Report No. 78320-HR CROATIA PUBLIC FINANCE REVIEW Restructuring Spending for Stability and Growth October, 2014 Macroeconomics and Fiscal Management Global Practice Europe and Central Asia Region Document of the World Bank ii January-December (Exchange Rate Effective as of October 20, 2014) EUR 1.00:USD 1.280:HRK 7.660 Weights and Measures: Metric System ABBREVIATIONS AND ACRONYMS ALMP Active-Labor Market Policies IPA Instruments for Pre-Accession Assistance APAFRD Agency for Payments in Agriculture, IPH Institute for Public Health Fisheries and Rural Development ASNC Areas of Special National Concern ISPA Instrument for Structural Policy Pre- Accession AWU Annual Work Units LFS Labor Force Survey CAP Common Agricultural Policy LGU Local Government Unit CARDS Community Assistance for Reconstruction, LTC Long-Term Care Development and Stabilization CCA Croatian Competition Agency LTO Large Taxpayer Office CIHI Croatian Health Insurance Institute MA Managing Authority CNB Croatian National Bank MAIC Multi-Annual Infrastructure Contract CNDP Complementary National Direct Payments MBO Management Buy-Out COFOG Classification of General Government MFF Multiannual Financial Framework Expenditure by Function CPI Consumer Price Index MIS Management Information System CRMS Compliance Risk Management Model MMATI Ministry of Maritime Affairs, Transport and Infrastructure CROSTAT Croatian Bureau of Statistics MoF Ministry of Finance CTA Central Tax Administration MoSPY Ministry of Social Policy and Youth EAFRD European Agricultural Fund for Rural MRDEUF Ministry of Regional Development and Development EU Funds EAGF European Agricultural Guarantee Fund NEETD Not in Employment, Education or Training, not Retired, and not Disabled EC European Commission NGO Non-Governmental Organization ECA Europe and Central Asia NHA National Health Accounts EDP Excessive Deficit Procedure NMS New Member States EFC Error, Fraud & Corruption NRDP National Rural Development Program ESI European Structural Investment Funds OECD Organization for Economic Cooperation and Development EU European Union OP Operational Program GDP Gross Domestic Product PAYGO Pay-As-You-Go GMB Guaranteed Minimum Benefit PPP Public-Private Partnership GMI Guaranteed Minimum Income PSC Passenger Service Contracts GNI Gross National Income R&D Research & Development HANFA Financial Services Supervisory Agency RoR Rates of Return HBS Household Budget Survey SAPARD Special Accession Program for Agriculture and Rural Development HMA Hilly and Mountain Areas SGEI Services of General Economic Interest HRMIS Human Resources Management Information SILC Survey of Income Living Conditions iii System HWV Homeland War Veterans SME Small and Medium Enterprises HZ Croatian Railways TAC Track Access Charge HZMO Croatian Pension Institute TOR Traditional Own Resources ICMS Integrated Case Management System VAT Value Added Tax IMF International Monetary Fund WEF World Economic Forum Vice President: Laura Tuck, ECAVP Country Director: Mamta Murthi, ECCU5 Director: Satu Kahkonen, GMFDR Practice Manager: Miria Pigato, GMFDR Task Team Leader: Sanja Madzarevic-Sujster, GMFDR iv ACKNOWLEDGEMENTS This report is the product of the World Bank team led by Sanja-Madzarevic-Sujster and comprised of Theo Thomas, Zeljko Bogetic, Gallina Vincelette, Ivan Kusen, Matija Laco, Roberto Fattal Jaef, Velimir Sonje, Julia Dhmitri, Raul Junquera, Ivana Maletic, Emilia Skrok, Ivana Jovic, Vasile Olievschi, Blanka Babic, Radoslaw Czapski, Jean-Francois Marteau, Lana Andricevic, Emil Tesliuc, Danijel Nestic, Vlatka Bilas, Marcelo Bortman, Agnes Couffinhal, Zoran Anusic, Anto Bajo, Irina Ramniceanu, Miroslav Bozic. Kim Murrel and Anne Grant provided the editorial support, while Dubravka Jerman provided the administrative support. The report was undertaken under the guidance of Mamta Murthi, Country Director, Satu Kahkonen, Practice Director, and Hongjoo Hahm, Country Manager. The team appreciated their comments, advice, and help throughout the course of this work as well as those from Elena Reitano (European Commission, DG ECFIN) and her Croatia team, and Johannes Wiegand (IMF Mission Chief) and his Croatia team. The responsibility for any data and/or opinion expressed in this paper remains exclusively that of the authors. The team would like to thank Mr. Maroje Lang (Assistant Minister of Finance), Ms. Jadranka Hajdinjak (Head of Unit, Ministry of Finance) for consolidating comments from the line ministries. The draft report was presented to the Government on June 4, 2014 and the team remains grateful for the feedback and comments received by line ministries. v Table of Contents EXECUTIVE SUMMARY ............................................................................................................ x I. INTRODUCTION ................................................................................................................ 1 II. CROATIA’S MACRO VULNERABILITIES AND FISCAL ADJUSTMENT CHALLENGES .............................................................................................................................. 5 A. Croatia’s Fiscal Performance ........................................................................................... 8 B. The Government’s Medium-Term Fiscal Program........................................................ 11 C. Debt Sustainability Analysis and Required Fiscal Adjustment ..................................... 14 D. Composition of Public Revenues and Spending ............................................................ 17 III. MAXIMIZING THE EFFICIENT USE OF EU FUNDS .................................................. 22 A. Pre-Accession Funds ...................................................................................................... 22 B. EU Financial Support for Croatia After Accession ....................................................... 25 C. Effective Utilization of EU Funds ................................................................................. 32 IV. REVENUE-SIDE ADJUSTMENT .................................................................................... 38 A. Revenue Reforms for Fiscal Consolidation ................................................................... 38 B. Croatia’s Tax System ..................................................................................................... 41 C. Strengthening Tax Administration to Improve Compliance .......................................... 44 V. RIGHTSIZING THE GOVERNMENT ............................................................................. 48 A. Entrenched Public Sector Management System ............................................................ 48 B. Local Government Finances .......................................................................................... 60 VI. IMPROVING THE EFFICIENCY AND EQUITY OF SOCIAL SPENDING ................. 70 A. Improving the Efficiency, Equity and Delivery of Health Care .................................... 72 B. Pension System Sustainability ....................................................................................... 83 C. Long-Term Care............................................................................................................. 94 D. Social Assistance ........................................................................................................... 99 VII. RATIONALIZING SUBSIDIES ..................................................................................... 109 A. Size and Structure of Croatia’s State Aid .................................................................... 109 B. Rationalizing Railway Subsidies ................................................................................. 115 C. Rationalizing Agriculture and Rural Development Subsidies ..................................... 124 VIII. ENSURING SUSTAINABILITY .................................................................................... 136 ANNEX 1: Literature .................................................................................................................. 140 Figures Figure 1. Croatia’s Per Capita Gross National Income and At-Risk-of-Poverty Rates, 2002-13 ................. 1 Figure 2. Croatia’s Macroeconomic Imbalances, 2002-13 ........................................................................... 2 Figure 3. Croatia’s Competitiveness Indicators ............................................................................................ 3 Figure 4. Effects of Job Finding and Exit Rates on Unemployment Dynamics ........................................... 7 Figure 5. Net Real Wages and Value Added Per Worker ............................................................................. 7 Figure 6. Croatia’s Fiscal Performance, 2008-13 ......................................................................................... 8 Figure 7. Evolution of Croatia’s Public Debt, 2008-13 (percent of GDP).................................................... 8 vi Figure 8. Croatia, EU10 and EU15: General Government Debt, Percent of GDP ....................................... 9 Figure 9. Croatia, EU10 and EU15: Interest Payments, Percent of GDP ..................................................... 9 Figure 10. 5-Year CDS Spreads: Croatia in the Eurozone Perspective ...................................................... 10 Figure 11. Public Debt and Sovereign Ratings, 2013 ................................................................................. 13 Figure 12. Public Debt and Economic Growth, 2000-07 ............................................................................ 14 Figure 13. Public Debt and Economic Growth, 2008-12 ............................................................................ 14 Figure 14. Cumulative Adjustment of Primary Expenditures/Revenues, 2014-17 ..................................... 15 Figure 15. Public Debt Sustainability Scenarios and Shock, Public Debt as % of GDP ............................ 16 Figure 16. Composition of Change in Public Debt, Percentage Points of GDP ......................................... 17 Figure 17. Evolution of Public Debt and Gross External Financing Requirements (in % of GDP) ........... 17 Figure 18. Social Sector Spending in 2012, percent of GDP ...................................................................... 19 Figure 19. Performance of Government Services ....................................................................................... 20 Figure 20. EU Funds Inflow to Croatia, Percent of GDP and Per Capita ................................................... 24 Figure 21. Net Impact of EU Funds on the Fiscal Deficit, ESA95, Percent of GDP .................................. 29 Figure 22. First-Round Demand Effect of EU Funds, Percent of GDP ...................................................... 31 Figure 23. Average Property Taxes in Selected OECD economies, 2000–11, Percent of GDP................. 39 Figure 24. Revenues in 2013, Percent of GDP ........................................................................................... 41 Figure 25. Social Contributions in 2013, Percent of GDP .......................................................................... 41 Figure 26. Indirect and Direct Taxes in 2013, Percent of GDP .................................................................. 42 Figure 27. Low Tax on Profits, Percent of Profits ...................................................................................... 42 Figure 28. Total tax revenues and per capita income GDP, selected EU countries, 2012 .......................... 42 Figure 29. The Inactivity Trap .................................................................................................................... 44 Figure 30. General Government Wage Bill, Percent of GDP ..................................................................... 48 Figure 31. Size of the Government ............................................................................................................. 49 Figure 32. Croatia’s Administration Effectiveness ..................................................................................... 50 Figure 33. Efficiency of government performance ..................................................................................... 51 Figure 34. Wage Differentials Based on Hourly Wages by Sector and Education Level ........................... 52 Figure 35. Recent (2009–13) and Historical (1979–2009) Wage Bill Consolidation Episodes1 ................ 54 Figure 36. Average Inhabitant per Local Government, EU, 2011 .............................................................. 61 Figure 37. Percentage of General Government Expenditures Executed at the Local Level ....................... 62 Figure 38. Local Spending, EU and Croatia, 2011 ..................................................................................... 62 Figure 39. Subnational Budget Items, per Capita, 2009–12 ....................................................................... 63 Figure 40. Expenditures, Unconsolidated (% of GDP) ............................................................................... 63 Figure 41. Fiscal Imbalances, 2002–2012 .................................................................................................. 67 Figure 42. Coverage of Subnational Spending by Subnational Tax Revenue ............................................ 67 Figure 43. Proportion of Population 65 and over, EU, 2012 ...................................................................... 71 Figure 44. Median Age of Population, EU, 2012 ....................................................................................... 71 Figure 45. Croatia-Proportion of Population aged 65+ and 80+................................................................. 71 Figure 46. Demographic Transition in Croatia (2011–61).......................................................................... 71 Figure 47. GDP and Life Expectancy at Birth in Selected Countries, 2010 ............................................... 72 Figure 48. Inequality in Reported Long-term Illness in Croatia and Selected EU Countries, 2010 ........... 73 Figure 49. Health Expenditure and GNI per capita, Croatia in the World (2010) ...................................... 76 Figure 50. Structure of Health Spending, Percent ...................................................................................... 76 Figure 51. Changes in the Share of Out-of-pocket Health Care Payments, 1997-2010 ............................. 80 Figure 52. Population above 65 and Old Age Dependency Ratios, Transition Countries .......................... 83 Figure 53. System Dependency Rates, High Income and Transition Countries ......................................... 83 Figure 54. Contributor Coverage, Croatia, High-Income and Transition Countries................................... 84 Figure 55. Beneficiary Coverage, Croatia and High Income and Transition Countries ............................. 85 Figure 56. Life Expectancy at Retirement, 2010, Europe ........................................................................... 85 Figure 57. International Comparisons of Social Insurance Contribution Rates .......................................... 86 Figure 58. Total PAYG Pension Expenditures, percent of GDP ................................................................ 89 vii Figure 59. Inverse Dependency Ratios, Croatia, 2011–50.......................................................................... 96 Figure 60. Targeting Accuracy in the Bottom Quintile, 2008-2011 ......................................................... 101 Figure 61. The Difference in Personal Income Tax Paid by Wage Earner With and Without Children (2012) ........................................................................................................................................................ 102 Figure 62. Croatia’s Total Non-Crisis State Aid and Its Structure, Percent of GDP ................................ 110 Figure 63. Quality of Railroad Infrastructure and Croatia’s State Aid ..................................................... 114 Figure 64. The Quality of Ports Infrastructure .......................................................................................... 114 Figure 65. Horizontal aid, as % of total aid less agriculture and fisheries and transport .......................... 115 Figure 66. Sectoral aid, as % of total aid less agriculture and fisheries and transport .............................. 115 Figure 67. Structure of Inland Transport by Types in Croatia, 2012 ........................................................ 115 Figure 68. Rail Traffic Intensity Compared to the EU (2012); Croatia=100 ............................................ 117 Figure 69. Age Structure of Infrastructure and Rolling Stock of the Croatian Railways ......................... 118 Figure 70. Evolution of State Budget Support to Railways ...................................................................... 119 Figure 71. Difference between the World Bank and HŽ Estimated Public Funds to Railway Sector, 2012- 2016 (in HRK thousands) ......................................................................................................................... 120 Figure 72. Budget Spending on Agriculture and Rural Development, 2004-12 ....................................... 124 Figure 73. Agriculture and Rural Development Spending by Program (million HRK, 2004-12) ............ 124 Figure 74. Farm Labor Productivity, Croatia and the EU (2005-11 average) .......................................... 125 Figure 75. Average Standard Output per Farm, by Agricultural Size Clusters (thousand EUR, 2010) ... 126 Figure 76. Agricultural Policy Costs........................................................................................................ 128 Figure 77. EU CAP Payment Requirements ............................................................................................. 130 Figure 78. Croatia’s Annual Financial Allocations under the CAP, 2014-20 (EUR million) .................. 131 Figure 79. Pre and Post-accession Simulation of EAGF Direct Payments and CNDPs ........................... 133 Figure 80. Fiscal Policy Adjustment in 2009-2012, percent of GDP, ESA 95 ......................................... 137 Figure 81. Fiscal Consolidation Among OECD Countries ....................................................................... 138 Tables Table 1. Croatia: Key Macroeconomic Indicators (Percent of GDP) ........................................................... 6 Table 2. Croatia: Disentangling Crisis from Structural Impact .................................................................... 9 Table 3. Public Debt Structure and Dynamics, Percent of GDP ................................................................. 10 Table 4. Fiscal Developments and Prospects 2010-2016............................................................................ 12 Table 5. Medium-Term Fiscal Scenarios, Percent of GDP ......................................................................... 14 Table 6. General Government Expenditures by Economic Classification, Percent of GDP....................... 18 Table 7. General Government Expenditures by Function (COFOG) in 2012, Percent of GDP ................. 19 Table 8. Implementation of EU assistance to Croatia, million EUR, March 2014 ..................................... 23 Table 9. Payment Appropriations 2014-2020, percent of GDP .................................................................. 26 Table 10. Payment Appropriations from the EU for Croatia in 2013 ......................................................... 27 Table 11. Estimated Payments from Croatia to the EU Budget in 2013-2020, Percent of GDP ................ 28 Table 12. Estimated National Co- and Pre-Financing, 2014-2020, Percent of GDP .................................. 28 Table 13. Net Fiscal Effect, Cash Basis, Percent of GDP........................................................................... 29 Table 14. Conventional Wisdom: Advice for the Revenue Side of Consolidation..................................... 38 Table 15. What Reform Measures Have Countries in the Region Introduced in 2010-2013?.................... 40 Table 16. Revenue Potentials, Difference in Potentials, and Tax Space, 2011 .......................................... 43 Table 17. Selected Countries Tax Effort, 1994-2009 ................................................................................. 43 Table 18. Public Administration in Croatia, 2000-2012 ............................................................................. 51 Table 19. Local Government Units and Population Distribution ............................................................... 61 Table 20. Distribution of Local Units According to Decentralized Function ............................................. 62 Table 21. Structure of Economic Classification of Local Spending, 2009-12 ............................................ 64 Table 22. Revenue Sources of Croatian Local Governments, 2009–12 ..................................................... 65 Table 23. Model of Income Tax Revenue-Sharing, 2013 ........................................................................... 66 Table 24. Health Indicators in Selected European Countries and Croatia, 2009–2010 .............................. 73 viii Table 25. Infant Mortality Rate by County, 2001-10.................................................................................. 74 Table 26. Issued Prescriptions by County Office, 2011.............................................................................. 75 Table 27. Surgeries Performed in Selected Hospitals, 2011 ....................................................................... 77 Table 28. Savings in Outpatient Costs Compared to Inpatient Costs for the Same Procedure with Proven Medical Evidence........................................................................................................................................ 78 Table 29. Prescriptions, 2005-2012 ............................................................................................................ 79 Table 30. Annualized Net Rates of Return (RoR) of Mandatory Second Pillar ......................................... 88 Table 31. Early Retirement Decrement, Old and New ............................................................................... 92 Table 32. Fiscal Impact and Replacement Rate—Simulation Results ........................................................ 93 Table 33. Number of Institutional LTC Beneficiaries ................................................................................ 96 Table 34. LTC Institutional Facilities by Type of Management ................................................................. 97 Table 35. LTC Employees by Type of Institution ...................................................................................... 97 Table 36. Trends in LTC Funding, 2003 and 2011 ..................................................................................... 98 Table 37. Projected Public Expenditures on Health and LTC .................................................................... 98 Table 38. Social Benefits Spending, 2004-2013 ....................................................................................... 100 Table 39. Targeting Accuracy of Social Protection Programs, 2011 ........................................................ 101 Table 40. Coverage: Proportion of Population Receiving the Transfer, 2011 .......................................... 105 Table 41. Generosity of Social Benefits, 2011 ......................................................................................... 105 Table 42. Coverage of the Working Age Population by Social Assistance, 2010 .................................... 106 Table 43. Eliminating Elite Capture in Non-Contributory Social Assistance Programs .......................... 107 Table 44. Croatia State Aid to Industry and Services per Aid Instrument (without Agriculture)............. 112 Table 45. Croatia State Guarantees, 2003-2012, in HRK million ............................................................ 112 Table 46. Default Risk for Invoking of State Guarantees, 2003-2011, in HRK million........................... 112 Table 47. State Aid to Transport Sector (including Railways) ................................................................. 113 Table 48. Key Performance Indicators of Railways ................................................................................. 117 Boxes Box 1. Structure of the Croatian General Government Debt ...................................................................... 10 Box 2. The Excessive Deficit Procedure for Croatia .................................................................................. 11 Box 3. EU Pre-Accession Assistance Programs (2001-2013) .................................................................... 23 Box 4. The Principle of Additionality ......................................................................................................... 25 Box 5. EU Transfers to Member States ...................................................................................................... 25 Box 6. EU-Related Outflow of National Funds .......................................................................................... 27 Box 7. Framework for Evaluating Fiscal Impact of EU Transfers ............................................................. 30 Box 8. Challenges to Management of Structural Funds ............................................................................. 34 Box 9. Public Sector Compensation is Being Cut across the EU ............................................................... 53 Box 10. The Draft Civil Service Salary Law .............................................................................................. 56 Box 11. The Austrian Performance Appraisal System ............................................................................... 57 Box 12. Top Level Appointments Procedures - Some EU Examples......................................................... 58 Box 13. Mergers and Coordination ............................................................................................................. 65 Box 14. Reform of the Portuguese General-Purpose Grant System ........................................................... 67 Box 15. Health Sector Finances: Data Issues ............................................................................................. 76 Box 16. Croatia’s Pension Reform in Perspective ...................................................................................... 84 Box 17. Basic Parameters of the Croatian Pension System, 2013/14 ......................................................... 86 Box 18. Reversing Low Fertility and Low Female Labor Force Participation ......................................... 102 Box 19. Welfare-to-Work Reforms in Selected EU Countries ................................................................. 106 Box 20. Varieties of State Aid and Its Instruments in the EU .................................................................. 110 Box 21. Services of General Economic Interest ....................................................................................... 111 Box 22. State Aid and National Programs in Agriculture......................................................................... 130 ix 1. On July 1, 2013, Croatia became the 28th member of the European Union. This achievement crowned more than a decade of macroeconomic and institutional reforms by the Croatian authorities and other stakeholders that yielded important development results. Croatia’s institutions are now stronger than a decade ago, reflecting broad and deep institutional adjustments that underpinned the pre-accession process and Croatia has become a high-income country within a decade. 2. Yet, the global economic crisis, with the loss of credit, exposed Croatia’s macroeconomic vulnerabilities. The stimulus for a significant share of Croatia’s pre-crisis growth has been withdrawn resulting in five consecutive years in recession through 2013. Prior to the crisis, large, relatively cheap capital inflows circulated into the economy, creating credit, consumption, and real estate booms that subsequently reversed. Such capital inflows are not expected to return, due to Croatia’s weak growth outlook and the more competitive, risk adverse post-crisis international environment. Unemployment rose to 17 percent in 2013, much higher than the Eurozone average (12 percent). Public debt has doubled since 2008 and remains on an upward trajectory, and the private sector has continued deleveraging. Credit agencies took note, reducing Croatia’s sovereign debt to speculative status in 2013. 3. Underpinning these macroeconomic imbalances, Croatia faces deep structural problems that are holding back a recovery of output, exports, and jobs:  Croatia has only recently begun to improve the flexibility of its labor market by making it easier to hire or release workers in the formal sector, which has been contributing to high and persistent unemployment.  The business climate for domestic and foreign investors remains cumbersome1. Perhaps because Croatia received relatively robust capital inflows before the global crisis, there was little pressure to improve the investment climate, which is weaker than in many high- income countries.  Croatia delayed fiscal adjustment in response to the global crisis in the hope that the country would “grow out” of the downturn as global economic conditions improved, despite growing deficit and debt levels. As a result, public debt and fiscal positions have become unsustainable; and  Public sector efficiency is at the low end and the cost of public services at the high end among the European club of countries to which it acceded. The fiscal footprint of the state remains comparatively large, both in terms of employment and involvement in productive sectors. This has been a drag on the fiscal deficit and debt levels. 4. The country now faces a twin challenge of strengthening macroeconomic stability while fostering recovery of growth and competitiveness. Reducing fiscal vulnerabilities will 1 Croatia ranks second at the bottom of European countries in the 2014 Doing Business rankings: www.doingbusiness.org x be key to macroeconomic stability and laying the foundation for stable growth. To promote growth and competitiveness, structural reforms in two areas will be critical: first, lightening of the regulatory burden in the labor market and measures to improve the investment climate; and, second, the rationalization of public expenditures and the reduction of the excessive size and cost of the public sector. Without urgent attention to these policies, the country is facing the threat of losing private sector confidence and crowding out private sector activity. There is also an inherent concern based on the earlier EU experience – that country may become a net contributor to the EU budget, instead of using EU grant funding to support growth and investments. 5. This report analyzes three interrelated issues needed to strengthen macroeconomic stability, and lay the foundation for a robust recovery:  First, it analyzes Croatia’s major fiscal weaknesses, risks, and alternative fiscal scenarios, and, on that basis, calculates that a fiscal adjustment of 4 percentage points of GDP will be needed over the medium term.  Second, it analyzes the institutional weaknesses and requirements for the efficient use of EU funds in the coming years. These funds, if used efficiently, will bolster growth and competitiveness. However, this requires additional fiscal space of 1.8 percentage points of GDP per year, and strengthened institutions to fully benefit from them. Otherwise, there is a risk that these funds will be underutilized.  Third, the report analyzes the structure of Croatia’s public finances and provides options for both short-term controls and a sustainable medium term fiscal adjustment. While revenue and tax administration measures can contribute, the adjustment will need to be predominantly through expenditure measures, both short term and structural. These measures are designed not only to reduce the fiscal deficit and public debt to sustainable levels, but also to improve the efficiency of the public sector and to create fiscal space for the productive use of EU funds, improving the economy’s competitiveness. These adjustments will entail difficult political choices, and some short-term adjustment costs. Maximizing the Efficient Use of EU Funds 6. While the overall effect of EU transfers on Croatia’s economic growth will be positive, fiscal space needs to be created to support their utilization, averaging up to 1.8 percent of GDP a year in 2014-2020. With joining the EU, Croatia benefits from large inflows of resources, averaging 3.7 percent of gross domestic product (GDP) per year during 2014-2020. However, Croatia will also need to transfer part of its revenues to contribute to the EU budget as well as secure funds for pre- and co-financing. Given Croatia’s large fiscal deficit, it is important to manage EU-related funds within the overall fiscal consolidation process through expenditure switching and substitution policies. EU-funds can be used to mitigate negative effects on growth from the consolidation required to meet the fiscal deficit reduction trajectory agreed under the EC Excessive Deficit Procedure. This can be done by maximizing the growth enhancing objectives of EU funds through planning procedures, and by switching away from lower priority spending or substituting for domestic spending where possible. 7. Key lessons for more efficient absorption of EU Funds for Croatia include: the need to streamline and integrate budget procedures for EU funds into the national budgeting system; enhance the institutional capacity of agencies absorbing EU transfers also at the local xi government levels; and improve project preparation. Croatia would be well advised to undertake additional institutional strengthening in order to: (i) Link better regional and national priorities with the Cohesion Policy priorities to maximize the substitution principle of EU funds; (ii) Develop a clear strategic vision at all levels through the alignment of strategies at national, regional and local levels to ensure that funds have cumulative and long-term impact on growth rather than being allocated to one-off projects; (iii) Build institutions and administrative capacities related to the identification, preparation and implementation of projects that can be supported by EU funds, especially at the local level; and (iv) Set up a sound financial management system at all levels (local, regional, national) and in all EU funds beneficiaries. Revenue-Side Adjustment 8. With the revenue burden already close to 42 percent of GDP, which is relatively high compared to other EU countries at a similar level of income (EU102), there appears to be limited scope for increasing fiscal space from additional revenues. Nonetheless, there is scope to modernize property taxation, further rationalize quasi-fiscal fees, broaden the tax base by eliminating exemptions, and to shift the burden from the high levels of social contributions to promote growth and employment. In addition, business and government could benefit from further simplification of the tax system and strengthening of the Croatian Tax Administration (CTA) to promote compliance. These measures could raise additional revenue of 2 percent of GDP per annum, while reducing adverse impacts on growth and employment. In particular:  A modern value-based property tax could add up to 1.5 percent of GDP in new tax revenues. Currently the property tax base is deeply eroded by a combination of legal exemptions, an incomplete cadastral registry and a real estate market that does not provide transparent information on valuations. However, in addition to the revenue potential the burden rests on middle- and upper-income families, and causes less distortion to business and consumer decisions than other taxes.  Reducing the large number of tax exemptions could bring an additional one percent of GDP in new revenues. For example, a recent review of child tax allowances shows the potential to collect an additional 0.5 percent of GDP. Similarly, the reinvested earnings tax relief for businesses reduced revenue by 0.6 percent of GDP in 2013 while the benefits on investment are questionable. Both could be redesigned to eliminate the waste. The design of taxes and social benefits should also reduce the disincentives for moving from either benefits or inactivity to work. 9. Strengthening and modernizing tax administration would help protect and expand the revenue base. As Croatia becomes further integrated into the global economy, the tax administration (CTA) will face additional taxpayer compliance risks of considerable complexity. For the quality of its services to meet modern standards for tax administrations the CTA would have to move from being procedure-oriented to results-based. The building blocks of modern tax 2 EU10 indicates the 2004 and 2007 acceding EU member states, except Malta and Cyprus. xii administration are: (i) a compliance risk management system; (ii) solid administration at headquarters with a focus on the new Large Taxpayer Office (LTO); (iii) a streamlined network of regional and local tax offices; and (iv) sound IT governance. Rightsizing the Government 10. Croatia spends more on public administration than EU10, but performs poorly in terms of effectiveness, rule of law and administrative barriers to doing business. While the public wage bill, at close to 12 percent of GDP, is already high, upward pressure is likely to come from: (i) further decentralization and the constant creation of new agencies, often with higher average wages at the local/agency rather than national level; and (ii) the need to build additional capacity to absorb sizeable European funds. Nonetheless, some 2 percent of GDP in cumulative savings could be achieved over the medium term through staff rationalization at local and national governments and wage system reform to create a leaner, but more effective administration. To achieve this the following could be done:  Rationalization of the wage bill. Freezing the wage bill in the short to medium term will reduce immediate budget pressures, while long-term structural reforms will help sustain cost improvements and improve public sector performance. The full application of the Human Resource Management Information System (HRMIS), after several years’ delay, can improve cost controls, combined with a targeted downsizing program (for local governments, judicial, teaching and administrative/auxiliary staffing in particular) to reduce the duplication and inefficiency in the system, aligning employment with the provision of improved and more streamlined services where possible.  Civil service salaries, job classification and appraisal. Long-term performance improvements could come from enhancing merit-based recruitment, improving the performance appraisal system (and ensuring it is applied consistently across the civil service), full implementation of the harmonized system to classify posts, including measures to mitigate the impact of large numbers of local level civil servants not meeting qualification requirements of their posts, if needed.  Depoliticization and performance management. The depoliticization of senior management posts would need to be extended to the heads of directorate level through the same recruitment and appraisal arrangements that now apply to lower grades. While this principle should not be compromised on, it would be useful for the government to look into management modalities for top level appointments used in other EU member states. For the new system to be well embedded in an overall system of performance management, stalled attempts to enhance strategic planning and the introduction of performance based management practices in civil service institutions need to be reinvigorated.  Territorial reorganization would be the best way to address current disproportionate numbers and resources assigned to local and regional self- government units (LGUs). If not pursued for the lack of political support, territorial reorganization could be replaced by persuasion, incentives and coordination to encourage joint provision of public services.  Spending responsibilities could be redefined to avoid duplication and overlap of functions and increase accountability of LGUs for the tasks allocated to them. Fiscal xiii decentralization should clearly specify responsibilities (for example, in such areas as education, social protection and health) and identify resources to be used to finance them.  Greater subnational reliance on own-source revenues would help promote funding adequacy and reduce central government transfers. The ideal LGU tax is one that can be levied on a relatively immobile and well demarcated local tax base—a property tax (see Tax section).  The fiscal operations of subnational governments need to be monitored to ensure fiscal prudence and alignment with the Excessive Deficit Procedure, including through reporting all balance sheet and off-budget activities. Improving the Efficiency and Equity of Social Spending 11. Despite some solid social outcomes, Croatia’s social spending is excessive, inefficient, and inequitable. Croatia can achieve equivalent or better outcomes with lower and more targeted spending. The urgency of reform arises from short-term fiscal pressures and long-term aging and its consequences for the demand for social services. Some 2-3 percent of GDP in cumulative savings could be achieved over the medium term if the most inefficient programs are eliminated or rationalized and targeting improved. Some key priorities include the following:  Croatia’s health spending is excessive, and remaining inefficiencies and arrears call for significant reform. Croatia spends about 9 percent of GDP on health, compared to an average of about 5.4 percent in the EU10. Although health outcomes are good, they come at a high cost and the system’s sustainability is complicated by rapid aging of the population. The reform agenda can be centered around three main areas: (i) there is significant room to rationalize and improve the service delivery model in the Croatian hospital system the largest cost segment of the system; (ii) the high referral rate of primary health care can also be reduced and controlled; and (iii) increased transparency and better accounting and measurement to engender lasting gains in efficiency.  Croatia is one of the oldest and rapidly aging transition economies, but pension contribution rates are low and early retirement incentives high, putting pressure on the pension system. A less generous PAYG valorization and indexation pattern would bring the pension system to a fiscally and socially sustainable path. Accelerating an increase in the retirement age and tightening the early retirement would be needed to offset declining labor force participation. However, Croatia also needs to decide if the multi-pillar system is to be supported as the transition, after decade and a half, is not even at the half way point.  Containing growing long term care costs (LTC) with a missing middle down the road. Croatia has the basic infrastructure, but needs a comprehensive LTC plan. Such a plan would address coordination of services managed by different ministries and agencies, and market incentives to encourage the private sector to provide more long- term social care in the community. A reliable costing system would also ensure accountability for public funding, while a formal strategy for monitoring LTC social services would help the system adapt to changing needs while containing costs.  Croatia’s generous social protection system relies on weakly targeted, categorical rather than needs-based benefits, which reduces efficiency and contributes to the persistence of poverty and social exclusion. At 3.8 percent of GDP, Croatia’s social xiv assistance is costly. The largest share accrues to war veterans and their survivors and to families with children; well-targeted means-tested social assistance programs (guaranteed minimum income) account for only 0.3 percent of GDP. An improvement in targeting within the current spending envelope should be able to eliminate absolute poverty. A single, unified set of criteria to assess eligibility for needs-based social assistance programs could help enhance both equity and efficiency. Lastly, a “make-work-pay” benefit reform that includes active labor market measures (employment subsidies, skills training, and measures to promote jobs for disabled workers and youth) targeted at the long-term unemployed and long-term social beneficiaries, combined with enhanced inspections could bolster growth as well as reduce long-term costs. Rationalizing Subsidies 12. Sector and producer subsidies through state aid have averaged 2.4 percent of GDP in recent years and generated considerable contingent liabilities. However, there is little evidence that they contributed to improving performance in the targeted sectors, especially for railroads, ports, and agriculture. Analysis of select subsidies suggests significant scope for rationalization, with potential savings of around 1 percent of GDP. This implies eliminating the most inefficient subsidies, such as product subsidies in agriculture and rationalizing sector subsidies, especially to railroads, ports, steel and shipyards, and tourism. This will help the fiscal consolidation program, improve efficiency, and raise welcome competitive pressure in these sectors. 13. Given the large imbalances in the size and structure of Croatia’s aid relative to EU standards, these reforms should quickly transition from sector-specific to horizontal types of aid (like R&D). Finally, with large additional resources available from the EU, especially in rail and agriculture, institutional strengthening and improvements in efficiency are required to raise absorption capacity, competitiveness and the overall performance of these sectors. Ensuring Sustainability 14. Croatia has entered European Union, achieving the dream of this generation and strengthening hopes for future prosperity, but its fiscal vulnerabilities pose substantial risks for that future. There is a case for sustained medium-term adjustment to substantially reduce those risks, putting its public debt on a downward trajectory. However, addressing this would be exceptionally difficult given the intertwined challenges Croatia is facing today: (i) to bring down the fiscal deficit and reverse adverse public debt dynamics; and (ii) create fiscal space for the absorption of large EU funds, preparing the ground for recovery and sustainable long-term growth. 15. To achieve debt sustainability, the government will need to turn a 1.8 percent primary deficit in 2013 into a balance in 2016. This will require an increase in revenues and/or a reduction in primary expenditures of 3.7 percentage points of GDP. While the adjustment is urgent, so as not to deepen vulnerabilities and risks of much higher borrowing costs, in a depressed environment, an adequate balance needs to be found between credible consolidation and the potential impact this may have on depressing growth and growth expectations. 16. At the same time, Croatia will need to create fiscal space averaging up to 1.8 percent of GDP a year in 2014-2020 to support EU funds absorption. This can be achieved through the efficient utilization of EU funds, combined with substitution of budget funding with EU xv financing (where allowed) and some switching of budget spending toward high return investments, to ensure both fiscal sustainability and foster income convergence with the rest of the EU. 17. Croatia’s spending and revenue pattern suggests that a sizeable fiscal adjustment of 4-5 percentage points of GDP could be implemented over the medium term. While revenue measures identified could raise additional revenue of 2 percent of GDP with limited adverse impacts on growth and employment, more fiscal space can be created on the spending side. 18. Croatia would need to cut its primary deficit by 1.8 percentage points of GDP per year over the next three years. Several countries in the EU have gone through fiscal consolidation episodes in the recent history. When corrected for the cycle, over the last four years Greece achieved an annual rate of primary deficit reduction of 4.1 percentage points of GDP on average (2009-2013), the highest in the developed world in recent years. Latvia had a similar experience over the same period. While one can argue that both countries experienced a large social cost due to such rapid adjustment, the required adjustment in Croatia would be similar to the consolidation episodes of Sweden, UK or Ireland. While such an adjustment is urgent, it is also achievable so Croatia can strengthen the hopes for future prosperity. 19. The consequences of not addressing twin challenges might be detrimental to Croatia’s short and long-term growth perspectives. By not insisting on more forceful absorption of EU funds, Croatia may face a possible loss of EU funds and more importantly an opportunity to finance its growth through foreign grants thus helping both its growth and fiscal consolidation objectives. By not reducing fiscal vulnerabilities, the country may face further deterioration in investors’ sentiment resulting in higher financing costs and/or lack of access to affordable long-term borrowing. This would deepen the recessionary trends and require much more urgent and deeper fiscal consolidation actions. Summary of Policy Options SHORT-TERM MEASURES MEDIUM-TERM MEASURES I. Maximizing the Efficient Use of EU Funds  Link regional and national priorities with the  Build institutions and administrative Cohesion Policy priorities. capacities related to the identification,  Develop a clear strategic vision for national preparation and implementation of projects priorities and EU funds’ allocation at all levels. that can be financed by EU funds, especially  Establish comprehensive budget management at the local level. systems to avoid different approaches to  Set up a sound financial management system financing sources. at all levels (local, regional, national) and in all entities benefitting from EU funds. II. Revenue-Side Adjustment  Eliminate a large number of tax exemptions  Introduce modern property taxation (savings given to households and businesses (savings up up to 1.5% of GDP). to 1.1% of GDP).  Strengthen and modernize Croatia’s tax administration (CTA) to protect and expand the revenue base. III. Rightsizing the Government Public administration  Finalize provisions regulating criteria for the  Adopt a targeted downsizing program to xvi Summary of Policy Options SHORT-TERM MEASURES MEDIUM-TERM MEASURES creation and management of agencies. contain the overall cost of public  Accelerate the full application of the HRMIS to administration, accompanied by the control staff costs. introduction of outsourcing, electronic service  Reform the wage system to reduce the public and staffing norms in particular sectors such wage bill and budget pressures in the short as education and judiciary (savings up to 1% term, and proceed with structural reforms to of GDP). enhance public sector performance over the  Reinvigorate attempts to enhance strategic long term (savings up to 0.15% of GDP). planning, professionalization and the introduction of performance based management practices in civil service Local and regional self-governments institutions.  Monitor fiscal operations of subnational  Redefine spending responsibilities of local governments to ensure fiscal prudence and governments to avoid duplication and overlap alignment with the Excessive Deficit of functions and to increase accountability of Procedure. LGUs for service delivery (savings up to  Increase LGUs’ reliance on own-source 0.25% of GDP only in social sector). revenues to ease financing constraints and  Proceed with territorial organization to reduce central government transfers (savings address current disproportionate number and up to 0.6% of GDP). resources assigned to LGUs. IV. Improving the Efficiency and Equity of Social Spending Health  Clear the stock of arrears, and strengthen  Consolidate health service networks by public financial management systems so that geographic areas to streamline services for payment delays are eliminated and arrears will acute cases (savings up to 0.5% of GDP). not reoccur (cost up by 1.1% of GDP).  Develop and implement national care  Conduct a review of exempt copayment pathways to enforce clinical guidelines in the categories with a view to reducing them and context of the adjusted networks. adjust the complementary health insurance  Expand public health services to reduce the premium with actuarial standards (savings up prevalence of behavioral risk factors. to 0.3% of GDP).  Reduce the referral rates in the primary health  Identify redundant capacity to merge services care. and facilities (savings up to 0.25% of GDP).  Expand eHealth systems.  Create high-frequency lower-cost specialized centers for ambulatory diagnosis and treatment (savings up to 0.25% of GDP). Pension System  Consider accelerating the retirement age  Consider gradually raising the second pillar increase before 2030 and tightening and contribution rate (transition cost of 1.7% of phasing out the early retirement (savings up to GDP). 0.3% of GDP).  Use means-testing for granting minimum  Rationalize the categories of privileged pensions and award pension points only for pensions and accelerate convergence of periods with contributions paid savings up to privileged pensions to PAYG by equalizing the 0.4% of GDP). maximum privileged pension with the old-age xvii Summary of Policy Options SHORT-TERM MEASURES MEDIUM-TERM MEASURES maximum pension (savings up to 0.4% of GDP in 2038).  Revisit the pension indexation/valorization (savings up to 0.8% of GDP by 2020). Long-term care  Develop a comprehensive LTC plan.  Favor community-based over institutional  Shift LTC services from the health to the social care. sector.  Explore a shift from the government  Decrease care fragmentation and increase producing LTC services to buying them from coordination. the private sector.  Explore the potential of cash benefits and vouchers for funding LTC. Social assistance  Establish one-stop-shop for all social benefits  Extend means-testing to most social  Introduce a single, unified set of criteria to assistance and family programs (savings up to assess eligibility for needs-based social 0.35% of GDP) assistance programs.  Implement “make-work-pay” benefit reforms.  Introduce a parametric redesign of the child tax  Start reducing losses due to error, fraud and allowance (savings up to 0.5% of GDP) corruption through strengthened oversight and inspection (savings of 0.2-0.4% of GDP). V. Rationalizing Subsidies Railways  Define an affordable level of funding for the  Adjust the level of services and the network sector, along with an overall transport size so that it matches resources available. investment program and funding limits.  Strengthen the contractual relationships of the  Set the structure and organization of the infrastructure manager and passenger financial support to railways through PSC and operator. MAIC.  Enforce the restructuring program in holding  Enhance the MMATI’s capacity to supervise companies to meet planned cost cutting the railway sector. targets (savings up to 0.3% of GDP).  Maximize EU Funds absorption for railways’ investments. Agriculture and rural development  Take advantage of the option to transfer Pillar  Strengthen and rationalize public services in 2 funds into the CNDP envelope for 2014-16 agriculture. (savings up to 0.13% of GDP).  Make fiscal discipline, budget transparency,  Make strategic decisions on the allocation of and streamlined budget planning priorities. the sector budget between the major nationally- funded expenditure categories and avoid duplicating nationally and EU-funded interventions.  Terminate “market interventions” on milk, mandarins and apples. xviii 1.1 On July 1, 2013, Croatia became the 28th member of the European Union. This achievement crowned more than Figure 1. Croatia’s Per Capita Gross National Income and At-Risk- a decade of macroeconomic and of-Poverty Rates, 2002-13 institutional reforms by the Croatian authorities and other stakeholders that have yielded important development results. This is even more noteworthy considering the extremely difficult initial conditions following the tragic regional armed conflicts in the 1990s. Croatia’s institutions are now far stronger than a decade ago, reflecting broad and deep institutional adjustments that underpinned the pre-accession process and Croatia has become a high-income country within one decade (Figure 1). Notes: Estimates of relative poverty for and prior to 2009 are derived 1.2 Other development from the household budget survey whereas estimates since 2010 are indicators have also improved based on survey of income and living conditions (SILC). Therefore, two markedly. The Croatia poverty series are not directly comparable. Source: CROSTAT and World Development Indicators. rate, as measured by the international line of moderate poverty at $5 in PPP terms, is estimated at 2.8 percent for 2012. The share of the population “at risk of poverty”, based on a higher national and relative poverty line, also declined substantially prior to the 2008 global financial crisis, although has subsequently increased. Croatia has the largest tourist industry in the region, capitalizing on its coastline and other historic and cultural assets. Infrastructure has improved substantially: Croatia’s roads and airports are now closer to EU standards3, helping move more people and products, and to produce more efficiently. 