBH, each canton has different contribution rates for farmers and self-employed so shifting responsibility for collecting them to the entity level may make the collection process more efficient and may increase the amount collected. 75 Improve Governance and Planning ï‚· Reduce fragmentation and duplicated functions. In FBH, considerable duplication of roles and functions exist concerning the administration of health care in the MoHs and IPHs. While this is a difficult political issue, the decentralized system hinders effective coordination and efficiency within the health system. Greater managerial skills at all levels and better planning of resources are also needed. Increase Efficiency of Service Delivery ï‚· Leverage family medicine reforms to increase PHC productivity and increase preventive care and health promotion. Family medicine reforms have been significant in reorienting the health system to provide more preventive services, and the health system should extend this to cover NCD screening and smoking prevention (a pilot project in FBH is testing incentives to increase these interventions). ï‚· Fully implement hospital and primary health care financing reforms. The BH is introducing the DRGs payment system for hospitals and the capitation system in PHC, but additional work is needed to complete the implementation. Improve Pharmaceutical Procurement and Policy ï‚· Centralize the procurement of drugs in FBH to the entity level. The procurement of drugs is highly fragmented, leading to substantial differences in the prices and affordability of medicines in different areas. Centralizing the procurement of drugs in FBH should reduce total drug expenditures by up to 20 percent. In addition, the government of FBH should increase the centralized regulation of quality control, produce a positive list of drug benefits, and regulate retail pricing. 76 CHAPTER 6: EDUCATION I. INTRODUCTION Political context 6.1. The Constitution of BH defines complex governance and management arrangements for the education sector. First, there is no Ministry of Education at the state level. There is a small education unit within the Ministry of Civil Affairs (MoCA), but its involvement in educational matters at the sub- national level is minimal. For instance, even though the country joined the Bologna Process45 in 2005, MoCA can only play a coordinating role, and it does not have enforcement authority at the entity or cantonal levels even for the implementation of internationally agreed actions. Second, governance structures for the education sector are different in the two entities. In RS, policy is basically formulated by a single Ministry of Education (MoE) at the entity level, while in FBH, it is at the cantonal level. In the latter, each of the 10 cantons is responsible for virtually all educational matters much like an independent country, and FBH MoE cannot intervene in cantons.46 Consequently, the rationale for the existence of FBH MoE itself is being questioned today. Even though the Conference of Education Ministers was recently established to develop a consensus on education policies among cantons, it is only an advisory body with no binding authority. The education department of BrÄ?ko District is yet another independent agency that sets education policy in BH. In total, there are 14 ministries that deal with education in BH— MoCA at the state level, one MoE in RS, one Federation and 10 cantonal MoEs in FBH, and the education department of BrÄ?ko District. Under this extremely fragmented education system, no state-level or federation-level reforms can be decided and implemented without consent of each entity and canton. Consequently, a dysfunctional institutional framework governs education policy in BH. Revisiting the recommendations of the 2006 PEIR 6.2. The 2006 Public Expenditure and Institutional Review (PEIR) noted that the highly decentralized education system, especially in FBH, has resulted in inequitable and inefficient spending on education. Lack of effective responses to the sharp declines in school-age population in the past decade has made the school system inefficient, with small class sizes and low student teacher ratios. Due to the relatively high teacher salaries compared to similar professions and to overstaffing, wages take up a high percentage of the education budget. Costly vocational schools do not seem to provide relevant skills required in the market. Access to upper secondary and tertiary education is inequitable with few mitigation measures. 6.3. Given these issues, the 2006 PEIR made a number of recommendations to improve the efficiency, equity, and effectiveness in education financing as follows: ï‚· Reduce the number of teachers proportionately to the decline in student numbers; ï‚· Increase the average class size for primary education; ï‚· Moderate the growth of teacher salaries; ï‚· Replace the input-based financing of schools with capitation financing; 45 The Bologna Process is the process of creating the European Higher Education Area (EHEA), and is based on cooperation between ministries, higher education institutions, students and staff from 47 countries, with the participation of international organizations (http://www.ond.vlaanderen.be/hogeronderwijs/bologna/). 46 It is currently discussed at FBH Parliament that based on the Law on Self-Governance, responsibilities over education may be further decentralized from cantons to municipalities and cities (between cantons and municipalities). If so decided, cantons will be responsible for legislation and curricula, municipalities for financing and planning for preschool and primary education, and cities for secondary education. 77 ï‚· Decrease the proportion of secondary vocational and technical education to (less costly) general education; ï‚· Unify management within each higher education institution; ï‚· Increase fees for repeating students in higher education; and ï‚· Raise higher education fees for foreign students. Unfortunately, very few actions have been taken to date and the same problems persist today. Focus of analysis in this PEIR 6.4. The fundamental challenge for the education sector is rooted in the existing constitutional setting. While there are many issues to be addressed nationwide, any analysis and policy reform recommendations targeting the state level will be of little practical consequence because the state government has no legal basis to take reform actions, as discussed above. Likewise, FBH government has no authority over cantons to introduce educational reforms without the latter‘s consent. Therefore, this PEIR will focus on the analysis of the education systems at the entity level for RS and at the cantonal level for FBH where administration and management of the education sector actually take place.47 It will focus on financing of primary and secondary education. It will not cover higher education financing because the subject was very recently studied by the European Union (see Annex 6.1),48 as well as in a World Bank study on skills development, which discussed the role of tertiary education at length.49 6.5. This chapter is organized as follows: The second section reviews the country‘s overall education system, public spending on education, and student performance. The third section compares the education systems and spending between the two entities. The fourth section examines equity and efficiency of education spending in RS and the cantons in FBH in detail. This includes the analyses of unit costs in education, efficiency measures, wage policies, and the standards and norms for staffing. The last section summarizes the findings and provides policy recommendations. II. OVERVIEW OF THE EDUCATION SYSTEM – STATE LEVEL50 6.6. The education system of BH consists of non-compulsory pre-school education, nine years of free, mandatory primary education, three to four years of secondary education, and higher education. Primary education is currently going through a transition from an eight-year to nine-year curriculum following the Law on Primary and Secondary Education of 2003. In RS, the transition started in the school year 2003/04, and all schools have already been shifted to the nine-year curriculum. In FBH, the transition began in the school year 2004/05, but not all of the cantons have begun the transition. In the school year 2009/10, three of the 10 cantons—Central Bosnia, Herzegovina-Neretva, and West Herzegovina Cantons—still enrolled new grade-one students in the eight-year curriculum.51 Secondary 47 The authors wished to analyze school-level data as well, but it was not possible due to data unavailability (see Annex 6.2). 48 European Union, 2009, Feasibility Study for Higher Education Financing Reform, Final Report, March 2009. 49 World Bank, 2009, Are Skills Constraining Growth in Bosnia and Herzegovina? , Report No. 54901-BA. 50 The population of the BrÄ?ko District is about 80,000, which is the size of a small canton in FBH. It consists of about 1.7 percent of the total population of BH. Due to lack of relevant data, the rest of this PER does not cover the BrÄ?ko District. 51 As of 2009/10, 55 percent of 234,000 primary students in FBH were still enrolled in the eight-year curriculum— ranging from 95 percent of 28,700 students in Central Bosnia canton to 41 percent of 39,500 students in Sarajevo canton. (Source: Federal Office of Statistics, 2010, Statistical Bulletin 2009 for Pre-school Education 2009, Primary Education End of 2008/2009 and Beginning of 2009/2010, Sarajevo). It would be logical to assume that any 78 education is offered in four-year gymnasiums (grammar schools), four-year technical schools, three-year vocational schools, as well as schools for art, religious, and special education. While students in technical schools have an option to proceed to university, those in vocational schools do not. There are two public universities in RS and six in FBH. 6.7. In 2009/10, there were approximately 625,000 students in BH, as shown in Table 6.1. Around 16,000 children were enrolled in 207 preschools, 350,000 in 1,914 primary schools, and 143,000 students in 306 secondary schools. There were 35 higher education institutions, enrolling 116,000 full-time and part-time students. Table 6.1: Enrollment by Level of Education, 2009/10 Number of institutions Enrollment RS FBH Total RS FBH Total Pre-school education 78 129 207 6,583 9,839 16,422 Primary education (grades 1-8/9) 754 1,160 1,914 108,736 240,877 349,613 Secondary education (grades 9/10-12/13) 94 212 306 48,225 95,134 143,359 Tertiary education 26 9 35 43,928 71,610 115,538 Colleges 17 0 17 5,520 0 5,520 Universities 9 9 18 38,408 71,610 110,018 Total 952 1,510 2,462 207,472 417,460 624,932 Sources: FBH Federal Office of Statistics; RS Institute of Statistics. 6.8. BH‟s enrollment rates are relatively high for primary and lower secondary, compared to other countries in the region, but they drop sharply at the upper secondary level, and further at the tertiary level where they are far below the regional average . Figure 6.1 compares BH‘s gross enrollment rates by level of education with those of the neighboring countries in the Europe and Central Asia (ECA) region. BH‘s primary education (grades 1-8/9) is equivalent to primary (grades 1-4/5) and lower secondary (grades 4/5-8/9) in this comparison. The pre-school enrollment rate is the second lowest in the region. particular school offers only one of the eight and nine-year curriculum for any particular grade, but without school- level data available, it is not possible to verify this. 79 52 6.9. 0 0 100 120 140 100 120 20 40 60 80 20 40 60 80 Tajikistan Tajikistan BiH Kyrgyz Rep. Turkey Turkey Kyrgyz Rep. Macedonia Macedonia Azerbaijan BiH Uzbekistan Armenia Belarus Serbia Armenia Kazakhstan Croatia Romania Poland Russian Fed. Georgia Lithuania Serbia Romania Pre-primary Kazakhstan Moldova Public spending on education Bulgaria Moldova Slovenia fees at the entity and cantonal levels. Croatia Hungary Latvia Source: World Bank EdStats. Bulgaria Russian Fed. Slovak. Rep. Slovak. Rep. Belarus Upper secondary Czech Rep. Ukraine Ukraine Czech Rep. Macedonia Latvia Uzbekistan Hungary Moldova Azerbaijan Lithuania Kyrgyz Rep. Lithuania Slovenia Poland Poland Croatia Russian Fed. Azerbaijan Slovenia Uzbekistan Ukraine Serbia Georgia Latvia Armenia 80 Uzbekistan Primary Belarus Azerbaijan Hungary Tajikistan Turkey Romania Georgia Bulgaria BiH Tajikistan Slovak. Rep. Figure 6.1: Gross Enrollment Rates, 2009 or latest Moldova Czech Rep. Turkey Kazakhstan Georgia Kazakhstan BiH Macedonia Russian Fed. Bulgaria Croatia Moldova Serbia Turkey Macedonia Armenia Kyrgyz Rep. Kyrgyz Rep. Azerbaijan Slovak. Rep. Tertiary Bulgaria Tajikistan Slovak. Rep. Georgia Slovenia Czech Rep. Belarus Hungary Ukraine Uzbekistan Romania Armenia Latvia Czech Rep. Hungary Lower secondary Poland Serbia Lithuania Belarus Poland Russian Fed. Germany Latvia Lithuania Croatia Ukraine Romania BiH Slovenia Kazakhstan Income Tax to the entities, cantons, and municipalities. In addition, the respective governments may levy various The state government redistributes various tax revenues, including VAT, Corporate Income Tax, and Personal cantonal level in FBH. Neither the entity governments nor the cantonal governments receive any The majority of the education budget is determined at the entity level in RS and at the their respective financial capacities52 and policy priorities. In total, BH spent 4.5 percent of GDP on earmarked allocations for education from the higher levels, so the education budget is decided based on education in 2009, close to the regional median of 4.2 percent, as shown in Figure 6.2. As a proportion of total public expenditures, BH spent less on education than other ECA countries: 8.8 percent versus a regional median of 11.7 percent. 6.10. It is important to note comparisons of education spending are affected by the fact that official GDP has been somewhat underestimated in BH due to the incomplete accounting of the “non-observed economy (NOE)â€?. To make BH‘s GDP comparable to other countries‘ GDP, the 2006 PEIR adjusted the officially estimated GDP upward by 30 percent.53 Since then, various tax reforms have taken place, resulting in reductions in the NOE. Today, the official GDP (compiled using a production approach) takes into account some estimates for the NOE. However, GDP figures produced using the expenditure approach potentially accounts for even more of the NOE. Differences between two methodologies (production and expenditure approaches), were 14 percent in 2008 and 16 percent in 2009. This suggests that BH‘s public spending on education in 2009 was somewhere between 3.9 percent of GDP (GDP using expenditure approach) and 4.5 percent (GDP using production approach). The lower end places BH slightly below the regional average of 4.2 percent (Figure 6.2). Figure 6.2: BH‟s Public Spending on Education in Comparison to Other ECA Countries (a) Public Expenditure on Education as % of GDP (b) Public Expenditure on Education 2009 or latest available year as % of Total Government Expenditure 10.0 2009 or latest available year 25.0 8.0 20.0 6.0 15.0 4.0 10.0 2.0 5.0 0.0 0.0 Sources: World Bank EdStats; Ministries of Finance of the respective BH entities and FBH cantons. Notes: 1/ Denotes GDP adjustment to account for approximate size of ―non-observed economy‖ (NOE). 6.11. Wages for educational staff relative to economy-wide wages have increased considerably in the last several years; between 2006 and 2008 alone, the average real wage for the education sector went up by 29 percent, while the economy-wide average wage rose by only 6 percent.54 As shown in Figure 6.3, by 2008, the average teachers‘ salary reached 108 percent of the average in the economy, but was still 32 percent lower than the average public sector salary. It is important to note, however, that the data on average pay in the private sector appear to be systematically under-reported in the employing organizations‘ surveys conducted by the statistical offices. As a result, it is most likely that public sector employees look more overpaid relatively to those in the private sector than might actually be the case. 53 BH‘s national account data were then not adjusted for the size of the NOE, unlike most other countries in the ECA region. The 2006 PEIR referred to different figures for the NOEs estimated by the OECD and IMF which ranged from 40 percent to 58 percent in the early to mid 2000s. The estimated NOE in the 2006 PEIR was a most conservative one. 54 World Bank, 2009, Are Skills Constraining Growth in Bosnia and Herzegovina? , Figure 4, Report No. 54901-BA. 81 Figure 6.3: Public Sector Wages in BH, 2008-2010 (a) Public sector salaries as % of average in the economy (b) Average gross monthly salaries (in KM) 2,000 140% 143% 139% 1,800 1,600 108% 109% 108% 1,400 1,200 1,000 800 600 400 200 0 2008 2009 2010 (3 Qs) 2008 2009 2010 (3 Qs) public admin, police & social security education average in economy public admin, police & social security education Source: BH State Statistical Office Student performance 6.12. The state and entity governments have aimed to establish a national standardized student assessment system since the mid 2000s, but no assessment tools have been developed to date to rigorously measure student performance nationally. As an alternative, BH participated in the Trends in International Mathematics and Science Study (TIMSS) in 2007.55 While TIMSS assesses mathematics and science achievements of grades 4 and 8 students, BH participated only in the grade 8 assessment. As illustrated in Figure 6.4, grade 8 students in BH did not perform as well as those in most other participating countries in the ECA region. It is particularly interesting to note that, apart from two countries with extremely low spending (Turkey and Georgia), there seems to be no clear correlation between public spending as a percentage of GDP and student performance. Whether the official or adjusted GDP is used for the x-axis, the performance of BH‘s students was only better than the two lowest spenders on education. The BH‘s students performed about the same as those in Ukraine which spends 5.3 percent of GDP on education and Bulgaria and Romania which spend about 4.2-4.3 of GDP, but much more poorly than those in the Russian Federation and Armenia whose spending was lower than BH‘s (much lower in the case of Armenia). Figure 6.4: TIMSS 2007: BH 8th Grade Performances in Comparison to Other ECA Countries TIMSS 2007 Mean Score TIMSS 2007 Mean Scores and education spending as 550 percentage of GDP 550 525 525 Russian Czech Hungary 500 Federation Republic Slovenia Lithuania 500 Armenia 475 475 Bulgaria Serbia Ukraine 450 450 Turkey BiH Romania BiH 425 Georgia (adj. (official 425 GDP) GDP) 400 400 3.0 3.5 4.0 4.5 5.0 5.5 6.0 8th grade mathematics 8th grade science 8th grade mathematics 8th grade science Souce: World Bank EdStats. 55 Unfortunately, BH will not participate in TIMSS in 2011. The country has not participated, or planned to participate in any other international assessments such as PISA or PIRLS yet. 82 6.13. The TIMSS results not only allow us to compare BH students‟ achievement with those in other participating countries, but reveal important characteristics of BH‟s education system that might indicate inefficient use of resources and help explain relatively low performance . Class sizes are relatively small in BH; the overall average class size for grade 8 mathematics was 24, while the international average (among TIMSS participants) was 29. In BH, 48 percent of the students were enrolled in a class with 1-24 students, and the rest in a class with 25-40 students. As a comparison, the international average is 30 percent in a class with 1-24 students, 59 percent with 25-40 students, and 11 percent with more than 40 students. Typically, students in a class with 25-40 students tend to perform better than those in a smaller class in many other countries, but in BH, there is very little difference in student performance between those in small classes (score 454) and those in larger classes (score 458). Mathematics teachers are less qualified than in the other countries surveyed: 91 percent of the students were taught by teachers who have completed only non-university post-secondary education, while only 9 percent were taught by teachers with a university degree or above. On international average, 21 percent of the students were taught by teachers with a postgraduate degree, 57 percent by those with an undergraduate degree, and only 18 percent by those with a non-university post-secondary education. Part of the reason for the large proportion of low qualified teachers might be because the teachers‘ age: 50 percent of the students were taught by teachers over 50 years old, compared to the international average at 23 percent. On the other hand, only 6 percent of students were taught by teachers 29 years old or under.56 6.14. The next two sections explore the potential areas where public funds could be more efficiently spent by reviewing entity and cantonal level data in detail. III. OVERVIEW OF THE EDUCATION SYSTEM – ENTITY LEVEL 6.15. Figure 6.5 illustrates the recent trends in enrollments in the two entities by level of education. While secondary enrollments have gradually declined throughout the 2000s in both entities, primary enrollments started to decline in 2003/04 in RS and in 2004/05 in FBH. Both declines primarily reflect the demographic change. Preschool enrollments have been extremely limited, though they started increasing around 2005. Tertiary enrollments have increased rapidly, especially in RS. The student-teacher ratios (STRs) for primary and secondary education continued to decline throughout the last decade in both entities, indicating that the governments have failed to adjust the number of teachers as enrollments declined. The BH‘s overall STR for primary education is around the ECA average, and that for secondary education is still above the ECA average today (Figure 6.6), but given the declining student population, without taking any action to stop the declining trend now, it will soon make the STRs too low. On the other hand, the STR for higher education increased sharply as enrollments increased, failing to meet the growing demand for higher education. The STR for preschool has increased slightly since the mid-2000s, again not to adjusting to the increasing enrollments. 