1.3 Yet, the global financial crisis, with the loss of credit, has exposed Croatia’s macroeconomic vulnerabilities (Figure 2). The stimulus for a significant share of Croatia’s pre- crisis growth has been withdrawn resulting in five consecutive years of recession through 2013. Prior to the crisis, large, relatively cheap capital inflows were circulated into the domestic economy, creating credit, consumption, and real estate booms that have subsequently reversed. Yet such capital inflows are not expected to return in the near term, due to weaknesses in Croatia’s growth outlook and the more competitive, risk adverse post-crisis international environment. Unemployment has risen to 17 percent in 2013, much higher than the Eurozone average (12 percent). Public debt has doubled since 2008 and remains on an upward trajectory, 3 The Croatia’s infrastructure is ranked 42nd out of 144 countries at the World Economic Forum’s Global Competitiveness Indicator. 1 whereby private sector continued deleveraging. Credit agencies took note, reducing Croatia’s sovereign debt to a speculative status in 2013. Figure 2. Croatia’s Macroeconomic Imbalances, 2002-13 8.0 120 Fiscal balance CAB Inflation GDP growth 6.0 Public debt Public and publicly guaranteed debt External debt 100 4.0 % of GDP 2.0 80 0.0 60 -2.0 40 -4.0 -6.0 20 -8.0 0 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013e -10.0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Source: CROSTAT, MoF, CNB. 1.4 Behind these macroeconomic imbalances, Croatia also faces deep structural problems that are holding back a recovery of output, exports, and jobs:  Croatia until recently had extremely rigid labor markets that made it difficult to hire or release workers in the formal sector, thus contributing to high and persistent unemployment.  The business climate for domestic and foreign investors remains cumbersome4. Perhaps because Croatia hosted relatively robust capital inflows before the global crisis, there was little pressure to improve investment climate which is weaker than in many high-income countries.  Croatia delayed fiscal adjustment in response to the global crisis with the hope that the country would “grow out” of the downturn as global economic environment improves despite growing deficit and debt levels. As a result, public debt and fiscal positions have become unsustainable and well above the thresholds set out in the new EU economic governance framework5; and  Public sector efficiency is at the low end and the cost of public services at the high end among the European club of countries to which it acceded. The fiscal footprint remains comparatively large, both in terms of employment and involvement in productive sectors. This has been a drag on the fiscal deficit and public debt levels, but also provides considerable scope for efficiency savings in the public sector that would not undermine performance. 1.5 In sum, Croatia’s pre-crisis growth has stalled, public debt ballooned and fiscal vulnerabilities escalated. Critically, public sector inefficiencies continue to stifle competitiveness and the country’s overall growth potential. What is the way out of this 4 Croatia ranks second at the bottom of European countries in the 2014 Doing Business rankings: www.doingbusiness.org 5 See: http://ec.europa.eu/economy_finance/economic_governance/index_en.htm 2 conundrum? And, given these issues, will Croatia be able to tap into the opportunities and convergence offered by the EU membership? 1.6 This report shows that without addressing macroeconomic weaknesses, through sustained fiscal adjustment and institutional reforms, Croatia will not be able to reignite higher growth and benefit fully from EU membership, and the quest for future prosperity may prove elusive. Without significant consolidation, the fiscal deficit and public debt levels will continue to breach the Maastricht criteria6 thus delaying the entry to Eurozone. Without significant institutional strengthening, public investment and consequently growth will likely remain low, despite the availability of significant EU cohesion and structural funds. Moreover, the cost of government services is particularly excessive, as will be shown in Chapter II – taxpayers are paying too high a cost for services they receive--and this also raises the cost to the private sector and the economy’s competitiveness. The relative cost of government services is among the highest in the EU, including the other member states that joined the EU after 2004. 1.7 Similarly, without accelerating structural reforms, especially in the area of investment climate, and public sector efficiency, Croatia will face further stifled competitiveness and any prospects for recovery of growth and jobs. Croatia’s labor markets were up until now among the most rigid in the EU (Figure 3) and this is a large part of the reason behind high and chronic unemployment7. Paradoxically, for a country that achieved so much and entered the EU, Doing Business indicators rank Croatia only 65th among 189 countries, similar to Albania and Moldova. Broader competitiveness indicators provide a more detailed but broadly the same unfavorable picture (Figure 3). Without additional improvements of the business environment, these weaknesses are also likely to undermine the potential benefits of European financial integration, in terms of lower interest rates and greater, more stable and productive capital inflows. Taken together, these risks are substantial and threaten to unwind much of the economic gains that Croatia achieved in the previous decade. Figure 3. Croatia’s Competitiveness Indicators Labor Market Efficiency Product Market Efficiency Source: World Economic Forum Global Competitiveness Index Source: World Economic Forum Global Competitiveness (2014-15). Index (2014-15). 6 General government debt at 60 percent of GDP and general government deficit at 3 percent of GDP. 7 World Bank (2011), “Employment Protection Legislation and Labor Market Outcomes: Theory, Evidence and Lessons for Croatia” 3 Ease of Doing Business Government Effectiveness Source: Doing Business (2014). Source: Worldwide Governance Indicators (2012). 1.8 Focusing on the fiscal and public sector related deficiencies outlined above, this report systematically analyzes three interrelated issues to assist the Croatian government in informing public policy, strengthening macroeconomic stability, and laying the foundation for a robust recovery:  First, it analyzes Croatia’s major fiscal weaknesses, risks, and alternative fiscal scenarios, and, on that basis, calculates the required fiscal adjustment needed over the medium term (Chapter II).  Second, it analyzes the institutional weaknesses and requirements for the efficient use of EU funds in the coming years (Chapter III). These funds, if used efficiently, will bolster growth and competitiveness. However, this requires fiscal space of 1.8 percentage points of GDP per year and strengthened institutions to fully benefit from these funds. Otherwise, there is a risk that these funds be underutilized.  Third, the report analyzes the structure of Croatia’s public finances and provides a blueprint of the fiscal adjustment of around 5 percentage points of GDP over the medium term (Chapter IV to VII). This is to be achieved through a combination of revenue, tax administration and, especially, expenditure measures, both short term and structural. These measures are designed not only to reduce the fiscal deficit and public debt to sustainable levels but also to improve the efficiency of public sector and create fiscal space for EU funds absorption, improving the cost competitiveness and contributing to the broader growth agenda. 4 2.1 Croatia has achieved major development results since mid-1990. Between 1998 and 2008, per capita national income (World Bank Atlas method) more than doubled––from $5,360 to $13,960––making Croatia a high-income country8. The at-risk-of-poverty rate fell to 16.3 percent in 2006 after reaching 18.2 percent in 2002, before rising after the start of the global economic crisis. Income inequality, measured by the Gini coefficient at 0.31 makes Croatia a country with moderate inequalities. Macroeconomic management improved, with a relatively low and stable fiscal deficit and sustainable debt levels, while inflation declined to low single digits. Important institutional reforms, such as reform of judiciary, regulatory framework, competition policies, supported the transition toward membership of the European Union (EU) that eventually came on July 1, 2013 when it became the 28th member. These achievements created a sense of optimism about Croatia’s future and the promise of further shared prosperity within the world’s richest economic block. 2.2 Yet despite these achievements, the global crisis exposed major macroeconomic and structural vulnerabilities. While the EU accession process helped Croatia’s income, competitiveness and living standards converge toward EU averages9, economic integration and connectivity also meant that the Eurozone recession was quickly transmitted through to Croatia. In late 2008, Croatia’s growth previously based on large capital inflows and local credit and real estate growth came to a halt. The economy went into a recession from which it is yet to recover. The cumulative output loss over the past five years (2009-13) is 12 percent of 2008 GDP. Unemployment rose from 8 percent in 2008 to 17 percent in 2013, with youth unemployment reaching 50 percent. Fiscal deficits increased to an average of 6 percent since 2009, resulting in a steady increase in public debt, which rose from 30 percent of GDP in 2008 to 75.7 percent in 2013. The effective interest rate-growth differential rose substantially since the onset of the crisis. General government interest payments now account for 3.5 percent of GDP, three-fourth of the overall capital budget. 2.3 The country has consequently entered the EU in a position of weakness, with an especially uncertain external environment, and high macroeconomic and fiscal imbalances. After the significant decline of GDP in 2009 there has been no economic recovery with domestic demand remaining depressed (Table 1). Weak external demand and uncertain economic prospects discouraged any substantial investment in production facilities, which was reflected in weak import, but also export dynamics. While the external current account narrowed, this was largely because of the weak economy with falling imports. This, however, helped slow the growth of external debt, which stayed elevated at 105 percent of GDP in 2013. On the structural front, weak export competitiveness and rigid labor markets held back the recovery of exports and jobs. And the overall investment climate remained difficult: Croatia is ranked 65th on the Doing 8 According to the World Bank classification. 9 In 2013, Croatia stood at 60.6 percent of GDP per capita in Purchasing Power Standards of EU27, up from 51 percent in 2001. 5 Business’ overall ease of doing business scale (out of 189 countries, and the second lowest in the EU). Table 1. Croatia: Key Macroeconomic Indicators (Percent of GDP) Actual Estimate Indicators 2008 2009 2010 2011 2012 2013 2014 National Accounts Real GDP growth 2.1% -7.4% -1.7% -0.3% -2.2% -0.9% -0.5% Total Investment 31.4 25.0 21.4 20.6 19.3 18.9 19.0 Gross National Savings 22.9 20.4 20.5 20.0 19.4 19.2 20.3 Foreign Savings 8.8 5.1 1.1 0.9 0.2 -0.9 -1.3 Public Sector Revenues 41.6 41.2 40.8 40.6 41.3 41.8 41.1 Expenditures 44.3 47.2 46.8 48.2 46.9 47.0 46.9 Interest payments 1.7 2.1 2.4 2.9 3.3 3.4 3.4 Fiscal Balance -2.7 -5.9 -6.0 -7.7 -5.6 -5.2 -5.8 Primary Balance -1.1 -3.8 -3.7 -4.8 -2.4 -1.8 -2.4 Balance of Payments Trade Balance -22.1 -16.0 -12.8 -13.7 -13.8 -14.6 -15.6 Current Account Balance -8.8 -5.1 -1.1 -0.9 -0.2 0.9 1.3 FDI 6.5 3.4 1.1 2.3 2.5 1.5 1.6 Debt Gross External Debt 85.1 100.4 103.3 102.6 102.1 104.7 104.4 Public Debt 29.6 36.5 52.8 59.9 64.4 75.7 81.5 Gross Internat. Res. (in months of imports of G&NFS) 4.6 7.0 7.2 7.2 7.2 8.2 7.5 Memo items: GDP (EUR millions) 48,135 45,093 45,022 44,737 43,959 43,591 43,577 GDP (US$ millions) 70,461 62,679 59,644 62,241 56,480 57,859 57,873 Inflation (p.a., %) 6.1 2.4 0.9 2.2 3.4 2.2 0.2 Debt service to export ratio 30.0 55.1 57.6 45.4 42.4 39.9 40.3 Exchange rate HRK:US$ (p.a.) 4.93 5.28 5.50 5.34 5.85 5.71 5.70 Source: CROSTAT, MoF, CNB, World Bank 2.4 The labor market worsened dramatically. After initial resilience due to labor market support measures, employment contracted sharply with a loss of 230,000 jobs. Many companies, especially those oriented to the domestic market, struggle to service debts and there have been notable bankruptcies, particularly in manufacturing, construction and trade. The survey-based unemployment rate rose from an average 8.4 percent in 2008 to above 17 percent at end-2013, the second highest among all EU1010 member states and Croatia. This has been coupled with labor force withdrawals, particularly among youth. At the same time, the share of long-term unemployment in total unemployment has risen to 65 percent. Labor force participation (15 years+) is among the lowest in EU--at 50.9 percent in 2013. 2.5 Downward wage rigidity and high relative costs explain in part the observed job destruction rates. The worsening labor market performance can be decomposed into the contribution of worker flows in and out of the unemployment pool (Figure 4).11 If the unemployment rate had been driven only by fluctuations in the job finding rate, actual 10 EU10 indicates the 2004 and 2007 acceding EU member states, except Malta and Cyprus. 11 The job exit rate is defined as the ratio between the number of newly registered unemployed and the total number of employed workers. The job finding rate is the ratio between the number of workers that were employed from the unemployment registry or employed for other reason, relative to the total number of unemployed. The assumption of constant labor force implies that all newly unemployed workers are treated as coming from the employment pool rather than from outside the labor force. 6 unemployment would have started to increase much sooner, and would have stalled at a lower level than the actual rate starting in 2011. Fluctuations in the job exit rate, or those workers entering the unemployment pool, account for the observed unemployment rate stronger, especially recently. The rigidity of wages (Figure 5) explains partly a high job destruction rate. Figure 4. Effects of Job Finding and Exit Rates Figure 5. Net Real Wages and Value Added Per on Unemployment Dynamics Worker Note: UR – unemployment rate; JEP - job exit rate; JFP - job finding rate Source: Croatian Employment Bureau, World Bank staff calculations 2.6 The country now faces a twin challenge of strengthening macroeconomic stability while fostering recovery of growth and competitiveness. Reducing fiscal vulnerabilities will be key to macroeconomic stability and laying the foundation for stable growth. To promote growth and competitiveness, structural reforms in two areas will be critical: first, lightening of the regulatory burden in the labor market and measures to improve investment climate; and, second, the rationalization of public expenditures and the reduction of the excessive size and cost of the public sector. 2.7 This chapter describes how to strengthen macroeconomic stability as a foundation for sustained recovery and growth. It outlines key fiscal vulnerabilities, policy requirements for debt sustainability and related risks, and makes the case for gradual medium-term consolidation. First, it analyzes the evolution of the fiscal position and public debt, which are currently on an unsustainable trajectory, in part because of the prolonged recession. Second, it highlights the implications and perils of this path for future debt sustainability, country ratings, cost of borrowing, and market access to low-cost, long-term finance and, indeed, long-term growth prospects. Third, it analyzes Croatia’s alternative debt sustainability scenarios under different adjustment and growth scenarios as well as risks related to shocks to growth, policy and contingent liabilities. And fourth, it compares Croatia’s composition of spending and revenues with those of the EU member states. On this basis, the chapter establishes the case for sustained medium-term adjustment of spending and revenues to address exceptionally high intertwined challenges: (i) to bring down fiscal deficit and reverse adverse debt dynamics; and (ii) create fiscal space for absorption of large EU funds (Chapter 3), preparing the ground for recovery and sustainable long-term growth. 7 A. Croatia’s Fiscal Performance 2.8 Croatia has experienced a significant deterioration in fiscal balances and public debt. Prolonged recession resulted in weakening revenues and rising expenditures, in part because of automatic stabilizer spending, but in a large part due to assumption of contingent liabilities. The fiscal deficit rose from 2.7 percent of GDP in 2008 to 7.7 percent in 2011 (Figure 6), out of which 2.4 percent are paid guarantees for public enterprises. Public debt, excluding state guarantees, escalated, from less than 30 percent of GDP in 2008 to 75.7 percent in 2013 (Figure 7). When guarantees were included, total public debt almost reached 81 percent of GDP in 2013. In late 2012 and early 2013, rating agencies (Moody’s, Standard & Poor’s) downgraded Croatia’s sovereign debt below the investment grade. Croatia’s cost of borrowing rose as sovereign spreads increased from 124 (at end-August 2008) to above 300 basis points in 2013. While many countries in similar situations responded with significant efforts to stem the rise of deficit and debt, Croatia delayed fiscal consolidation until late 2013 through the second budget revision. Figure 6. Croatia’s Fiscal Performance, 2008-13 Figure 7. Evolution of Croatia’s Public Debt, (percent of GDP) 2008-13 (percent of GDP) 50 -9 Revenues -8 48 Expenditures -7.7 Deficit -7 46 -5.9 -6.0 -6 -5.6 44 -5.2 -5 42 -4 -3 40 -2.7 -2 38 -1 36 0 2008 2009 2010 2011 2012 2013 prel Source: MoF, EUROSTAT, World Bank staff calculations 2.9 Disentangling the fiscal accounts shows that only a portion of the fiscal deterioration is caused by the crisis, while large part of it is being of a structural nature (Table 2). This means that even if growth rates return to more ‘normal’ long-run averages, fiscal consolidation will still be required. The recession and actions to alleviate it involve fiscal costs through three channels: (i) automatic stabilizers; (ii) other non-discretionary effects going beyond the normal impact of the cycle; and (iii) discretionary fiscal stimulus. Some of these impacts will be short- lived; others will be longer lasting or even permanent. If such effects are disentangled, like the assumption of shipyards’ liabilities12 and repayment of pensioners’ debt which have nothing to do with the recent crisis, the structural deficit remains high, on average close to 4 percent over the last four years. Therefore, out of the average general government deficit of around 6 percent in the 2010-2013 period, two-thirds are due to structural fiscal problems. This is important to recognize, as remedies to be applied are those with longer-term effect than those that would cure the short-term crisis impact. 12 Significant contingent liabilities (around 5 percent of GDP) in the form of issued guarantees from 2005-2008 have been assumed by the state in 2009-2011 under the EC-approved restructuring program of five former state-owned shipyards. 8 Table 2. Croatia: Disentangling Crisis from Structural Impact (in percent of GDP) 2010 2011 2012 2013e Real GDP, % -1.7 -0.3 -2.2 -0.9 Potential GDP, % -0.5 -1.9 -1.6 -0.8 Output gap, % of potential GDP -2.8 -2.1 -2.6 -2.1 General government balance, % GDP -6.0 -7.7 -5.6 -5.2 Primary balance, % GDP -3.7 -4.8 -2.4 -1.8 One-off and other temporary measures -1.1 -2.6 -0.2 -0.2 o/w shipyards’ debt assumption and repayment -0.8 -2.4 0 -0.1 o/w pensioners’ debt repayment -0.3 -0.2 -0.2 0.0 Government gross fixed capital formation 3.3 3.5 3.5 3.3 Cyclically-adjusted balance -5.0 -7.1 -4.5 -4.0 Cyclically-adjusted primary balance -2.6 -4.2 -1.3 -0.6 Structural balance* -3.9 -4.5 -4.3 -3.9 Structural primary balance* -1.6 -1.6 -1.1 -0.5 *Cyclically-adjusted balance excluding one-off and other temporary measures. Source: EC, Eurostat, WB staff estimates 2.10 While Croatia’s fiscal trends since 2009 have been similar to other EU countries, its public debt and interest payments are now higher than in EU1013 countries. Comparison with EU10 countries is also indicative of Croatia’s rising vulnerabilities. Croatia’s public debt rose much faster than that of EU10 from the lower level in 2009 (Figure 8). Reflecting the rising strain public debt is putting on the budget, interest expenditures also increased significantly and much faster than in EU10. In 2013, interest spending was 3.4 percent of GDP, three-fourths of total capital investments (Figure 9). It is indicative that even though Croatia has lower public debt than mature EU economies (EU15), its interest payments are higher. This suggests that Croatia’s public debt generates a comparatively larger pressure on the budget, squeezing other productive expenditures. Figure 8. Croatia, EU10 and EU15: General Figure 9. Croatia, EU10 and EU15: Interest Government Debt, Percent of GDP Payments, Percent of GDP Note: General Government debt, as defined by Maastricht criteria, does not include guaranteed debt. Source: EUROSTAT, MoF, World Bank staff calculations and estimates. 13 EU10 represents EU countries from 2014 and 2017 enlargement without Malta and Cyprus. EU15 includes Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, Netherlands, Portugal, Spain, Sweden, United Kingdom. 9 2.11 These developments are reflected in Croatia’s debt Figure 10. 5-Year CDS Spreads: Croatia in the Eurozone Perspective Croatia Estonia Germany Italy Latvia Slovenia profile and uncomfortably 600 Basis points high cost of sovereign borrowing (Figure 10). 500 Because of its fiscal position 400 and macro vulnerabilities, Croatia still does not borrow 300 very long-term in sovereign bond markets (mostly 5-10 200 years). A large share of Croatia’s debt is denominated in 100 foreign currency (Box 1) 0 making it highly sensitive to 1/3/2011 2/3/2011 3/3/2011 4/3/2011 5/3/2011 6/3/2011 7/3/2011 8/3/2011 9/3/2011 1/3/2012 2/3/2012 3/3/2012 4/3/2012 5/3/2012 6/3/2012 7/3/2012 8/3/2012 9/3/2012 1/3/2013 2/3/2013 3/3/2013 4/3/2013 5/3/2013 6/3/2013 7/3/2013 8/3/2013 9/3/2013 1/3/2014 10/3/2011 11/3/2011 12/3/2011 10/3/2012 11/3/2012 12/3/2012 10/3/2013 11/3/2013 12/3/2013 changes in the exchange rate. Medium-term bonds recently Source: Bloomberg. carried a large premium over the German bund of about 300 basis points. Even this premium, however, exceeds that of some EU countries with significant fiscal challenges, indicating the precariousness of Croatia’s sovereign borrowing conditions. Finally, Croatia’s sovereign rates clearly and significantly exceed its medium term growth rate under any reasonable scenario, implying unsustainable debt dynamics. Box 1. Structure of the Croatian General Government Debt The largest share of public debt is in long-term securities (above 60 percent), followed by loans and T-bills. A significant share of debt is denominated in foreign currency, where Euro-denominated debt accounts for two thirds of the overall debt (Eurobonds, project loans from international financial institutions, and borrowing in the domestic market through syndicated loans denominated in Euro). A share of the debt with fixed interest rate amounts to 81 percent. Table 3. Public Debt Structure and Dynamics, Percent of GDP 2008 2009 2010 2011 2012 2013 General government gross debt (Maastricht definition) 29.6 36.5 52.8 59.9 64.4 75.7 1. breakdown by residents - domestic 19.6 23.6 29.1 34.7 36.3 42.3 - foreign 10.0 12.9 15.3 16.6 19.3 24.3 2. breakdown by original maturity - short-term (T-bills) 4.4 5.9 6.3 6.2 5.8 7.1 - long-term 25.2 30.6 46.4 53.8 58.6 68.9 Change of General government gross debt -3.3 6.9 16.3 7.2 4.5 11.2 Implicit interest rate 5.4 6.9 6.4 5.6 5.4 5.2 Contributions to change of general government gross debt: Primary balance 1.1 3.8 3.7 4.8 2.4 1.8 Snowball effect -0.7 3.6 2.7 2.2 3.6 3.4 Interest payments 1.7 2.1 2.4 2.9 3.3 3.4 Growth of nominal GDP -2.4 1.5 0.3 -0.7 0.4 0.1 Stock-flow adjustment -3.6 -0.5 9.9 0.2 -1.5 6.0 Guaranteed debt 13.5 15.7 9.4 7.5 12.3 5.1 Public debt (including guaranteed debt) 43.1 52.2 62.2 67.4 76.8 80.8 Source: MoF, CNB, CROSTAT, World Bank staff calculations and estimates. The data are preliminary after the ESA2010 harmonization. 10 As for the structure of debt according to the levels of government, the largest portion of debt referred to the central government (94 percent), followed by the debt of extra-budgetary users (5 percent) and the debt of local government units (1 percent). Importantly, after a full assumption of Croatia’s guaranteed debt of SOEs (of 11.7 percent of GDP), the guaranteed debt of the State Development Bank (4.9 percent of GDP) represents another contingent liability. B. The Government’s Medium-Term Fiscal Program 2.12 The fiscal consolidation effort in 2012 was short-lived as spending level in both 2013 and 2014 remained unchanged. The revised 2013 budget targeted even a fiscal expansion with a fiscal deficit of 5.5 percent of GDP, higher than in the previous year, reflecting lower revenues from deteriorated economic situation, and an increase in overall spending (by 0.2 percent of GDP). The proposed expenditure cuts of the wage bill (0.4 percent of GDP) and current transfers (0.1 percent of GDP) fell short of the needed adjustment given the increased health spending to clear arrears in the health sector (0.9 percent of GDP) and payments into the EU budget (0.6 percent of GDP). The 2013 deficit outturn, however, turned better than expected due to the local government surplus. The original 2014 budget left the fiscal deficit target largely unchanged as interest payments and the payment into the EU budget squeezed the fiscal space further by 1.1 percent of GDP. The Excessive Deficit Procedure (EDP), launched in January 2014 by the EU Council, sought the subsequent budget revision to reach the fiscal deficit level of 4.6 percent of GDP in 2014 (Box 2). Box 2. The Excessive Deficit Procedure for Croatia As the EU Member State with excessive deficit, Croatia was put under the EDP. The EDP for Croatia lays out a three-year fiscal consolidation program that will bring fiscal deficit levels down to Maastricht threshold by 2016. A credible and sustainable adjustment path will require Croatia to reach a headline general government target deficit of 4.6 percent of GDP in 2014, 3.5 percent of GDP in 2015 and 2.7 percent of GDP in 2016. This is consistent with an annual improvement in the structural balance of 0.5 percent of GDP in 2014, and 0.9 and 0.7 of GDP in 2015 and 2016, respectively. The Council of the European Union set a deadline of April 30, 2014 for Croatia to take effective action in 2014 and to report in detail on the consolidation strategy to achieve these targets. The EC is in close dialogue with the Croatian authorities and will continue conducting frequent consultations on the consolidation policy mix and its impact. If necessary, the Council will recommend Croatia to make further adjustments in its budget. In a Commission Communication from June 2, 2014, the EC provided an official assessment of effective action taken by the Croatian authorities. Source: http://ec.europa.eu/economy_finance/economic_governance/sgp/corrective_arm/index_en.htm 2.13 The Government’s medium-term fiscal framework targeted the EDP-consolidation path. The 2014 revised budget, triggered by the launch of the EDP, targeted a general government fiscal deficit of 4.5 percent of GDP before the harmonization with the ESA2010 took place in October 2014. The new deficit data suggest a deficit increase to 5.4 percent of GDP in 2014 on the account of broader general government coverage (highways, railway infrastructure, public television, etc.). According to the medium-term fiscal plans, by 2016, primary expenditures would need to be reduced by 2 percentage points of GDP (Table 4). The fiscal consolidation measures that are underpinning the medium-term consolidation process include the following: 11 i. Retaining the relative level of revenues through one-off measures (the withdrawal of state- owned enterprises profits) and a transfer of the second pillar savings on extended work service pensions) and permanent measures (the introduction of a gambling tax, concession fees, excises, telecommunication fees, the increase in the health insurance contribution rate from 13 to 15 percent, and, from 2015, the introduction of capital gain, and interest income taxes and a property tax in 2016); ii. A reduction in the public sector wage bill by 0.7 percentage points of GDP through the abolishment of a public sector loyalty bonus, wage freeze, and staffing rationalization; iii. A further reduction in material expenses by around 0.8 percent of GDP, particularly on current and investment maintenance, through the introduction of centralized procurement and the pharmaceutical policy reform; iv. A reduction in subsidies by 0.2 percent of GDP primarily for agriculture and shipbuilding; v. A reduction in current transfers by 0.7 percentage points of GDP through the pension (privileged and disability), health (the hospital network rightsizing, pharmaceuticals and sick leave control) and social welfare reforms (further consolidation of programs and the extension of means-testing to target support); and vi. Retaining the capital spending at an average 4.6 percentage points of GDP. Table 4. Fiscal Developments and Prospects 2010-2016 (percent of GDP unless otherwise indicated, General Government) 2013 2014 2015 2016 2010 2011 2012 prel plan plan plan Total Revenue 40.8 40.6 41.3 41.8 41.5 41.3 40.9 Direct Taxes 18.2 17.8 17.6 17.6 17.7 17.6 17.4 Indirect Taxes 17.9 17.4 18.3 18.8 18.2 18.1 17.8 Nontax Receipts 4.5 5.0 4.6 5.2 5.3 5.2 5.5 Capital revenues 0.1 0.3 0.8 0.2 0.3 0.3 0.3 Expenditures and Net Lending 46.8 48.2 46.9 47.0 46.9 46.0 44.8 Consumption 7.2 7.4 7.2 7.7 7.6 7.3 6.8 Wage bill 12.2 12.4 12.2 12.0 12.1 11.8 11.3 Interest 2.4 2.9 3.3 3.4 3.4 3.4 3.4 Current Transfers 16.7 16.8 17.2 17.3 17.1 16.9 16.6 Subsidies 2.5 2.4 2.2 2.0 2.1 1.9 1.9 Capital expenditures 5.9 6.3 4.9 4.6 4.5 4.5 4.7 Overall balance -6.0 -7.7 -5.6 -5.2 -5.4 -4.7 -3.9 Primary balance -3.7 -4.8 -2.4 -1.8 -2.0 -1.3 -0.5 External financing 5.9 7.1 4.4 9.7 1.1 0.8 0.5 Domestic financing 0.2 0.6 1.3 -4.5 5.0 4.1 3.5 o/w privatization 0.0 0.1 0.0 0.1 0.7 0.2 0.2 Notes: 2014-2016 data show the April government MTEF corrected for the ESA2010 coverage. Source: MoF, EUROSTAT, World Bank staff estimates. 2.14 A short-term challenge for fiscal policy will be financing large liabilities coming due; the medium-term challenge is to reduce fiscal vulnerability and reverse adverse debt dynamics through sufficient, sustained, and smart adjustment. Gross financing needs14 in 14 Without Croatian Highways and Rijeka-Zagreb Highway debt that were kept off budget. 12 2014 stood at 18 percent of GDP (at 11 percent of GDP without the short-term debt rollover). The easing of the financial crisis in the Eurozone may create space for financing the budget, although additional efforts will be needed to strengthen government solvency. To reverse adverse debt dynamics, Croatia will need to implement more ambitious, sustained fiscal adjustment over the medium term. 2.15 Risks of delayed adjustment as well as implementation risks are significant. Unless Croatia implements a sustained and sufficient medium-term fiscal adjustment, fiscal and macroeconomic risks could be escalated.  First, the rise of the fiscal deficit and public debt would likely result in a further downgrade of Croatia’s sovereign ratings (Figure 11), raising its relative cost of external borrowing, and squeezing other productive expenditures in the budget. Croatia’s rating is already notably worse than that of several EU countries with significant fiscal problems (e.g., Italy, Spain, and Slovenia) even if its public debt is lower. Figure 11. Public Debt and Sovereign Ratings, 2013 Source: Standard & Poor’s Ratings Services, EUROSTAT, World Bank staff estimates.  Second, high sovereign costs adversely affect private sector borrowing costs, stifling investments and adding a drag to the overall economic recovery.  Third, unchecked increase in public debt resulted in breaching of the 60 percent debt ceiling in Croatia’s budget law (Article 47 in the Budget Act) as well as the Maastricht debt ceiling, creating a debt “overhang” that constitutes a drag on growth. Additionally, there are around 5 percent of GDP in contingent liabilities that pose additional risk to rising public debt overhang. Research15 has shown that levels of debt that Croatia is already facing are consistent with particularly weak growth performance and rising fiscal and debt risks, especially after the global crisis (Figure 12 and Figure 13). 15 Baldacci, Gupta et al., 2010 13 Figure 12. Public Debt and Economic Growth, Figure 13. Public Debt and Economic Growth, 2000-07 2008-12 Source: World Development Indicators. C. Debt Sustainability Analysis and Required Fiscal Adjustment 2.16 Given Croatia’s fiscal and debt position and associated risks, what are the implications of various policy scenarios for country’s debt sustainability? And what are the possible risks and shocks to these scenarios that could derail the needed adjustments? To answer these questions, a “no-policy change” scenario, which effectively simulates the current gradual consolidation approach, is analyzed. The consequences of a “likely scenario” based on the EDP- consistent medium-term fiscal strategy is then examined and compared with the “no-policy change scenario.” Finally, the report assesses the impact of various shocks that might complicate fiscal adjustment as well as the implications of a large contingent liability shock arising, for example, from a sudden call on the state guarantees that would then need to be converted into public debt, putting major pressure on the current budget. 2.17 A scenario of no-policy change assumes delayed expenditure adjustment, resulting in the steady increase in public debt over the medium term (Table 5). This scenario illustrates the peril of gradual adjustment process. It assumes a cut in primary expenditures of 2 percentage points of GDP by 2016. With slow movement towards a primary balance and the interest rate significantly exceeding growth, public debt rises to 86 percent of GDP in 2016. Both scenarios do not take into account a potential contingent liability shock. Table 5. Medium-Term Fiscal Scenarios, Percent of GDP No policy change scenario Likely scenario 2012 2013 2014p 2015p 2016p 2017p 2014p 2015p 2016p 2017p Total revenues 41.3 41.8 41.5 41.3 40.9 40.1 41.5 42.1 42.1 42.1 Total expenditures 46.9 47.0 46.9 46.0 44.8 43.3 46.9 45.9 44.8 43.8 Interest 3.3 3.4 3.4 3.4 3.4 3.3 3.4 3.3 3.2 3.0 Fiscal balance -5.6 -5.2 -5.4 -4.7 -3.9 -3.2 -5.4 -3.8 -2.7 -1.7 Primary balance -2.4 -1.8 -2.0 -1.3 -0.5 0.1 -2.0 -0.5 0.5 1.3 General government debt 64.4 75.7 81.1 84.8 85.8 85.5 81.1 83.7 83.5 82.5 Note: Likely scenario builds on the EDP-consistent fiscal framework, while the no-policy change scenario largely follows the original 2014-2016 government fiscal framework. 14 2.18 The likely scenario, based on the EDP, assumes 2 percentage points of GDP in primary expenditure cuts by 2016 and 0.6 percentage points of GDP in additional revenues that should lead to a correction of the excessive deficit by 2016. The required fiscal effort in terms of the compliance with the fiscal deficit (below the 3 percent of GDP) and debt criterion (sufficiently diminishing debt-to-GDP ratio), still puts the debt-to-GDP ratio above the Maastricht threshold. Under this scenario, public debt is forecast to peak in 2015 at 83.7 percent of GDP to decline thereafter. 2.19 To achieve debt Figure 14. Cumulative Adjustment of Primary sustainability, the government will Expenditures/Revenues, 2014-17 need to turn the 1.8 percent of GDP primary deficit in 2013 into a balance by 2016. This adjustment is equal to a gradual reduction in primary expenditures and/or rise in revenues of 3.7 percentage points of GDP, with expenditures being reduced down to 43.8 percent of GDP by 2017 (Figure 14). While the adjustment is urgent not to deepen vulnerabilities and risks of much higher borrowing costs, in a depressed environment, an adequate Source: MoF, Eurostat, CROSTAT, staff calculation balance needs to be found between credible consolidation and a burden that an over-excessive consolidation may have on depressing growth and growth expectations. The burden of adjustment should largely fall on unproductive public expenditures while protecting public investments. 2.20 Stress tests point to still high downside risks (Figure 15). Debt outcomes remain very sensitive to growth. A negative growth shock of an average -1.4 percent in 2014-17 leads to an increase in public debt to 101 percent in 2017. A combined shock scenario of a delayed primary balance beyond 2016, a rise in the implicit interest rate to 6.2 percent, and lower GDP growth (at an average of 0.2 percent in 2014-17) raises public debt to 92 percent of GDP in 2017. Finally, both a strong depreciation shock, given the debt currency structure, and a contingent liability shock would push public debt towards 101 and 92 percent of GDP, respectively, at the end of the projection period. 15 Figure 15. Public Debt Sustainability Scenarios and Shock, Public Debt as % of GDP a) Likely vs. growth shock b) Likely vs. combined shock c) Baseline vs. real exchange rate shocks d) Baseline vs. contingent liabilities shocks 1/ Growth shock refers to permanent one-half standard deviation shock in projection years. 2/ Combined shock applies permanent 1/4 standard deviation shock to real interest rate, growth rate, and excluding interest payment current account balance. 3/ One-time (2014) real depreciation of 30 percent is defined as nominal depreciation (measured by percentage fall in EUR value of local currency) minus domestic inflation (based on GDP deflator). 4/ Contingent liability shock tests a sudden call of state guarantees in the amount of 5 percent of GDP that are converted into public debt in 2014. Source: MoF, EUROSTAT, CROSTAT and WB staff estimates. 2.21 Different factors will drive the public debt change in the medium term. As expected, the most critical factor is an achievement of sufficient primary surpluses (Figure 16). As the primary deficit moves closer to a balance in 2015, the public debt starts falling in 2016, ensuring a sustainable, downward debt trajectory. Relatedly, the highly unfavorable interest-rate-growth differential in the early projection years falls rapidly at the end of the period with lower debt, lower interest payments, and more robust growth. 16 Figure 16. Composition of Change in Public Debt, Percentage Points of GDP Source: MoF, Eurostat, CROSTAT, staff calculation 2.22 Even with debt on a Figure 17. Evolution of Public Debt and Gross External sustainable path, the debt Financing Requirements (in % of GDP) and interest rate levels will require constant vigilance to deepen and sustain the gains of adjustment. In 2016, debt levels will remain at close to thrice the level before the crisis (Figure 17). Interest spending on public debt at somewhat above 4 percent will still exceed the growth of nominal output. 2.23 Additionally, Croatia will need to apply the expenditure switching to Notes: Implicit interest rate on public debt is derived as nominal interest support recovery. The fiscal expenditure divided by previous period debt stock. adjustment will need to be Source: MoF, Eurostat, CROSTAT, staff calculation accompanied by spending reallocation to those projects which could be substituted by the EU grant funds as well as privatization, both of which would lead to a faster decline in public debt. D. Composition of Public Revenues and Spending 2.24 The good news is that current spending patterns offer significant scope for rationalization. At 47 percent of GDP in 2013, Croatia’s spending level was 5.2 percentage points of GDP higher than in its EU10 peers (Table 6). At the same time, total revenues were about 3 percentage points higher than in EU10. 2.25 Public spending is particularly excessive in areas such as subsidies, public wages, and consumption: 17  Overall subsidies, mostly to railways, shipyards, and agriculture, at 2.1 percent of GDP are double the EU15 and EU10 spending.  The public sector wage bill at 12 percent of GDP is 2.6 percentage points of GDP higher than in EU10 or 1.6 percentage point of GDP higher than in EU15.  At 7.7 percent of GDP that Croatia allocates to current consumption, there is at least 1-2 percentage point space for further rationalization of these costs. Average spending on operations and maintenance in Croatia is higher than the levels observed in the comparator countries which may also reflect inefficient consumption of inputs (e.g., energy consumption, space renting) or higher unit prices resulting from insufficiently competitive public procurement. Table 6. General Government Expenditures by Economic Classification, Percent of GDP EU15 EU10 Croatia 2009-11 2012 2013 2009-11 2012 2013 2009-11 2012 2013 Total Revenues 44.5 44.9 45.6 38.0 38.6 38.7 40.5 41.3 41.8 Direct taxes 12.8 12.9 13.2 6.9 6.6 6.6 6.6 6.1 6.3 Indirect taxes 12.8 13.1 13.2 12.9 13.3 13.3 17.5 18.3 18.8 Social contributions 14.1 13.6 13.7 11.9 12.3 12.4 11.8 11.5 11.3 Sales 2.6 3.0 3.1 2.4 2.7 2.8 2.6 2.7 3.0 Other current revenue 2.2 2.2 2.5 3.9 3.7 3.6 2.4 2.7 2.4 Total Expenditures 50.6 49.3 48.9 43.8 41.9 41.8 47.4 46.9 47.0 Current Expenditures 46.3 44.5 44.7 37.9 36.6 36.7 41.5 41.9 42.1 Consumption 6.9 6.4 6.4 6.1 5.8 5.8 7.2 7.2 7.7 Wage bill 11.1 10.5 10.4 9.7 9.3 9.4 12.3 12.2 12.0 Interest 2.8 3.0 2.8 2.2 2.3 2.2 2.5 3.3 3.4 Subsidies 1.2 1.2 1.2 0.9 1.1 1.1 2.5 2.2 2.0 Social benefits 21.8 21.1 21.4 16.9 15.9 16.0 16.0 16.3 15.9 Current transfers 2.4 2.3 2.4 2.1 2.2 2.2 0.8 0.9 1.4 Capital Expenditures 4.2 4.8 4.3 5.9 5.4 5.1 6.2 4.9 4.6 Capital transfers 1.7 1.9 1.5 1.1 1.0 1.0 2.8 1.8 1.6 Investments 2.6 2.9 2.8 4.7 4.4 4.1 3.4 3.1 3.0 Deficit -5.8 -4.3 -3.2 -5.7 -3.3 -3.2 -6.5 -5.6 -5.2 Gross GG Debt 82.1 89.2 91.7 46.0 49.7 50.9 49.7 64.4 75.7 Note: Data reflect ESA2010 harmonization and are still preliminary. Source: MoF, Eurostat, CROSTAT, staff calculation 2.26 Capital expenditures are on par with EU15, but below the EU10 comparators. After years of excessive growth of capital spending on highways before the crisis, Croatia is today, at 4.6 percent of GDP in 2013, below the level of public resources allocated for capital expenditures in EU10. 2.27 Looking at the functional breakdown of spending, five distinct differences to the comparator groups prevail:  Spending on general administration services is 0.3-1.6 percentage points of GDP higher in Croatia than in EU15 and EU10, respectively. 18  Spending on public order and safety is 0.7-0.8 percentage points of GDP higher than in comparator groups, also reflecting higher spending on judiciary.  Spending on economic affairs is 0.4-1.2 percentage points of GDP higher than in comparator groups, suggesting high subsidization of economic sectors.  Health and social protection spending taken together is 3 percentage points of GDP above the EU10, while still behind the EU15 group (by 5 percentage points). Health spending looked at separately is 1.7-4.4 percentage points of GDP above the comparators, partly reflecting a different basic insurance coverage.  In parallel, spending on environment protection and housing is half of the comparators’ spending, indicating an area of future spending pressures to align environmental standards with those of the EU. Table 7. General Government Expenditures by Function (COFOG) in 2012, Percent of GDP EU15 EU10 Croatia Total 51.7 42.6 47.0 General public services 6.8 5.5 7.1 Defense 1.5 1.0 1.5 Public order and safety 1.9 1.8 2.6 Economic affairs 4.1 4.9 5.3 Environment protection 0.8 0.8 0.4 Housing and community amenities 0.8 0.8 0.4 Health 7.5 4.8 9.2 Recreation, culture and religion 1.1 1.4 1.2 Education 5.2 4.6 5 Social protection 20.2 14.3 13.1 Transport 1.5 2.5 1.2 R&D Economic affairs 0.3 0.1 .. Note: Difference in total expenditure as per Table 6 due to net lending and rounding. Source: Eurostat, CROSTAT, staff calculation 2.28 Social spending (health, education, and social protection) remains buoyant. Education spending mostly stayed on par Figure 18. Social Sector Spending in 2012, percent of with the comparators in 2012. However, GDP public health spending at 9.2 percent of 25 Health Education Social protection GDP in 2012 was an outlier. This also 20 reflects the growing health sector arrears as Croatia adjusted its payment practices 15 with the EU Directive of financial payments. Social protection, although 10 much lower than in EU15, is on par with EU10. Croatia’s social welfare system 5 also provides substantial benefits to war veterans', a type of spending not found in 0 other comparator countries. EU15 EU10 Croatia Source: EUROSTAT, MoF, World Bank staff calculations. 2.29 Apart from addressing the 19 spending level, Croatia’s effectiveness of key public expenditures leaves plenty of room for improvement (Figure 19). Comparative indicators of effectiveness or government performance16 composed by the EU for education and health show that Croatia lags significantly behind most EU countries. The education system performance indicator ranks Croatia sixth from below, next to Greece, Portugal and Slovakia.17 Croatian health system exhibits slightly better performance– ranks tenth in Europe from below, slightly better than Slovenian, German and Portuguese.18 Overall government effectiveness, a general governance indicator providing a summary assessment of the quality of public administration services, its regulatory system, and its impartiality, sets Croatia at the seventh place from below in Europe, better than in Romania, Bulgaria, Greece, Italy, Poland and Hungary. This contradicts the amount of public resources allocated to the sector that is above EU10 and EU15 countries. Figure 19. Performance of Government Services Education performance score Health system performance score Note: Education system performance is measured by: (i) pupils of 15y of age with PISA literacy score below 1; (ii) secondary schooling attainment (% of adults 25-64y of age who completed at least upper secondary level education), (iii) tertiary schooling attainment (% of adults 30-34y of age who completed any form of tertiary level education) and (iv) life-long learning attainment measured by % of adults (25y-64y) stating they attended any form of learning program in 4 weeks preceding the labour force survey. PISA based indicator for primary education is weighted by 40% and the rest of three indicators are weighted by 20% each. Health system performance is measured by equally weighted healthy life expectancy of males and healthy life expectancy of females at age 65. Two indicators are equally weighted. Source: Eurostat, OECD, World Bank and Sonje (background paper) 16 Government performance is a broader concept than quality of governance (which includes rule of law, absence of corruption, strength of democratic and electoral institutions, and administrative efficiency) and includes measures of government output, utilizing proxy measurements of public administration, education and health services. 17 Result is broadly similar to Sopek’s (2001) who did not measure performance per se, but rather, efficiency, by comparing real public expenditures on education with outcomes of PISA tests. 18 This result is broadly in line with efficiency frontier measurements done by Badjun et al. (2011). 20 Government effectiveness Source: Eurostat, http://ec.europa.eu/europe2020/pdf/themes/34_quality_of_public_administration_final.pdf 21 With joining the EU, Croatia has the opportunity to benefit from large committed resources from the EU, averaging 3.7 percent of gross domestic product (GDP) per year during 2014-2020. However, Croatia will also need to transfer part of its revenues to fund the EU budget. While the overall effect of EU-related transfers on Croatia’s economic growth should be positive, fiscal space needs to be created to support their absorption, averaging up to 1.8 percent of GDP a year in 2014-2020. This can be achieved through improved implementation of EU funded- projects, combined with some switching of budget spending toward high return investments, to ensure fiscal sustainability and foster income convergence with the rest of the EU. 3.1 European Union (EU) membership will allow Croatia to benefit from sizeable EU transfers to its national budget. The Union has committed a considerable amount of assistance to Croatia after its accession in July 2013. The EU Structural and Cohesion funds available to Croatia are close to €1.6 billion a year (or €1.8 billion with the direct payments and market measures for agriculture). The aim of these transfers is to facilitate Croatia’s income convergence with the rest of the EU. However, EU membership is also associated with member states’ contributing to the EU budget, which will bring new fiscal challenges as resources exit from its budget accounts. Croatia’s contribution to the EU budget will be about €530 million annually. While EU Structural and Cohesion Funds are expected to provide the bulk of the investment financing, the short and medium-term financing gap arising from the cash flow management will put extra fiscal pressure at a point when further fiscal consolidation should be a priority. 3.2 Therefore, a key question for Croatia is how to create the fiscal space to best absorb EU funds and to ensure their efficient use. Put differently, the issue is how the budget and EU funds management can be rationalized and made more efficient to maximize the positive fiscal and economic effects of EU membership. These questions are approached by assessing: (i) Croatia’s pre-accession absorption record as a signal for the capacity of the country to utilize EU resources; (ii) the estimated effect of the EU transfers on the revenue and expenditure sides of the budget; and (iii) the overall effect of the EU funds on Croatia’s fiscal stance, their demand impact and the challenges to the efficient utilization of EU funds. A. Pre-Accession Funds 3.3 Since 2001, Croatia has been allotted €1.13 billion in European Union pre-accession assistance (Box 3). The EU provided pre-accession assistance to Croatia through several instruments, including the Community Assistance for Reconstruction, Development and Stabilization (CARDS) program, from which the country received €76 million through 2004. In 2005 and 2006, Croatia received around €221 million from pre-accession programs PHARE, ISPA and SAPARD. In addition, Croatia had a financial allocation of €836 million under the Instrument for Pre-Accession (IPA) program within the period 2007-2013. 22 Box 3. EU Pre-Accession Assistance Programs (2001-2013) The EU provides financial assistance for candidate countries and potential future members in order to support their efforts to enhance the political, economic and institutional reforms needed to fulfill the EU’s membership criteria. This assistance is also aimed at helping to strengthen administrative capacity in preparation for managing higher levels of EU funding after accession. In the early 2000s, the PHARE program was the Commission’s main pre - accession instrument and was complemented by the Instrument for Structural Policies and Pre-Accession (ISPA) which financed infrastructure projects in the transport and environment sectors (a precursor to the EU Cohesion Fund). The Special Accession Program for Rural Development (SAPARD) was similarly designed as a precursor to the European Agricultural Fund for Rural Development (EAFRD). Other pre-accession instruments included the Community Assistance for Reconstruction, Development and Stabilization (CARDS) program. In 2007, these four pre-accession assistance programs (CARDS, PHARE, ISPA and SAPARD) were replaced by the Instrument for Pre-Accession (IPA). The five IPA components which have been available to Croatia are as follows:  Transition Assistance and Institution Building (IPA I)  Cross- Border Co- operation (IPA II)  Regional Development (IPA III)  Human Resources Development (IPA IV)  Rural Development (IPA V). 3.4 There were mixed results in terms of the absorption of pre-accession funds. While the first generation programs (CARDS, PHARE, ISPA and SAPARD) and elements of the IPA demonstrated relatively high absorption rates, the rates declined for the remaining second generation (IPA) assistance (Table 8). Contracting under the first generation assistance was 89 percent. However, absorption for IPA funds as well as the European Structural and Investment (ESI) Funds 2007-2013 that followed the same procedures has been slower, particularly for the transport (IPA IIIa), human resources operational programs (IPA IV) and the IPARD program (IPA V). However, the contracting period is still ongoing and the rates may further improve from on average 61 percent contracting rate. Table 8. Implementation of EU assistance to Croatia, million EUR, March 2014 Paid, % Contracted, % Paid, % of Programs Allocated of of allocated contracted allocated CARDS 2003 29.4 97.7% 94.9% 92.7% CARDS 2004 46.6 94.6% 94.1% 89.1% Phare 2005 73.1 87.4% 91.5% 80.0% Phare 2006 64.1 85.2% 94.1% 80.2% ISPA 59.0 97.2% 87.8% 85.4% SAPARD 25.0 61.7% 75.4% 46.5% TOTAL 1ST GENERATION 297.2 88.9% 91.1% 81.0% ASSISTANCE IPA I 2007-2013 249.0 69.0% 81.8% 56.5% IPA II 2007-13 105.0 77.0% 53.0% 40.9% IPA IIIb 2007-13 131.3 64.1% 34.9% 22.4% IPA IIIc 2007-13 82.0 79.1% 54.5% 43.1% IPA IV 2007-13 94.4 59.5% 71.0% 42.2% IPA V 2007-13 144.3 56.4% 43.0% 24.3% TOTAL IPA 806.0 75.5% 59.8% 45.1% 23 OP Transport 2007-2013* 237.0 33.9% 68.4% 23.2% OP Environment 2007-2013 281.1 34.3% 44.1% 15.1% OP Regional Competitiveness 2007-2013 187.8 36.9% 66.5% 24.6% OP Human Resources Development 2007- 152.4 48.9% 75.7% 37.0% 2013 Total ESI 2007-2013 858.3 37.3% 62.4% 23.3% TOTAL 1,961.5 60.8% 67.4% 41.0% *IPA IIIa and OP Transport were merged into the IP Transport 2007-2013 into an overall amount of EUR237 million. Source: MRDEUF, www.strukturnifondovi.hr 3.5 Despite delays in implementation of multiannual IPA programs, which led to some reductions in allocated funds19, the overall impact of pre-accession funds has been significant. Under these multiannual programs, contracting deadlines are not set, but funds allocated in a particular year that are not spent (i.e. not paid to final beneficiaries and not certified to the European Commission) by the end of the third year (n+3) are automatically decommited (i.e. withdrawn from the allocation)20. However, the European Commission has taken measures, including with advanced Figure 20. EU Funds Inflow to Croatia, Percent of payments, to reduce the loss of funds for GDP and Per Capita Croatia and the problem of late start has not been critical because projects were few and amounts relatively small. Figure 20 shows that the realized flow of funds from the EU to Croatia—which excludes national co-financing—grew close to 0.8 percent of GDP per year by 2013. As discussed later, the economic impact of such flows will depend on how well the funds are spent, rather than simply focusing on absorption levels that may overlook such negative factors as low Source: MoF, World Bank staff calculations. social or economic rates of return or the displacement of private investments. 3.6 The experience with EU pre-accession assistance offers valuable lessons for the post- accession of Croatia. First, to avoid implementation delays, the country needs to be ready to submit proposals still in 2014 to minimize the decommitment of EU funds. Second, in the implementation of the Cohesion Policy, it is necessary to ensure: a multiannual strategic approach; partnership (the objectives should be in the interest of all parties: the central state, local and regional self-government units, the civil society, entrepreneurs, social partners); the principle of additionality (Box 4); systematic control; and monitoring and evaluation. Third, a 19 European Court of Auditors (2011) “Has EU assistance improved Croatia's capacity to manage post - accession funding?” Special Report No 14. 20 The first decommitment of funds due to the n+3 rule took place at the end of 2011, specifically in IPA III and V, whereas funds from component IV were timely spent by the end of 2011. Thereby, there remained unspent funds at the end of 2011 and allocations were reduced in the amount of 9.3 million euros for IPA IIIa - Operational Program Transport, in the amount of 5.2 million euros for IPA IIIb – Operational Program Environmental Protection, in the amount of 0.1 million euros for IPA IIIc Operational Program Regional Competitiveness, and in the amount of 25.32 million euros for IPARD. 