56 Mullis, I.V.S., Martin, M. & Foy, P, 2008, TIMSS 2007 International Mathematics Report: Findings from IEA’s Trends in International Mathematics and Science Study at the Fourth and Eighth Grades , TIMSS & PIRLS International Study Center, Boston. 83 Figure 6.5: Enrollments and Student-Teacher Ratios by Level of Education, 2001/02-2009/10 Sources: FBH Federal Office of Statistics; RS Institute of Statistics. Figure 6.6: Student-Teacher Ratios by Level of Education – International Comparison, 2008 or Latest Student-Teacher Ratio in Primary Education Student-Teacher Ratio in Secondary Education latest available year latest available year 25.0 25.0 20.0 20.0 15.0 15.0 10.0 10.0 5.0 5.0 0.0 0.0 Source: EdStats database 6.16. Figure 6.7 shows the recent trends in secondary enrollment by type of education. In both entities, three-year vocational education has played a diminishing role over the past decade, accounting for 20 percent of all secondary school students in 2009/10 (down from 34 percent in 2001/02). Four-year secondary programs, meanwhile, have grown in importance: enrollment in gymnasiums has increased from 18 to 23 percent in RS over the past decade and from 24 to 27 percent in FBH (up from 22 to 26 percent for the country as a whole). However, the majority of enrollments is in four-year programs in technical fields in secondary education. Enrollment in the fields of economics, law, 84 public health, engineering and related disciplines now accounts for 52 percent of all secondary students (up from 43 percent in 2001/02); the proportion is even higher in RS, where 59 percent of students are enrolled in the technical fields (compared to 49 percent in FBH). Figure 6.7: Share of Secondary Enrollment by Type of Education, Academic Year 2001/02-2009/10 Sources: FBH Federal Office of Statistics; RS Institute of Statistics. Public spending on education 6.17. As discussed above, the mechanisms for education financing differ between the two entities. In RS, based on the new regulation on primary and secondary education financing adopted in 2008, the entity government pays all expenses for primary education and only wages for secondary education. Operation and maintenance expenses for the latter are paid by municipalities based on the Law on Local Self-Governance. In 2010, the total education budget was KM 347 million (€174 million) and was allocated to schools based on the size of schools, the condition of facilities, the number of students and the number of classes. Of the education budget, 91 percent was spent on wages. Given this allocation, the only way to achieve budget savings would be to reduce the teaching force and/or consolidate small inefficient schools. 6.18. In FBH, public spending on education is the responsibility of cantons. Less than one percent of education spending occurs at the level of FBH. In 2009, FBH MOE spent its budget of KM 2 million (€1 million) on activities dealing with: (i) curriculum and textbook approval; (ii) promotion of education reforms; (iii) some infrastructure improvements; and (iv) minor equalization measures —e.g., providing free textbooks to students from underprivileged communities and subsidizing room and board for some university students. Most education spending in FBH (over KM 800 million (€400 million)) takes place at the cantonal level with a very small amount at the municipal level. Cantonal governments are typically responsible for paying staff salaries—which account for about 90 percent of all education spending—as well as most non-salary operational expenses. In some cantons, municipalities finance capital improvements and utility costs, especially for primary schools. 6.19. Figure 6.8 illustrates the recent trends in education spending in the two entities. The RS increased its allocation of total government expenditure to education from 17.3 percent in 2007 to 22.0 percent in 2008, primarily reflecting a sharp increase in teacher salaries. As a result, education spending as a percentage of RS‘s GDP increased considerably from 3.3 percent in 2007 to 4.3 percent by 2009. On the other hand, FBH steadily spent around 21.5 percent of government expenditure on education between 2005 and 2008, before experiencing a slight decrease to 20.2 percent in 2009 due to the economic crisis. 85 As a percentage of GDP, education spending remained high at 4.6 to 5.3 percent. If adjusted GDP is applied, these figures are lowered to 3.7 percent for RS and 4.1 percent for FBH in 2009. Figure 6.8: Public Spending on Education by Entity, 2005-2009 Sources: FBH Federal Office of Statistics; RS Institute of Statistics; Ministries of Finance of the respective entities and FBH cantons. 6.20. Education spending as a percentage of GDP for the entire country is a weighted average of that for the two entities, which is close to the ECA average (Figure 6.9a). However, education spending as a percentage of the total government expenditure for the entire country (8.8 percent) is not an average of that for the two entities (22 percent for RS and 20 percent for FBH) because the sum of entity and cantonal government budgets comprises only 42 percent of the country‘s total government expenditure. Indeed, the total government expenditure for BH is between 45 percent and 52 percent of GDP—which is one of the highest in the ECA region. Meanwhile the total budget expenditure of each entity is between 17 percent and 20 percent of GDP for RS and between 20 percent and 24 percent of GDP for FBH (Figure 6.9b). Given the high level of public spending in the country as a whole, the relatively low education spending as a percentage of the total government expenditure (8.8 percent) probably implies that the government is spending enough on education. On the other hand, it is difficult to judge whether or not each entity and canton is spending a reasonable proportion of its budget on education because that determination would depend on the functions assigned to them, financial capacities, policy priorities, and geographical, demographic, and other factors. Figure 6.9: Public Expenditure on Education as % of Total Government Expenditure, 2009 or Latest (a) Compared to education spending as % of GDP (b) Compared to total government expenditure as % of GDP 25.0 25.0 as % of total government expenditure as % of total government expenditure RS (adj. GDP) RS RS (adj. GDP) RS Public expenditure on education Public expenditure on education FBiH FBiH 20.0 FBiH (adj. GDP) 20.0 FBiH (adj. GDP) 15.0 15.0 10.0 10.0 BiH BiH (adj. GDP) BiH BiH (adj. GDP) 5.0 5.0 0.0 0.0 0.0 2.0 4.0 6.0 8.0 10.0 0.0 10.0 20.0 30.0 40.0 50.0 60.0 Public expenditure on education as % of GDP Total government expenditure as % of GDP Source: World Bank EdStats. 86 6.21. The differences in education spending between the two entities are also illustrated in terms of per-student spending (Figure 6.10). On average, RS spends a lower amount per student than FBH does. In particular, RS spends a lower amount per student for secondary education than for primary education, which is contrary to the normal situation as secondary education has a higher number of specializations and is consequently more expensive. One reason for this is the role of municipal governments in financing operational expenditures for secondary schools in RS, which is not reflected here. However, this does not fully explain the lower spending in RS on education in general, and secondary education in particular, because operational expenditures only make up about 10-15 percent of the total recurrent spending. Figure 6.10: Per Student Spending on Primary and Secondary Education by Entity, 2005-2009 Per student spending 2,500 2,000 FBiH secondary 1,500 FBiH primary KM RS primary 1,000 RS secondary 500 0 2005 2006 2007 2008 2009 Sources: FBH Federal Office of Statistics; RS Institute of Statistics; Ministries of Finance of the respective entities and FBH cantons. 6.22. These observations lead to a series of questions that must be examined: What accounts for the differences in financing levels between the two entities? Is RS spending insufficient amounts on education or is it managing adequate resources more efficiently than FBH? Is the secondary school network more efficient than the primary one in RS? How different are teacher salaries between the entities? The next section analyzes the efficiency and equity of education spending in RS and each canton in FBH. IV. ANALYSIS OF THE EQUITY AND EFFICIENCY OF EDUCATION SPENDING IN RS AND FBH CANTONS Drivers of unit costs 6.23. As summarized in Table 6.2, while secondary class sizes are comparable across the two entities-an average of 24.3 students in FBH and 25.0 in RS-class sizes in primary education vary substantially between the entities, ranging from 19.4 in RS to 21.6 in FBH‟s primary schools. As a result, small schools and classes in primary education lead to higher unit costs in RS, where 754 primary schools serve just over 100,000 students—an average of 144 students per school. Meanwhile, unit costs in 87 secondary education are substantially lower in RS: KM 1,471 per student,57 compared to KM 2,024 in FBH. Table 6.2: Education Statistics for Primary and Secondary Schools by Entity, 2009 Summary Statistics Efficiency Measures Number Number Number Number Non- Average Student- % Part- % Non- Spending % Spending of of of of Teaching School Average Teacher Time Teaching per Student on Personnel Schools Classes Students Teachers 1 Staff 2 Size Class Size Ratio Teachers Staff (KM) Costs FBH Primary schools 1,160 10,570 240,877 14,575 618 207.7 22.8 16.5 20.7% 4.1% 1,670 91.0% Secondary schools 212 3,911 95,134 6,629 366 448.7 24.3 14.4 35.5% 5.2% 2,024 86.3% Total primary and secondary schools 1,372 14,481 336,011 21,203 984 244.9 23.2 15.8 25.3% 4.4% 1,770 89.5% RS Primary schools 754 5,609 108,736 7,560 n/a 144.2 19.4 14.4 39.9% n/a 1,753 90.4% Secondary schools 94 1,926 48,225 3,159 160 513.0 25.0 15.3 27.8% 4.8% 1,471 98.9% Total primary and secondary schools 848 7,535 156,961 10,718 n/a 185.1 20.8 14.6 36.4% n/a 1,666 92.7% FBH + RS Primary schools 1,914 16,179 349,613 22,134 n/a 182.7 21.6 15.8 27.3% n/a 1,696 90.8% Secondary schools 306 5,837 143,359 9,787 526 468.5 24.6 14.6 33.0% 5.1% 1,838 89.7% Total primary and secondary schools 2,220 22,016 492,972 31,921 n/a 222.1 22.4 15.4 29.0% n/a 1,737 90.5% Sources: FBH Federal Office of Statistics; RS Institute of Statistics; Ministries of Finance of the respective entities and FBH cantons. Notes: 1 Number of teachers is reported as full-time equivalent (FTE), calculated as the number of full-time teachers plus half of the number of part-time teachers. 2 Non-teaching staff include individuals who are involved in the educational process but do not participate directly in student instruction (e.g., pedagogues, psychologists, health workers, librarians). This category does not include auxiliary staff, such as janitors, food preparation workers, etc. 3 n/a = data is not available. 6.24. Variations across the cantons within FBH are even wider, as shown in Table 6.3 In 2009, spending on primary education in FBH ranged from KM 1,450 per student in Central Bosnia Canton to KM 2,039 per student in Canton 10, compared to KM 1,753 per student in RS for primary education. Spending on secondary education varied even more across cantons—from KM 1,651 per student in Bosnian Podrinje Canton to KM 2,568 in Canton 10. 57 True costs of RS secondary education are likely higher by approximately 10 percent, due to the fact that non- salary expenses are paid out of municipal budgets. 88 Table 6.3: Education Statistics for Primary and Secondary Schools by FBH Canton, 2009 Summary Statistics Efficiency Measures Number Number Number Number Non- Average Student- % Part- % Non- Spending % Spending of of of of Teaching School Average Teacher Time Teaching per Student on Personnel Schools Classes Students Teachers 1 Staff2 Size Class Size Ratio Teachers Staff (KM ) Costs Primary and Secondary Schools A B C D E F=C/A G=C/B H=C/D I J K L Una-Sana 189 1,750 41,358 2,459 67 219 23.6 16.8 19% 2.7% 1,579 94.0% Posavina 25 260 5,304 341 16 212 20.4 15.6 19% 4.5% 1,981 76.4% Tuzla 261 2,901 70,619 4,355 197 271 24.3 16.2 28% 4.3% 1,738 89.1% Zenica-Doboj 246 2,462 59,330 3,665 144 241 24.1 16.2 30% 3.8% 1,628 89.6% Bosnian Podrinje 18 179 3,863 240 16 215 21.6 16.1 11% 6.3% 1,880 82.8% Central Bosnia 170 1,660 40,279 2,271 109 237 24.3 17.7 26% 4.6% 1,511 85.5% Herzegovina-Neretva 175 1,543 32,526 2,267 94 186 21.1 14.3 32% 4.0% 1,953 94.4% West Herzegovina 88 705 14,598 1,005 41 166 20.7 14.5 18% 3.9% 1,761 90.5% Sarajevo 131 2,536 58,940 3,938 279 450 23.2 15.0 23% 6.6% 2,075 87.4% Canton 10 69 485 9,194 663 21 133 19.0 13.9 15% 3.1% 2,192 93.6% Total FBH 1,372 14,481 336,011 21,203 984 245 23.2 15.8 25% 4.4% 1,770 89.5% cf. Total RS 848 7,535 156,961 10,718 n/a 185 20.8 14.6 36% n/a 1,666 92.7% Primary Schools Una-Sana 166 1,313 30,811 1,713 35 186 23.5 18.0 13% 2.0% 1,455 93.5% Posavina 23 195 3,837 251 13 167 19.7 15.3 15% 4.9% 1,860 80.2% Tuzla 225 2,099 49,245 2,970 135 219 23.5 16.6 26% 4.3% 1,670 90.6% Zenica-Doboj 211 1,822 43,364 2,545 86 206 23.8 17.0 25% 3.3% 1,469 90.8% Bosnian Podrinje 15 136 2,775 172 13 185 20.4 16.1 12% 7.0% 1,970 86.6% Central Bosnia 144 1,218 29,591 1,629 69 205 24.3 18.2 22% 4.1% 1,450 87.6% Herzegovina-Neretva 145 1,127 23,093 1,564 46 159 20.5 14.8 24% 2.9% 1,767 98.0% West Herzegovina 78 547 10,596 747 24 136 19.4 14.2 15% 3.1% 1,741 94.2% Sarajevo 93 1,743 41,025 2,497 185 441 23.5 16.4 19% 6.9% 2,034 88.5% Canton 10 60 370 6,540 490 12 109 17.7 13.4 3% 2.4% 2,039 93.6% Total FBH 1,160 10,570 240,877 14,575 618 208 22.8 16.5 21% 4.1% 1,670 91.0% cf. Total RS 754 5,609 108,736 7,560 n/a 144 19.4 14.4 40% n/a 1,753 90.4% Secondary Schools Una-Sana 23 437 10,547 746 32 459 24.1 14.1 33% 4.1% 1,940 95.0% Posavina 2 65 1,467 91 3 734 22.6 16.2 32% 3.2% 2,296 68.2% Tuzla 36 802 21,374 1,385 62 594 26.7 15.4 33% 4.3% 1,893 86.0% Zenica-Doboj 35 640 15,966 1,121 58 456 24.9 14.2 40% 4.9% 2,058 87.3% Bosnian Podrinje 3 43 1,088 68 3 363 25.3 16.0 9% 4.2% 1,651 71.4% Central Bosnia 26 442 10,688 642 40 411 24.2 16.6 39% 5.9% 1,678 80.7% Herzegovina-Neretva 30 416 9,433 703 48 314 22.7 13.4 49% 6.4% 2,406 87.9% West Herzegovina 10 158 4,002 259 17 400 25.3 15.5 25% 6.2% 1,815 81.3% Sarajevo 38 793 17,915 1,442 94 471 22.6 12.4 29% 6.1% 2,168 84.9% Canton 10 9 115 2,654 174 9 295 23.1 15.3 48% 4.9% 2,568 93.6% Total FBH 212 3,911 95,134 6,629 366 449 24.3 14.4 35% 5.2% 2,024 84.0% cf. Total RS 94 1,926 48,225 3,159 160 513 25.0 15.3 28% 4.8% 1,471 98.9% Sources: FBH Federal Office of Statistics; Ministries of Finance of the respective FBH cantons. Notes: 1 Number of teachers is reported as full-time equivalent (FTE), calculated as the number of full-time teachers plus half of the number of part-time teachers. 2 Non-teaching staff include individuals who are involved in the educational process but do not participate directly in student instruction (e.g., pedagogues, psychologists, health workers, librarians). This category does not include auxiliary staff, such as janitors, food preparation workers, etc. 6.25. The differences in class sizes and student-teacher ratios account for much of the variation in unit costs within FBH. It comes as no surprise that cantons with the smallest classes also have the highest cost per student. In primary education, in particular, the differences are dramatic. As shown in Figure 6.11, while Central Bosnia Canton groups its primary students, on average, in classes of 24.3, Canton 10 can only muster average classes of 17.7. As a result, Canton 10‘s unit costs in primary education are 40 percent higher than those in Central Bosnia Canton. Because Canton 10 maintains a network of 60 primary schools to cater to its 6,500 students, it is unable to group students in larger classes under the current arrangement. 89 Figure 6.11: Per-Student Expenditure and Average Class Size in Primary and Secondary Schools by FBH Canton, 2009 Source: World Bank staff calculations based on data from FBH Federal Office of Statistics and Ministries of Finance of the respective FBH cantons. 6.26. In all, about half of the variation in unit costs among cantons can be explained by the differences in average class sizes. While the educational norms of most cantons stipulate that the optimum class size in primary education should be at least 24 students (see Table 6.11 below), only one of ten cantons manages to maintain average classes of that size. In fact, half of the cantons maintain average classes below 21 students in primary education, thus incurring unit costs at least 12 to 15 percent higher than the cantons with primary classes near the optimum size. On average, increasing the class size by one student is associated with an approximate reduction in per-student expenditures of 4 percent in primary education and 8 percent in secondary education. While these figures do not take into account other factors that influence unit costs—such as the demographic and geographic characteristics of the individual schools and communities—a more thorough analysis of the drivers of unit costs could be conducted if relevant school-level data were available (see Annex 6.2 for detailed recommendations on improving data collection practices in education). Wages and allowances 6.27. Education employees on average, including not only teachers but also administrators and support staff, might have been slightly underpaid until recently, but now appear to be reasonably paid or even slightly overpaid relative to other employees in the economy. Table 6.5 and Figure 6.11 show the average wages between 2006 and 2009 for the education sector and total economy in the two entities. In 2006, education employees in RS earned only 88 percent of the economy-wide average, while those in FBH earned 102 percent. The inter-entity gap reversed in 2008 and further widened in 2009: education employees in RS earned 114 percent of the economy-wide average, while those in FBH made 104 percent. Without more detailed data on types of jobs, qualifications of employees, and so on, it is difficult to judge whether education employees are, on average, underpaid, reasonably paid, or overpaid. On the one hand, given that a much higher proportion of the employees in the education sector have a university degree than those in the entire economy (42 percent vs. 16 percent in the case of FBH in 2009), the slightly higher average wage for the former may be justified. On the other hand, according to OECD data, even teachers with many years of experience earn much less than workers with tertiary education in other sectors, as illustrated in Table 6.6, so the higher proportion of university graduates among education employees might not be a sufficient reason for higher salaries for them. 90 Table 6.4: Average Wages for the Education Sector and Total Economy by Entity, 2006-2009 Net Wage as % of Education Sector Total Economy Education as % of Total Economy National Average Average Average Number of % with Gross Average Number of % with Gross Average Number of % with Average Persons Higher Wage Net Wage Persons Higher Wage Net Wage Persons Higher Gross Average Education Total 1 1 2 Employed Education (KM) (KM) Employed Education (KM) (KM) Employed Education Wage Net Wage Sector Economy 2009 FBiH total 36,446 41.8% 1,263 827 426,556 15.9% 1,204 792 8.5% +25.9% 104.9% 104.4% 97.0% 100.2% RS 18,886 n/a 1,377 901 258,634 15.8% 1,204 788 7.3% n/a 114.4% 114.3% 105.8% 99.7% FBiH + RS 55,332 n/a 1,302 852 685,190 15.8% 1,204 791 8.1% n/a 108.1% 107.8% 100.0% 100.0% 2008 FBiH total 35,413 38.2% 1,170 796 430,745 14.4% 1,105 751 8.2% +23.8% 105.9% 105.9% 93.4% 95.0% RS 18,196 n/a 1,243 832 259,205 14.1% 1,132 755 7.0% n/a 109.8% 110.2% 97.7% 95.5% FBiH + RS 53,609 n/a 1,195 808 689,950 14.3% 1,115 753 7.8% n/a 107.2% 107.4% 94.8% 95.2% 2007 FBiH total 34,751 35.3% n/a 683 413,676 14.3% 974 662 8.4% +21.1% n/a 103.2% 80.2% 83.8% RS 17,692 n/a 730 491 258,236 13.1% 875 585 6.9% n/a 83.4% 83.9% 57.6% 74.0% FBiH + RS 52,443 n/a n/a 618 671,912 13.8% 936 632 7.8% n/a n/a 97.7% 72.6% 80.0% 2006 FBiH total 34,526 33.0% n/a 613 389,601 13.5% 887 603 8.9% +19.5% n/a 101.7% 72.0% 76.3% RS 17,111 n/a 697 458 248,139 12.3% 793 521 6.9% n/a 87.9% 87.9% 53.8% 65.9% FBiH + RS 51,637 n/a n/a 562 637,740 13.1% 850 571 8.1% n/a n/a 98.4% 65.9% 72.3% Sources: FBH Federal Office of Statistics; RS Institute of Statistics. Notes:1 Average gross wage (net wage plus taxes and contributions) for FBH cantons is calculated on the assumption that the ratio of net wage to gross wage is constant across cantons. 2 Difference in percentage points between education sector and the total economy in terms of percent of workers with a university degree. 3 n/a = data is not available. 6.28. Figure 6.13a illustrates the Figure 6.12: Ratio of the Education Sector Average Wage ratio of the education sector average to Total Economy Average Wage, 2006-2009 wage to GDP per capita between 2006 1.2 and 2009 in both entities and the ratio RS of unweighted teachers‟ average 1.1 BiH wage58 to GDP per capita in FBH. The FBiH unweighted teachers‘ average wage was slightly lower than the overall average 1.0 for education sector employees in FBH throughout the period. Assuming that 0.9 this was also the case in RS, the teachers‘ average wages both in FBH and RS were about 80 percent higher 0.8 2006 2007 2008 2009 than GDP per capita in 2009. This is considerably higher compared to the Sources: FBH Federal Office of Statistics; RS Institute of Statistics international standards: the ratios of annual teachers‘ salaries after 15 years of experience (with minimum training) to GDP per capita were 58 The wage data for primary teachers originally provided by the trade union was presented in terms of the net wages, instead of the base wages and coefficients, but the net wages for primary teachers exactly match with those for secondary teachers which were presented in terms of the base wages and coefficients. Due to lack of detailed data on the teacher profile, simple, unweighted average wages were used in this report without taking into account the number of teachers in each wage category. 