24 high level of fund utilization does not automatically mean better growth and development. Funds need to be targeted and used strategically in accordance with sound financial management rules. Box 4. The Principle of Additionality The principle of additionality was adopted for the Community structural funds in 1989. This prescribes that for EU structural funds to make real impact on eligible regions, EU funding may not replace national structural spending of the Member States. Thus, Member States are required to maintain at least the same level of national structural expenditure in real terms from one programming period to another. The principle of additionality implies that EU funds should augment the available national resources in order to enhance the welfare of the Europeans. Thus, EU money might contribute to the implementation of certain large scale projects that would not have been possible to execute with national funds only. However, although there is already certain expertise in managing EU funds due to the pre-accession experience, effectively using EU funds remains a key challenge for beneficiaries. As various EC reports note, the focus on financial rather than strategic issues tends to lead to funds being spent where they are most easily absorbed instead of where they might be most effective. Source: http://ec.europa.eu/regional_policy/archive/policy/how/additionality_en.htm B. EU Financial Support for Croatia After Accession 3.7 Accession to the EU has a number of effects on the general government budget of a new member state. A new member is obliged to transfer part of its revenue to the EU budget according to predefined criteria, but it can also draw funds from the EU budget. Inflows of EU Funds 3.8 Upon the accession of Croatia to the EU, the Croatian budget is expected to encounter numerous structural changes that will impact the budget cash flows. In addition to the pre-accession multiannual assistance, Croatia will receive transfers from the EU budget that can be divided into: (i) transfers not related to projects (i.e. direct inflow to the national budget); (ii) transfers that depend on the absorption capacity of an individual Member State;21 and (iii) other transfers (Box 5). Box 5. EU Transfers to Member States There are three types of transfers from the EU budget. First, transfers (not related to projects) that comprise direct aid, agricultural market-related expenditure and transfers on the basis of internal policies. Direct support schemes for farmers not related to projects come from the European Agricultural Guarantee Fund (EAGF). Transfers by means of internal policies include a variety of EU programs aimed at increasing the co-operation between MS in the conduct of common policies: citizenship, freedom, security, justice, education, environmental protection, research, energy efficiency. The second group of transfers (project-related) includes transfers from the Structural Funds, the Cohesion Fund and European Agricultural Fund for Rural Development (EAFRD). The EU Structural and Investment (ESI) Funds have three main objectives: (i) promotion of the development and structural adjustment of regions whose development is lagging behind; (ii) economic and social assistance to areas with structural difficulties; and (iii) assistance to adaptation and modernization of policies and systems for 21 The absorption capacity of an individual Member State is primarily measured by administrative absorption capacity as a key determinant of the successfulness of implementation of EU structural policies, as well as financial absorption capacity as a measure of capability to co-finance projects at both state and local levels. 25 education, training and employment. Structural Funds cover exclusively regions whose GDP per capita is below 75 percent of the EU average and projects are co-financed by the EU up to maximum 75 percent of the eligible cost amount. Countries eligible for cohesion funding are those EU MS’, such as Croatia, with a gross national income lower than 90 percent of the EU average. The Cohesion Fund finances action on the trans-European transport networks, priority projects of special interest, as well as some other transport and environmental activities and are complemented by the Connecting Europe Facility (CEF). Projects are co-financed by the EU up to maximum 85 percent of the amount of the eligible costs. EAFRD relates to rural development projects financing. Accounting for over a third of the EU budget for 2014-2020 (amounting to €325.1 billion), the objective of the EU Cohesion Policy 2014-2020 (that includes the Structural and Cohesion Funds and was approved in November 2013) is to deliver the Europe 2020 objectives of smart, sustainable and inclusive growth. The third group of transfers from the EU budget includes other pre-accession assistance, special arrangements and budgetary compensation. These transfers have a relatively small significance in the total EU budget, but they can comprise sizeable allocations to new Member States in the first years of EU membership. These compensations are introduced in order to prevent new Member States becoming net contributors to the EU budget in their first years of membership. 3.9 For the period 2014-2020, Croatia is expected to disburse from the EU budget around 2.4 percent of GDP on average per year (Table 9). The EU has made commitment appropriations of about 3.7 percent of Croatia’s GDP annually for 2014-2020 through the Partnership Agreement)22, which comprises precommitted funds of €12.5 billion (including direct payments and market measures for agriculture). A sizeable part of this budget will be allocated to the Cohesion Fund (€2.6 billion) and for regional development (€4.3 billion). Table 9. Payment Appropriations 2014-2020, percent of GDP 2014 2015 2016 2017 2018 2019 2020 Pre-accession assistance 0.5 0.5 Budget compensation 0.1 0.0 0.0 0.0 0.0 0.0 0.0 EU funds 1.1 1.6 2.0 2.3 2.6 2.9 3.1 Agriculture 0.4 0.5 0.5 0.5 0.6 0.7 0.7 Structural Funds 0.2 0.6 0.9 1.2 1.4 1.5 1.6 Cohesion Funds 0.1 0.3 0.4 0.4 0.5 0.5 0.6 Internal Policies 0.4 0.2 0.2 0.2 0.2 0.2 0.2 TOTAL receipts 1.7 2.1 2.0 2.3 2.7 2.9 3.1 Source: World Bank staff estimates 3.10 Croatia received around 0.8 percent of GDP in 2013 from the MFF 2007-2013 (Table 10). The major share of these funds, in the amount of EUR 167.4 million (0.38 percent of GDP), relates to activities that promote Sustainable Growth (including Competitiveness for Growth and Employment and Cohesion for Growth and Employment). Project-related funds amount to EUR149 million (0.34 percent of GDP) and include Structural Funds (EUR90 million or 0.2 percent of GDP) and Cohesion Fund resources (EUR60 million or 0.13 percent of GDP). 22 The Partnership Agreement (PA) for 2014-2020 (http://europa.eu/rapid/press-release_IP-14-1223_en.htm) was adopted by the EC on October 30, 2014--a year later than the EU Parliament adopted the Multiannual Financial Framework with an envelope of €14.4 billion. The PA allows the new generation of EU spending programs to be implemented as soon as Operational Programs are adopted. 26 Table 10. Payment Appropriations from the EU for Croatia in 2013 Payments Payments 2013 2013 (€ mill.) (% GDP) 1. Sustainable growth (1a +1b) 167.4 0.38 1a Competitiveness for growth and employment* 17.6 0.04 1b Cohesion for growth and employment 149.8 0.34 of which Structural Funds 89.9 0.20 European Social Fund (ESF) - Convergence 18.0 0.04 European Regional Development Fund (ERDF) – 68.5 0.15 Convergence European Regional Development Fund (ERDF) - 3.36 0.01 European territorial cooperation of which Cohesion Fund 59.9 0.13 2. Preservation and management of natural resources 2.6 0.01 European Fisheries Fund 2.2 0.00 Other CFP support, Life + * 0.4 0.00 3. Citizenship, freedom, security and justice 42.2 0.09 3a Freedom, security and justice * 1.1 0.00 Schengen facility 40.0 0.09 3b Citizenship * 1.1 0.00 Transition Facility 4. EU as a global player 86.8 0.20 Instrument for Pre-accession Assistance for Rural 27.7 0.06 Development (IPARD) Transition and institution-building assistance to candidate 59.1 0.13 countries 6. Compensations (Cash-flow facility) 75.0 0.17 TOTAL 374.0 0.84 Note: Excludes direct payments in the amount of EUR 96 million. These funds will be paid in 2014 for liabilities towards farmers per hectare of eligible area relevant in 2013. Source: COM (2013) Draft amending budget no. 1 to the general budget 2013, Brussels, 18.3.2013 Flow of Funds to the EU 3.11 Expenditures on own resources. Own resources of the EU budget are automatically transferred from the Member State (MS) into the EU budget. Own resources of the EU budget are: Traditional Own Resources (TOR), revenue from VAT and revenue based on Gross National Income (GNI). A special part of the EU’s own revenue consists of various corrections, of which the largest is the UK rebate (Box 6). Box 6. EU-Related Outflow of National Funds Own resources comprise the following categories:  TOR consist mainly of customs, agricultural duties and sugar levies, whereby 75 percent of all collected revenue on this basis is automatically transferred to the EU budget, while the remaining 25 percent is kept to defray the costs of collection.  VAT-based revenues are calculated as a predefined percentage of the VAT base, which has to be harmonized with EU rules. In order to prevent disproportional payments into EU budget, the VAT base is capped by 50 percent of a GNI. Since 1 January 2007, a uniform rate of 0.3 percent has been applied on the VAT base, or 50 percent of the GNI, with some exceptions.  GNI-based revenues are the largest contributors to the EU budget–calculated as the difference between 27 total EU budgetary expenditure and revenue collected from other sources, this patches any gaps in the EU budget on the basis of the relative size of a MS’s GNI.  An additional cost to the budget is also for corrections. Notably, after joining the EU the UK became the largest contributor to the EU budget, mostly thanks to the low level of transfers from the Common Agricultural Policy (CAP) due a relatively small agricultural sector. Since 1985 the UK has been refunded a part of its payment to the EU budget, in the amount of 66 percent of its net position. The loss of these revenues is made up by all MS’s, with the provision that the largest contributors bear only one quarter of the share. 3.12 In the case of Croatia in the period 2014-2020, expenditures on own resources are estimated at 1.1 percent of GDP. Out of around EUR530 million, the bulk of expenditures relates to the GNI-based estimated payments in the amount of 0.7 percent of GDP (Table 11). Table 11. Estimated Payments from Croatia to the EU Budget in 2013-2020, Percent of GDP 2014 2015 2016 2017 2018 2019 2020 GNI –based 0.7 0.7 0.7 0.7 0.7 0.7 0.7 VAT- based 0.2 0.1 0.1 0.1 0.1 0.1 0.1 TOR – based 0.2 0.1 0.1 0.1 0.1 0.1 0.1 UK correction 0.1 0.1 0.1 0.1 0.1 0.1 0.1 TOTAL expenditures 1.1 1.1 1.1 1.1 1.1 1.1 1.0 Source: World Bank staff estimates 3.13 Co-financing and pre-financing. In addition to direct payments to the EU budget, there will be expenditures related to co-financing and pre-financing of projects at the state and local level. The minimum prescribed co-financing rate of a member state corresponds to 25 percent of total funds from the Structural Funds and Rural Development Funds and 15 percent of total funds from the Cohesion Fund. In addition, the EU has historically co-financed projects in the range of 50 to 85 percent, while the financing of the remaining part up to the total project value has remained the responsibility of the member state. The estimated amount of co-financing averages to 0.5 percent of GDP per year. The pre-financing requirements range from 0.1 to 0.3 percent of GDP between 2014 and 2020 and include pre-financing of Structural and Cohesion Funds for the entire period. Table 12. Estimated National Co- and Pre-Financing, 2014-2020, Percent of GDP 2014 2015 2016 2017 2018 2019 2020 Co-financing 0.1 0.5 0.4 0.5 0.5 0.6 0.6 Agriculture 0.0 0.2 0.0 0.0 0.0 0.0 0.0 Structural Funds 0.1 0.2 0.3 0.4 0.4 0.5 0.5 Cohesion Funds 0.0 0.0 0.1 0.1 0.1 0.1 0.1 Pre-financing 0.1 0.3 0.2 0.2 0.2 0.3 0.3 TOTAL 0.3 0.8 0.6 0.7 0.8 0.9 1.0 Source: World Bank staff estimates Impact of EU-Related Fund Flows on the Fiscal Accounts 3.14 The simple net impact of EU-related fund flows on the budget accounts is evident in widened cash deficit. Many countries with high cash deficits had to switch spending to accommodate the additional spending. Assuming total compliance with the ‘additionality principle’ for all EU funds—i.e. that no EU funds replace national spending—the flow of EU- related funds is likely to generate a budget deficit, on a cash basis. 28 3.15 In Croatia’s case, the cash deficit could increase by an average of around 1. 8 percent of GDP annually in the Figure 21. Net Impact of EU Funds on the Fiscal Deficit, first six years of accession (Figure ESA95, Percent of GDP 21). However, this may be an upper Poland Hungary Lithuania Slovakia bound as the additionality principle is 0 difficult to verify and consequently -0.2 enforce. The results are consistent -0.4 with previous research (Figure 21)23, -0.6 which suggests that government will -0.8 have to accommodate EU funds with -1 2004 either a higher deficit or expenditure -1.2 2005 -1.4 switching. 2006 -1.6 -1.8 Note: Substitution as reported by the authorities for HU; maximum substitution according to EU rules for other. Source: Rosenberg, Sierhej (2007) Table 13. Net Fiscal Effect, Cash Basis, Percent of GDP 2014 2015 2016 2017 2018 2019 2020 EU related revenues 1.7 2.2 2.0 2.3 2.6 2.9 3.1 Budget compensation 0.1 0.0 0.0 0.0 0.0 0.0 0.0 Refunds on EU projects 1.6 2.2 2.0 2.3 2.6 2.9 3.1 Agriculture 0.4 0.5 0.5 0.5 0.6 0.7 0.7 Structural funds 0.2 0.6 0.9 1.2 1.4 1.5 1.6 Cohesion funds 0.1 0.3 0.4 0.4 0.5 0.5 0.6 Internal Policies 0.4 0.2 0.2 0.2 0.2 0.2 0.2 Pre-accession assistance 0.5 0.5 0.0 0.0 0.0 0.0 0.0 EU related expenditures 3.0 4.0 3.7 4.1 4.5 4.8 5.2 Contribution to EU 1.1 1.1 1.1 1.1 1.1 1.1 1.0 National co- and pre-financing 0.2 0.6 0.6 0.7 0.8 0.9 1.0 Agriculture 0.0 0.2 0.0 0.0 0.0 0.0 0.0 Structural Funds 0.1 0.2 0.3 0.4 0.4 0.5 0.5 Cohesion funds 0.1 0.2 0.2 0.3 0.3 0.4 0.4 Pre-accession assistance 0.2 0.1 EU projects (refunded by EU) 1.6 2.2 2.0 2.3 2.6 2.9 3.1 Net fiscal impact -1.3 -1.8 -1.7 -1.8 -1.9 -1.9 -2.0 Net Inflow of EU Funds 0.3 0.4 0.4 0.5 0.7 0.9 1.1 Notes: Agricultural funds are fully transferred to non-government beneficiaries. The calculations assume no substitution or switching of expenditure between national and EU related spending, i.e. the strict additionality principle. The maximum possible substitution rate is at 50 percent according to the EU rules. Source: World Bank staff estimates, based on data from PA 2014-2020 23 Kopits and Szekely (2002) calculate the impact on the national budget of the Czech Republic, Estonia, Hungary, Poland and Slovakia, which varies between 3 and 5 percent of GDP. Backe (2002), Antczak (2002), Vincelette and Vassileva (2006) also obtain similar immediate effects of EU accession on national budgets. 29 3.16 Given Croatia’s large fiscal deficit, it is important to manage EU-related funds within the overall fiscal consolidation process through expenditure switching and substitution policies. EU-funds can be used to mitigate some of the negative effects on growth from the consolidation required to ensure sustainability of public finances now pursued under the deficit-reduction trajectory agreed under the EC Excessive Deficit Procedure. This can be done by maximizing the growth enhancing objectives of EU funds through the agreed strategic plan, and switching away from lower priority spending (as discussed elsewhere) or substituting for domestic spending where possible. It is therefore important to quantify and monitor the effect of the flow of EU funds on the size and direction of the fiscal impulse on the economy, as both revenues and expenditures contain substantial transactions with a non-domestic entity (i.e. the European Union). The change in the headline fiscal deficit from one year to the other is no longer a good approximation of the demand-side impact of fiscal policy. Box 7. Framework for Evaluating Fiscal Impact of EU Transfers Using ESA95 accounting, the net effect of EU funds on fiscal (1 ) EU related receipts stance can be estimated by adding unconditional budget Budget compensation transfers received from the EU and substituted spending, and Refunds on EU projects/policies subtracting contributions to the EU and budgetary co-financing (2 ) EU related expenditures of projects. Measuring the impact of EU funds on the fiscal Contribution to EU accounts is fraught with a number of methodological Spending on EU projects/policies difficulties. The main include: National co-financing (3) Substituted spending  Status of the ultimate user of funds. Under ESA95 rules, Net fiscal impact (1)-(2)+(3) only funds that end up with “government units as final beneficiaries” are recorded as an expenditure and offsetting revenue item in the fiscal accounts. In practice, funds for agricultural support virtually all go to the private sector, while those for internal policies and cohesion go to the public sector. The status of the ultimate user is the most uncertain for structural funds. Information obtained from earlier wave of accessions countries suggest that 45 percent of regional development funds, 70 percent of social funds and 100 percent of community initiative funds end up in the public sector.  Rate of co-financing: Under EU rules, countries need to co-finance every project from national resources, at rates ranging from 15 percent for cohesion funds to 25-50 percent for structural funds.  Extent of substitution of spending: Member states are allowed to use EU-funds to substitute national spending for some purposes (e.g., agriculture), but not for others (e.g., structural). In practice it is virtually impossible to establish how much a government would have spent on a certain expenditure item if it had not had access to EU funds. Estimates of the impact of EU funds, however, crucially hinge on getting the amount of additionality right. Source: Rosenberg, Sierhej (2007) 30 Growth Potential of EU Funds 3.17 The main goal of the EU funds is to foster economic development and income convergence by supporting public and private physical as well as human capital. The use of structural funds, particularly during periods when output is below potential, can be considered as a fiscal stimulus. These funds, along with domestic co-financing, constitute a significant share of public investment in the newer EU member states, and can therefore have a major impact on growth in countries like Croatia where public investment has been significantly reduced during the last five years. During 2009-11, under the previous funding round, Cohesion funding (including national co-financing) represented over seventy percent of public investment in Hungary, Slovakia, Bulgaria, Lithuania and Estonia, and over 50 percent on Poland and the Czech Republic, and as such it was a major contributor to growth and employment24. 3.18 If effectively used, the EU-funds should contribute to economic growth both in the short and long term. In the short term, Figure 22. First-Round Demand Effect of EU Funds, they would create higher domestic Percent of GDP demand. In the long term, if invested in 2.5 the improvement of the physical and 2.0 human capital, they should contribute to α=0.9 economic growth from the supply side. 1.5 α=0.65 Importantly, there are notable second- round effects that EU transfers aim to 1.0 achieve on trade integration, connectivity, investment and growth. 0.5 3.19 The impact on growth is 0.0 2013 2014 2015 2016 2017 2018 expected to be positive over the -0.5 medium term, but can be lagged. Figure 22 presents a simple approach Note: The demand impact can be defined as : D=α(T+NC)-C-A that suggests that in the first two years Where demand (D) depends on transfers from the EU (T), national co-financing (NC), contributions paid (C), and of EU membership the demand impact advances received (A). α is a measure of substitution between is estimated to be rather modest (less the EU transfers and domestic spending (when α=1 if there is than ½ percent GDP) and could be much no substitution). lower (and even negative) depending on Source: World Bank Staff, based on Rosenberg and Sierhej the level of “crowding out” of (2007) domestically funded projects by EU funded ones.25 Similarly in Bulgaria “during the first three years of Bulgaria’s EU membership, the overall demand effect was negative mainly due to the very low absorption and the contribution to the EU budget. However, as absorption rates improved, the demand impact was projected at around 3 percent of GDP per year”26. 24 World Bank, December 2013, EU11 Regular Economic Report, No.28. 25 The assumptions made are as follows: expenditures financed with structural and rural development funds do not replace domestic spending while other EU transfers such as Cohesion, Common Agriculture Policy, Schengen replace about half of previously domestically-funded spending (this implies values for α at around 0.9, meaning that almost all EU funds are assumed to be additional to domestically-funded spending (α=1). While Rosenberg and Sierhej (2007) found that α typically ranges between 0.55-0.65, this could be higher now that the additionality principle applies to all the MFF funds. 26 Paliova and Lybek (2014) use a range of quantitative methods that suggest the growth impact of EU funds in Bulgaria is in the range of 1.5-3 percent over the medium-term. 31 3.20 The results depend crucially on the effects of expenditure switching and maximizing the quality of EU-related spending. The experience of the new EU members points to significant potential for fiscal consolidation whilst utilizing EU funds. As discussed elsewhere in this report, Croatia can create fiscal space for EU-funds from reducing public expenditure on direct or indirect subsidies, rationalizing social transfer programs, enhancing the efficiency of health care, and streamlining public administration. The net impact on growth, in the short and long run, will depend on the relative growth elasticity of different types of spending—which enhances the need to carefully program EU funds to promote growth. C. Effective Utilization of EU Funds 3.21 Changes have been introduced for the 2014-2020 program period to boost the impact of the EU funds on growth. This builds on the lessons from the previous program period. The funds must be aligned with the Europe 2020 Strategy to better focus resources on a limited number of key priorities with high growth potential. An increased focus on results requires governments to show whether public investments are achieving their desired effect through enhanced transparency, accountability and performance measurement. 3.22 Countries are also required to fulfill a range of specific pre-conditions to access the funds. For example, all EU Member States have to produce a “smart specialization” strategy that identifies particular strengths and opportunities, or take measures to improve public procurement systems and demonstrate compliance with environmental laws. Strategies to fight youth employment or to promote gender equality and non-discrimination are also preconditions. 3.23 Cohesion Policy is also set to be linked to the EU’s economic governan ce framework through macroeconomic conditionality. The areas supported by the Cohesion policy will have to be consistent with National Reform Programs that address reforms identified through the EC’s annual country-specific recommendations in the process known as the European Semester27. The EC can now ask Member States to modify programs to support key structural reforms, which should be consistent with Cohesion Policy investments. In addition, the EC has the authority to suspend funds if macroeconomic problems, such as excessive deficits, are not satisfactorily addressed. 3.24 To successfully restructure public spending to accommodate EU-related expenditures Croatia can learn from its pre-accession experience and from other EU member states. Lessons include streamlining budget procedures for EU funds, enhancing the institutional capacity of agencies absorbing EU transfers, and improving the preparation of project activities. Other key institutional reforms include the following:  Linking regional and national priorities with the Cohesion Policy priorities. The coordination of strategic priorities at the local, regional and national level as well as harmonization with the objectives of the Cohesion Policy present key challenges. The process of EU funds planning is based on: Operational programs (OP) and the application of the accrual accounting. Only well designed OPs are financed. The planning process is undertaken by line ministries and coordinated by the Ministry of Regional Development and EU Funds. Croatia has prepared the first seven-year strategic document – Partnership 27 For a description see: http://ec.europa.eu/europe2020/making-it-happen/ 32 Agreement for the period 2014-2020, which establishes the priorities for utilizing structural funds, but clear, implementable OPs need to be developed.  Developing a clear strategic vision at all levels. The development of the Partnership Agreement aims to align strategies at national, regional and local levels to ensure that funds have cumulative and long-term impact on growth rather than being allocated to one-off projects. However, Croatia’s strategic planning function is still being developed at the national, regional and local level to set clear priorities, objectives and performance indicators. Without such a strategy-based and results-oriented programs, the EU will likely finance projects in European corridors passing through Croatia, and environmental infrastructure. Thereby, Croatia needs to intensify its work on developing coherent sectoral strategies (from transport, environmental protection, energy via science, R&D, primary education, pre-school education, health care, social welfare to the economy and agriculture). It is important that such sectoral strategies are focused on a specific, limited number of objectives that foster coordination the central/local/regional levels.  Building institutions and administrative capacities related to the EU-funds’ identification, preparation and implementation, especially at the local level. Experience gained from the implementation of pre-accession assistance programs shows that the preparation of large projects, such as railways, waste management centers, water treatment equipment and water supply networks, can take several years. For example, the capacity for designing technical documentation satisfying the EU standards has proven weak and resulted in lengthy delays. In addition, the process for solving property-related legal issues and obtaining the required permits is currently lengthy, with few environmental impact studies. This lack of technical capacity and efficient processes underscore the importance of connecting with design firms, universities and other institutions in the preparation process of projects.  Setting up a sound financial management system at all levels (local, regional, national) and in all EU funds beneficiaries. The annual allocation from pre-accession programs was relatively small (about 150 million euro) in relation to the allocations expected from European Funds (about 1.8 billion euro) upon accession, but its absorption proved demanding for the administration and final beneficiaries. As a consequence, the reform of Croatia’s financial and budget regulations to align with the practices of the EU is a priority. A significant step in strengthening financial management and control systems was made with the Fiscal Responsibility Act and the Directive on drawing up and delivery of the Statement on Fiscal Responsibility. It is important to continue to develop this system, and to extend its scope to include the less-controlled parts currently outside the budget, specifically companies owned by the state and/or local and regional self- government units, which are potentially significant beneficiaries of EU funds.  Establishing comprehensive budget management systems to avoid different approaches to financing sources. Strategic documents being prepared for EU financing should be in line with national strategic documents. The existing budget regulations separate EU funds management from the national budget management. Moreover, the legislation is not based on definitions and budget classifications set by the Organic Budget Act and subordinate acts. This complicates Croatia’s ability to switch resources to the highest priorities through the budget, while as highlighted above, the fiscal 33 consolidation process will require significant expenditure switching within a comprehensive budget. 3.25 Programming, institutional design, and implementation are the elements of the management of EU funds that matter most for its effective use. Taking stock of the challenges other countries confronted in Structural Funds (SF) programing, building up institutions, and putting the funds to work (Box 8) offers the following lessons: (i) ensuring the strategic focus of the programming documents and their selectivity; (ii) establishing a legal framework that addresses every aspect of EU funds management (especially procurement); (iii) avoiding frequent changes to laws; (iv) investing in human resources (using the ESI technical assistance allocation); (v) coordinating actively with the EU; (vi) using existing systems, institutions, and procedures but clarify their roles; and (viii) putting in place a comprehensive management information system (MIS) to ensure effective monitoring. Box 8. Challenges to Management of Structural Funds There are three important aspects of EU funds management that influence the size and quality of their utilization: programming, institutional design, and implementation: Source: McMaster and Bachtler 2005: 2. Programming is about both strategic planning—allocating scarce and limited resources among multiple objectives, stakeholders, and instruments—and designing systems for monitoring and evaluation and for implementation. The latter requires considerable cooperation and coordination between the EU and both national and subnational governments. Programming also means making difficult political choices from a limited set of funding priorities that not only address local issues but are also in tune with European objectives. Another programming challenge that Member States (MS) have to tackle is the balance between positive discrimination for less developed regions and supporting general country development—promoting “growth poles”. In other words, should more resources be channeled to lagging regions to give them a chance to catch up with the rest of the country, or should more financing go to better developed regions that have more potential which then can spill over to lagging regions? Coordination of national and European policies is a central aspect of programming. Not only must national goals reflect European targets, but financial decisions also have to show the complementarity of national resources and EU funds, since MS must incorporate into their strategies and action plans the fact that EU funds do not cover all fields of state activity. Financial coordination is to some extent automatically achieved by a co- financing mechanism because MSs must contribute their own resources to structural and cohesion funds, which are related to EU priorities. This limits MS autonomy in shaping their financing priorities but it may also lead to more concentrated and thus more effective action. MSs mainly determine the institutional structure for implementing EU cohesion policy nationally . Yet the 34 EU defines both broad principles and specific implementation and management rules. The national institutional framework for using EU funds might be characterized by two dimensions: centralized/decentralized (with respect to the center-regions relationship) and integrated/not integrated (whether EU funds are managed by an existing administrative, financial, and economic system or by a separate system created specifically to administer them). With decentralization, fund management is closer to the ultimate beneficiaries, which is generally desirable but more difficult to manage; centralized systems arguably have efficiency advantages. In 2004 –06 most new MSs decided on a centralized system. Poland applied a mixed-integrated system; some smaller MSs, like Lithuania, chose a centralized system with a mix of integrated/not integrated approaches. Since 2007 there has been a trend toward regionalization, but some MSs chose rationalization —simplification of national management structures — by limiting a number of managing authorities (Finland, Greece, Hungary, Sweden) and instruments, e.g. cutting programs (Italy). Governance Structure Options Chosen, 2007–13 Governance Structure Country Centralized Finland, Denmark, Baltic States, Slovenia Mixed Poland (MRD + regions), France, Spain Decentralized (regionalized) Germany, Netherlands, Austria, Belgium, Portugal Source: Ferry et al. 2007. The vast spectrum of national and subnational institutions managing EU funds demands a coordination mechanism that ensures that funds are distributed efficiently. This has proved difficult because there is a significant amount of institutional inertia that is hard to mitigate even where special task forces or inter- institutional agencies or committees have been created. Insufficient formalization of relationships, ministry resistance to being steered, opaque lines of responsibility, in addition to lack of coordination between regional and national government, have been identified as major important impediments to efficient coordination. The shape of the implementation system is unique to each MS; there is no single model although the EC prescribes some elements in detail. When a system is hampered by bottlenecks and excessive concentration on risks, achievement of the expected cohesion result is jeopardized. The main concerns about implementation system pertain to such aspects as ensuring both a high quality and a large quantity of applications; making project selection more efficient; building implementation capacity; and monitoring the entire process. Campaigns focused on raising awareness about available funds and incentives for providing support services via specialist consultancies were used to improve project generation and selection. Finally, new MSs have often had problems with monitoring and evaluation, which can be relatively new tools in the national policy tool box. To address that, some MSs, such as the Czech Republic, have established special evaluation units to carry out comprehensive reviews. Human Resources. New MSs tend to have a shortage of experienced staff and knowledge. This holds true for both management authorities and beneficiaries. High staff turnover also undermines efficient funds administration. Inefficient systems tend to disproportionately burden weak and inexperienced applicants, like small municipalities, which often lack resources to muddle through a complex system. Thus adverse selection may mean that those most in need of EU assistance do not obtain it because they are incapable of going through the application process (Maniokas 2006, 5). 3.26 Many practical recommendations for more effective use of EU funds can be derived from the experience of EU members, both old and new, for all stages of EU funds implementations. Project generation could be enhanced by:  The national authorities issuing early and clear guidelines for SF programs. The information must be abundant and precise (without, however, discouraging beneficiaries or making the process too complicated).  Generating fewer but larger projects by introducing funding restrictions or targeting projects with spillover potential (e.g., a minimum funding threshold, as in Spain and Portugal). 35  Geographical targeting: growth poles (Slovakia, Romania); second-order towns (Latvia); projects for lagging regions (Denmark, Spain, Germany).  Differentiated allocations: more resources for the least developed areas (Austria, Latvia, Germany).  Broad information campaigns with a large number of information sources (all member states).  Financing project promoters and program advisers (Poland, Italy, Greece) or via specialist consulting companies (Latvia, Romania).  Funding project preparation from national resources (Sweden, France, Hungary, Austria). 3.27 Critical issues for quick and effective project submission include:  Introducing a single, simple application form for certain projects (England, France).  Allowing electronic submission of project applications (Sweden, Belgium, Portugal).  Offering applicants more trust, reinforced by an effective but light monitoring mechanism (fewer and less burdensome information obligations – less red tape).  Empowering applicants, especially those with the greatest need, by offering them tailored training or the services of specialists.  Avoiding more rigorous and complicated application regulations than those required by the EU (e.g., Poland 2004–06). 3.28 Project appraisal can greatly benefit from:  The managing authority (MA) prescreening the needs of potential beneficiaries— they can submit letters of interest based on which the MA can adjust the programs envisaged (Wales, Scotland).  Adjusting the timeline of calls for applications: no deadlines (Greece, Italy); specific timetables (Sweden, Latvia); coordinated calls for several funds, such as the European Regional Development Fund and European Social Fund (Slovakia, Belgium).  Set appraisal procedures based on continuous open call (Germany, Italy); competition-based assessment (Greece, Italy); immediate appraisal of large-scale projects (Greece); a faster track for smaller projects (Denmark, Slovakia); or distinguishing between core and supporting interventions (Italy).  Outsourcing large and technical projects (Italy, Germany).  Dividing the appraisal process into stages of initial appraisal (technical) and detailed appraisal (engagement of experts, specific working groups and committees, local representatives), with the second stage perhaps followed by applicants providing recommendations (Scotland, Denmark, Austria, Sweden). 3.29 Project selection can be improved by:  Choosing sound selection criteria, e.g., adding to universal eligibility some quality criteria, like envisaged impact (Portugal, Spain); strategic fit (Latvia, Belgium, the Slaskie Region of Poland); linkage with horizontal criteria, such as equality, integration, and environmental sustainability (Germany, Scotland, Sweden); or efficiency criteria, e.g., quick launching of projects (Greece, the Slaskie Region of Poland. 36  Applying methods for appraising criteria, such as an automatic method (Italy, Germany); a scoring system (Scotland, Portugal, the Slaskie Region of Poland); weighting of different criteria (Greece, Belgium); or a qualitative approach (Finland, France, Austria). 3.30 While effective EU funds utilization will be of a paramount importance for supporting growth, fiscal space will need to be created to pre- and co-finance the EU- funded projects. As discussed in the next Chapter, there is a limited scope for creating fiscal space through raising revenues. A combination of strengthening tax compliance, reducing tax expenditures and introducing modern property taxation could provide space to support the EU funds absorption. 37 With a revenue burden already at close to 42 percent of GDP, which is relatively high compared to EU10 countries, there is limited scope for increasing fiscal space from additional revenues. Nonetheless, there is scope to introduce a modern property tax, further rationalize quasi-fiscal fees, broaden the tax base by eliminating exemptions and shift the burden from the high levels of social contributions to promote growth and employment. In addition, business and government could benefit from further simplification of the tax system and strengthening of the Croatian Tax Administration (CTA) to promote compliance. These measures could raise additional revenue of 2 percent of GDP more efficiently and fairly while reducing adverse impact on growth and employment. A. Revenue Reforms for Fiscal Consolidation 4.1 Revenue measures have been an important component of many countries fiscal consolidation efforts, which seek to balance measures for additional revenue with support for growth, efficiency and fairness. While the global financial crisis has affected the structure of economies in various ways—for example, the reduction in domestic demand has led to large declines in income and consumption taxes in many Eastern European Countries28—according to the IMF a “broad consensus emerged on a set of measures…preference was to be given to minimizing distortions (through, for instance, broadening the tax base by eliminating inappropriate exemptions or tax expenditures before increasing the rate), targeting negative externalities, and strengthening tax compliance”29. This consensus is detailed in Table 14, and highlights some key features, discussed below, that include the potential for switching from taxes on employment toward consumption, given concerns about high levels of unemployment in many countries, and the opportunity that many countries have for introducing or increasing taxes of immovable property as being more growth friendly and efficient. Table 14. Conventional Wisdom: Advice for the Revenue Side of Consolidation Recommendation Rationale Exploit consumption taxes more fully, expanding the base of the Most rate differentiation under the VAT is rationalized by value-added tax (VAT) before raising standard rates (using the distributional concerns that could be better achieved by direct social transfer system to protect the most vulnerable as needed), transfers; excises better handle environmental and other and reviewing excise levels. concerns requiring differentially high tax rates. Look for opportunities to broaden the base of the personal income Exemptions and deductions remain significant in many tax—a first step being to quantify all tax expenditures—and, while countries, and their cost should be transparent; raising recognizing that increased inequality might call for increased effective rates can have strongly adverse effects on progressivity, avoid very high marginal effective tax rates. incentives, in terms of both real and avoidance activities. Resist increasing social contributions and consider combining a Unless increased contributions are perceived as carrying cut in the employers’ contribution with an increase in consumption matching increased benefit entitlement, they can have strong taxation—a fiscal devaluation. incentive and employment effects. With a fixed exchange rate, a fiscal devaluation can boost net exports— temporarily—by reducing the foreign currency price of 28 World Bank, EU11 Regular Economic Report, Promoting Shared Prosperity during a Weak Recovery in Central and Eastern Europe, December 2013 29 IMF Fiscal Monitor, Taxing Time, October 2013. 38 exports and increasing the domestic relative consumer price of imports. For the corporate income tax, quantify and review tax Intense international tax competition is likely to continue, and expenditures, resisting further inappropriate base erosion and addressing it will require strong international cooperation; tax pressure to cut statutory rates; reduce the tax bias toward debt distortions can jeopardize financial stability by encouraging finance. excess leverage. Increase property taxes, especially recurrent charges on residential Property taxes appear to be relatively growth-friendly and properties; scale back transaction taxes. can serve equity and accountability objectives; transaction taxes can impede efficient trades. Implement effective carbon pricing either by carbon taxation or Pricing measures are essential to encourage efficient by full auctioning under cap-and-trade schemes; eliminate fossil mitigation and so are a particularly efficient source of fuel subsidies and review environmental taxes more generally. revenue; fuel subsidies are very poorly targeted to distributional aims. In the financial sector, adopt tax measures to discourage volatile These measures would ensure a "fair and substantial financing as well as adopt financing improved resolution contribution" of financial institutions to the fiscal costs of mechanisms; counteract the VAT exemption for financial services their potential distress and failure; as a tax on the sum of by adopting a financial activities tax (FAT). wages and profits of financial institutions, a FAT would provide a fix, albeit an imperfect one, for a major distortion in the VAT. Strengthen tax compliance by identifying and acting on Improving tax compliance would promote fairness and compliance gaps, aggressive tax planning, and offshore tax abuse. reduce distortions. Source: IMF (2013), Fiscal Monitor 4.2 Payroll taxes and social contributions can suppress employment and labor-intensive growth. In OECD countries, the burden of labor taxes can account for up to 40 percent of labor costs, particularly for lower-wage individuals. Labor taxes are also disproportionally high in European countries. For instance, in middle-income Poland, Romania, and the Slovak Republic, social security contributions made up 45–49 percent of gross wage; in Chile they were 13 percent and in the Republic of Korea 16 percent. High levels of labor taxation combined with tax incentives for capital investment help explain the substitution of technology, especially for unskilled labor, that can exacerbate high unemployment rates. Shifting the tax Figure 23. Average Property Taxes in Selected OECD economies, 2000–11, Percent of GDP burden from labor to indirect taxes is 2.5 often known as a ‘fiscal devaluation’ as it Recurrent taxes on immovable property Recurrent taxes on net wealth shifts relative prices in favor of labor 2.0 Estate, inheritance, and gift taxes even if it is revenue neutral. Taxes on financial and capital transactions 4.3 Many countries also have scope 1.5 to introduce or increase property taxes. The burden of property taxes is 1.0 often seen as fairer as it falls mainly on middle- and upper-income families that 0.5 own property, and is less distortionary for businesses and consumer decisions 0.0 EST CZE SVK AUT SVN TUR HUN DEU POL GRC DNK NLD IRL ESP than other taxes. Yet there is a wide variation in revenue raised, with Source: OECD, Development Revenue Statistics. From IMF developing and transition countries (2013). raising around 0.6 percent of GDP30 while in OECD countries it accounted on average for 1.8 percent of GDP in 200831 (Figure 23). The property tax is relatively straightforward to 30 USAID (2009): “Property Tax Reform in Developing and Transition Countries” 39 administer, given the fixed nature of the assets being taxed, although it requires a comprehensive cadaster, well defined ownership rights and a well-defined, transparent process of property valuations32. The revenue, and sometimes the administration, are also often decentralized to local levels of government. Currently Croatia is preparing the introduction of a property tax from 2016, 33 which might also have the benefit of moving many properties that are currently used for holiday lets from the informal to the formal sector. 4.4 The region’s response to falling revenues has varied significantly, although most have increased VAT and excise duty rates during the crisis. As domestic demand has fallen throughout the region, this has impacted the composition of revenue as income and consumption taxes have fallen faster than overall GDP, which was largely supported by export growth in 2012 and 2013. Increasing VAT rates was seen as one of the least distortionary and growth reducing reforms, with rates rising in many countries, including Poland, Bulgaria, Hungary, Slovenia while excise duties were raised in many countries. Countries like the Czech R. and Ireland have also sought to rebalance the tax structure from labor (social security contributions) towards VAT. 4.5 Similarly, Croatia’s recent reforms were spurred by a decline in VAT and excise tax collections, largely due to falling income taxes34. The recent recession has been characterized by declining domestic demand and increasing unemployment, which has had a particularly large impact on consumption and income-related taxes that have fallen by more than overall GDP in many countries across Europe. Croatia responded by increasing the basic VAT rate from 23 to 25 percent, and replacing the zero rate with a 5-percent rate, which was only partly offset by a reduction of compulsory health insurance contributions from 15 to 13 percent (subsequently reversed in 2014). The government also provided some personal income tax reliefs, paid for by a new 12 percent tax on dividends. Table 15. What Reform Measures Have Countries in the Region Introduced in 2010-2013? Note: An upward (downward) arrow indicates a revenue-increasing (decreasing) change. Source: European Commission, OECD, IMF and World Bank staff. Adapted from IMF (2013). 31 OECD (2010): “Revenue Statistics 1965-2009”. 32 Norregaard, J. (2013), Taxing Immovable Property, Revenue Potential and Implementation Challenges, IMF WP/13/129 33 IMF (2012a): “Croatia: Options for Modernizing the Property Tax”. 34 IMF (2012b): “Croatia-Concluding Statements of the IMF Staff Visit”. 40 B. Croatia’s Tax System 4.6 Croatia has the third highest revenue burden, at close to 42 percent of GDP, among the EU10 countries (Figure 24). The large overall tax burden is not due to social security contributions that collect 11.3 percent of GDP—some 2.5 percentage points below the EU15 and one percent below the EU10 level ( Figure 25). The same hold for direct taxes (Figure 26) which are eroded by large tax exemptions (tax expenditures) in the case of households (disability, tax allowance on dependent family members and children, underdeveloped regions) as well as businesses (underdeveloped regions, reinvested earnings, R&D, investment promotion tax reliefs). In fact, the effective rate for personal income tax is only around 9 percent, despite the top marginal rate being 40 percent.35 Figure 24. Revenues in 2013, Percent of GDP Figure 25. Social Contributions in 2013, Percent of GDP Note: HR stands for Croatia; the country abbreviations follow the international standard. Source: Eurostat 4.7 Croatia has the highest indirect tax to GDP ratio among the EU countries (Figure 36 26). Croatia’s VAT rate at 25 percent is also among the highest in EU. With that tax burden and already weak competitiveness, there is not much fiscal space for raising the tax rates further. In recent decades Croatia’s penetration of EU markets has been among the lowest in emerging Europe. This suggests that there is minimal space for tax rate increases—which in any case would further erode competitiveness. In fact, both business and government can benefit from tax systems that are simple to administer and that encourage self-compliance. From 2012, the total corporate profit tax rate as percent of profits declined substantially due to reinvested earnings tax relief, as well as generous investment promotion-related corporate tax relief (Figure 27). 35 Kesner-Škreb, Madžarević-Šujster (2004) 36 It is followed by Hungary and Denmark that collect 18.6 and 17.1 percent of GDP, respectively, also having the highest standard VAT rate. 41 Figure 26. Indirect and Direct Taxes in 2013, Percent Figure 27. Low Tax on Profits, Percent of Profits of GDP Source: Doing Business 2014 Source: Eurostat 4.8 While Croatia’s tax burden is below the EU27 average of 43 percent of Figure 28. Total tax revenues and per capita income GDP, it is relatively high compared to GDP, selected EU countries, 2012 60 many of its peers. The ability of countries to raise additional taxes is often measured 55 FI SE through benchmarking taxes against its 50 DK AU peers, although as economies have very different structures this is an imprecise 45 DE method (e.g. studies must differentiate for 40 HR SK CZ EE PL differing levels of GDP per capita, 35 LV openness to foreign trade, the share of BG RO elderly or agricultural population). A simple 30 comparison between the size of the tax 30.0 50.0 70.0 90.0 110.0 130.0 burden in Croatia and the EU27 average Note: GDP at PPS (EU28=100) shows that Croatia is below the average, Source: Eurostat which might indicate that there is still room for tax increases. However, a comparison of the tax burden in Croatia to that in EU10 with similar GDP per capita indicates that the tax burden is relatively high, and suggests that there is little scope for increasing aggregate revenues without losing further competitiveness. 4.9 Indeed, Croatia appears to be collecting more revenue than suggested by its economic capacity. A comparative regional study37 used a dual approach, looking at “revenue potential (economic)” and “revenue potential (legal)”.38 The difference between revenue potential (legal) and actual revenue collected is the “tax gap.” The difference between revenue potential (economic) and actual revenue collected is the “tax space”—i.e. the amount of revenue a country can afford to collect given its economic strength rather than what the legislature has mandated. It appears that legally mandated revenue potential is often higher than suggested by country’s economic fundamentals. At 3.83 percent of GDP Croatia’s difference between legal 37 Khwaja, S. and Iyer, I (2013). 38 “Revenue potential (economic)” refers to inherent economic capacity and fundamentals of a country, while “revenue potential (legal)” refers to what the legal framework prescribes. 42 and economic potential is only similar to Poland (Table 16); while its negative tax space is still troubling also manifested in high tax arrears. Table 16. Revenue Potentials, Difference in Potentials, and Tax Space, 2011 Percent of GDP Potential Difference Tax Space Actual Tax Ratio Econ Legal Legal-Econ (Econ-Actual) Hungary 38.61 34.95 37.23 2.28 -3.66 Slovenia 37.96 36.10 38.83 2.73 -1.86 Czech Republic 34.58 34.74 35.64 0.90 0.16 Poland 32.81 32.36 36.20 3.84 -0.45 Estonia 31.51 35.41 34.54 -0.88 3.90 Slovak Republic 30.94 35.80 36.99 1.19 4.86 Bulgaria 30.64 30.79 31.59 0.80 0.16 Lithuania 28.87 32.41 31.38 -1.04 3.54 Latvia 28.71 33.51 32.21 -1.29 4.80 Romania 28.58 28.01 30.09 2.08 -0.57 Croatia 34.67 31.02 34.85 3.83 -3.64 B&H 37.57 30.01 30.44 0.43 -7.56 Serbia 36.65 26.49 27.31 0.82 -10.16 Montenegro 35.16 31.12 29.56 -1.56 -4.04 Macedonia 28.79 28.91 29.42 0.51 0.12 Albania 22.16 23.77 26.01 2.23 1.61 Source: Khwaja and Iyer 2013. 4.10 Similarly, other studies suggest that Croatia has a relatively high tax effort. Table 17. Selected Countries Tax Effort, 1994-2009 Estimates of Croatia’s relative tax capacity Actual Estimated Tax Effort Tax/GDP (%) Tax/GDP (%) Index that control for differences in GDP per capita, 1 2 3 = 1/2 old age dependency ratios, the openness to Poland 28.42 27.85 1.02 foreign trade, the share of the agriculture Hungary 35.03 31.19 1.12 sector and the quality of public administration Slovenia 34.91 30.81 1.13 (as measured by the corruption perception Croatia 33.84 28.72 1.18 Portugal 31.32 30.53 1.03 index) find that Croatia has a relatively high Finland 34.83 33.09 1.05 tax effort. Table 17 shows the tax effort as an Austria 35.5 32.54 1.09 index of the ratio between the shares of Greece 33 28.99 1.14 actually collected taxes in GDP and the Italy 36.12 28.98 1.25 estimated tax capacity. Croatia has a ratio of France 38.77 30.05 1.29 1.18, which is similar to that observed in Source: Le, Moreno-Dodson, Bayraktar (2012). Hungary and Slovenia and is around the average for advanced economies. 4.11 Introducing a modern value-based property tax and further rationalizing quasi- fiscal fees would also help, but the revenue potential is likely to be low in the medium term. As noted above, in OECD countries the property tax represents 1.8 percent of GDP39, in developing countries it averages 0.3–0.7 percent40, and in Croatia it accounts for about 0.3 percent of the GDP41. According to the Law on Local and Regional Self-Government Financing, in Croatia the tax on real estate is shared 40 percent by the central government and 60 percent by 39 OECD (2010): “Revenue Statistics 1965-2009”. 40 Slack 2009, Bahl and Martinez 2008, Norregaard 2014. 41 IMF (2012): “Croatia: Government Opportunities for Strengthening Tax Administration”. 43 local governments (municipalities and towns). However, the property tax base is deeply eroded by a combination of legal exemptions, undervaluation, and failure to cover all properties. The problem is exacerbated by low collection rates and poor tax management. The property tax has great revenue potential; its burden rests on middle- and upper-income families, and it causes less distortion to business and consumer economic decisions than other taxes do. Croatia is currently preparing for an introduction of a new ad valorem property tax at a uniform rate of 1.5 percent, to replace existing ‘utility fees’ and the second home tax. 4.12 Continuing to shift the tax burden from labor to indirect taxes could also help support growth. Recent IMF studies have suggested that “revenue-neutral rebalancing that reduces income taxes, while increasing consumption and property taxes is associated with faster long-term growth”, while labor taxes have a negative effect on growth 42. Given the high level of health insurance contribution of 15 Figure 29. The Inactivity Trap percent on the gross wage (almost double 95 compared to EU average), Croatia has 85 considerable scope for this type of fiscal 75 devaluation. However, the fall in revenue 65 55 could be offset by eliminating a large 45 number of tax exemptions. The recent 35 review of child tax allowances shows a 25 potential to collect 0.5 percent of GDP 15 Bulgaria Croatia Romania Lavia Lithuania Hungary Poland Estonia NMS, ECA Slovak Republic Czech Republic Slovenia more by its parametric redesign. Similarly, reinvested earnings tax relief for businesses led to taxes foregone of 0.6 Average wage earner Low-wage/part-time earner percent of GDP in 2013. The design of taxes should also reduce the disincentives Note: The effective tax rate (tax and benefits foregone) for moving from either benefits or applied for average and low-wage earners incomes. Source: World Bank (2013c) and WB staff estimates for inactivity to work (Figure 29). Croatia C. Strengthening Tax Administration to Improve Compliance 4.13 Significant progress has been made in comprehensive reform of the tax administration. CTA was created in 1994 within the Ministry of Finance (MOF). Its primary responsibility is to apply revenue legislation, such as the General Tax Act, the Tax Administration Act, and other substantive tax laws. It is responsible for collecting national and some local taxes, social insurance contributions, and a large number of levies, fees, and other charges. CTA is now being organized along functional lines at the central and regional levels; a full-fledged Large Taxpayers Office (LTO) has been established at the national level in 2012 along with a Taxpayer Service Unit and a unit for Sector Strategy and Development; and all processes are being organized to provide wider coverage of electronic services. 4.14 Although the government counts on better revenue collections and lower compliance costs to regain fiscal stability and international competitiveness, CTA still faces major challenges. For the quality of its services to meet the standards seen in modern tax administrations CTA would have to move from being procedure-oriented to being results-based. The building blocks of modernization process of tax administration are: (i) a compliance risk 42 IMF Fiscal Monitor, October 2013. 44 management system; (ii) solid administration at headquarters and the strengthened LTO; (iii) a streamlined network of regional and local tax offices; and (iv) a sound IT governance. 