91 Table 6.5: Ratio of Salary After 15 Years of 1.16 (primary), 1.22 (lower secondary), and Experience (Minimum Training) to Earnings for 1.29 (upper secondary) for the OECD Full-Time Full-Year Workers with Tertiary average, and 1.12, 1.18, and 1.29, Education Aged 25 To 64 respectively, for the EU 19 average. Among the OECD countries, only upper secondary Lower Upper teachers with 15 years of experience in Primary secondary secondary German (1.82) and teachers at all levels in education education education Korea (2.02) earn as much as or more than Czech Republic 0.49 0.50 0.53 the teachers in BH in terms of their relative Finland 0.87 0.93 1.02 wages to GDP per capita. Hungary 0.50 0.50 0.60 6.29. However, the high level of non- Poland 0.59 0.68 0.78 observed economy in BH partially explains OECD avg. 0.77 0.79 0.86 the high wage levels relative to GDP per EU 19 avg.* 0.77 0.81 0.89 capita. Figure 6.13b shows the ratio of the Estonia 0.70 0.70 0.70 education sector average wage to adjusted Source: OECD, Education at a Glance 2010. * EU19 = These GDP per capita. It now appears that teachers 19 countries are Austria, Belgium, the Czech Republic, in FBH were paid at slightly higher than the Denmark, Finland, France, Germany, Greece, Hungary, Italy, OECD and EU 19 average in 2009, whereas Ireland, Luxembourg, the Netherlands, Poland, Portugal, the those in RS were still paid much higher. With Slovak Republic, Spain, Sweden and the United Kingdom. the most recent salary increases and tax exemptions for the public servants, including teachers (effective January 2011), teachers both in FBH and RS are now paid much higher than the OECD average relative to the size of their economy. Figure 6.13: Ratio of the Education Sector Wage to GDP Per Capita, 2006-2009 (a) Using official GDP figures (b) Using adjusted GDP figures 2.0 2.0 RS RS 1.8 1.8 BiH BiH 1.6 FBiH 1.6 FBiH 1.4 FBiH teachers FBiH teachers 1.4 only only 1.2 1.2 1.0 1.0 0.8 0.8 2006 2007 2008 2009 2006 2007 2008 2009 Sources: FBH Federal Office of Statistics and RS Institute of Statistics for education sector wages; and FBH teacher trade unions for teachers‘ salaries. 6.30. In order to grasp the adequacy of BH‟s wage levels in general, it would be useful to compare the ratio of the average wage in the economy to GDP per capita between the two entities, BH as a country, and two neighboring countries (Serbia and Croatia). As in the case for the education sector wage, Figures 6.14a and 6.14b illustrate the ratio of the economy-wide average wage to GDP per capita and to adjusted GDP per capita. Even when the adjusted GDP per capita is used, it is striking that workers in BH seem to be paid much more than those in Serbia and Croatia relative to their respective economies. Possible reasons for the relatively high wages in BH might include: the adjustment to GDP might not be sufficient and actual GDP could be higher; the labor supply for the formal sector jobs is limited due to the low enrollments at the upper secondary and tertiary levels, so that wages need to 92 be relatively high to attract them; the very high proportion of public sector jobs whose wage level is much higher than the economy-wide average is driving the overall average wage high. Figure 6.14: Ratio of the Economy-Wide Wage to GDP Per Capita, 2006-2009 (a) Using official GDP figures (b) Using adjusted GDP figures 1.8 1.8 1.6 1.6 RS RS 1.4 1.4 BiH BiH 1.2 FBiH 1.2 FBiH 1.0 Serbia 1.0 Serbia Croatia Croatia 0.8 0.8 2006 2007 2008 2009 2006 2007 2008 2009 Sources: Database of Economic Indicators of RS (http://www.irbrs.net/Statistika.aspx?tab=2&lang=eng) for the average wages and GDP per capita. 6.31. Within FBH, the average wages for education employees-both in absolute and relative terms-differ considerably between the cantons. Table 6.7 compares the average wage for education employees with the economy-wide average wage in each canton in 2009. Education employees on average are better paid than total employees on average within the respective cantons, but the differences vary considerably. On the one hand, even though the net average wage for the education sector was the highest in Sarajevo Canton (KM 951), it was only equivalent to 98 percent of the economy average. On the other hand, Central Bosnia Canton‘s net wage for education employees was one of the lowest (KM 755), but it was equivalent to 116 percent of the economy average. Table 6.6: Average Wages for the Education Sector and Total Economy by Canton, 2009 Education as % of Net Wage as % of Education Sector Total Economy Total Economy National Average Number of Average Average Number of Average Average Number of Persons Gross Wage Net Wage Persons Gross Wage Net Wage Persons Average Education Total Employed (KM)1 (KM) Employed (KM)1 (KM) Employed Net Wage Sector Economy FBiH cantons Una-Sana 3,950 1,216 796 32,794 1,123 739 12.0% 107.7% 93.4% 93.5% Posavina 459 1,129 739 5,166 1,060 698 8.9% 105.9% 86.8% 88.3% Tuzla 7,143 1,266 829 81,882 1,103 726 8.7% 114.2% 97.3% 91.8% Zenica-Doboj 6,011 1,065 697 70,907 1,004 661 8.5% 105.5% 81.8% 83.6% Bosnian Podrinje 409 1,207 790 4,896 1,085 714 8.4% 110.7% 92.8% 90.3% Central Bosnia 3,387 1,153 755 37,298 989 651 9.1% 116.0% 88.6% 82.3% Herzegovina-Neretva 4,008 1,295 847 42,018 1,355 892 9.5% 95.0% 99.5% 112.8% West Herzegovina 915 1,160 759 14,310 1,067 702 6.4% 108.1% 89.1% 88.8% Sarajevo 9,180 1,453 951 119,532 1,478 973 7.7% 97.8% 111.7% 123.1% Canton 10 984 1,278 837 9,753 1,148 755 10.1% 110.7% 98.2% 95.6% FBiH total 36,446 1,263 827 426,556 1,204 792 8.5% 104.4% 97.0% 100.2% RS 18,886 1,377 901 258,634 1,204 788 7.3% 114.3% 105.8% 99.7% FBiH + RS 55,332 1,302 852 685,190 1,204 791 8.1% 107.8% 100.0% 100.0% Sources: FBH Federal Office of Statistics; RS Institute of Statistics. Notes:1 Average gross wage (net wage plus taxes and contributions) for FBH cantons is calculated on the assumption that the ratio of net wage to gross wage is constant across cantons 93 6.32. Education employees‟ wages differ greatly between the two entities and across the cantons because the base salary, allowances, and coefficients for various types of job categories are annually determined based on the “collective agreementâ€? between the respective Ministries of Education and trade unions at the entity level in RS and at the cantonal level in FBH. The process is closely linked with the wage policy for civil servants. While all the allowances have been absorbed into wages in RS, a number of allowances such as meal per diem and winter clothing allowances remain in FBH, making the total recurrent spending on staff less obvious.59 6.33. In RS, salary coefficients for teachers are defined based on their qualifications and the level of education that they teach (primary or secondary) within the matrix for the education sector employees. Table 6.8 summarizes the salary coefficients for all primary and secondary education sector employees from principals of large schools at the highest, to night watchmen and cleaning staff at the lowest. There are only three wage categories for primary and secondary teachers, respectively, depending on their qualifications: (i) VII-level of professional training—university degree; (ii) VI-level of professional training—two-year college degree; and (iii) IV-level of professional training—secondary education. 59 The amount for meal per diem was one of the most easily adjusted budget items during the economic crisis: all cantonal governments negotiated with the trade unions to cut it by 20-60 percent in 2009 to cope with the budget shortage. 94 Table 6.7: Salary Coefficients for Primary and Secondary Education Sector Employees, RS, Effective Since 2008 Salary Salary sub- Salary Salary sub- Job Title Qualifications Coefficient Job Title Qualifications Coefficient group group group group Primary Education Sector Employees Secondary Education Sector Employees 1 1 Principal of a primary school with 800< pupils Univ. degree 12.0 1 1 Principal of a primary school with 800< pupils n.a. 12.0 ditto 2-year degree 11.5 2 Principal of a primary school with 401-800 pupils Univ. degree 11.5 2 Principal of a primary school with 401-800 pupils n.a. 11.5 ditto 2-year degree 11.0 3 Principal of a primary school with <400 pupils Univ. degree 11.0 3 Principal of a primary school with <400 pupils n.a. 11.0 ditto 2-year degree 10.5 2 1 Assistant principal VII level of prof. training - 2 1 Assistant principal VII level of prof. training - 9.5 9.5 univ. degree univ. degree 2 Pedagogue, psychologist, logopedist, special VII level of prof. training - 2 Secondary school teacher , pedagogue, VII level of prof. training - education teacher, social worker univ. degree 9.0 psychologist, logopedist, special education univ. degree 9.0 teacher, social worker 3 Secondary school teacher , librarian, secretary, VII level of prof. training - 3 Librarian, secretary, accountant, out-of-school VII level of prof. training - 8.5 8.5 accountant univ. degree practice coordinator univ. degree 3 1 Assistant principal VI level of prof. training - 3 8.0 2-year degree 2 Primary school teacher, teacher (who has VI level of prof. training - 1 Practice teacher VI level of prof. training - completed a teacher’s academy and a trained 2-year degree 7.5 2-year degree 7.5 teacher for music and art classes) 3 Librarian, secretary, accountant VI level of prof. training - 2 Librarian, secretary, accountant, out-of-school VI level of prof. training - 7.0 7.0 2-year degree practice coordinator 2-year degree 4 1 Librarian, secretary, accountant, teacher (who IV level of prof. training - 4 1 Librarian, secretary, accountant IV level of prof. training - 6.0 6.0 has not completed a teacher’s academy) secondary educ. secondary educ. Practice teacher , out-of-school practice IV level of prof. training - coordinator secondary educ.; or 6.0 V level of prof. training - highly qualified 2 Administrative-finance worker, assistant librarian, IV level of prof. training - 2 Assistant librarian, programmer, operator (EMIS IV level of prof. training - operator secondary educ. 5.5 and other jobs), administrative worker, finance secondary educ. 5.5 worker 3 Janitor, driver, heating maintenance mechanic V level of prof. training 3 Procurement cleark, janitor, heating maintenance V level of prof. training (highly qualified) or mechanic (highly qualified) or 5.0 5.0 IV level of prof. training IV level of prof. training - (qualified) secondary educ. 5 1 Janitor, heating maintenance mechanic III level of prof. training 5 1 Janitor, heating maintenance mechanic III level of prof. training 4.5 4.5 (qualified) (qualified) 2 Night watchman, cleaning lady I level of prof. training - 2 Night watchman, cleaning lady I level of prof. training - primary educ. or non- 3.5 primary educ. or non- 3.5 qualified qualified Source: Republic Srpska, Official Gazette of Republica Srpska, No. 28/94, I issue, Decree on Promulgation of the Law on Salaries of Republica Srpska Education and Culture Sector Employees dated December 29, 2007. Notes: 1/ The labor cost is the value of the simplest work and determined by the Governent of Republic Srpska and the Union of Education, Science and Culture of Republica Srpska. 2/ The basic salary is the product of the labor cost and a coefficient for the corresponding salary group and sub-group. 3/ The basic salary includes allowances for meal, food provisions for winter, firewood/fuel, and leaves. 4/ The basic salary increases by 0.5 percent for each full year of experience. 95 6.34. In FBH, the wage categories for teachers are determined based on their qualifications in all cantons except Central Bosnia Canton where the qualification categories were replaced with the year of experience (less than 10 years; 10-20 years; and more than 20 years) in 2008. Unlike in RS, there is no difference in the salary coefficients between primary and secondary teachers. While the wage categories are the same across the cantons, the base wage and coefficients, and, as a result, the net wages vary considerably between cantons, as shown in Table 6.9. For instance, in 2009, the base wage for teachers was KM 104 with coefficients between 6.7 and 9.0 (or the net wage of KM 697-986) in Sarajevo Canton, compared to KM 130 with coefficients between 4.5 and 6.0 (or the net wage of KM 585-780) in Posavina Canton. Table 6.8: Avg. Wage for Teachers, All Employees in School, and All Employees in Canton by Canton, 2009 Coefficients Wages for teachers Avg. wage Avg. wage for all for all Avg. wage for Avg. wage employees employees Avg. wage for all employees for all in canton in canton teachers in school VI level VI level Ratio of the Avg. wage Avg. wage employees Minimum Avg. wage relative to relative to relative to relative to IV level (2-year VII level IV level (2-year VII level highest Unweighted for teachers for all in school wage for all for all all all avg. wage for avg. wage for (secondar college (univ. Base (secondar college (univ. wage to the avg. for all relative to employees relative to employees employees employees employees all employees all employees y educ.) degree) degree) wage y educ.) degree) degree) lowest teachers FBiH avg. in school FBiH avg. in school in canton in FBiH in 8 cantons in canton in canton Canton A B C D DxA DxB DxC C/A D D/783 E E/697 F G G/785 G/765 D/G E/G Una-Sana - 1.61 1.83 490 - 789 897 - 843 1.08 745 1.07 549 720 0.92 0.94 1.17 1.04 Posavina 4.50 5.10 6.00 130 585 663 780 1.33 676 0.86 595 0.85 351 690 0.88 0.90 0.98 0.86 Tuzla - - - - - - - - - - - - - - - - - - Zenica-Doboj 1.70 1.70 1.70 400 680 680 680 1.00 680 0.87 610 0.88 400 660 0.84 0.86 1.03 0.92 Bosnian Podrinje 1.60 2.10 2.40 360 576 756 864 1.50 732 0.94 648 0.93 396 686 0.87 0.90 1.07 0.94 Central Bosnia - - - - - - - - - - - - - - - - - - Herz.-Neretva 1.85 1.85 1.85 477 882 882 882 1.00 882 1.13 802 1.15 559 878 1.12 1.15 1.00 0.91 West Herz. - - - - - - - - - - - - - - - - - - Sarajevo 6.70 7.60 9.00 104 697 790 936 1.34 808 1.03 728 1.04 489 962 1.23 1.26 0.84 0.76 Canton 10 2.40 2.40 2.60 344 827 827 895 1.08 850 1.09 749 1.08 448 759 0.97 0.99 1.12 0.99 Year of experience <10 yrs 10-20 yrs 20 yrs< Central Bosnia 1.68 1.72 1.77 459 771 790 813 1.05 791 1.01 731 1.05 550 658 0.84 0.86 1.20 1.11 Total FBIH 785 1.00 1.03 0.00 c.f.) Avg. 8 cantons 717 772 843 783 1.00 697 1.00 765 0.97 1.00 1.02 0.91 Source: Federation Primary Teacher Trade Union. 6.35. The wage differences between cantons can partially, but not fully, be explained by the differences in the level of economic development as measured in the economy-wide average wage in each canton. Figure 6.15 illustrates the trends in the unweighted average wages for teachers relative to the cantonal-wide average wages for each canton. It is interesting that while the cantonal-wide average wages grew steadily in almost all cantons between 2006 and 2009 except for a slight slowdown in 2009 due to the economic crisis, teacher salaries did not grow in the same manner in all cantons. As a result, the ratios of the average wage for teachers to the average wage for all employees in the respective cantons are quite different between cantons. While the ratio of teachers‘ average wage to the cantonal-wide average wage dropped in Sarajevo and Posavina Cantons, that in Una-Sana Canton and Canton 10 increased between 2006 and 2008 before dropping slightly in 2009. This may suggest that teachers in the latter group may be overpaid compared to those in other cantons. 96 Figure 6.15: Unweighted Average Wage for Primary and Secondary Teachers Relative to Cantonal- Wide Average Wage by Canton, 2006-2009 Unweighted avg. wage for primary and secondary Average wage for all employees in canton by canton, teachers, 2006-2009 2006-2009 1,000 1,000 Sarajevo Sarajevo 900 Tuzla 900 Herz.-Neretva Herz.-Neretva FBiH average 800 800 Canton 10 Posavina KM KM Una-Sana Canton 10 700 700 FBiH average Una-Sana 600 600 Central Bosnia West Herz. Posavina Bosnian Podrinje 500 500 Zenica-Doboj Zenica-Doboj 400 400 Bosnian Podrinje Central Bosnia 2006 2007 2008 2009 2006 2007 2008 2009 Ratio of unweighted avg. wage for teachers to avg. wage for all employees in canton, 2006-2009 1.3 Ratio of avg. teacher wage to avg. cantonal wage Sarajevo Tuzla 1.2 Herz.-Neretva Canton 10 1.1 Una-Sana 1.0 FBiH average Central Bosnia 0.9 Posavina Zenica-Doboj 0.8 Bosnian Podrinje 2006 2007 2008 2009 Source: Federation teacher unions. 6.36. Table 6.9 above also illustrates the degree of wage differentiation among the teaching profession. Pay scales for teachers need to be designed in such a way that they can attract and retain highly qualified teachers, but the current salary coefficients for teachers in BH seem to be too compressed to fairly reflect their qualifications and teaching requirements . First, secondary teachers are generally paid more than primary teachers, but there is no wage difference between primary and secondary teachers in FBH, and a less than five percent difference in RS. This is not in line with international standards. For instance, the ratio of salary per teaching hour of upper secondary to primary teachers (after 15 years of experience) is 1.39 for the OECD average, and 1.43 for the EU 19 average, ranging from 0.91 for England, 1.00 for Scotland to 1.81 for Netherlands and 1.70 for Luxembourg. Second, the differences based on the qualifications of teachers may be appropriate on average, but they vary considerably across the cantons in FBH. Some cantons (Herzegovina-Neretva and Zenica-Doboj Cantons) have a flat pay scale, whereas the highest paid teachers in Bosnian Podrinje Canton earn 50 percent more than the lowest. Standards and norms for staffing 6.37. The RS and each canton in FBH set standards and norms on staffing, defining the minimum, optimal, and maximum class sizes, the number of teaching hours, the number of support staff primarily based on the number of students, recurrent expenses by different teaching system. Table 6.10 shows the staffing standards and norms which in turn define school budgets in RS. If these norms are strictly applied, for instance, schools with 60 students are expected to form only two classes with 30 students each, instead of three classes with 20 students each. However, the guidelines also note that ―in exceptional cases, the Ministry can permit a class with less than 18 pupils, if it is the only class in 97 one grade and a combined class has more than the norms‖. Without detailed analysis of school-level data, it is difficult to form a clear picture as to how these norms are applied in practice, but given that the average school size for RS is very small at 144 (or 16 students per grade on average) (see Table 6.9 above), it is likely that many schools are given an ―exceptional‖ status. Table 6.9: Factors Determining School Budget for Primary Schools 1 No. of classes 2. No. of pupils 1. No. of 2 No. of pupils 1/ classes a No. of pupils in classes 1 18-32 2 33-60 3 61-90 4 91-120 5 121-150 6 151-180 7 181-210 8 211-240 9 241-279 b No. of pupils in combined classes 1 merging two grades up to 18 2 merging three grades up to 12 c No. of pupils in special education classes 1 up to 10 3 No. of lessons in a class in accordance with the curriculum No. of teachers = Class hours (curriculum) / Teacher's weekly work hours (40 4 No. of lessons planned per class hours) 5 No. of working hours of non-teaching staff in line with the Coefficients Pedagogue 1 for school with 16 or more classes 0.05 per class < 0.5 for school with 16 or less classes Psychologist 1 for school with 24 or more classes May be combined with special education teacher/logopedist, or social worker 0.05 per class < 0.5 for school with 24 or less classes Assistant principal 0.025 per class exceeding 24 classes or 8 or more branch classes 0.5 for school with 2 or more nine-grade branch classes Librarian 1 for school with 16-32 classes and 5,000 units of literary and non-literary 0.05 per class <0.5 for school with 16 or less classes Assistant librarian 0.05 per class exceeding 32 classes and 10,000 units of literary and non-literary Secretary 1 for school with 16 or more classes 0.05 per class <0.5 for school with 16 or less classes Accountant 1 for school with 16 or more classes 0.05 per class <0.5 for school with 16 or less classes Administrative-finance worker 1 for school with 24 or more classes 0.05 per class exceeding 24 classes Janitor 1 for school with 16 or more classes and an area between 2,000-5,000 m2 0.05 per 100m2 or per each class below 16 < 0.5 or less than 2,000m2 0.05 per 100m2 or per each class exceeding 32 < 1 or more than 5,000 m2 Night watchman 1 per school facility Heating maintenance mechanic 1 for school with central heating for up to 2,000m2 2 for school with central heating for more than 4,000m2 Transportation 1 for school with no organized transportation Hygeine 0.25 per class Maintenance 1 per school facility 6 Years of past work experience of employees 7 Salary coefficients 8 Labor cost 9 Increase according to working conditions Source: Ministry of Education and Culture of the Republika Srpska, Rules on Primary School Funding. Notes: 1/ In exceptional cases, the Ministry can permit a class with less than 18 pupils if it is the only class in one grade and a combined class has more than the norms. 2/ Schools in mountainous areas and extremely undeveloped municipalities may form smaller class sizes if formed in the "most cost-effective manner". 98 6.38. Table 6.11 summarizes the minimum, optimum, and maximum class sizes and teaching loads by subject in six cantons in FBH. There are three important issues to be addressed. First, it is interesting that all cantons (where data are available) Table 6.10: Standards for Class Sizes and Teaching Loads by specify the ―optimum‖ class Canton, 2011 Class size (regular with no Teaching loads (hours), excl. preparation, size, in addition to the minimum combined classes) correction, evaluation hours and maximum sizes. This Foreign custom seems to discourage languages, History, National math, Biology, geography, schools from forming classes Min. Optimum Max. languages chemistry informatics music, etc. beyond the optimum size even Una-Sana 18 27 35 18 19 20 21 if they are below the maximum Posavina 17 25 33 18 19 19 20 size. Since there is no clear Tuzla 22 28 34 20 20 21 22 Zenica-Doboj 18 26 36 18 19 20 21 evidence to suggest that these Bosnian Podrinje 16 24 32 - - - - optimum sizes result in better Central Bosnia - - - - - - - education than anywhere Herz.-Neretva - - - - - - - between the optimum and West Herz. - - - - - - - Sarajevo 18 24 32 18 19 20 21 maximum sizes, by removing Canton 10 - - - - - - - the optimum sizes, schools may Source: Ministry of Education of the respective cantons. form slightly larger classes Note: - indicates that data are not available. efficiently without harming quality. Second, there is no clear rationale for the wide variations in the minimum class size from 16 (Bosnian Podrinje Canton) to 22 (Tuzla Canton). Cantons which set the minimum class size lower than others could increase it, unless there is clear justification such as geographical constraints. Third, teaching loads are almost the same for all cantons except for Tuzla Canton. By increasing the teaching hours by one or two hours, other cantons would be able to reduce the number of teachers, and therefore the wage bills for teachers, by 5-10 percent. In order to analyze the correlations between costs and quality, more detailed school-level data such as per student spending, wage bills, and student performance are needed. V. RECOMMENDATIONS 6.39. Total BH spending on education as a percentage of GDP is close to the regional average and the scale of spending does not need to be adjusted significantly. However, the available resources could be spent much more efficiently. Altogether, education spending appears to be adequate or only slightly lower than it should be, given the current level of economic development in BH. However, BH‘s highly fragmented government structure makes it difficult for this expenditure to translate into effective and efficient provision of education. The room for efficiency gains is reflected in the declining average class sizes and student-teacher ratios. Even though these efficiency indicators are not yet at an alarming level compared to other countries in the region, as enrollments will continue declining it is now time to improve the school network and staffing. So far progress has been very slow. Since the 2006 PEIR made various recommendations on educational reforms, very few actions have taken place. As in other sectors part of the problem rests with constitutional framework that does not empower the State government and FBH government in the case of FBH to implement nationwide or entity-wide reforms. However, the need for reforms is becoming more urgent and these should be built around the following recommendations: 6.40. Limit further wage increases. The recent increase in education budget has primarily been driven by the sharp increases in wages for primary and secondary education sector employees, particularly in RS. This is not only fiscally unsustainable, but also has prohibited non-wage spending and capital investment from growing. It appears that teachers are already slightly (FBH on average) or significantly (RS) overpaid in terms of the ratio of their wages to GDP per capita, and therefore, there is no 99 justification for further increases in the wage bills. It is the time for RS and cantonal governments to stop increasing wages and start spending the available budget more efficiently and effectively to improve the quality. 6.41. Develop a more decompressed wage structure. The new structure should take into account teachers‘ qualifications, experience, performance and other criteria to better attract and retain highly qualified teachers. 6.42. Rebalance the spending between wage and non-wage spending for primary and secondary education to improve the quality and between levels of government to enhance pre-school and tertiary education. While the country as a whole does not overspend on education there is a clear need to reallocate resources within the sector. Wage expenditures should be restrained and non-wage expenditures increased. Also, it is time to improve the efficiency of the system, especially by responding to the declining student enrollments at the general education level. More should be invested in pre-school and tertiary education which has been neglected, as clearly indicated in the extremely low enrollment rates at those levels. In particular, investment in rapidly growing higher education is essential for economic growth. 6.43. Address the access issues in upper secondary which are at the root of the low enrolment rates there. While the primary and lower secondary enrollment rates are reasonably high, the rate for upper secondary is very low compared to the neighboring countries. The government needs to investigate the reasons for such low rates at the upper secondary level and take necessary actions to improve enrollments. 6.44. Implement the 2006 PEIR recommendation of introducing per capita financing of schools to replace the existing input-based financing.60 School financing based on the number of students (output- based) is expected to help enhance efficiency, equity, and transparency in school financing, if implemented rigorously. While the entity and some cantonal governments were interested in introducing it at some point and were provided various pieces of analysis and training on the new financing scheme, it was never implemented possibly because the entity and cantonal governments did not feel the urgency to economize education spending. The entity and cantonal governments may wish to reconsider introducing this financing scheme to achieve the various objectives which the new financing scheme is expected to support. 6.45. Establish an education management information system. There are several differences in the characteristics of education spending and efficiency measures between the entities and between the cantons in FBH that should be reflected upon in forming future policies, and this will require significant new efforts at data collection. Some cantons are more efficient (e.g., higher student-teacher ratios, higher average class sizes, higher teaching loads, slower wage increases, etc.) than others. In order to improve the efficiency and equity in school financing in practice, more detailed analysis of school-level data is essential. For instance, without knowing geographic, socio-economic, and demographic conditions, it is difficult to judge what would be an appropriate class size or student-teacher ratio for any particular school or canton. Also, information on students‘ learning outcomes based on standardized national student assessment is essential to measure how efficient and effective the sector is. At present, both RS and FBH publish statistical bulletins, but more detailed data should be collected and analyzed so that useful policy recommendations can be made for policy makers. As the country‘s education system is likely to continue 60 The new financing scheme has indeed been widely used to improve efficiency, effectiveness, and transparency of school financing that many countries in the region have introduced, or have been trying to introduce. See Alonso, Juan Diego and Sanchez, Alonso, eds., forthcoming, Reforming Education Finance in Transition Countries: Six Case Studies in Per Capita Financing System, The World Bank: Washington, DC. 100 being fragmented, it is essential that the policy makers at the entity and cantonal levels have access to the most complete, accurate, and up-to-date information in order to make informed-decisions to improve the quality of education (see Annex 6.2 for detailed recommendations on improving data collection practices in education). 101 ANNEX 6.1: MAIN FINDINGS AND RECOMMENDATION ON HIGHER EDUCATION FINANCING 6.46. Public expenditure on higher education is comparable with other countries in the region. In 2007, BH spent 0.5 percent of GDP on higher education. This is low by EU-19 standards (around 1 percent of GDP), but is in line with EU candidate and potential candidate countries, including Albania, Bulgaria, Croatia, Macedonia and Romania.61 6.47. In FBH, six public universities are located across five cantons. Although university students may come from any canton, it is the receiving cantons that finance most of the costs associated with higher education. The sending cantons transfer a relatively small amount to the receiving cantons typically to subsidize for boarding costs of their students. As a result, there is a large amount of ―cross-subsidy‖ between cantons. According to the EU study, two Sarajevo and Tuzla end up subsidizing other cantons while the others are subsidized to a different extent.62 6.48. Four policy options to modify the canton-based system of higher education financing in FBH include: (i) Ensure that cantons are funded through an appropriate change in the revenue distribution formula so that there is an equitable and efficient allocation for funding higher education from each canton; (ii) Transfer the funding of FBH universities to the entity, removing the need to finance universities through the cantons; (iii) Transfer the funding of all universities in FBH (and RS) to a State agency or ministry, so that neither entities nor cantons are involved in university financing; (iv) Move from funding institutions to funding students, so that each student receives an earmarked amount of money equal to average tuition costs which he/she is free to use at the university of his or her choice. Source: European Union, 2009, Feasibility Study for Higher Education Financing Reform, Final Report, March 2009. 61 Linden, T. and Arnhold, N. with Vasiliev, K., 2008, From Fragmentation to Cooperation: Tertiary Education, Research and Development in South Eastern Europe, The World Bank, Education Working Paper Series, No. 13. 62 The method for estimating the amount of cross-subsidies may not be most accurate, but the study provides a general picture of the cross-subsidies between the cantons. 102 ANNEX 6.2: RECOMMENDATIONS FOR IMPROVING DATA COLLECTION PRACTICES IN EDUCATION 6.49. During the course of this PEIR, the education team has identified numerous shortcomings in the way in which authorities in BH collect and publish data that are relevant for analyzing the country‘s education sector. This annex presents an overview of the current state of education data collection; discusses additional kinds of data that would be useful for policymakers and sector specialists; and provides some examples of analytical work that can be conducted if additional data were available. Current state of education data collection in Bosnia 6.50. At present, the main actors tasked with collecting and disseminating education sector data in BH are the official statistical agencies of the two entities—the Federal Office of Statistics in FBH and the Institute of Statistics in RS. These organizations collect information from schools and other educational institutions and publish annual bulletins on each education sub-sector (pre-school, primary, secondary, and higher education). These bulletins are available on the agencies‘ respective websit es63 (in PDF files up to 150 pages in length) and contain data on the numbers of institutions, students, and staff in aggregate, as well as in geographic and other aspects for each entity. Such information is useful but far from sufficient for a thorough analysis of the education sector. 6.51. In the early 2000s, an attempt to establish a national Education Management Information System (EMIS) was made with support from the World Bank. However, at present a functional EMIS exists in only two cantons of FBH (Sarajevo and Tuzla) and a less comprehensive one in RS. In most cantons, the EMIS was either never fully implemented or has fallen out of use. The most advanced EMIS in BH is maintained by the Sarajevo Canton; the system collects data at school level and allows for ―online access [to data] on the condition of the schools, classes and each student at the end of the semester, and at the end of the school year,‖ as well as for printing of relevant certificates.64 6.52. Outside of the two or three Ministries of Education that rely on EMIS software, no detailed information on school characteristics is available in BH. Furthermore, BH lacks a national system of standardized student assessment, so no comparisons are possible of the quality of educational outcomes across schools, cantons, entities, or over time. Due to these shortcomings, it is impossible for policymakers to relate information on educational inputs (including school-level expenditures) to information on sector outcomes. Therefore, in order to examine the education sector in a serious analytical way, and to formulate better-informed policy decisions, additional data are required. Additional data needed to inform better education policies 6.53. To formulate the best policies governing education, policymakers‘ decisions must be informed by rigorous analyses of the data obtained from the sector. Because aggregate-level statistical data are not sufficient for a thorough analysis, detailed school-level data must be collected and made available. A set of comparable indicators must be gathered at least once a year from educational institutions across entities and cantons. This can be achieved through a full implementation of a national EMIS or (as a second-best approach) through the expansion of statistical data collection and improvement of data dissemination practices by the respective statistical agencies.65 63 See: http://www.fzs.ba/ and http://www.rzs.rs.ba/. 64 See: http://www.emis.edu.ba/. 65 Data dissemination can be improved by sharing data in Microsoft Excel files or other database formats; it can also benefit from the development of a national or entity-level statistical database from which relevant data can be 103 6.54. Several types of education-related data must be gathered in one database or a series of databases that can easily be linked to each other. These include, but are not limited to: ï‚· School characteristics: in addition to the total numbers of students, classes, and teachers, more detailed information must be collected on school facilities and equipment, school- based staff (with details on staff positions, qualifications, pay grades, etc.), and student backgrounds (including ethnicity, language of instruction, demographics, etc.) ï‚· Education financing data: school-level expenditure data must be gathered and analyzed in conjunction with information on other educational inputs. Such information is often collected through EMIS reporting or, ideally, through linking of education systems to official treasury data. ï‚· Demographic and community data: information on community demographics is essential for developing sound education policy. For example, knowing the poverty or relative deprivation levels of students from different communities can help better target educational resources to reach disadvantaged students from communities with the greatest need. Such data can be collected through EMIS reporting or, ideally, through linking of education systems to official statistical data (such as community-level census or survey data). ï‚· National student assessments: a system of standardized student assessments must be instituted to allow measurment of the quality of education being delivered in BH. Without comparable, consistent, and comprehensive assessments, policymakers will never know what kind of results are being achieved as a return on education investments. 6.55. In the absence of centralized national institutions, some of the data collection practices described above can be implemented at the entity or canton level. However, state-level coordination is desirable for setting data collection standards and minimizing duplication of efforts that will arise if each Ministry of Education intends on developing its own distinct data collection framework. However, some key systems (such as standardized student assessments) must be developed at the national level if they are to serve a useful purpose for decision-making in different parts of BH. Analytical implications of better data collection 6.56. If the systems described above are successfully implemented, policymakers throughout BH will be able to base their decisions concerning education on better information rooted in more thorough sector analysis. Examples focusing on several areas of relevant analyses are outlined below. 6.57. Equity: linking resource allocation with need. In order to examine the equity of educational resource allocation, policymakers must have sufficient information on the background characteristics of the children enrolled in school (as well as those who are of school age but do not attend school). By collecting such information directly though an EMIS or linking information on educational inputs to community data from the census or from statistical surveys, policymakers can ascertain whether a particular allocation of resources manifests in an equitable distribution of effort to provide quality education to children from all areas and backgrounds. One approach is to analyze the variation in per- student expenditures across communities with different demographic or socioeconomic characteristics. queried by users in a simple and straightforward manner. For an example of one such database, see the Local Data Bank from Poland‘s Central Statistical Office: http://www.stat.gov.pl/bdlen/app/strona.html?p_name=indeks. 104 6.58. Another is to study whether the most qualified teachers are evenly distributed across schools in urban and rural areas. To do this, detailed information is needed on children and their communities. 6.59. Allocative efficiency: linking resource allocation with demand. Due to the changing demand for education service provision in BH, policymakers must be able to examine the alignment of resource allocation with demands of the various sub-sectors. Currently, BH is faced with growing enrollment in pre-schools and in institutions of higher learning; meanwhile, the number of students in primary and secondary education has begun a slow decline in line with demographic trends exhibited by many countries in the region. Education officials, therefore, must align sector priorities—and sector resources— with the shifting demands of the population. In order to do this, information must be available on current resource allocations and on the recent trends and future projections of enrollment, birth rates, labor market conditions, etc. 6.60. Technical efficiency: linking resource allocation with performance. Knowing whether existing resources are being used efficiently in providing quality education is a question on the minds of all policymakers. Therefore, analyses of whether the numbers of schools and staff—as well as the levels of expenditures on salaries, teaching materials, technology, building maintenance, and capital improvements—produce better results in some places than in others can be highly informative. Conducting relative efficiency analyses66 across different schools, municipalities, and cantons can shed light on areas where the mix of educational inputs—and the policies that govern them—contribute to higher learning outcomes. High performing and highly efficient schools can then serve as examples for others from which valuable lessons about best practices can be learned. To do this, however, detailed information on the level of inputs (including school-level spending) and outcomes (such as standardized student assessments) is required. 6.61. Mentioned above are but a few examples of issues that need to be kept in mind by BH‘s policymakers. To conduct rigorous analyses of these issues, more detailed information about the education sector is required. Therefore, better data collection practices must be put in place to ensure that future researchers and decision-makers are able to easily link sector inputs to performance; conduct comparisons across cantons and municipalities on the levels of need, resource provision, and efficiency of resource utilization; and provide useful advice on how to rationally manage budgets and sector resources in response to changing demographic and economic trends. 66 Data envelopment analysis (DEA) is one technique that allows for comparisons of relative efficiency across sub- national decision-making units. 105 CHAPTER 7: FORESTRY I. INTRODUCTION 7.1. A recent survey (the 2010 State Forest Inventory) indicates that BH forest cover is much larger, and the forests in much better condition, than previously thought. Consequently the sustainable yield is now reckoned to be about 11 million m3 compared with a current annual cut of only about 6–7 million m3. Hence the forestry sector contribution to the economy could be increased by increasing timber harvesting. However the government has a major presence in the sector and a number of changes in how it operates are needed to unlock this potential. The benefits of this should not be overstated. Forestry presently contributes about 2.5 percent of GDP, including the wood industry, so even a significant scaling up of harvesting would not add more than 1.5 percent to GDP. However this is not a trivial amount and moreover the problems that exist in the sector are representative of challenges of public expenditure and management more broadly. I - Forestry Sector 7.2. Forests and forest land cover account for Figure 7.1: Forest Cover approximately 52 percent of the overall land area of BH. Broadleaves predominate (68.8 percent) with the main species being beech and oak. Indications from the completed but unpublished State Forest Inventory are that the total forest area is circa 2.7 million ha and that the growing stock is underestimated. The annual allowable cut (AAC) for publicly owned forests in Bosnia and Herzegovina (BH) is approximately 6 million m3 being almost identical within FBHFBH and RS. This compares with an estimated annual increment of 4.17 million in FBH and 5.06 million in RS. Removals represent 56 percent of increment in comparison with an Source: Dennison 2006.67 EU-27 average of 60 percent. Private forests represent 18.2 percent of the forest area in FBH and 22 percent in RS. Private forests, with over 500,000 owners, are extremely fragmented with an average area of less than 1 ha and contribute relatively little to harvested volumes. Private owners are not organized. Table 7.1: Forest and Forest Land Area (ha) – (Dennison 2006 and Tosterud and Nordberg68 2011) FBiH RS BiH Fore st Ca te gory Are a Are a Are a % High Forest 519,140 464,507 983,647 43% Degraded High Forest 18,732 22,302 41,034 2% Plantations 65,121 59,726 124,847 5% Low (Coppice) Forest 279,436 178,379 457,815 20% Available for Refor 188,503 206,793 395,296 17% Unvailabale for Refor 102,041 50,711 152,752 7% Mined Area 128,217 18,263 146,480 6% Tota l 1,301,190 1,000,680 2,301,870 100% 67 Dennison, S. (2006). Forest and Wood Sector Review – Bosnia and Herzegovina. USAID Cluster project, January 2006, Sarajevo. 68 Tosterud, A. and Nordberg, M. (2011) Sawlog Market Assessment. USAID/SIDA FIRMA, Sarajevo, March 2011. 