4.15 CTA should improve corporate governance by streamlining the organizational structure and building up the role of HQ in designing programs and providing strategic guidance. There are too many field offices and they rely heavily on manual procedures and intense and frequent contacts with taxpayers, especially at the local level. This not only overwhelms staff; it prevents them from doing more complex and important tasks, such as audits and collection enforcement. The inefficiencies associated with a large network of local offices will worsen as business processes become more highly automated and e-filing (including CTA pre-filing of tax declarations using third party data) and e-enforcement is extended. 4.16 The CTA headquarters’ has too few staff to adequately direct the large number of local offices, which operate with considerable autonomy and minimal effective oversight. The way CTA is organized does not correspond with principles of modern tax administrations, which distinguish between the role of HQ in designing programs and providing strategic guidance and the operational role of local and regional offices. Consequently, corporate governance should move toward a clear allocation of responsibilities between HQ, regional, and local offices. 4.17 CTA could be given more operational autonomy to effectively and efficiently administer the tax system, while at the same time should be faced with stronger accountability provisions. CTA staffing levels have been eroded due to recruitment restrictions. Additionally, CTA is lagging modern tax administrations with respect to assigning flexibility and autonomy in relation to: (i) designing and implementing its own organizational structure; (ii) determining the level and mix of staff and influencing or negotiating their remuneration; (iii) influencing the staff recruitment process; (iv) allocating resources to respond effectively to changed priorities and risks; (iv) making tax rulings; and (v) setting service standards. Therefore, giving CTA semi-autonomous status and enough power to administer the tax system effectively could be considered. 4.18 A new approach to understanding the taxpayer base and compliance trends, with a modern Compliance Risk Management Model (CRMS), could help improve tax compliance and efficient enforcement. Introducing a CRMS to systematically identify, assess, rank, and deal with tax compliance risks is a priority. Its overarching objective is to stimulate and facilitate voluntary compliance with tax laws and prevent noncompliance by understanding the taxpayers’ base and taxpayer risk profiles. Pursuing a risk management approach would help CTA to: (i) craft a compliance strategy that directs efforts at major compliance risks (e.g., large taxpayers who are the highest compliance risk to revenues); (ii) reflect the strategy in annual operational plans and instructions (e.g., national audit plans, audit selection parameters, taxpayer service plans, and debt collection plans); and (iii) improve compliance tools (e.g., audit organization and methods, taxpayer services, and enforced collection) and the skills of tax administration management and staff to efficiently realize strategies and operational plans. 4.19 Building capacity of CTA staff in the LTO is also a high priority for improving revenue collection. This is often a complex task given the various corporate models and requests for highly specialized skills. CTA has made significant progress in getting a full-fledged national LTO in operation in 2012. The new office has been assigned 700+ large taxpayers who account 45 for about half of tax revenue. The LTO is fully aligned with the taxpayer segmentation approach, which is a core strategy of modern tax administration. 4.20 Building capacity in the new Office for Taxpayer Services and the Sector Strategy and Development Unit will also reduce voluntary compliance costs. Croatia has recently established a dedicated unit for taxpayer services and assistance, with a call center, and tailoring services to unique characteristics and risk profiles of taxpayers (the taxpayer segmentation approach) will be critical for enhancing voluntary compliance. Strengthening this unit, and using the call center to control non-filers and delinquent taxpayers is a priority to enhance compliance. 4.21 CTA has an impressive range of strategic planning documents, but implementation proved to be challenging. The Sector Strategy and Development Unit should enhance the institutional awareness and understanding of strategic directions and manage coordination efforts across CTA to ensure that strategic and business planning processes are linked and translated into instructions and procedures critical for the day-to-day work of local and regional offices. Other priorities could include strengthening IT governance to putting in place an integrated management information system, and improving the planning and exploitation of information. CTA should regain control of its IT systems and information by progressively phasing out the current outsourcing arrangements. Human resource capacity is central to modernizing tax administration. The CTA training strategy therefore needs to be aligned with the reform priorities, current organizational conditions, business reengineering processes, and IT developments. Recommendations 4.22 With a revenue burden already at close to 42 percent of GDP, there is limited scope for increasing fiscal space from additional revenues. Nonetheless, there is scope to raise additional revenues of 2 percent of GDP more efficiently and fairly while reducing adverse impact on growth and employment through modern property taxation, broadening the tax base by eliminating tax exemptions and strengthening tax compliance. 4.23 Croatia would be well advised to introduce modern property taxation. However, tapping the property tax revenue potential would require stronger local governments and a comprehensive reform in four steps: (i) identification of the property being taxed; (ii) assessment of the property value and the tax base; (iii) establishment of the tax rates; and (iv) abolishment of other property-related taxes. It is unlikely that such a tax would generate significant revenue initially until cadastral registry is updated and the real estate market better developed so that valuations can be more transparently calculated. 4.24 Croatia has considerable scope for fiscal devaluation through labor taxes. This is mostly being the case with health insurance contribution. The fall in revenue could be offset by eliminating a large number of tax exemptions given to households as well as businesses. The design of income taxes should also reduce the disincentives for work. 4.25 Strengthening and modernizing CTA would help protect and expand the revenue base. Tax noncompliance risks like tax avoidance and evasion in Croatia are exacerbated by any weaknesses in the tax administration. Weaknesses in the tax administration also make it difficult for CTA to keep pace with national and international developments. As Croatia becomes further integrated into the global economy, CTA will face a range of additional taxpayer compliance risks of considerable complexity. The key areas of focus in modernizing tax administration are: 46 (i) setting up a compliance risk management system; (ii) solid administration at headquarters and the LTO strengthening; (iii) streamlining network of regional and local tax offices; and (iv) establishing sound IT governance. 4.26 While there is a limited space for raising revenues further, a sizeable government spending level has space for downward adjustment. Such an adjustment is needed urgently to support fiscal sustainability. Around 4 percentage points of GDP could be found over the medium term through rationalizing spending on public administration, social sectors and subsidies as discussed in the following Chapters V-VII. 47 Croatia spends more on public administration than most other EU countries, but performs poorly in indicators of public administration effectiveness, the rule of law and administrative barriers to doing business. While the public wage bill, at close to 12 percent of GDP, is already high, further upward pressure is likely to come from two sources in the medium term: (i) further decentralization and the creation of new agencies, often with higher average wages than of national government; and (ii) the need to build additional capacity to absorb sizeable European funds. Nonetheless, some 2 percent of GDP in cumulative savings could be achieved over the medium term through staff rationalization in local/regional and national governments to create a leaner, but more effective administration. A. Entrenched Public Sector Management System 5.1 Public administration in Croatia is characterized by its high cost, large size, but relatively low effectiveness (Figure 30). Figure 30. General Government Wage Bill, Percent Improving efficiency has proven difficult as a of GDP result of a combination of factors. One is the high level of rigidity in the organizational structures and remuneration system in the public administration. Another is the fragmentation of local and regional self- government units43, which are expensive, but not self-sustaining, and proliferation of agencies. Finally, public administration is still highly politicized, which hinders the emergence of a professional managerial cadre. 5.2 Adequate public administration capacities also constitute one of the key Source: Eurostat, MOF. Weighted average for EU15 and requirements for EU membership. The EU EU10. recognizes that “strengthening institutional and administrative capacity, reducing the administrative burden and improving the quality of legislation underpins structural adjustments and fosters economic growth, as well as employment” and produces a range of indicators comparing the performance of government, the judiciary, public administration etc. across EU members44. The experience of the 2004 EU entrants45 suggests that public administration reforms require sustained support across the governments, combined with a professional, merit-based, and independent civil service. The administrative reforms require persistence and transparent monitoring over several political cycles to show results, which has been challenging in Croatia. For example, a government regulatory ‘guillotine’ effort aimed at simplifying business regulations was initiated in late 2006, with almost half of the recommendations for simplification 43 Local government or local government units further in the text. 44 EC. Europe 2020, Key areas: comparing Member States' performances, Modernizing Public Administration, http://ec.europa.eu/europe2020/making-it-happen/key-areas/index_en.htm 45 World Bank (2007), 2006 EU8 Fiscal Studies. 48 or elimination of unnecessary legislation implemented within a year and a half; however, the process was stopped soon after new elections started. Professionalization of top civil servants’ positions that started in 2007 and was hampered by lack of transparency in some cases had been reversed already in early 2012. 5.3 The excessive size hampers decision-making and service delivery. There has been a proliferation of agencies and subordinate entities with operational autonomy and revenue-raising capabilities. This has given rise to many instances of overlapping functions, weak coordination and lack of clarity about accountability among ministries, agencies and other state subordinate entities46. This hampers policy-making and implementation, places a burden on coordination and management and ultimately adversely impacts value-for-money and the quality of service delivery. The functional review of the central administration, carried out in 2007-2008, provided specific recommendations, which could have saved some 7-10 percent of the wage bill47, equivalent to up to 1.2 percent of GDP, and would go some way to moving it toward the average size in EU10 that stands 2.6 percentage points of GDP below Croatia. 5.4 The cost of government services is high in Croatia. The cost of government’s services (e.g., the “size” of government) is assessed taking into account the cost of government services relative to the general price level (a relative price effect). The analysis shows that the size of the government in Croatia is high not only in nominal terms, but also once controlled for its per capita real income, and compared for the intermediate consumption and the wage bill in particular (Figure 31). Figure 31. Size of the Government Real intermediate consumption Real wage bill 14 18 Real intermediate consumption, % of GDP IT BG DE 12 16 EE Real wage bill in % of GDP HU FI LT 10 RO LV EE BG RO LV HU 14 8 PL 12 LT NL R² = 0.3246 6 R² = 0.212 10 IE NL 4 AT AT 8 DE DE 2 6 30 50 70 90 110 130 150 0 30 50 70 90 110 130 150 GDP pc at PPS in % of EU27 average GDP pc at PPS in % of EU27 average Note: Croatia is red marked. Source: EUROSTAT and MOF, Sonje (2013) background paper. 5.5 Moreover, Croatia is rated comparatively low with respect to the effectiveness of its administration (Figure 32). World Bank governance indicators point to general weaknesses of 46 Some examples being the Center for Monitoring Business Activities in the Energy Sector and Investments, the Agency of Environmental Protection and Fund for Energy Efficiency and Environmental Protection, Agency for Vocational Education, Agency for Tertiary Education, Agency for Adult Education, also see Chapter 7 on agriculture. 47 Institute for International Relations (2008), “Functional Reviews and Assistance in the Restructuring of State Administration Bodies and their Subordinate Agencies in Croatia”, World Bank -funded Public Administration Reform Project, November 2008. 49 Croatian public administration in particular in terms of corruption, the rule of law, regulatory quality, and government effectiveness. Croatia lags behind the EU15 and EU10 by around 27 percentage points and 15 percentage points, respectively, on the Governance Indicators ranking. The World Economic Forum's Competitiveness Report 2013-2014 ranked inefficient public administration as the top barrier for doing business in Croatia. Selected Doing Business indicators similarly show the degree to which Croatia lags other EU countries with respect to selected administrative/judicial processes.48 While some improvement has been observed in various dimensions, notably as regards the enforcement of debt contracts, tax administration as well as company registration, wide gaps remain in terms of the number of days required to complete administrative procedures for starting a business, dealing with building licenses, registering a property, and enforcing contracts. Figure 32. Croatia’s Administration Effectiveness Composite Governance Indicator Global Competitiveness Index, Institutions’ Score EU-15 EU-10 Croatia 2008 2012 SEE5 0 20 40 60 80 100 Country's percentile rank Source: WB Governance Matters Composite Governance Indicators Note: Includes Voice and Accountability, Political Stability, Government Effectiveness, Rule of Law, Control of Corruption, Regulatory Quality Note: SEE5 refers to South-East European countries: Serbia, Montenegro, FYR Macedonia, Albania and BiH. Source: World Bank (http://data.worldbank.org) and the World Economic Forum (http://www.weforum.org). 5.6 Croatia’s structure and the low efficiency of key public expenditures leave plenty of scope for improvement. Croatia’s government performance score is the third lowest in the EU, only somewhat better than the performance of governments in Romania and Slovakia (Figure 33). Similarly, Croatia’s public administration is fourth from the bottom in Europe, better only than Malta, Greece and Lithuania. Given the high cost, close to the level allocated in the Scandinavian countries, Croatia’s taxpayers are paying high price for public services, which also tend to be of lower quality. These relative inefficiencies indicate an opportunity for large savings and efficiency improvements without affecting the quality of service delivery. 48 Croatia is ranked the second lowest amongst the EU28 in the Doing Business indicators and has a global rank of 65, only ahead of Malta (94). 50 Figure 33. Efficiency of government performance Public administration performance Government performance score Source: Eurostat 2013. Source: Eurostat 2013. 5.7 The high wage bill reflects the growing number of public sector employees, rather than high wages. The rationalization of military staff has been more than offset by growing employment at local government levels, and in education and health services. At around 293,000 employees, the general government sector employment represents 17 percent of the labor force and is the major employer in the country (Table 18). This number excludes the additional 80 thousands+ staff employed by public enterprises. The total number of people employed in the public sector amounts to 22 percent of the Croatian labor force, some 4 percentage points above the EU average. In specific areas the growth in the number of public servants also does not seem well aligned with the demand for services. For example, Croatia’s public education staffing continues to grow despite a significant decline in the number of students enrolled. Table 18. Public Administration in Croatia, 2000-2012 2000 2006 2008 2010 2012 Civilian 219,313 234,004 245,851 242,185 246,205 Non-civilian 63,939 46,841 45,128 45,910 46,558 Public administration, % of population 6.4 6.3 6.6 6.5 6.9 Public administration, % of labor force 15.3 15.7 16.3 16.5 17.0 Public administration, % of employment 18.2 17.7 17.8 18.7 20.2 Civil servants 73,294 72,004 76,472 67,863 71,390 Central 47,776 38,373 38,981 28,309 30,729 - ministries 23,194 17,375 17,493 18,550 18,193 - agencies 18,633 16,011 17,119 6,265 9,033 - central administration working in LGUs 5,435 4,413 3,759 3,077 3,069 - outside the executive 514 574 610 417 434 Local 25,518 33,631 37,491 39,554 40,661 Public servants 146,019 162,000 169,379 174,322 174,815 - teachers and other education staff 66,440 84,037 88,288 89,136 90,060 - health employees 58,028 56,525 59,201 60,169 62,418 - cultural workers 1,289 1,765 1,838 2,020 1,609 - social workers 9,093 6,430 6,633 9,266 6,775 - judiciary 11,169 13,243 13,419 13,731 13,953 Military employment 63,939 46,841 45,128 45,910 46,558 Total 283,252 280,845 290,979 288,095 292,763 Source: Ministry of Finance 51 5.8 Croatia’s public sector enjoys relatively favorable pay; except for the top caliber staff. Based on labor force survey data, Croatia’s civil and public servants are better off than their private sector peers when it comes to salary levels (higher than in the private sector for most of the public service) and working hours (shorter than in the private sector); the exception being the graduate and post-graduate posts where private sector offers some margin (Figure 34). For highly educated and top caliber staff, public sector wages are around 30 percent below private sector peers. Another symptom is that while labor turnover within public administration is generally low, the only segment where mobility occurs is at the top end as younger and more ambitious public servants leave for employment in the private sector after a few years of acquisition of know-how. Past that point the wage differential with the private sector becomes too high to be compensated by the additional public sector job security. 5.9 Public sector pay is largely based on seniority; not on performance. The reliance on benefits for years of service places an Figure 34. Wage Differentials Based on Hourly Wages by additional burden on the wage bill and Sector and Education Level increases the incentives for older 70 workers to stay regardless of 62.2 60 performance. Croatia consequently has a large cohort of middle-aged staff that 50 47.9 Hourly wage (Kuna) entered the service around the time of 40 36.2 39.7 32.4 32.1 independence. Since there is a built-in 30 26.4 incentive to stay in the system (i.e. staff 23.5 20.2 23.0 20 17.6 17.5 receiving 0.5 percent increments per year of service and a strong bargaining 10 power of public sector unions), the rate 0 Primary and less Vocational General 2-year or Non- College Post-graduate* of natural attrition is low. At the same Secondary* Secondary* University College Graduate* time, a staff retrenchment program, Public Sector PrivateSector with severance packages based on years of service, would be costly (for Notes: Figures refer to full-time employees in paid employment. * indicates that the difference between the average wage in the public someone with 30 years of service it and private sector is statistically significant at the 95% confidence level would equal around EUR15,000), and Source: Staff estimates based on the 2011 LFS. further increase as this cohort ages. In addition to the years-in-service allowance49, there are 12 different bonuses that range from 4 to 150 percent of basic salary, are not mutually exclusive, and could be larger than the basic pay. None of them though is performance related. This non- standard approach across the administration leads to an upward pressure on wages and relates total remuneration to the strength of bargaining power rather than qualifications, merit or performance. 5.10 Although previous Croatian governments recognized the importance of civil service and administrative reforms, the process has been slow and sometimes reversed. The Civil Service Law, adopted in 2005, was to be the basis for a fundamental reform of the public management system, to be followed by legislation on the broader public service, on the public and civil service wage system and by the introduction of new public management tools such as 49 These include bonuses for overtime (50 percent), loyalty bonus (4, 8 and 10 percent, depending on the years of service), work at night (40 percent), work on Saturdays and Sundays (25-35 percent), working on shifts (10 percent), work on holidays (150 percent), special sector bonuses (working in health sector), duty roster bonus, on-call bonus, Christmas bonus, holiday bonus, transportation allowance, and child gift. 52 fiscal, social and environmental impact assessment and strategic planning.50 The government also moved to put in place e-government, reduce the number of ministries, and open a Civil Service Training Center. Seven years after the reform strategy was drafted, however, pay and promotion decisions are not merit-based and the functional review recommendations have only been partially implemented. In contrast, top civil servant positions continue to be largely political appointments. 5.11 Given their impact on the budget, reforms to government employment and compensation are an unavoidable component of spending reforms. Following the current crisis, reforms in public employment and compensation have been undertaken by economies at all income levels throughout the EU (Box 9). Recent wage bill consolidation efforts across Europe have included some short-term measures—such as wage or hiring freezes—combined with structural measures aimed at reforming the public wage formation, performance management systems or reorganizing government (Figure 53). The experience of such episodes suggests that while wages cuts and recruitment freezes may have a positive short-term fiscal impact, structural reforms that improve the effectiveness of recruitment and salary setting tend to be more sustainable (and are often used in combination), while effective downsizing tends to be more targeted than generalized51. Croatia has done similar short-term measures (a 3-percent wage cut, bonuses reduction); however, the problem remains acute. Box 9. Public Sector Compensation is Being Cut across the EU Many countries in Europe have introduced measures to: (i) reduce public sector compensation, either through actual pay cuts (e.g. in Ireland and Spain) or through a nominal pay freeze (e.g. UK, Netherlands); (ii) reforms to performance pay systems (e.g. Italy has revised its collective bargaining arrangements); (iii) cuts in recruitment and (iv) reorganizing public sector bodies. In Poland, the total budget for public administration (which includes the civil service) has been capped since 2009 with a freeze in individual pay rates. This has led to significant reductions in real pay rates. In Slovenia, measures included a reduction of 4 percent in basic salaries for all civil servants and a 1 percent reduction in basic salary for directors, until real growth in GDP exceeds 2.5 percent, combined with a freeze in high-level promotions and the cancellation of some bonuses. The Czech Republic and Hungary have cut public sector wage rates. In Portugal, pay rates have been frozen, along with promotions, performance bonuses and mobility salary changes, while salaries for higher earners were reduced by an average of 5 percent. Source: World Bank, various and OECD (2012). 50 The main innovative features of the law were: (i) the de-politicization of senior management posts, including ‘assistant minister’ positions; (ii) linking grade and pay to merit and performance; (iii) enhancing professional capacities through a code of ethics and training; and (iv) the development of a central managerial capacity. 51 IMF Fiscal Monitor, April 2014, Public Expenditure Reform, making difficult choices 53 Figure 35. Recent (2009–13) and Historical (1979–2009) Wage Bill Consolidation Episodes1 1 Historical episodes are: Austria (1996–97), Belgium (1982), Belgium (1992, 1994), Canada (1991 –92), Italy (1993– 95), Portugal (2005–07), the United Kingdom (1994), Denmark (1983 –84), Germany (1983–84), Germany(1995– 2000), Ireland (1982), Ireland (1987 –88), the Netherlands (1984–86), the Netherlands (2005), Portugal (2000 –03), Spain (1997). Recent ones (from 2009) include: Bosnia, Bulgaria, Croatia, Cyprus, the Czech Republic, France, Greece, Hungary, Ireland, Italy, Japan, Latvia, FYR Macedonia, the Netherlands, Poland, Portugal, Romania, Serbia, the Slovak Republic, Slovenia, Spain, the United Kingdom, and the United States. Note: For examples of structural measures and social dialogue, see IMF Fiscal Monitor, April 2014. Wage bill reduction is calculated over five years, following the first year of reform for the historical episodes; it is calculated as the change from 2009–12 for the recent episodes. We use potential GDP whenever available, otherwise nominal GDP. Source: IMF Fiscal Monitor, April 2014, Public Expenditure Reform, making difficult choices. For historical episodes, Devries and others (2011); IMF Staff Reports for Article IV consultations for various countries and years. For recent episodes, OECD (Public Sector Compensation in Times of Austerity, 2012); IMF Country teams; and IMF Staff Reports for Article IV consultations Reforming public employment and compensation 5.12 To contain overall costs, Croatia will need to introduce a targeted downsizing program. The program needs to target reducing the duplication and inefficiency in the system, aligning employment with the provision of improved and more streamlined and electronic services where possible. The costs and expected savings of the program should also be considered in advance, such as retrenchment payments, and transparently monitored. Some areas, where staffing numbers might be reduced without affecting the delivery of public services, include the following:  Local governments52: A high fragmentation of local government units (over 570 LGUs) bloats employment numbers (as discussed in the following chapter). Large efficiency gains could be generated if some functions are amalgamated or grouped at municipalities/cities/counties’ level. Further decentralization also needs to be followed by respective rationalization of functions at the central government level.  Civilian Staff: Recently, there have been increases in the number of public service employees, including teachers, judicial and medical personnel. In 2012, these three sectors accounted for 57 percent of total public administration. There is a potential for efficiency 52 Local government term is used alternately to local and regional self-government term throughout the report. 54 gains in bringing pupil to teacher ratios back to 1990s’ levels 53 and with rightsizing hospital network.  Judicial sector: Croatia has one of the largest numbers of judges and court personnel per capita in the Europe and Central Asia region, while this overstaffing does not translate into higher effectiveness. Croatia has 42.8 judges per 100,000 inhabitants against the EU10 average of 27.154 and while one can argue differences in their jurisdictions, a removal of non-adjudicative processes away from judges might bring significant efficiency gains.  Administrative support: Functional reviews carried out in 2008 that are still largely relevant, suggested that large inefficiencies exist in line ministries as staff spends an inordinate amount of time on filing and retrieving documents, instead of dealing with clients or policy issues. Electronic services for citizens and businesses, document and case management systems should help reduce labor-intense procedures, improve communication, increase transparency in dealing with the clients, and reallocate staff to higher value added tasks.  Auxiliary services: There is also lot of staff time being spent on auxiliary services that could be outsourced, such as IT services, cleaning, catering and the like on the basis of public tenders. The savings estimated from those amounts to over 0.15 percent of GDP per year. 5.13 A noticeable trend in Croatia's public administration has been the increase in the number of agencies. In 2013, there were still around 70 plus various agencies/institutes at the central level. The "mushrooming" of agencies (a term commonly used) is a concern for the following reasons: (i) there is a lack of clear criteria for deciding whether an agency should be created, and for what purpose. Consequently some agencies have been created with policy responsibilities that overlap with those of ministries; (ii) there is no clear framework determining the relationship between an agency and its "parent" ministry leading to an array of differing arrangements that can obscure accountability; (iii) agencies often have recruited their own staff, rather than take on the staff previously responsible for the function in the parent ministry, and even where a group of staff with scarce technical skills have been transferred, new support staff have often been recruited rather than transferred from the ministry. This has resulted so far always with an increase in civil service numbers, and the annual approval of civil service admission plans has been unable to prevent this. There has been a recent effort to harmonize agency salary arrangements with those of the civil service salary framework and to streamline the number of agencies with some initial fiscal gains achieved (of around 0.2 percent of GDP). 5.14 A modern legislation to regulate the civil service wage system should be introduced. While the modalities for regulating public sector wages are subject to a reassessment process, and a move towards having a single law regulating all public sector wages is being considered, instead of the previously pursued sector-by-sector approach (Box 10), the importance of adopting a new legal framework remains. The Law regulating the wage system was supposed to 53 In 2012/13, pupil to teacher ratio stood at 9, while in 1998/99 it stood at 14. From 1998, the number of students enrolled in primary, secondary and tertiary education declined by 6 percent, while the number of teaching staff increased by 47 percent. 54 World Bank (2014) Croatia: Justice Sector Expenditure and Institutional Review or http://europa.eu/rapid/press- release IP-14-273 en.htm 55 help operationalize the principles of performance-based career advancement inherent in the new civil service system. Unfortunately, such legislation has still not been enacted, which is blocking implementation of significant parts of the Civil Service Law. Special salary schemes and special supplements made the salary system non-transparent and detached from the “equal pay for equal work” principle. The Civil Service Law does not cover local self-government employees, which have their own systems for personnel management and wage setting. This has led to distortions, as especially larger cities pay significantly higher wages to their staff than central ministries. While studies have been conducted on the scope for harmonization between the two systems, this has not as yet led to concrete results. Box 10. The Draft Civil Service Salary Law The 2009 draft Civil Service Salary Law proposed to establish a wage grid system in which horizontal progression in steps (grades) is based on performance appraisal only, while vertical progression in brackets (class) is based on internal or open competition. The draft law also rationalizes supplements and allowances, which is compensated for by more advantageous conditions for career progression for those that perform well. The draft law foresees a 3- percent step increase every five years, as long as performance is at least ‘good’ for four of the five years of this period. For outstanding performers, a 3-percent step can be granted every year while for excellent performers this can happen every two years. Career progression is also possible through obtaining higher qualifications. However, dismissal is also possible for those that rate unsatisfactory twice in consecutive assessments. In terms of the compression ratio, the Law proposes a compression ratio of 1:5.8 between the highest position (former Directors, now Head of Sectors) and the lowest clerical grade included in the civil service system. This is on the lower end of European practice but well within reasonable limits. The ratio would be comparable to that in France (1:5.7) but well below the Netherlands (1:8.2) or the UK (1:9.8), though above Italy, where the ratio is slightly below 1:5. An increase in the compression ratio would also help in addressing the relatively disadvantaged position of senior level civil servants vis-a-vis their peers in the private sector. 5.15 The practice of annual review of staffing numbers and plans by the Ministry of Administration and the Ministry of Finance needs to be mandatory. It is not clear how profound the degree of analysis has been in past years, although certainly some bodies have been asked to justify and reduce their staffing. In view of the fiscal crisis, a freeze on all but the most essential increases was imposed. This has, however, led to only a marginal decline in employment in central ministries and state administrative bodies, while increased in public sector agencies. Because there will be an obvious need to control tightly civil service numbers for the foreseeable future, this exercise should in future be undertaken in relation to a set of priorities that should be agreed with the Ministry of Finance. The strategic plans now being submitted by all budget units and the functional reviews covering a number of state administration bodies provide an obvious starting point for this. Performance appraisal and job classification 5.16 The new performance appraisal system has been applied; however, the delay in the adoption of a law regulating the wage system55 has created a situation in which the new system is rapidly being discredited even before its full implementation. The Civil Service Law requires all posts to be classified using the following criteria: professional qualification required, complexity of tasks, independence in work, degree of co-operation with other state 55 The law was introduced in Sabor in 2008, for implementation with the 2009 budget, and in view of the importance of the law in the EU accession context the Government opted for the use of the fast track procedure. However, following vehement protests from trade unions, the draft law was withdrawn from Sabor at the request of the trade unions representing civil servants, which argued that the Law should not be adopted through the fast track procedure. Since then, union opposition also remains strong. 56 bodies and communication with clients, degree of responsibility and influence on decision- making implementation. Some public administration bodies have done this, while others are part way through this process. A revised staff appraisal system has been adopted with training delivered to managerial staff, new forms and regulations were produced. However, there are several pitfalls to this process:  There is considerable skepticism amongst the staff about the new arrangements, particularly suspicion about the objectivity of assessments. This is not particular to Croatia: such suspicions arise in any system in which new, performance-based appraisal arrangements are introduced. Experience in other countries shows that it is necessary to deal with this proactively, by providing information to staff, reinforced by briefings, seminars and question and answer sessions.  Second, the staff appraisal system will face an equally grave challenge to its credibility from the general tendency of ministries to grade staff too generously. From a sample of ministries it is clear that the number of civil servants being graded “unsatisfactory” is almost inexistent, which is not credible, although there are some instances reported of staff who have lost promotion prospects due to poor appraisal results. There is also a widespread tendency to over-grade staff: over two thirds of staff have been graded "extraordinary" (defined as “performance and efficiency of the highest quality, ensuring optimum and consistent service”). In many bodies, an implausibly high number of staff have been rated “outstanding” or “excellent”. Clearly this cannot provide a serious basis for performance-related pay, even if the element of pay related to performance is small. 5.17 The operation of the staff appraisal system would need to be improved by (a) linking it to pay and promotion prospects. In addition, the Ministry of Administration needs to (b) continue providing more detailed and better guidance, including a quota system, to managers on the application and interpretation of the new appraisal grading, and (c) ensure that information about the new system is proactively provided to staff. This should be done in advance of the implementation of a link between salary and appraisal. 5.18 Many countries have experienced challenges when it comes to the introduction of a functioning performance appraisal system, especially where these are linked to career progression or salary enhancements. Box 11 describes the way the introducing of performance appraisal mechanisms was handled in the Austrian federal system, where, like in the case of Croatia, there had been a strong concern among civil servants and the trade unions about the introduction of this tool. While in Austria career progression decisions take into account a combination of factors, performance appraisal has become a key element in such decisions. Box 11. The Austrian Performance Appraisal System The Austrian Federal Administration created a good foundation for a merit-based reward system with the staff appraisal carried out annually. The performance appraisal is based on individual employee objectives that are set each year in agreement between the civil servant and his/her direct supervisor. Staff appraisal. As an instrument of management by objectives, staff appraisal constitutes an occasion for setting agreed objectives for the coming year on the basis of what was achieved in the previous year. Such MBO agreements are reached through cooperation, with due regard to the objectives set by the next higher management level (top-down approach) and taking into account the interests and qualifications of each staff member (bottom-up approach). Staff appraisals focus on mutual feedback on the quality of cooperation and represent a common quest for possible improvements or, in case of conflict, for solutions to such conflicts. Staff appraisal constitutes also a staff development tool. On the basis of the performance attained and the specific interests and skills of each staff 57 member, options for career development are to be discussed and concrete development measures agreed upon. Performance appraisal. The performance appraisal is based on individual employee objectives that are set each year in agreement between the civil servant and his/her direct supervisor. The laws for civil and public servants provide special possibilities for awards. It is necessary to provide an allocation for performance awards for each ministry in the annual budget. The minister is free to set the modalities of allocation and amount. For example, Federal Chancellery provides 400 Euro for each staff member for extraordinary performance. One person can get a maximum of 550 Euro per year. If you give one staff member the maximum you have to shorten the awards for other staff members. In the case of Federal Ministry of Finance, a staff member can get the maximum of a half monthly salary as an award. But such awards can be provided only to 30 percent of all staff members. Source: Verheijen (2012), background paper. Based on input from Gerhard Ungersboeck, former SG of the Austrian State Chancellery. Depoliticization and performance management 5.19 The special nature of managing top level appointment processes is well recognized both in administrative practice and in academic literature and the related challenges are many. Solutions could include defining specific procedures for the management of top level appointments, while remaining true to merit principles. Examples of special procedures for managing top level appointments in the civil service can be found in several EU Member States. Two examples are the Irish Top Level Appointments Committee and the Dutch General Top Civil Service initiative (Box 12). Box 12. Top Level Appointments Procedures - Some EU Examples Ireland instituted in the 1984 the Top Level Appointment Committee structure. Through this mechanism, access to top level appointments was to be broadened (Millar and McKevvitt in Bekke and van der Meer (2001)). The mechanisms used were a change in procedures as well as a change in Committee composition. While the Irish example is a response to perceived bias in favor of staff from the recruiting ministry, and did not directly relate to the issue of politicization, it might be an option to look into as for as its mechanisms are concerned. In the Irish case the result was the opening up of top level positions for talented staff (which previously had been virtually impossible), as well as a decrease in the bias in favor of staff from the same ministry. Similarly, the Netherlands broadened out recruitment processes for top level staff in the late 1990s, in part in response to the same problems as the one described in the Irish case, but in part also as a response to what was considered ‘creeping politicization’. Like in the Irish case, top level appointments became the responsibility of inter- ministerial teams, with added measures to facilitate rotation of senior officials between ministries (see van der Meer and Roborgh in Bekke van der Meer (2001)). The Dutch case also involved the creation of a dedicated Secretariat to monitor the application of the new recruitment and appointment principles. Source: Verheijen (2012), background paper 5.20 Croatia needs to secure a more stable and professional middle and lower management layer. Every political cycle so far has destabilized the institutional memory of the administration. Apart from the Minister, his/her Deputy and all Assistant Ministers are required to leave the office. Croatia could follow the example of European countries where political appointments get limited to only first two layers of the ministries’ management (Ministers and Deputies). The professionalization of middle and lower management is crucial going forward. While politicization and a high staff turnover might not have overly visible adverse effects on the delivery of public services in the past, the EU membership has raised demand on administration significantly. 5.21 The depoliticization of senior management posts would need to be extended to the Assistant Ministers/Heads of Directorate level bodies through the same recruitment and appraisal arrangements that now apply to lower grades. While this principle should not be 58 compromised on, it would be useful for the government to look into some of the management modalities for top level appointments used in other EU member states, countries in which ministerial autonomy in making top level appointments was previously highly prized, even if not for reasons of political patronage. Such mechanisms could be introduced without affecting the principle of de-politicization, which is not the case in current practices. 5.22 Countries that have introduced transformative reforms in recent years have had to strengthen central civil service management functions. The Ministry of Administration and senior managers need to have a clear picture of developments in the administration and what capacities and structures need to be built if the reforms are to succeed. This would be partially supported by the introduction of the central HR Management Information System (HRMIS), but it also needs to be complemented by a series of indicators on, for example, absenteeism, recruitment, retention, aggregate appraisal data, training and the capacities of HR units in state administrative bodies. 5.23 Public administration reform remains difficult and highly sensitive issue in Croatia . Adherence to current systems and practices remains strong, regardless that these have created a civil service system that is costly and structurally underperforms. This sets Croatia apart from most EU10 countries in that while these states also (continue to) have underperforming civil service systems; they do not exercise the same high fiscal burden (except Slovenia). If reducing the size of the state administration has been politically difficult, and is likely to remain so, improving the performance of the system is essential if Croatia is to become a performing EU Member State. There are interesting lessons to pick up from the EU10. For example, key success factors in Latvia was the creation of a Reform Administration Council—a body that was created at the height of the crisis and included representatives from trade unions, employers federation, local government officials and other key stakeholders. The body reached consensus on key decisions and then tasked the various representatives with communicating the changes to their constituents. Recommendations 5.24 At 12 percent of GDP, Croatia spends more on public administration than other EU countries, but performs poorly in indicators of public administration effectiveness, the rule of law and administrative barriers to doing business. Nonetheless, some 2 percent of GDP in cumulative savings could be achieved over the medium term through staff rationalization to create a leaner administration. This would need to be combined with strengthened performance management to create a more effective administration. The Civil Service Law contains all the necessary ingredients to make this move towards creating a more performance oriented civil service, provided it is complemented by the relevant legislation on wages. In a sense, the legal framework as prepared is significantly more advanced than that of many of the new EU member states. 5.25 A targeted downsizing program needs to be adopted to contain the overall cost of public administration accompanied by outsourcing, electronic service introduction and introducing staffing norms in particular sectors such as education, health and judiciary. 5.26 Provisions regulating criteria for the creation and management of agencies should be finalized. While there are different modalities for regulating the status of agencies, the following elements could be considered as part of any move toward a more orderly management of the creation and management of agencies: (i) Set clear criteria for the circumstances in which 59 an agency may be created; (ii) Create a clear and standardized framework of management relations and accountability (especially in relation to budgetary matters) between ministries and agencies and ensure the transparent reporting of key administration and performance metrics; (iii) Bring agencies fully within the framework of civil or public service legislation including legislation on salaries; (iv) Create immediately a mechanism to review, harmonize or abolish all agencies created in the past five years with powers to ensure that agencies and their staffing conform to the new framework; and (v) Require that, before any future agencies are created, a statement of the rationale for doing so and the anticipated benefits should be published, along with a fiscal impact assessment. 5.27 Reforming the wage system to reduce the wage bill in the short term, and proceeding with long-term structural reforms that will enhance public sector performance is recommended. In terms of the new wage system, any new legal draft should include the linkage between performance and pay, to make the necessary corrections to the performance appraisal system, as well as to ensure that it is applied in a consistent manner across the public and civil service, and to allow for proper implementation of the classification of posts, including the consideration of measures to mitigate the impact of large numbers of local level civil servants not meeting qualification requirements of their posts, in the event that this scenario indeed materializes. This should include the design of and budgeting for social mitigation measures. 5.28 The full application of the HRMIS needs to be accelerated to improve cost controls . Until the HRMIS is fully implemented across the public sector, the Government lacks one of the fundamental tools for controlling staffing numbers and salaries. At present, through the Central Payroll Accounting System, the Ministry of Finance has means to verify the number of civil servants employed in ministries and state administrative organizations, but not yet public servants, military and agencies’ employment. Staff appraisal information, absenteeism, training and education, professional career development is currently missing information, while being critical for the proper HR management. The absence of this control mechanism carries serious fiscal risks. 5.29 Finally, stalled attempts to enhance strategic planning, professionalization and the introduction of performance-based management practices in civil service institutions need to be reinvigorated. It is difficult to see how a system such as the one designed in Croatia could function without being part of a broader reform process to put public sector management on a more performance-based footing. While these reforms are less controversial and visible, they are not less important to Croatia achieving its strategic objective of becoming a well-integrated and well performing EU Member State. B. Local Government Finances 5.30 Decision-making, financing and public services delivery is highly centralized in Croatia, compared to most EU10 countries. In Croatia the subnational government budget corresponds to 5 percent of GDP; in EU10 it averaged about 8 percent during the pre-accession year. The government is aware of the importance of giving citizens more voice and transparency and accountability in matters of local interest, in order to deliver public services more efficiently. However, the current fragmentation of local government units (LGUs) (Figure 36) also makes it 60 hard to serve citizens effectively while fragmentation also limits LGUs’ financial and human capacity required to maximize the use of EU funds. 5.31 Croatia has a two-tier subnational government system; municipalities and cities at the local level and counties at the regional level. There are 20 counties plus the capital city, Zagreb, which is considered both a county and a Figure 36. Average Inhabitant per Local Government, EU, city. It also has 127 cities and 428 2011 60 municipalities for a total of 576 Thousand subnational units. Their size varies 50 considerably. Zagreb has most people, 779,000, and the smallest area, 641 km2. 40 More than half of municipalities have 30 fewer than 3,000 citizens56. About two- 20 thirds of Croatians live in cities, though about half the cities have fewer than 10 10,000 people. Croatia also has Areas of 0 Special Concern (ASSC) and Hilly and PL BG PT CZ DFR SK BE CY AT LU IT HU ES DK SI FI LV SE NL IE EE MT LT DK RO HR EL Mountain Areas (HMA). Croatia has one Source: Eurostat. of the lowest numbers of inhabitants per LGU in the EU (Figure 36) with one of the largest concentration of citizens in the capital city (Table 19). Table 19. Local Government Units and Population Distribution Note: FBiH stands for Federation of BiH, while BiH RS stands for Republic of Srpska. Source: World Bank (2013a), South East Europe Municipal Finance Review. 5.32 There have been several initiatives since 2001 to build up local fiscal and management authority. However, LGUs continue to execute a small fraction of general government spending, and the share has not changed much in the last decade: it rose from 13 percent in 2000 to about 17 percent in 2008, but it fell back to 14 percent after the global financial crisis (Figure 37). Croatia’s subnational spending, at about 5 percent of GDP, is similar to that of Portugal and Slovakia, and far below medium or highly-decentralized two-tier countries like the Czech Republic, Hungary, or Denmark (Figure 38). The Baltic States and Finland, with local spending of more than 25 percent of total general government spending, are much more decentralized than Croatia. 56 Bajo, Skrok (2013) background paper. 61 Figure 37. Percentage of General Government Figure 38. Local Spending, EU and Croatia, Expenditures Executed at the Local Level 2011 Subnational spending (in % of general government total) Subnational spending (in % of GDP) 60 50 40 30 20 10 0 MT CY EL IE LU BE PT ES AT HR DE SK BG SI FR HU EU27 EE RO CZ LT UK LV IT PL NL FI SE Source: MOF. Source: MOF Spending assignments are unclear 5.33 Counties and larger cities have taken responsibility for a wide range of public services. In the first phase of decentralization in 2002, the fiscally strongest 32 cities and all counties were given responsibilities to provide some public services locally. Counties took responsibility for activities of regional importance, such as primary and secondary education, health care, urban planning, economic development, and traffic and transport infrastructure. Cities and municipalities took responsibility for accommodating the immediate needs of their residents, such as housing and community amenities, municipal services, culture, sports, protection and promotion of the natural environment, fire-fighting, and local transport, as well as partially child, social and primary health care, as well as primary education. Responsibilities were later gradually transferred to other units on a discretionary basis. About 152 local government units have taken over one or more decentralized functions (Table 20)—usually starting with fire protection. Table 20. Distribution of Local Units According to Decentralized Function Health Social Secondary Primary Fire Issuing Total Care Care Education Education Protection Building Number of Permits Unitsb Counties 20 20 20 20 20 20 20 Towns/Cities - - - 34 64 30a 66 Municipalities - - - - 66 - 66 Total 20 20 20 54 130 50 152 Source: MOF, Annual Report 2012. Notes: a Number of the major towns; more towns have since taken on this responsibility, but their exact number is unknown. b Those that have taken over at least one of the decentralized functions. 5.34 Spending responsibilities among local government units are complex and unclear. All municipalities and cities except the very largest have the same responsibilities for providing public services no matter what their development level, fiscal capacity, or size. Cities that have more than 35,000 inhabitants or that are county seats are exceptions because they can also perform tasks otherwise allocated to counties. This creates a risk of duplication of tasks and inequality in the services provided as small municipalities and cities with little fiscal capacity cannot give their residents the same kind and quality of public service as larger cities. 62 5.35 Large cities provide most of decentralized public services, questioning the existence of other structures. Although counties have been given wide responsibilities for public services, they often have much less fiscal capacity than cities. Thus, the role of counties in providing public services, as measured by spending in per capita terms, is very small (Figure 39). That is also true for municipalities, which do not perform many public functions. Thus cities are most active in providing services. However, only a quarter of all Croatian cities finance some primary education. Likewise, 60 LGUs that have more than 8,000 inhabitants have not yet committed to performing decentralized functions like primary education. Figure 39. Subnational Budget Items, per Capita, Figure 40. Expenditures, Unconsolidated (% of 2009–12 GDP) 8.0 Social protection 6,000 All LGs 7.0 Education 5,000 Municipalities Recreation, culture and 6.0 religion Cities 4,000 Health 5.0 Counties Housing and community 3,000 4.0 amenities Environmental protection 2,000 3.0 Economic affairs 1,000 2.0 Public order and safety 0 1.0 Defence tax revenue expenditure wage bill investment 0.0 General public services 2009 2010 2011 Source: MOF. 5.36 The provision of public services by LGUs is limited. Of the four public functions (education, health care, welfare, and fire protection) that after 2002 were partly transferred to subnational governments, only fire protection is almost completely decentralized. Although spending on education accounts for about 20 percent of total local government expenditure and is the largest spending item, the national budget covers teacher and staff wages and other important line items and national legislation specifies teacher salaries, teaching loads, and rules for operation of educational institutions. As for public health, since only a small part of the responsibility for primary health care was transferred to counties, health spending is still a very small share of local spending (Figure 40). For social protection, too, responsibilities and spending are shared by the local and central levels. In effect, subnational governments spend mostly on community development (housing and communal infrastructure) and general public services, which in reality means mostly financing the functioning of LGU executive and legislative bodies’ i.e. local administration. 