107 7.3. The management of forests is carried out in accordance with valid 10-year forest management plans (FMPs) irrespective of size or ownership. The FMPs in FBH are prepared by specialized forest management planning companies while in RS, the Research and Development and Project Centre within JPS Å ume RS prepares the FMPs. Plans are approved by the Entity forest authority (Federal Forest Office within FBH and Forest Agency within RS) and their implementation is mandatory. In 2009 all Forest Management Units within JPS Å ume RS received Forest Stewardship Council (FSC) certification. Progress in FBH has been relatively slow and an Action Plan for forest certification was adopted in May 2010. Three cantonal forest management companies (CFMCs) have recently undergone certification pre-assessment and assessment. 7.4. The regulatory framework is somewhat complicated. Within RS, the Forest Law (2008) provides the overall framework and is supported by a series of regulations relating to timber sales and technical norms of forest management. In FBH, the Forest Law was abolished by the Constitutional Court in 2009 following a petition by the mayor of the Konjic municipality and the Council of Municipalities and Cities of FBH for the protection of the right to local self-governance. Forest management in FBH is currently governed under a temporary regulation. A new forest law has been prepared but needs to go through Parliamentary procedure. 7.5. The harvesting infrastructure is relatively poor in both Entities with a high proportion of small scale harvesting contractors. There is a lack of investment in new machines and technology due to (a) limited access to finance, (b) lack of continuity of work and (c) low profit margins. The average road density of 9.4 m/ha (10.1 m/ha in FBH and 9.05 m/ha in RS) is significantly below other European countries with broadly similar topography e.g. Austria 36 m/ha and Switzerland 40 m/ha. The low density and condition of the road network has consequences for harvesting and transport resulting in higher costs and can lead to environmental damage. 7.6. The wood industry has undergone a dramatic change in structure and ownership since the war. The sector was privatized resulting in the current position where there are an estimated 1,500 sawmills compared with 61 before the war and significant processing overcapacity (Tosterud and Nordberg 2011). The smaller sawmills are typically unable to produce high quality sawnwood for export, but are mainly for the local market. 7.7. The contribution of the forestry sector, including wood industry, to GDP has fallen in recent years in line with a downturn in the economy and for 2009 was of the order of 1.9 percent (FBH) to 2.5 percent (RS) (Table 7.2). The sector is a significant employer, especially in rural areas, representing 5 percent of total employment in BH in 2009. Although employment fell in 2009, the decrease was modest given the downturn in the overall economy. The BH is a net exporter of forest products and the balance of exports over imports is of the order of KM 400 million for the period 2008-2009. This is in marked contrast to the overall economy where imports exceed exports by a factor of circa 2.4 in recent years. Sawnwood is the major export followed closely by furniture. Table 7.2: Value Added as percent GDP 2007 2008 2009 Sector FBiH RS FBiH RS FBiH RS Forestry Logging and Related Services 0.85 1.7 0.85 1.7 0.66 1.2 Manufactire of Wood and Wood Products 0.77 1.6 0.78 1.6 0.65 0.9 Manufacture of Furniture 0.63 0.5 0.58 0.4 0.56 0.4 Source: World Bank consultant estimates. 108 II. STRATEGIC, POLICY AND INSTITUTIONAL FRAMEWORK (a) Strategic and Policy Framework 7.8. Under BH Constitution, the ownership of public forests rests with the two Entities. There is no BH national forest policy and responsibility for policy and strategy rests at the Entity level. 7.9. In FBH, the preparation of a Forest Program (FP) began in December 2009. It has been in part funded by the Federal Ministry of Agriculture, Water Management and Forestry (MAWMF) and the World Bank Forest Development and Conservation Project (FDCP). The FP process, based on identified issues, commissioned 22 studies each providing a detailed action plan, monitoring indicators. The expected completion date is September 2011. 7.10. In RS, a slightly different approach has been adopted for the preparation of the Strategy for the Development of Forestry 2011-2020 The draft FP was prepared by a panel of experts appointed by the Ministry of Agriculture, Forestry and Water Management (MAFWM) under the direction and supervision of a Council which also evaluated the reports prepared by the panel of experts. A draft was completed in 2010 that was revised in 2011 to incorporate feedback. The final version will need wherever possible to quantify goals and include a formal procedure or protocol for monitoring and reporting on implementation. (b) Institutional Arrangements 7.11. There have been a number of changes in recent years. The more significant is in RS with the transformation of the Public Enterprise into a corporation in 2005, the reduction in the number of Forest Management Units (FMU) from 44 to 23 in 2008, the establishment of the Forest Agency as an intermediary between the Ministry and the JPS Å ume RS in 2008 and the incorporation of the Forest and Hunting Inspection into the overall Inspection Service. 7.12. In FBH changes have been less significant. The Federal Forest Inspection was transferred from the Federal Forest Office to the Federal Inspection Service in 2007 and at the Cantonal level, the Forest Inspection was transferred from the Cantonal Forest Office to the Cantonal Inspection Service in the same year. (c) Institutional Arrangements FBH 7.13. The ownership of the public forest resource rests with FBH which transfers the right to manage this resource to each of the Cantons. The Cantons transferred this right to the management companies which they established in compliance with the Law on Forests (2002). At the Federal level there is a Forestry Department within the Ministry of Agriculture, Water Management and Forestry with a unit responsible for legal matters and all aspects relating to forest law and related legislation and a Federal Forest Office (FFO) which deals with development and support and has an overall monitoring role. The FFO prepares an annual plan for the provision of financial support to forestry based on the premise that the share of the OKFÅ  due to the FFO will be available within the overall MAWMF budget allocation from the Ministry of Finance (MoF). Potential projects are evaluated by an in-house committee and the beneficiary enters into a contract with the MAWMF to perform the project activities with monitoring of implementation being undertaken by the relevant Cantonal Forest Office. 7.14. The Federal Forest Inspection (FFI) forms part of the overall Inspection Service. Its primary role is to monitor the implementation of all actions relating to the law on forests within FBH including the 109 payment or non-payment of the OKFÅ . The FFI has no formal link with the Cantonal Forest Inspection although they both collaborate on an informal basis. 7.15. At the Cantonal level, responsibility for forestry rests with the relevant Ministry within which there is a Cantonal Forest Office (CFO) whose main function is to control the activities of the CFMC and provide advice and support to private forest owners. Typically the CFO has a headquarters and a number of regionally based units encompassing one or more municipalities depending on area and extent of forest cover. The CFO has a major role in relation to the guarding and protection of the forest resource and this work is undertaken by the forest guards which they employ. The Cantonal Forest Inspection (CFI) forms part of the Cantonal Inspection Service. Their role is essentially the same as that of the FFI. 7.16. The staffing within the various units of the administration but in particular the Forestry Department is currently very low and represents only 60 percent of the planned compliment of staff. This has implications for the implementation of the FP. 7.17. Eight Cantonal Forest Management Companies (CFMC) have been established, the exceptions being Posavski which has a small forest area and HercegovaÄ?ko-Neretvanski which has a number of forest enterprises which are equivalent to regional management units (RMUs). Seven of these companies are established as 100 percent public companies owned by the Canton. The exception is Tuzla Canton where the CFMC JP Å ume TK is established as a joint stock company. The administrative basis for establishing these forest companies placed a number at a significant disadvantage in terms of forest resources due to the extent and composition (quality) of the underlying resource. The organizational structure of the CFMCs follows both formal (area based management units) and traditional (functional headquarter based units) lines. (d) Institutional Arrangements Republika Srpska 7.18. The Forestry Department within the Ministry of Agriculture, Forestry and Water Management is responsible for forests and forestry. It is assisted by the Forestry Council which is a forum for high level discussion on forestry and forestry related issues and developments and comprises representatives of the Ministry, other state bodies, institutions and organizations which are related with forest sector. 7.19. The Forest Agency established under the Law on Forest 2008 has a broad remit. This includes technical aspects of forestry, cadastre, record keeping, forest health monitoring, allocation of special funds, promotion of forestry, and support for private owners. It has a role in relation to the control and monitoring of executive works of public forest enterprises. While the Forest Agency has a broad remit, in practice it essentially acts as an intermediary between the Ministry and JPS Å ume RS. As an organization it is still in a developmental stage and has not as yet a full staff compliment. 7.20. The Forest and Hunting Inspection (FHI) is now within the general Inspection Service. It carries out forest control measures for both public and private owned forests based on ten year and annual forest management plans. 7.21. JPS Å ume RS manages the public forest. It has a traditional organizational structure with headquarters, twenty three Forest Management Units, a Research Development and Design Centre which undertakes forest management planning, a Centre for Seedling Production and a Karst Management Centre. 110 7.22. The FMUs report to the headquarters and are managed on an areas basis, comprising a number of forest Districts which in turn comprise a number of management sub units. The functional units providing services e.g. finance and accounting are also repeated in the headquarters. Each FMU has a sub account but all financial flows essentially go through the headquarters. Figure 7.2: Overview of Institutional Arrangements in BH State Office within Ministry of Foreign Trade and Economic Relations Entity Republic of Srpska (RS) Federation of Bosnia and Herzegovina (FBiH) Inspection Inspection Entity Ministries Ministry Agriculture, Forestry Other RS Other FBiH Ministry Agriculture, Water & Water Management Ministries Ministries Management & Forestry Forestry Department Forestry Department Federal Forest Forest Council Inspection Forest & Forest Agency Hunting Inspection Canton C1 C2 C10 Forest Inspection Cantonal CFO 1 CFO 2 CFO 10 Ministries Forest Guards Forest JPS Å ume RS R&D Centre CFMC 1 CFMC 2 CFMC 8 Companies Karst Mgmt Forest FMU 1 FMU 23 Nursery Prod Units RMU 1 RMU n Private 250,000 Owners 250,000 Owners Forest Owners Source: Dennison, S., 2006. III. FINANCIAL ARRANGEMENTS AND PERFORMANCE (a) Costs and Operational Efficiency (i) Unit Cost Analysis 7.23. Costs and efficiency vary considerably between the different CFC and between FBH and RS. A data questionnaire was collected to compare unit costs and the results are presented in Table 7.3 below. The first obvious comment is on the range of unit costs which varies by a large margin from each other as well as the averages are considerably higher than those in Ireland. Some of these higher costs could partially be attributed to the relative low average road density and the general poor condition of the road network. An additional constraint is the large costs associated with staffing which is further discussed below. 111 Table 7.3: Analysis of Unit Costs New Road Thinnings Fellings Soil Prep Planting Transport Road Reconst Constr Company (KM/m) (KM/m) (KM/m3) (KM/m3) (KM/ha) (KM/1000) (KM/m3) JPS Å ume RS 61 39 - 28 390 - - Sarajevo 110 25 40 - 70 30 – 45 2,500 2,250 - Tuzla 34 7 39 32 300 530 - USK 106 14 - 28 964 260 18.5 Ireland69 70 27 32 19 820 290 14.7 Source: World Bank consultant estimates. 7.24. There is a trade-off between road density and harvesting costs: the higher the road density, the lower the average extraction distance to roadside and the lower the average harvesting costs. A reasonable average road density for the overall BH estate would possibly lie within the range of 12-15m/ha depending on local conditions. The current density varies across CFMCs and the individual FMUs within JPS Å ume Republika Srpska and is within the range of 7 to 14m/ha (Sokolović and Bajrić, 2011 and interviewees). This lower than optimum density, coupled with the poor condition of the roads in parts of the network, results in a lowering of the overall operational efficiency for harvesting and transport and a decrease in the wood paying capability of the sector. (ii) Staff Costs 7.25. The data questionnaire was designed to ascertain how staff resources were utilized across the main range of forest activities. A high proportion of staff have either a forestry degree or technical forestry qualification. The employment levels for both workers and professional staff are high given the size of the resource and activities undertaken. There are political pressures to retain and employ high numbers of staff in the essentially rural areas where forest activities take place. Even though the FDCP supported extensive training including workshops on various aspects of institutional and financial management, and short courses in basic computer skills and public relations skills, there is no structured human resource development program and the approach to training is ad hoc and focused heavily on technical aspects of forestry. Furthermore, there has been a policy to provide employment to people who were injured during the war. Table 7.4 shows as the wood manufacturing industry was generally reducing labor (2,215 jobs lost between 2007 and 2009) as a response to the financial crisis, employment in the logging sector was stable or slightly increased in RS. This is further illustrated in Table 7.5 in-which staffing in all job categories in JPS Å ume RS has increased by 15 percent between 2007 and 2010. Table 7.4: Employment in the Forestry Sector 2007 2008 2009 Sector FBiH RS FBiH RS FBiH RS Forestry Logging and Related Services 5,586 5,154 5,525 5,276 5,232 5,475 Manufactire of Wood and Wood Products 8,813 6,432 8,666 7,194 7,052 6,461 Manufacture of Furniture 5,609 2,035 5,624 1,987 4,927 2,234 Totals 20,008 13,621 19,815 14,457 17,211 14,170 Source: World Bank consultant estimates. 69 Phillips, H (2008) Standard Costing – Report commissioned by Coillte, Newt own mount kennedy, Ireland 112 Table 7.5: Staffing in JPS Å ume RS Year Forest Forest Clerical + Qualified Total Engineer Technician Other Workers 2007 372 1,116 1,142 1,056 3,678 2008 401 1,199 1,134 1,228 3,983 2009 416 1,229 1,155 1,324 4,136 2010 423 1,254 1,178 1,377 4,232 Source: World Bank consultant estimates. 7.26. There has been little exposure of staff to forestry practices outside of BH or to the management of state-owned forest companies with an essentially commercial mandate or to the broader discipline of management of natural resources. The immediate impact of this is a public forest sector which is ―inward looking‖ and continues along what can be termed traditional lines even though these may no longer have relevance in the changing operating environment. When BH performance was compared against a number of state forest companies within the EU (m3 produced per employee and cost of forest management per m3), BH compares poorly with other countries averaging less than 1,200 m3 per employee against an average of circa 5,000m3. However, low staff costs has allowed BH to perform close to the average of €9/m3 regarding cost of management although there is a wide variation across the individual CFMCs. Figure 7.3: Timber Output per Employee BiH RS RS FBiH (Good) FBiH FBiH (Poor) Austria Austria Bavaria Bavaria Czech Rep Czech Rep Estonia England Estonia Finland France France Ireland Ireland Netherlands N-Ireland N-Ireland Romania Romania Slovakia Slovakia Average Average - 2,000 4,000 6,000 8,000 10,000 12,000 0.0 5.0 10.0 15.0 20.0 m3/Employee € /m3 Source: World Bank consultant estimates based on Phillips, H. and Popa, B. (2011) 70 and Savcor Indufor (2007)71. (iii) Taxing Revenue 7.27. Since 2009, each CFMC is required to pay 7 percent of its revenue from timber and secondary forest products by way of tax, comprising 1 percent to FBH budget, 5 percent to the municipality where the timber originated and 1 percent to the Canton budget. Both of these 1 percent payments are intended to be used for improvements to the general purpose functions of forests. Prior to 2009, each CFMC was obliged to pay 15 percent of timber revenues into a special in-company account, to support simple biological reproduction. Within RS the JPS Å ume RS is obliged to pay 10 percent of its timber revenues, based on stumpage, to a special simple biological reproduction account in the MAFWM. Prior to 2008, this account was held internally. The company is also obliged to pay 10 70 Phillips, H. and Popa, B. (2011). Functional review of forestry sector. Report prepared for World Bank. 71 Savcor Indufor (2007) A Way Forward with National Forest Administration – Romsilva. Report prepared under the World Bank Forest Development Program. 113 percent of its timber revenues, based on log yard price, to the municipality where the timber originated. These costs are referred to as taxes but they are levied more like concession fees that are based on total revenue rather than profit. Even though these fees are generally intended to support rural development in the more needy areas,most of these fees end up as part of general funds and go to pay for staffing and other recurrent costs. (b) Logging and Other Revenues (i) Funding for General Purpose Functions of Forests – OKFÅ  7.28. In FBH all legal entities (i.e. enterprises72) are required to pay a green tax, popularly known as the OKFÅ  tax, as compensation for utilization of the general purpose functions of forests.73 This tax is assessed at the rate of 0.07 percent of their total income. The OKFÅ  is paid to the Federal MoF from which some 80 percent of the total should be transferred to the respective Canton within which the legal entity generated its income. Within RS, a similar scheme applies but the rate is 0.1 percent of income and is payable to a special account within the MAFWM. Within FBH, the Financial Police and Federal and/or Cantonal Forestry Inspections undertake the control over the calculation and payment of OKFÅ  while in RS it is the responsibility of the Financial Police.74 7.29. At the Federal level, the Federal Forest Office prepares an outline annual plan for the year ahead comprising a series of “projectsâ€? which are to be funded from the OKFÅ . This draft plan is based on submissions mainly from the CFMCs but also includes other projects e.g. research. The plan is then submitted to the MAWMF for approval. At the Cantonal level, a broadly similar procedure applies only this time it is the Cantonal Forest Office that prepares the annual plan for utilization of the OKFÅ . Table 7.6 summarizes the value of the OKFÅ  collected at the Federal and Cantonal level as well as expenditure at the Federal level. In FBH, the KM 10 million expenditure of OKFÅ  at Federal level represents 52 percent of the value collected over the 2004-2010 period. There was no expenditure for 2009 and 2010 due to the non-inclusion of projects to be funded from the OKFÅ  within the par ent Ministry‘s budget. The low value for OKFÅ  collected in 2010 in part reflects the uncertainty surrounding the validity of payments following the Forest Law being deemed unconstitutional. Table 7.6: Tab OKFÅ  Collection and Federal Expenditure FBH (‘000KM) Year Collected Expenditure Collected Federal Budget Federal Level Cantonal Budgets 2004 2,160.9 1,070.2 6,792.9 2005 2,742.1 1,215.6 9,887.1 2006 2,981.0 1,732.2 11,221.4 2007 2,961.1 2,980.0 12,303.2 2008 4,073.0 3,055.4 14,700.8 2009 3,471.3 0.0 12,176.5 2010 985.0 0.0 - Source: World Bank consultant estimates. 72 Some enterprises are exempt. These include users of state forests, public institutions, humanitarian associations and organizations, foundations and their subsidiaries 73 A broadly similar tax scheme applies in the neighboring countries of Croatia and Serbia. 74 This arrangement whereby the Forest Inspection is involved in tax compliance is unique to FBH and results in two Federal organizations being responsible for the same function relating to tax compliance. 114 7.30. A large proportion of the CFO budget relates to recurrent costs, the majority of which are the wages for the Forest Guards. Table 7.7 shows the division of costs at a Cantonal forest office from 2007 until 2010. As time evolved, fewer funds were allocated to most expenditure and more to the forest guards. Hence, the current arrangements regarding the OKFÅ  do not ensure that (a) the mo nies are used for the purpose intended, (b) that within FBH, the proportion due to MAWMF gets allocated or that each Canton receives its full amount and finally that (c) the prioritization of disbursement at project is clear and transparent. Table 7.7: Example of Cantonal Forest Office Expenditure (KM) Description 2010 2009 2008 2007 Cantonal Forest Office 133,802 142,310 146,100 129,850 Expenditures of forest guards service 2,528,007 952,638 845,349 722,288 Management plans 29,216 233,727 384,159 138,000 Supervision management plans 16,167 16,494 15,841 - Projects for afforestation, nursing and forest protection 10,000 25,892 234,107 136,197 Improvement of biodiversity 68,000 45,439 110,057 153,237 Development of undeveloped parts of municipalities 141,908 307,529 297,954 255,192 Forest road construction and maintainence - 75,551 - 253,123 Total 2,927,100 1,799,580 2,033,566 1,787,886 Source: World Bank consultant estimates. (ii) Timber Revenues 7.31. There are three types of timber sales common to FBH and RS: (a) annual contract (successive delivery), (b) auction and, (c) local (needs for local population). Under the successive delivery system, timber is sold using a standard pricelist. Companies purchasing timber must first comply with a range of eligibility criteria. Where the demand is greater than supply, then a second series of criteria are used to allocate volume among the eligible companies. These include (a) number of employees, (b) history of previous contracts, (c) financial data (status) and (d) level of finalization (capacity to add value). A CFMC internal commission (committee) decides upon the allocation. Under the 2008 decision on the sale of wood assortments from state forests in BH, all companies were expected to move from a pricelist based system to a more market orientated approach. In practice this has not happened due to the regulation being interpreted as only a guideline and the lower prices received at auctions during the economic downturn. Sales of timber and wood assortments to the local population (mainly firewood) are by way of the pricelist. In RS an additional ―submission‖ system is used for the highest value assortments. Potential purchasers are invited to submit a price offer. The highest offer is then set as a basis among potential purchasers to determine if an increased price is possible. Due to the fall in demand, this system has not been used for the past two years. 7.32. All CFMCs and JPS Å ume RS maintain that their pricelist is updated based on market demand but exactly how is unclear. The methodology is inherently flawed as it uses a non market series of values as its starting point with only a pro rate amendment for those assortments sold at auction. It is not possible in either Entity to enter into any sales contract for longer than a year. This, while not an immediate problem, could act as a barrier to potential new entrants in such areas as wood pellet production where a secure supply of raw material is a pre-requisite for investment. (iii) Lost Revenues – Illegal Logging 7.33. Whilst it is possible to get some general indications of illegal harvesting, it is always much more difficult to get an estimate of what is sometimes called the “legal-illegalâ€? timber trade. Pružan and Konjalić (2011) provide the most recent estimate of illegal logging, albeit only for the Srednjobsanski Canton. Their analysis indicates that while the level of illegal logging had decreased on an annual basis 115 from 2003 to 2007, there was an increase in 2008 and that levels have remained more or less stable at that level until 2010. On average over the 2003-2010 period, illegal harvesting accounted for 5.37 percent of total volumes harvested which was equivalent to 27,717 m3 per year. The authors maintain that socio- economic needs are the main driver for illegal logging with the perpetrators being in the main members of the local communities harvesting timber for firewood needs. (c) Financial Performance 7.34. A summary of the 2008 and 2009 outturn for the CFMCs in FBH shows (a) a significant fall in revenues in 2009, (b) an almost complete dependence on timber revenues, (c) a low level of profitability with an overall loss in 2009 and (d) a significant increase in payables and almost doubling of receivables. This is despite only a 2.5 percent fall in average timber prices compared with 10-20 percent across many European countries due to the practice of using pricelists rather than markets to determine timber prices. During 2009, only four of the CFMCs reported a profit (Table 7.8). Profit expressed as a percent of turnover was low ranging from 0.3 percent in Unsko-sanski to 4 percent in Canton 10. Table 7.8: Financial Performance CFMCs 2009 - FMAWMF (2010) Timber Operating Production Average No. Average Receivables Payables Canton Total Revenue Profit Loss Revenue Expenses Costs Employees Salary 31.12.2009 31.12.2009 Unsko-sanski 22,882 21,490 22,805 18,232 78 0 578 749 1,302 4,503 Posavski 0 0 0 0 0 0 0 0 0 0 Tuzlanski 12,317 11,964 12,571 10,527 0 254 525 643 4,356 1,378 ZeniÄ?ko-dobojski 27,876 21,706 31,636 11,555 0 3,759 957 815 7,298 18,626 Bosansko-podrinjski 1,383 1,013 1,359 1,085 24 0 40 728 262 63 Srednjobosanski 26,919 25,109 28,604 28,604 0 1,685 850 625 7,640 8,855 Hercego.-neretvanski 0 0 0 0 0 0 0 0 0 0 ZapadnohercegovaÄ?ki 530 385 403 120 8 0 12 667 237 80 Sarajevski 12,902 10,209 13,106 10,385 0 203 485 759 797 4,202 Kanton 10 30,221 29,805 29,007 22,512 1,213 0 523 1,005 20,749 5,242 Totals 135,031 121,680 139,490 103,020 1,323 5,902 3,970 759 42,642 42,949 Source: World Bank consultant estimates. 7.35. The performance of JPS Å ume RS has varied over time and is affected by non-commercial pressures. Data for it are in Table 7.9 for the period 2007 to 2010 and include the planned outturn for the current year. The company showed a steady performance throughout 2007 and 2008. The downturn in the economy and economic crisis is reflected in the 2009 results which saw the revenues significantly reduced with associated knock on impact on costs. The company however reported a small profit before tax of KM 1.4 million equivalents to 0.9 percent of revenues. Of concern is the almost doubling of payables from KM 28.82 million in 2008 to KM 52.83 million in 2009. The official payment terms are 90 days credit for large customers e.g. Natron Hayat and between 30-60 days credit for small and medium enterprises (SME). The value for payables increased in 2010 by one third to KM 70.77 million. JPS Å ume RS is unable to enforce its credit terms and at this stage is in essence providing cheap credit to its customers in the expectation that once the economy recovers it will be able to return to more normal levels for payables. The company, as the dominant supplier of round wood, is also under some political pressure to maintain supplies to customers. 116 Table 7.9: Profit and Loss - JPS Å ume RS ('000KM) 2011 2010 2009 2008 2007 Planned Planned Actual Planned Actual Planned Actual Planned Actual 1 Operational Revenue 158,613 159,612 136,149 164,822 126,131 160,818 153,241 139,640 133,872 Timber 154,649 159,439 135,945 162,819 125,214 157,551 152,212 139,610 133,314 Non Timber Products 117 173 3 100 3 26 30 Other 3,847 201 1,903 914 3,241 1,029 558 2 Financial revenues (e.g. Interest) 3 Other revenues 23,758 28,291 39,626 34,545 31,002 28,938 28,546 23,043 25,244 4 Total revenue (1+2+3) 182,371 187,903 175,775 199,367 157,133 189,756 181,787 162,683 159,116 5 Operating Expenses 70,612 74,488 61,735 74,853 60,650 60,242 63,349 45,408 52,983 Salaries 68,935 72,376 60,010 71,811 58,573 57,012 59,552 45,408 49,183 Wages (Physical Workers) 1,677 2,112 1,725 3,042 2,077 3,230 3,797 3,800 6 Financial Expenses 449 458 486 202 395 438 515 430 609 7 Other Expenses 109,756 111,095 112,488 121,572 94,681 122,106 112,425 108,165 100,252 8 Total Expenses (5+6+7) 180,817 186,041 174,709 196,627 155,726 182,786 176,289 154,003 153,844 9 Gross Profit (4-8) 1,554 1,862 1,066 2,740 1,407 6,970 5,498 8,680 5,272 10 Tax 155 186 259 274 226 697 377 868 797 Net Profit (9-10) 1,399 1,676 807 2,466 1,181 6,273 5,121 7,812 4,475 Gross Profit as % Revenue 0.85% 0.99% 0.61% 1.37% 0.90% 3.67% 3.02% 5.34% 3.31% 2010 2009 2008 2007 11 Receivables as on 31 December 44,027 34,774 31,913 34,844 12 Payables as on 31 December 70,773 52,830 28,823 24,226 Source: World Bank consultant estimates. 7.36. In general, all of the CFMCs and JPS Å ume RS can to a large degree be considered as timber management companies. They do not have daughter companies engaged in other activities e.g. panel board production, wind energy etc. Their business is essentially that of implementing forest management plans. This being so, it is possible to compare their profitability with other state owned forest enterprises. The data used for comparison was the 2007 outturn as this is the most recent data available for the other sources and the report from Savcor Indufor (2007) together with more recent analysis Figure 7.4: Profitability of State Forest Companies for Romania (Phillips and Popa, 2011). The RS profitability, based purely on timber CFMC (Poor) management for both CFMCs and JPS CFMC (Good) Å ume RS is less than the average for the Austria selected range of companies. However, the Bavaria Czech Rep better CFMCs did outperform Romania, Estonia Czech and Bavaria. Caution is essential in Finland interpreting the results as during 2007, France timber prices were not market based in BH Ireland whilst they were in the comparison Norway countries. What the comparison shows is Romania that there is potential to increase Slovakia Average profitability to closer to the average. 0 5 10 15 20 25 30 35 40 45 Profit as % Turnover (Timber Management) IV. RECOMMENDATIONS Source: World Bank consultant estimates based on Phillips, H. and Popa, B. (2011) and Savcor Indufor (2007). 7.37. The forestry sector could make a more significant contribution to the economy if it were more efficiently managed. The government has a major presence in the sector and a number of changes in how it operates are needed to unlock this potential. The scale of harvesting could be stepped up and the institutional arrangements need to be considerably streamlined. Further policy changes are needed to give the institutions an appropriate 117 operational framework. At this stage it would not seem appropriate to expand the private sector role although this may be possible in the more distant future. The following are the recommendations for the forestry sector: ï‚· Improve road access and conditions to help increase production, improve harvesting efficiency and increase forestry contribution to GDP. ï‚· Identify and reduce unproductive labor in public forest sector institutions in order to improve productivity and profitability. ï‚· Invest significantly in human resources through the provision of additional skills in the areas of business and organizational development, change and change management, investment appraisal, cost benefit analysis, robust budgeting processes, public awareness and communication. ï‚· Implement institutional changes in FBH: transfer the responsibility for security and protection of the forest resource Cantonal Forest Management Companies; have the Cantonal Forest Offices report formally to the Federal Forest Office and operate effectively as daughter organizations; give the Financial Police the sole responsibility for OKFÅ  (timber tax) compliance; increase the staff within the Forestry Department in FBH in line with its responsibilities. In RS review the staffing and roles of both Forestry Department and Forest Agency. ï‚· Have the larger Cantonal Forest Management Companies within FBH and JPS Å ume RS establish a comprehensive Business Process Review (BPR) of all the major processes within their organization and have the BPR critically analyze existing business processes to identify opportunities to improve compliance, improve quality and reduce costs. ï‚· Develop a transparent and market-based mechanism for timber sales in a manner that allows independent verification with a definite timeframe and indicators for implementation. ï‚· Set the overall level of the OKFÅ  tax based on an in-depth cost-benefit analysis and reflecting the true value of public good functions being provided. The system for allocation of the OKFÅ  collected and its subsequent disbursement needs to be made clear and transparent with complete traceability and accountability. It would appear that the OKFÅ  is increasingly being treated as part of the general budget and as a consequence only a proportion gets used for forestry. The overall level of OKFÅ  should be based on an in -depth cost benefit analysis and reflect the true value of public good functions being provided. 7.38. At this stage, privatization is not judged to be a sustainable option in BH. This is due to (a) the absence of any sound market price to determine the value of the harvesting rights / timber concession and (b) the presence of weak institutions and lack of capacity to effectively implement the legal framework; and the risk that privatization could lead to over exploitation resulting in forest degradation, soil erosion and environmental damage. 118 CHAPTER 8: ENERGY I. SECTOR OVERVIEW 8.1. The energy sector has recovered from war and the turmoil of the early 1990s, fueled by strong growth in energy demand that came along with renewed economic growth. In 2009, Total Primary Energy Supply (TPES) stood at 5.9 million tons oil equivalent (Mtoe), about only 80 percent of the peak of 7 Mtoe reached in 1990 (see Figure 8.1 below). TPES fell significantly in the 1990s as a result of war and economic contraction, but also reflecting structural changes of the economy. However, this downward trend was reversed in the early 2000s fueled by strong economic growth. Between 2000 and 2009, the expansion in primary energy demand (6.2 percent on average per year) outstripped GDP growth (4.6 percent per year). Over the same period, primary energy growth in OECD Europe was only of 0.7 percent per year. 8.2. The country‟s reliance on domestic coal has remained almost unchanged since 1990 (between 60 to 65 percent of TPES) with the total share of all fossil fuels reaching over 94 percent in 2009. The energy mix is dominated by coal, which provides about 64 percent of the TPES. Oil products account for about 19 percent of TPES, while hydro and natural gas account for about 6 percent each. Given the country‘s reliance on domestic coal and hydropower, dependence on energy imports (37.5 percent) is significantly lower than that of OECD Europe (which stands at 48 percent, see table 8.1 below). However, it is worth noting that almost 100 percent of oil products and natural gas are imported, thus indicating a high import dependence on these fossil fuels. 8.3. As a result, BH‟s carbon intensity is one of the highest in the Western Balkans (See Table 8.1). In 2007, the International Energy Agency (IEA) estimated total CO2 emissions at 16.3 Mt, broken down as follows: energy sector (52 percent), construction and agriculture (16 percent), transport (14 percent), industry (13 percent), and residential (5 percent). Emissions have tripled since 1995, and increased by 21 percent between 2000 and 2004 as energy consumption increased. At the same time, a switch to lower-carbon fuels did not occur; coal accounts for 75 percent of emissions, followed by oil products (21 percent) and natural gas (4 percent). Figure 8.1: Primary Energy Supply 8 Ktoe 7 6 5 4 3 2 1 0 Coal Oil Gas Hydro Other Renewables Source: IEA. World Energy Balances, 2011. 119 8.4. In contrast, the structure of final consumption has changed over the past decade; industry witnessed a sharp decline, while consumption in the transport and the buildings sectors expanded significantly. Total final energy consumption (TFC) in Figure 8.2: Total Final Energy Consumption the country reached 2,736 6 Ktoe in 2009 (see Figure 8.2). Ktoe Along with economic growth, 5 came more demand for personal mobility and lifestyle 4 improvement. Back in 1990 3 the industrial sector consumed almost twice as much as the 2 transport and buildings sector combined, with a share of 45 1 percent in TFC. In 2009, 0 however, the transport sector (31 percent) accounted for the largest share in TFC followed Industry Transport Residential by buildings (20 percent), Commerce and public services Non-specified (other) Non-energy use industry (21 percent) and Source: IEA. World Energy Balances, 2011. services and public buildings (4 percent) sectors. Due to the reliance on oil and electricity in the transportation and buildings sector, their share in BH‘s final energy consumption has risen over time to 42 percent and 23 percent, respectively. Coal and gas each accounted for around or above 10 percent, with the remainder coming from fuelwood and others. 8.5. Energy Intensity (EI) has improved since 2000 (by about 10 percent) but it remains high due to inefficient use of energy. In 2008, EI was estimated at 0.71 toe per thousand USD of GDP (in 2000 USD), about 4 times above the average for OECD Europe and about 30 percent above the average in the Western Balkans. This reflects high levels of losses in energy transformation, which are estimated at about 50 percent of TPES. Inefficient conversion of primary energy (particularly in power generation) is due largely to outdated equipment and technologies (see below), and to the limited share of combined heat and power (CHP) in the energy mix. 8.6. Energy end-use is also inefficient, thus indicating a huge potential for reducing the burden of energy expenses in both public and private sectors. The buildings sector (private, public, and residential) account for one third of TFC. Buildings are generally poorly insulated, causing heat losses of more than 30 percent. In addition, since the gas distribution network is not fully connected, the majority of the heating requirement of the households is fulfilled by the inefficient use of electricity. As a result, about 65 percent of the total electricity consumed in 2009 went to the residential and public buildings sector. By contrast, annual electricity consumption per capita was 2467 kWh, about 2.5 times lower compared to average levels of 6287 kWh in OECD Europe in 2009 (see Table 8.1 ). 120 Table 8.1: Western Balkans Select Energy Indicators BH Western Balkans OECD Europe Total primary energy supply (Mtoe) 5.99 30.3 1821.5 Energy consumption (toe) per capita 1.59 1.78 3.35 Electricity consumption (kWh) per capita 2,467 3747 6,287 Energy intensity of GDP (toe/thousand 2000 0.71 0.56 0.17 US$) Carbon intensity (kg CO2/2000 US$) 2.32 1.53 0.38 Net imports as percent of TPES 37.5% ** 48% (Dependence) Source: All data are from IEA 2008 selected energy indicators. * Includes: Albania, Serbia, Croatia, and Macedonia, FYR. Data for Montenegro is not available; ** Not calculated to avoid double counting due to intra- regional trade. 8.7. The country‟s energy infrastructure (particularly power) is well developed but significant investments are needed to complete its rehabilitation and enable the country to compete in export markets. The country‘s 11 power plants are owned by three government-owned companies operating in two Entities: Elektroprivreda Bosne i Hercegovine in FBH of Bosnia and Herzegovina (EPBH), (installed capacity of 1.8 GW), Elektroprivreda Republike Srpske (EPRS), (2.6 GW) in RS and Elektroprivreda Hrvatske Zajednice Herceg Bosne (EPHZHB), (0.8 GW) in the FBH serving the area of Herzegovina. EPBH and EPRS have both hydropower and coal-fired power plants, but EPHZHB has only hydropower capacity. Generation levels have increased since the early 1990s with exports reaching 500 GWh, or 37 percent of electricity generated during that year. However, significant investments are required to complete the rehabilitation of power infrastructure. In fact, generation has not recovered from its pre-war level (14.6 TWh in 1990) and its efficiency is still low (about 30 percent) due to lack of maintenance and spare parts, and the slow replacement of obsolete units. Moreover, several expansion and greenfield projects are envisioned to keep up with demand growth and eventually, increase exports. These investments are expected to be financed by the three power utilities. 8.8. Power utilities operate as three small vertically integrated monopolies in three separate irregularly shaped geographical areas giving rise to inefficiencies. The BH power sector is being restructured to reduce these inefficiencies and to comply with the provisions of the EC Directive 2003/54 (electricity), which promotes competition in the electricity sector. The first step was taken in 1999 with the establishment of the Joint Power Coordination Center (ZEKC). ZEKC was owned by the three Elektroprivredas, who were represented on its management board. Its roles were to coordinate electricity exchanges between the Elektroprivredas and with foreign countries, to coordinate scheduled generation and transmission outages, and to maintain the operating reliability of the transmission system as a whole. 8.9. More recently, the merger between coal companies and EPBH along with the needed modernization of coal mining infrastructure led to a significant fiscal burden and it is also likely to put additional pressure on the sector‟s financial viability. The combination of low productivity (due to overstaffing) and lack of investment in advanced technologies has led to financial difficulties in the coal sector. Mining companies continued to accumulate losses as their costs exceed the economic break-even point for almost a decade. As a result, they were unable to cover pension schemes and remediation costs. To address these issues, the government in FBH decided in 2010 to merge its coal mines with the power utility EPBH. Subsequent to the merger, the Government took on the obligation for pensions and income taxes while other accumulated and unfunded past liabilities were transferred to EPBH (see below). 121 8.10. As a result, FBH faces the twin challenge of improving the efficiency of public expenditure in energy services and reducing the potential risks to increasing public financing in the power sector. To address these challenges, the chapter assesses the fiscal impact of the merger of coal companies with the power utility EPBH and analyzes the financial viability of the power utilities vis-à-vis their investment plans. Also, the chapter assesses the potential for increasing energy efficiency in the public buildings sector, as a key measure to improve the efficiency of expenditures in energy services. II. COAL SECTOR RESTRUCTURING AND INVESTMENT NEED IN THE POWER SECTOR 8.11. Bosnia‟s energy sector is characterized by a high degree of government ownership and large capital expenditure requirements in energy infrastructure. In the power and coal sectors, investment needed to meet demand reliably is estimated at over €4 billion by 2020.75 Up to now, the power sector has been financially viable and creditworthy, and as such has been able to cover its costs and finance its capital expenditures. The section below assesses the impact of the coal sector restructuring and these large investment plans on the sector‘s commercial and financial viability, and ultimately on public expenditure. A. RESTRUCTURING OF THE COAL MINING SECTOR IN FBH 8.12. The coal industry is a key sector in the country: it accounts for 65 percent of electricity generation and energy supply and provides employment to about 16,000 people. Proven reserves of coal (both brown coal and lignite) are estimated at 5.7 billion tons. These are located, along with production capacities, in FBH (4,779 Mt, and 62 percent of production) and in RS (0.845 Mt of reserves and 38 percent of coal production). Coal production (about 9 Mt in 2009) is derived from 11 open cast mines located in FBH (employing 14,000 miners) and the remaining 38 percent from two open cast mines in RS (employing 2,000 miners). Prior to their restructuring, the mines in FBH were fully state-owned and sold most of their production to the nearby power plants of the Power Utility of Bosnia and Herzegovina (EPBH). The RS mines of Ugljevik and Gacko are integrated with the nearby power plants of (EPRS). Overall, more than 75 percent of the coal produced in BH is consumed by electric utilities; the rest is sold to industry and households. 