5.37 Half of LGUs spending is allocated on wages and operational costs, while a little less than a quarter of subnational financing goes to investment. While before the crisis, about 23 percent of subnational budgets were directed to investment (Table 21), in 2011–12 that fell to about 17 percent. As a proportion of their budgets, municipalities spend the most on investment, counties the least (only 16 percent of their budgets). LGUs spend a relatively high proportion of their budgets—just under one third—on goods and services, mainly energy, utilities, maintenance, and other material expenditures. Although the national government pays teachers and doctors, another major category of local spending is salaries. Subsidies, mostly to large 63 public commercial companies, account for a small share of subnational spending except in the City of Zagreb. Table 21. Structure of Economic Classification of Local Spending, 2009-12 Municipalities Towns and Cities City of Zagreb Counties (20) Total LGs (428) (127) Operating Spending 66.4 76.5 79.4 83.9 76.8 Compensation of employees 11.9 21.5 19.9 14.2 18.4 Goods and services 29.7 29.4 28.2 39.2 30.5 Interest 0.8 1.3 0.8 1.0 1.0 Subsidies 1.8 2.4 12.7 3.7 5.6 Grants to foreign governments and 2.0 0.6 0.8 6.7 1.8 to other general government units Social transfers 5.3 4.4 3.8 6.2 4.6 Other expenses 15.0 16.9 13.3 12.8 14.9 Capital spending 33.6 23.5 20.6 16.1 23.2 TOTAL SPENDING 100.0 100.0 100.0 100.0 100.0 Source: MOF, WB staff calculations. Excessive fragmentation 5.38 Service delivery is hampered by excessive fragmentation. The small size of many municipalities, relative to their wide range of responsibilities, raises the question of whether the scale of administrative overhead and public services is efficient. Though it is difficult to discern the precise relationship between local government size and efficiency, research shows that efficiency begins to drop off significantly below 5,000 inhabitants.57 Croatia has a relatively large number of municipalities below that size. In countries where there are too many subnational entities, or at least too many that are too small to be viable, issues of spending and tax assignments cannot be properly addressed without territorial reorganization or introducing incentives for municipalities to merge voluntarily or at least to cooperate over service provision. 5.39 After territorial reorganization, the second best option to reap economies of scale would be through mergers or coordination. In OECD countries several approaches have been tried. Some countries have encouraged amalgamation. Incentive mechanisms have been introduced in other countries to encourage cooperation (Box 13). 57 Fox and Gurley (2006). 64 Box 13. Mergers and Coordination Economic theory suggests that local efficiency gains can be achieved through merger or coordination, which internalize spillover effects and produce economies of scale. While the latter option is easier, it may end up being less transparent and missing clear Typology of Municipal Merger Policies accountability lines. Still, cooperation might be feasible for countries where municipalities have wide responsibilities. It might also be a prelude to eventual merger of municipalities. Municipality mergers could be voluntary or mandatory. If voluntary, municipalities are free to join together. In Finland, to promote municipal mergers, grants were offered to merging municipalities. The amount of the grant depended on the number of municipalities merging, the size of the municipality after the merger, and previous sizes of the merging municipalities. France and Finland promote cooperation. France recognizes inter-communal structures as legal entities and gives them partial Source: OECD (2006), Efficiency of sub-central spending, workshop subsidies. In some countries, like Turkey and proceedings. Japan, municipalities may join forces in formal associations to perform such joint tasks as school management and waste disposal. Mechanisms to Encourage or Discourage Amalgamation or Cooperation are: (i) progressively increasing grants (bigger entities obtain larger grants); (ii) Threshold requirements for investment grants (to obtain financing there is a minimum population size); (iii) One-off and formula-based compensations; (iv) Deterrents (promoting smaller municipalities by giving them larger grants). Limited fiscal independence 5.40 Subnational governments rely heavily on transfers from the national government, which will have to be reduced given the fiscal pressures. Although LGUs can collect their own taxes, almost two-thirds of their revenues accrue from central government transfers, including shared taxes, which is the most significant source of their income (Table 22). The second most significant local government revenue is non-tax revenue, mainly from communal fees, followed by other current transfers. For counties, government transfers constitute more than 85 percent of their revenues. Table 22. Revenue Sources of Croatian Local Governments, 2009 –12 Revenue All Local Counties Municipalities Cities Governments Local revenues 33.7% 13.9% 40.3% 37.9% Taxes 2.8% 5.0% 3.0% 2.3% Nontax revenues 30.9% 8.9% 37.3% 35.6% Transfers from central 66.3% 86.1% 59.7% 62.1% government Shared taxes 55.6% 66.4% 39.0% 57.1% Other transfers 10.7% 19.7% 20.8% 5.1% Current 6.7% 16.2% 9.7% 3.8% Capital 3.9% 3.5% 11.0% 1.2% Source: MOF, World Bank Staff 65 5.41 The tax-sharing system is complex. Taxes on income and real estate transactions are shared based on taxes actually collected in that constituency. Though county and local levels have their own tax bases, the national government sets or restricts their tax rates. Table 23 shows the general shares of different levels of government. Some income tax revenues go to an Equalization Fund, which is used to transfer resources to subnational governments that have assumed responsibility for decentralized functions that revenue from the standard tax share cannot fully finance. Municipalities and towns in Areas of Special State Concern (ASSC) and Hilly and Mountainous Areas (HMA) receive higher shares of income tax revenue (90 percent, with the share of counties at 10 percent). Since 2001 cities and municipalities can add surtax to personal income in their jurisdiction, payable in that jurisdiction; the central government sets the maximum surtax rate. Table 23. Model of Income Tax Revenue-Sharing, 2013 Central Counties Municipalities City of Zagreb Decentralized Government and Towns Functions*) (Equalization Fund) General model of income tax-sharing 0% 16% 56.5% 72.5% 12% (15.5%) Tax revenue-sharing on islands Personal income 0% 16% 56.5% 12%**) - Tax revenue-sharing with ASSC and H&M areas** - 10% 90% - - Real estate tax 40% 0 60% 60% Note: Transferred functions are elementary education (3.1%), secondary education (2.2%), health care (3.2%), social welfare (2.2%), and firefighting (1.3%). Revenue from these shares remains in the subnational governments that have assumed responsibility for the function in addition to a share of joint funding of capital project investment equal to 15.5% .** ASSC (Areas of Special State Concern); HMA (Hills and Mountain Areas) 5.42 Subnational governments have relatively limited taxing powers, while the equalization system plays as disincentive to collect taxes. Local tax sources (inheritance and gifts; taxes on motor vehicles, boats and vessels, and gambling machines) are often subject to central government caps. Nontax revenues are therefore an important subnational revenue source. In this respect, Croatia stands out from the other EU countries. Income from property tax is also significant for all subnational units. Equalization of fiscal and service capacity is realized in Croatia through tax-sharing, the Equalization Fund, and grants. The system lacks transparency. A different tax-sharing arrangement concerns ASSCs which are from 2013 grouped per criteria of lagging regions. The equalization formulas are usually supplemented with additional ad-hoc transfers. As a result, they cannot act as an incentive. Furthermore, grants that have an equalization purpose for decentralized functions are not necessarily used for this purpose. 5.43 The current transfer design does not relieve fiscal capacity disparities either. The system should be based on tax capacity rather than on historical spending and revenue. Full equalization removes the incentive to increase the tax base of a jurisdiction by attracting new economic activity or increasing the efficiency in service delivery. To retain this incentive, the national government could opt for less than complete equalization, so that the gap between the jurisdictional and the average tax base is only partially compensated. The grant system could, for instance, be designed so that after equalization the tax capacity of the poorest jurisdiction is 66 within 10–20 percent of average tax capacity. Another possibility is to design equalization, as in Sweden, to reduce the difference between a minimum guaranteed tax capacity and actual tax capacity. Equalization reforms introduced recently in Portugal illustrate well-designed equalization systems (Box 14). Box 14. Reform of the Portuguese General-Purpose Grant System The reform reinforced equalization and created a horizontal component of equalization. Allocation criteria were changed by increasing the weight of population and environmentally protected areas, reducing the lump sum transfer to municipalities, and reducing the weight of the number of parishes in the allocation formula (thus penalizing municipal fragmentation). A Social Municipal Fund was created to disburse earmarked grants to compensate for the cost of responsibilities transferred to municipal governments for education, health care and social assistance. The “minimum guaranteed growth” clause was abolished; it had guaranteed that municipalities would not receive less in real terms than in the previous year. Finally, to ensure fiscal neutrality for the national government, the tax shares were reduced to compensate for the increase of the share of the municipalities in the personal income tax.  Source: OECD 2011. Supporting fiscal consolidation 5.44 Fiscal imbalances of municipalities and counties raise questions about their viability. Unconsolidated local government balance ranges from -0.5 to 0.2 percent of GDP. However, consolidated balances show persistently high vertical fiscal imbalances, in particular for municipalities and counties (Figure 41). Today more than half of subnational spending is covered by subnational tax revenues—in the EU countries only Sweden and Austria had higher ratios in 2011 (Figure 42). This is, however, a simplistic measure of vertical fiscal imbalances because it does not capture own revenues only. This share has not changed much over the last decade, pointing to a structural problem. Large vertical imbalances may relax fiscal discipline. A common view is that high reliance on intergovernmental transfers or borrowing “softens” the local government budget constraint, mainly because the cost of spending is not adequately internalized. Figure 41. Fiscal Imbalances, 2002–2012 Figure 42. Coverage of Subnational Spending by Subnational Tax Revenue 70.0% 2011 2000 80 60.0% 70 50.0% 60 Municipalities 40.0% 50 Towns and Cities 30.0% Counties 40 20.0% Total LGs 30 20 10.0% 10 0.0% 2003 2007 2010 2011 0 EL NL RO BG UKHU IE CY LT PL LU PT DK BE DE CZ IT SI ES FI EE SK FR LV HR SE AT Source: Eurostat and MOF. Source: MOF and Alibegović 2010. 5.45 Local debt is low but there may be fiscal risks arising from indirect borrowing. While their debt is low (around 1.5 percent of GDP), there are possible fiscal risks stemming from off-budget activities of subnational governments, such as guarantees (explicit and implicit) for municipal companies and any arrears they incur. How much local government units in 67 Croatia can borrow is defined by the national government annually. The two main limitations on local borrowing are a general limitation on the borrowing of all local units and then a specific limitation on the borrowing of each local unit. These borrowing limits do not include borrowing limits for communal utilities and/or guarantees issued by LGUs. The main problem that needs to be underscored is the lack of transparent criteria for a local government unit to get approval from the central Government for borrowing. 5.46 While being an opportunity, the EU Funds will put additional pressure on local finances. The heavy subnational functional and financial dependence on the central government will restrict the possibility of their matching the funds (the EU co-financing element), which will effectively limit their financial absorption capacity. The capacity problem relates partly to the small size of many LGUs. The experiences of Estonia and Latvia demonstrate that there is a definite correlation between the small size and weak capacity. The requirement that a small municipality employs qualified people proved to be a serious problem for the Baltics and the Czech Republic because the local pool of experts was small and the administrations could not afford qualified staff from elsewhere. One solution—that the national government creates local development centers, such as the regional development agencies in Croatia, which give advice and support to local governments—have not shown the expected results either. The financial absorption capacity of Croatian local governments could be built up, however, by creating incentives or mandatory rules for LGUs to merge or otherwise unite to provide services; and attracting private capital inflows and involvement through public-private partnerships (PPPs). Recommendations 5.47 The current fragmentation of LGUs makes it hard to serve citizens effectively while fragmentation also limits LGUs’ financial and human capacity to effectively absorb EU funds. Croatia has one of the lowest numbers of inhabitants per LGU in the EU with one of the largest concentration of citizens in the capital city. Large cities provide most of decentralized public services, thus questioning the existence of other LGU structures. Addressing the optimal territorial organization and functional and fiscal decentralization matrix are priorities for Croatia for both fiscal and EU funds’ absorption reasons. 5.48 Territorial organization would be the best way to address current disproportionate number and resources assigned to LGUs. If not pursued for the lack of political support, territorial organization should be replaced by persuasion and incentives to encourage joint provision of public services. 5.49 Spending responsibilities need to be redefined to avoid duplication and overlap of functions and to increase accountability of LGUs for the tasks allocated to them. Fiscal decentralization should clearly specify responsibilities (for example, in such areas as education, social protection and health) and identify resources to finance them. Accountability and efficiency concerns imply that expenditure assignments should have clear lines of demarcation between various fiscal units, and there should be transparent reporting of progress in carrying out tasks. Closely related activities should be assigned to the same level of government. 5.50 Greater subnational reliance on own-source revenues would help promote funding adequacy and reduce central government transfers. The ideal LGU tax is one that can be levied on a relatively immobile and well-demarcated local tax base—a property tax (see Tax section). Administrative capacity is also an important consideration in defining revenue assignments. Croatia could start by simplifying the framework guiding tax-sharing, eliminating a 68 number of local taxes and relaxing the upper limits for them, and improving the design and administration of property taxes that provide stable yield and are indirectly linked to income. 5.51 The fiscal operations of subnational governments need to be monitored to ensure fiscal prudence and alignment with the Excessive Deficit Procedure. It is critical that subnational governments be fully transparent in reporting all balance sheet and off-budget activities. Timely and comprehensive information is needed on both subnational budget operation and debt under the new EU fiscal governance rules. In this vein, some countries (e.g., Portugal) require that municipal accounts be consolidated with those of their public enterprises and submitted to external audit. 5.52 Finally, priority should be given to strengthening subnational administrative capacity, among else to assess and execute investment projects, especially considering the expected large inflows of EU funds. 69 Despite some solid social outcomes, Croatia’s social spending is inefficient, and inequitable. Croatia can achieve equivalent or better outcomes with lower and more targeted spending. The urgency of reform arises from short-term fiscal pressures and long-term aging of the population and its consequences for the demand for social services. About 2-3 percent of GDP in cumulative savings could be achieved over the medium term if the most inefficient programs were eliminated or rationalized and targeting improved. This would help support medium-term fiscal consolidation, ensure adequate buffers to cover for age-related spending and contribute to poverty reduction and reversing the trend of rising social exclusion of disadvantaged groups. 6.1 Croatia’s combined social spending (on health and social protection) at 22.3 percent of GDP is three percentage points higher than of its EU peers, with considerable scope for rationalization and improvements in efficiency and equity. Also, as Croatia’s population ages, pressures for spending on pensions and long-term care will increase. These long-term issues, along with the macroeconomic need for medium-term fiscal consolidation, imply an urgent need for a comprehensive reform. This chapter reviews the main issues in each major area of social spending and, where possible, quantifies potential savings from rationalization, reallocations, and improvements in targeting of various programs. It concludes that social spending in Croatia could be significantly streamlined, especially in the areas of health care and social assistance, but less so in the pension and long-term care systems. The combined savings from comprehensive reforms could range from cumulative 2-3 percent of GDP over the medium term, depending on the boldness and speed of reforms. Some proposed reform measures for government’s consideration have been in the pipeline and may just need to be firmed up or accelerated; others will require bold new efforts––health, privileged pensions and social assistance––and their sustained implementation over the medium term. 6.2 In Croatia, as in its neighbors, the largest population cohort is already between 50 and 59, relatively close to retirement (Figure 43 and Figure 44). The baby boomer generation was born slightly later—mainly between 1951 and 1960—in Croatia than in Western European countries. It is also Croatia’s largest population cohort ever. In most Western countries, the largest cohort is the Generation X, the children of the baby boomers, who are currently between 40 to 49. Croatia’s Generation X is slightly younger, currently 25 to 34, but its members are significantly smaller than their parents. As a consequence, the elderly population in Croatia will fairly soon expand dramatically. 70 Figure 43. Proportion of Population 65 and over, EU, Figure 44. Median Age of Population, EU, 2012 2012 Source: Eurostat. 6.3 The demographic transition is not a gradual process; it comes in waves. For Croatia, these waves set in earlier than in other countries, and are larger. Each successive decade will bring growth in the share of the old and oldest people in the total population. In Figure 45. Croatia-Proportion of Population aged 65+ EU10 by 2050 the share of people over and 80+ 65 will already exceed 30 percent, which will not happen in the EU15 until after 2060. Croatia is more like the EU15 average (Figure 45). The share of people aged 80 or more, who have a higher risk of dependence, will exceed 10 percent in the EU27 in 2050 and in Croatia about a decade later, when the EU27 average will be 12 percent (Figure 46). Source: UN Figure 46. Demographic Transition in Croatia (2011–61) 2011 2061 85+ male female 85+ 80-84 80-84 75-79 75-79 70-74 70-74 65-69 65-69 60-64 60-64 55-59 55-59 50-54 50-54 45-49 45-49 40-44 40-44 35-39 35-39 30-34 30-34 25-29 25-29 20-24 20-24 15-19 15-19 10-14 10-14 5-9 5-9 0-4 0-4 -10.0 -8.0 -6.0 -4.0 -2.0 0.0 2.0 4.0 6.0 8.0 10.0 -8.0 -6.0 -4.0 -2.0 0.0 2.0 4.0 6.0 8.0 Source: CROSTAT 71 A. Improving the Efficiency, Equity and Delivery of Health Care 6.4 Croatia’s health system spending is excessive, and remaining inefficiencies and arrears call for a significant reform. In 2012, Croatia spent about 9 percent of GDP on health58 (out of which around 7.9 percent of GDP from the central budget), compared to an average of about 5.4 percent in EU10. Although health outcomes are moderately good, they come at a high cost. Widespread underreporting of wages means that wage-based contribution revenues to finance the system are low. However, the contribution rate of 15 percent to finance the public system is also high. Hospital admission and occupancy rates are high by Western standards and no longer suited to the needs because the system was built mainly to deal with acute illnesses. Arrears in the health system, mainly from hospitals and pharmaceuticals, are around 1 percent of GDP, indicating ongoing weaknesses in public financial management. 6.5 The health system’s sustainability is complicated by rapid aging. Croatia has reached stage 4 of its demographic transition–the contracting phase--with a negative natural population growth (-0.7 percent in 2007-2012). Since 2001, the 65+ age group has grown to be larger than the population aged 15 years and below. Non-communicable, chronic diseases and morbidity will continue increasing, with attendant need for additional health and long-term care services (LTC). Health outcomes 6.6 At first blush, Croatian citizens enjoy comparatively good and improving overall health outcomes (Figure 47). Croatia’s life expectancy is one to two years longer than that of comparator countries such as Hungary, Slovakia, and Estonia. Past burden of infectious diseases has now been largely replaced by chronic non-communicable diseases, with leading causes of death now being heart and blood vessel diseases. Between 1990 and 2010, life expectancy increased from 72.5 to 76.5 years, infant Figure 47. GDP and Life Expectancy at Birth in Selected Countries, 2010 82 mortality fell from 10.7 ESP ITA FRA to 4.4 infant deaths per GRC GBR Life expectancy at birth, in years 80 SVN 1,000 live births, and PRT GER the age-standardized 78 76.5 CZE mortality rate decreased 76 ALB CROATI A from 1,060 to 790 BIH POL EST SVK R² = 0.8404 MKD deaths per 100,000 74 ARM MNE SRB TUR HUN LTU inhabitants. Finally, GEO BGR ROM LVA 72 some 70 percent of responders are satisfied 70 UKR with the quality and MDA RUS efficiency of the public 68 health system, just 0 10000 20000 30000 40000 US$ per capita below Western Europe, GDP Turkey, and Estonia Source: World Development Indicators. EU selected countries with GDP lower (80, 79, and 78 percent, than US$40,000 58 Eurostat, COFOG methodology 72 respectively) and at the same level as Slovenia and Latvia59. 6.7 But Croatia scores much worse in important health areas, such as diabetes, cancer, and smoking and alcohol-related diseases. Diabetes death rates are about 50 percent higher than in EU10. Death rates due to cerebrovascular diseases, female breast cancer, traffic accidents, and smoking-related diseases in Croatia are also higher than the average for EU10. This is mainly a result of life-style factors such as unhealthy diet, physical inactivity, obesity, and, in particular, smoking, and alcoholism. Smoking in Croatia is the leading cause of the disease burden, while consumption of alcohol is also higher than in the EU. Some 27.4 percent of people aged 18 and over are smokers (33.8 percent men and 21.7 percent women), more than in either the EU15 (25.6 percent) or EU10 (25.4 percent) (Table 24). Table 24. Health Indicators in Selected European Countries and Croatia, 2009–2010 Lithuania Romania Slovakia Slovenia Hungary Croatia Estonia Poland Latvia Serbia Czech EU10 Rep. 2010 2009 Life expectancy at birth, in years 76.9 77.5 75.3 74.5 73.3 73.2 75.9 73.6 74.1 75.4 79.5 75.2 Infant deaths per 1,000 live births 4.4 2.9 3.55 5.1 7.8 4.9 5.6 10.1 7 5.7 2.4 6.3 Maternal deaths p/100,000 live 9.2 2.5 0 18.7 46.1 0 1.9 21.1 19.9 11.4 4.6 9.1 births Age-Standardized Mortality rates per 100,000 All causes, all ages 790 744 840 915 952 963 810 959 972 860 625 863 Female breast cancer, all ages 27.4 20.1 22.1 28.1 25.2 24.2 20.3 22.6 30.2 21.3 25.5 22.2 External causes, all ages 52.7 48.2 87.8 59 86.7 116 57.6 53.9 42.8 51 61.1 57.7 Motor v. traffic accidents, all ages 9.6 6.9 7.2 8.5 9.6 11.4 10.7 12.7 9 7.2 8.1 10.1 Diabetes, all ages 20.2 13.2 11 17.9 14.9 6.9 13.7 8.2 27 10.8 7.5 12.8 Selected alcohol related causes 86.3 71.3 119 114 116 153 84.6 109 63.7 84.9 95.6 93.8 Selected smoking related causes 350 316 325 429 443 490 248 434 350 427 193 342 Source: http://data.euro.who.int/hfadb/, July 2012 Equity in health sector 6.8 There remain wide socio- Figure 48. Inequality in Reported Long-term Illness in economic and geographic disparities Croatia and Selected EU Countries, 2010 in health indicators in Croatia. Statistics on income and living conditions show that disparities in reported long-term illness between the rich and the poor are higher in Croatia than almost anywhere else in the EU (Figure 48). Together with Bulgaria, Croatia has the largest gaps between the richest and poorest, both for long-term illness and self-reported health status. Source: Eurostat, SILC. While in rich districts such as Zagreb, 59 EBRD and World Bank, Life in Transition, 2010. 73 Zadar, Bjelovar and Bilogora, and Karlovac, infant mortality rates have been comparable to the EU average or even lower, in poorer areas such as Lika and Senj, Osijek and Baranja, Virovitica and Podravina, they are higher than the country average and almost twice as in other richer areas (Table 25). Table 25. Infant Mortality Rate by County, 2001-10 Overall Overall County 2001 2005 2010 2001-2010 County 2001 2005 2010 2001-2010 period period Croatia 7.7 5.7 4.4 5.78 Sisak and Moslavina 5.1 8 4.5 5.82 Primorje - Gorski Krapina and Zagorje 6.5 3.3 1.6 4.34 Kotar 7.1 5.5 4.4 5.88 Zagreb 7 3.5 2.7 4.58 Virovitica - Podravina 5.8 7.7 7 6.05 Zadar 7.3 2.2 4.5 4.7 City of Zagreb 8.4 4.9 4.8 6.09 Bjelovar and Bilogora 8.5 4 5.6 4.89 Vukovar and Srijem 5.8 7.3 2.2 6.28 Karlovac 8.1 8 2.7 4.9 Osijek and Baranja 5.8 8.8 7 6.41 Dubrovnik and Neretva 4.7 5.6 5.6 5.02 Požega and Slavonija 8.2 4.8 5.1 6.57 Koprivnica and Križevci 3.3 6.5 3.4 5.22 Brod and Posavina 9.2 5.1 5.8 6.93 Split and Dalmatia 8.3 6.6 4 5.26 Istria 10.2 5.6 5.3 7.04 Varaždin 11.7 4.9 4 5.54 Međimurje 12.6 7.3 3.1 7.62 Šibenik and Knin 5.2 6.3 4 5.73 Lika and Senj 11.5 4.4 7.4 8.35 Note: The last column, which represents mortality computed over the 2001-2001 period, provides a more accurate picture of differences across counties than yearly data which randomly fluctuates due to the statistical rarity of the event considered (for instance, Lika and Senj 2005 is a random fluctuation) Source: Croatian Health Service Yearbooks and Author’s computations. 6.9 Access to pharmaceuticals appears to be more equitable across regions. Still, in 2011, the difference between the lowest and the highest county values in terms of prescriptions per insured individual, expenditures per insured individual, and expenditure per prescription was about 20 percent (Table 26). 74 Table 26. Issued Prescriptions by County Office, 2011 County number of number of number of expenditures total drugs insured prescriptions prescriptions per expenditure per insured per individuals insured individual individual prescription Bjelovar 121.652 1.511.286 12,42 775,6 62,44 94.359.901 Čakovec 111.373 1.367.703 12,28 714,9 58,21 79.615.654 Dubrovnik 129.347 1.443.857 11,16 707,0 63,34 91.451.690 Gospić 50.409 595.146 11,81 736,5 62,38 37.124.416 Karlovac 129.772 1.634.151 12,59 785,6 62,39 101.947.002 Koprivnica 116.155 1.410.641 12,14 674,3 55,52 78.318.770 Krapina 132.083 1.753.993 13,28 748,1 56,33 98.806.720 Osijek 309.572 3.628.929 11,72 674,0 57,50 208.663.310 Pazin 213.841 2.310.824 10,81 708,9 65,60 151.590.323 Požega 79.370 921.430 11,61 604,5 52,07 47.981.370 Rijeka 303.691 3.671.612 12,09 764,9 63,27 232.302.459 Sisak 175.434 2.201.664 12,55 738,2 58,82 129.505.454 Slavonski Brod 165.793 1.741.282 10,50 632,1 60,18 104.798.639 Split 472.721 4.803.508 10,16 610,9 60,12 288.766.310 Šibenik 110.248 1.380.959 12,53 743,2 59,34 81.941.440 Varaždin 174.456 2.169.607 12,44 695,1 55,89 121.263.860 Vinkovci 185.129 2.162.354 11,68 700,4 59,97 129.671.608 Virovitica 86.835 1.103.246 12,71 703,8 55,39 61.114.234 Zadar 171.240 1.918.125 11,20 679,7 60,68 116.386.078 Zagreb 1.123.474 12.508.745 11,13 688,9 61,87 773.942.699 Croatia 4.362.595 50.239.062 11,52 694,44 60,30 3.029.551.937 Source: HZZO Efficiency concerns 6.10 Health financing is organized according to social health insurance principles. A single fund, the Croatian Institute for Health Insurance (CIHI), covers the entire population (1.52 million active workers, 1.05 million pensioners, 1.15 million family members and 0.63 million individuals covered by special programs). A particular feature of the health insurance scheme is that it is divided into a mandatory/compulsory part60 and a complementary part, which is also managed by the CIHI but to which participation is not mandatory. 6.11 Croatia’s health system is facing a massive and growing gap between available public resources and expenditures. Croatia spends significantly more than countries in the region with similar gross national income (GNI) per capita (Figure 49), but, as discussed above, delivers only average outcomes expected for countries at its income level. The generosity of health benefits, a low proportion of private financial contributors, and the legacy of a publicly financed system with no significant pressure for efficiency are putting an increasing strain on the budget. Health accounts for 18 percent of public expenditures, higher than the average for all EU countries (around 16 percent). Past reforms have mostly concentrated on mobilizing additional financial resources and shifting health expenditures from public to private sources but the reforms so far have not focused enough on the efficiency and control of public health expenditures. The system remains hospital-centric (Figure 50), where wage bill consumes over 60 percent of the overall cost. 60 The current mandatory system includes a very large package of personal health services (virtually universal) with a limited list of negative services. While expanding the list of negative services would be very difficult to implement, the system would benefit from establishing evidence-based Croatian-specific protocols by level of care. These Croatian protocols could be used to reform the basic package of personal services so that it emphasizes not a positive list of diseases but rather equity and ease of access and use of “care pathways.” 75 Box 15. Health Sector Finances: Data Issues In the absence of National Health Accounts (NHA) data, getting a good picture of the sector’s finances from a public finance perspective is a challenge. The main source of information on revenues and expenditure is the CIHI data. However, this has a number of important limitations:  The CIHI data does not provide a complete picture of public expenditure on health. Indeed, public health expenditure needs to include expenditure incurred on health from public funds, beyond the CIHI (so direct expenditure by other institutions (eg Ministry of Health, local governments), are missing.  Part of the CIHI revenues and expenditures, which pertain to a voluntary coverage scheme, are technically not public but private health expenditure.  Part of the CIHI revenues and expenditures are social protection expenditures, not health expenditures.  Private, out-of-pocket, health expenditures are not well captured by the Household Budget Survey/SILC conducted annually by the CROSTAT. As the following sections show, estimates of public expenditures also lack precision because of payment delays prevalent in the system. A large and increasing share of payments is made with delays (and if technically arrears are low, it is in part because creditors are forced to accept ever increasing delays in payments). In addition, all institutions do not necessarily agree on what constitutes their respective and mutual payment obligations. The net effect is thus difficult to assess, but clearly, the envisaged NHA exercise will be useful. 6.12 The stock of payment arrears increased over the last ten years. At the end of 2013 health institutions overdue liabilities amounted to HRK 3.3 billion (1 percent of GDP), or about 15 percent of their revenues that year. This is a result of several problems having to do with reporting and accounting, spending control as well as the contracting arrangements. First, receivables are likely to be overestimated as many invoiced services are outside the scope of the contracts signed with the health institutions. Second, the distinction between arrears and liabilities is somewhat fluid with liabilities representing nearly 44 percent of institutions’ revenue in 2013 (HRK 9.7 billion or 2.9 percent of GDP), about a third of which are arrears and are differentiated by the length of the contracted payment period. Third, most of the arrears are liabilities for pharmaceuticals, utilities, medical materials, blood, and blood derivatives. Bulk of arrears has been overdue by more than a year. Figure 49. Health Expenditure and GNI per Figure 50. Structure of Health Spending, Percent capita, Croatia in the World (2010) 5.6 15.6 Primary health care 10.4 17.6 Drugs by prescriptions Polyclinic-consultative specialist health care Hospital care Compensations 41.4 9.5 Other Source: WHO, Global Health Expenditure Database. Source: MoF, HZZO Countries are ranked by per capita income. Croatia is red-shaded diamond. 76 Generating savings and increasing the efficiency of hospital care 6.13 While capacity in hospitals––which absorbs about half of health spending if properly accounted for––at the first sight, is not excessive,61 there is significant scope for efficiency improvements. There is room to optimize hospital inpatient capacity for acute and long-term care and to adjust the health service delivery model to the changing needs of the population. An analysis of the efficiency and use of beds shows that in 2012 clinical hospital centers, clinical hospitals, and clinics had the highest utilization (81 percent), followed by general hospitals (72 percent). Special hospitals were the least efficient, with 70 percent bed occupancy in 2012, at the level of 2011. 6.14 Overall, current bed occupancy rates are within the EU average, but the length of treatment is excessive. The length of treatment in any type of hospital was nine days, one day over the EU average. Difficult areas of hospital reforms have been their design, organizational structure, and their institutional behavior. While a new performance payment system (Diagnosis- Related Groups system) has been put in place, structural and cultural constrains have limited the benefits of the new system. 6.15 Ambulatory units’ services provide moderate amount of good but low-cost services, with potential budgetary saving from their greater use. Analysis done for this report shows significant cost savings could be obtained by conducting a proportion of surgeries on an outpatient basis. These surgeries were undertaken on an inpatient basis but current international protocols highlight the benefits of conducting them on an ambulatory basis. Data was collected from four hospitals in Croatia, namely clinical hospital center Osijek, clinical hospital center Split, general hospital Slavonski Brod, and general hospital Zadar (Table 27). Out of the 9,787 surgeries included in the sample (all surgeries for seven conditions performed in these hospitals in 2011), only 176 (1.8 percent) were conducted as specialized ambulatory services. Table 27. Surgeries Performed in Selected Hospitals, 2011 Hospital Osijek Split Slavonski Brod Zadar Total Selected Surgery Inpatient Amb Inpatient Amb Inpatient Amb Inpatient Amb Inpatient Amb Cataract surgery 971 1938 547 777 4233 0 Surgery of varicose veins 107 41 43 118 309 0 Surgery of inguinal/femoral hernia 487 715 182 36 305 1689 36 Anus (hemorrhoids) surgery 65 113 57 41 276 0 Testicular surgery 242 341 3 85 12 110 778 15 Removal of osteosyntethic material 211 3 236 77 91 43 84 2 622 125 Tonsillectomy 406 893 216 365 1880 0 Total 2489 3 4277 80 1221 91 1800 2 9787 176 Source: MoH, Staff collected for this study 6.16 Shifting select surgeries that under international protocol should be outpatient to ambulatory units would result in significant savings. Cross-country evidence shows that performing these surgeries as outpatient procedures could reduce their unit costs by about 30 to 70 percent (Table 28). 61 Croatia has 1.6 hospitals and 565 hospital beds per 100,000 inhabitants, whereas older EU member states have close to an average of 2.7 hospitals per 100,000 inhabitants and about the same hospital bed rate as Croatia . 77 Table 28. Savings in Outpatient Costs Compared to Inpatient Costs for the Same Procedure with Proven Medical Evidence Unit cost Source Country Procedure(s) saving Babson 1972 UK Hernia repair & varicose vein surgery 40–44% Prescott et al. 1978 UK Hernia repair & varicose vein surgery 65% Evans and Robinson 1980 Canada Pediatric surgery 70% Coe 1981 United States Hernia repair 65% Flanagan and Bascom 1981 United States Hernia repair 70% Rockwell 1982 United States Hernia repair 45% Caldamone and Rabinowitz 1982 United States Orchidopexy 56% Pineault et al. 1985 Canada Hernia repair & tubal Ligation 12–26% Heath et al. 1990 UK Laparoscopy, arthroscopy & cystoscopy 49–68% Arregui et al. 1991 United States Laparoscopic cholecystectomy 46% Mitchell and Harrow 1994 United States Hernia repair 36% Kao et al. 1995 United States Anterior cruciate ligament Repair 58% Mowschenson and Hodin 1995 United States Thyroidectomy & para-thyroidectomy 30% van den Oever and Debbaut 1996 Belgium Inguinal hernia repair 43% Zegarra et al. 1997 United States Laparoscopic cholecystectomy 25% Levy and Mashoof 2000 United States Open Bankart repair 56% Kumar et al. 2001 UK Anterior cruciate ligament repair 20–25% Rosen et al. 2001 United States Laparoscopic cholecystectomy 11% Lemos et al. 2003 Portugal Laparoscopic sterilization 62.4% Source: European Observatory Controlling utilization 6.17 The efficiency of primary health care sector could be improved. As in most countries, primary health care in Croatia plays two key roles in the health system: the entry point and a “gate keeper” for higher level services. In 2012, there were 2,532 primary care teams that provided services to 3.4 million patients out of the 4.5 million beneficiaries of mandatory health insurance. Rates of referrals from primary to secondary care vary greatly across settings for epidemiological as well as organizational reasons and there are no formal standards for referral rates. Yet, cross-country studies show that they tend to range between 5 and 12 percent. Croatia’s 25 percent appears on the high side, which in turn suggests that primary health care might need further strengthening. 6.18 The following factors help explain Croatia’s high rates of referrals to the secondary level:  Increasing need for technological support for diagnoses and treatment;  Still inadequate financial incentives to resolve cases. Currently the payment mechanism includes some payments linked to specific services (for example, for EKG, sutures); nevertheless, the main component of the payment system is still primarily per capita;  The practice of defensive medicine with excessive medical tests mostly caused by perverse budgeting and a lack of control of resource utilization;  Increasing demand from patients for specialized services;  A lack of updated care pathways to define the specific role and scope of primary care services in the implementation of clinical guidelines; 78  An excessive focus on curative services that reduces the extent to which primary care doctors can focus on disease prevention and health promotion activities. 6.19 The need to differentiate long-term care from acute services is increasing. The need for long-term care for chronic diseases and social care as well as for terminally ill patients is growing due to changes in the age structure and health of population who has gone through the demographic and epidemiologic transition. So does the need for non-hospital facilities, that are designed, organized, and funded to provide these specific services, grow. Existing inpatient services need to be rationalized while, on the other hand, ambulatory services, day care services, and ambulatory surgeries need to be significantly expanded and, in many geographical areas, entirely set-up. 6.20 It is important to implement interventions for orthopedic prosthetic and other medical devices and materials similar to those that have been implemented for pharmaceuticals. Expenditures on medical devices have increased over the years, and a new set of regulations should be devised to define standards, reduce costs, and increase their availability. As in the case of pharmaceuticals, clearly defining the list of standard devices included in the basic package as well as standard reference prices could result in significant savings for the sector. Central procurement of some selected products already appeared to save around 40 percent. CIHI now uses several types of risk sharing agreements62 in the case of pharmaceuticals to ensure that suppliers have a financial interest in ensuring that the volumes prescribed are in line with budget limits. Through these ordinances, Croatia was the first country in the world to implement national ethical marketing practices of pharmaceutical companies (through an innovative revolving fund63), restricting their practices of promoting drugs through doctors’ offices and requiring them to be transparent as to how they are spending their marketing budgets. However, spending on pharmaceuticals is still on rise both for volume and the price reasons (Table 29). Table 29. Prescriptions, 2005-2012 2005 2006 2007 2008 2009 2010 2011 2012 Number of issued prescriptions (Million) 34.8 36.2 39.8 43.2 41.8 49.2 50.2 52.9 Ratio of number of prescriptions per insured individuals 8.1 8.3 9.1 9.9 9.6 11.3 11.5 12.2 Average expenditure per issued prescription (HRK) 89.4 89.8 78.9 78.6 79.3 57.6 72.5 74.4 Source: CIHI, staff calculation 6.21 By regional and international standards, the share of private expenditure is low in Croatia and consideration should be given to increasing out-of-pocket payments. Over the last decade, out-of-pocket payments slightly decreased and have amounted to 14.5 percent since 200864. This is far below the average of East European countries and is even lower than in EU15 countries (Figure 51). Patients in Croatia have to pay out of pocket for services provided by private health facilities, which have no contracts with the CIHI. They also have to pay out of 62 Specifically, Pay Back Agreements define maximum expenditure limits for a particular drug and trigger reimbursements or donations of the pharmaceutical company if exceeded; and Cross Product Agreements allow the listing of one drug (usually an innovative drug) on condition of reducing the price of a second drug resulting in a neutral or positive budget impact for CIHI. 63 To be reimbursed each quarter, pharmaceutical companies must deposit their marketing budget in a fund with CIHI and report on the use of those funds. Entering into ethical marketing agreements is voluntary, but mandatory for adding new drugs to the market. 64 Eurostat 79 pocket for any services that are not covered by mandatory health insurance, unless they have taken out additional insurance (either through the CIHI scheme or privately) to cover those services. Figure 51. Changes in the Share of Out-of-pocket Health Care 6.22 Careful consideration Payments, 1997-2010 will have to be given to private participation in health costs to ensure this does not erode financial protection for the vulnerable groups and that it avoids error of inclusion. At the same time, the private insurance market should be fostered and allowed to compete on all market segments and at the same time relieve some burden of the public sector. In 2013, around 2.5 million persons, close to 60 percent of Source: WHO, Global Health Expenditure Database. the population, were covered by CIHI’s complementary health insurance scheme. This complementary insurance covers co-payments imposed by the mandatory insurance but it does not expand the scope of services insured beyond the basic package. The premium for the complementary health insurance is set by CIHI. There are 3 levels of premium, between HRK 840 and 5,108 annually. A person’s premium category is primarily determined by their income and working status (the retirees pay less than the insured at a given income level)65. In addition, the state covers the premium of certain categories (persons with disabilities, organ and blood donors, regular students older than 18 years old, persons with total income per family member lower than 45.6 percent of the basic minimum wage, veterans) that in 2012 accounted for 22 percent of all beneficiaries. In addition, children (around 18 percent of beneficiaries) are also covered for by the state. While some narrowing of the categories entitled to coverage by the state has happened since 2010, there is a need to introduce means-testing targeting rules for exempted categories. Improving accounting and management 6.23 The transparency of public health expenditure accounting and management needs immediate attention. Currently, Croatia does not have accurate estimates of public health expenditure, inter alia, because of under-reporting of expenditures and various liabilities and arrears. A welcome effort to build National Health Accounts will contribute to greater 65 There are seven categories regarding the computation of the complementary health insurance contributions: (A) retired persons with monthly income lower than 5.108 HRK which have to contribute for the voluntary complementary health insurance 840 HRK annually; (B) retired persons with monthly income higher than 5.108 HRK have to contribute 960 HRK annually; (C) insured persons with monthly income lower than 5.108 HRK have to contribute 960 HRK annually; (D) insured persons with monthly income higher than 5.108 HRK have to contribute 1.560 HRK annually; (E) insured persons who are family members of the insured persons from categories A, B, C and D have to contribute 960 HRK annually; (F) others insured persons have to contribute 960 HRK annually; and (G) for insured people who comply with the article 14a of the Voluntary health insurance law the premium is paid by the Croatian state. 80 transparency and better monitoring of resources. At the CIHI level, activities which pertain to mandatory and complementary health insurance should be separated, along with the separation of various social protection benefits such as maternity leave. Consideration could also be given to mainstreaming the management of these benefits into social protection programs which might be better equipped to benefit from economies of scale or have better fraud-detection programs. 6.24 Health indicators could be usefully disaggregated by geographical areas (regions or counties) and socio-economic status and there is scope to improve ambulatory service in poor areas. While health indicators for regions or counties are being collected, they are not analyzed, and differences in health status or in the distribution of health factors seem to be hidden in the national averages. Greater use of ambulatory services in poor areas could also help reduce the service gap with richer areas. 6.25 The financial accountability of public health facilities must be strengthened as a matter of priority, so that they are under increasing pressure to balance their budgets. Hospital managers should have more flexibility, but at the same time should be accountable for the expenditures and liabilities incurred by their hospitals (fiscal discipline). Incentives should also be put in place to better cap public health expenditure (global budgets). Recommendations 6.26 There is significant room to rationalize and improve the service delivery model in the Croatian hospital system and primary health care, while at the same time address private participation in the health costs. There is a sizeable reform agenda in health care centered on three main areas. First, significant budget savings and efficiency improvements should be made in hospitals—the largest cost segment of the system. Second, short-term and long-term measures must be put in place to improve continuous control of utilization. And third, these improvements must be accompanied by increased transparency and better accounting and measurement to engender lasting gains in efficiency. Given considerable waste in the system, it is estimated that the following measures could generate 1 percent of GDP in saving without adversely affecting the level and equity of service:  Consolidating health service networks by geographic and functional areas to streamline services for acute cases. Each regional network should have a “head” regional hospital, and county and local service hospitals and ambulatory services and a referral and counter-referral network. The National Plan for the Development of Clinical Hospital Centers, Clinical Hospitals, Clinics and General Hospitals for 2014-16 could address the excessive and fragmented network and should address the compliance with hospital care standards (in particular in hospitals with well overdue maintenance) and accreditation protocols.  Identifying redundant capacity to merge services and facilities. The proposed realignments should pay particular attention to special health facilities (like spas) and mono- profile services/hospitals and reassess their role in light of current and planned needs and new technical developments. Consideration should also be given to outsourcing the delivery of some services to the private sector and more broadly to a range of public-private partnerships, which could leverage more private financing to serve the systems’ strategic goals. Converting or creating services for long-term care, including those for chronic diseases and palliative care, out of the excess hospital capacity would be advisable. Concentrating 81 specialist services (such as vascular surgery and stroke) in fewer hospitals could reduce mortality66 and co-morbidities.  Creating high-frequency lower-cost specialized centers for ambulatory diagnosis and treatment. The reporting system should be adjusted to report both the different services, but also introduce specific incentives to conduct the services as inpatient services based on updated care pathways.  Reducing and controlling the referral rate of primary health care. Building on the initial steps taken to introduce performance-based payment systems that take into account the percentage of cases solved, the extent of prevention and promotion activities undertaken, and fees-for-service for specific practices and/or goals that trigger bonus payments, interventions might include: (i) implementing communication campaigns to advertise the key roles played by primary care; (ii) updating the clinical guidelines and create care pathways that define the specific role and scope of primary care services and train the primary care teams in the use of the new care pathways; and (iii) implementing quality control procedures (technical audits) based on care pathways to ensure that payments are actually linked to results and good practices.  Developing and implementing national care pathways and putting them to practice in clinical guidelines in the context of the adjusted networks. These protocols should be used to reform the basic package of personal services applying standard protocols to specified context and promoting organized and efficient patient care based on the evidence-based practice.  Expanding public health services to reduce the prevalence of behavioral risk factors. In addition to the social costs associated with high mortality rates, all these public health problems significantly increase the cost of health services.  The stock of arrears and payment delays should be eliminated and procedures and safeguards put in place to prevent their reoccurrence. The accumulation of arrears and liabilities in the form of payment delays only offer a temporary financial relief to the CIHI or for hospitals. Further prolongation of payment deadlines may jeopardize the supply of health services by these institutions and, thus, the implementation of any agreed contracted program with the CIHI.  A review of exempt copayment categories needs to be undertaken and the complementary health insurance premium adjusted with actuarial standards. For poor households, the exemptions should be based on the means-testing targeting procedures. Other categories of exempt population should be rationalized.  Review a potential for sin taxes, earmarked for health sector. This would go beyond the current relatively low excises on tobacco.  eHealth systems needs to be expanded. The e-prescription mechanism has proved to be an effective way to supervise and control the use of pharmaceuticals. Extending this application to other areas of public health and service management would create additional opportunities to control and rationalize the use of the sector’s resources. In addition, eHealth systems 66 Naylor et al (2012) 82 would make it easier to conduct technical audits to increase quality and reduce quality differences, but also to reduce useless interventions with no cost-benefits. B. Pension System Sustainability 6.27 Croatia is one of the oldest and rapidly aging transition economies with significant challenges faced by the pension system. At 17.3 percent of the population over 65 and an old- age population dependency ratio of 23.8 percent (Figure 52)—it is older even than the average high-income OECD country, which averages 16.6 percent over 65 and a 23.1 percent old-age dependency rate. In 2013, the pension system had 1.4 million contributors and 1.2 million pensioners––a system dependency rate of 1.17––one of the worst ratios in the region (Figure 53). Figure 52. Population above 65 and Old Age Dependency Ratios, Transition Countries Population 65+ in total Old age dependency ratio (65+/15-64) 30.0 20.0 18.0 25.0 16.0 14.0 20.0 12.0 10.0 15.0 8.0 10.0 6.0 4.0 5.0 2.0 0.0 0.0 Bosnia and Herzegovina Russian Federation Slovenia Uzbekistan Georgia Bosnia and Herzegovina Montenegro Turkmenistan Armenia Bulgaria Kosovo Kazakhstan Latvia Russian Federation Azerbaijan Georgia Serbia Romania Belarus Estonia Uzbekistan Lithuania Croatia Montenegro Tajikistan Moldova Slovenia Albania Latvia Armenia Kazakhstan Bulgaria Hungary Turkmenistan Kosovo Poland Turkey Azerbaijan Slovak Republic Moldova Serbia Romania Ukraine Lithuania Belarus Croatia Tajikistan Kyrgyz Republic Albania Estonia Macedonia, FYR Czech Republic Hungary Poland Turkey Slovak Republic Ukraine Kyrgyz Republic Macedonia, FYR Czech Republic Figure 53. System Dependency Rates, High Income and Transition Countries System dependency rate (contributors/pensioners) System dependency rate (contributors/pensioners) developed countries and Croatia transition countries 4.5 3.5 4 3 3.5 2.5 3 2 2.5 2 1.5 1.5 1 1 0.5 0.5 0 0 Bulgaria Azerbaijan Croatia Poland Romania Kazakhstan Belarus Hungary Armenia Uzbekistan Montenegro Turkey Slovenia Lithuania Ukraine Russian Federation Latvia Czech Republic Macedonia, FYR Kyrgyz Republic Serbia Slovak Republic Albania Moldova Estonia Bosnia and Herzegovina Bosnia Republika Srpska Netherlands Finland Japan Belgium Croatia Ireland France Germany New Zealand Norway Denmark Sweden United States Italy Canada Spain Austria Portugal Luxembourg Greece Source: CROSTAT, EUROSTAT, staff calculations Dependency concerns 6.28 The system dependency rate is expected to worsen to 1 in 2032 and to 0.77 contributors for each pensioner by 2058. Currently it compares to the countries in the EU with the lowest coverage, such as Spain and Greece. Compared to other transition countries, however, Croatia’s system dependency rate is about average. Certain progress has been made recently with the pension reform but many issues remain unresolved (Box 16). 83 Box 16. Croatia’s Pension Reform in Perspective The main challenges for Croatia’s pension system are how to deal with low replacement rates, and inequities even as the multipillar system is being redesigned to respond to a rapidly aging environment. Demographic trends and rising life expectancy caused substantial deficits in the pay-as-you go (PAYG) system during the 1990s; replacement rates declined from over 75 percent to less than 50 percent. The 1998 pension reform aimed at scaling back the PAYG system to create fiscal space for introducing a second pillar (private mandatory) and a third pillar (private voluntary) to make up for declining PAYG benefits. Croatia launched the mandatory fully funded second pillar in 2002. However, with numerous revisions of the PAYG system between 2004 and 2011 fiscal pressures built up, prolonging transition to the multipillar system and leading individuals to opt out of the second pillar. The interventions also created significant inequities in the benefits of various pensioner cohorts. In October 2010, Parliament raised the female retirement age and the early retirement age, and introduced a modest late retirement bonus. These measures were not enough to assure long-run sustainability and compensate for the pension supplement extended to the PAYG-only pensioners of 27 percent. This supplement severely discriminated multipillar participants leading them to exit from the multipillar system. In 2013, a further increase in the retirement age to 67 was legislated, along with the relaxation of the early retirement rules and more generous indexation. At the same time, authorities launched a more serious convergence path of privileged pensions with the PAYG to address the equity concerns. 