8.13. Other major challenges faced by the coal mining companies in FBH include the severe lack of investments in modern technology, below-cost pricing, and high costs because of low productivity. Most of the investments made in coal companies in FBH over the past decade were for repairs and some maintenance but did not fully cover depreciation. As a result, the asset base of the coal companies had been shrinking and by the end of 2005, the value of tangible assets (including primarily equipment) had been written-off by 93 percent. As a result, coal mines are operating with obsolete technology which decreases their productivity. In addition, the average price for coal paid by electricity plants (1.8 2004 EUR/GJ in 2006) did not cover their costs (2.0 2004 EUR/GJ), which are high because mainly of overstaffing, despite recent efforts to reduce costs. As a result, labor productivity is low in comparison with international standards, although there are significant differences between mines. 8.14. By 2009, coal companies had been registering losses and accumulating large debts. Mining companies suffered a loss of about 20.2 and 13.1 million euros in 2005 and 2006 respectively. Direct state subsidies covered these losses and as a result indirectly subsidized electricity prices. Apart from operating costs, the liabilities arising from contributions owed by the mining companies to pension, health 75 BH Energy Sector Study. Energy Institute Hrvoje Požar, Croatia Soluziona, Spain Economics Institute Banjaluka, BH Mining Institute Tuzla (2008). 122 insurance and other funds resulted in a high level of indebtedness. As it can be seen in Table 8.3, by end- 2008, total liabilities from coal companies in FBH amounted to over BAM 265 million (135 million euro). 8.15. To address the long-standing economic difficulties in the coal sector, the Government of FBH decided to merge the coal mines with the government-owned electric utility EPBH. On 14 January 2009, the Government of FBH took the decision76 to transfer 100 percent of the shares in seven coal mines77 to EPBH. On September 30, 2009, the Company received approval from the Competition Council of BH for the transfer of ownership, and following a valuation performed by an independent auditing firm, EPBH issued shares in the amount of BAM 81,804 thousand (equivalent to the estimated value of the coal mines) to the Government of FBH, thereby increasing its share in the Company. Table 8.2: Overview of Write-offs by the Government of FBH & Liabilities of the Coal Mines Transferred to EPBH Amount in 000‘ BAM Write- off of liabilities by the Government of FBH 133,686 Salary income taxes 22,191 Liabilities for pensions 111,495 Total Liabilities (after corrections and write-offs) 264,297 Long Term liabilities 88,982 o/w Loan liabilities 65,293 Short Term Liabilities 175,315 o/w Suppliers 50,341 Health and unemployment insurance 80,059 contributions & other taxes Source: Value appraisal of Federation of Bosnia and Herzegovina at 31.12.2008. PWC, 2009 8.16. In addition to institutional Table 8.3: Overview of Government Liability Payment implications, the restructuring of the Schedule. coal sector had a fiscal impact on the Year Liability paid in 000s BAM/year Government of FBH. Following the 2009 19,526 merger, the Government of FBH took 2010 19,526 over the obligations for income taxes and 2011 19,526 pension contributions. According to the 2012 19,526 law on financial consolidation of coal 2013 19,526 mines78, the fiscal burden was equivalent 2014 19,526 to BAM 133.7 million (see Table 8.2). 2015 19,526 These liabilities will be financed directly Total 133,686 from the budget over a seven-year period as given in Table 8.3. 76 No. 12/09, Official Gazette of FBH, No. 4 / 09 77 RMU―Kakanj‖d.o.o.,Kakanj; RMU―Breza‖d.o.o.,Breza; RU ―GraÄ?anica‖ d.o.o., Gornji Vakuf -Uskoplje; RMU ―Abid Lolić‖ d.o.o., Travnik-Bila; Rudnici ―Kreka‖ d.o.o., Tuzla; RMU ―Ä?urÄ‘evik‖ d.o.o., Ä?urÄ‘evik; RMU ―Zenica‖ d.o.o., Zenica. 78 Reference 123 8.17. The merger also had wider financial consequences for EPBH, in the form of transfers of accumulated liabilities and large investment needs. Although pension contributions and income tax liabilities represented a large proportion of accumulated debts of coal companies, other significant obligations remained on their books as mentioned above (see Table 8.2). Furthermore, investments to maintain the levels of production are urgently needed within the next few years. In the medium- to long- term, further capital expenses would need to be financed to bring the sector to sustainable production and cost levels. Table 8.4 provides figures for the investment needs for FBH coal mining sector up to 2030. Overall, the liabilities transferred from coal as well as the medium and long term investment needs in the sector, weaken the financial situation of EPBH. These investments would have to be financed by the consolidated company – and ultimately by the Government of FBH as owner of the power company. Table 8.4: Investment Needs for Coal Mines in FBH by 2030. Level of Investments (000s BAM) Entity 2010-2016 2017-2030 Total FBH 319,070 218,662 537,731 Source: Value appraisal of Federation of Bosnia and Herzegovina at 31.12.2008. PWC, 2009 B. INVESTMENT OUTLOOK AND FINANCIAL ANALYSIS OF POWER UTILITIES 8.18. Power supply is expected to satisfy the consumption growth in the medium term, although the supply and demand balance is expected to get tighter before new generation comes on stream. Based on information provided by power utilities, Figures 8.3, 8.4 and 8.5 present the projected demand and supply balance with and without new generation capacity installed by the new investment. 8.19. Power demand is expected to increase steadily in the three entities at an annual growth rate raging from 1 percent to 3 percent by 2015. It is expected that such an increase in demand would be met by existing plants with exports/imports declining/increasing until new infrastructure is in place: ï‚· EPBH will require the new generation capacities to start producing from 2012 in order to maintain its export levels. Although domestic demand is expected to be fully satisfied by current capacities, exports are likely to decrease as the supply and demand balance tightens and/or there is any delay in new investments. ï‚· EPHZHB is heavily dependent on the new hydro and wind capacities to be able to fully cover its domestic demand. In 2014 and 2015, the new generation capacities would provide about 30 percent of the total electricity demand of Mostar. Consequently, any delay in financing the planned investments in EPHZHB would result in higher imports, which so far have been more expensive than the cost of domestic production. ï‚· EPRS expects none of the new planned generation units to become operational by 2015. Therefore, the demand and supply is expected to remain balanced in 2015, with exports declining slightly. 124 Figure 8.3: Demand and Supply Balance in FBH Figure 8.4: Demand and Supply Balance in Mostar 8 6 TWh 4 2 - 2008 2009 2010 2011 2012 2013 2014 2015 TPPs - existing HPPs - existing New generation Domestic purchase Figure 8.5: Demand and Supply Balance in RS 6 4 TWh 2 0 2008 2009 2010 2011 2012 2013 2014 2015 Existing TPPs Existing HPPs Domestic purchase Import Domestic distribution Domestic direct customers + Export Source: World Bank based on projections by the power utilities 8.20. To achieve self-sufficiency and maintain export levels, significant investments in coal production, generation, and distribution are being planned by all three power utilities. The three government-owned utilities have developed rehabilitation and upgrading plans for existing plants, and investments in new generation. Some of the key projects include: the rehabilitation and upgrading of thermal power projects (TPP) Kakanj and Tuzla (EPBH) and about 1.5 GW of hydropower plant capacity. EPRS plans the expansion projects for the TPP Gacko II (600 MW) and Ugljevik II (600 MW) although they are not expected to be operational by 2015. New projects for export are planned in Bugojno (EPBH) and Kongora (EPHZHB). New HPP projects include Buk Bijela (RS), Vrilo and Kablic (EPHZHB). Also, wind projects are being built in Borova Glava, Mesihovina, Velika Vlajna, and PokleÄ?ani (EPHZHB) with an installed capacity of 200MW. Additional investments are also planned to rehabilitate and rebuild distribution networks, since these are owned and operated by the power utilities. Annex 1 provides a snapshot of the total planned investments by the utilities until 2015 which amount roughly to about BAM 6.1 billion (equivalent to 3.1 billion euros). 8.21. Such large investment plans are expected to be fully financed by the electricity producers. According to information provided by the power utilities, the bulk of investment will be financed through their own resources and additional foreign loans – which may include some type of sovereign guarantee. No major private sector involvement in the above projects is currently being envisioned. To assess the 125 projected impact of these investments, a financial model has been developed for each one of the utilities for the purpose of this PEIR.79 8.22. Given the foreseen borrowing and tariff levels, the financial position of all three utilities would be weakened as they would face liquidity constraints and eventually, difficulties in fulfilling their future debt service obligations. None of the three power utilities would be able to cover the costs of financing their overall investment programs given the current and foreseen tariff levels (see below for a discussion on tariffs). They would also face difficulties in meeting their debt service obligations under the earlier and future externally-financed projects, as evidenced by the sharp deterioration of their self- financing and a debt service coverage ratio in Figures 8.6 and 8.7 below. Figure 8.6: Self Financing Ratio of Power Utilities Figure 8.7: Debt Service Coverage Ratio 80% 60% 40% 20% 0% 2010 2011 2012 2013 2014 2015 -20% EPBiH EPHZHB EPRS Source: World Bank staff calculations 8.23. So far, electricity tariffs have enabled power utilities to carry out their operations reasonably efficiently –although they did not fully cover depreciation costs. Over the years, the financial strategy of the three utilities has aimed at generating adequate cash from internal generation to cover cash operating expenses and debt service, and to provide a reasonable share of self-financing for carrying out capital investments. So far, electricity tariffs were set at levels that enabled the utilities to achieve this strategy, but did not necessarily cover the full amount of the annual depreciation charges. This has led in some years to net losses in accounting terms, but has not prevented the utilities from being able to carry out their operations efficiently. 8.24. However, the existing and expected future tariff levels would become inadequate to sustain the needed investments, thus putting at risk the sector‟s financial and commercial viability. Using the information provided by three EPs, Figures 8.5a, b and c compare the existing and expected tariff levels with the capital expenditure planned in each year and the resulting year-end cash balance: ï‚· EPBH registered net losses of BAM 14 million in 2010 as a result of increased production costs and lower power generation. Tariffs were recently revised by the regulatory body in FBH and will come into effect in 2012. According to the utility forecast, these are assumed to stay constant until 2015. As investments increase every year, the tariff levels in 2012 would not allow the company to keep pace with the capital expenditure resulting in a negative year- end cash balance to the utility of almost BAM 200 million by 2015 as evident from Figure 79 The model includes financial projections, including projected balance sheets, income statements, and cash flows. The base year is 2010 and includes information from the companies‘ audited financial statements from 2008 -2010. The main assumptions in terms of investment plans, projected sales and costs, financing plans, and tariff levels have been provided by the companies. 126 8.8. It must be noted that these investments do not include the investments EPBH may make in the coal mines. ï‚· EPHZHB improved markedly its financial performance with net income increasing from BAM 7.3 million in 2008 to BAM 40.5 million in 2010. The cash flows also improve reaching BAM 14 million in 2014. The Company forecast that average electricity tariffs would increase in 2012 by 26 percent as EPHZHB would stop supplying the Aluminum Plant in Mostar, to which EPHZHB has the obligation to supply power at below cost. However, given the foreseen debt-service requirements, year-end cash balance is still expected to become negative by 2012 and reach about (minus) BAM 400 million in 2015 (see Figure 8.9). ï‚· EPRS. Electricity tariffs have been increasing over the past three years at an annual rate between 3 to 4 percent. Coupled with strong sales, EPRS has registered net profits of around BAM 30 million and strong cash flows of BAM 47 million in 2010. Under EPRS (conservative) assumptions, tariffs are would increase yearly by about 1 percent. At these levels, the year-end cash balance turns negative in 2011, reaching about (minus) BAM 670 million in 2015. (see Figure 8.10). Figure 8.8: Domestic Tariffs and Year-End Figure 8.9: Domestic Tariff and YECB in Cash Balance in EPBH EPHZHB 1,500 0.1600 0.1500 1,000 KM million BAM/kWh 0.1400 500 0.1300 - 0.1200 (500) 0.1100 2008 2009 2010 2011 2012 2013 2014 2015 CAPEX Year end cash balance Domestic tariff Figure 8.10: Domestic Tariffs and Year-End Cash Balance in EPRS 600 0.130 400 0.125 200 KM million BAM/kWh - 0.120 (200) 0.115 (400) 0.110 (600) (800) 0.105 2008 2009 2010 2011 2012 2013 2014 2015 CAPEX Year end cash balance Domestic tariff Source: World Bank staff calculations 127 8.25. Compounding these problems there is fragmented regulation of generation and distribution at the state and entity level. There are three separate electricity regulators. First, the State Electricity Regulatory Commission (DERK) based in Tuzla, is responsible for electricity transmission tariffs. Transmission is now performed by a separate transmission system operator (TSO), Elektroprenos BH. Then there are separate regulators at the Entity level. The FBH Regulatory Commission for Electricity (FERK) is based in Mostar responsible for electricity generation and distribution in FBH. The RS Regulatory Commission for Electricity (REERS) based in Trebinje is responsible for electricity generation and distribution in RS. The entity regulators are responsible for licensing, proposing network access tariffs and end-user energy prices, and protecting customers. 8.26. The projections of the financial performance indicate that the utilities would face challenges in meeting their debt servicing requirements and the required counterpart funding for investment projects. In this context, access to financing is likely to become more difficult and costly if the tariff and productivity issues analyzed above are not solved. Some government support - most likely some credit enhancements - would be needed. This would in turn create contingent liabilities for the governments. 8.27. To allow utilities to access financing for the needed investments without being a drag on public finances, BH may consider the following measures: ï‚· Continue the restructuring and modernization of the coal mines within the new institutional setup. Efforts should focus on improving efficiency by adopting modern technologies and improving labor productivity. ï‚· Continue restructuring of electricity sector towards EU standards for corporate governance; and set tariffs at their full cost recovery levels, while putting in place targeted social protection mechanisms to protect vulnerable population. ï‚· Prepare and adopt an investment strategy for each of the power utilities based on a comparative least-cost investment plan which takes into account the power sector demand forecast at the entity, national, and regional levels. Options for private sector involvement should also be considered. III. IMPROVING THE EFFICIENCY OF PUBLIC SPENDING BY TAPPING THE ENERGY SAVINGS POTENTIAL 8.28. There is significant potential for reducing the burden of expenditure in energy services in the public and private sectors by improving Energy Efficiency (EE). Improving EE would also yield positive economic value through increased employment, more competitive economies, and eventually increased power exports (since it would free up domestic energy supplies). Energy end-use is inefficient; public buildings are generally poorly insulated and the estimated potential for energy savings ranges between 30 percent and 35 percent. Despite this significant potential, several market barriers to energy efficiency persist. This section: (i) examines the status of demand side energy efficiency (EE), (ii) assesses the technical EE potential in public buildings based on a survey undertaken for the purpose of this PEIR, and (iii) identifies market barriers and offers recommendations on the way forward. 128 A. OVERALL TRENDS AND INSTITUTIONAL FRAMEWORK FOR ENERGY EFFICIENCY 8.29. Heating of buildings (residential, public, and services) is a major contributor to energy consumption and it is expected to grow significantly over the next two decades. Between 2000 and 2009, energy consumption in the buildings sector has been growing strongly at an average rate of 8.2 percent per year. Its share in total final energy consumption has also increased over time to reach one third in 2009. Looking forward, end-use energy consumption, including the buildings sector, is expected to increase by around 70 percent during 2006-27. 80 This translates in annual growth rates of more than 3 percent. The most rapid growth is expected in the commercial sector (140 percent) followed by industry (100 percent) and the residential sector (60 percent). 8.30. Most building heating needs are satisfied by natural gas and by the much more inefficient use of electricity. Access rates are of about 20 percent for natural gas, more than 99 percent for electricity, while only about 12 percent of household and public buildings are connected to district heating (DH) systems. Only about 17 cities have access to DH networks, all of which face stiff competition from subsidized electricity, fuelwood and natural gas. Because of low regulated energy prices in the past, there was an excessive use of electricity for heating, the least fuel-efficient way of space heating. About 70 percent of total annual sales of electricity are to households and the public sector, which use about half of it for heating. 8.31. So far, actions to improve demand-side EE have been limited. Up to now, efforts to improve EE have concentrated on the supply-side, through the rehabilitation of updated infrastructure in the power and DH sectors. However, progress in improving EE on the demand-side is very limited. The lack of progress is compounded by the cumulative effect of policies that channeled subsidies to firms, and the population, through cheap energy, which stimulated excessive and wasteful consumption. 8.32. Significant efforts are required to create an institutional framework that would enable implementation of EE measures. The implementation, support, and diffusion of EE services and technologies will require establishment of an institutional framework comprising: (i) a policy framework, (ii) supporting structures; (iii) implementation and know-how, (iv) incentives, and (v) information. The country has made little progress in creating such an institutional framework, and it compares poorly to its regional neighbors (see Figure 8.6). 80 Source: International Resources Group, Final report of the Regional Energy Demand Planning Project – Future Energy Scenarios in South East Europe and the Potential for Energy Efficiency, USAID/IRG, 2008. 129 Figure 8.11: Institutional Framework in countries in South East Europe Source: Status of Energy Efficiency in the Western Balkans. The World Bank, 2010. 8.33. Some of the key missing elements of an institutional framework for EE in BH as seen in Figure 8.7 above include: ï‚· Legal and regulatory framework. The country needs to define administrative responsibilities, develop and approve EE legislation, set national goals for EE, and develop an EE strategy and action plans. ï‚· Supporting structures. BH needs to create an energy institute to advise the Government on energy policy, and provide and analyze statistical data to support policy formulation. Institutions to assess and implement EE measures are also needed; this function is now left to line ministries which lack this capacity. ï‚· Implementation. BH has adopted some EU energy efficiency standards, but most are not yet obligatory. EE hardware and software are not widely available, and neither are trained experts to install them; training programs are needed to develop such expertise. 130 ï‚· Incentives. Most energy efficiency incentives are lacking. BH needs to introduce financial incentives for EE. Energy utilities are not yet promoting energy savings through demand- side management programs. ï‚· Public information. A public information system to educate consumers and promote EE is yet to be created. 8.34. In addition, energy prices do not fully cover cost and residential consumers are being cross- subsidized by large industrial/commercial consumers. As discussed in the previous section, power tariffs are set below their full cost-recovery level. In addition, residential consumers are being cross- subsidized by non-residential consumers as can be seen in Figure 8.7 below. The gas and heat sectors also lack adequate regulation. There is no regulation of gas prices at the national level and tariffs are set by transmission and distribution companies. The tariffs for heat are also set by companies owned by the municipalities at below cost resulting in deterioration in service and infrastructure. Apart from the negative effect on energy efficiency investments, such tariff distortions also cause costly distortions in energy infrastructure investments. Figure 8.12: Electricity Prices in South East Europe in 2009 Residential Non-Residential 25 20 USD cents / KWh 15 10 5 0 Croatia Montenegro Romania Bulgaria Bosnia and Albania Macedonia Serbia UNMIK Herzegovina Kosovo Source: ERRA database B. TECHNICAL POTENTIAL FOR ENERGY SAVINGS IN PUBLIC BUILDINGS 8.35. There is a significant energy savings potential to be realized through EE measures in all end-use sectors, although it should be noted that end-use sectors like transport and residential, are particularly hard to deal with. Estimates for the South East Europe region by 2020 are shown below since there is no available data on energy savings potential in the country. Although, these estimates range widely by sector and country due to country differences and above all, scarcity of reliable data, they indicate a large potential for improving the efficiency of public expenditure by implementing end-use energy efficiency measures. 131 8.36. There is a significant scope to rationalize public spending in energy services by improving energy efficiency in public sector buildings. Electricity sales for public Table 8.5: Energy Savings Potential in South East lighting amounted BAM 12.4 million Europe by Sector according to EPBH and BAM 8.1 million Sector Energy Savings Potential (percent) according to EPRS in 2010. Regarding Transport 10 heat energy, two main District Heating Residential 10-35 utilities exist in the municipalities of Public 35-40 Sarajevo and Banja Luka. These regions Industrial 10-30 also represent the largest number of Service 5-25 public buildings as can be seen in Table Source: Status of Energy Efficiency in the Western Balkans. The World Bank, 2010. 8.6 below. According to data obtained from the Sarajevo and Banja Luka DH utilities, DH bills in the health, education, public administration, and consumption by military/police amounted to BAM 1.5 million in Sarajevo and BAM 2.8 million in Banja Luka in 200?. Table 8.6: Public Building Stock Cities/Municipalities Number of buildings including Number of buildings not including ―Institutions‖ and ―Funds‖* ―Institutions‖ and ―Funds‖* Federation of Bosnia and 3,673 3,457 Herzegovina Sarajevo 2,717 2,610 Mostar 889 781 Jablanica 67 66 Republika Srpska 2,167 1,867 Banja Luka 1,693 1,475 Pale 140 118 Prijedor 334 274 Source: FBH Tax Administration; RS Institute of Statistics *Buildings included in the category ―Institutions― and ―Funds‖ can be both public and private 8.37. However, the EE potential in public buildings is difficult to estimate due to lack of reliable data. Systematic data on energy consumption in the public buildings sector was not available in BH, despite considerable efforts to obtain them. Drawing from data obtained in a recent survey undertaken for the purpose of this PEIR and international benchmarks, EE savings potential was estimated for public and commercial buildings. Although figures were calculated by using savings potential based on implemented EE projects in the region (see Annex 3), differences in construction technologies and costs imply limitation of the accuracy of conclusions. However, they provide orders of magnitude for policymakers to rationalize public expenditure and improve the efficiency of energy use. 132 Box 8.1: Data Sources To Estimate EE Potential In Public Buildings Bosnia and Herzegovina lacks comprehensive and recent sectoral data on energy efficiency and consumption. There is no agency responsible for collecting this information on a consolidated basis. The two agencies for statistics the Federal Office of Statistics (in FBH) and the Institute of Statistics – (in RS) – are responsible for collecting data but no reliable, comprehensive national energy balance or data sets are available on a regular basis. To assess the weight of energy expenses in public buildings and estimate the potential for energy savings, a survey was undertaken in the context of this PEIR. The information was gathered from District Heating utility companies in Sarajevo and Banja Luka as well as information from public buildings in both FBH and RS in the following municipalities: Centar, Ilidža, Mostar, Jablanica, Pale and Prijedor. Based on insights from this information as well as regional benchmarks, sectoral estimates for energy savings potential have been estimated. The sources for the estimates are summarized below: ï‚· Energy consumption: Data series based on IEA energy balance figures. No disaggregated information for public vs. commercial sectors was available at the national/entity level. ï‚· Energy saving potential and investment requirements: Energy saving potentials and investment requirements are based on energy savings obtained from selected EE projects implemented in public buildings and lighting by the World Bank in the South East Europe region. ï‚· Building stock: Public building stock was provided by Tax Administration of FBH of Bosnia and Herzegovina and Republika Srpska Institute of Statistics. Given that neither of the aforementioned administration/institute has ―public buildings‖ as a single category, the following type of buildings wer e included in the stock: institution, state company, public company, social company, chamber of commerce, cooperative, local communities, administration bodies and administration organizations, administration institutions, citizens‘ associations, cooperatives‘ unions, and funds. 8.38. According to rough estimates, energy savings in public and commercial buildings amount to 85 thousand of tons of oil equivalent or 30 percent of the sector‟s energy consumption. This estimate is slightly lower than the estimates for the South East Europe region of 35-40 percent mentioned above since more conservative assumptions were used (see Annex 2 for details). Most energy saving potential is associated with increased use of thermal insulation to avoid heat loss. Thermal energy consumption in Bosnia and Herzegovina in 2005 was high at about 200 kWh/m2, compared to about 150 KWh/m2 in Western Europe. The large majority of buildings are old and lack maintenance; therefore, their insulation will be the major contributing measure in realizing this savings potential. Table 8.7: EE Saving Potential in Commercial and Public Buildings Parameter Total Non-Electricity Energy Consumption ('000 toe 2008) 142.3 EE-based Saving Potential percent 40 Consumption after EE Potential is Realized ('000 toe 2008) 85.4 Electricity Energy Consumption ('000 toe 2008) 139.9 EE-based Saving Potential percent 20 Consumption after EE Potential is Realized ('000 toe 2008) 28.0 Total Energy Consumption ('000 toe 2008) 282.3 EE-based Saving Potential, percent 30 Consumption after EE Potential is Realized ('000 toe 2008) 197.3 EE-based Saving Potential ('000 toe 2008) 84.9 133 C. BARRIERS TO EE IN PUBLIC BUILDINGS AND POLICY OPTIONS Below are some identified market barriers that need to be removed to jump-start EE investments: 8.39. Lack of reliable energy efficiency data: Reliable updates of energy consumption data are absent in BH, even the most basic data, including national energy balances and sectoral EE indicators. This has practical implications. First, it is difficult to calculate estimates for EE potential by sector, or to establish EE priorities. Second, it is difficult to prepare high-quality Energy Efficiency Action Plans with monitorable and realistic interim targets. 8.40. Inadequate legal framework and lack of coordinated EE policies: EE policy falls under the responsibility of the ministries responsible for energy. Implementation of such policies is further devolved to the Cantonal level in FBH and the Municipal level in the Republika Srpska. However, a formal policy is not in place related to EE. In addition, the country‘s complex political structure poses additional challenges to coordination. The legal and regulatory framework to support EE measures have not been adopted either, although there is a growing requirement for the country to address EE issues through its adoption of the Energy Community Treaty and Energy Charter. Energy Efficiency is only indirectly covered in other legislation. Regulators, for example, have the responsibility of considering both environmental and energy efficiency issues in their tariff making and investment approval regulations and decisions. 8.41. Low awareness and lack of information: Consumers at all levels, including asset managers in local governments and line ministries responsible for building operation and maintenance, lack information to support EE measures and behavior changes. For example, they often lack meters or heat cost allocators for DH and billing information (all forms of energy) that would make them aware of their patterns of energy consumption. As a result, they have no idea how their consumption compares to that of their peers, how their consumption could be reduced, or what the benefits of reduced consumption are. Training and know-how on EE measures are also scarce. Few architects, engineers, plumbers, and installers have the technical skills or knowledge to exploit energy savings potential and training is unavailable. 8.42. Insufficient Resources and Support: There is a need to build capacity on EE within Government institutions and to allocate adequate resources to encourage EE investments. There is no designated Energy Institute and/or Agency to provide governments with the data to formulate policies, and to be the ―change agency‖ or ―champion‖ to help implement policy and establish and operate the dissemination support scheme for EE services and technologies. The country is also running behind schedule in the preparation of its National Energy Efficiency Action Plan, which is required by the European Community Directive 2006/32 due to: (i) a lack of energy data, including EE indicators; (ii) a lack of qualified staff; and (iii) limited government support due to lack of good data; hence, governments are unaware of the extent of the problems they face in the energy sector. 8.43. Incentives of all kinds are lacking, including low energy prices: In addition to price distortions in energy (see above discussion on power tariffs), the lack of consumption-based billing, although it is applied by a few DH Companies, creates a disincentive for those buildings connected to DH systems to reduce consumption and at the same time does not enable heat suppliers to improve the efficiency of production and delivery systems. In addition, there are split incentives in investing in EE; the direct financial benefits of EE measures would not in most cases be captured by administrators of EE buildings (schools, hospitals) since energy bills are usually paid out of local and/or entity budgets. 8.44. Inadequate access to financing; The budgets available for addressing energy efficiency issues are very limited and compete with other critical areas of the economy for scarce funds. The primary 134 source of financing for energy efficiency investments, for the time being, is regular commercial loans which are usually short-term and high-cost. To jump-start the market, longer maturities would be required to allow more comprehensive initial investments which can both renew buildings and achieve significant energy savings. This fact, coupled with low level awareness and the difficult financial situation of many governments rank energy efficiency investments low in priority. 8.45. To address some of the barriers identified above, the Government may consider the following policy options based on international experience in the region: ï‚· The public sector should adopt an exemplary role by introducing EE demand-side measures in all its buildings, and facilities. Taking the lead would include the following: (i) implement public sector EE projects, especially in buildings frequented by the general public, to maximize the demonstration effect and create demand; (ii) establish building codes and equipment standards and ensure enforcement. This would also help BH conform to the EU Directive on building energy performance; (iii) enable start-up and facilitate demand for an adequate number of energy services companies; and (iv) establish programs to train and certify energy auditors, managers, and contractors. ï‚· In the context of an EE program in public buildings, schools and hospitals are priorities. In the short- to medium-term, prospects are slim for major, sustainable EE programs in the presence of distorted prices and limited financing. But public-sector programs for schools, hospitals, and government buildings can promote comfort, health, as well as energy savings that will increase over time as energy prices rise. Experience in the region has shown that prioritizing EE investments in schools and hospitals helps generate acceptance by the local communities and promotes political support for EE measures. ï‚· Develop a basic legal framework and set up institutional arrangements for the implementation of EE programs: There is a need to develop a policy on EE as well as a legal and regulatory framework at the state and at the entity levels. However, experience has shown that the creation of institutional arrangements, such as EE agencies, by which EE policy is formulated and implemented is key to the successful implementation of comprehensive EE policies. The Government may consider selecting and providing funding to a ministerial department or EE agency as the responsible authority to, among others, implement policies and programs, and improve data collection. ï‚· Set up a comprehensive and systematic energy data gathering and reporting system: The establishment of such system, consistent with those of EUROSTAT, is an EU accession requirement. Together with Odyssee, a detailed database on EE data and indicators for the EU-27 plus Norway and Croatia, these institutions could provide useful information on how to set up basic energy data systems (www.odyssee-indicators.org). ï‚· Energy efficiency funds and other supporting mechanisms could be considered. Many EE measures require investments with a payback period that is longer than the financing that may be available in commercial terms. The creation of EE funds and/or other financing support mechanisms could provide subsidies for implementing EE investments, shortening the payback periods, and help provide access to below-market rate financing where necessary. The Funds for Environmental Protection in FBH and RS could for instance support EE investments in public buildings. ï‚· Launch targeted information campaigns to raise awareness of the benefits of EE among consumers. It takes a long time to educate consumers, certify energy auditors, manufacturers and contractors, and introduce and enforce energy building codes and efficiency standards for homes, appliances, and vehicles. Such programs should be started as soon as possible. 135 IV. RECOMMENDATIONS 8.46 The energy sector is already having a significant impact on public finances and steps are needed to ensure that sector does not become a greater burden in the future. The key to controlling the fiscal impact lies largely with the power and coal sector. The efficiency of both needs to be sharply improved and new investments made to meet rising demand. This will only be possible if the authorities confront the need to overhaul the electricity tariff regime. These reforms need also to be supplemented with energy efficiency measures particularly in the buildings sector which accounts for full one third of total final energy consumption and is highly inefficient. The following recommendations are directed at meeting these objectives. ï‚· Continue the restructuring and modernization of coal mines within the new institutional setup. Efforts should focus on improving efficiency by adopting modern technologies and improving labor productivity. ï‚· Continue the restructuring of the electricity sector towards EU standards for corporate governance; and set tariffs at their full cost-recovery levels, while putting in place targeted social protection mechanisms to protect vulnerable population. ï‚· Prepare and adopt an investment strategy for each of the power utilities based on a comparative least-cost investment plan which takes into account the power sector demand forecast at the entity, national, and regional levels. Options for private sector involvement should also be considered. ï‚· Adopt an exemplary role by implementing an EE program in public buildings, focusing for instance in schools and hospitals. ï‚· Develop a basic legal framework and set up institutional arrangements, including establishing a “change agencyâ€? to champion the implementation of EE programs. ï‚· Set up a comprehensive and systematic energy data gathering and reporting system and launch targeted information campaigns. ï‚· Establish a financing support mechanism for EE investments, such as Energy Efficiency Funds. 136 ANNEX 8.1: PLANNED INVESTMENTS BY THE STATE UTILITIES TILL 2015 (in million BAM) EPBH EPHZHB EPRS Generation Distribution Generation Distribution Generation Distribution New New New Rehabilitati Capacity Rehabilitatio Construc Rehabilita Capacity Rehabilita Construc Rehabilit Capacity Rehabilita Constructi on Addition n tion tion Addition tion tion ation Addition tion on - - - - - - - - Until 2009 - 5.7 - 3.0 - - - - - 2010 8.1 31.3 98.4 - 1,375.7 - 122.7 - - - 2011-2013 1,172.9 68.8 239.3 - - 161.1 531.0 63.8 197.0 - - - - - - - 2014-2015 1,909.3 - 111.6 - - Total 3639.7 1368 1075.6 Source: World Bank based on data provided by the power utilities. 137 ANNEX 8.2: ENERGY SAVING POTENTIAL BY ENERGY EFFICIENCY IN COMMERCIAL AND PUBLIC BUILDINGS (detailed calculations) Parameter Source / Note Figure Buildings Total Energy Consumption ('000 toe, 2008) IEA 936.4 Share of Non-Electrical Fuels in Building Consumption IEA 76 percent (2008) Buildings Non-Electrical Energy Consumption ('000 toe, Calculated 711.7 2008) Share of Commercial and Public in Total Building Area BH Energy Sector Study. Energy Institute Hrvoje 20 percent (2008) Požar, Croatia Soluziona, Spain Economics Institute Banjaluka, BH Mining Institute Tuzla (2008) – Module 14 Commercial and Public Buildings Non-Electrical Energy Assuming area breakdown of residential non 142.3 Consumption** ('000 toe, 2007) residential buildings is a good approximation of breakdown of non-electrical consumption. Saving Potential by Insulation percent Data from EE projects implemented in the South 40 percent East Europe Region Saving Potential by Insulation in Commercial and Public Disregarding heating by electrical appliances and 56.9 Buildings ('000 toe, 2008) disregarding non-heating uses of non-electricity fuels Electricity Consumption in Commercial and Public Buildings IEA, completed by data from EPs 139.9 Commercial and Public Buildings Electricity EE potential Estimate based on the 20percent saving potential 20 percent suggested by the Energy Sector Study (mentioned 28.0 above) which is in line with the general statement that buildings in BH have 35-40percent EE saving potential Total Saving Potential in Commercial and Public 84.9 Buildings ('000 toe, 2008) Share of Energy savings Potential in Total Energy 3.1 percent Consumption of BH 138 ANNEX 8.3: EXAMPLES OF ENERGY EFFICIENCY PROJECTS IMPLEMENTED IN SOUTH EAST EUROPE Serbia Energy Efficiency Project. The first phase of the World Bank-funded Serbian Energy Efficiency Project aimed to improve EE in public hospitals and schools—27 buildings with a heated area of 104,969 m2 were renovated. Specific building heat consumption dropped from 266 kWh/m2/year to 162 kWh/m2/year (or 39percent) at investment costs of around 35 EUR/m2. Hospitals‘ specific heat consumption was reduced from 329 kWh/m2/year to 210 kWh/m2/year. With investments with costs ranging from 21.1 to 58.8 EUR/m2, annual energy demand for space heating could be further reduced 44 percent. Most energy savings derived from roof insulation and new windows and doors. The project is financed by a US$21 million IDA credit (2004). Additional financing of US$28 million was approved in 2007 and is financing similar projects in other hospitals and schools in Serbia. Bulgaria Energy Efficiency Project. The Global Environmental Facility and the World Bank approved a $10 million grant in 2005 to provide the bulk of initial capitalization for the Bulgarian Energy Efficiency Fund (BEEF), a revolving fund. The seed capital was $16 million, including contributions from the Bulgarian and Austrian governments and private Bulgarian firms. The fund has a flexible combination of financial products, public private partnership in its capitalization and management, strong country ownership, a manager with strong financial skills and local knowledge, and a solid pipeline of finance- ready projects. The total leveraged investment by BEEF‘s loans was US$24 million, mainly in investments in public buildings and public lighting. Demonstrated savings potential in public school buildings was of 40 percent. Croatia Energy Efficiency Project. HEP Energy Service Company, part of Croatia‘s national electricity utility, was created in 2003, with support from a World Bank loan and a Global Environmental Facility grant. Overall, 31 EE projects on a commercial, for-profit basis, for a total cumulative value of US$29.5 million in EE investments were implemented, showing the potential of utility-based energy service companies. Initially focused on implementing EE projects in public lighting and schools, HEP ESCO diversified its business activities and is now the leading EE Company in Croatia, active in six market segments: public lighting, schools, hospitals, cogeneration, industry and commercial buildings. 139