6.29 A relatively low active coverage rate is one of the causes for Croatia’s unfavorable system dependency rate—only 50 percent of Croatians aged 15 to 65 contributes (Figure 54). Among the several reasons for low coverage are:  Low formal labor participation, with the labor force less than 55 percent of the prime age population, is among the lowest in the EU.  A large shadow economy, estimated at 20–25 percent of GDP, contributes to both low labor participation and low coverage rates.  Rising unemployment, especially for the youth, has been depressing the active coverage rate.  Early exit from the labor force, either through early retirement or disability, affects the coverage. Figure 54. Contributor Coverage, Croatia, High-Income and Transition Countries Contributors as % of Working Age Contributors as % of Working Age Population - Developed and Croatia Population - Transition 100% 100% 90% 90% 80% 80% 70% 70% 60% 60% 50% 50% 40% 30% 40% 20% 30% 10% 20% 0% Slovenia Latvia Serbia Turkey Poland Armenia Moldova Lithuania Bulgaria Bosnia Republika Srpska Hungary Azerbaijan Georgia Belarus Albania Montenegro Croatia Czech Republic Macedonia, FYR Romania Slovak Republic Bosnia and Herzegovina Estonia Ukraine Uzbekistan Kazakhstan Russian Federation Kyrgyz Republic 10% 0% Greece Luxembourg Austria Netherlands Ireland Croatia United Kingdom Belgium Israel Sweden Japan Italy Iceland Denmark Australia France Finland Canada Switzerland Germany Portugal Spain Norway New Zealand United States Source: CROSTAT, EUROSTAT, staff calculations 6.30 Beneficiary coverage, or the ratio of pensioners to population was 27 percent, similar to other transition economies of similar income and population age but higher than in many older OECD countries (Figure 55). There are a number of reasons for this, among them: (i) a low effective retirement age; (ii) pervasive and abundant early retirement programs; 84 (iii) easily obtainable disability certification; and, most of all, (iv) almost 100 percent active coverage in the pre-independence working cohorts currently retiring or soon to retire. Figure 55. Beneficiary Coverage, Croatia and High Income and Transition Countries Pensioners as % of Total Population, High Income Pensioners as % of Total Population, Transition Countries and Croatia Countries 40% 40% 35% 35% 30% 30% 25% 25% 20% 20% 15% 15% 10% 10% 5% 5% 0% 0% Slovenia Latvia Moldova Turkey Serbia Poland Armenia Bulgaria Lithuania Czech Republic Azerbaijan Hungary Albania Belarus Bosnia Republika Srpska Croatia Montenegro Russian Federation Romania Kazakhstan Macedonia, FYR Ukraine Tajikistan Estonia Kyrgyz Republic Uzbekistan Slovak Republic Bosnia and Herzegovina France Luxembourg Portugal Greece Norway Spain Ireland Finland Japan Austria Germany Sweden Croatia Canada Denmark New Zealand United States Netherlands Italy Belgium Source: CROSTAT, EUROSTAT, staff calculations 6.31 The retirement age has not been adequately adjusted to capture rising life expectancy. With life expectancy at retirement now 14.5 years for men (who retire at 65) and 21.8 for women (who in 2013 retired at 61 year), Croatia is at the higher end of new EU countries (Figure 56). In 2010 the female retirement age was increased to reach 65 in 3-month-a- year increments. Yet by 2030, when the transition to the statutory retirement age of 65 is complete, female life expectancy would already have increased by two to three years, annulling half of the retirement age increase. And recent studies suggest that life expectancy may be accelerating. If so, retirement age for women needs to rise faster and start increasing for men before the two equalize. Figure 56. Life Expectancy at Retirement, 2010, Europe Panel A. Men Panel B. Women 24 28 22 26 24 20 22 18 20 16 18 16 14 14 12 12 Portugal (65) Switzerland (64) Poland (60) Italy (60) Spain (65) Ireland (66) Greece (60) Estonia (65) Austria (60) Romania (65) Hungary (65) Latvia (62) Croatia (61) Bulgaria (60) Slovenia (61) UK (65) France (60) Czech Republic (65) Lithuania (60) Netherlands (65) Sweeden (65) Slovakia (62) Montenegro (60) 10 Note: For Croatia data are for 2012. Source: World Bank Database, Eurostat. 6.32 Early retirement incentives are too high while incentives to postpone retirement too low (Box 17). The progressive early retirement decrement of 1.8 percent a year for service of 40 years is too low––below the actuarially neutral level of 3–4 percent estimated for Croatia and the level in many other European countries. It is unlikely to encourage employment in older age. In EU countries, this increment exceeds 3 percent a year. Furthermore, early retirement by five years is at the longer end of most OECD and EU countries and could be significantly shortened. 85 Box 17. Basic Parameters of the Croatian Pension System, 2013/14 Retirement age/vesting period: i. Old-age pension (2013): Men 65/15; Women 61 years (rising to 65 by 2030)/15 ; ii. Early retirement pension (2013): Men 60/35; Women 56 (rising to 60 by 2030/30 (35 by 2030). Calculation period for old-age: Life-time average earnings. Pension determination For those in PAYG only: point formula. For those in both pillars: i. Pre-2002 years of service: point formula; ii. Post-2002 years of service: Basic pension + second pillar annuity. Indexation: Twice a year, 70 percent of the higher of CPI/wage rate in previous 6 months + 30 percent CPI/wage rate of lower of in 6 months before that. Early pension decrement: 1.8–4 percent a year, depending on years of service; decrement is permanent. No decrement for those with 40 years of service Late retirement bonus: 0.15 percent for each month of late retirement. Minimum pension: 0.825 percent of 1999 gross wage for each year of service, indexed same as point value. Maximum pension: maximum average annual points of 3.8. Contribution rate: 20 percent of gross wage; for those in both PAYG and second pillar, 15 percent goes to first pillar, 5 percent on individual account in second pillar. Minimum contribution base: 35 percent of average wage in previous year. Maximum monthly contribution base: 6 monthly average gross wages in Croatia. 6.33 The pension contribution rate of 20 percent is among the lowest in the region, though the social insurance contribution rate as a whole is among the highest (Figure 57). This is mostly because health insurance contributions are exceptionally high (15 percent). Given already high labor cost in Croatia and the deficit in the health system, increasing contributions for pensions and reducing those for health would require major reforms in the health system (see the section on health care system). Figure 57. International Comparisons of Social Insurance Contribution Rates 50 45 40 35 30 25 20 15 10 5 0 Slovenia Belgium Turkey Sweden Luxembourg Serbia France Iceland Bulgaria Azerbaijan Lithuania Latvia Moldova Poland Canada Armenia Kosovo Hungary Spain Portugal Belarus Greece Germany Austria Albania Georgia Finland Ireland Netherlands Norway Croatia United States Turkmenistan Estonia Russian Federation Czech Republic Macedonia, FYR Japan Ukraine Slovak Republic Romania United Kingdom Italy Kazakhstan Switzerland Uzbekistan Kyrgyz Republic Pension Other social insurance Source: World Bank database. Equity concerns 6.34 The Croatian pension formula and the minimum pension are highly redistributive. The PAYG system awards pension points for each year based on the individual’s earned income. 86 The average wage earns one point, which was 0.8 percent of the average gross wage at the end of 2012. The minimum pension per year of service is 0.825 percent of the 1998 average gross wage indexed annually (as pension point value is), which because of indexation was 0.74 percent of average gross wage in 2012. Minimum wage earners with 40 years of service would earn a minimum pension higher than their last net wage (the net individual replacement rate is over 100 percent); and average wage earners with 40 years of service would receive a pension benefit about 3 percent higher. On the other hand, the maximum pension is limited to 3.8 annual points, i.e., the pension formula is proportional up to 3.8 times the average wage, while contributions are proportional up to 6 times the average wage. As a result, the net individual replacement rate for someone earning 6 times the average wage would be around 30 percent. 6.35 The pension benefit formula undervalues historical earnings. The initial old-age pension benefit is obtained by multiplying the sum of personal points for all years of service by the point value, which is adjusted by the modified mean of inflation and the wage rate (the modified Swiss formula, see Box 17), as are pension payments. Swiss valorization and indexation were introduced to prevent large differences in benefits for new and old pensioners. Most OECD countries and some European and Central Asian countries have introduced 100 percent wage valorization and 100 percent price indexation. This is considered international best practice because it preserves the purchasing power of the elderly, it yields both fiscally and socially sustainable benefits; and it strengthens the link between lifetime contributions and benefits. 6.36 Large cross-cohort pension differences have emerged because of ad hoc pension supplements, which have affected the balance of the pension pillars. Targeted supplements have created large cross-cohort differences and triggered discussions about new supplements. In July 2007 the Law on Pension Supplement was passed to alleviate differences in pensions caused by previous supplements. Pensions of those who retired in 1999 were raised by 4 percent and from 2010 onwards the pensions of PAYG-only retirees increased by 27 percent. Multipillar participants were left out, although they worked most of their service in the PAYG-only system. Instead of extending the supplement to the PAYG-only service of multipillar participants or withdrawing it from all participants, the Government allowed multipillar participants to return to PAYG-only system and transfer their second pillar accumulations in exchange for the full PAYG benefit. 6.37 Since 2011 multipillar participants have been returning to the PAYG pillar only. In the vast majority of cases, the calculated second pillar annuity has been insufficient to compensate for the loss of the 27 percent PAYG supplement. More than 4,000 multipillar retirees have returned since 2011 and only those remained who had large annuities and few PAYG-only years of service. Such a drive-through second pillar brings benefits only to second pillar fund managers who collect fees for pension savings that are transferred to the PAYG pillar, while their clients' savings do not materialize in the appropriate annuity level. Because the 2007 supplement discriminates against multipillar participants, addressing it should be a pension policy priority. 6.38 More than a dozen special and privileged pension schemes have emerged in the last two decades, fiscally exhausting the pension system and creating serious inequities. Out of the 2 percent of GDP distributed for privileged pensions, the largest pension program (1.5 percent of GDP in 2013) was for Homeland War Veterans (HWV) disability and survivor pensions. The average HWV disability and survivors benefit was 2.4 times higher than the old- 87 age pension, yet the average age of beneficiaries is only 51. Furthermore, the minimum initial pension for non-disabled HWVs who served more than 100 days at 45 percent of the average net wage in Croatia, regardless of length of service, has reduced incentives for active HWVs to participate in formal employment and report full incomes. The minimum pension granted to HWVs (so far around 500) with an average of 5 years and 9 months of service (insufficient contributory period for a regular pension), was 10 percent above the average pension earned with more than 33 years of service. Partial reforms to military and police pension systems were launched early in the 2000s to extend the vesting period to 35 service years for men and 30 for women, raising the retirement age to the regular retirement age, and widening the calculation period to the 10 best-earning years. The number of pension beneficiaries who participated in the World War II or had been political prisoners is declining gradually. In 2012, privileged pension benefits for members of Parliament, government officials, and constitutional court judges were abolished, and the list of internal affairs personnel subject to accelerated service periods was trimmed to cover only hazardous positions. On the other hand, nothing was done about the largest privileged group, HWVs. In December 2012, the Government published the HWV Registry, believing that would reveal fraudulent cases and help withdraw their pensions. However, that mechanism has not brought any savings so far. Long-run sustainability concerns 6.39 Returns on second pillar funds are fluctuating but exceeding wage growth. By 2013 second pillar membership reached 1.65 million and its assets reached 16 percent of GDP, with pension funds becoming the largest non-bank institutional investor in the capital market. In its first six years (2002–07), the rates of return on second-pillar funds significantly outgrew average wage growth (Table 30). In 2008 second pillar accounts lost 12.5 percent on average, while wages continued to grow. On the contrary, by 2013, the real rate of return stood at 4 percent, outgrowing the average real gross wage increase by less than one percent which strengthens the argument that the second pillar gives higher value-for-money than the first pillar. Table 30. Annualized Net Rates of Return (RoR) of Mandatory Second Pillar Year/Period Annualized RoR net of Management Fees Nominal Gross Wage Rate (percent) (percent) 2002 13.2 6.0 2003 7.4 4.8 2004 7.4 6.3 2005 7.1 4.4 2006 5.7 6.2 2007 6.8 6.2 2008 –12.5 7.0 2009 8.7 2.2 2010 8.6 -0.4 2011 0.5 1.5 2012 12.7 1.0 2013 4.5 0.8 2013/2002 (annual rate) 7.5 3.8 Source: Croatian Agency for Financial Sector Supervision (HANFA) and CROSTAT. 6.40 Cost of administering the second pillar has dropped considerably. Pension fund management companies charge: (i) a maximum 0.8 percent front-end fee on paid-in 88 contributions; (ii) a maximum annual fee of 0.45 percent of total asset value; and (iii) a switching or exit fee. Current fees are among the lowest for the second pillar internationally. Furthermore, since 2011 pension funds have been required to remunerate REGOS for individual account management service (HRK30 per account annually). To engage pension fund management companies in supervising account management costs, consideration could be given to having those companies participate in REGOS ownership structure and decision-making. That has had positive results in countries as diverse as Estonia and Mexico. 6.41 Numerous interventions in the PAYG system have caused pension reform to diverge from its initial course, eroding the system’s fiscal sustainability. That has halted progress on the second pillar, which was intended to support future pension growth as the society ages. With first-pillar PAYG expenditures Figure 58. Total PAYG Pension Expenditures, percent of GDP having risen for the last three years 11.0 (to 10.9 percent of GDP in 2013, Figure 58), Croatia’s current 10.5 pension system must deal simultaneously with worsening 10.0 demographic ratios, low labor participation rates, low and 9.5 declining replacement rates, inadequate pensions for multipillar 9.0 cohorts, large cross-cohort differences in pensions, and overly 8.5 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 generous privileged pension schemes. Source: Ministry of Finance, staff calculations 6.42 Though in the medium and long run the Croatian pension system appears fiscally sustainable, it is socially unsustainable because of its numerous inequities. Unless there are further reforms,67 average replacement rates would continue to decline along with PAYG expenditures, so that the PAYG balance would improve significantly, and implicit pension debt would decline—but the structural problems of the entire pension system would worsen. The most prominent are: (i) cross-cohort differences in pension benefits; (ii) a long-run deterioration of replacement rates, making future pensions inadequate; (iii) redistribution toward special occupations and low-income earners; and (iv) low second pillar annuities because of its low contribution. The Government’s Reform Plans and Their Likely Impact 6.43 The government plans to fiscally consolidate the pension system, improve social equity by converging privileged pensions to the general system, and make progress with all three pillars simultaneously. At the beginning of its term, the current government abolished privileged pensions for government officials and MPs and in March 2012 resumed pension indexation, which had been suspended during 2010 and 2011. However, it has not yet undertaken short-term measures to consolidate the pension system and to address its medium- and long-term issues. As a consequence, the pension spending to GDP ratio was again high in 2012-13 and is likely to rise further in 2014 to 11 percent of GDP. Putting pension spending on the downward path in a stagnant economic environment will require forceful savings measures, such as raising 67 World Bank 2011, Croatia: Policy Options for Further Pension System Reform, World Bank, July 2011 89 the retirement age faster, tightening early retirement requirements, decisively reducing privileged pensions, decreasing or means testing minimum pensions, awarding pensions only for contributions paid, and restricting eligibility for disability pensions. 6.44 Effective from January 2014, the Government introduced the following measures:  Raised the retirement age for both genders to 67 by 2038 years (by 3 months a year to offset population aging and gradually increase the effective retirement age.  Tightened the early retirement decrement for service from 37 to 40 years and eliminated it for service of 40 or more years.68 Although minor, increasing the decrement toward a more actuarially neutral level is useful. However, eliminating the decrement for service beyond 40 years stimulates early exit regardless of age. It discriminates against those who enter service after attaining more education or after a period of childcare and contradicts the decrement automatically charged in the second pillar regardless of age.69 Actuarially unbalanced decrement in the first pillar distorts the balance with the other two pillars. According to the Pension Institute, this measure would result in net fiscal annual savings only about 0.05 percent of GDP over the next two decades.  Pension indexation with 70 percent of the higher of wage-or-price-inflation and 30 percent of the lower of wage-or-price-inflation. This indexation pattern, also applied to valorize past incomes, aims to improve pension adequacy regardless of economic environment. It increases real pensions at 70 percent of the real wage rate when real wages grow and prevents a drop when real wages decline. Such a protective pattern is not applied in any other European country. This measure would introduce extreme uncertainty in planning pension expenditures, generate large implicit liabilities, and distort the balance between pillars in favor of PAYG. As shown later, it would also increase PAYG expenditures by an annual average of 1 percent for the next 50 years and reduce fiscal room for further development of the second pillar.  Separated “earned” from “privileged” pension with plans to introduce a separate indexation rule for privileged ones. The intent is to freeze the privileged part of pensions for all categories until real GDP reaches 2.5 percent, which could yield annual savings of 0.03 percent of GDP by 2030. However, the alternative indexation rule should not be subject to discretional changes across political cycles.  Reduced privileged pensions higher than HRK5,000 by 10 percent;70  Tightened the system of early retirement with an accelerated service period for some occupations. In August 2012 the eligibility list for military staff was reduced; the reform stopped with that group of privileged cohort.  Established a central disability assessment body that is charged with disability assessment according to the remaining work capacity, and recertification of disability pensions each 3 years.  Further consolidated pension system administrative costs by restructuring HZMO (Croatian Pension Insurance Institute)-the PAYG administrator. 68 For example, early retirement with 39 years of service increased from 2.9 to 3.6 percent a year. 69 The current second-pillar annual decrement is about 4 percent for every year of early retirement. 70 The reduction is automatically revoked when real GDP growth exceeds 2 percent and budget deficit shrinks below 3 percent. HWVs with 100 percent disability and HWV's survivor children were excluded. 90  Transferred second pension pillar assets of individuals covered by the special early retirement (through accelerated service period) scheme for internal affairs personnel and the Croatian Army members. rently eligible for extended service period and those with more than 20 years of service in such occupations.71 6.45 The Government further announced long-term policy measures to be implemented in the next stage of pension reform:  A PAYG pension supplement of 27 percent would be extended to all beneficiaries in proportion to their PAYG service;72 and  The list of hazardous occupations eligible for accelerated early retirement would be revised and the early retirement concept reformed from 2015. 6.46 The pension policy measures adopted are a mix of measures in both right and wrong directions. Differentiation between regular and special/privileged pensions by indexation rules in line with different sources of financing initiated the needed convergence of privileged pensions to PAYG. However, a more generous new indexation for PAYG (“earned”) pensions distributes a larger share of growth to current pensioners, costs more and is pro-cyclical. Privileged pensions, set to be frozen until economic conditions improve, will be indexed in parallel with generously indexed regular pensions, slowing their convergence to regular in the long run. A reduction of some 35,000 highest privileged pensions (HWVs, MPs, military) by 10 percent would also support the benefit convergence. 6.47 An increase in the retirement age is a natural policy response to aging population, but the transition period is too long (by 2038). The adjustment of basic pension is recommended to reduce the discrepancy between PAYG and the second pension pillar benefits. However, its implementation without a simultaneous increase in the second pillar contribution rate boosts PAYG costs in the medium and long run and requires further savings measures. Measures that reduce the penalty for early retirement go in the wrong direction, contradict the objectives to achieve a closer link between benefits and contributions and goal of raising among the lowest in Europe participation rate. However, they are not expected to generate significant costs. 6.48 A disability assessment to identify the remaining work capacity and the recertification of disability pensions are positive long-run measures which aim to reduce the provision of disability pensions to those capable to work, and reduce the error and fraud in the disability assessment system. A transfer of second pension pillar assets of insured individuals in occupations currently eligible for an extended service period is a significant move in the wrong direction. Instead of continuing to divert only the new pensioners with an extended service period back to the first pillar until the new law on hazardous occupations’ service period is adopted, the law opted for a transfer of almost all assets of currently insured individuals back to the first pillar regardless of their age or service period. This measure will have a negative impact on the labor market, create unequal social insurance treatment between occupations, lock-in the occupations with an extended service period, and delay the reform of the early retirement of hazardous occupations. An extension of 27-percent supplement would eliminate the 71 The second pillar assets of other hazardous and hard-condition occupations are scheduled for transfer in 2015. 72 The supplement raises PAYG pensions by 27 percent only of those that do not participate in the second pillar and discriminates against the second pillar participants. 91 discrimination of the second pillar participants and eliminate their exit from the second pillar at retirement due to inequity created in the past. However, extending the 27-percent supplement to all would require additional 1-2 percent of GDP per year that could only be found in the stronger PAYG reform. 6.49 The government does not envisage tightening the early retirement over the medium term. The early retirement period remains to be 5 years below regular retirement age with 35 years of service for both genders from 203073. The early retirement decrement remains largely unchanged, below 4 percent per year, depending on the service period (Table 31). The late retirement bonus remained the same–1.8 percent per year of late retirement–which remains low and unattractive. Table 31. Early Retirement Decrement, Old and New Early retirement decrement – Decrement per year of early Decrement per year of early years of service retirement - Old PIL retirement - December 2013 Law 35 0.34 0.34 36 0.34 0.32 37 0.29 0.30 38 0.24 0.25 39 0.19 0.15 40 0.15 0.10 41 0.15 0 Early retirement decrement for 0.15-0.34 0 unemployed more than 2 years (and (scheme above) eligible for early retirement) Source: Official Gazette, www.nn.hr 6.50 The pension reform package does little to improve the system in the short run and only improves it marginally in the long run through the rise in retirement age. In the short run the measures are too weak to reduce PAYG expenditures and the pension deficit. The inequities between privileged and regular pensions and especially between PAYG-only and multi-pillar pensions although reduced, have remained embedded in the system. In the long run, the package moderately raises average PAYG replacement rates and improves the poor adequacy perspectives, but at the cost of significantly higher PAYG expenditures and deficit (Table 32). 73 Currently the service period requirement for early retirement for women is transitioning from 30 to 35, currently standing at 31 years. 92 Table 32. Fiscal Impact and Replacement Rate—Simulation Results 2013 2014 2015 2016 2017 2023 2033 2043 2053 2063 2073 PAYG pension expenditures, %GDP Baseline 10.5 10.6 10.6 10.5 10.3 9.5 7.7 6.2 5.5 5.1 4.7 Basline + indexation 10.5 10.6 10.7 10.6 10.5 9.9 8.2 6.5 5.9 5.6 5.3 Baseline + retirement age 10.5 10.6 10.6 10.5 10.3 9.5 7.6 5.9 5.0 4.6 4.2 Baseline + basic pension 10.5 10.6 10.7 10.5 10.3 9.6 8.4 7.7 7.7 7.4 6.9 Baseline + privileged 10% cut 10.5 10.5 10.5 10.4 10.3 9.5 7.7 6.2 5.5 5.1 4.7 Baseline + freeze of privileged pension part 10.5 10.5 10.4 10.3 10.1 9.4 7.7 6.2 5.5 5.1 4.7 Disability pensions reform* 10.5 10.6 10.5 10.3 10.0 9.1 7.2 5.6 4.9 4.4 4.0 Transfer of extended service period from pilar 1 to 2 10.5 10.6 10.6 10.5 10.3 9.7 8.0 6.5 5.8 5.4 5.1 All measures 10.5 10.6 10.7 10.5 10.4 9.9 8.7 7.2 7.1 6.9 6.6 Overall PAYG deficit, %GDP Baseline -4.0% -3.9% -3.9% -3.8% -3.7% -3.2% -1.7% -0.2% 0.3% 0.7% 1.3% Basline + indexation -4.0% -4.0% -4.0% -3.9% -3.9% -3.6% -2.1% -0.5% -0.1% 0.2% 0.7% Baseline + retirement age -4.0% -3.9% -3.9% -3.8% -3.7% -3.2% -1.5% 0.5% 1.2% 1.5% 2.0% Baseline + basic pension -4.0% -4.0% -3.9% -3.8% -3.7% -3.3% -2.3% -1.6% -1.9% -1.6% -0.9% Baseline + privileged 10% cut -4.0% -3.8% -3.8% -3.7% -3.7% -3.2% -1.7% -0.2% 0.3% 0.7% 1.3% Baseline + freeze of privileged pension part -4.0% -3.8% -3.7% -3.6% -3.5% -3.2% -1.7% -0.2% 0.3% 0.7% 1.3% Disability pensions reform* -4.0% -3.9% -3.8% -3.6% -3.4% -2.8% -1.2% 0.4% 0.9% 1.4% 2.0% Transfer of extended service period from pilar 1 to 2 -4.0% -3.8% -3.8% -3.7% -3.6% -3.1% -1.6% -0.2% 0.3% 0.7% 1.3% All measures -4.0% -3.9% -3.8% -3.8% -3.7% -3.2% -2.1% -0.6% -0.6% -0.5% -0.1% Average PAYG replacement rate, women Baseline 30.3% 29.7% 29.0% 28.5% 28.0% 25.3% 20.4% 15.9% 13.2% 12.0% 11.4% Basline + indexation 30.3% 29.8% 29.3% 28.9% 28.6% 26.7% 22.1% 17.4% 14.5% 13.4% 12.9% Baseline + retirement age 30.3% 29.7% 29.0% 28.5% 28.0% 25.3% 20.4% 15.4% 12.1% 10.7% 10.2% Baseline + basic pension 30.3% 29.7% 29.1% 28.6% 28.2% 25.9% 22.6% 20.2% 18.7% 17.7% 17.0% Baseline + privileged 10% cut 30.3% 29.7% 29.0% 28.5% 28.0% 25.3% 20.4% 15.9% 13.2% 12.0% 11.4% Baseline + freeze of privileged pension part 30.3% 29.5% 28.8% 28.2% 27.5% 25.0% 20.4% 15.9% 13.2% 12.0% 11.4% Disability pensions reform* 30.3% 29.7% 28.7% 28.0% 27.4% 24.4% 19.0% 14.3% 12.1% 10.5% 10.0% Transfer of extended service period from pilar 1 to 2 30.3% 29.7% 29.0% 28.5% 28.0% 25.3% 20.4% 15.9% 13.2% 12.0% 11.4% All measures 30.3% 29.7% 29.3% 28.9% 28.4% 26.9% 24.4% 20.0% 18.1% 17.0% 16.5% * Guesstimate, the disability certification reform details are not y et know n and legislated Recommendations 6.51 At 11 percent of GDP spent on pensions, further aging pressures would further deteriorate sustainability of Croatia’s pension system. Low and worsening system dependency rate due to lax early retirement rules and disability options, as well as a large pool of privileged pensioners continue to exacerbate the problem. At the same time, adequacy of multi- pillar pension beneficiaries continues to be of concern due to constant delays of strengthening the second-pillar contributions. 6.52 Government would need to ensure a broad consensus on the multi-pillar system and finally create the appropriate balance between the first and the second pillar. Several options could be considered to respond to these quests:  Consider gradually raising the second pillar contribution rate to 10 percent. Increase in the second pillar contribution rate would raise future replacement rates, and reduce future PAYG costs and implicit pension debt. However, it would require additional fiscal space today, created through a more decisive PAYG saving. Gradually raising the second pillar contribution rate from the current 5 percent to 10 percent by 2016 and reducing the PAYG rate at the same scale would improve replacement rates for future generations but substantially increase the PAYG deficit. Budget transfers for lost revenues are expected to reach 1.7 percent of GDP a year in 2016 and stay there from then on. Due to declining 93 contributions, without mandatory budget transfers the PAYG balance will not be restored even by 2070.  Consider accelerating the retirement age increase before 2030 and tightening and phasing out the early retirement. This could bring around 0.3 percent of GDP in savings already five years after the full introduction.  Rationalize the categories of privileged pensions and accelerate convergence of privileged pensions to PAYG by equalizing the maximum privileged pension with the old-age maximum pension. This would improve the equity of the pension system. If the maximum pension converges to the general maximum level and the minimum privileged pension of 45 percent of the net average wage is abolished pension expenditures would be progressively reduced by about 0.03 percent of GDP a year in 2017 to 0.44 percent in 2038.  Use means-testing for granting minimum pensions and award pension points only for periods with contributions paid. A reduction in the minimum pension by 50 percent would separate the minimum from the average pension and thus reduce incentives to evade contribution payment. It would lower the PAYG deficit by up to 0.4 percent of GDP and implicit pension debt by 12 percent of GDP, but because it would also lower the average old age replacement rate by about 1 percentage point, it might exacerbate poverty.  Revisit the pension indexation. An alternative is application of 100 percent wage valorization and 100 percent price indexation, as is done in most OECD countries. This approach is considered international best practice because: (a) it preserves the purchasing power of the elderly as they age; (b) in the long run it yields both fiscal sustainability and socially sustainable benefits; and (c) it tightens the link between contributions and benefits. Such a policy generates a fiscally more sustainable system; whereby annual fiscal savings would reach 0.8 percent of GDP by 2020. C. Long-Term Care 6.53 Croatia will soon experience sudden large expansion of its elderly population and thus growing long-term care (LTC) costs. First, a significant expansion of the 65-74 age group will happen over the next 10 years, with annual growth rates of up to 4 percent for this particular age group. This significant expansion is due to the retirement of the large baby boomer generation. During the 2020s, as the same cohort ages further, Croatia will experience a strong expansion of the 75+ age group. After that, the next strong expansion will take place during the 2040s, when the relatively young Generation X will start to retire. Throughout the next 40 years (and beyond), the younger age groups—in particular the working age population—will continuously decline. As Croatian population grows older, more and more people will become dependent on the support from others for their activities of daily living. At the same time, less and less healthy people—potential caregivers—will be available. In addition, the working age population will decrease while the population in retirement age will increase. 6.54 The demographic transition is likely to cause considerable fiscal pressures: public spending on LTC will surge in future in three ways: First, public spending per LTC beneficiary is likely to increase faster than GDP per capita as standards of living increase. Second, because the share of the dependent population that actually demands LTC services is currently low in most Eastern European countries, over time the dependent population is more 94 likely to demand them. Third, as already discussed the dependent population itself will increase considerably. The net effect is strong growth in public expenditures on LTC. Also, financing LTC will become increasingly unsustainable because LTC is mainly financed through pay-as- you-go systems. 6.55 At the moment, Croatia’s severely dependent74 population is growing by more than one percent annually. The annual growth rate is expected to decrease for the next 10 years, then again turn upward. The severely dependent population is forecast to grow from about 300,000 today to 350,000 in 2050 (+15 percent). Meanwhile, the healthy population is expected to fall from 2.8 million to 2.3 million (–19 percent). What all this means is that between today and 2060, the old-age dependency ratio will double, from about 26 percent today to more than 50 percent (in some countries even more). 6.56 LTC in Croatia is provided by a variety of formal and informal institutions. These include homes for the elderly and infirm, homes for the disabled, and homes for people with mental illnesses. They can be state-owned and private or run by NGOs and religious organizations. The capacity of these institutions has been increasing in line with demand. The main non-institutional social services for the elderly are home care (both nursing and housekeeping assistance), day care in local institutions, transportation, and the availability of 24- hour assistance to persons living at home. Also, the Ministry of Social Policy and Youth (MoSPY) is creating a national network of county and community gerontological centers to establish a network for non-institutional care for the elderly, in consultation with family physicians, to monitor and coordinate all activities related to the needs of elderly. Centers for nursing and care also provide non-institutional LTC, including day-care and home care for the elderly and infirm. These centers may be founded by county and local governments, private companies, NGOs, in particular Caritas and the Red Cross. Foster care is a form of community non-institutional social care for the elderly and infirm, particularly targeted persons without family, home, or income. Finally, there are informal caregivers for the elderly, including family, friends and neighbors although data are scant on this category of providers. 6.57 Two questions arise: Who will provide care for the increasing dependent population in the future? And how much and who will pay for the increased cost of this care? The answer to the first question is provided above––Croatia does have basic infrastructure for the provision of long-term care. The issue is to improve its efficiency and prepare for the rising demand and pressures on the system. To understand the implications of the second question, it is useful to look at inverse dependency ratios (Figure 59). While today, there are more than nine healthy people—nine potential care providers—for each severely dependent person, by 2050 there will be only six. And there will be fewer and fewer people of working age and more and more people who are retired. While today, there are four people of working age—four potential taxpayers—for each person of retirement age, by 2050 there will be only two. In other words, the burden on young and healthy people to care and pay for dependent people will be significantly heavier. 74 The severely dependent population is defined as people who need significant support from others in their activities of daily living (dressing, personal hygiene, cleaning, shopping, food preparation, legal matters, mobility, and participation in social activities). It mainly comprises mentally and physically disabled, dement, and frail elderly people, but can also include convalescent people and people in need of rehabilitation who are in need of LTC services for a limited period of time, not permanently. 95 6.58 The question of how much and who will pay for the increased care needed is particularly grave for Figure 59. Inverse Dependency Ratios, Croatia, 2011–50 Croatia. Like the rest of Eastern Europe, Croatia will grow old before it grows rich. Western countries were able to fully take advantage of low dependency ratios when the demographic transition began and over the last 60 years because large population cohorts were not burdened with high numbers of elderly and children. They were able to accumulate significant wealth, which they can now use to pay for their retirement and Source: Koettl (2010) care needs. Meanwhile, countries in Eastern Europe, including Croatia, were deprived of a demographic dividend because for much of time their economies and productivity levels were severely depressed. Not only do Croatia’s largest population cohorts retire earlier than in Western countries, they also retire with less wealth, and there is not a lot of time left for the country to accumulate a significant amount of wealth. LTC Trends in Croatia 6.59 There was a noticeable increase in institutional capacity for providing LTC , as is evidenced by a 29- percent rise in the total number of beneficiaries and a 29 percent rise in the number of institutions for the disabled, the elderly, and people with mental illnesses. Generally, the elderly and infirm made up 84 percent of all beneficiaries in 2012, a proportion that has risen since 2003; the percentage of the disabled receiving institutional care services similarly rose to 26 percent (from 20.7 percent) and of persons with mental illnesses, fell slightly to below 16 percent in 2012 (Table 33). Table 33. Number of Institutional LTC Beneficiaries 2003 2008 2009 2010 2011 2012 Homes for people with disabilty 4,055 4,822 4,673 5,643 5,805 6,151 - state 3,086 3,656 3,543 4,369 4,513 4,692 - non-state 969 1,166 1,130 1,274 1,292 1,459 Homes for elderly and infirm persons 11,794 14,150 14,688 14,919 15,396 15,291 - state 78 177 168 167 - non state - county 9,965 10,327 10,502 10,532 10,700 10,574 - other non-state 1,829 3,823 4,108 4,210 4,528 4,550 Homes for people with menatal illnesses 3,784 3,882 4,027 4,264 3,999 3,857 - state 3,590 3,168 3,222 3,292 3,019 3,008 - non-state 194 714 805 972 980 849 TOTAL 19,633 22,854 23,388 24,826 25,200 25,299 Source: MoSPY; various Annual statistical report 6.60 The other notable trend is the rise in the number of institutions not run by the national government, which has grown by 36 percent, compared to a rise of 14 percent for 96 national institutions (Table 34). In 2012, more than two-thirds of all beneficiaries of institutional LTC for the three main categories of services were accommodated in facilities managed by regional or local governments and private organizations rather than national. Table 34. LTC Institutional Facilities by Type of Management 2003 2008 2009 2010 2011 2012 Homes for people with disabilty 38 37 37 41 41 41 - state 24 26 26 28 28 28 - non-state 14 11 11 13 13 13 Homes for elderly and infirm persons 94 123 129 127 129 131 - state 1 2 2 2 - non state - county 46 47 46 46 45 45 - other non-state 48 76 82 79 82 84 Homes for people with menatal illnesses 21 27 27 28 28 27 - state 18 18 18 18 18 18 - non-state 3 9 9 10 10 9 TOTAL 153 187 193 196 198 199 Source: MoSPY; various Annual statistical report 6.61 More beneficiaries and institutions inevitably mean more employees (Table 35). Between 2003 and 2012 period total LTC employees increased by 29 percent; employees in national institutions by 12 percent, and in other types of institutions by 41 percent. The average indicator for national institutions is 2.4 beneficiaries per employee, and in other types 2.9. Table 35. LTC Employees by Type of Institution 2003 2008 2009 2010 2011 2012 Homes for people with disabilty 2,293 2,442 2,465 2,511 2,553 2,389 - state 1,817 1,967 1,976 2,063 2,110 1,939 - non-state 476 475 489 448 443 451 Homes for elderly and infirm persons 3,720 4,728 4,812 5,052 5,104 5,372 - state 20 52 53 51 - non state - county 2,923 3,248 3,289 3,431 3,465 3,521 - other non-state 797 1,481 1,503 1,569 1,586 1,800 Homes for people with menatal illnesses 1,120 1,396 1,443 1,505 1,512 1,463 - state 1,073 1,189 1,208 1,253 1,258 1,243 - non-state 47 207 235 252 254 220 TOTAL 7,133 8,566 8,720 9,067 9,169 9,224 Source: MoSPY; various Annual statistical report 6.62 The most obvious trend in funding of LTC services is the rise in the financial participation of beneficiaries. Funding for LTC services is based on the MoSPY classification; including full coverage by the state, full coverage by beneficiaries, state-beneficiary cost-sharing and coverage from “other sources” not specified (Table 36). In 2012, half of beneficiaries were able to fully finance their social service costs, compared to 37 percent in 2003. Funding that includes national state subsidies declined gradually from 51 percent in 2003 to 42 percent in 2012. 97 Table 36. Trends in LTC Funding, 2003 and 2011 2003 2012 beneficary + beneficary + Sources of funding beneficiary state state other beneficiary state state other Type of homes State 733 2,099 3,639 205 931 2,548 4,054 334 Homes for people with disabilty 4 972 2,069 41 98 1,519 2,930 145 Homes for elderly and infirm persons 104 34 6 23 Homes for people with menatal illnesses 729 1,127 1,570 164 729 995 1,118 166 Non-state 6,477 2,269 2,076 2,135 11,803 2,308 1,760 1,561 Homes for people with disabilty 22 122 818 7 28 625 572 234 Homes for elderly and infirm persons 6,430 2,084 1,164 2,116 11,611 1,347 883 1,283 Homes for people with menatal illnesses 25 63 94 12 164 336 305 44 Total beneficiaries 7,210 4,368 5,715 2,340 12,734 4,856 5,814 1,895 Total % 36.7% 22.2% 29.1% 11.9% 50.3% 19.2% 23.0% 7.5% Source: MoSPY; various Annual statistical report 6.63 Public spending on LTC will grow from the current 0.15 percent of GDP to about 1.3 percent of GDP in the medium variant with a range between 0.8 and 1.8 percentage points in other demographic projection variants (Table 37). This outcome is somewhat more optimistic than projections for the EU countries, where aging is projected to raise health and LTC expenditures from 1.7 to 3.2 percentage points of GDP (2.2 percentage points on average) between 2000 and 2050. Table 37. Projected Public Expenditures on Health and LTC Recommendations 6.64 The demographic transition is likely to cause considerable fiscal pressure: public spending on LTC will surge in future from the current 0.15 percent of GDP to about 1.3 percent of GDP in the medium variant. Croatians will have to start saving for their future LTC needs either through a fully-funded pillar within the social security system; or like in the case of France, private LTC insurance based on cash benefits. 6.65 Croatia has the basic infrastructure for LTC but it needs a comprehensive LTC plan. Such a plan would address coordination of services managed by different ministries and agencies, market incentives to encourage the private sector to provide more long-term social care in the community, a reliable costing system to ensure accountability for public funding, and a formal strategy for monitoring LTC social services. As part of this plan, Croatia needs to shift demand for LTC services from the health to the social sector because most of the needs will be in social, not health services, and this would come at a lower fiscal cost. 6.66 Expanding services to favor community-based over institutional care could be considered. Services like assisted living, day care, and home-based care allow recipients to continue living relatively independent lives and delay institutionalization. Community-based care complements informal care; both are more cost-efficient than institutional care and offer a wider range of options for those needing support. The potential to incentivize demand for and supply of 98 non-institutional care through the payment side, for example with lower co-pays in non- institutional services could be considered. 6.67 Care fragmentation needs to be decreased and coordination increased. Fragmentation of financing and managing LTC results in cost shifting and inadequate care. The government could begin with a joint assessment of needs by interdisciplinary teams —doctors and social workers-to identify an applicant’s degree of dependency and tailor medical and social services to the individual’s needs. Such an assessment would take into account the applicant’s family situation, the local availability of informal care and formal care services, and the applicant’s financial situation. 6.68 Croatia is well advised to explore a shift from the government producing LTC services to buying them from the private sector. Public sector cannot produce all of the care so government will have to level the playing field for public and private providers. Results-based financing that also extends to the private sector could help achieve this. But government has to define its core competencies, such as providing LTC in sparsely populated areas and to beneficiaries like the mentally ill, where there is little incentive for private providers. The government should license care and define and monitor care standards. 6.69 The potential of cash benefits and/or vouchers could be explored. Cash benefits put beneficiaries in control, able to buy the services they consider are most relevant to their own situation—including informal care or care from private providers. It might be a tool to support informal care—the most cost-efficient type—and spur private sector response. If there are concerns about the abuse of cash benefits, the potential of vouchers could be explored. 6.70 Given the increasing fiscal pressures, the government could benefit from a review of financing options. Croatia could not afford in the medium term to dedicate LTC insurance, but could consider carving it out of the currently exorbitant health insurance contribution rate. D. Social Assistance 6.71 Croatia’s generous social protection system relies on poorly targeted, categorical rather than needs-based benefits, which waste public resources and contributes to the persistence of poverty and social exclusion. At 3.8 percent of GDP, Croatia’s social assistance is costly. The largest share accrues to war veterans and their survivors and to families with children; well-targeted means-tested social assistance programs (guaranteed minimum benefit) account for less than 0.4 percent of GDP. As a result, poverty has risen along with social exclusion75. 6.72 Coverage of the poorest 20 percent is low. More than 60 percent of beneficiaries have been poor for more than three years and are considered socially excluded. Almost the same percentage of able-bodied individuals has been inactive. There is almost no integration of employment and social assistance policies, although some efforts have been made to integrate able-bodied beneficiaries into public works initiatives. Failure to coordinate local and national policies leads to overlaps and disincentives for re-entry into the labor market. Social reforms are also necessary to improve the labor force participation rate, among the lowest in Europe, and 75 Absolute poverty rate is estimated to grow from 13.3 percent in 2008 to 20.1 percent in 2013 using the poverty line fixed at 60% of median equivalent consumption in 2008. 99 make providers more accountable so that corruption, waste, and informal payments can be reduced. 6.73 Croatia spends a large share of GDP on non-contributory programs and the system is too complex and poorly targeted. After five years of recession, the demand for these programs is on rise, while the resources to finance them are increasingly scarce. Based on the broad definition of social assistance Croatia spent about 3.8 percent of GDP in 2013 on social assistance (Table 38). Table 38. Social Benefits Spending, 2004-2013 % of GDP 2004 2009 2012 2013 Social benefits 3.55 3.78 3.78 3.77 o/w child benefits 0.61 0.54 0.52 0.51 o/w child tax allowance1/ 1.00 1.00 1.00 1.00 o/w local government spending 0.32 0.47 0.44 0.43 o/w means-tested benefits 0.31 0.35 0.38 0.39 Means-tested benefits in total 8.75 9.24 10.00 10.34 1/ estimate based on staff simulations Source: MoF 6.74 Even using the narrower definition which focuses on traditional cash transfer programs for the poor, vulnerable and families with children, excluding the tax allowance–Croatia’s spending is still sizeable at about 2.4 percent of GDP. Also, the social assistance system has become more complex and more costly over time. The system includes a large number of cash and in-kind benefit for relatively small target groups. Recent efforts towards consolidation of benefits under the GMB are a move in the right direction but there is a large scope for the consolidation of other small entitlements. 6.75 During the crisis, the overall targeting accuracy of social assistance programs improved (Figure 60). While the targeting accuracy of social assistance in kind and cash, as well as pensions strengthened, the targeting accuracy of unemployment benefits and child allowance declined. The targeting accuracy of the last-resort program for the poor, the support allowance program, increased from 69 percent in 2008 to 77 percent in 2011. In the case of child allowance, the decline in the targeting accuracy was from 57 down to 50.6 percent for the bottom quintile. The error of inclusion of higher quintile groups in the child allowance program increased to 27 percent from 16 percent in 2008. 6.76 Croatia has a number of well-targeted social assistance programs whose performance ranges from average (child allowances) to very good (the social support allowance program). Results from the Household Budget Survey (HBS) 2011 indicate that 59 percent of total social assistance (Table 39) reaches the poorest 20 percent of the population, which is close to an average of the Europe and Central Asia countries. This result reflects in particular the targeting performance of the cash and in-kind poverty benefit programs (77 and 89 percent, respectively). The poverty benefit program (or GMB) has one of the best regional targeting results and has even improved over time. 100 Figure 60. Targeting Accuracy in the Bottom Quintile, 2008-2011 (percent of total benefit expenditures distributed to the poorest 20) Notes: Targeting accuracy or benefits' incidence is the transfer amount received by the group as a percent of total transfers received by the population. Source: Staff estimates based on the 2008–2011 HBS. Table 39. Targeting Accuracy of Social Protection Programs, 2011 Quintiles of consumption per adult equivalent, net of each social transfer Total Q1 Q2 Q3 Q4 Q5 Overall social protection 100.0 48.2 24.7 14.3 8.4 4.3 Social insurance 100.0 53.2 23.5 12.9 6.6 3.9 Old-age pension 100.0 65.0 16.7 10.3 5.4 2.7 Disability and survivors pension 100.0 64.7 16.4 8.8 5.0 5.2 Sickness benefit incl nursing, disability 100.0 67.3 6.5 8.8 11.8 5.6 Unemployment benefit 100.0 53.6 17.8 17.8 8.1 2.7 Social assistance programs 100.0 59.2 17.6 12.4 5.5 5.3 Social assistance in cash 100.0 76.7 7.8 6.6 4.9 4.0 SA in kind (food, firewood, clothes) 100.0 88.5 10.9 0.0 0.6 0.0 Family allowances (child allowance, 100.0 50.6 22.1 15.5 5.8 6.0 maternity leave, layette) Remittances and private transfers 100.0 32.3 15.5 16.9 15.9 19.4 Source: Estimates based on the 2011 HBS. 6.77 The child allowance program targeting, however, has deteriorated over time and has space for further improvement. It is targeted to a broader group of poor families with children than those in the poorest quintile, using a self-declared, imperfect income test. The targeting accuracy of family allowances (child benefit, maternity leave benefit, layette) suggests that slightly more than 50 percent of total allowances is received by the poorest 20 percent of the population, whereas 6 percent of total allowance is received by the richest 20 percent. The program accounts for 0.5 percent of GDP. However, to get the complete picture of family benefits, one also needs to look into the child tax allowance--a tax rebate offered for families with children that results in forgone taxes of about one percent of GDP. 101 6.78 The largest savings Figure 61. The Difference in Personal Income Tax Paid by Wage and potential for Earner With and Without Children (2012) reallocation of funds toward transparent social 2500 welfare programs could Average gross wage come from improving the 2000 equity effects of the child Difference (HRK) tax allowance. The largest 1500 delinquent among various social assistance programs 1000 in terms of targeting accuracy as well as 500 effectiveness is the child tax allowance program. This 0 11,200 12,400 14,200 15,400 18,400 10,000 10,600 11,800 13,000 13,600 14,800 16,000 16,600 17,200 17,800 19,000 19,600 5,200 6,400 8,200 9,400 3,400 4,000 4,600 5,800 7,000 7,600 8,800 MW=2,814 allowance is one of the most Gross monthly wage (HRK) generous allowances in the One child Two children Three children personal income tax (PIT). It is also one of the most Source: Staff calculations. regressive social assistance policies in effect, with only 10.8 percent of its “net benefit” going to the poorest quintile of the population. The poor and lower-middle income individuals do not earn enough income to qualify for or benefit fully from the child tax allowance. This program duplicates child allowance program, but by design is less transparent. It is a quasi-fiscal program, which absorbs (or forgoes tax revenues) the larger share of public funds among all social assistance programs (after war veterans’ benefits). The revenues foregone are assessed at 1 percent of GDP in 2012. Unlike the child allowance, the benefits of the child tax allowance are not transparently budgeted and provided to beneficiaries; thus the policy escapes parliamentarian oversight (Figure 61). Box 18. Reversing Low Fertility and Low Female Labor Force Participation International evidence on fertility dynamics has demonstrated that a discussion of pro-natalist policies must include an analysis of labor market dynamics, especially the participation of women. Croatia is faced with the double challenge of having an aging population that relies on a relatively small economically active working-age population; it has among the lowest female labor force participation rates–45 percent in 2013, a reduction from 63 percent in 1980. The fertility rate in Croatia has been steadily declining from around 2 children per woman in 1975 to 1.5 children in 2012, well below the replacement level of fertility (2.1). This combination of very low fertility with low participation in the labor market by women may reflect an incompatibility between motherhood and working that is also observed in the other lowest-low fertility countries (OECD, 2007). Figure below highlights one of the primary reasons why demographers and economists are increasingly focused on policies that can make work and child rearing compatible. In 1980 (left panel), proportionally more women participated in the labor market in countries such as the Nordic countries, but birth rates in these countries were lower. This is consistent with standard economic theories of fertility which emphasize the trade-off between time spent raising children and time spent working. However, while thirty years later (right panel) female labor force participation has increased in nearly all countries, some countries have succeeded in combining this greater participation with overall higher levels of fertility, while others, including Croatia, have seen fertility drop, sometimes well below replacement fertility. Because Croatia is experiencing both low fertility and relatively low female labor force participation, an evaluation of Croatia’s family policies should consider the likely impact on their pro-natalist objectives as well on the labor market participation, particularly of women. 102 Total Fertility Rate and Female Labor Force Participation (1980-2011) Source: Eurostat and World Bank staff calculations Basic economic theory suggests that family size will go down when the direct cost of children (education, clothing) increase, while it will go up if household incomes increase . Micro-economic theory also predicts that there is an inverse relationship between the family size and the female labor force participation due to indirect costs of having children, woman’s opportunity cost in terms of foregone employment earnings, but also leisure. Hence, a prediction of the theory is that family sizes will generally decrease when female education levels rise and women’s wages and labor market attachment increase. The relationship between the direct and indirect (opportunity) cost of children is modified by institutional factors such as social norms for higher desired family sizes, norms that prescribe greater participation by men in childcare duties and housework, but also technological advances that reduce housework. Greater participation in childcare by men and technological advances that reduce housework are also examples of factors that enable women to combine work with raising families. As such, these factors are not only expected to increase family size but also increase the female labor force participation. The opportunity cost of raising children is higher when women face barriers to (re-)entry into the labor market after a period of childcare following childbirth. Some studies have focused on a combination of high unemployment together with rigidities in the labor market in the mid 80's and 90's to explain the low fertility levels in Southern European countries. The opportunity cost of raising children is also higher when women (as well as men) face employment barriers to combine work with continued child care. Del Boca (2002) shows empirical evidence for Italy suggesting that the low availability of part-time employment coupled with a limited number of slots and non-compatible hours available in the Italian public child care system constitute two important factors that account for the low employment (and fertility) rates of married women. Also Apps and Rees (2005) and Del-Boca, Pasqua, and Pronzato (2009), show that the availability of subsidized child care is likely to benefit both female labor supply and fertility. 103 Pro-natal measures provided by Croatian government include financial support to a system of child benefits, birth grants, tax deductions for children, and maternity/parental leave. Including foregone revenues76 through the child tax deduction; Croatia spent approximately 1.9 percent of GDP (2012) on family benefits, comparable to other OECD countries. Croatia’s child benefits and birth grants policies may have a positive impact on the family size while reducing labor force participation in families with lower income women, but are unlikely to impact the family size or the labor force participation of higher income wo men. Croatia’s parental leave policies benefit working women by facilitating their re-entry. As such, this policy is expected to have a positive effect on both the family size and the participation of women in the labor market. The child-based tax deduction benefits higher income tax payers more, and thus benefits wealthier parents. As a result, the prediction with regards to its pro-natalists impact is ambiguous; while on the one hand this policy reduces the direct costs of having children, the wealthier parents who are most to benefit from this policy are also the least likely to be affected by financial incentives in their choice of family size. This policy also does not address the issue of compatibility between family and work. Croatia’s maternal leave policy falls under the category of pro-natalist policies, which aim to increase the compatibility between work and child rearing, reduce the mother’s opportunity cost of children, and thus may lead to a larger family size at higher levels of women’s participation in the labor market. Other examples of these kinds of policies include: availability of affordable childcare facilities, availability of part-time work opportunities, flexible work arrangements for women and men, primary school hours that are compatible with working hours, greater participation of men in childrearing and housework, improving re-entry into the labor market (e.g. low unemployment and more flexible hiring/firing) and parental leave arrangement. Croatia would benefit from a greater focus on policies that allow families to balance the work and family. The experience of other countries has shown that effectively addressing the compatibility between the work and family can enable countries to enjoy relatively high fertility levels and labor market participation by women. Croatia’s current mix of family policies provides lower -income families with direct financial incentives to have larger families; however with no results on increased fertility rates and at the expense of low labor force participation. Source: Based on Joost de Laat (World Bank 2010) and staff analysis 6.79 The coverage of social assistance programs in Croatia is relatively low. Only 16 percent of the poorest quintile receives the social assistance in cash (the support allowance) (Table 40). Around 5 percent of the poorest quintile has access to social assistance in kind, which is mostly delivered by local governments. Such a low coverage undermines achievements in terms of the targeting efficiency of social assistance (the support allowance or GMI) in cash and in kind. The coverage is slightly higher for family allowances, with more than a third of the poorest quintile being beneficiaries of the program. Sickness and disability benefits go to about 5 percent of the poorest quintile. Overall, around 20 percent of the population participates in some kind of social assistance programs. 76 World Bank staff estimates 104 Table 40. Coverage: Proportion of Population Receiving the Transfer, 2011 Quintiles of consumption per adult equivalent, net of each social transfer Total Q1 Q2 Q3 Q4 Q5 Overall social protection 67.3 99.9 92.5 72.6 49.0 22.4 Social insurance 52.7 96.4 74.4 50.1 27.7 15.0 Old-age pension 32.3 80.4 38.5 25.0 12.2 5.6 Disability and survivors pension 24.0 61.9 26.8 15.7 8.6 7.1 Sickness benefit incl nursing, disability 2.1 5.2 1.6 1.1 2.0 0.7 Unemployment benefit 6.1 11.8 7.3 7.1 3.4 0.9 Social assistance programs 20.3 46.9 25.1 17.1 7.7 4.8 Social assistance in cash 4.9 16.2 2.5 2.2 1.4 2.1 SA in kind (food, firewood, clothes) 1.2 4.6 0.9 0.0 0.5 0.0 Family allowances (child benefit, maternity 16.7 36.7 22.1 15.3 6.5 3.2 leave, layette) Remittances and private transfers 43.2 46.7 44.6 43.6 39.8 41.5 Source: Estimates based on the 2011 HBS. 6.80 Social assistance programs in Croatia are not particularly generous, at least compared to social insurance programs. For the bottom quintile, social assistance programs provides for about 27 percent of overall resources (Table 41). The family allowance programs contribute 23 percent of funds at the disposal to the poorest quintile, while social assistance in cash accounts for 26 percent. Social insurance programs are more generous and they cover around 87 percent of the resources of the bottom quintile. Table 41. Generosity of Social Benefits, 2011 Quintiles of income per capita distribution, net of each social transfer Total Q1 Q2 Q3 Q4 Q5 Overall social protection 46.1 89.8 47.9 30.4 22.0 16.5 Social insurance 50.1 86.6 46.5 32.1 24.7 18.8 Old-age pension 51.5 76.4 39.1 30.9 26.8 20.4 Disability and survivors pension 39.5 62.0 29.6 22.9 18.8 17.9 Sickness benefit incl nursing, disability, 18.2 34.9 8.0 12.1 8.4 8.7 Unemployment benefit 15.6 29.5 12.1 10.3 8.3 6.1 Social assistance programs 14.6 27.3 10.4 8.8 7.1 6.6 Social assistance in cash 15.6 26.1 10.5 8.2 7.6 3.2 SA in kind (food, firewood, clothes) 10.5 15.5 4.7 n.a. 0.4 n.a. Family allowances (child benefit, maternity 12.9 22.2 10.6 8.8 6.6 8.3 leave, layette) Remittances and private transfers 5.4 14.2 5.5 4.9 4.1 3.1 Source: Estimates based on the 2011 HBS. 6.81 Some evidence of the disincentive employment effect is found for family-based social assistance (child allowance, maternity leave, layette assistance), but not in case of unemployment benefits and needs-based social assistance. The relatively low coverage and generosity of social assistance might explain this. A higher proportion of the working age population not employed, not in education or training, not retired, and not disabled (NEETD) population are social assistance beneficiaries. Over a quarter (27 percent) of the NEETD 105 population lives in a household where someone receives family allowances. The larger is the share of NEETDs in a household the lower is the probability for him/her to find a job. But, having more children and more working adults in NEETD’s households is found to significantly increase probability of employment within a year period. Of the total working age population, more than 18 percent of population lives in a household where someone receives family-related assistance and almost 8 percent lives in a household where someone receives needs-based social assistance (Table 42). Table 42. Coverage of the Working Age Population by Social Assistance, 2010 In In Employed Retired Disabled NEETD Total education training Direct and indirect beneficiaries in proportion of WA population (in %) Total SA 21.4 27.6 36.7 16.3 71.2 36.5 24.5 Family 17.9 21.7 23.9 8.4 16.6 27.0 18.5 allowances Needs-based SA 4.7 7.7 25.0 8.9 63.1 14.5 7.9 Direct and indirect beneficiaries in proportion of WA population in the poorest quintile (in %) Total SA 36.7 50.2 63.5 24.2 60.0 48.1 40.8 Family 30.7 39.7 63.5 14.5 6.5 34.1 29.9 allowances Needs-based SA 9.4 17.6 34.3 11.5 60.0 22.2 16.2 Source: Estimates based on the 2010 HBS. 6.82 A social assistance system that includes a large number of programs offering generous benefits can generate work disincentives and reduce the supply of labor. In Croatia, a number of factors that create a culture of dependency and work disincentives exist. First, the pool of working age individuals between 15 and 64 who directly or indirectly receive social assistance transfers is large (688,000 out a total population of 4.2 million). Second, some social assistance programs are, in fact, income-replacement programs offering generous benefits, such as the disability programs. Third, some individual-level benefits stipulate that the beneficiary cannot work, which has a strong work disincentive effects. For example, the social assistance in cash (GMI program) has an implicit marginal tax rate on earnings (100 percent) after three months. Some families also lose the child allowance for their children if their income goes up over the eligibility threshold of the program. Finally, the work disincentive impact of all social assistance benefits could be compounded at the household level given the possibility of accumulating numerous benefits. Many EU countries face similar problems and in order to prevent the long-term dependence on social assistance benefits have reformed their social assistance systems to promote work incentives (Box 19). Box 19. Welfare-to-Work Reforms in Selected EU Countries Hungary modified the design of its regular social assistance benefit so that beneficiaries could continue to receive some benefits for up to 6 months after getting a job. Similarly, Latvia has introduced a GMI benefit of limited duration that can be received in reduced amounts after getting a salaried job. In Slovakia, the program of reforms to mitigate any adverse impact on incentives to join the labor market included tax policy reforms, active-labor market policies as well as the social assistance benefit itself. Through active labor market policies, the unemployed can work in locally organized works to increase their social assistance benefits. In the UK, France and the US, the benefit formula for last resort programs has been modified to reduce the marginal tax on earnings from 100 percent to a lower level: France’s Revenue de solidarite active has reduced the rate to 38 percent; in the UK the new Universal Credit program uses a marginal tax rate of 65 percent. 106 6.83 The Government faces a serious fiscal consolidation challenge, which it needs to address in the context of the EU Excessive Deficit Procedure. In the case of social assistance spending, the primary objective should be to strengthen the effectiveness of social assistance programs through improved targeting. This could be done through the introduction of means- testing across a wide range of programs. The objective would be to restrict the access of the relatively well-off households, e.g. the top 60 percent of the population, to these programs. A sizable share of social assistance spending reaches high and high-middle income households (top 40 percent of population), which does not contribute to poverty alleviation. 6.84 Improved targeting can generate noteworthy fiscal savings, but also improve fairness and spur activation. Restricting the eligibility for social assistance programs and policies primarily to the poor and the lower-middle income population could generate fiscal savings amounting to 0.85 percent of GDP (Table 43), depending on the types of reforms pursued. Table 43. Eliminating Elite Capture in Non-Contributory Social Assistance Programs Potential Non-contributory SA % of GDP, Elite capture % savings Political programs and policies 2012 Benefits to Q4-Q5 (% of GDP) difficulty Child tax allowance 1.0 54% 0.5 Moderate Child allowance 0.5 12% 0.1 Moderate Support allowance 0.2 9% 0.02 Small Other programs 2.2 11% 0.23 Moderate Total 3.8 0.85 Source: Estimates based on HBS 2011. Recommendations 6.85 Significant saving of about 0.85 percent of GDP could be generated by eliminating waste and improving the targeting of Croatia’s social assistance. With the current level of social assistance spending and some improvement in targeting, Croatia should be able to eliminate relative poverty for the poorest 10 percent of the population. In 2010, assuming perfect targeting, all the poor could have been lifted out of poverty at the cost of approximately 0.4 percent of GDP (or HRK 1.34 billion). In reality though, the cost would have been somewhat higher because households would have changed their behavior, expecting to be fully subsidized. Also, an allowance for the administrative costs of delivering cash transfers should be added to this total. 6.86 There is a strong rationale for using a single, unified set of criteria, to assess eligibility for needs-based social assistance programs. First, unified set of criteria ensures horizontal equity: all Croatian citizens are assessed using the same criteria. Differences in the scope of different programs could still apply, for example by offering different programs to different target groups (e.g. poor families with children versus disabled) or by varying the eligibility thresholds across programs. Second, it could reduce the administrative costs as the frontline staff use the same criteria across all programs. It can also minimize the private cost of applicants, as they have to present the same documentation for different benefits. Finally, harmonizing the eligibility criteria for need-based programs could reduce error and fraud. 6.87 To extend means-testing procedures to other social assistance and family programs than the GMI, Croatia could (i) adopt a single legislation for the assessment of household 107 means (using the example of some EU countries like Slovenia), based on the eligibility criteria for the GMI; (ii) to amend the eligibility criteria of individual programs to match the criteria of the GMI; or (iii) to consolidate programs using the platform of the GMI program. The Government has already embarked on this route: it consolidated four programs for low-income households using the support allowance program as a base for a new program––the Guaranteed Minimum Benefit/Income program. 6.88 The parametric redesign of the child tax allowance is recommended. This program does not have the poverty alleviation objective, and is highly regressive, while it is costly in terms of taxes foregone. Introducing a flat tax allowance of the amount close to the child benefit allowance would reduce current inequity and increase tax revenues. 6.89 The government should implement “make-work-pay” benefit reforms. This should include targeting active labor market measures (employment subsidies, labor market training, and measures to promote jobs for disabled workers and youth) to the long-term unemployed and long-term social beneficiaries. This also includes intensifying “activation” measures for those groups, including by introducing compulsory job-search workshops and improving basic skills of the long-term unemployed. 6.90 Finally, Croatia’s costly and complex social assistance system makes it imperative that it starts reducing losses due to error, fraud & corruption (EFC). Over the next years Croatia needs to put in place an effective and cost-efficient system that will prevent, detect, deter and monitor EFC in the benefit system. The MoSPY and other ministries and agencies involved in social policy would need to put in place an effective mechanism to reduce EFC in the benefit system. Generous benefits and infrequent recertification invite fraud. The savings could range between 0.2-0.4 percent of GDP, equivalent to the overall spending on the GMI program or half of the spending on the child allowance program. The experience of OECD countries has shown that through developing the right institutions, sufficient staff, and mechanisms to prevent, detect, deter and monitor EFC governments can reduce these losses substantially. 108 Croatia’s sector and producer subsidies through state aid have averaged 2.4 percent of GDP in recent years. However, there is little evidence that they have contributed to improving performance in the targeted sectors. On the contrary, low and sometimes deteriorating performance indicators suggest the ineffectiveness of state aid, especially for railroads, ports, and agriculture. Analysis of select subsidies suggests significant scope for rationalization, in the order of about 1 percent of GDP. This implies eliminating the most inefficient subsidies, such as product subsidies in agriculture and rationalizing sector subsidies, especially to railroads, ports, steel and shipyards, and tourism. This will help fiscal consolidation, improve efficiency, and raise welcome competitive pressure in these sectors. Given the large imbalances in the size and structure of Croatia’s state aid relative to EU standards, these reforms should quickly transition from sector-specific to horizontal types of state aid (like R&D). Finally, with large additional resources available from the EU, especially in rail and agriculture, institutional strengthening and improvements in efficiency are required to raise absorption capacity, competitiveness and the overall performance of these sectors. A. Size and Structure of Croatia’s State Aid 7.1 Croatia’s large sector subsidies need significant rationalization, especially in the context of the EU frameworks on subsidies for member countries. Croatia’s subsidies represented on average 2.4 percent of GDP during 2008-13. With significant fiscal consolidation required to bring public debt to a sustainable trajectory, subsidies of this magnitude are no longer affordable. Moreover, there are questions about their effectiveness, compared to other EU member countries, and the policy options to streamline them. To answer these questions, this chapter aims to: (i) document the size and structure of these subsidies in the context of the EU frameworks; (ii) present evidence on the productivity and efficiency of sectors they support and draw preliminary judgments about their effectiveness; and (iii) outline the scope for cuts in unproductive subsidies as well as policy actions towards their rationalization. 7.2 The EU regulatory framework recognizes potential benefits as well as costs of state aid. On the one hand, state aid can promote economic integration, supporting efficient use of public resources, boost growth and jobs. For example, private investors may find activities with potentially important social spillovers too costly or risky (e.g. environmental protection, use of renewable energy or research and development), which could justify the use of state aid. On the other hand, when it supports inefficient businesses and/or results in unfair competition between subsidized and non-subsidized enterprises, state aid becomes costly, discriminatory, and wasteful. Because all aid is distortionary, EU state aid is well regulated to ensure that government interventions minimize distortions to competition and trade within the EU. This has kept governments from giving selective advantages to specific firms (Box 20). 7.3 State aid is a matter of balance. Well-designed support and effective state aid instruments can help efficient and innovative companies grow stronger while fostering the orderly exit of inefficient firms. However, effective state aid must also be affordable and not threaten fiscal sustainability; it must recognize the inherent difficulty in picking “winning” industries or enterprises; it must resist pressures from interest groups receiving ineffective aid to 109 continue that aid; and, finally, it must be subject to a critical review and challenge from international partners. Box 20. Varieties of State Aid and Its Instruments in the EU The new EU State Aid Modernization program, launched in May 2012, sets out an ambitious state aid reform program. Its objectives are threefold: (i) to foster sustainable, smart, and inclusive growth in a competitive internal market; (ii) to focus the Commission’s ex-ante scrutiny on cases with the biggest impact on the internal market while strengthening Member States cooperation in the enforcement of state aid; and (iii) to streamline the rules and provide for faster decisions. According to the EU definition, state aid is any aid that: (i) is granted from state resources, (ii) constitutes an economic advantage to a certain company, economic sector or region, (iii) is selective as it affects the balance between companies receiving the aid and their competitors, and (iv) has an adverse effect on competition and trade between member states. In some cases, the actual aid element may differ from the nominal amount as in the case of subsidies, loans or guarantees. State aid in the EU is classified into four main categories: horizontal (allocated to all enterprises in the economy), sectoral (provided to select companies in select sectors), regional, and aid to agriculture and fisheries. Aid can be provided through various instruments: grants, tax exemptions and relief, equity participation, soft loans, tax deferrals (liabilities of taxpayers to tax authorities) and guarantees (issued and revoked). Source: European Commission, http://ec.europa.eu/competition/state_aid/modernisation/index_en.html 7.4 Yet Croatia’s state aid is anything but in balance: relative to GDP, Croatia spends double compared to the EU countries. In 2013, the total state aid, including agriculture, in Croatia amounted to 2 percent of GDP, about two times higher than the EU average, or the comparators like EU10 or EU15 countries (Figure 62). Figure 62. Croatia’s Total Non-Crisis State Aid and Its Structure, Percent of GDP Source: EUROSTAT, Croatian Competition Agency (CCA). 7.5 Not only is the size of Croatian state aid significantly higher than EU standards, its structure is dominated by selective, sectoral and firm-specific state aid. Nor has this structure of aid changed much over time. While in EU27 state aid is fairly evenly spread across sectors, with coal and manufacturing sectors, for example, accounting just over 50 percent of sectoral aid and financial services and other non-manufacturing close to 40 percent, in Croatia it is dominated by two privileged sectors/companies: shipbuilding and RTV broadcasting (HTV). Croatia’s heavy reliance on sectoral aid (40 percent) compared with the EU’s (12 percent) suggests a considerable scope to reform state aid to bring it in line with the EU. It also implies 110 significant cuts in state aid for the currently privileged sectors, e.g. steel, shipbuilding, transport and tourism. 7.6 Contrary to a clear upward trend in EU27 and especially in EU15, aid for R&D in Croatia has been stagnant. There are other disparities between Croatia and other EU Member States in the share of aid allocated to the various horizontal objectives. For example, around half of the horizontal aid in Croatia is allocated to regional and local aid with additional 36 percent to R&D and SME. By contrast, EU15 horizontal aid is more equally distributed between environmental and energy saving, regional aid and R&D. As a share in total aid to industries and services, R&D aid in Croatia is half of the similar aid granted in EU10 or approximately one- quarter of the EU15 average. 7.7 In addition, Croatia’s share of regional aid is approximately half that of EU10. Regarding aid to industry and services, regional aid was sizeable and more important in the EU10 countries (30 percent of the total) than in the EU15 countries (23 percent) because these member states have more eligible regions. By contrast, regional development aid in Croatia is mostly subject to special regulations (e.g. the Act on Areas of Special State Concern) and state aid schemes, and only exceptionally awarded as individual state aid.77 7.8 Since 2003, when the State Aid Act was first adopted in Croatia, the regime has been gradually adjusting toward support for “services of general economic interest” (Box 21). Notwithstanding that sector aid for specific companies continues to be significant, and with the exception of the 2006-7 rescues of shipyards and the Radio-TV company, less aid is gradually being directed to sector specific objectives, with more aid for horizontal objectives. Box 21. Services of General Economic Interest Services of general economic interest (SGEI) is predominantly state aid for public services and is mainly directed towards land, maritime and air transport, airport services and public service broadcasting (Hrvatska radiotelevizija-Croatian RTV). In 2012 SGEI’s share of total aid awarded was around 23 percent. The new package of EU state aid rules for SGEI clarifies key state aid principles and introduces a diversified and proportionate approach with simpler rules for SGEIs that define them as small, local in scope or pursuing a social objective, while better accounting for competition considerations for large cases. Member States are largely free to define which services are of general interest, but, the EC ensures that public funding of SGEIs does not unduly distort competition in the internal market. Whenever possible, the SGEI should be selected through an open and transparent public tender to ensure the best quality at the cheapest cost for taxpayers who fund these services. Source: http://ec.europa.eu/competition/state_aid/legislation/sgei.html 7.9 The instruments most frequently and extensively used by EU member states, including Croatia, are subsidies and guarantees (Table 44). Croatia’s most frequently and extensively used aid instruments are subsidies, at roughly 60 percent of state aid to industry and services, followed by state guarantees at approximately 30 percent. On average, subsidies represented 5.5 percent of general government expenditures. This is one of the highest shares of subsidies to general government expenditures among EU27 countries and twice as high as the EU27 average. 77 In 2012, regional aid providers were the Ministry of Economy, the Croatian Bank for Reconstruction and Development, and the Croatian Agency for Postal Services and Electronic Communications. At the local level, the biggest aid provider is the City of Zagreb. 111 Table 44. Croatia State Aid to Industry and Services per Aid Instrument (without Agriculture) Source: CCA. 7.10 The discrepancy between Croatia and the EU in the use of guarantees can be attributed to shipbuilding, transport and tourism. State guarantees are often less transparent but extensively used in Croatia, and related to loan or other financial obligations. However, only a small part of this was registered as state aid (only HRK 1.1 billion or 11 percent of issued total guarantees), before the whole amount was assumed by the state (Table 45). Almost all aid to shipbuilding is directed to large shipyards. In 2011, the Competition Agency authorized restructuring aid to shipyards in the amount of HRK16.4 billion (or 5 percent of GDP), which covers the period from March 2006 until the closure of the restructuring process. Subsequently the Kraljevica shipyard went into bankruptcy, reducing the amount of state aid that might be recovered to HRK14.8 billion or roughly 4.2 percent of GDP. Finally, tax exemptions, which are more often used in EU, accounted for only around 11 percent of aid in Croatia. Table 45. Croatia State Guarantees, 2003-2012, in HRK million 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Total guarantees issued 11,724.4 10,579.1 7,168.1 10,835.8 14,989.1 7,866.7 9,392.1 9,628.5 10,346.0 5,590.9 - do not constitute aid 6,467.2 3,081.3 2,767.5 6,176.7 10,284.9 5,386.6 6,780.9 6,398.7 9,283.8 4,501.7 - guarantees containing aid elements 5,257.2 7,497.8 4,400.6 4,659.1 4,704.2 2,480.1 2,611.2 3,229.8 1,062.2 1,089.2 Aid involved in issued state guarantees 194.7 269.4 235.5 2,363.3 2,739.2 1,516.0 928.5 1,388.2 860.0 674.9 Invoked state guarantees containing aid elements* 430.0 778.8 374.5 391.8 279.2 152.3 209.8 323.3 241.9 196.3 Other guarantees** 3.8 5.0 15.7 23.5 16.1 19.4 13.3 22.9 Total amount state aid contained in state guarantees 624.7 1,048.2 613.9 2,760.1 3,034.1 1,691.8 1,154.4 1,730.9 1,115.2 894.1 % of total gurantees issued 5.3 9.9 8.6 25.5 20.2 21.5 12.3 18.0 10.8 16.0 Source: CCA. 7.11 Railway, road and port infrastructure account for a lion’s share of all outstanding guarantees (90 percent). These state guarantees were given to the Croatian railway and shipyard companies, which are considered firms “in financial difficulties,” meaning that the full amount of the issued guarantee is considered state aid. However, the guarantees issued to shipyards in difficulty to cover advance payments are not considered state aid in their full amount, but only for the estimated value of interest charged below the market price. Many of these guarantees were recently called, especially for shipyards, becoming explicit, ex-post subsidies (Table 46) thus increasing government direct debt. Table 46. Default Risk for Invoking of State Guarantees, 2003-2011, in HRK million 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Guarantees with state aid element 5,257.2 7,497.8 4,400.6 4,659.1 4,704.2 2,480.1 2,611.2 3,229.8 1,062.2 1,089.2 Called guarantees shipyards 0 0 0 0 0 289.7 1,898.9 772.2 1,268.3 288.3 Other called guarantees 429.97 778.83 374.53 391.8 279.2 152.3 209.8 323.3 241.9 196.3 Total called guarantees 429.97 778.83 374.53 391.8 279.2 442.0 2,108.7 1,095.5 1,510.2 484.6 Risk rate w/o shipyards 8.18% 10.39% 8.51% 8.41% 5.94% 6.14% 8.03% 10.01% 22.77% 18.02% Risk rate total 8.18% 10.39% 8.51% 8.41% 5.94% 17.82% 80.76% 33.92% 142.18% 44.49% Source: CCA. 7.12 State aid in the Croatian transport sector is allocated between railways, maritime transport and air transport. Most of this aid was awarded in the form of grants and guarantees. 112 Aid to railways78 represents around 8 percent of total state aid and around 56 percent of aid to the transport sector. Aid to maritime transport represents around 4.5 percent of total state aid and around 25 percent of aid to transport sector. Aid is granted under the aid schemes “Grants for transport connectivity of the islands with the mainland and between the islands” and through de minimis aid has been granted to a specific beneficiary not exceeding the ceiling of HRK 1,500,000 (EUR 200,000) over any three fiscal years. Aid to air transport represents around 3 percent of total state aid and around 19 percent of aid to transport sector. Croatia differs from most EU countries which focus their aid almost exclusively on railways. Table 47. State Aid to Transport Sector (including Railways) Source: EC and CCA. Effectiveness of subsidies 7.13 Despite sizeable public funds allocated to the sector, the quality of Croatia’s infrastructure in railroads and ports has been significantly below EU standards. No standard and detailed data are available EU wide or in Croatia for a comprehensive assessment, but indicators suggest that public funds have not been particularly effective, especially in terms of their objective of improving competitiveness. For example, the quality of railroad infrastructure in Croatia remains low not only compared with EU15 but also compared with EU10—and the gap is growing (Figure 63). Compared to the large and relatively stable level of public funds allocated to the sector, the rapidly deteriorating quality of railroad infrastructure suggests that funds are not being used efficiently. In port infrastructure, the picture is somewhat better in that the quality of infrastructure has been improving, closing the gap with EU1379, but still remains far below the EU15 (Figure 64). 78 The railway company receives more aid from State budget than the above mentioned but these amounts were particularly aimed at reconstruction and investment in infrastructure and as such do not constitute state aid within the meaning of the State aid act. Also, they receive aid from providers at local and regional level. 79 EU13 refers to all countries that acceded the EU after 2004, including Croatia. 113 Figure 63. Quality of Railroad Infrastructure and Croatia’s State Aid Score: 1 - 7 (best) 5.0 3.5 0.35 4.8 4.6 3.4 0.3 Croatia 4.4 EU13 0.25 4.2 EU15 3.3 4.0 0.2 3.8 3.2 3.6 0.15 3.4 3.1 Quality of infrastructure 0.1 3.2 State aid to railways (right) 3.0 3 0.05 2009-2010 2010-2011 2011-2012 2012-2013 2013-2014 2.9 0 2006 2007 2008 2009 2010 2011 2012 2013 Figure 64. The Quality of Ports Infrastructure Score: 1 - 7 (best) 6.0 5.5 5.0 4.5 4.0 Croatia 3.5 EU13 EU15 3.0 Source: CCA and World Competitiveness Index. Aligning Croatia’s State Aid Subsidies with the EU Frameworks 7.14 Given the large imbalances in the size and structure of Croatia’s aid relative to EU standards, the challenge is to quickly transition from sector-specific to horizontal aid (Figure 65 and Figure 66). Before accession, most Central and Eastern European countries had state aid policies significantly different from EU15. After the initial divergence before accession, the average state aid in EU10 has rapidly converged toward the EU15 average. 7.15 State aid rules will be more rigorously applied to Croatia as an EU member state. In that context, questions arise about how this reform would be implemented.80 Which subsidies and in which sectors will need to adjust most and by how much? The following two sections take an in-depth look at two large types of subsidies that have significant potential for rationalization– –railways and agriculture subsidies. It is estimated that the potential savings in these subsidies, from efficiency gains from proposed measures, are in the order of one percent of GDP. 80 A complete assessment of state aid effectiveness is beyond the scope of this report. In 2012, for the first time the Competition Agency collected data on the effectiveness of granted aid in 2011. Although data was requested from all aid providers, only a handful of institutions responded. While the effects of aid schemes were assessed positively, the submitted data and assessment made by aid providers were rather general and lacked a detailed breakdown. In future, comprehensive monitoring of effectiveness of aid is needed to more rigorously quantify its effects. 114 Figure 65. Horizontal aid, as % of total aid less Figure 66. Sectoral aid, as % of total aid less agriculture and fisheries and transport agriculture and fisheries and transport 100,00 90,00 90,00 80,00 80,00 70,00 70,00 60,00 60,00 50,00 50,00 40,00 40,00 30,00 30,00 20,00 20,00 10,00 10,00 EU15 EU12 HR EU15 EU12 HR 0,00 0,00 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Source: EC, CCA (horizontal aid includes regional and local level state aid) B. Rationalizing Railway Subsidies 7.16 Demand for railway transport has been declining since 2007, although it still accounts for a large share of national traffic (Figure 67). Railway transport in Croatia is predominantly transit: the country benefits from its strategic location at the junction of two main European transport corridors, Corridor X and Corridor Vb.81 More than EUR 600 million82 has been invested in Adriatic seaports over the past decade, in an attempt to develop ports as entry points for international goods and passengers to markets in Central and South East Europe and for cruise and tourism industries. However, because of poor connectivity with other transport modes and neighboring countries, their potential has not been fully utilized. In addition, around HRK50 billion (15 percent of GDP) has been invested into the motorway network since 2002, constraining public funds for railways to HRK19 billion (5.8 percent of GDP) over the period. Figure 67. Structure of Inland Transport by Types in Croatia, 2012 Source: Croatian Bureau of Statistics. 81 Corridor X connects Salzburg, Austria and Thessaloniki, Greece via Zagreb, while Corridor Vb connects the port of Rijeka with Budapest, Hungary via Zagreb. The new EU infrastructure policy establishes a core transport network on nine-major corridors. The Mediterranean Corridor links the Iberian Peninsula with the Hungarian-Ukranian border and in Croatia follows the route of the former Corridor Vb; while the Rhine-Danube Corridor includes the Danube branch in Croatia. 82 “Survey on Status of Preparedness for EU Accession and Support to Transport Sector”, WB, November 2012 115 7.17 Croatia’s accession to the EU in July 2013 has integrated the national transport market into the European transport network. As the EU pursues an active policy of promoting railways as an efficient, reliable, safe and environmentally friendly mode of transport, Croatia will have at its disposal around EUR1.5 billion83 in the financing period 2014-2020 to modernize its transport sector, most of which is supposed to be allocated to international rail corridors. Railway efficiency and development is recognized as one of the drivers of transport and economic growth in the Government’s 2011-2015 Program. 7.18 In preparation for European membership, Croatia completed the following reforms in the railway sector:  Harmonization of the legal and institutional framework of the railway sector with the EU Acquis Communautaire. However, rail companies still need to adapt to a post-accession environment that requires opening the domestic market to new entrants and services. This requires a gradual transformation to a business-oriented industry, independent from state intervention; and  The Ministry of Maritime Affairs, Transport and Infrastructure (MMATI) adopted a five-year restructuring plan for Croatian Railways in June 2012, an important step to increase operational and financial performance. Previously organized as a holding company with four lines of business84, Croatian Railways are now split into three independent companies - HZ Infrastructure, HZ Cargo and HZ Passenger Transport. Importantly, the Restructuring Plan resulted in the adoption of key policy decisions, in accordance with EU rules, such as the dissolution of the holding company structure, improvement of the contract between the government and the infrastructure manager, elimination of subsidies to the cargo company and a decrease of subsidies’ share in operating revenues. All restructuring plans are being regularly updated to track progress of the key reform elements, and in conformity with the EU state aid guidelines prepared for formal submission to the EC. 7.19 The railway reform program, however, may fail to achieve its objectives if the government does not provide a credible medium term financial and institutional strategy. While the short-term financial situation of the railway companies is being constrained, their medium-term financial sustainability is not secure. They still rely on large state support for investments. Important decisions outside the area of HZ’s control are yet to be addressed by transport and finance ministries, including the rightsizing of the network and services, charging for infrastructure or investment funding strategies. Subsidies also need to be revised in the Restructuring Plan to sustainable levels. As an important step in this direction, the MMATI and HZ Passenger Transport have recently prepared a 10-year Passenger Service Contract (PSC) which includes a significant reduction of railway services. 7.20 The urgent need for restructuring is increasing against the backdrop of a significant decline in demand, poor operating performance, and an aging rolling stock. The rail traffic in Croatia is much lower than in the rest of the EU (Figure 68). Measured in million traffic units, it has declined by 36 percent since 2007, to 3,525 million traffic units in 2012 (Table 48). Cargo traffic volumes were strongly affected by the international financial crisis, as well as internal 83 Latest figure is EUR 1.5 bn (260mn ERDF+1,117mn CF+150mn extra from CF that was recently added), to which one can add Connecting Europe Facility (CEF) allocation of 465mn EUR. 84 HZ Cargo, HZ Passenger Transport, HZ Traction and HZ Infrastructure 116 problems such as low utilization of freight services that were implicitly paying higher track access charge (TAC) than the passenger company while operating only a quarter of all train- kilometers in Croatia, and an unsuccessful attempt to privatize the company in 2013.85 Also, less than 10 percent of rail network is double tracked with only 35 percent allowing speeds of more than 80km/h, down from 65 percent in 2008. Figure 68. Rail Traffic Intensity Compared to the EU (2012); Croatia=100 Note: Traffic intensity is the ratio of traffic units per km. Source: Eurostat, World Bank Calculations Table 48. Key Performance Indicators of Railways 2005 2007 2009 2010 2011 2012* Traffic units (mill pass-km + ton-km) 4,372 5,482 4,706 4,474 4,249 3,525 Traffic intensity [traffic units/km] 1,603,778 2,013,924 1,728,876 1,643,644 1,560,985 1,295,004 Total staff 14,152 13,411 12,843 12,491 12,468 11,493 Labor productivity [traffic units/staff] 308,925 408,761 366,425 358,178 340,792 306,708 Labor cost as % of operating revenue 76.20% 64.60% 71.20% 70.60% 78.6% 71.3% Average unit operating Cost less 0.084 0.066 0.077 0.078 0.081 0.113 depreciation [Eurocents] Operating ratio with state support 1.25 1.27 1.12 1.12 1.04 1.33 Operating ratio without state support 66% 79% 72% 72% 69% 58% * HZ Holding separated into three independent companies in 2012, some figures may not be comparable Source: HZ, World Bank Calculations 7.21 The majority of assets are more than 30 years old and keeping the present level of operations will require significant infrastructure investments (of at least HRK 425 million or 0.13 percent of GDP annually for the next 10 years). Since 2007, on average 30 km of track has been rehabilitated per year, nearly all being funded by national sources. Wagon productivity deteriorated by 40 percent between 2007 and 2011 to just 0.47 million ton-km per wagon, while 85 While the latter issue was somewhat addressed though a dissolution of the holding company structure, a transit oriented railways like Croatian Railways should explore ways to fine-tune TAC in relation with neighboring countries and other modes of transport to stimulate traffic growth for that segment. 117 most European transit operators achieve 5-7 times higher wagon productivity (Figure 68). The wagon fleet is more than 30 years old (Figure 69). On the other hand, HZ Passenger Transport’s coach productivity was only slightly down since the 2008 crisis and in 2011 still surpassed that of the EU27 average by 40 percent, indicating that the current fleet size and usage in this segment is good. One reason for it is a regular replacement of depleted assets with new passenger coaches, as it has been done in early 2014 when HZ Passenger Transport signed up an order of 44 new train units for a total of HRK1.6 billion. Figure 69. Age Structure of Infrastructure and Rolling Stock of the Croatian Railways Source: Eurostat, World Bank Calculations The Low Efficiency of the Railway Sector 7.22 Labor productivity of railways has been low and falling since 2008, mainly because of overstaffing. Labor productivity is almost half of that realized by European peers, and more than four times lower than that of Finland, Latvia and Lithuania. This suggests that the size of the labor force is not optimal for the significantly reduced level of traffic, despite a continuous staff decrease from 2005. The largest scope for productivity improvements exists in HZ Infrastructure. Its 2012 level of 2.36 staff per track-km is similar to other South East European infrastructure managers (which are all in a dire need for reform), but less than half of that in Western Europe. Improving staff productivity to 1.9 staff per track-km would bring around HRK175 million (0.05 percent of GDP) of savings per year. HZ Infrastructure laid-off over 8 percent of its staff in 2012, but improving productivity will require stronger efforts in this direction along with the introduction of performance management systems. 7.23 The financial performance of the Croatian Railways system is weak. There is a substantial need for cost-cutting and business rationalization in order to achieve restructuring goals within the next 3-4 years. Significant reductions has meant that HZ Holding’s has been operating with losses since 2008 and ended 2011 with HRK 127 million (about 17 million euros) of net losses before its dissolution into three separate companies. The companies consequently have relied on significant state support, typically covering only 69 percent of their operating costs from operating revenues and with almost no margin left for capital expenditures86. Following the holding dissolution, the financial situation of the companies has remained weak, altogether accumulating losses of over HRK 600 million in spite of overall subsidies exceeding 86 Figures for 2011 on a holding level, calculated by the World Bank based on data provided by HZ. 118 HRK 1.9 Billion in 201287. The situation is most bleak for HZ Cargo, which cannot rely on state support any more due to state aid rules. 7.24 Overstaffing and the wage bill remain unsustainably high. In 2012, HZ Holding used more than 70 percent of its operating revenues just to pay labor costs, while most European railways operate with a wage bill of around 40 percent. According to World Bank calculations, the average annual cost per employee reached HRK 134,881 in 2011 and the cost of staff increased by 21 percent over 2005-2011. HZ Holding reported an average monthly gross wage of HRK 8,559 for 2011; only 2 percent less than the public sector average, but 10 percent more than the average in the education sector or the national average, despite a relatively less qualified labor force. 7.25 Not surprisingly, the railway sector has been heavily reliant on the budgetary support. Despite a 36-percent decrease Figure 70. Evolution of State Budget Support to Railways in comparison to 2005, public transfers to railways amounted to HRK1.9 billion in 2012, or 0.6 percent of GDP (Figure 70). More than half of these funds were earmarked for the infrastructure operations, and slightly less than 20 percent for passenger subsidies through Passenger Service Contracts (PSCs). Part of the 2012 reform program envisages an increase in public funds to railways to support the restructuring processes in the railway companies, totaling HRK14 billion over period 2012-2016. However, railways have Source: HZ, State budget, World Bank calculations been resorting to significant external loans in the short term to fund investment, retrenchment and for liquidity. Elements of a Viable Restructuring Plan 7.26 Considering low labor productivity, systematically weak financial performance, overstaffing, and the high and increasing wage bill, the rail companies are in need of serious and urgent restructuring. Elements of such necessary restructuring are clear: adjusting the staff size to the actual demand for railway services, and further renegotiating the collective agreements with Unions, which in current shape are still inflexible and put a heavy burden on companies’ financial positions. HZ Infrastructure and HZ Cargo started in this direction in 2012 , partially relieving some financial pressure, but much more will be needed to put these companies on a sound financial footing. HZ Passenger Transport resorted to high level of public funds to finance its services and has launched a technological modernization of the company, while additional cost cutting measures are likely possible. 7.27 The Restructuring Plan prepared by HZ Holding in 2012 proposed a continued and increased reliance on budget and external resources. Given the budgetary situation, the 87 All companies ended 2012 with losses, amounting to HRK 232 million, 248.5 million and 127.9 million for HZ Infrastructure, HZ Cargo and HZ Passenger Transport, respectively. 119 Restructuring Plan would need to be strengthened to reduce reliance on budgetary resources (Figure 71) and tap into the EU funds for capital investments. The Restructuring Plans also counted on rather optimistic levels of State support for investments, reaching HRK 6.3 billion (around 2 percent of GDP) per year. Total investments foreseen by the Restructuring Plan were close to HRK 24.5 billion and were to be financed through a combination of sources: HRK 10 billion through commercial loans (most probably with a State guarantee); HRK 6.3 billion from EU funds; while own sources should cover for the residual. It is not clear whether the estimated amount of state contribution includes the national co-financing share for the EU funds; whereas the investment part assumes having well-prepared projects that could be approved and could enable early disbursement. This plan is being reviewed following the approval of the convergence program in relation to Croatia’s EDP and will need to be finalized after a thorough check of financial means at the disposal of the government and the sector. Figure 71. Difference between the World Bank and HŽ Estimated Public 7.28 Moreover, the Funds to Railway Sector, 2012-2016 (in HRK thousands) railway investment program needs to derive 3656 from the overall 3048 3038 transport investment 2697 2627 strategy that should 2464 foremost prioritize the 2494 2552 2430 transport investments 1612 over the medium term and link investment needs with the sources of financing based on more realistic assumptions. This work is now in the making with an 2012 2013 2014 2015 2016 interim transport strategy Total subsidies, capital and interest (HZ) Total subsidies, capital and interest (WB) currently being prepared Source: HZ, State budget, World Bank calculations by MMATI. 7.29 A viable restructuring plan must also establish prudent investment policies. State allocations for railway investments have been shrinking since 2008, to just below HRK450 million for all ex-holding companies. The World Bank estimates that companies should invest at least EUR130 million (HRK975 million) annually to sustain current level of operations. The government has decided to increase again the level of subsidies for both passenger transport and infrastructure for 2014-2016, but has not clearly decided whether part of it should be dedicated to recurring maintenance and which should be focused on operational costs. 7.30 A significant staff retrenchment over a five-year period is planned, but bolder actions are needed to make companies profitable. For the period 2012-2016, the combined staff size of the three companies could be reduced by 35 percent, which according to projections should result in HRK 520-760 million (0.15-0.23 percent of GDP) in savings. Given the low labor productivity and poor financial position, however, these measures are probably not ambitious enough. Around 4,700 staff is still employed in subsidiaries of the main companies, which mean that some of the operational indicators related to productivity would have been much worse if reported figures had accounted for them. Since the adoption of the plan, HZ 120 infrastructure did some significant retrenchment, and HZ Cargo is currently undertaking a significant downsizing process. 7.31 A staff downsizing of 48 percent, from 2011 levels, would represent the cost cutting needed to bring staff productivity to EU27 levels by 2016-2017. However, the average annual cost of retrenchment ranges from HRK105 to HRK150 million and would place additional pressure either on the companies’ financial situation or on an already tight state budget. Any additional staff cuts therefore need to be well planned, taking into account current borrowing costs (as retrenchments in 2012-13 were predominantly financed through expensive commercial loans), and budgetary impacts. Further savings could also be made by cutting external services and/or modifying the level of services provided. HZ Infrastructure seems to plan on such an accelerated cost cutting scheme according to its reform program, as well as the restructuring of subsidiaries: keeping profitable business activities within a single company for track maintenance while privatizing, liquidating or transferring others to the state. 7.32 Considering that both investment and operating subsidies had a decreasing trend in the past (with the latter essentially used for covering operating losses), companies started exploring the EU funding options. Given a declining state aid and little margin left to invest from own resources, HZ Infrastructure and to a lesser extent HZ Passenger Transport have started to explore opportunities with EU structural funds to finance their investment programs. However, they would need to complement it with strong savings on costs, either to directly invest or get some borrowing space. HZ Cargo, which in 2012 in accordance with the EU regulations was not a beneficiary of state support, will not be eligible for any kind of external assistance, except that coming from private sources. Following a failed privatization of HZ Cargo, the company has started financial restructuring in 2014 and is in the process of finalizing a restructuring program that has been submitted to the EC by mid-2014. EU Funding and Related Challenges of the Croatian Railway Sector 7.33 In addition to formidable internal restructuring challenges, EU funding brings its own set of demands. Effective and efficient utilization of EU financial assistance under the 2014-20 perspective is subject to different criteria that need to be satisfied, at both strategic and programmatic levels. In particular, fulfillment of ex-ante conditionalities should be given attention. In the railway sector, these include: (i) the preparation of a transport plan or framework for the sector, which fulfills legal requirements for (ii) an environmental assessment, (iii) a mature project pipeline, and (iv) to ensure capacity of intermediary and beneficiary bodies to deliver the project pipeline. Although the EC may grant a temporary waiver on some of them, ex-ante conditionalities represent a necessary prerequisite for using EU support and should be addressed at the earliest. 7.34 Croatia must therefore develop a comprehensive transport strategy to guide the work on the rail sub-sector strategy and multi-annual planning documents, setting clear, consistent and specific objectives. During the previous programming period 2007-2013, around 92 percent of EU (IPA) funds for transport infrastructure projects were streamlined to railways, a flagship sector in terms of allocations. However, while 82 percent of railway funds were contracted, only 44 percent was disbursed as of end-2013, signaling that work on project preparation needs to be seriously stepped-up within the new financial perspective 2014-2020. 7.35 With access to unprecedented amounts of funds for railways, considerable emphasis should be put on preparing a comprehensive and multimodal transport strategy . The 121 current transport strategy dates back to 1999 and has not been updated since, while the five-year railway sub-sector plan expired in 2012. Prioritization should be made between and within modes in relation to transport strategy and status of pipeline of projects upfront. An “interim” transport strategy has been prepared to satisfy the ex-ante requirement in the first years of 2014- 2020 perspective. In parallel, MMATI is working on a new transport strategy for Croatia that is most likely due in two-three years and is supported through IPA funds. 7.36 EU Funds should be the main source of financing for future rail investments. Based on the proposed EU budget, EUR8.4 billion from ESI Funds is available to Croatia in the 2014- 2020 programming period, and approximately 20 percent is expected to be allocated for transport. The government’s efforts and resources should therefore be concentrated around maximizing the use of EU funds, while state and other support to rail investments should be limited to counterpart financing, restructuring and critical infrastructure overhauls that would enable an exit from the sector. Although the maximum counterpart financing rate for ESI Funds for Croatia is 15 percent, experience of other member states is of a somewhat higher share of, varying between 30 and 50 percent, so meeting counterpart financing might be challenging and therefore requires careful planning and sequencing. 7.37 The total national contribution for railway projects in the next financing perspective may reach close to EUR600 million. This scenario is based on the EUR 1.95 billion of funds that might be earmarked to railways, representing 80 percent of the total transport allocation. Funds utilization in the first years is generally lower than expected for most countries during their first years after accession, and it will take some time for new structures to work in Croatia. The disbursement rate is therefore expected to be small in the first years, at around 2 percent, and gradually rise to 20 percent by 2020. National co-financing is forecasted at around 30 percent. Based on these assumptions, the country is projected to contribute with at least EUR 30 million in 2014 for financing EU projects in the rail sector, expected to increase towards the end of the financial perspective up to EUR 140 million. Recommendations 7.38 Compared to the current level of subsidies, railway operations subsidies could be decreased by 0.3 percent of GDP without significantly affecting the service if cost-cutting measures and other improvements described above are made. The government needs to decide on the affordable level of funding for railways, defining the overall transport investment program and setting clear funding limits. Financing should be transparent with sufficient funding for operations and investments and limits on state guarantees for state companies and state contributions to loan repayments. 7.39 Once the overall level of public funding (excluding EU funds) is settled, there is a need for setting the structure and organization of the financial support to railways. Two main tools are the Public Service Contract (PSC) and the Multi-annual Infrastructure Contract (MAIC). A first option is to provide public financial support to the infrastructure manager, therefore subsidizing the track access charge (TAC). This provides incentives for traffic growth, as rail operators pay lower charges for using the infrastructure, but the TAC is set below the level of cost recovery. The alternative is subsidizing passenger transport through PSCs and maintaining the price of access in relation to its cost. The choice between the two should be based on the amount of funding available for railways, traffic elasticity following the EU accession and market opening vis-à-vis the access charge, and the elasticity of demand for 122 passenger transport with regard to cost and quality of services. Another issue to consider is how to subsidize capital investments for the passenger segment. With the policy choices made in 2013, the government opted for subsidizing passenger transport, including its capital investments, but the TAC still needs to be optimized. 7.40 The affordable level of services and network size needs to be determined in relation to modes and resources available for subsidizing the sector. Over the medium-term the operators have to maintain or improve the efficiency of service provision in proportion to the level of funds they receive from the state. If there are plans to modify amounts for PSCs, the size of services should be reorganized and this might entail reducing provision. Likewise, the network should be right-sized based on the traffic intensity and the level of funding for its modernization and overhaul. 7.41 The government needs to strengthen the contractual relationships with the infrastructure manager and passenger operator: (i) MAIC should be established for a period of 4-5 years and determine the rights and obligations of infrastructure manager, making it also accountable for the quality and safety of operation and maintenance of the railway network. In particular, the MAIC should stipulate the annual targets for cost cutting and network downsizing, performance indicators to be achieved by HZ Infrastructure and allocation of funds for entire duration of the contract; (ii) One Multi-annual PSCs for up to 10 years or more should be put in place, allowing the passenger company to develop long-term investment plans that could be translated into improved services. The government may also look into having several regional PSCs that are centrally managed in contrast to one fully centralized PSC. 7.42 Former holding companies must become financially viable and self-sustaining. The ultimate precondition is timely enforcement and reinforcement of the restructuring program that meets planned cost cutting targets. However, despite the efforts to implement reform measures, lower than expected traffic intensity has significantly undermined the already weak financial situation of all companies. The Restructuring Plans will most likely have to be revised to account for the current conditions and the constrained budget. Converging to performance levels of the EU27 countries requires stronger cost-cutting measures and a more agile approach to reform. 7.43 Success of actions to be implemented for improving sector and corporate governance, management and performance of the railway operators, are contingent on the MMATI’s capacity to supervise the railway sector. The leading role in steering the sector reform should be shifted from the rail companies to the transport ministry. While the MMATI needs to closely coordinate its work with other partners in the government, namely MOF and MRDEUF, it remains solely in charge of broad spectrum of activities that require adequate human resources in terms of numbers, skills and qualifications in order to define the affordable level of service that the government wants to purchase both in terms of infrastructure and in terms of passenger services. 7.44 Last but not least, significant efforts need to be devoted to maximizing EU Funds absorption. This involves a high level of preparedness of beneficiaries and implementing bodies. From the government perspective, guidance should be provided with the adoption of a new transport strategy for Croatia with a well-articulated Transport section of the OP. 123 C. Rationalizing Agriculture and Rural Development Subsidies 7.45 Over the last decade, the agriculture and rural development budget has seen an increase in both absolute and relative terms (Figure 72). Relative to the entire state budget, agriculture and rural development has increased from about 2.5 percent in 2004 to roughly 3.5 percent, and grown more than twice as fast as the overall state budget (11.9 and 5.3 percent annual increase rates, respectively). The sector budget currently slightly exceeds 1 percent of GDP. This is comparable to public spending for agriculture and rural development in some EU10 countries (e.g. Bulgaria, Latvia or Lithuania), but is higher than in most other EU countries, where levels generally range between 0.3 and 0.8 percent. Within the agricultural budget, transfer programs, and in particular direct payments have preserved a sizeable share but have decreased, in relative terms, over time (Figure 73). Figure 72. Budget Spending on Agriculture and Rural Figure 73. Agriculture and Rural Development Development, 2004-12 Spending by Program (million HRK, 2004-12) Source: Ministry of Finance and Croatian Bureau of Statistics and staff calculations 7.46 However, there are several challenges with the execution of various transfer programs that need to be better addressed in the future. First, Pillar 1-type payments (direct payments, market interventions and income support) have experienced some over-commitment and deferred disbursement. Over 2004-11, budgetary commitments were often higher than the available finances; therefore, the actual payments to farmers deviated from what was planned, generating discontent among the beneficiaries and the general public. Over time, insufficient political will to ensure financial discipline as regards agricultural payments led to sizeable budgetary commitments and growing difficulty to meet payment deadlines. The basic piece of legislation regulating the agricultural support is the Act on State Support in Agriculture and Rural Development. The Act contains provisions on financial discipline but these have not been implemented consistently. 7.47 Second, the non-enforcement of farm subsidy modulation has weakened budget revenue collection as well as equitable distribution of support. The objectives of modulation in Croatia are identical to those in the EU. The first goal is to achieve a more balanced distribution of agricultural support, with a higher share of support to small and medium farmers. This is justified by the fact that the largest farmers benefit from economies of scale. Another goal 124 is the redistribution of total budgetary finances towards increasing the share of structural (primarily rural development) measures, and reducing the share of direct payments. The income forgone from not applying modulation was about HRK75 million in 2012. Another deficiency in execution of direct payments for the 2012 production year is that farmers did not receive their individual bills which points to a lack of transparency in executing direct payments. 7.48 Third, subsidies for “market intervention” have not been sufficiently transparent. It is not possible to accurately determine in advance the commitments for market intervention support for a given fiscal or production year. The Croatian agricultural budget lacks transparency in this respect. Moreover, the data of the Ministry of Finance on the payments from the state budget for market interventions in agriculture do not match the total commitments nor is it clear which measures they belong to. 7.49 Finally, national pillar 2-type payments (notably investment support) have crowded out the EU co-funded program (SAPARD), while being subject to major disbursement delays. The implementation of the national investment support scheme over the 2004-09 period enjoyed abundant and growing budgetary support. However, the more favorable terms of this national program, compared to SAPARD, was one of the main reasons for the latter’s failure. In a relatively short period of time since the scheme was launched, and until its abolition in November 2009, large budget commitments were made; some of these had to be paid through 2013. The deadlines for reimbursing participants in the investment support program kept being extended due to a growing number of competing requests on the budget, coupled with budgetary limitations. In the last few years, the delays in refunding farmers were longer than two years, leading to dissatisfaction. Productivity concerns of agricultural sector in Croatia 7.50 Croatia continuously suffers from low small farm productivity. Overall, and despite its small share in the Croatian economy, agriculture remains a key source of income (including subsistence) for part of the population. Figure 74. Farm Labor Productivity, Croatia and Agriculture, forestry and fisheries account for the EU (2005-11 average) only about 4 percent of GDP, slightly above 2 percent of GDP of EU27 average. The sector, however, employs about 11 percent of the total workforce in 2012, compared to 5 percent in EU27, with a sizeable share of 65 years and older. 7.51 With almost two thirds of the agricultural area operated by holdings larger than 10 hectares, Croatia displays relatively good overall land productivity. Across the new EU member states, average sector performance is ultimately driven by the diversity in the distribution of agricultural land Note: EU12 are EU countries that joined after between “small” and “large” farmers. Yet, some 2004 without Croatia. productivity gaps remain to be closed for certain Source: Eurostat and own calculations commodities, mainly in the livestock sector. 125 7.52 However, a high share of small and subsistence-oriented agricultural holdings weakens overall labor productivity (Figure 74). Three quarters of all farms operate on less than 5 hectares, and almost 60 percent of these (or half of the total number) consume more than half of their agricultural output88. Compared to EU27, Croatia pursues a much more labor-intensive agriculture, and utilizes twice as many annual work units (AWU) per hectare (i.e. 0.16 AWU/ha, as a 2005-11 average). This suggests that Croatia has ample scope to obtain further productivity gains in agriculture subject to promoting alternative incomes outside farming. 7.53 Croatian farmers face significant income variability, depending on their farm size . The standard output89 of an average Croatian farm is still about 5 times smaller than that of the average EU15 agricultural holding (EUR 9,065 versus EUR 45,334 in 2010). Yet, it compares favorably to the average farm in EU1290, which is producing only EUR 6,941 worth of output per year. Nevertheless, there is significant variation by farm clusters, which needs to be considered when targeting current and future policy measures. The smallest Croatian farmers face the most significant income gaps compared to the rest of the EU, but the income gap closes as the farms grow in size. The largest farmers in Croatia (i.e. operating on more than 100 hectares of land) are already generating more income than their EU peers (Figure 75). Figure 75. Average Standard Output per Farm, by Agricultural Size Clusters (thousand EUR, 2010) 30 100 400 EU-12** 350 25 80 Croatia 300 20 250 60 15 EU-15* 200 40 150 10 20 100 5 50 0 0 0 0-2 ha 2-4.9 ha 5-9.9 ha 10-19.9 ha 20-29.9 ha 30-49.9 ha 50-99.9 ha >100 ha No land Small Medium Large Source: Eurostat, Agricultural Census 2010, and own calculations 7.54 Other constraints affecting primarily the small and medium farm segments are related to farmer mobility, irrigation, and pensions. There is an unfinished land agenda that limits the mobility of farmers, especially smallholders, and locks the sector in a suboptimal structure. An insufficiently developed irrigation sector further limits farm production and productivity, while the pension system that excludes farmers limits farm restructuring. 88 Eurostat, Agricultural Census 2010 89 Standard output (SO) measures the economic size of the farm, and, starting 2010, it replaces the standard gross margins for describing the economic size of European farms. Eurostat defines the SO as the average monetary value of the agricultural output at farm-gate price, in euro per hectare or per head of livestock. 90 EU12 are EU countries that joined in 2004 and 2007. 126 Spending on public services should be rationalized 7.55 In contrast to sector transfers, spending for administration and provision of public services91 in the agriculture sector has increased significantly in absolute terms, and doubled in relative terms (accounting for as much as 25 percent of total sector spending in 2013). These expenditures are among the largest growing budgetary expenses for administration compared to other segments of the state administration over the last decade. The average rate of increase of expenses for public services in the period 2005-11 was above 20 percent per annum, which is as much as six times the rate of increase of the total expenditures of the state budget during the same period. 7.56 Within the agriculture and rural development administration, the paying agency plays a central role in the future implementation of the Common Agricultural Policy (CAP). A paying agency performs well-defined and tightly regulated functions: it delivers the envelope of CAP support to eligible beneficiaries, and performs administrative and on-the-spot controls to ensure that payments are made according to the CAP rules. The Croatian Agency for Payments in Agriculture, Fisheries and Rural Development (APAFRD) shows the fastest growth of budgetary expenses within public services in the last few years. This trend is common across EU10 countries, and is dictated by the requirement to comply with strict EU fiduciary rules. In 2012, the share of APAFRD in the total expenses for public services was only slightly higher than 13 percent. More importantly, its administrative costs in the total transfer envelope fell below 2.5 percent, which is low by EU standards (in EU10 they tend to range between 4-8 percent). APAFRD was accredited in 2013 and proved sufficient capacity for tapping into the CAP accession benefits. The current amount of HRK134 million for APAFRD, almost 30 percent higher than in previous years, and the share of paying agency costs is at a level comparable to EU benchmarks. 7.57 Still, there is scope for improving the efficiency and organization of these services. The share of spending for public services in agriculture in Croatia (25 percent of the total sector budget in 2013) is high compared to the EU, and calls for rationalization. The increase in the expenditures for administration and public services can be only partially explained by the adaptation to the CAP. The multiplication of activities performed by different public services in agriculture is evident, especially regarding assistance services to farmers. Market interventions not fully compatible with the EU 7.58 Going forward, intervention measures to regulate agricultural markets, including refunds for exports of agricultural products to non-EU countries, will be financed entirely by the European Agricultural Guarantee Fund (EAGF) in a context of shared management between the EC and Croatia. Croatia’s financial obligations will, therefore, be limited to national program management and administration. 7.59 These intervention measures will, however, provide some potential fiscal savings in relation to pre-accession support. Prior to accession, market interventions occupied a variable, though non-negligible, share in Croatia’s agricultural budget. Direct market interventions 91 Expenses for public services include expenses for administration and management of the Ministry of Agriculture and expenses of other services under the auspices of the Ministry: Agency for Payments in Agriculture, Fisheries and Rural Development; the Croatian Agricultural Agency; Croatian Center for Agriculture, Food and Village; Institute for Agricultural Extension; Food agency; Agency for Agricultural Land. These expenses include also expenses for veterinary and food safety within the Ministry. 127 accounted annually and on average for HRK153 million (or EUR 21 million) in the period 2004- 12, or some 5 percent of total agricultural expenditure. They peaked at 13 percent in 2007 but they fell to only 1 percent (EUR 3 million) in 2012. Public interventions grew twice as fast as direct payments. After accession, the ceilings for EAGF market-related expenditures are set in the EU’s multi-annual financial framework 2014-20. Deducting direct payments, the expected EU-funded ceilings are at about EUR 33 million in 2013, and subject to a sharp decline thereafter as direct payments phase in. 7.60 During pre-accession, the highest share of expenditures went to programs dealing with ‘adverse weather conditions’. During two extremely unfavorable years (2007 and 2011) the Government decided to compensate farmers for adverse weather impact not only with funds of the Ministry of Agriculture but also by significant means from the budgetary reserve. Further, 2012 was the driest year and had the most damaging effects to yields in agriculture than any of the previous years. However, due to accumulated commitments and budgetary limitations, the Government did not provide compensation for drought damage. 7.61 Other market interventions during the last decade could hardly be justified, and Croatia continues to implement not fully EU-compatible market intervention measures. The basic administrative framework for intervention measures similar to that in the EU was established by the Act on Organization of Agricultural Markets adopted in 2009. It was expected that, after its adoption, any measures would be applied in line with the legal framework of the EU. One provision of the Act enabled the application of mechanism of “special intervention measures” in cases of “serious market disturbances” for a wide range of products. This provision had not been harmonized with the EU regulations, and one of the conditions for concluding the accession negotiations was to amend this Act. Figure 76. Agricultural Policy Costs At the beginning of 2011, the application of the Score: 1 - 7 (best) provision on special intervention measures was 4.5 narrowed to the sectors of sugar, hops, beef and 4.0 veal and sheep and goats (in cases of significant increase or reduction of prices) and to sectors of 3.5 Croatia pig meat, eggs and poultry meat in cases of EU13 3.0 EU15 significant increase of prices of these products. However, regardless of such harmonization 2.5 with the EU acquis, market interventions 2.0 continued to be applied for products for which the existing legal framework provides no legal basis. Note: Agricultural policy costs are assessed on the 7.62 Finally, comparative estimates suggest scale of 1 being excessively burdensome for the that the quality and efficiency of agricultural economy to 7– balancing interests of taxpayers, policies is low in Croatia (Figure 76). consumers, and producers. EU13 average stands for Measuring the burden of agricultural policy new member states without Croatia. from the viewpoint of producers, consumers, Source: Global Competitiveness Index, World Economic Forum and taxpayers, using the Global Competitiveness Index, Croatia’s performance is significantly lower than in other EU countries—and the gap is growing. Overall, therefore, the scope for rationalization of rural policy, institutions, and subsidies appears to be substantial. 128 Low pre-accession absorption and foregone aid 7.63 Like EU10, Croatia benefitted from EU pre-accession funds to support rural investments; these emulate the Common Agriculture Policy (CAP) rural development pillar. The pre-accession support is a simplified version of CAP, involving fewer support schemes and much lower funding, but resting on the same basic implementation principles. 7.64 Prior to accession, Croatia benefitted from both SAPARD and its successor IPARD, but absorptions rates were poor. The SAPARD program was implemented during the period 2006-09, and involved a total allocation of EUR 33.33 million. Of this, the EU share was EUR 25 million, while the national public contribution amounted to EUR 8.33 million. The final rate of SAPARD utilization was only 47 percent, as Croatia forfeited approximately EUR 13 million of the available EU funds. This was due to the automatic de-commitment rule (“n+3”) governing EU funds, according to which monies not disbursed by the end of the third year following their commitment will be returned to the EU budget. 7.65 Due to the low absorption, the fiscal cost of the national counterpart financing for SAPARD and IPARD92 was well below the commitment appropriations. The budgetary efforts for co-financing the SAPARD and IPARD programs were over-shadowed by the payments made under various national rural development support schemes that were approved under less rigid requirements. The most financially prominent of these were a domestic investment scheme in agriculture, formally terminated in 2009, but still causing budgetary outlays few years later. 7.66 While IPARD absorption improved in 2013 to 39 percent (contracting rate), further de-commitment of EU funds is expected. Croatia already de-committed EUR46 million by end-2013. The IPARD program for 2007-13 was approved by the EC in 2008 with an allocation of EUR183 million. However, the start was delayed until 2010, due to a lengthy accreditation process. Given the EU-funded project life cycle (involving beneficiary pre-financing, then reimbursement), it is expected that disbursements will be made for a series of projects contracted during the first three years of implementation. In 2013, 10.2 percent of original allocation was paid, almost 25 percent more than in the preceding three years. However, to fully utilize the remaining EU funds for the current programming period, Croatia will have to disburse some EUR46 million each year between 2014 and 2016. It will be difficult to reach that goal, mainly because there will be no more tenders for the rural infrastructure measure, which is characterized by a high unit value of projects. Therefore it is estimated that by the end of the implementation of the IPARD Croatia will have returned to the EU budget 35 percent of the allocated funds. Overview of the EU Common Agricultural Policy and Resources Croatia Will Receive 7.67 As an EU member, Croatia is able to finance its agricultural and rural development from three93 main sources:  European Agricultural Guarantee Fund (EAGF): fully financing direct payments and market management mechanisms under CAP Pillar 1.  European Agricultural Fund for Rural Development (EAFRD): financing the EU’s contribution to the rural development program implemented by Croatia. Public 92 Instrument for Pre-accession Assistance for Agriculture and Rural Development 93 The European Fisheries Fund is not covered by this analysis. 129 counterpart financing to the EAFRD allocation has to be provided in the national budget for agriculture and rural development.  Public budget allocation for the agriculture and rural development sector: financing from national budgetary sources applies to (i) the public counterpart financing matching financial inflows under the EAFRD; (ii) Complementary National Direct Payments (CNDP), or “top-up” payments, should Croatia opt to apply them; (iii) state aid to agriculture and rural development. 7.68 Under Pillar 1 of the CAP, direct payments are the most important type of income support both in Croatia and the EU (Figure 77). Direct payments have traditionally taken the heaviest share of the CAP Figure 77. EU CAP Payment Requirements expenditure, accounting for almost 80 percent in 2007-13 for the entire CAP EU27, dropping to 75 percent in Pillar 1: Pillar 2: 2014-20. Despite a relative decline Income and market support Rural development as a share in total agricultural support (from 85 percent in 2004-05 Direct income support for Increased agricultural to 66 percent in 2010-12) direct farmers competitiveness Sustainable management of payments in Croatia have generally Market interventions (export natural resources and climate increased in absolute terms. They refunds, intervention) action now account for HRK 2.3 billion, or Balanced territorial development of rural areas EUR 302 million. In general, the high level of agricultural support EAGF EAFRD was driven by preparations for the implementation of the CAP and the desire to increase domestic competitiveness. 7.69 The EU has a common rural development policy, institutionalized as the CAP’s rural development pillar (Pillar 2). It is aimed at meeting the investment support and restructuring needs of rural regions through a coherent set of sectoral and territorial support measures. Within the CAP framework, it accompanies and complements the market and income support policy implemented under CAP Pillar 1. During the recent evolutions of the CAP, this rural development pillar has been gaining increasing importance. As is the case in other new EU member states, assistance under CAP Pillar 2 will account for about 46 percent of the total CAP financial support to Croatia. Box 22. State Aid and National Programs in Agriculture State aid rules in the agricultural sector are based on three different perspectives . Firstly, the agricultural State Aid rules follow the general principles of the EU's competition policy. Secondly, State Aid rules in the agriculture sector have to be coherent with the provisions of the CAP. Finally, the rules have to be compatible with the Community's international obligations, in particular the WTO Agreement on Agriculture. Subject to compliance with the EU rules, state aids are paid from national budget resources . The EU definition of state aid applies to any measure providing national businesses with an "advantage in any form whatsoever conferred on a selective basis to undertakings by national public authorities". The EU has therefore pronounced a general prohibition of State Aid under Article 87(1) of the Treaty. However, it also recognizes that "in some circumstances, government interventions are necessary for a well-functioning and equitable economy" and therefore allows for "a number of policy objectives for which State aid can be considered compatible" under Articles 87(2) and 87(3), the so-called ‘exemptions’. However, these State Aid measures are tightly regulated, monitored and assessed by the European Commission through a system of rules and regulatory acts. They fall under three categories: 130  Existing State Aid: State Aid measures in place before accession and approved by the EC to be extended until mid-2016 ("sunset clause");  New State Aid: State Aid measures put in place after accession which require EC approval (on a no- objection basis) and registration and which can be extended beyond mid-2016;  De Minimis: State Aid measures which, for simplification, can be put in place without prior registration and approval by the EC within a maximum authorized envelope and a limit per beneficiary. Croatia will have to make a strategic decision as to how to allocate its domestic public resources between the key national expenditure categories in the sector, i.e.: (i) CNDPs; (ii) state aid, and (iii) counterpart financing of its rural development program. Income support scheme (EUR 11.7 million or 0.02 percent of GDP) needs to be terminated, while the sector still has at its disposal domestically-funded market-related expenditures of EUR440 million (HRK3.3 million or 1 percent of GDP). As per the latest Act on Support in Agriculture and Rural Development from October 2012, the existing state support in the Republic of Croatia covers:  payments in particularly sensitive sectors;  payments within the framework of national rural development measures;  support to agriculture and rural development financed from the State Budget and governed by special regulations; and  support to agriculture and rural development financed from the budgets of regional and local administration units and governed by special regulations. 7.70 Croatia can now leverage significant financial resources for its agriculture and rural development sector through the CAP (Figure 78). Over the next financial perspective, the CAP available for Croatia is estimated at about Figure 78. Croatia’s Annual Financial Allocations EUR3.5-3.7 billion. This accounts for about 25 under the CAP, 2014-20 (EUR million) percent of the country’s total expected receipts from the EU budget, under the 2014-2020 multiannual financial framework. Croatia joined the EU just before the new programming period, and opted not to benefit from the 6- month rural development allocation under the CAP 2007-13, but, instead, for a full IPARD annual allocation throughout 2013, amounting to EUR 27.7 million. This saved a considerable administrative burden stemming from the drafting and launching of a 6-month rural Note: Pillar I payments include direct payments and development program (RDP). de-mining reserve but exclude market-related expenditures. 7.71 Questions arise, what specific issues Source: EU multi-annual financial perspective for and constraints in the sector will the Croatia and own calculations based on the Treaty of government need to resolve to fully take accession. advantage of much larger EU funds in the future. These issues relate to farm productivity, excessive spending and lack of transparency on sector transfers and other subsidies and the EU funds programming to which we turn next. Programming the first post-accession National Rural Development Program 7.72 The annual financial envelope that Croatia expects from EAFRD during 2014-20 is EUR333 million, thus qualifying the country for a total of EUR 2.3 billion of potential assistance for its future national rural development program. With the new rural development program introduced in 2014, all the national rural development schemes need to be absorbed into the EAFRD-financed program, converted into state aid or terminated. This could save around (EUR35.9 million) 0.08 percent of GDP from 2014. 131 7.73 The 2014-20 regulation on rural development and Croatia’s Accession Treaty to the European Union foresee certain options and requirements with regard to the allocation of the future rural development funds:  Transfer of funds from Pillar 2 to CNDPs: Just like all the other new member states, Croatia has negotiated its right to use up to 20 percent of its annual EAFRD allocation for 2014- 2016 to contribute to the top-up envelope but not exceeding 45 percent of phasing- in funds in any given year. This is justified through the phasing in calendar of the new entrants’ EAGF payments, and is also appealing to countries facing domestic budget constraints and/or low absorption capacity of rural development funds. Croatia needs to reflect this in its National Rural Development Program (NRDP) 2014-20, and also provide a 20-percent matching counterpart financing from its national budget. This will mean, for 2014-16, an annual contribution to the CNDP envelope of EUR46.6 million, of which EUR37.3 million would come from EAFRD and the rest from the national budget.  Minimum allocations per priority/measure: Certain rural development priorities or measures are subject to a minimum allocation. Thus, LEADER will have to receive at least 2.5 percent of the total program envelope (or EUR 8.3 million annually out of the EAFRD allocation). A minimum of 30 percent of the total EAFRD contribution will have to be used for climate change mitigation and adaptation and land management (through agri-environment-climate and organic farming measures, as well as through payments to areas facing natural and other specific constraints). Given Croatia’s agri-environmental endowments as well as its existing pre-accession experience with such measures, it might be advisable to even consider going beyond this threshold. Up to 0.25 percent of the EAFRD annual allocation (EUR 830,000) can be dedicated for technical assistance. 7.74 The amount of domestic funds needed to match the EAFRD contributions will depend on Croatia’s rural development programming choices. An additional factor to consider is the actual date of the future NRDP approval by the EC, as delays may impact on the calendar of domestic budget fund disbursements. Delays with NRDPs are not uncommon in either old or new member states, and it is not excluded that the program might start only in early 2015. Regardless of that, and given the regulated co-financing requirements for most of the rural development measures that might be reflected in Croatia’s NRDP, it could be expected that the matching funds that need to be committed would be 15 percent of the total public contribution. Under this assumption and assuming transfer of 15 percent from EAFRD to direct payment envelope as already notified by Croatia to the EC, EUR53.6 million could need to be committed annually to co-finance the EAFRD-sourced payments. 7.75 However, the national co-financing needs will also be driven by the future program absorption rates. As seen above, the pre-accession absorption track record is not too favorable. Still, one needs to acknowledge that SAPARD and IPARD focused primarily on investment measures, which can be associated with specific implementation complexities (beneficiary pre- and co-financing), further exacerbated in the Croatian context (difficulties in collateralizing farm land and assets and in accessing loans). In turn, the full-blown post-accession rural development menu contains measures that can be more easily disbursed – notably the transfers to the CNDP envelope and several of the area-based environmentally-focused payments (such as payments for areas with specific handicaps or agri-environmental measures). Should Croatia opt for a sizable representation of these in its NRDP, the prospects for improving its absorption rates despite the much higher volume of EU support are not unrealistic. Assuming a full utilization of the CNDP 132 transfer option over 2014-16 and at least 30 percent of the EAFRD funds spent on agri- environment-climate payments, it could be expected that at the very least some EUR44 million would need to be disbursed annually, to match the EAFRD between 2014 and 2016. Implications for the post-accession ARD budget 7.76 Direct payments for Croatia in 2013 will start at 25 percent of their full level, increase yearly by five percentage points until 2016 and then increase by ten percentage points per year to reach 100 Figure 79. Pre and Post-accession Simulation of EAGF Direct Payments percent in 2022. and CNDPs Effectively, this would mean that EU-funded direct payments in the first year of accession would be only EUR 93 million versus EUR 303 million paid from the national budget in 2013 (Figure 79). 7.77 To safeguard its farmers’ income support, Croatia may consider temporarily utilizing Source: Ministry of Finance and own calculations based on Croatia’s Complementary National accession documents Direct Payments (CNDP). Topping-up represents a transitional option and is subject to an EU-agreed phasing out schedule, as EU-funded direct payments are being phased in. Unlike the other EU new member states, Croatia was allowed to utilize CNDPs from the national budget up to 100 percent of the EU direct payment ceiling, even from its first year of membership. The top-ups can be funded from both the national budget and also through an up to 20 percent reallocation of EAFRD funds. In order to safeguard farmers’ incomes, top-ups could be considered up to at least 60 percent of the full direct payment ceiling, equivalent to an amount up to EUR224 million in the first year of accession. These top-ups could be then gradually phased-out over 2014-20, as the EAGF-funded direct payments are being gradually increased. The fiscal space could be made available by, for instance, reallocating the sizable domestic pre-accession direct payment envelope. Recommendations 7.78 At one percent of GDP allocated on public spending for agriculture and rural development Croatia has a room to reduce the spending by 0.2-0.5 percentage points of GDP to align its spending level to most other EU countries. Within the agricultural budget, transfer programs, and in particular direct payments have preserved a sizeable albeit declining share. Subsidy modulation, rationalization of public service and administration cost, as well as elimination of duplicative programs to EU-funded ones would be among key areas for cost- cutting interventions. 7.79 Croatia will have to make strategic decisions on the allocation of its sector budget between the major nationally-funded expenditure categories. Some of the pre-accession 133 schemes will have to be terminated (e.g. the income scheme benefiting small farmers), and market-related expenditures will be taken over by the EU budget. Besides CNDPs, these categories include co-financing of the rural development programs funded under the second CAP pillar from the EAFRD, state aid and administrative costs associated with the implementation of national and EU-funded support measures. State aid in agriculture is entirely funded from the national budget and presents the opportunity of promoting, within tightly regulated limits, measures that are otherwise non-eligible under the CAP. Such measures may include supports of public interest but also aids targeted to sensitive sectors (which Croatia will continue to apply in the first three years of accession under its sunset clause), or aids that can address market failures and/or support CAP fund absorption (especially under Pillar 2). It is not clear at this point, besides the annual EUR 22.2 million allocations for sensitive sectors, which other measures are planned to be included in this support category and qualify for national funding. 7.80 Public services in agriculture need strengthening and rationalization. While Paying Agency needs to be further strengthened taking into account recommendations of the European Commission, including ensuring the necessary budget for its functioning and enough educated staff to implement the requirements of the CAP, there is scope for service rationalization. Given the size of the sector and the number of beneficiaries of agricultural support, the administrative system appears oversized and a certain amount of cost cutting is recommended (estimated at approximately 20 percent of its existing public services budget). 7.81 Fiscal discipline, budget transparency, and streamlining budget planning are priorities. Specifically: (i) Strict implementation of the legal provisions on financial discipline. Croatia needs to be very careful as regards the deadlines for all the previously committed and overdue payments in the first year of EU membership. Any support that will not qualify as existing state aid will not be paid. For all agricultural subsidies, commitments made in a certain year should be followed by a clear budgetary projection of the disbursements in the next year, with precise payment deadlines. (ii) Increasing transparency, classify budgetary lines to enable a simple functional and economic classification of the agricultural budget. For the same reason, it is recommended that Croatia also clearly separates its ARD commitment and payment appropriations. This would be in line with the multiannual CAP – and, more broadly, the EU budget–and will help Croatia improve tracking and managing national and EU expenditures. 7.82 The most important “market interventions” which include milk, mandarins and apples should be terminated. Such subsidies distort markets and incentives and favor certain groups of producers and land use and are at the same time a major burden to the state budget. 7.83 Lessons learned in other EU member states demonstrate the need for well- integrated programs in order to effectively leverage, mobilize, and eventually absorb available funds. Selectivity and coherence as well as timeliness in programming the specific provisions of Pillar 2 are equally important as their communication to beneficiaries. Simplicity of application and approval processes and procedures reduces transaction costs during the preparation, submission, and processing of support applications and reimbursement claims. 134 7.84 It would also be advisable for Croatia to take advantage of the option to transfer Pillar 2 funds into the CNDP envelope for 2014-16 an additional 15 percent exploiting the flexibility between two pillars. This presents a double advantage. On the one hand, it alleviates the pressure on the domestic resources, as such a transfer would involve an average annual allocation of EUR87.1 million from EAFRD in 2014-2016 matched by EUR37.3 million national counterpart financing, plus another EUR74 million for ensuring the safeguarding of pre- accession direct payment advantages. In total, this would result in a national top-up of EUR179 million, freeing up EUR58.6 million (0.13 percent of GDP) annually of public resources over 2014-20 for alternative uses. It would also enhance the weak absorption of rural development funds that was such a challenge during the pre-accession period. 135 8.1 Croatia has entered the EU, achieving the dream of this generation and strengthening hopes for future prosperity. But its fiscal weaknesses and vulnerabilities pose substantial risks for that future. The previous chapters established the case for sustained medium- term adjustment to substantially reduce those risks, putting its public debt on a downward trajectory. However, addressing this will be difficult given the intertwined challenges Croatia is facing today: (i) to bring down the fiscal deficit and reverse adverse public debt dynamics; and (ii) create fiscal space for the absorption of large EU funds preparing the ground for recovery and sustainable long-term growth. A predictable and credible policy framework effectively dealing with the twin challenges would significantly improve consumer and investor sentiment, thus facilitate a stronger recovery. 8.2 To achieve debt sustainability, the government will need to turn 1.8 percent primary deficit in 2013 into a balance in 2016. This will require an increase in revenues or a reduction in primary expenditures of 3.7 percentage points of GDP. While the adjustment is urgent not to deepen vulnerabilities and risks of much higher borrowing costs, in a depressed environment, an adequate balance needs to be found between credible consolidation and an over-excessive burden consolidation may have on depressing growth and growth expectations. 8.3 At the same time, Croatia will need to create fiscal space averaging up to 1.8 percent of GDP a year in 2014-2020 to support EU funds absorption. This can be achieved through the efficient utilization of EU funds, combined with some switching of budget spending toward high return investments that would support long-term growth, to ensure both fiscal sustainability and foster income convergence with the rest of the EU. 8.4 Croatia’s spending and revenue pattern suggests that a sizeable fiscal adjustment of 4-5 percentage points of GDP could be implemented over the medium term. This would make the EDP targets achievable and would ensure that Croatia follows the likely debt- sustainability scenario presented in Chapter 2. While revenue measures identified could raise additional revenue of 2 percent of GDP efficiently and fairly while reducing the adverse impact on growth and employment, more fiscal space can be created on the spending side. There are several areas identified that have room for savings: (i) In public administration costs, some 2 percent of GDP in cumulative savings could be achieved over the medium term through staff rationalization at local and national governments to create a leaner, but more effective administration. (ii) In social transfers and health cost, some 2-3 percent of GDP in cumulative savings could be achieved over the medium term if the most inefficient programs were eliminated or rationalized and targeting improved. This would, however, be partially offset in the medium term by aging pressures. (iii) Croatia has significant scope to rationalize subsidies, in the order of about 1 percent of GDP. This would imply eliminating the most inefficient subsidies, such as product subsidies in agriculture and rationalizing sector subsidies, especially to railroads, ports, steel and shipyards, and tourism. This would also raise 136 competitive pressure in these sectors while creating room for horizontal aid (such as R&D). 8.5 Fiscal adjustment will need to be accompanied by institutional strengthening in order to absorb substantial EU funds in the coming years as well as structural reforms to generate savings, reduce waste and improve the overall efficiency of the public sector. The fiscal adjustment will need to be accompanied by spending reallocation to those projects which could be substituted by the EU grant funds as well as privatization, both of which would lead to a faster decline in public debt. 8.6 What failed and what worked in past attempts at fiscal adjustment? The lessons learned from EU member states of past fiscal consolidation episodes that can help Croatia avoid costly mistakes can be divided in three dimensions:  Rationale for and design of fiscal adjustments plans: Since the mid-1980s fiscal consolidations have usually been introduced in response to high and rising public debt and due to a relatively large initial size of government in EU countries most plans have focused on spending cuts (Figure 80).  Implementation record and underlying macroeconomic factors: Average annual planned improvement in the structural fiscal balance equaled 1.7 percent of GDP cumulative over three years, while the outturn was on average lower at 0.9 percent improvement per year. Also, expenditure cuts rarely materialized to the extent envisaged, while revenues often outperformed the expectations. Finally, deviations of economic growth from initial expectations were a key factor underlying success or failure. Figure 80. Fiscal Policy Adjustment in 2009-2012, percent of GDP, ESA 95 Note: Fiscal policy adjustment since the crisis peak has beeen calculated using the maximum values of revenue, expenditure and deficit indicator in 2009-2011 period compared to 2012. Source: EUROSTAT, MoF, World Bank staff calculations and estimates.  Fiscal institutions and political factors: As medium-term fiscal limits were made more legally binding, actual compliance with spending targets improved. Coordination among different levels of government was an important determinant of general government balance improvement. And finally, broad public support for the adjustment plan improved adjustment outcomes. 137 8.7 What are the implications for planned adjustment in Croatia? Consolidation plans should include binding and well-defined medium-term expenditure ceilings, which are endorsed by the legislature to ensure they become binding. To allow for policy response to unexpected shocks, plans need to permit some flexibility while credibly preserving the medium-term consolidation objectives. Using cyclically adjusted targets would allow the automatic stabilizers operate in response to cyclical fluctuations; however, uncertainties on potential output movement may render such targeting difficult. 8.8 Targets need to be based on sound information on the initial state of public finances. Improved fiscal transparency has also other positive effects: it tends to result in better credit ratings and hence lower borrowing costs. Fiscal consolidation targets need to be based on credible macroeconomic assumptions which need to be conservative to avoid downward revisions. Also, the revenue-expenditure mix in fiscal consolidation plans should reflect country- specific societal preferences and structural fiscal characteristics. 8.9 Can such adjustment be done? Most EU countries have gone through fiscal consolidation episodes in the recent history. When corrected for the cycle, over the last four years Greece achieved an annual rate of primary deficit reduction of 4.1 percentage point of GDP on average (2009-2013), the highest in the developed world in recent years (Figure 81). Latvia had a similar experience over 2009-2012 period. Figure 81. Fiscal Consolidation Among OECD Countries (Average annual cyclically adjusted primary balance improvement, percent of GDP) 4.5 4.1 4.0 3.5 3.3 3.0 2.8 2.5 2.5 1.9 1.8 1.8 2.0 1.6 1.6 1.5 1.5 1.5 1.3 1.2 1.0 1.0 0.5 0.0 Note: Fiscal consolidation episodes as defined in OECD Economic Outlook 81, May 2007 Source: EUROSTAT, World Bank staff calculations. 8.10 Required annual adjustment in Croatia is much smaller than in Greece and thus significantly easier. Croatia would need to cut its primary deficit by 1.5 percentage points of GDP per year over the next three years to bring its public debt to sustainable levels. While one can argue that Greece has experienced a large social cost due to such an adjustment, Croatia’s adjustment would be similar to fiscal consolidation episodes of Sweden, the UK, Romania or Ireland. While such an adjustment is urgent, it is also achievable so Croatia can strengthen the hopes for future prosperity as an EU member state. 8.11 The consequences of not taking these measures might be detrimental to Croatia’s short and long-term growth perspectives. By not insisting on the more forceful absorption of 138 EU funds, Croatia may face a possible loss of EU funds and more importantly an opportunity to finance its growth through foreign grant funds thus helping its growth and fiscal consolidation objectives. By not addressing fiscal vulnerabilities, the country may face further rise in the cost of financing as well as a lack of access to affordable long-term borrowing. This would on the other hand, deepen the recessionary trends and require much more urgent and deeper fiscal consolidation actions. 139 Adsera, A. (2004): Changing Fertility Rates in Developed Countries. The Impact of Labor Market Institutions. Journal of Population Economics, 17. Adsera, A. (2005). Vanishing Children: From High Unemployment to Low Fertility in Developed Countries, American Economic Review, 95. Ahmad E. and G. Brosio (2006). Decentralization and Local Service Provision: What Do We Know?. LSE Asia Research Centre Working Papers. Alibegovic D. (2010). “Innovative Municipal Development Financing. The Case of Three Croatian Cities” Antczak, M. (2003). Do Acceding Countries Need Higher Fiscal Deficits?. Warsaw. Apps, P., and R. Rees (2005). Gender, Time Use, and Public Policy over the Life Cycle. Oxford Review of Economic Policy, 21(3), 439-461. Backé, P. 2002, “Fiscal Effects of EU Membership for Central European and Baltic EU Accession Countries”, Oesterreichische Nationalbank (ed.): Focus on Transition 2/2002, Vienna. Bađun M. (2011). “Why are there so many disability pensions beneficiaries in Croatia?”, Institute of Public Finance Baldacci E., S. Gupta et al. (2010). Restoring Debt Sustainability After Crises: Implications for the Fiscal Mix, IMF Working Paper WP/10/232 Becker, G. S. (1981). A Treatise on the family. Cambridge, MA: Harvard University Press. Bloom, D. E., D. Canning, G. Fink, and J. E. Finlay (2009). The Cost of Low Fertility in Europe, Working Paper 14820, National Bureau of Economic Research. Cigno, A. (1991). Economics of the family. Oxford, England: Clarendon Press. Croatia, Government of the Republic of. (2006). National Population Policy. Ministry of Family, Verterans' Affairs and Intergenerational Solidarity. Zagreb. de Laat, J., A. Sevilla-Sanz. (2009) Female Labor Force Participation, Men's Home Production, and Fertility: Evidence from the OECD. Feminist Economics. Forthcoming Del-Boca, D. (2002). The effect of child care and part time opportunities on participation and fertility decisions in Italy. Journal of Population Economics, 15(3), 549–573. Del-Boca, D., and M. Locatelli (2006): The Determinants of Motherhood and Work Status: a Survey. CHILD Working Papers 15, Centre for Household, Income, Labour and Demographic economics. Del-Boca, D., S. Pasqua, and C. Pronzato (2009): Motherhood and Market Work Decisions in Institutional Context: A European Perspective, Oxford Economic Papers, Forthcoming. Deloitte (2013). “Preliminary Tax Gap Estimates”. Croatia Revenue Administration and Modernization Project. 140 EBRD and World Bank. (2010). Life in Transition Survey. European Commission. http://europa.eu/rapid/press-release_MEMO-13-1124_en.htm?locale=en European Commission. http://ec.europa.eu/economy_finance/economic_governance/index_en.htm European Court of Auditors. (2011). “Has EU assistance improved Croatia's capacity to manage post- accession funding?” Special Report No 14. European Parliament. 2007. Structural Funds' Implementation in Poland--Challenges for 2007- 2013. Brussels: DG Internal Policies of the Union, Budgetary Support Unit, 2007. Ferry M., F. Gross, J. Bachtler, and I. McMaster. 2007. "Turning Strategies into Projects: The Implementation of 2007-13 Structural Funds Programmes." European Policy Research Centre. IQ-Net Thematic Paper 20, no. 2. Fox, W. and T. Gurley (2006). “Will Consolidation Improve Sub-National Governments?”. World Bank Policy Research Working Papers, May 2006 Gauthier, A. (2007) The impact of family policies on fertility in industrialized countries: a review of the literature. Population Research Policy Review (26) 323–346 Hoem, J.M. (2008). The impact of public policies on European fertility. Demographic Research. Vol.19, Article 10, 249-260 IMF (2012a), Croatia: Options for Modernizing the Property Tax. IMF (2012b). Croatia: Concluding Statement of the IMF Staff Visit. IMF (2013). IMF Fiscal Monitor, Taxing Time Kesner-Škreb, Marina; Madžarević-Šujster, Sanja. (2004). Income Tax Progressivity in Croatia (1995-2002). International Review of Economics & Business. VII (2004). Koettl (2010). “Financing, Providing and Regulating Long-term Care for the Elderly in Croatia“. Thee World Bank Kopits, G. and I. Székely (2002). Fiscal Policy Challenges of EU Accession for the Baltics and Central Europe, November 2002. Khwaja M. and I. Iyer (2013). “Revenue Potential, Tax Space, and Tax Gap in ECA Countries: A Comparative Analysis”. Le T.M., B. Moreno-Dodson and N. Bayraktar (2012). Tax Capacity and Tax Effort: Extended Cross-Country Analysis from 1994 to 2009. World Bank Policy Research Working Paper No. 6252 Malekovic, S., J. Puljiz and S. Tisma (2011). “New Opportunities for Regional and Local Actors in Croatia in Supporting Socio-Economic Development” Maniokas, K. 2006. "Improving the User-friendliness of the Structural Funds Management System in Lithuania: Challenges and opportunities." Lithuania: European Social, Legal and Economic Projects. McMaster, Irene, and John Bachtler. 2005. Implementing Structural Funds in the New Member States: Ten Policy Challenges. Glasgow: European Policies Research Centre. 141 Ministry of Regional Development. 2005. „Program naprawczy zwiększający absorpcję funduszy strukturalnych w ramach Narodowego Planu Rozwoju 2004-2006.” 2005. [[which country? also, title needs to be translated. Naylor C. et al (2012). “Transforming our Health Care System”, The King’s Fund, Ideas that change health care. Norregaard J. (2013). Taxing Immovable Property, Revenue Potential and Implementation Chellenges. IMF Working Paper, WP/13/129, May 2013. OECD (2006). “The efficiency of sub-central spending”, workshop proceedings. OECD (2007). Babies and Bosses: Reconciling Work and Family Life OECD (2010). Revenue Statistics 1965-2009. Paliova, I and Lybek, T. (2014). Bugaria’s EU Funds Absoprtion: Maximizing the Potential!. IMF Working Paper. WP/14/21. International Monetary Fund Rosenberg C. and Sierhej R (2007). Interpreting EU Funds Data for macroeconomic Analysis in the New Member States, IMF Working Paper WP/07/77, International Monetary Fund Sopek P. (2013). Budget perspective in Croatia after accession to the European Union. Financial Theory and Practice, 37 (1) Taylor, S., and P. Raines. 2001. "Implementing EU Cohesion Policy and the Pre-Accession Instruments." Forest Hills, Loch Ard, September 12. USAID (2009). “Property Tax Reform in Developing and Transition Countries” Vincelette G. and I. Vassileva. (2006). Untangling The Maze Of European Union Funds To Bulgaria. World Bank Policy Research Working Paper 3962, July 2006 Weil, D. (1999). Population Growth, Dependency, and Consumption: Why Has Fertility Fallen below Replacement in Industrial Nations, and Will It Last? American Economic Review, 89, 251-255, Papers and Proceedings of the One Hundred Eleventh Annual Meeting of the American Economic Association. World Bank. World Development Indicators, various. World Bank. www.doingbusiness.org World Bank (2008). “Croatia: A Public Finance Review; Restructuring Public Finance to Sustain Growth and Improve Public Services”, Report No. 37321 World Bank (2011). “Employment Protection Legislation and Labor Market Outcomes: Theory, Evidence and Lessons for Croatia” World Bank (2013a). “South East Europe Municipal Finance Review, Local Government Finance in the Western Balkans” World Bank (2013b). EU11 Regular Economic Report, No. 28, December 2013. World Bank (2013c). Growing with Jobs in Europe and Central Asia. World Bank (2014). Croatia: Justice Sector Expenditure and Institutional Review. 142